diff --git "a/checkpoint-10830/trainer_state.json" "b/checkpoint-10830/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-10830/trainer_state.json"
@@ -0,0 +1,75843 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 10830,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002770083102493075,
+      "grad_norm": 1.493026852607727,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 1.0538,
+      "step": 1
+    },
+    {
+      "epoch": 0.000554016620498615,
+      "grad_norm": 1.3199710845947266,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 0.9806,
+      "step": 2
+    },
+    {
+      "epoch": 0.0008310249307479224,
+      "grad_norm": 1.3620953559875488,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 1.0318,
+      "step": 3
+    },
+    {
+      "epoch": 0.00110803324099723,
+      "grad_norm": 1.4497380256652832,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.0472,
+      "step": 4
+    },
+    {
+      "epoch": 0.0013850415512465374,
+      "grad_norm": 1.3085776567459106,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.9745,
+      "step": 5
+    },
+    {
+      "epoch": 0.0016620498614958448,
+      "grad_norm": 1.4101017713546753,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.0561,
+      "step": 6
+    },
+    {
+      "epoch": 0.0019390581717451524,
+      "grad_norm": 1.4515964984893799,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 1.0348,
+      "step": 7
+    },
+    {
+      "epoch": 0.00221606648199446,
+      "grad_norm": 1.2317475080490112,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.9423,
+      "step": 8
+    },
+    {
+      "epoch": 0.002493074792243767,
+      "grad_norm": 1.3780522346496582,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 1.0005,
+      "step": 9
+    },
+    {
+      "epoch": 0.002770083102493075,
+      "grad_norm": 1.424789309501648,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.0221,
+      "step": 10
+    },
+    {
+      "epoch": 0.0030470914127423824,
+      "grad_norm": 1.352807641029358,
+      "learning_rate": 5.5e-07,
+      "loss": 1.0366,
+      "step": 11
+    },
+    {
+      "epoch": 0.0033240997229916896,
+      "grad_norm": 1.4198808670043945,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.9874,
+      "step": 12
+    },
+    {
+      "epoch": 0.003601108033240997,
+      "grad_norm": 1.37282133102417,
+      "learning_rate": 6.5e-07,
+      "loss": 0.9976,
+      "step": 13
+    },
+    {
+      "epoch": 0.003878116343490305,
+      "grad_norm": 1.3420826196670532,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 1.0269,
+      "step": 14
+    },
+    {
+      "epoch": 0.004155124653739612,
+      "grad_norm": 1.3467051982879639,
+      "learning_rate": 7.5e-07,
+      "loss": 1.0208,
+      "step": 15
+    },
+    {
+      "epoch": 0.00443213296398892,
+      "grad_norm": 1.336922526359558,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.0465,
+      "step": 16
+    },
+    {
+      "epoch": 0.004709141274238227,
+      "grad_norm": 1.2495678663253784,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 0.981,
+      "step": 17
+    },
+    {
+      "epoch": 0.004986149584487534,
+      "grad_norm": 1.3265974521636963,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 0.949,
+      "step": 18
+    },
+    {
+      "epoch": 0.005263157894736842,
+      "grad_norm": 1.204640507698059,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 0.9917,
+      "step": 19
+    },
+    {
+      "epoch": 0.00554016620498615,
+      "grad_norm": 1.1768105030059814,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.0069,
+      "step": 20
+    },
+    {
+      "epoch": 0.005817174515235457,
+      "grad_norm": 1.1372712850570679,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 0.9638,
+      "step": 21
+    },
+    {
+      "epoch": 0.006094182825484765,
+      "grad_norm": 1.0769680738449097,
+      "learning_rate": 1.1e-06,
+      "loss": 0.9549,
+      "step": 22
+    },
+    {
+      "epoch": 0.006371191135734072,
+      "grad_norm": 1.0383652448654175,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 0.9953,
+      "step": 23
+    },
+    {
+      "epoch": 0.006648199445983379,
+      "grad_norm": 0.8858652114868164,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.9546,
+      "step": 24
+    },
+    {
+      "epoch": 0.006925207756232687,
+      "grad_norm": 0.7955514192581177,
+      "learning_rate": 1.25e-06,
+      "loss": 0.9369,
+      "step": 25
+    },
+    {
+      "epoch": 0.007202216066481994,
+      "grad_norm": 0.7989655137062073,
+      "learning_rate": 1.3e-06,
+      "loss": 0.9543,
+      "step": 26
+    },
+    {
+      "epoch": 0.007479224376731302,
+      "grad_norm": 0.7056751251220703,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 0.9063,
+      "step": 27
+    },
+    {
+      "epoch": 0.00775623268698061,
+      "grad_norm": 0.6925091743469238,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.9554,
+      "step": 28
+    },
+    {
+      "epoch": 0.008033240997229917,
+      "grad_norm": 0.7046273350715637,
+      "learning_rate": 1.45e-06,
+      "loss": 0.954,
+      "step": 29
+    },
+    {
+      "epoch": 0.008310249307479225,
+      "grad_norm": 0.7233896255493164,
+      "learning_rate": 1.5e-06,
+      "loss": 1.0106,
+      "step": 30
+    },
+    {
+      "epoch": 0.008587257617728532,
+      "grad_norm": 0.6680523157119751,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 0.9682,
+      "step": 31
+    },
+    {
+      "epoch": 0.00886426592797784,
+      "grad_norm": 0.624336302280426,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.9258,
+      "step": 32
+    },
+    {
+      "epoch": 0.009141274238227148,
+      "grad_norm": 0.6681525111198425,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 0.9115,
+      "step": 33
+    },
+    {
+      "epoch": 0.009418282548476454,
+      "grad_norm": 0.641791820526123,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.9337,
+      "step": 34
+    },
+    {
+      "epoch": 0.009695290858725761,
+      "grad_norm": 0.6529171466827393,
+      "learning_rate": 1.75e-06,
+      "loss": 0.9515,
+      "step": 35
+    },
+    {
+      "epoch": 0.009972299168975069,
+      "grad_norm": 0.5708187222480774,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.9184,
+      "step": 36
+    },
+    {
+      "epoch": 0.010249307479224376,
+      "grad_norm": 0.5597304105758667,
+      "learning_rate": 1.85e-06,
+      "loss": 0.9142,
+      "step": 37
+    },
+    {
+      "epoch": 0.010526315789473684,
+      "grad_norm": 0.5888805389404297,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.9374,
+      "step": 38
+    },
+    {
+      "epoch": 0.010803324099722992,
+      "grad_norm": 0.5895810723304749,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 0.968,
+      "step": 39
+    },
+    {
+      "epoch": 0.0110803324099723,
+      "grad_norm": 0.5828868746757507,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.8874,
+      "step": 40
+    },
+    {
+      "epoch": 0.011357340720221607,
+      "grad_norm": 0.5907834768295288,
+      "learning_rate": 2.05e-06,
+      "loss": 0.9543,
+      "step": 41
+    },
+    {
+      "epoch": 0.011634349030470914,
+      "grad_norm": 0.5174127221107483,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.95,
+      "step": 42
+    },
+    {
+      "epoch": 0.011911357340720222,
+      "grad_norm": 0.5172901749610901,
+      "learning_rate": 2.15e-06,
+      "loss": 0.9246,
+      "step": 43
+    },
+    {
+      "epoch": 0.01218836565096953,
+      "grad_norm": 0.5019023418426514,
+      "learning_rate": 2.2e-06,
+      "loss": 0.8369,
+      "step": 44
+    },
+    {
+      "epoch": 0.012465373961218837,
+      "grad_norm": 0.5965269207954407,
+      "learning_rate": 2.25e-06,
+      "loss": 0.9166,
+      "step": 45
+    },
+    {
+      "epoch": 0.012742382271468145,
+      "grad_norm": 0.5445163249969482,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.8671,
+      "step": 46
+    },
+    {
+      "epoch": 0.01301939058171745,
+      "grad_norm": 0.5398634076118469,
+      "learning_rate": 2.35e-06,
+      "loss": 0.9209,
+      "step": 47
+    },
+    {
+      "epoch": 0.013296398891966758,
+      "grad_norm": 0.49044713377952576,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.8799,
+      "step": 48
+    },
+    {
+      "epoch": 0.013573407202216066,
+      "grad_norm": 0.5420831441879272,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.8706,
+      "step": 49
+    },
+    {
+      "epoch": 0.013850415512465374,
+      "grad_norm": 0.49241095781326294,
+      "learning_rate": 2.5e-06,
+      "loss": 0.8907,
+      "step": 50
+    },
+    {
+      "epoch": 0.014127423822714681,
+      "grad_norm": 0.4719524383544922,
+      "learning_rate": 2.55e-06,
+      "loss": 0.8677,
+      "step": 51
+    },
+    {
+      "epoch": 0.014404432132963989,
+      "grad_norm": 0.5064590573310852,
+      "learning_rate": 2.6e-06,
+      "loss": 0.9435,
+      "step": 52
+    },
+    {
+      "epoch": 0.014681440443213296,
+      "grad_norm": 0.46574175357818604,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.8723,
+      "step": 53
+    },
+    {
+      "epoch": 0.014958448753462604,
+      "grad_norm": 0.4822365641593933,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.8311,
+      "step": 54
+    },
+    {
+      "epoch": 0.015235457063711912,
+      "grad_norm": 0.4746630787849426,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.8785,
+      "step": 55
+    },
+    {
+      "epoch": 0.01551246537396122,
+      "grad_norm": 0.49930208921432495,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.9016,
+      "step": 56
+    },
+    {
+      "epoch": 0.015789473684210527,
+      "grad_norm": 0.4789316952228546,
+      "learning_rate": 2.85e-06,
+      "loss": 0.8373,
+      "step": 57
+    },
+    {
+      "epoch": 0.016066481994459834,
+      "grad_norm": 0.4359954297542572,
+      "learning_rate": 2.9e-06,
+      "loss": 0.8478,
+      "step": 58
+    },
+    {
+      "epoch": 0.016343490304709142,
+      "grad_norm": 0.4391110837459564,
+      "learning_rate": 2.95e-06,
+      "loss": 0.8584,
+      "step": 59
+    },
+    {
+      "epoch": 0.01662049861495845,
+      "grad_norm": 0.42015522718429565,
+      "learning_rate": 3e-06,
+      "loss": 0.8128,
+      "step": 60
+    },
+    {
+      "epoch": 0.016897506925207757,
+      "grad_norm": 0.4458240866661072,
+      "learning_rate": 3.05e-06,
+      "loss": 0.9213,
+      "step": 61
+    },
+    {
+      "epoch": 0.017174515235457065,
+      "grad_norm": 0.43013569712638855,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.8891,
+      "step": 62
+    },
+    {
+      "epoch": 0.017451523545706372,
+      "grad_norm": 0.4295575022697449,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.8786,
+      "step": 63
+    },
+    {
+      "epoch": 0.01772853185595568,
+      "grad_norm": 0.4423756003379822,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.8812,
+      "step": 64
+    },
+    {
+      "epoch": 0.018005540166204988,
+      "grad_norm": 0.4237767457962036,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.8272,
+      "step": 65
+    },
+    {
+      "epoch": 0.018282548476454295,
+      "grad_norm": 0.41085582971572876,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.8848,
+      "step": 66
+    },
+    {
+      "epoch": 0.0185595567867036,
+      "grad_norm": 0.4319153130054474,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.874,
+      "step": 67
+    },
+    {
+      "epoch": 0.018836565096952907,
+      "grad_norm": 0.44288167357444763,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.8799,
+      "step": 68
+    },
+    {
+      "epoch": 0.019113573407202215,
+      "grad_norm": 0.4345931112766266,
+      "learning_rate": 3.45e-06,
+      "loss": 0.9089,
+      "step": 69
+    },
+    {
+      "epoch": 0.019390581717451522,
+      "grad_norm": 0.41736096143722534,
+      "learning_rate": 3.5e-06,
+      "loss": 0.8675,
+      "step": 70
+    },
+    {
+      "epoch": 0.01966759002770083,
+      "grad_norm": 0.4338845908641815,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.8776,
+      "step": 71
+    },
+    {
+      "epoch": 0.019944598337950138,
+      "grad_norm": 0.4050421118736267,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.8871,
+      "step": 72
+    },
+    {
+      "epoch": 0.020221606648199445,
+      "grad_norm": 0.40696731209754944,
+      "learning_rate": 3.65e-06,
+      "loss": 0.8587,
+      "step": 73
+    },
+    {
+      "epoch": 0.020498614958448753,
+      "grad_norm": 0.40157297253608704,
+      "learning_rate": 3.7e-06,
+      "loss": 0.832,
+      "step": 74
+    },
+    {
+      "epoch": 0.02077562326869806,
+      "grad_norm": 0.42174622416496277,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.9153,
+      "step": 75
+    },
+    {
+      "epoch": 0.021052631578947368,
+      "grad_norm": 0.44821983575820923,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.8115,
+      "step": 76
+    },
+    {
+      "epoch": 0.021329639889196676,
+      "grad_norm": 0.4175020158290863,
+      "learning_rate": 3.85e-06,
+      "loss": 0.8775,
+      "step": 77
+    },
+    {
+      "epoch": 0.021606648199445983,
+      "grad_norm": 0.3963990807533264,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.823,
+      "step": 78
+    },
+    {
+      "epoch": 0.02188365650969529,
+      "grad_norm": 0.4085932672023773,
+      "learning_rate": 3.95e-06,
+      "loss": 0.8255,
+      "step": 79
+    },
+    {
+      "epoch": 0.0221606648199446,
+      "grad_norm": 0.39907416701316833,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.8611,
+      "step": 80
+    },
+    {
+      "epoch": 0.022437673130193906,
+      "grad_norm": 0.41605669260025024,
+      "learning_rate": 4.05e-06,
+      "loss": 0.8628,
+      "step": 81
+    },
+    {
+      "epoch": 0.022714681440443214,
+      "grad_norm": 0.41773635149002075,
+      "learning_rate": 4.1e-06,
+      "loss": 0.8363,
+      "step": 82
+    },
+    {
+      "epoch": 0.02299168975069252,
+      "grad_norm": 0.4385448694229126,
+      "learning_rate": 4.15e-06,
+      "loss": 0.8849,
+      "step": 83
+    },
+    {
+      "epoch": 0.02326869806094183,
+      "grad_norm": 0.4114312529563904,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.8031,
+      "step": 84
+    },
+    {
+      "epoch": 0.023545706371191136,
+      "grad_norm": 0.41531598567962646,
+      "learning_rate": 4.25e-06,
+      "loss": 0.8371,
+      "step": 85
+    },
+    {
+      "epoch": 0.023822714681440444,
+      "grad_norm": 0.4031124413013458,
+      "learning_rate": 4.3e-06,
+      "loss": 0.8606,
+      "step": 86
+    },
+    {
+      "epoch": 0.02409972299168975,
+      "grad_norm": 0.40101325511932373,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.8568,
+      "step": 87
+    },
+    {
+      "epoch": 0.02437673130193906,
+      "grad_norm": 0.4076603949069977,
+      "learning_rate": 4.4e-06,
+      "loss": 0.877,
+      "step": 88
+    },
+    {
+      "epoch": 0.024653739612188367,
+      "grad_norm": 0.4312387704849243,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.8396,
+      "step": 89
+    },
+    {
+      "epoch": 0.024930747922437674,
+      "grad_norm": 0.38493359088897705,
+      "learning_rate": 4.5e-06,
+      "loss": 0.7502,
+      "step": 90
+    },
+    {
+      "epoch": 0.025207756232686982,
+      "grad_norm": 0.4111658036708832,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.7986,
+      "step": 91
+    },
+    {
+      "epoch": 0.02548476454293629,
+      "grad_norm": 0.40969961881637573,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.8424,
+      "step": 92
+    },
+    {
+      "epoch": 0.025761772853185594,
+      "grad_norm": 0.3882950246334076,
+      "learning_rate": 4.65e-06,
+      "loss": 0.7825,
+      "step": 93
+    },
+    {
+      "epoch": 0.0260387811634349,
+      "grad_norm": 0.4332575798034668,
+      "learning_rate": 4.7e-06,
+      "loss": 0.824,
+      "step": 94
+    },
+    {
+      "epoch": 0.02631578947368421,
+      "grad_norm": 0.4235541820526123,
+      "learning_rate": 4.75e-06,
+      "loss": 0.8532,
+      "step": 95
+    },
+    {
+      "epoch": 0.026592797783933517,
+      "grad_norm": 0.408812940120697,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.8803,
+      "step": 96
+    },
+    {
+      "epoch": 0.026869806094182824,
+      "grad_norm": 0.40831273794174194,
+      "learning_rate": 4.85e-06,
+      "loss": 0.8272,
+      "step": 97
+    },
+    {
+      "epoch": 0.027146814404432132,
+      "grad_norm": 0.4236239492893219,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.841,
+      "step": 98
+    },
+    {
+      "epoch": 0.02742382271468144,
+      "grad_norm": 0.40624263882637024,
+      "learning_rate": 4.95e-06,
+      "loss": 0.8278,
+      "step": 99
+    },
+    {
+      "epoch": 0.027700831024930747,
+      "grad_norm": 0.41792380809783936,
+      "learning_rate": 5e-06,
+      "loss": 0.842,
+      "step": 100
+    },
+    {
+      "epoch": 0.027977839335180055,
+      "grad_norm": 0.4306834936141968,
+      "learning_rate": 4.999999973459309e-06,
+      "loss": 0.8569,
+      "step": 101
+    },
+    {
+      "epoch": 0.028254847645429362,
+      "grad_norm": 0.40494856238365173,
+      "learning_rate": 4.999999893837233e-06,
+      "loss": 0.818,
+      "step": 102
+    },
+    {
+      "epoch": 0.02853185595567867,
+      "grad_norm": 0.3964189291000366,
+      "learning_rate": 4.999999761133774e-06,
+      "loss": 0.8353,
+      "step": 103
+    },
+    {
+      "epoch": 0.028808864265927978,
+      "grad_norm": 0.4005384147167206,
+      "learning_rate": 4.999999575348937e-06,
+      "loss": 0.8338,
+      "step": 104
+    },
+    {
+      "epoch": 0.029085872576177285,
+      "grad_norm": 0.41835519671440125,
+      "learning_rate": 4.999999336482725e-06,
+      "loss": 0.8211,
+      "step": 105
+    },
+    {
+      "epoch": 0.029362880886426593,
+      "grad_norm": 0.4398205280303955,
+      "learning_rate": 4.999999044535142e-06,
+      "loss": 0.8708,
+      "step": 106
+    },
+    {
+      "epoch": 0.0296398891966759,
+      "grad_norm": 0.4256204664707184,
+      "learning_rate": 4.999998699506195e-06,
+      "loss": 0.8362,
+      "step": 107
+    },
+    {
+      "epoch": 0.029916897506925208,
+      "grad_norm": 0.4092758893966675,
+      "learning_rate": 4.999998301395892e-06,
+      "loss": 0.793,
+      "step": 108
+    },
+    {
+      "epoch": 0.030193905817174516,
+      "grad_norm": 0.4022393822669983,
+      "learning_rate": 4.99999785020424e-06,
+      "loss": 0.855,
+      "step": 109
+    },
+    {
+      "epoch": 0.030470914127423823,
+      "grad_norm": 0.39198121428489685,
+      "learning_rate": 4.9999973459312505e-06,
+      "loss": 0.781,
+      "step": 110
+    },
+    {
+      "epoch": 0.03074792243767313,
+      "grad_norm": 0.40529149770736694,
+      "learning_rate": 4.999996788576932e-06,
+      "loss": 0.8363,
+      "step": 111
+    },
+    {
+      "epoch": 0.03102493074792244,
+      "grad_norm": 0.4149932563304901,
+      "learning_rate": 4.999996178141298e-06,
+      "loss": 0.8145,
+      "step": 112
+    },
+    {
+      "epoch": 0.031301939058171746,
+      "grad_norm": 0.4344826340675354,
+      "learning_rate": 4.99999551462436e-06,
+      "loss": 0.8545,
+      "step": 113
+    },
+    {
+      "epoch": 0.031578947368421054,
+      "grad_norm": 0.41700538992881775,
+      "learning_rate": 4.999994798026133e-06,
+      "loss": 0.8152,
+      "step": 114
+    },
+    {
+      "epoch": 0.03185595567867036,
+      "grad_norm": 0.4254540205001831,
+      "learning_rate": 4.999994028346633e-06,
+      "loss": 0.8452,
+      "step": 115
+    },
+    {
+      "epoch": 0.03213296398891967,
+      "grad_norm": 0.4131767451763153,
+      "learning_rate": 4.999993205585875e-06,
+      "loss": 0.8019,
+      "step": 116
+    },
+    {
+      "epoch": 0.032409972299168976,
+      "grad_norm": 0.4246065616607666,
+      "learning_rate": 4.999992329743877e-06,
+      "loss": 0.8517,
+      "step": 117
+    },
+    {
+      "epoch": 0.032686980609418284,
+      "grad_norm": 0.3999440670013428,
+      "learning_rate": 4.999991400820658e-06,
+      "loss": 0.7985,
+      "step": 118
+    },
+    {
+      "epoch": 0.03296398891966759,
+      "grad_norm": 0.41130325198173523,
+      "learning_rate": 4.999990418816236e-06,
+      "loss": 0.8455,
+      "step": 119
+    },
+    {
+      "epoch": 0.0332409972299169,
+      "grad_norm": 0.40182363986968994,
+      "learning_rate": 4.999989383730634e-06,
+      "loss": 0.8081,
+      "step": 120
+    },
+    {
+      "epoch": 0.03351800554016621,
+      "grad_norm": 0.4479239583015442,
+      "learning_rate": 4.999988295563873e-06,
+      "loss": 0.8775,
+      "step": 121
+    },
+    {
+      "epoch": 0.033795013850415515,
+      "grad_norm": 0.41858404874801636,
+      "learning_rate": 4.999987154315977e-06,
+      "loss": 0.7986,
+      "step": 122
+    },
+    {
+      "epoch": 0.03407202216066482,
+      "grad_norm": 0.4002572000026703,
+      "learning_rate": 4.999985959986968e-06,
+      "loss": 0.7207,
+      "step": 123
+    },
+    {
+      "epoch": 0.03434903047091413,
+      "grad_norm": 0.40468156337738037,
+      "learning_rate": 4.999984712576874e-06,
+      "loss": 0.7843,
+      "step": 124
+    },
+    {
+      "epoch": 0.03462603878116344,
+      "grad_norm": 0.42578715085983276,
+      "learning_rate": 4.99998341208572e-06,
+      "loss": 0.761,
+      "step": 125
+    },
+    {
+      "epoch": 0.034903047091412745,
+      "grad_norm": 0.4391866624355316,
+      "learning_rate": 4.9999820585135335e-06,
+      "loss": 0.8479,
+      "step": 126
+    },
+    {
+      "epoch": 0.03518005540166205,
+      "grad_norm": 0.4086853861808777,
+      "learning_rate": 4.999980651860345e-06,
+      "loss": 0.7987,
+      "step": 127
+    },
+    {
+      "epoch": 0.03545706371191136,
+      "grad_norm": 0.4147675335407257,
+      "learning_rate": 4.999979192126181e-06,
+      "loss": 0.8087,
+      "step": 128
+    },
+    {
+      "epoch": 0.03573407202216067,
+      "grad_norm": 0.40620920062065125,
+      "learning_rate": 4.999977679311076e-06,
+      "loss": 0.8191,
+      "step": 129
+    },
+    {
+      "epoch": 0.036011080332409975,
+      "grad_norm": 0.41266223788261414,
+      "learning_rate": 4.999976113415059e-06,
+      "loss": 0.8403,
+      "step": 130
+    },
+    {
+      "epoch": 0.03628808864265928,
+      "grad_norm": 0.4243127405643463,
+      "learning_rate": 4.999974494438166e-06,
+      "loss": 0.8207,
+      "step": 131
+    },
+    {
+      "epoch": 0.03656509695290859,
+      "grad_norm": 0.4049611985683441,
+      "learning_rate": 4.99997282238043e-06,
+      "loss": 0.7541,
+      "step": 132
+    },
+    {
+      "epoch": 0.03684210526315789,
+      "grad_norm": 0.43927672505378723,
+      "learning_rate": 4.999971097241887e-06,
+      "loss": 0.8645,
+      "step": 133
+    },
+    {
+      "epoch": 0.0371191135734072,
+      "grad_norm": 0.4410925805568695,
+      "learning_rate": 4.999969319022573e-06,
+      "loss": 0.8064,
+      "step": 134
+    },
+    {
+      "epoch": 0.037396121883656507,
+      "grad_norm": 0.42407283186912537,
+      "learning_rate": 4.999967487722527e-06,
+      "loss": 0.7815,
+      "step": 135
+    },
+    {
+      "epoch": 0.037673130193905814,
+      "grad_norm": 0.4339797794818878,
+      "learning_rate": 4.9999656033417865e-06,
+      "loss": 0.7834,
+      "step": 136
+    },
+    {
+      "epoch": 0.03795013850415512,
+      "grad_norm": 0.4262239634990692,
+      "learning_rate": 4.999963665880392e-06,
+      "loss": 0.8049,
+      "step": 137
+    },
+    {
+      "epoch": 0.03822714681440443,
+      "grad_norm": 0.41159674525260925,
+      "learning_rate": 4.999961675338384e-06,
+      "loss": 0.8195,
+      "step": 138
+    },
+    {
+      "epoch": 0.03850415512465374,
+      "grad_norm": 0.39753422141075134,
+      "learning_rate": 4.999959631715806e-06,
+      "loss": 0.7023,
+      "step": 139
+    },
+    {
+      "epoch": 0.038781163434903045,
+      "grad_norm": 0.4160650968551636,
+      "learning_rate": 4.999957535012701e-06,
+      "loss": 0.7488,
+      "step": 140
+    },
+    {
+      "epoch": 0.03905817174515235,
+      "grad_norm": 0.4177490174770355,
+      "learning_rate": 4.999955385229113e-06,
+      "loss": 0.8098,
+      "step": 141
+    },
+    {
+      "epoch": 0.03933518005540166,
+      "grad_norm": 0.441594660282135,
+      "learning_rate": 4.999953182365088e-06,
+      "loss": 0.825,
+      "step": 142
+    },
+    {
+      "epoch": 0.03961218836565097,
+      "grad_norm": 0.44137829542160034,
+      "learning_rate": 4.999950926420673e-06,
+      "loss": 0.7645,
+      "step": 143
+    },
+    {
+      "epoch": 0.039889196675900275,
+      "grad_norm": 0.44875359535217285,
+      "learning_rate": 4.999948617395916e-06,
+      "loss": 0.8157,
+      "step": 144
+    },
+    {
+      "epoch": 0.04016620498614958,
+      "grad_norm": 0.4129585325717926,
+      "learning_rate": 4.999946255290865e-06,
+      "loss": 0.8188,
+      "step": 145
+    },
+    {
+      "epoch": 0.04044321329639889,
+      "grad_norm": 0.4289536476135254,
+      "learning_rate": 4.999943840105571e-06,
+      "loss": 0.7128,
+      "step": 146
+    },
+    {
+      "epoch": 0.0407202216066482,
+      "grad_norm": 0.4285975992679596,
+      "learning_rate": 4.999941371840084e-06,
+      "loss": 0.7924,
+      "step": 147
+    },
+    {
+      "epoch": 0.040997229916897505,
+      "grad_norm": 0.4204922318458557,
+      "learning_rate": 4.99993885049446e-06,
+      "loss": 0.7936,
+      "step": 148
+    },
+    {
+      "epoch": 0.04127423822714681,
+      "grad_norm": 0.43389734625816345,
+      "learning_rate": 4.999936276068749e-06,
+      "loss": 0.8296,
+      "step": 149
+    },
+    {
+      "epoch": 0.04155124653739612,
+      "grad_norm": 0.45425310730934143,
+      "learning_rate": 4.999933648563005e-06,
+      "loss": 0.8428,
+      "step": 150
+    },
+    {
+      "epoch": 0.04182825484764543,
+      "grad_norm": 0.44408005475997925,
+      "learning_rate": 4.9999309679772875e-06,
+      "loss": 0.8612,
+      "step": 151
+    },
+    {
+      "epoch": 0.042105263157894736,
+      "grad_norm": 0.44268134236335754,
+      "learning_rate": 4.999928234311651e-06,
+      "loss": 0.8082,
+      "step": 152
+    },
+    {
+      "epoch": 0.042382271468144044,
+      "grad_norm": 0.43825775384902954,
+      "learning_rate": 4.999925447566154e-06,
+      "loss": 0.8004,
+      "step": 153
+    },
+    {
+      "epoch": 0.04265927977839335,
+      "grad_norm": 0.42596134543418884,
+      "learning_rate": 4.999922607740855e-06,
+      "loss": 0.833,
+      "step": 154
+    },
+    {
+      "epoch": 0.04293628808864266,
+      "grad_norm": 0.42418310046195984,
+      "learning_rate": 4.999919714835816e-06,
+      "loss": 0.8072,
+      "step": 155
+    },
+    {
+      "epoch": 0.043213296398891966,
+      "grad_norm": 0.4111458361148834,
+      "learning_rate": 4.999916768851096e-06,
+      "loss": 0.7352,
+      "step": 156
+    },
+    {
+      "epoch": 0.043490304709141274,
+      "grad_norm": 0.4288739860057831,
+      "learning_rate": 4.99991376978676e-06,
+      "loss": 0.8314,
+      "step": 157
+    },
+    {
+      "epoch": 0.04376731301939058,
+      "grad_norm": 0.4291544556617737,
+      "learning_rate": 4.999910717642871e-06,
+      "loss": 0.8112,
+      "step": 158
+    },
+    {
+      "epoch": 0.04404432132963989,
+      "grad_norm": 0.42698702216148376,
+      "learning_rate": 4.9999076124194925e-06,
+      "loss": 0.7727,
+      "step": 159
+    },
+    {
+      "epoch": 0.0443213296398892,
+      "grad_norm": 0.3970215320587158,
+      "learning_rate": 4.999904454116691e-06,
+      "loss": 0.753,
+      "step": 160
+    },
+    {
+      "epoch": 0.044598337950138504,
+      "grad_norm": 0.4505182206630707,
+      "learning_rate": 4.9999012427345345e-06,
+      "loss": 0.8088,
+      "step": 161
+    },
+    {
+      "epoch": 0.04487534626038781,
+      "grad_norm": 0.4358536899089813,
+      "learning_rate": 4.9998979782730905e-06,
+      "loss": 0.8246,
+      "step": 162
+    },
+    {
+      "epoch": 0.04515235457063712,
+      "grad_norm": 0.4245203733444214,
+      "learning_rate": 4.999894660732428e-06,
+      "loss": 0.7818,
+      "step": 163
+    },
+    {
+      "epoch": 0.04542936288088643,
+      "grad_norm": 0.4643189609050751,
+      "learning_rate": 4.999891290112619e-06,
+      "loss": 0.8118,
+      "step": 164
+    },
+    {
+      "epoch": 0.045706371191135735,
+      "grad_norm": 0.4445042312145233,
+      "learning_rate": 4.999887866413732e-06,
+      "loss": 0.8134,
+      "step": 165
+    },
+    {
+      "epoch": 0.04598337950138504,
+      "grad_norm": 0.47488442063331604,
+      "learning_rate": 4.999884389635843e-06,
+      "loss": 0.8216,
+      "step": 166
+    },
+    {
+      "epoch": 0.04626038781163435,
+      "grad_norm": 0.44003528356552124,
+      "learning_rate": 4.999880859779024e-06,
+      "loss": 0.8301,
+      "step": 167
+    },
+    {
+      "epoch": 0.04653739612188366,
+      "grad_norm": 0.45998671650886536,
+      "learning_rate": 4.99987727684335e-06,
+      "loss": 0.7673,
+      "step": 168
+    },
+    {
+      "epoch": 0.046814404432132965,
+      "grad_norm": 0.43222859501838684,
+      "learning_rate": 4.999873640828897e-06,
+      "loss": 0.8208,
+      "step": 169
+    },
+    {
+      "epoch": 0.04709141274238227,
+      "grad_norm": 0.45207300782203674,
+      "learning_rate": 4.999869951735744e-06,
+      "loss": 0.8097,
+      "step": 170
+    },
+    {
+      "epoch": 0.04736842105263158,
+      "grad_norm": 0.4396827220916748,
+      "learning_rate": 4.999866209563967e-06,
+      "loss": 0.8436,
+      "step": 171
+    },
+    {
+      "epoch": 0.04764542936288089,
+      "grad_norm": 0.4444088935852051,
+      "learning_rate": 4.999862414313647e-06,
+      "loss": 0.822,
+      "step": 172
+    },
+    {
+      "epoch": 0.047922437673130196,
+      "grad_norm": 0.4522409439086914,
+      "learning_rate": 4.999858565984864e-06,
+      "loss": 0.8165,
+      "step": 173
+    },
+    {
+      "epoch": 0.0481994459833795,
+      "grad_norm": 0.42866262793540955,
+      "learning_rate": 4.999854664577699e-06,
+      "loss": 0.7353,
+      "step": 174
+    },
+    {
+      "epoch": 0.04847645429362881,
+      "grad_norm": 0.43448740243911743,
+      "learning_rate": 4.999850710092236e-06,
+      "loss": 0.7896,
+      "step": 175
+    },
+    {
+      "epoch": 0.04875346260387812,
+      "grad_norm": 0.42707017064094543,
+      "learning_rate": 4.999846702528559e-06,
+      "loss": 0.7476,
+      "step": 176
+    },
+    {
+      "epoch": 0.049030470914127426,
+      "grad_norm": 0.4463941156864166,
+      "learning_rate": 4.999842641886752e-06,
+      "loss": 0.88,
+      "step": 177
+    },
+    {
+      "epoch": 0.049307479224376734,
+      "grad_norm": 0.44734442234039307,
+      "learning_rate": 4.999838528166902e-06,
+      "loss": 0.7961,
+      "step": 178
+    },
+    {
+      "epoch": 0.04958448753462604,
+      "grad_norm": 0.4721640944480896,
+      "learning_rate": 4.999834361369096e-06,
+      "loss": 0.8119,
+      "step": 179
+    },
+    {
+      "epoch": 0.04986149584487535,
+      "grad_norm": 0.4124251902103424,
+      "learning_rate": 4.999830141493423e-06,
+      "loss": 0.7843,
+      "step": 180
+    },
+    {
+      "epoch": 0.05013850415512466,
+      "grad_norm": 0.4647778868675232,
+      "learning_rate": 4.999825868539971e-06,
+      "loss": 0.7736,
+      "step": 181
+    },
+    {
+      "epoch": 0.050415512465373964,
+      "grad_norm": 0.4526465833187103,
+      "learning_rate": 4.999821542508833e-06,
+      "loss": 0.7312,
+      "step": 182
+    },
+    {
+      "epoch": 0.05069252077562327,
+      "grad_norm": 0.4479784369468689,
+      "learning_rate": 4.9998171634001e-06,
+      "loss": 0.768,
+      "step": 183
+    },
+    {
+      "epoch": 0.05096952908587258,
+      "grad_norm": 0.4574298560619354,
+      "learning_rate": 4.9998127312138645e-06,
+      "loss": 0.8163,
+      "step": 184
+    },
+    {
+      "epoch": 0.05124653739612189,
+      "grad_norm": 0.470766544342041,
+      "learning_rate": 4.99980824595022e-06,
+      "loss": 0.7886,
+      "step": 185
+    },
+    {
+      "epoch": 0.05152354570637119,
+      "grad_norm": 0.46268603205680847,
+      "learning_rate": 4.999803707609264e-06,
+      "loss": 0.7692,
+      "step": 186
+    },
+    {
+      "epoch": 0.051800554016620495,
+      "grad_norm": 0.45423924922943115,
+      "learning_rate": 4.999799116191091e-06,
+      "loss": 0.8334,
+      "step": 187
+    },
+    {
+      "epoch": 0.0520775623268698,
+      "grad_norm": 0.4589080512523651,
+      "learning_rate": 4.9997944716957985e-06,
+      "loss": 0.7835,
+      "step": 188
+    },
+    {
+      "epoch": 0.05235457063711911,
+      "grad_norm": 0.4934840202331543,
+      "learning_rate": 4.999789774123486e-06,
+      "loss": 0.7404,
+      "step": 189
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 0.4710850119590759,
+      "learning_rate": 4.999785023474253e-06,
+      "loss": 0.8659,
+      "step": 190
+    },
+    {
+      "epoch": 0.052908587257617726,
+      "grad_norm": 0.44491663575172424,
+      "learning_rate": 4.9997802197482e-06,
+      "loss": 0.8241,
+      "step": 191
+    },
+    {
+      "epoch": 0.05318559556786703,
+      "grad_norm": 0.43371081352233887,
+      "learning_rate": 4.9997753629454295e-06,
+      "loss": 0.8145,
+      "step": 192
+    },
+    {
+      "epoch": 0.05346260387811634,
+      "grad_norm": 0.4385470151901245,
+      "learning_rate": 4.999770453066043e-06,
+      "loss": 0.7779,
+      "step": 193
+    },
+    {
+      "epoch": 0.05373961218836565,
+      "grad_norm": 0.46287888288497925,
+      "learning_rate": 4.999765490110148e-06,
+      "loss": 0.7731,
+      "step": 194
+    },
+    {
+      "epoch": 0.054016620498614956,
+      "grad_norm": 0.4724902808666229,
+      "learning_rate": 4.999760474077847e-06,
+      "loss": 0.8538,
+      "step": 195
+    },
+    {
+      "epoch": 0.054293628808864264,
+      "grad_norm": 0.4405905604362488,
+      "learning_rate": 4.999755404969247e-06,
+      "loss": 0.7764,
+      "step": 196
+    },
+    {
+      "epoch": 0.05457063711911357,
+      "grad_norm": 0.45440441370010376,
+      "learning_rate": 4.9997502827844565e-06,
+      "loss": 0.7488,
+      "step": 197
+    },
+    {
+      "epoch": 0.05484764542936288,
+      "grad_norm": 0.4569130539894104,
+      "learning_rate": 4.999745107523583e-06,
+      "loss": 0.7604,
+      "step": 198
+    },
+    {
+      "epoch": 0.05512465373961219,
+      "grad_norm": 0.44485560059547424,
+      "learning_rate": 4.999739879186739e-06,
+      "loss": 0.7815,
+      "step": 199
+    },
+    {
+      "epoch": 0.055401662049861494,
+      "grad_norm": 0.46380650997161865,
+      "learning_rate": 4.999734597774033e-06,
+      "loss": 0.8247,
+      "step": 200
+    },
+    {
+      "epoch": 0.0556786703601108,
+      "grad_norm": 0.43232858180999756,
+      "learning_rate": 4.999729263285578e-06,
+      "loss": 0.7638,
+      "step": 201
+    },
+    {
+      "epoch": 0.05595567867036011,
+      "grad_norm": 0.4762643277645111,
+      "learning_rate": 4.999723875721486e-06,
+      "loss": 0.8308,
+      "step": 202
+    },
+    {
+      "epoch": 0.05623268698060942,
+      "grad_norm": 0.4557073414325714,
+      "learning_rate": 4.999718435081873e-06,
+      "loss": 0.7677,
+      "step": 203
+    },
+    {
+      "epoch": 0.056509695290858725,
+      "grad_norm": 0.4623831808567047,
+      "learning_rate": 4.999712941366854e-06,
+      "loss": 0.7769,
+      "step": 204
+    },
+    {
+      "epoch": 0.05678670360110803,
+      "grad_norm": 0.4392584264278412,
+      "learning_rate": 4.999707394576547e-06,
+      "loss": 0.7662,
+      "step": 205
+    },
+    {
+      "epoch": 0.05706371191135734,
+      "grad_norm": 0.46204671263694763,
+      "learning_rate": 4.999701794711067e-06,
+      "loss": 0.7529,
+      "step": 206
+    },
+    {
+      "epoch": 0.05734072022160665,
+      "grad_norm": 0.4556841254234314,
+      "learning_rate": 4.999696141770535e-06,
+      "loss": 0.7985,
+      "step": 207
+    },
+    {
+      "epoch": 0.057617728531855955,
+      "grad_norm": 0.4767592251300812,
+      "learning_rate": 4.9996904357550705e-06,
+      "loss": 0.7712,
+      "step": 208
+    },
+    {
+      "epoch": 0.05789473684210526,
+      "grad_norm": 0.4875626266002655,
+      "learning_rate": 4.9996846766647935e-06,
+      "loss": 0.8396,
+      "step": 209
+    },
+    {
+      "epoch": 0.05817174515235457,
+      "grad_norm": 0.48093894124031067,
+      "learning_rate": 4.999678864499828e-06,
+      "loss": 0.7904,
+      "step": 210
+    },
+    {
+      "epoch": 0.05844875346260388,
+      "grad_norm": 0.4370277225971222,
+      "learning_rate": 4.999672999260297e-06,
+      "loss": 0.7666,
+      "step": 211
+    },
+    {
+      "epoch": 0.058725761772853186,
+      "grad_norm": 0.500248372554779,
+      "learning_rate": 4.9996670809463245e-06,
+      "loss": 0.8361,
+      "step": 212
+    },
+    {
+      "epoch": 0.05900277008310249,
+      "grad_norm": 0.47237977385520935,
+      "learning_rate": 4.999661109558036e-06,
+      "loss": 0.8191,
+      "step": 213
+    },
+    {
+      "epoch": 0.0592797783933518,
+      "grad_norm": 0.4805334210395813,
+      "learning_rate": 4.99965508509556e-06,
+      "loss": 0.8315,
+      "step": 214
+    },
+    {
+      "epoch": 0.05955678670360111,
+      "grad_norm": 0.4623453617095947,
+      "learning_rate": 4.999649007559022e-06,
+      "loss": 0.8125,
+      "step": 215
+    },
+    {
+      "epoch": 0.059833795013850416,
+      "grad_norm": 0.4652257561683655,
+      "learning_rate": 4.999642876948553e-06,
+      "loss": 0.7935,
+      "step": 216
+    },
+    {
+      "epoch": 0.060110803324099724,
+      "grad_norm": 0.47447672486305237,
+      "learning_rate": 4.9996366932642814e-06,
+      "loss": 0.8014,
+      "step": 217
+    },
+    {
+      "epoch": 0.06038781163434903,
+      "grad_norm": 0.45881807804107666,
+      "learning_rate": 4.99963045650634e-06,
+      "loss": 0.7345,
+      "step": 218
+    },
+    {
+      "epoch": 0.06066481994459834,
+      "grad_norm": 0.4540233612060547,
+      "learning_rate": 4.99962416667486e-06,
+      "loss": 0.8386,
+      "step": 219
+    },
+    {
+      "epoch": 0.060941828254847646,
+      "grad_norm": 0.461611807346344,
+      "learning_rate": 4.999617823769977e-06,
+      "loss": 0.7832,
+      "step": 220
+    },
+    {
+      "epoch": 0.061218836565096954,
+      "grad_norm": 0.4760967791080475,
+      "learning_rate": 4.999611427791823e-06,
+      "loss": 0.8167,
+      "step": 221
+    },
+    {
+      "epoch": 0.06149584487534626,
+      "grad_norm": 0.476432740688324,
+      "learning_rate": 4.999604978740535e-06,
+      "loss": 0.8037,
+      "step": 222
+    },
+    {
+      "epoch": 0.06177285318559557,
+      "grad_norm": 0.4600237309932709,
+      "learning_rate": 4.999598476616251e-06,
+      "loss": 0.7826,
+      "step": 223
+    },
+    {
+      "epoch": 0.06204986149584488,
+      "grad_norm": 0.45525020360946655,
+      "learning_rate": 4.999591921419107e-06,
+      "loss": 0.7856,
+      "step": 224
+    },
+    {
+      "epoch": 0.062326869806094184,
+      "grad_norm": 0.47345778346061707,
+      "learning_rate": 4.999585313149244e-06,
+      "loss": 0.755,
+      "step": 225
+    },
+    {
+      "epoch": 0.06260387811634349,
+      "grad_norm": 0.46238973736763,
+      "learning_rate": 4.999578651806801e-06,
+      "loss": 0.8153,
+      "step": 226
+    },
+    {
+      "epoch": 0.0628808864265928,
+      "grad_norm": 0.48064154386520386,
+      "learning_rate": 4.99957193739192e-06,
+      "loss": 0.7965,
+      "step": 227
+    },
+    {
+      "epoch": 0.06315789473684211,
+      "grad_norm": 0.5076362490653992,
+      "learning_rate": 4.9995651699047436e-06,
+      "loss": 0.8075,
+      "step": 228
+    },
+    {
+      "epoch": 0.06343490304709141,
+      "grad_norm": 0.4526555836200714,
+      "learning_rate": 4.999558349345416e-06,
+      "loss": 0.8074,
+      "step": 229
+    },
+    {
+      "epoch": 0.06371191135734072,
+      "grad_norm": 0.4602035880088806,
+      "learning_rate": 4.999551475714081e-06,
+      "loss": 0.7598,
+      "step": 230
+    },
+    {
+      "epoch": 0.06398891966759003,
+      "grad_norm": 0.46630504727363586,
+      "learning_rate": 4.999544549010885e-06,
+      "loss": 0.7549,
+      "step": 231
+    },
+    {
+      "epoch": 0.06426592797783934,
+      "grad_norm": 0.46300217509269714,
+      "learning_rate": 4.999537569235975e-06,
+      "loss": 0.7216,
+      "step": 232
+    },
+    {
+      "epoch": 0.06454293628808865,
+      "grad_norm": 0.46852821111679077,
+      "learning_rate": 4.9995305363895e-06,
+      "loss": 0.7966,
+      "step": 233
+    },
+    {
+      "epoch": 0.06481994459833795,
+      "grad_norm": 0.49596261978149414,
+      "learning_rate": 4.999523450471609e-06,
+      "loss": 0.8022,
+      "step": 234
+    },
+    {
+      "epoch": 0.06509695290858726,
+      "grad_norm": 0.48228102922439575,
+      "learning_rate": 4.99951631148245e-06,
+      "loss": 0.7834,
+      "step": 235
+    },
+    {
+      "epoch": 0.06537396121883657,
+      "grad_norm": 0.46538177132606506,
+      "learning_rate": 4.999509119422178e-06,
+      "loss": 0.7692,
+      "step": 236
+    },
+    {
+      "epoch": 0.06565096952908588,
+      "grad_norm": 0.4842844009399414,
+      "learning_rate": 4.999501874290944e-06,
+      "loss": 0.7869,
+      "step": 237
+    },
+    {
+      "epoch": 0.06592797783933518,
+      "grad_norm": 0.4862155616283417,
+      "learning_rate": 4.999494576088903e-06,
+      "loss": 0.8026,
+      "step": 238
+    },
+    {
+      "epoch": 0.06620498614958449,
+      "grad_norm": 0.4713565409183502,
+      "learning_rate": 4.999487224816208e-06,
+      "loss": 0.7665,
+      "step": 239
+    },
+    {
+      "epoch": 0.0664819944598338,
+      "grad_norm": 0.4564686715602875,
+      "learning_rate": 4.9994798204730165e-06,
+      "loss": 0.791,
+      "step": 240
+    },
+    {
+      "epoch": 0.0667590027700831,
+      "grad_norm": 0.4710893929004669,
+      "learning_rate": 4.999472363059485e-06,
+      "loss": 0.7406,
+      "step": 241
+    },
+    {
+      "epoch": 0.06703601108033241,
+      "grad_norm": 0.4848516583442688,
+      "learning_rate": 4.999464852575774e-06,
+      "loss": 0.8032,
+      "step": 242
+    },
+    {
+      "epoch": 0.06731301939058172,
+      "grad_norm": 0.4644072651863098,
+      "learning_rate": 4.9994572890220395e-06,
+      "loss": 0.7975,
+      "step": 243
+    },
+    {
+      "epoch": 0.06759002770083103,
+      "grad_norm": 0.46789732575416565,
+      "learning_rate": 4.999449672398445e-06,
+      "loss": 0.7652,
+      "step": 244
+    },
+    {
+      "epoch": 0.06786703601108034,
+      "grad_norm": 0.4857868552207947,
+      "learning_rate": 4.99944200270515e-06,
+      "loss": 0.7853,
+      "step": 245
+    },
+    {
+      "epoch": 0.06814404432132964,
+      "grad_norm": 0.4426226019859314,
+      "learning_rate": 4.999434279942319e-06,
+      "loss": 0.6664,
+      "step": 246
+    },
+    {
+      "epoch": 0.06842105263157895,
+      "grad_norm": 0.4759196937084198,
+      "learning_rate": 4.9994265041101155e-06,
+      "loss": 0.8032,
+      "step": 247
+    },
+    {
+      "epoch": 0.06869806094182826,
+      "grad_norm": 0.4856152832508087,
+      "learning_rate": 4.999418675208704e-06,
+      "loss": 0.7578,
+      "step": 248
+    },
+    {
+      "epoch": 0.06897506925207757,
+      "grad_norm": 0.4679270088672638,
+      "learning_rate": 4.999410793238252e-06,
+      "loss": 0.7618,
+      "step": 249
+    },
+    {
+      "epoch": 0.06925207756232687,
+      "grad_norm": 0.48782986402511597,
+      "learning_rate": 4.999402858198925e-06,
+      "loss": 0.7189,
+      "step": 250
+    },
+    {
+      "epoch": 0.06952908587257618,
+      "grad_norm": 0.44610273838043213,
+      "learning_rate": 4.999394870090894e-06,
+      "loss": 0.784,
+      "step": 251
+    },
+    {
+      "epoch": 0.06980609418282549,
+      "grad_norm": 0.4906792938709259,
+      "learning_rate": 4.9993868289143255e-06,
+      "loss": 0.7445,
+      "step": 252
+    },
+    {
+      "epoch": 0.0700831024930748,
+      "grad_norm": 0.4803594946861267,
+      "learning_rate": 4.9993787346693924e-06,
+      "loss": 0.7581,
+      "step": 253
+    },
+    {
+      "epoch": 0.0703601108033241,
+      "grad_norm": 0.47042086720466614,
+      "learning_rate": 4.999370587356267e-06,
+      "loss": 0.7409,
+      "step": 254
+    },
+    {
+      "epoch": 0.07063711911357341,
+      "grad_norm": 0.5199850797653198,
+      "learning_rate": 4.999362386975121e-06,
+      "loss": 0.7776,
+      "step": 255
+    },
+    {
+      "epoch": 0.07091412742382272,
+      "grad_norm": 0.5027449131011963,
+      "learning_rate": 4.999354133526127e-06,
+      "loss": 0.808,
+      "step": 256
+    },
+    {
+      "epoch": 0.07119113573407203,
+      "grad_norm": 0.5312409996986389,
+      "learning_rate": 4.999345827009464e-06,
+      "loss": 0.8147,
+      "step": 257
+    },
+    {
+      "epoch": 0.07146814404432134,
+      "grad_norm": 0.4875726103782654,
+      "learning_rate": 4.9993374674253065e-06,
+      "loss": 0.7487,
+      "step": 258
+    },
+    {
+      "epoch": 0.07174515235457064,
+      "grad_norm": 0.46339771151542664,
+      "learning_rate": 4.999329054773832e-06,
+      "loss": 0.6929,
+      "step": 259
+    },
+    {
+      "epoch": 0.07202216066481995,
+      "grad_norm": 0.4794660210609436,
+      "learning_rate": 4.999320589055218e-06,
+      "loss": 0.7913,
+      "step": 260
+    },
+    {
+      "epoch": 0.07229916897506926,
+      "grad_norm": 0.481033056974411,
+      "learning_rate": 4.999312070269647e-06,
+      "loss": 0.7211,
+      "step": 261
+    },
+    {
+      "epoch": 0.07257617728531857,
+      "grad_norm": 0.49464985728263855,
+      "learning_rate": 4.999303498417296e-06,
+      "loss": 0.7695,
+      "step": 262
+    },
+    {
+      "epoch": 0.07285318559556787,
+      "grad_norm": 0.5082013607025146,
+      "learning_rate": 4.99929487349835e-06,
+      "loss": 0.7431,
+      "step": 263
+    },
+    {
+      "epoch": 0.07313019390581718,
+      "grad_norm": 0.45051777362823486,
+      "learning_rate": 4.999286195512992e-06,
+      "loss": 0.7167,
+      "step": 264
+    },
+    {
+      "epoch": 0.07340720221606649,
+      "grad_norm": 0.47732335329055786,
+      "learning_rate": 4.999277464461405e-06,
+      "loss": 0.7561,
+      "step": 265
+    },
+    {
+      "epoch": 0.07368421052631578,
+      "grad_norm": 0.4845159649848938,
+      "learning_rate": 4.999268680343775e-06,
+      "loss": 0.7316,
+      "step": 266
+    },
+    {
+      "epoch": 0.07396121883656509,
+      "grad_norm": 0.522322952747345,
+      "learning_rate": 4.999259843160289e-06,
+      "loss": 0.7979,
+      "step": 267
+    },
+    {
+      "epoch": 0.0742382271468144,
+      "grad_norm": 0.4906967282295227,
+      "learning_rate": 4.9992509529111335e-06,
+      "loss": 0.7278,
+      "step": 268
+    },
+    {
+      "epoch": 0.0745152354570637,
+      "grad_norm": 0.4843719005584717,
+      "learning_rate": 4.999242009596498e-06,
+      "loss": 0.7645,
+      "step": 269
+    },
+    {
+      "epoch": 0.07479224376731301,
+      "grad_norm": 0.4734237790107727,
+      "learning_rate": 4.9992330132165725e-06,
+      "loss": 0.7667,
+      "step": 270
+    },
+    {
+      "epoch": 0.07506925207756232,
+      "grad_norm": 0.49938181042671204,
+      "learning_rate": 4.999223963771548e-06,
+      "loss": 0.825,
+      "step": 271
+    },
+    {
+      "epoch": 0.07534626038781163,
+      "grad_norm": 0.5260105729103088,
+      "learning_rate": 4.999214861261616e-06,
+      "loss": 0.776,
+      "step": 272
+    },
+    {
+      "epoch": 0.07562326869806094,
+      "grad_norm": 0.4987190365791321,
+      "learning_rate": 4.99920570568697e-06,
+      "loss": 0.7815,
+      "step": 273
+    },
+    {
+      "epoch": 0.07590027700831024,
+      "grad_norm": 0.5055657625198364,
+      "learning_rate": 4.9991964970478055e-06,
+      "loss": 0.7818,
+      "step": 274
+    },
+    {
+      "epoch": 0.07617728531855955,
+      "grad_norm": 0.49878621101379395,
+      "learning_rate": 4.999187235344316e-06,
+      "loss": 0.7492,
+      "step": 275
+    },
+    {
+      "epoch": 0.07645429362880886,
+      "grad_norm": 0.5403164625167847,
+      "learning_rate": 4.9991779205767e-06,
+      "loss": 0.8327,
+      "step": 276
+    },
+    {
+      "epoch": 0.07673130193905817,
+      "grad_norm": 0.518882155418396,
+      "learning_rate": 4.999168552745154e-06,
+      "loss": 0.7789,
+      "step": 277
+    },
+    {
+      "epoch": 0.07700831024930747,
+      "grad_norm": 0.5157930254936218,
+      "learning_rate": 4.999159131849878e-06,
+      "loss": 0.7889,
+      "step": 278
+    },
+    {
+      "epoch": 0.07728531855955678,
+      "grad_norm": 0.4988757371902466,
+      "learning_rate": 4.999149657891071e-06,
+      "loss": 0.7961,
+      "step": 279
+    },
+    {
+      "epoch": 0.07756232686980609,
+      "grad_norm": 0.5131459832191467,
+      "learning_rate": 4.999140130868935e-06,
+      "loss": 0.7952,
+      "step": 280
+    },
+    {
+      "epoch": 0.0778393351800554,
+      "grad_norm": 0.5224674344062805,
+      "learning_rate": 4.999130550783672e-06,
+      "loss": 0.8077,
+      "step": 281
+    },
+    {
+      "epoch": 0.0781163434903047,
+      "grad_norm": 0.5174897313117981,
+      "learning_rate": 4.999120917635485e-06,
+      "loss": 0.8272,
+      "step": 282
+    },
+    {
+      "epoch": 0.07839335180055401,
+      "grad_norm": 0.4849293529987335,
+      "learning_rate": 4.999111231424579e-06,
+      "loss": 0.7451,
+      "step": 283
+    },
+    {
+      "epoch": 0.07867036011080332,
+      "grad_norm": 0.4857077896595001,
+      "learning_rate": 4.999101492151159e-06,
+      "loss": 0.7637,
+      "step": 284
+    },
+    {
+      "epoch": 0.07894736842105263,
+      "grad_norm": 0.4597320854663849,
+      "learning_rate": 4.999091699815434e-06,
+      "loss": 0.7693,
+      "step": 285
+    },
+    {
+      "epoch": 0.07922437673130193,
+      "grad_norm": 0.4780363440513611,
+      "learning_rate": 4.999081854417608e-06,
+      "loss": 0.7371,
+      "step": 286
+    },
+    {
+      "epoch": 0.07950138504155124,
+      "grad_norm": 0.4885082244873047,
+      "learning_rate": 4.999071955957894e-06,
+      "loss": 0.8039,
+      "step": 287
+    },
+    {
+      "epoch": 0.07977839335180055,
+      "grad_norm": 0.5139564871788025,
+      "learning_rate": 4.9990620044365e-06,
+      "loss": 0.7936,
+      "step": 288
+    },
+    {
+      "epoch": 0.08005540166204986,
+      "grad_norm": 0.4996238350868225,
+      "learning_rate": 4.999051999853637e-06,
+      "loss": 0.7985,
+      "step": 289
+    },
+    {
+      "epoch": 0.08033240997229917,
+      "grad_norm": 0.4952002763748169,
+      "learning_rate": 4.9990419422095185e-06,
+      "loss": 0.8066,
+      "step": 290
+    },
+    {
+      "epoch": 0.08060941828254847,
+      "grad_norm": 0.5024906992912292,
+      "learning_rate": 4.999031831504358e-06,
+      "loss": 0.7374,
+      "step": 291
+    },
+    {
+      "epoch": 0.08088642659279778,
+      "grad_norm": 0.4709744453430176,
+      "learning_rate": 4.9990216677383695e-06,
+      "loss": 0.7183,
+      "step": 292
+    },
+    {
+      "epoch": 0.08116343490304709,
+      "grad_norm": 0.509341835975647,
+      "learning_rate": 4.99901145091177e-06,
+      "loss": 0.7757,
+      "step": 293
+    },
+    {
+      "epoch": 0.0814404432132964,
+      "grad_norm": 0.49265819787979126,
+      "learning_rate": 4.999001181024775e-06,
+      "loss": 0.7329,
+      "step": 294
+    },
+    {
+      "epoch": 0.0817174515235457,
+      "grad_norm": 0.5245534181594849,
+      "learning_rate": 4.998990858077602e-06,
+      "loss": 0.7452,
+      "step": 295
+    },
+    {
+      "epoch": 0.08199445983379501,
+      "grad_norm": 0.519623339176178,
+      "learning_rate": 4.998980482070473e-06,
+      "loss": 0.7812,
+      "step": 296
+    },
+    {
+      "epoch": 0.08227146814404432,
+      "grad_norm": 0.5161734819412231,
+      "learning_rate": 4.998970053003607e-06,
+      "loss": 0.7834,
+      "step": 297
+    },
+    {
+      "epoch": 0.08254847645429363,
+      "grad_norm": 0.48719218373298645,
+      "learning_rate": 4.998959570877224e-06,
+      "loss": 0.7662,
+      "step": 298
+    },
+    {
+      "epoch": 0.08282548476454293,
+      "grad_norm": 0.5277923941612244,
+      "learning_rate": 4.998949035691548e-06,
+      "loss": 0.7422,
+      "step": 299
+    },
+    {
+      "epoch": 0.08310249307479224,
+      "grad_norm": 0.49356892704963684,
+      "learning_rate": 4.998938447446803e-06,
+      "loss": 0.7372,
+      "step": 300
+    },
+    {
+      "epoch": 0.08337950138504155,
+      "grad_norm": 0.5243819952011108,
+      "learning_rate": 4.998927806143212e-06,
+      "loss": 0.7854,
+      "step": 301
+    },
+    {
+      "epoch": 0.08365650969529086,
+      "grad_norm": 0.5223947763442993,
+      "learning_rate": 4.998917111781003e-06,
+      "loss": 0.7182,
+      "step": 302
+    },
+    {
+      "epoch": 0.08393351800554016,
+      "grad_norm": 0.49004465341567993,
+      "learning_rate": 4.998906364360401e-06,
+      "loss": 0.7404,
+      "step": 303
+    },
+    {
+      "epoch": 0.08421052631578947,
+      "grad_norm": 0.502047598361969,
+      "learning_rate": 4.998895563881637e-06,
+      "loss": 0.7783,
+      "step": 304
+    },
+    {
+      "epoch": 0.08448753462603878,
+      "grad_norm": 0.4903964400291443,
+      "learning_rate": 4.9988847103449376e-06,
+      "loss": 0.7503,
+      "step": 305
+    },
+    {
+      "epoch": 0.08476454293628809,
+      "grad_norm": 0.5018570423126221,
+      "learning_rate": 4.998873803750534e-06,
+      "loss": 0.7641,
+      "step": 306
+    },
+    {
+      "epoch": 0.0850415512465374,
+      "grad_norm": 0.5024906396865845,
+      "learning_rate": 4.998862844098659e-06,
+      "loss": 0.8005,
+      "step": 307
+    },
+    {
+      "epoch": 0.0853185595567867,
+      "grad_norm": 0.5039277076721191,
+      "learning_rate": 4.998851831389544e-06,
+      "loss": 0.7907,
+      "step": 308
+    },
+    {
+      "epoch": 0.08559556786703601,
+      "grad_norm": 0.5008938312530518,
+      "learning_rate": 4.998840765623424e-06,
+      "loss": 0.774,
+      "step": 309
+    },
+    {
+      "epoch": 0.08587257617728532,
+      "grad_norm": 0.4881752133369446,
+      "learning_rate": 4.998829646800533e-06,
+      "loss": 0.6848,
+      "step": 310
+    },
+    {
+      "epoch": 0.08614958448753463,
+      "grad_norm": 0.5107519030570984,
+      "learning_rate": 4.998818474921107e-06,
+      "loss": 0.7868,
+      "step": 311
+    },
+    {
+      "epoch": 0.08642659279778393,
+      "grad_norm": 0.5051405429840088,
+      "learning_rate": 4.998807249985382e-06,
+      "loss": 0.7591,
+      "step": 312
+    },
+    {
+      "epoch": 0.08670360110803324,
+      "grad_norm": 0.5206391215324402,
+      "learning_rate": 4.9987959719935995e-06,
+      "loss": 0.7905,
+      "step": 313
+    },
+    {
+      "epoch": 0.08698060941828255,
+      "grad_norm": 0.5223249793052673,
+      "learning_rate": 4.998784640945997e-06,
+      "loss": 0.7769,
+      "step": 314
+    },
+    {
+      "epoch": 0.08725761772853186,
+      "grad_norm": 0.5234482884407043,
+      "learning_rate": 4.998773256842816e-06,
+      "loss": 0.7348,
+      "step": 315
+    },
+    {
+      "epoch": 0.08753462603878116,
+      "grad_norm": 0.5284384489059448,
+      "learning_rate": 4.998761819684297e-06,
+      "loss": 0.7819,
+      "step": 316
+    },
+    {
+      "epoch": 0.08781163434903047,
+      "grad_norm": 0.5017664432525635,
+      "learning_rate": 4.998750329470683e-06,
+      "loss": 0.7495,
+      "step": 317
+    },
+    {
+      "epoch": 0.08808864265927978,
+      "grad_norm": 0.5360183119773865,
+      "learning_rate": 4.998738786202219e-06,
+      "loss": 0.8078,
+      "step": 318
+    },
+    {
+      "epoch": 0.08836565096952909,
+      "grad_norm": 0.4968814253807068,
+      "learning_rate": 4.9987271898791486e-06,
+      "loss": 0.7568,
+      "step": 319
+    },
+    {
+      "epoch": 0.0886426592797784,
+      "grad_norm": 0.5032696723937988,
+      "learning_rate": 4.99871554050172e-06,
+      "loss": 0.7403,
+      "step": 320
+    },
+    {
+      "epoch": 0.0889196675900277,
+      "grad_norm": 0.5070774555206299,
+      "learning_rate": 4.998703838070179e-06,
+      "loss": 0.7752,
+      "step": 321
+    },
+    {
+      "epoch": 0.08919667590027701,
+      "grad_norm": 0.5313223600387573,
+      "learning_rate": 4.9986920825847746e-06,
+      "loss": 0.7908,
+      "step": 322
+    },
+    {
+      "epoch": 0.08947368421052632,
+      "grad_norm": 0.5166332125663757,
+      "learning_rate": 4.998680274045756e-06,
+      "loss": 0.7316,
+      "step": 323
+    },
+    {
+      "epoch": 0.08975069252077562,
+      "grad_norm": 0.5434812307357788,
+      "learning_rate": 4.998668412453375e-06,
+      "loss": 0.7771,
+      "step": 324
+    },
+    {
+      "epoch": 0.09002770083102493,
+      "grad_norm": 0.5094904899597168,
+      "learning_rate": 4.998656497807882e-06,
+      "loss": 0.7731,
+      "step": 325
+    },
+    {
+      "epoch": 0.09030470914127424,
+      "grad_norm": 0.5478783249855042,
+      "learning_rate": 4.998644530109531e-06,
+      "loss": 0.8333,
+      "step": 326
+    },
+    {
+      "epoch": 0.09058171745152355,
+      "grad_norm": 0.5198971629142761,
+      "learning_rate": 4.998632509358575e-06,
+      "loss": 0.7141,
+      "step": 327
+    },
+    {
+      "epoch": 0.09085872576177285,
+      "grad_norm": 0.5152432322502136,
+      "learning_rate": 4.998620435555271e-06,
+      "loss": 0.8345,
+      "step": 328
+    },
+    {
+      "epoch": 0.09113573407202216,
+      "grad_norm": 0.4936472475528717,
+      "learning_rate": 4.9986083086998745e-06,
+      "loss": 0.7765,
+      "step": 329
+    },
+    {
+      "epoch": 0.09141274238227147,
+      "grad_norm": 0.5072058439254761,
+      "learning_rate": 4.998596128792642e-06,
+      "loss": 0.7071,
+      "step": 330
+    },
+    {
+      "epoch": 0.09168975069252078,
+      "grad_norm": 0.5457636713981628,
+      "learning_rate": 4.998583895833834e-06,
+      "loss": 0.7662,
+      "step": 331
+    },
+    {
+      "epoch": 0.09196675900277008,
+      "grad_norm": 0.51166170835495,
+      "learning_rate": 4.998571609823708e-06,
+      "loss": 0.8131,
+      "step": 332
+    },
+    {
+      "epoch": 0.09224376731301939,
+      "grad_norm": 0.5171675682067871,
+      "learning_rate": 4.998559270762527e-06,
+      "loss": 0.7898,
+      "step": 333
+    },
+    {
+      "epoch": 0.0925207756232687,
+      "grad_norm": 0.5349574685096741,
+      "learning_rate": 4.9985468786505515e-06,
+      "loss": 0.7768,
+      "step": 334
+    },
+    {
+      "epoch": 0.09279778393351801,
+      "grad_norm": 0.5129192471504211,
+      "learning_rate": 4.9985344334880455e-06,
+      "loss": 0.7694,
+      "step": 335
+    },
+    {
+      "epoch": 0.09307479224376732,
+      "grad_norm": 0.5073570609092712,
+      "learning_rate": 4.998521935275272e-06,
+      "loss": 0.7798,
+      "step": 336
+    },
+    {
+      "epoch": 0.09335180055401662,
+      "grad_norm": 0.5064870119094849,
+      "learning_rate": 4.998509384012499e-06,
+      "loss": 0.7825,
+      "step": 337
+    },
+    {
+      "epoch": 0.09362880886426593,
+      "grad_norm": 0.5043659806251526,
+      "learning_rate": 4.99849677969999e-06,
+      "loss": 0.7532,
+      "step": 338
+    },
+    {
+      "epoch": 0.09390581717451524,
+      "grad_norm": 0.5378217101097107,
+      "learning_rate": 4.998484122338014e-06,
+      "loss": 0.7881,
+      "step": 339
+    },
+    {
+      "epoch": 0.09418282548476455,
+      "grad_norm": 0.48902422189712524,
+      "learning_rate": 4.998471411926841e-06,
+      "loss": 0.7362,
+      "step": 340
+    },
+    {
+      "epoch": 0.09445983379501385,
+      "grad_norm": 0.49571239948272705,
+      "learning_rate": 4.998458648466738e-06,
+      "loss": 0.7496,
+      "step": 341
+    },
+    {
+      "epoch": 0.09473684210526316,
+      "grad_norm": 0.47406813502311707,
+      "learning_rate": 4.9984458319579775e-06,
+      "loss": 0.7248,
+      "step": 342
+    },
+    {
+      "epoch": 0.09501385041551247,
+      "grad_norm": 0.5102288722991943,
+      "learning_rate": 4.998432962400832e-06,
+      "loss": 0.773,
+      "step": 343
+    },
+    {
+      "epoch": 0.09529085872576178,
+      "grad_norm": 0.5285248160362244,
+      "learning_rate": 4.9984200397955755e-06,
+      "loss": 0.7537,
+      "step": 344
+    },
+    {
+      "epoch": 0.09556786703601108,
+      "grad_norm": 0.5389485359191895,
+      "learning_rate": 4.99840706414248e-06,
+      "loss": 0.7941,
+      "step": 345
+    },
+    {
+      "epoch": 0.09584487534626039,
+      "grad_norm": 0.49833494424819946,
+      "learning_rate": 4.9983940354418225e-06,
+      "loss": 0.782,
+      "step": 346
+    },
+    {
+      "epoch": 0.0961218836565097,
+      "grad_norm": 0.5146127939224243,
+      "learning_rate": 4.998380953693879e-06,
+      "loss": 0.7467,
+      "step": 347
+    },
+    {
+      "epoch": 0.096398891966759,
+      "grad_norm": 0.5298901200294495,
+      "learning_rate": 4.998367818898928e-06,
+      "loss": 0.7928,
+      "step": 348
+    },
+    {
+      "epoch": 0.09667590027700831,
+      "grad_norm": 0.49955227971076965,
+      "learning_rate": 4.998354631057248e-06,
+      "loss": 0.7364,
+      "step": 349
+    },
+    {
+      "epoch": 0.09695290858725762,
+      "grad_norm": 0.510553777217865,
+      "learning_rate": 4.99834139016912e-06,
+      "loss": 0.7328,
+      "step": 350
+    },
+    {
+      "epoch": 0.09722991689750693,
+      "grad_norm": 0.530810534954071,
+      "learning_rate": 4.998328096234823e-06,
+      "loss": 0.7356,
+      "step": 351
+    },
+    {
+      "epoch": 0.09750692520775624,
+      "grad_norm": 0.5010079741477966,
+      "learning_rate": 4.998314749254641e-06,
+      "loss": 0.7729,
+      "step": 352
+    },
+    {
+      "epoch": 0.09778393351800554,
+      "grad_norm": 0.5428636074066162,
+      "learning_rate": 4.998301349228858e-06,
+      "loss": 0.7817,
+      "step": 353
+    },
+    {
+      "epoch": 0.09806094182825485,
+      "grad_norm": 0.530292809009552,
+      "learning_rate": 4.998287896157757e-06,
+      "loss": 0.7581,
+      "step": 354
+    },
+    {
+      "epoch": 0.09833795013850416,
+      "grad_norm": 0.4924919903278351,
+      "learning_rate": 4.998274390041622e-06,
+      "loss": 0.749,
+      "step": 355
+    },
+    {
+      "epoch": 0.09861495844875347,
+      "grad_norm": 0.5401570796966553,
+      "learning_rate": 4.998260830880744e-06,
+      "loss": 0.8132,
+      "step": 356
+    },
+    {
+      "epoch": 0.09889196675900278,
+      "grad_norm": 0.5332610011100769,
+      "learning_rate": 4.998247218675407e-06,
+      "loss": 0.7535,
+      "step": 357
+    },
+    {
+      "epoch": 0.09916897506925208,
+      "grad_norm": 0.5515699982643127,
+      "learning_rate": 4.998233553425903e-06,
+      "loss": 0.7602,
+      "step": 358
+    },
+    {
+      "epoch": 0.09944598337950139,
+      "grad_norm": 0.5187147259712219,
+      "learning_rate": 4.998219835132521e-06,
+      "loss": 0.8135,
+      "step": 359
+    },
+    {
+      "epoch": 0.0997229916897507,
+      "grad_norm": 0.537226676940918,
+      "learning_rate": 4.998206063795551e-06,
+      "loss": 0.7067,
+      "step": 360
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5453082919120789,
+      "learning_rate": 4.998192239415287e-06,
+      "loss": 0.7277,
+      "step": 361
+    },
+    {
+      "epoch": 0.10027700831024931,
+      "grad_norm": 0.5305824875831604,
+      "learning_rate": 4.998178361992022e-06,
+      "loss": 0.7555,
+      "step": 362
+    },
+    {
+      "epoch": 0.10055401662049862,
+      "grad_norm": 0.5143697261810303,
+      "learning_rate": 4.998164431526052e-06,
+      "loss": 0.7336,
+      "step": 363
+    },
+    {
+      "epoch": 0.10083102493074793,
+      "grad_norm": 0.48405060172080994,
+      "learning_rate": 4.99815044801767e-06,
+      "loss": 0.73,
+      "step": 364
+    },
+    {
+      "epoch": 0.10110803324099724,
+      "grad_norm": 0.537443995475769,
+      "learning_rate": 4.998136411467174e-06,
+      "loss": 0.7398,
+      "step": 365
+    },
+    {
+      "epoch": 0.10138504155124654,
+      "grad_norm": 0.563195526599884,
+      "learning_rate": 4.9981223218748635e-06,
+      "loss": 0.7933,
+      "step": 366
+    },
+    {
+      "epoch": 0.10166204986149585,
+      "grad_norm": 0.532027006149292,
+      "learning_rate": 4.9981081792410365e-06,
+      "loss": 0.7176,
+      "step": 367
+    },
+    {
+      "epoch": 0.10193905817174516,
+      "grad_norm": 0.533951997756958,
+      "learning_rate": 4.998093983565994e-06,
+      "loss": 0.7919,
+      "step": 368
+    },
+    {
+      "epoch": 0.10221606648199447,
+      "grad_norm": 0.5181298851966858,
+      "learning_rate": 4.998079734850036e-06,
+      "loss": 0.7518,
+      "step": 369
+    },
+    {
+      "epoch": 0.10249307479224377,
+      "grad_norm": 0.5276827812194824,
+      "learning_rate": 4.998065433093465e-06,
+      "loss": 0.7567,
+      "step": 370
+    },
+    {
+      "epoch": 0.10277008310249308,
+      "grad_norm": 0.5201783776283264,
+      "learning_rate": 4.998051078296586e-06,
+      "loss": 0.7873,
+      "step": 371
+    },
+    {
+      "epoch": 0.10304709141274238,
+      "grad_norm": 0.5155224204063416,
+      "learning_rate": 4.998036670459704e-06,
+      "loss": 0.7725,
+      "step": 372
+    },
+    {
+      "epoch": 0.10332409972299168,
+      "grad_norm": 0.5198693871498108,
+      "learning_rate": 4.998022209583124e-06,
+      "loss": 0.7613,
+      "step": 373
+    },
+    {
+      "epoch": 0.10360110803324099,
+      "grad_norm": 0.5042403936386108,
+      "learning_rate": 4.998007695667153e-06,
+      "loss": 0.6974,
+      "step": 374
+    },
+    {
+      "epoch": 0.1038781163434903,
+      "grad_norm": 0.5788574814796448,
+      "learning_rate": 4.9979931287121e-06,
+      "loss": 0.7692,
+      "step": 375
+    },
+    {
+      "epoch": 0.1041551246537396,
+      "grad_norm": 0.51310133934021,
+      "learning_rate": 4.997978508718273e-06,
+      "loss": 0.7307,
+      "step": 376
+    },
+    {
+      "epoch": 0.10443213296398891,
+      "grad_norm": 0.5450568199157715,
+      "learning_rate": 4.997963835685983e-06,
+      "loss": 0.7569,
+      "step": 377
+    },
+    {
+      "epoch": 0.10470914127423822,
+      "grad_norm": 0.49824953079223633,
+      "learning_rate": 4.997949109615542e-06,
+      "loss": 0.7723,
+      "step": 378
+    },
+    {
+      "epoch": 0.10498614958448753,
+      "grad_norm": 0.5454652905464172,
+      "learning_rate": 4.9979343305072625e-06,
+      "loss": 0.756,
+      "step": 379
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.5060559511184692,
+      "learning_rate": 4.997919498361458e-06,
+      "loss": 0.7974,
+      "step": 380
+    },
+    {
+      "epoch": 0.10554016620498614,
+      "grad_norm": 0.5230852365493774,
+      "learning_rate": 4.997904613178443e-06,
+      "loss": 0.776,
+      "step": 381
+    },
+    {
+      "epoch": 0.10581717451523545,
+      "grad_norm": 0.5657520890235901,
+      "learning_rate": 4.997889674958535e-06,
+      "loss": 0.7845,
+      "step": 382
+    },
+    {
+      "epoch": 0.10609418282548476,
+      "grad_norm": 0.5440294146537781,
+      "learning_rate": 4.9978746837020495e-06,
+      "loss": 0.7658,
+      "step": 383
+    },
+    {
+      "epoch": 0.10637119113573407,
+      "grad_norm": 0.5455194115638733,
+      "learning_rate": 4.997859639409306e-06,
+      "loss": 0.7923,
+      "step": 384
+    },
+    {
+      "epoch": 0.10664819944598337,
+      "grad_norm": 0.5205777287483215,
+      "learning_rate": 4.997844542080623e-06,
+      "loss": 0.7595,
+      "step": 385
+    },
+    {
+      "epoch": 0.10692520775623268,
+      "grad_norm": 0.5009412169456482,
+      "learning_rate": 4.9978293917163225e-06,
+      "loss": 0.7421,
+      "step": 386
+    },
+    {
+      "epoch": 0.10720221606648199,
+      "grad_norm": 0.5138939619064331,
+      "learning_rate": 4.997814188316725e-06,
+      "loss": 0.7442,
+      "step": 387
+    },
+    {
+      "epoch": 0.1074792243767313,
+      "grad_norm": 0.5283079147338867,
+      "learning_rate": 4.997798931882153e-06,
+      "loss": 0.7864,
+      "step": 388
+    },
+    {
+      "epoch": 0.1077562326869806,
+      "grad_norm": 0.522305965423584,
+      "learning_rate": 4.997783622412932e-06,
+      "loss": 0.6957,
+      "step": 389
+    },
+    {
+      "epoch": 0.10803324099722991,
+      "grad_norm": 0.5620875954627991,
+      "learning_rate": 4.997768259909385e-06,
+      "loss": 0.7834,
+      "step": 390
+    },
+    {
+      "epoch": 0.10831024930747922,
+      "grad_norm": 0.5754278898239136,
+      "learning_rate": 4.997752844371839e-06,
+      "loss": 0.7981,
+      "step": 391
+    },
+    {
+      "epoch": 0.10858725761772853,
+      "grad_norm": 0.5440781712532043,
+      "learning_rate": 4.997737375800623e-06,
+      "loss": 0.7658,
+      "step": 392
+    },
+    {
+      "epoch": 0.10886426592797784,
+      "grad_norm": 0.5162774920463562,
+      "learning_rate": 4.997721854196063e-06,
+      "loss": 0.7367,
+      "step": 393
+    },
+    {
+      "epoch": 0.10914127423822714,
+      "grad_norm": 0.5624562501907349,
+      "learning_rate": 4.9977062795584895e-06,
+      "loss": 0.7703,
+      "step": 394
+    },
+    {
+      "epoch": 0.10941828254847645,
+      "grad_norm": 0.5338923931121826,
+      "learning_rate": 4.997690651888234e-06,
+      "loss": 0.8014,
+      "step": 395
+    },
+    {
+      "epoch": 0.10969529085872576,
+      "grad_norm": 0.5490716695785522,
+      "learning_rate": 4.997674971185627e-06,
+      "loss": 0.8036,
+      "step": 396
+    },
+    {
+      "epoch": 0.10997229916897507,
+      "grad_norm": 0.4992327094078064,
+      "learning_rate": 4.997659237451002e-06,
+      "loss": 0.7433,
+      "step": 397
+    },
+    {
+      "epoch": 0.11024930747922437,
+      "grad_norm": 0.5591122508049011,
+      "learning_rate": 4.997643450684692e-06,
+      "loss": 0.756,
+      "step": 398
+    },
+    {
+      "epoch": 0.11052631578947368,
+      "grad_norm": 0.5397447347640991,
+      "learning_rate": 4.997627610887035e-06,
+      "loss": 0.7525,
+      "step": 399
+    },
+    {
+      "epoch": 0.11080332409972299,
+      "grad_norm": 0.5162704586982727,
+      "learning_rate": 4.997611718058365e-06,
+      "loss": 0.7558,
+      "step": 400
+    },
+    {
+      "epoch": 0.1110803324099723,
+      "grad_norm": 0.5250886678695679,
+      "learning_rate": 4.997595772199019e-06,
+      "loss": 0.7266,
+      "step": 401
+    },
+    {
+      "epoch": 0.1113573407202216,
+      "grad_norm": 0.5203819274902344,
+      "learning_rate": 4.9975797733093386e-06,
+      "loss": 0.6951,
+      "step": 402
+    },
+    {
+      "epoch": 0.11163434903047091,
+      "grad_norm": 0.5280457139015198,
+      "learning_rate": 4.997563721389661e-06,
+      "loss": 0.701,
+      "step": 403
+    },
+    {
+      "epoch": 0.11191135734072022,
+      "grad_norm": 0.5666073560714722,
+      "learning_rate": 4.997547616440327e-06,
+      "loss": 0.7444,
+      "step": 404
+    },
+    {
+      "epoch": 0.11218836565096953,
+      "grad_norm": 0.5710530877113342,
+      "learning_rate": 4.997531458461678e-06,
+      "loss": 0.8137,
+      "step": 405
+    },
+    {
+      "epoch": 0.11246537396121883,
+      "grad_norm": 0.5098825097084045,
+      "learning_rate": 4.99751524745406e-06,
+      "loss": 0.7014,
+      "step": 406
+    },
+    {
+      "epoch": 0.11274238227146814,
+      "grad_norm": 0.5209449529647827,
+      "learning_rate": 4.997498983417814e-06,
+      "loss": 0.768,
+      "step": 407
+    },
+    {
+      "epoch": 0.11301939058171745,
+      "grad_norm": 0.5273720622062683,
+      "learning_rate": 4.997482666353287e-06,
+      "loss": 0.7811,
+      "step": 408
+    },
+    {
+      "epoch": 0.11329639889196676,
+      "grad_norm": 0.5411760807037354,
+      "learning_rate": 4.997466296260826e-06,
+      "loss": 0.7593,
+      "step": 409
+    },
+    {
+      "epoch": 0.11357340720221606,
+      "grad_norm": 0.5214300155639648,
+      "learning_rate": 4.997449873140776e-06,
+      "loss": 0.7749,
+      "step": 410
+    },
+    {
+      "epoch": 0.11385041551246537,
+      "grad_norm": 0.5203116536140442,
+      "learning_rate": 4.9974333969934886e-06,
+      "loss": 0.7535,
+      "step": 411
+    },
+    {
+      "epoch": 0.11412742382271468,
+      "grad_norm": 0.5743964910507202,
+      "learning_rate": 4.997416867819312e-06,
+      "loss": 0.7537,
+      "step": 412
+    },
+    {
+      "epoch": 0.11440443213296399,
+      "grad_norm": 0.5519621968269348,
+      "learning_rate": 4.997400285618597e-06,
+      "loss": 0.8026,
+      "step": 413
+    },
+    {
+      "epoch": 0.1146814404432133,
+      "grad_norm": 0.5224972367286682,
+      "learning_rate": 4.997383650391698e-06,
+      "loss": 0.7503,
+      "step": 414
+    },
+    {
+      "epoch": 0.1149584487534626,
+      "grad_norm": 0.561909556388855,
+      "learning_rate": 4.997366962138965e-06,
+      "loss": 0.7419,
+      "step": 415
+    },
+    {
+      "epoch": 0.11523545706371191,
+      "grad_norm": 0.5283569097518921,
+      "learning_rate": 4.997350220860755e-06,
+      "loss": 0.6599,
+      "step": 416
+    },
+    {
+      "epoch": 0.11551246537396122,
+      "grad_norm": 0.5303184390068054,
+      "learning_rate": 4.997333426557422e-06,
+      "loss": 0.7193,
+      "step": 417
+    },
+    {
+      "epoch": 0.11578947368421053,
+      "grad_norm": 0.5405933856964111,
+      "learning_rate": 4.997316579229322e-06,
+      "loss": 0.7526,
+      "step": 418
+    },
+    {
+      "epoch": 0.11606648199445983,
+      "grad_norm": 0.5390035510063171,
+      "learning_rate": 4.997299678876813e-06,
+      "loss": 0.7556,
+      "step": 419
+    },
+    {
+      "epoch": 0.11634349030470914,
+      "grad_norm": 0.5486871600151062,
+      "learning_rate": 4.997282725500256e-06,
+      "loss": 0.7899,
+      "step": 420
+    },
+    {
+      "epoch": 0.11662049861495845,
+      "grad_norm": 0.5667344927787781,
+      "learning_rate": 4.997265719100009e-06,
+      "loss": 0.7683,
+      "step": 421
+    },
+    {
+      "epoch": 0.11689750692520776,
+      "grad_norm": 0.5894953608512878,
+      "learning_rate": 4.9972486596764335e-06,
+      "loss": 0.8083,
+      "step": 422
+    },
+    {
+      "epoch": 0.11717451523545706,
+      "grad_norm": 0.5495768785476685,
+      "learning_rate": 4.997231547229892e-06,
+      "loss": 0.7519,
+      "step": 423
+    },
+    {
+      "epoch": 0.11745152354570637,
+      "grad_norm": 0.5191949009895325,
+      "learning_rate": 4.997214381760746e-06,
+      "loss": 0.7611,
+      "step": 424
+    },
+    {
+      "epoch": 0.11772853185595568,
+      "grad_norm": 0.5151809453964233,
+      "learning_rate": 4.997197163269363e-06,
+      "loss": 0.7357,
+      "step": 425
+    },
+    {
+      "epoch": 0.11800554016620499,
+      "grad_norm": 0.5261116623878479,
+      "learning_rate": 4.997179891756107e-06,
+      "loss": 0.701,
+      "step": 426
+    },
+    {
+      "epoch": 0.1182825484764543,
+      "grad_norm": 0.5604850649833679,
+      "learning_rate": 4.997162567221344e-06,
+      "loss": 0.7442,
+      "step": 427
+    },
+    {
+      "epoch": 0.1185595567867036,
+      "grad_norm": 0.553929328918457,
+      "learning_rate": 4.997145189665443e-06,
+      "loss": 0.7543,
+      "step": 428
+    },
+    {
+      "epoch": 0.11883656509695291,
+      "grad_norm": 0.547197163105011,
+      "learning_rate": 4.997127759088772e-06,
+      "loss": 0.6987,
+      "step": 429
+    },
+    {
+      "epoch": 0.11911357340720222,
+      "grad_norm": 0.546122133731842,
+      "learning_rate": 4.997110275491702e-06,
+      "loss": 0.7578,
+      "step": 430
+    },
+    {
+      "epoch": 0.11939058171745152,
+      "grad_norm": 0.5426105260848999,
+      "learning_rate": 4.997092738874605e-06,
+      "loss": 0.7829,
+      "step": 431
+    },
+    {
+      "epoch": 0.11966759002770083,
+      "grad_norm": 0.5304821133613586,
+      "learning_rate": 4.9970751492378514e-06,
+      "loss": 0.7703,
+      "step": 432
+    },
+    {
+      "epoch": 0.11994459833795014,
+      "grad_norm": 0.5384340286254883,
+      "learning_rate": 4.997057506581816e-06,
+      "loss": 0.7285,
+      "step": 433
+    },
+    {
+      "epoch": 0.12022160664819945,
+      "grad_norm": 0.5669154524803162,
+      "learning_rate": 4.9970398109068716e-06,
+      "loss": 0.7767,
+      "step": 434
+    },
+    {
+      "epoch": 0.12049861495844875,
+      "grad_norm": 0.5595332384109497,
+      "learning_rate": 4.997022062213397e-06,
+      "loss": 0.7556,
+      "step": 435
+    },
+    {
+      "epoch": 0.12077562326869806,
+      "grad_norm": 0.5652584433555603,
+      "learning_rate": 4.997004260501767e-06,
+      "loss": 0.7491,
+      "step": 436
+    },
+    {
+      "epoch": 0.12105263157894737,
+      "grad_norm": 0.5273657441139221,
+      "learning_rate": 4.996986405772359e-06,
+      "loss": 0.7251,
+      "step": 437
+    },
+    {
+      "epoch": 0.12132963988919668,
+      "grad_norm": 0.536659836769104,
+      "learning_rate": 4.996968498025553e-06,
+      "loss": 0.712,
+      "step": 438
+    },
+    {
+      "epoch": 0.12160664819944599,
+      "grad_norm": 0.5609496235847473,
+      "learning_rate": 4.99695053726173e-06,
+      "loss": 0.7873,
+      "step": 439
+    },
+    {
+      "epoch": 0.12188365650969529,
+      "grad_norm": 0.5493842363357544,
+      "learning_rate": 4.9969325234812705e-06,
+      "loss": 0.7245,
+      "step": 440
+    },
+    {
+      "epoch": 0.1221606648199446,
+      "grad_norm": 0.5588358640670776,
+      "learning_rate": 4.996914456684556e-06,
+      "loss": 0.7185,
+      "step": 441
+    },
+    {
+      "epoch": 0.12243767313019391,
+      "grad_norm": 0.5604621767997742,
+      "learning_rate": 4.996896336871973e-06,
+      "loss": 0.7619,
+      "step": 442
+    },
+    {
+      "epoch": 0.12271468144044322,
+      "grad_norm": 0.5636209845542908,
+      "learning_rate": 4.996878164043903e-06,
+      "loss": 0.7588,
+      "step": 443
+    },
+    {
+      "epoch": 0.12299168975069252,
+      "grad_norm": 0.5372119545936584,
+      "learning_rate": 4.996859938200733e-06,
+      "loss": 0.7461,
+      "step": 444
+    },
+    {
+      "epoch": 0.12326869806094183,
+      "grad_norm": 0.5723452568054199,
+      "learning_rate": 4.996841659342852e-06,
+      "loss": 0.7263,
+      "step": 445
+    },
+    {
+      "epoch": 0.12354570637119114,
+      "grad_norm": 0.5590176582336426,
+      "learning_rate": 4.996823327470644e-06,
+      "loss": 0.7499,
+      "step": 446
+    },
+    {
+      "epoch": 0.12382271468144045,
+      "grad_norm": 0.5287463665008545,
+      "learning_rate": 4.996804942584502e-06,
+      "loss": 0.7078,
+      "step": 447
+    },
+    {
+      "epoch": 0.12409972299168975,
+      "grad_norm": 0.5544666647911072,
+      "learning_rate": 4.9967865046848155e-06,
+      "loss": 0.7296,
+      "step": 448
+    },
+    {
+      "epoch": 0.12437673130193906,
+      "grad_norm": 0.6168888211250305,
+      "learning_rate": 4.996768013771974e-06,
+      "loss": 0.7747,
+      "step": 449
+    },
+    {
+      "epoch": 0.12465373961218837,
+      "grad_norm": 0.550980269908905,
+      "learning_rate": 4.996749469846372e-06,
+      "loss": 0.7419,
+      "step": 450
+    },
+    {
+      "epoch": 0.12493074792243768,
+      "grad_norm": 0.5525485277175903,
+      "learning_rate": 4.996730872908403e-06,
+      "loss": 0.7427,
+      "step": 451
+    },
+    {
+      "epoch": 0.12520775623268698,
+      "grad_norm": 0.5857051014900208,
+      "learning_rate": 4.9967122229584614e-06,
+      "loss": 0.7936,
+      "step": 452
+    },
+    {
+      "epoch": 0.12548476454293628,
+      "grad_norm": 0.5720162391662598,
+      "learning_rate": 4.996693519996943e-06,
+      "loss": 0.8068,
+      "step": 453
+    },
+    {
+      "epoch": 0.1257617728531856,
+      "grad_norm": 0.539708137512207,
+      "learning_rate": 4.996674764024247e-06,
+      "loss": 0.7568,
+      "step": 454
+    },
+    {
+      "epoch": 0.1260387811634349,
+      "grad_norm": 0.5840601921081543,
+      "learning_rate": 4.996655955040768e-06,
+      "loss": 0.7781,
+      "step": 455
+    },
+    {
+      "epoch": 0.12631578947368421,
+      "grad_norm": 0.5401661396026611,
+      "learning_rate": 4.996637093046907e-06,
+      "loss": 0.7179,
+      "step": 456
+    },
+    {
+      "epoch": 0.1265927977839335,
+      "grad_norm": 0.5300602912902832,
+      "learning_rate": 4.996618178043067e-06,
+      "loss": 0.7329,
+      "step": 457
+    },
+    {
+      "epoch": 0.12686980609418283,
+      "grad_norm": 0.5564039349555969,
+      "learning_rate": 4.996599210029645e-06,
+      "loss": 0.7799,
+      "step": 458
+    },
+    {
+      "epoch": 0.12714681440443212,
+      "grad_norm": 0.59776371717453,
+      "learning_rate": 4.996580189007049e-06,
+      "loss": 0.7401,
+      "step": 459
+    },
+    {
+      "epoch": 0.12742382271468145,
+      "grad_norm": 0.5383924245834351,
+      "learning_rate": 4.996561114975678e-06,
+      "loss": 0.754,
+      "step": 460
+    },
+    {
+      "epoch": 0.12770083102493074,
+      "grad_norm": 0.5681918859481812,
+      "learning_rate": 4.996541987935939e-06,
+      "loss": 0.706,
+      "step": 461
+    },
+    {
+      "epoch": 0.12797783933518006,
+      "grad_norm": 0.5702113509178162,
+      "learning_rate": 4.996522807888239e-06,
+      "loss": 0.754,
+      "step": 462
+    },
+    {
+      "epoch": 0.12825484764542935,
+      "grad_norm": 0.5508577227592468,
+      "learning_rate": 4.996503574832984e-06,
+      "loss": 0.7765,
+      "step": 463
+    },
+    {
+      "epoch": 0.12853185595567868,
+      "grad_norm": 0.5382973551750183,
+      "learning_rate": 4.996484288770582e-06,
+      "loss": 0.7074,
+      "step": 464
+    },
+    {
+      "epoch": 0.12880886426592797,
+      "grad_norm": 0.5670721530914307,
+      "learning_rate": 4.996464949701444e-06,
+      "loss": 0.7513,
+      "step": 465
+    },
+    {
+      "epoch": 0.1290858725761773,
+      "grad_norm": 0.5926334261894226,
+      "learning_rate": 4.9964455576259795e-06,
+      "loss": 0.8037,
+      "step": 466
+    },
+    {
+      "epoch": 0.12936288088642658,
+      "grad_norm": 0.5309609174728394,
+      "learning_rate": 4.9964261125446015e-06,
+      "loss": 0.7443,
+      "step": 467
+    },
+    {
+      "epoch": 0.1296398891966759,
+      "grad_norm": 0.5415372848510742,
+      "learning_rate": 4.99640661445772e-06,
+      "loss": 0.7404,
+      "step": 468
+    },
+    {
+      "epoch": 0.1299168975069252,
+      "grad_norm": 0.5469106435775757,
+      "learning_rate": 4.996387063365753e-06,
+      "loss": 0.7584,
+      "step": 469
+    },
+    {
+      "epoch": 0.13019390581717452,
+      "grad_norm": 0.5746037364006042,
+      "learning_rate": 4.9963674592691125e-06,
+      "loss": 0.7732,
+      "step": 470
+    },
+    {
+      "epoch": 0.13047091412742381,
+      "grad_norm": 0.5430501699447632,
+      "learning_rate": 4.996347802168216e-06,
+      "loss": 0.6859,
+      "step": 471
+    },
+    {
+      "epoch": 0.13074792243767314,
+      "grad_norm": 0.5458232164382935,
+      "learning_rate": 4.996328092063481e-06,
+      "loss": 0.7435,
+      "step": 472
+    },
+    {
+      "epoch": 0.13102493074792243,
+      "grad_norm": 0.5637723207473755,
+      "learning_rate": 4.996308328955325e-06,
+      "loss": 0.7916,
+      "step": 473
+    },
+    {
+      "epoch": 0.13130193905817175,
+      "grad_norm": 0.603699803352356,
+      "learning_rate": 4.996288512844169e-06,
+      "loss": 0.7256,
+      "step": 474
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 0.539997935295105,
+      "learning_rate": 4.996268643730433e-06,
+      "loss": 0.7704,
+      "step": 475
+    },
+    {
+      "epoch": 0.13185595567867037,
+      "grad_norm": 0.5679884552955627,
+      "learning_rate": 4.996248721614538e-06,
+      "loss": 0.7806,
+      "step": 476
+    },
+    {
+      "epoch": 0.13213296398891966,
+      "grad_norm": 0.52909255027771,
+      "learning_rate": 4.996228746496909e-06,
+      "loss": 0.7311,
+      "step": 477
+    },
+    {
+      "epoch": 0.13240997229916898,
+      "grad_norm": 0.5588310956954956,
+      "learning_rate": 4.996208718377968e-06,
+      "loss": 0.7754,
+      "step": 478
+    },
+    {
+      "epoch": 0.13268698060941828,
+      "grad_norm": 0.5584151744842529,
+      "learning_rate": 4.9961886372581425e-06,
+      "loss": 0.7034,
+      "step": 479
+    },
+    {
+      "epoch": 0.1329639889196676,
+      "grad_norm": 0.5524438619613647,
+      "learning_rate": 4.9961685031378575e-06,
+      "loss": 0.7404,
+      "step": 480
+    },
+    {
+      "epoch": 0.1332409972299169,
+      "grad_norm": 0.5850812196731567,
+      "learning_rate": 4.99614831601754e-06,
+      "loss": 0.7978,
+      "step": 481
+    },
+    {
+      "epoch": 0.1335180055401662,
+      "grad_norm": 0.5962361693382263,
+      "learning_rate": 4.996128075897619e-06,
+      "loss": 0.7731,
+      "step": 482
+    },
+    {
+      "epoch": 0.1337950138504155,
+      "grad_norm": 0.5720769166946411,
+      "learning_rate": 4.996107782778526e-06,
+      "loss": 0.7761,
+      "step": 483
+    },
+    {
+      "epoch": 0.13407202216066483,
+      "grad_norm": 0.5690904855728149,
+      "learning_rate": 4.996087436660689e-06,
+      "loss": 0.7837,
+      "step": 484
+    },
+    {
+      "epoch": 0.13434903047091412,
+      "grad_norm": 0.5276780128479004,
+      "learning_rate": 4.996067037544542e-06,
+      "loss": 0.7025,
+      "step": 485
+    },
+    {
+      "epoch": 0.13462603878116344,
+      "grad_norm": 0.5334899425506592,
+      "learning_rate": 4.996046585430517e-06,
+      "loss": 0.705,
+      "step": 486
+    },
+    {
+      "epoch": 0.13490304709141274,
+      "grad_norm": 0.5671269297599792,
+      "learning_rate": 4.99602608031905e-06,
+      "loss": 0.712,
+      "step": 487
+    },
+    {
+      "epoch": 0.13518005540166206,
+      "grad_norm": 0.5702267289161682,
+      "learning_rate": 4.9960055222105745e-06,
+      "loss": 0.7627,
+      "step": 488
+    },
+    {
+      "epoch": 0.13545706371191135,
+      "grad_norm": 0.5774746537208557,
+      "learning_rate": 4.995984911105527e-06,
+      "loss": 0.7708,
+      "step": 489
+    },
+    {
+      "epoch": 0.13573407202216067,
+      "grad_norm": 0.626246452331543,
+      "learning_rate": 4.995964247004346e-06,
+      "loss": 0.7916,
+      "step": 490
+    },
+    {
+      "epoch": 0.13601108033240997,
+      "grad_norm": 0.5762872099876404,
+      "learning_rate": 4.99594352990747e-06,
+      "loss": 0.7308,
+      "step": 491
+    },
+    {
+      "epoch": 0.1362880886426593,
+      "grad_norm": 0.5570892691612244,
+      "learning_rate": 4.9959227598153395e-06,
+      "loss": 0.7184,
+      "step": 492
+    },
+    {
+      "epoch": 0.13656509695290858,
+      "grad_norm": 0.5843603610992432,
+      "learning_rate": 4.995901936728394e-06,
+      "loss": 0.7467,
+      "step": 493
+    },
+    {
+      "epoch": 0.1368421052631579,
+      "grad_norm": 0.5812234878540039,
+      "learning_rate": 4.9958810606470775e-06,
+      "loss": 0.7576,
+      "step": 494
+    },
+    {
+      "epoch": 0.1371191135734072,
+      "grad_norm": 0.5652825832366943,
+      "learning_rate": 4.995860131571832e-06,
+      "loss": 0.7052,
+      "step": 495
+    },
+    {
+      "epoch": 0.13739612188365652,
+      "grad_norm": 0.5675123929977417,
+      "learning_rate": 4.995839149503103e-06,
+      "loss": 0.7658,
+      "step": 496
+    },
+    {
+      "epoch": 0.1376731301939058,
+      "grad_norm": 0.565321147441864,
+      "learning_rate": 4.995818114441335e-06,
+      "loss": 0.7434,
+      "step": 497
+    },
+    {
+      "epoch": 0.13795013850415513,
+      "grad_norm": 0.5747477412223816,
+      "learning_rate": 4.9957970263869735e-06,
+      "loss": 0.7441,
+      "step": 498
+    },
+    {
+      "epoch": 0.13822714681440443,
+      "grad_norm": 0.5597822070121765,
+      "learning_rate": 4.9957758853404685e-06,
+      "loss": 0.7484,
+      "step": 499
+    },
+    {
+      "epoch": 0.13850415512465375,
+      "grad_norm": 0.5810073018074036,
+      "learning_rate": 4.995754691302267e-06,
+      "loss": 0.7952,
+      "step": 500
+    },
+    {
+      "epoch": 0.13878116343490304,
+      "grad_norm": 0.5751741528511047,
+      "learning_rate": 4.995733444272821e-06,
+      "loss": 0.8118,
+      "step": 501
+    },
+    {
+      "epoch": 0.13905817174515236,
+      "grad_norm": 0.5670412182807922,
+      "learning_rate": 4.99571214425258e-06,
+      "loss": 0.7356,
+      "step": 502
+    },
+    {
+      "epoch": 0.13933518005540166,
+      "grad_norm": 0.5851835608482361,
+      "learning_rate": 4.995690791241997e-06,
+      "loss": 0.794,
+      "step": 503
+    },
+    {
+      "epoch": 0.13961218836565098,
+      "grad_norm": 0.5708001255989075,
+      "learning_rate": 4.995669385241525e-06,
+      "loss": 0.7258,
+      "step": 504
+    },
+    {
+      "epoch": 0.13988919667590027,
+      "grad_norm": 0.5651164054870605,
+      "learning_rate": 4.995647926251619e-06,
+      "loss": 0.7173,
+      "step": 505
+    },
+    {
+      "epoch": 0.1401662049861496,
+      "grad_norm": 0.580865204334259,
+      "learning_rate": 4.995626414272734e-06,
+      "loss": 0.8015,
+      "step": 506
+    },
+    {
+      "epoch": 0.1404432132963989,
+      "grad_norm": 0.5580934882164001,
+      "learning_rate": 4.995604849305328e-06,
+      "loss": 0.7708,
+      "step": 507
+    },
+    {
+      "epoch": 0.1407202216066482,
+      "grad_norm": 0.5962908267974854,
+      "learning_rate": 4.995583231349857e-06,
+      "loss": 0.7617,
+      "step": 508
+    },
+    {
+      "epoch": 0.1409972299168975,
+      "grad_norm": 0.5531542897224426,
+      "learning_rate": 4.995561560406781e-06,
+      "loss": 0.7422,
+      "step": 509
+    },
+    {
+      "epoch": 0.14127423822714683,
+      "grad_norm": 0.5768089294433594,
+      "learning_rate": 4.995539836476561e-06,
+      "loss": 0.8083,
+      "step": 510
+    },
+    {
+      "epoch": 0.14155124653739612,
+      "grad_norm": 0.5697327256202698,
+      "learning_rate": 4.995518059559656e-06,
+      "loss": 0.7677,
+      "step": 511
+    },
+    {
+      "epoch": 0.14182825484764544,
+      "grad_norm": 0.5646145939826965,
+      "learning_rate": 4.995496229656531e-06,
+      "loss": 0.7015,
+      "step": 512
+    },
+    {
+      "epoch": 0.14210526315789473,
+      "grad_norm": 0.5464889407157898,
+      "learning_rate": 4.995474346767647e-06,
+      "loss": 0.7104,
+      "step": 513
+    },
+    {
+      "epoch": 0.14238227146814406,
+      "grad_norm": 0.6004087924957275,
+      "learning_rate": 4.995452410893471e-06,
+      "loss": 0.8018,
+      "step": 514
+    },
+    {
+      "epoch": 0.14265927977839335,
+      "grad_norm": 0.6363440155982971,
+      "learning_rate": 4.9954304220344676e-06,
+      "loss": 0.7857,
+      "step": 515
+    },
+    {
+      "epoch": 0.14293628808864267,
+      "grad_norm": 0.5825870633125305,
+      "learning_rate": 4.9954083801911025e-06,
+      "loss": 0.7574,
+      "step": 516
+    },
+    {
+      "epoch": 0.14321329639889196,
+      "grad_norm": 0.5635371804237366,
+      "learning_rate": 4.995386285363846e-06,
+      "loss": 0.7117,
+      "step": 517
+    },
+    {
+      "epoch": 0.1434903047091413,
+      "grad_norm": 0.5889778137207031,
+      "learning_rate": 4.995364137553166e-06,
+      "loss": 0.7625,
+      "step": 518
+    },
+    {
+      "epoch": 0.14376731301939058,
+      "grad_norm": 0.5800310969352722,
+      "learning_rate": 4.995341936759533e-06,
+      "loss": 0.7419,
+      "step": 519
+    },
+    {
+      "epoch": 0.1440443213296399,
+      "grad_norm": 0.5781015157699585,
+      "learning_rate": 4.995319682983417e-06,
+      "loss": 0.738,
+      "step": 520
+    },
+    {
+      "epoch": 0.1443213296398892,
+      "grad_norm": 0.5652474761009216,
+      "learning_rate": 4.995297376225293e-06,
+      "loss": 0.7867,
+      "step": 521
+    },
+    {
+      "epoch": 0.14459833795013852,
+      "grad_norm": 0.5604120492935181,
+      "learning_rate": 4.995275016485633e-06,
+      "loss": 0.7324,
+      "step": 522
+    },
+    {
+      "epoch": 0.1448753462603878,
+      "grad_norm": 0.5692135095596313,
+      "learning_rate": 4.995252603764913e-06,
+      "loss": 0.7839,
+      "step": 523
+    },
+    {
+      "epoch": 0.14515235457063713,
+      "grad_norm": 0.5682529211044312,
+      "learning_rate": 4.995230138063607e-06,
+      "loss": 0.7833,
+      "step": 524
+    },
+    {
+      "epoch": 0.14542936288088643,
+      "grad_norm": 0.6351110339164734,
+      "learning_rate": 4.995207619382194e-06,
+      "loss": 0.7737,
+      "step": 525
+    },
+    {
+      "epoch": 0.14570637119113575,
+      "grad_norm": 0.5999937057495117,
+      "learning_rate": 4.995185047721151e-06,
+      "loss": 0.7681,
+      "step": 526
+    },
+    {
+      "epoch": 0.14598337950138504,
+      "grad_norm": 0.592374861240387,
+      "learning_rate": 4.995162423080957e-06,
+      "loss": 0.7981,
+      "step": 527
+    },
+    {
+      "epoch": 0.14626038781163436,
+      "grad_norm": 0.5485160946846008,
+      "learning_rate": 4.995139745462092e-06,
+      "loss": 0.7414,
+      "step": 528
+    },
+    {
+      "epoch": 0.14653739612188366,
+      "grad_norm": 0.5628599524497986,
+      "learning_rate": 4.99511701486504e-06,
+      "loss": 0.7589,
+      "step": 529
+    },
+    {
+      "epoch": 0.14681440443213298,
+      "grad_norm": 0.5511693954467773,
+      "learning_rate": 4.9950942312902805e-06,
+      "loss": 0.6909,
+      "step": 530
+    },
+    {
+      "epoch": 0.14709141274238227,
+      "grad_norm": 0.5663548111915588,
+      "learning_rate": 4.995071394738299e-06,
+      "loss": 0.8017,
+      "step": 531
+    },
+    {
+      "epoch": 0.14736842105263157,
+      "grad_norm": 0.5832469463348389,
+      "learning_rate": 4.995048505209581e-06,
+      "loss": 0.7032,
+      "step": 532
+    },
+    {
+      "epoch": 0.1476454293628809,
+      "grad_norm": 0.5971035361289978,
+      "learning_rate": 4.995025562704611e-06,
+      "loss": 0.7705,
+      "step": 533
+    },
+    {
+      "epoch": 0.14792243767313018,
+      "grad_norm": 0.5898804068565369,
+      "learning_rate": 4.995002567223877e-06,
+      "loss": 0.7446,
+      "step": 534
+    },
+    {
+      "epoch": 0.1481994459833795,
+      "grad_norm": 0.5439696311950684,
+      "learning_rate": 4.994979518767866e-06,
+      "loss": 0.7333,
+      "step": 535
+    },
+    {
+      "epoch": 0.1484764542936288,
+      "grad_norm": 0.5540503263473511,
+      "learning_rate": 4.994956417337069e-06,
+      "loss": 0.7951,
+      "step": 536
+    },
+    {
+      "epoch": 0.14875346260387812,
+      "grad_norm": 0.5638033747673035,
+      "learning_rate": 4.994933262931976e-06,
+      "loss": 0.7685,
+      "step": 537
+    },
+    {
+      "epoch": 0.1490304709141274,
+      "grad_norm": 0.5664800405502319,
+      "learning_rate": 4.994910055553079e-06,
+      "loss": 0.7569,
+      "step": 538
+    },
+    {
+      "epoch": 0.14930747922437673,
+      "grad_norm": 0.6010585427284241,
+      "learning_rate": 4.994886795200869e-06,
+      "loss": 0.756,
+      "step": 539
+    },
+    {
+      "epoch": 0.14958448753462603,
+      "grad_norm": 0.5838546752929688,
+      "learning_rate": 4.994863481875842e-06,
+      "loss": 0.7701,
+      "step": 540
+    },
+    {
+      "epoch": 0.14986149584487535,
+      "grad_norm": 0.6126722097396851,
+      "learning_rate": 4.994840115578491e-06,
+      "loss": 0.7468,
+      "step": 541
+    },
+    {
+      "epoch": 0.15013850415512464,
+      "grad_norm": 0.5752817392349243,
+      "learning_rate": 4.994816696309314e-06,
+      "loss": 0.7532,
+      "step": 542
+    },
+    {
+      "epoch": 0.15041551246537396,
+      "grad_norm": 0.5906909704208374,
+      "learning_rate": 4.994793224068807e-06,
+      "loss": 0.7719,
+      "step": 543
+    },
+    {
+      "epoch": 0.15069252077562326,
+      "grad_norm": 0.606719434261322,
+      "learning_rate": 4.994769698857469e-06,
+      "loss": 0.7709,
+      "step": 544
+    },
+    {
+      "epoch": 0.15096952908587258,
+      "grad_norm": 0.5538843870162964,
+      "learning_rate": 4.9947461206757995e-06,
+      "loss": 0.7749,
+      "step": 545
+    },
+    {
+      "epoch": 0.15124653739612187,
+      "grad_norm": 0.5870587825775146,
+      "learning_rate": 4.994722489524299e-06,
+      "loss": 0.6922,
+      "step": 546
+    },
+    {
+      "epoch": 0.1515235457063712,
+      "grad_norm": 0.5842876434326172,
+      "learning_rate": 4.994698805403469e-06,
+      "loss": 0.7525,
+      "step": 547
+    },
+    {
+      "epoch": 0.1518005540166205,
+      "grad_norm": 0.5563871264457703,
+      "learning_rate": 4.994675068313813e-06,
+      "loss": 0.7227,
+      "step": 548
+    },
+    {
+      "epoch": 0.1520775623268698,
+      "grad_norm": 0.5730799436569214,
+      "learning_rate": 4.994651278255835e-06,
+      "loss": 0.754,
+      "step": 549
+    },
+    {
+      "epoch": 0.1523545706371191,
+      "grad_norm": 0.6045185327529907,
+      "learning_rate": 4.994627435230039e-06,
+      "loss": 0.7342,
+      "step": 550
+    },
+    {
+      "epoch": 0.15263157894736842,
+      "grad_norm": 0.5501013398170471,
+      "learning_rate": 4.994603539236932e-06,
+      "loss": 0.659,
+      "step": 551
+    },
+    {
+      "epoch": 0.15290858725761772,
+      "grad_norm": 0.5863350033760071,
+      "learning_rate": 4.994579590277021e-06,
+      "loss": 0.7128,
+      "step": 552
+    },
+    {
+      "epoch": 0.15318559556786704,
+      "grad_norm": 0.5850765109062195,
+      "learning_rate": 4.9945555883508155e-06,
+      "loss": 0.7499,
+      "step": 553
+    },
+    {
+      "epoch": 0.15346260387811633,
+      "grad_norm": 0.61758953332901,
+      "learning_rate": 4.994531533458824e-06,
+      "loss": 0.7482,
+      "step": 554
+    },
+    {
+      "epoch": 0.15373961218836565,
+      "grad_norm": 0.5901806354522705,
+      "learning_rate": 4.994507425601559e-06,
+      "loss": 0.7402,
+      "step": 555
+    },
+    {
+      "epoch": 0.15401662049861495,
+      "grad_norm": 0.5781405568122864,
+      "learning_rate": 4.994483264779529e-06,
+      "loss": 0.7703,
+      "step": 556
+    },
+    {
+      "epoch": 0.15429362880886427,
+      "grad_norm": 0.5744783282279968,
+      "learning_rate": 4.994459050993251e-06,
+      "loss": 0.7451,
+      "step": 557
+    },
+    {
+      "epoch": 0.15457063711911356,
+      "grad_norm": 0.5938632488250732,
+      "learning_rate": 4.9944347842432365e-06,
+      "loss": 0.7549,
+      "step": 558
+    },
+    {
+      "epoch": 0.15484764542936288,
+      "grad_norm": 0.6398091316223145,
+      "learning_rate": 4.994410464530001e-06,
+      "loss": 0.757,
+      "step": 559
+    },
+    {
+      "epoch": 0.15512465373961218,
+      "grad_norm": 0.5992589592933655,
+      "learning_rate": 4.994386091854061e-06,
+      "loss": 0.7152,
+      "step": 560
+    },
+    {
+      "epoch": 0.1554016620498615,
+      "grad_norm": 0.6135343313217163,
+      "learning_rate": 4.994361666215935e-06,
+      "loss": 0.6954,
+      "step": 561
+    },
+    {
+      "epoch": 0.1556786703601108,
+      "grad_norm": 0.6189139485359192,
+      "learning_rate": 4.99433718761614e-06,
+      "loss": 0.7436,
+      "step": 562
+    },
+    {
+      "epoch": 0.15595567867036012,
+      "grad_norm": 0.5707741379737854,
+      "learning_rate": 4.994312656055198e-06,
+      "loss": 0.6516,
+      "step": 563
+    },
+    {
+      "epoch": 0.1562326869806094,
+      "grad_norm": 0.5933526754379272,
+      "learning_rate": 4.9942880715336284e-06,
+      "loss": 0.7263,
+      "step": 564
+    },
+    {
+      "epoch": 0.15650969529085873,
+      "grad_norm": 0.598362922668457,
+      "learning_rate": 4.994263434051953e-06,
+      "loss": 0.7285,
+      "step": 565
+    },
+    {
+      "epoch": 0.15678670360110802,
+      "grad_norm": 0.5745391249656677,
+      "learning_rate": 4.994238743610695e-06,
+      "loss": 0.7279,
+      "step": 566
+    },
+    {
+      "epoch": 0.15706371191135735,
+      "grad_norm": 0.5936552286148071,
+      "learning_rate": 4.994214000210379e-06,
+      "loss": 0.7818,
+      "step": 567
+    },
+    {
+      "epoch": 0.15734072022160664,
+      "grad_norm": 0.6095513701438904,
+      "learning_rate": 4.99418920385153e-06,
+      "loss": 0.7492,
+      "step": 568
+    },
+    {
+      "epoch": 0.15761772853185596,
+      "grad_norm": 0.6010913252830505,
+      "learning_rate": 4.994164354534675e-06,
+      "loss": 0.7537,
+      "step": 569
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.6135689616203308,
+      "learning_rate": 4.994139452260341e-06,
+      "loss": 0.772,
+      "step": 570
+    },
+    {
+      "epoch": 0.15817174515235458,
+      "grad_norm": 0.6023951172828674,
+      "learning_rate": 4.994114497029058e-06,
+      "loss": 0.7703,
+      "step": 571
+    },
+    {
+      "epoch": 0.15844875346260387,
+      "grad_norm": 0.5701375603675842,
+      "learning_rate": 4.994089488841354e-06,
+      "loss": 0.7329,
+      "step": 572
+    },
+    {
+      "epoch": 0.1587257617728532,
+      "grad_norm": 0.5752888917922974,
+      "learning_rate": 4.994064427697761e-06,
+      "loss": 0.7459,
+      "step": 573
+    },
+    {
+      "epoch": 0.15900277008310248,
+      "grad_norm": 0.6101413369178772,
+      "learning_rate": 4.994039313598812e-06,
+      "loss": 0.7899,
+      "step": 574
+    },
+    {
+      "epoch": 0.1592797783933518,
+      "grad_norm": 0.5689857602119446,
+      "learning_rate": 4.994014146545039e-06,
+      "loss": 0.7274,
+      "step": 575
+    },
+    {
+      "epoch": 0.1595567867036011,
+      "grad_norm": 0.5898393392562866,
+      "learning_rate": 4.9939889265369755e-06,
+      "loss": 0.7446,
+      "step": 576
+    },
+    {
+      "epoch": 0.15983379501385042,
+      "grad_norm": 0.6104997396469116,
+      "learning_rate": 4.993963653575159e-06,
+      "loss": 0.6932,
+      "step": 577
+    },
+    {
+      "epoch": 0.16011080332409972,
+      "grad_norm": 0.6198656558990479,
+      "learning_rate": 4.993938327660125e-06,
+      "loss": 0.7027,
+      "step": 578
+    },
+    {
+      "epoch": 0.16038781163434904,
+      "grad_norm": 0.5969960689544678,
+      "learning_rate": 4.993912948792412e-06,
+      "loss": 0.7283,
+      "step": 579
+    },
+    {
+      "epoch": 0.16066481994459833,
+      "grad_norm": 0.5853732824325562,
+      "learning_rate": 4.993887516972558e-06,
+      "loss": 0.7232,
+      "step": 580
+    },
+    {
+      "epoch": 0.16094182825484765,
+      "grad_norm": 0.5809240341186523,
+      "learning_rate": 4.9938620322011035e-06,
+      "loss": 0.7308,
+      "step": 581
+    },
+    {
+      "epoch": 0.16121883656509695,
+      "grad_norm": 0.6255086064338684,
+      "learning_rate": 4.993836494478589e-06,
+      "loss": 0.7621,
+      "step": 582
+    },
+    {
+      "epoch": 0.16149584487534627,
+      "grad_norm": 0.5679871439933777,
+      "learning_rate": 4.993810903805557e-06,
+      "loss": 0.715,
+      "step": 583
+    },
+    {
+      "epoch": 0.16177285318559556,
+      "grad_norm": 0.6073582172393799,
+      "learning_rate": 4.993785260182552e-06,
+      "loss": 0.7685,
+      "step": 584
+    },
+    {
+      "epoch": 0.16204986149584488,
+      "grad_norm": 0.6174919009208679,
+      "learning_rate": 4.9937595636101165e-06,
+      "loss": 0.7825,
+      "step": 585
+    },
+    {
+      "epoch": 0.16232686980609418,
+      "grad_norm": 0.5922225117683411,
+      "learning_rate": 4.993733814088798e-06,
+      "loss": 0.6968,
+      "step": 586
+    },
+    {
+      "epoch": 0.1626038781163435,
+      "grad_norm": 0.5925888419151306,
+      "learning_rate": 4.993708011619142e-06,
+      "loss": 0.7288,
+      "step": 587
+    },
+    {
+      "epoch": 0.1628808864265928,
+      "grad_norm": 0.5409933924674988,
+      "learning_rate": 4.9936821562016965e-06,
+      "loss": 0.6759,
+      "step": 588
+    },
+    {
+      "epoch": 0.1631578947368421,
+      "grad_norm": 0.5877025127410889,
+      "learning_rate": 4.99365624783701e-06,
+      "loss": 0.7661,
+      "step": 589
+    },
+    {
+      "epoch": 0.1634349030470914,
+      "grad_norm": 0.5677634477615356,
+      "learning_rate": 4.993630286525634e-06,
+      "loss": 0.7698,
+      "step": 590
+    },
+    {
+      "epoch": 0.16371191135734073,
+      "grad_norm": 0.6122007369995117,
+      "learning_rate": 4.993604272268119e-06,
+      "loss": 0.7112,
+      "step": 591
+    },
+    {
+      "epoch": 0.16398891966759002,
+      "grad_norm": 0.5659086108207703,
+      "learning_rate": 4.993578205065017e-06,
+      "loss": 0.746,
+      "step": 592
+    },
+    {
+      "epoch": 0.16426592797783934,
+      "grad_norm": 0.6072430610656738,
+      "learning_rate": 4.993552084916883e-06,
+      "loss": 0.7886,
+      "step": 593
+    },
+    {
+      "epoch": 0.16454293628808864,
+      "grad_norm": 0.599383533000946,
+      "learning_rate": 4.993525911824268e-06,
+      "loss": 0.711,
+      "step": 594
+    },
+    {
+      "epoch": 0.16481994459833796,
+      "grad_norm": 0.5857592821121216,
+      "learning_rate": 4.993499685787732e-06,
+      "loss": 0.7318,
+      "step": 595
+    },
+    {
+      "epoch": 0.16509695290858725,
+      "grad_norm": 0.5840777158737183,
+      "learning_rate": 4.99347340680783e-06,
+      "loss": 0.7209,
+      "step": 596
+    },
+    {
+      "epoch": 0.16537396121883657,
+      "grad_norm": 0.55238938331604,
+      "learning_rate": 4.9934470748851196e-06,
+      "loss": 0.7329,
+      "step": 597
+    },
+    {
+      "epoch": 0.16565096952908587,
+      "grad_norm": 0.5710905194282532,
+      "learning_rate": 4.99342069002016e-06,
+      "loss": 0.7157,
+      "step": 598
+    },
+    {
+      "epoch": 0.1659279778393352,
+      "grad_norm": 0.6297195553779602,
+      "learning_rate": 4.993394252213512e-06,
+      "loss": 0.7414,
+      "step": 599
+    },
+    {
+      "epoch": 0.16620498614958448,
+      "grad_norm": 0.6218348741531372,
+      "learning_rate": 4.993367761465736e-06,
+      "loss": 0.7627,
+      "step": 600
+    },
+    {
+      "epoch": 0.1664819944598338,
+      "grad_norm": 0.6218968629837036,
+      "learning_rate": 4.993341217777396e-06,
+      "loss": 0.7518,
+      "step": 601
+    },
+    {
+      "epoch": 0.1667590027700831,
+      "grad_norm": 0.6355380415916443,
+      "learning_rate": 4.993314621149054e-06,
+      "loss": 0.7509,
+      "step": 602
+    },
+    {
+      "epoch": 0.16703601108033242,
+      "grad_norm": 0.5781269669532776,
+      "learning_rate": 4.9932879715812745e-06,
+      "loss": 0.7405,
+      "step": 603
+    },
+    {
+      "epoch": 0.1673130193905817,
+      "grad_norm": 0.5942702293395996,
+      "learning_rate": 4.993261269074625e-06,
+      "loss": 0.7434,
+      "step": 604
+    },
+    {
+      "epoch": 0.16759002770083103,
+      "grad_norm": 0.6177949905395508,
+      "learning_rate": 4.9932345136296725e-06,
+      "loss": 0.7514,
+      "step": 605
+    },
+    {
+      "epoch": 0.16786703601108033,
+      "grad_norm": 0.6140938997268677,
+      "learning_rate": 4.993207705246983e-06,
+      "loss": 0.7241,
+      "step": 606
+    },
+    {
+      "epoch": 0.16814404432132965,
+      "grad_norm": 0.5944777727127075,
+      "learning_rate": 4.9931808439271275e-06,
+      "loss": 0.7917,
+      "step": 607
+    },
+    {
+      "epoch": 0.16842105263157894,
+      "grad_norm": 0.610851526260376,
+      "learning_rate": 4.993153929670676e-06,
+      "loss": 0.7748,
+      "step": 608
+    },
+    {
+      "epoch": 0.16869806094182827,
+      "grad_norm": 0.5631710886955261,
+      "learning_rate": 4.9931269624782e-06,
+      "loss": 0.6437,
+      "step": 609
+    },
+    {
+      "epoch": 0.16897506925207756,
+      "grad_norm": 0.5906318426132202,
+      "learning_rate": 4.993099942350271e-06,
+      "loss": 0.7635,
+      "step": 610
+    },
+    {
+      "epoch": 0.16925207756232688,
+      "grad_norm": 0.6264278888702393,
+      "learning_rate": 4.993072869287464e-06,
+      "loss": 0.7851,
+      "step": 611
+    },
+    {
+      "epoch": 0.16952908587257617,
+      "grad_norm": 0.5960770845413208,
+      "learning_rate": 4.993045743290353e-06,
+      "loss": 0.7507,
+      "step": 612
+    },
+    {
+      "epoch": 0.1698060941828255,
+      "grad_norm": 0.5849323272705078,
+      "learning_rate": 4.993018564359514e-06,
+      "loss": 0.7203,
+      "step": 613
+    },
+    {
+      "epoch": 0.1700831024930748,
+      "grad_norm": 0.6004356741905212,
+      "learning_rate": 4.992991332495525e-06,
+      "loss": 0.7446,
+      "step": 614
+    },
+    {
+      "epoch": 0.1703601108033241,
+      "grad_norm": 0.6141982078552246,
+      "learning_rate": 4.992964047698964e-06,
+      "loss": 0.7275,
+      "step": 615
+    },
+    {
+      "epoch": 0.1706371191135734,
+      "grad_norm": 0.589536190032959,
+      "learning_rate": 4.992936709970409e-06,
+      "loss": 0.7654,
+      "step": 616
+    },
+    {
+      "epoch": 0.17091412742382273,
+      "grad_norm": 0.6536529660224915,
+      "learning_rate": 4.992909319310442e-06,
+      "loss": 0.7798,
+      "step": 617
+    },
+    {
+      "epoch": 0.17119113573407202,
+      "grad_norm": 0.5964363217353821,
+      "learning_rate": 4.992881875719644e-06,
+      "loss": 0.7158,
+      "step": 618
+    },
+    {
+      "epoch": 0.17146814404432134,
+      "grad_norm": 0.6039339900016785,
+      "learning_rate": 4.992854379198598e-06,
+      "loss": 0.7219,
+      "step": 619
+    },
+    {
+      "epoch": 0.17174515235457063,
+      "grad_norm": 0.6138789057731628,
+      "learning_rate": 4.9928268297478866e-06,
+      "loss": 0.7485,
+      "step": 620
+    },
+    {
+      "epoch": 0.17202216066481996,
+      "grad_norm": 0.6011034250259399,
+      "learning_rate": 4.992799227368096e-06,
+      "loss": 0.7412,
+      "step": 621
+    },
+    {
+      "epoch": 0.17229916897506925,
+      "grad_norm": 0.6099687814712524,
+      "learning_rate": 4.992771572059812e-06,
+      "loss": 0.7477,
+      "step": 622
+    },
+    {
+      "epoch": 0.17257617728531857,
+      "grad_norm": 0.6124699711799622,
+      "learning_rate": 4.992743863823621e-06,
+      "loss": 0.7584,
+      "step": 623
+    },
+    {
+      "epoch": 0.17285318559556787,
+      "grad_norm": 0.6268585920333862,
+      "learning_rate": 4.992716102660114e-06,
+      "loss": 0.6723,
+      "step": 624
+    },
+    {
+      "epoch": 0.1731301939058172,
+      "grad_norm": 0.567093551158905,
+      "learning_rate": 4.9926882885698765e-06,
+      "loss": 0.7189,
+      "step": 625
+    },
+    {
+      "epoch": 0.17340720221606648,
+      "grad_norm": 0.635819673538208,
+      "learning_rate": 4.992660421553501e-06,
+      "loss": 0.7455,
+      "step": 626
+    },
+    {
+      "epoch": 0.1736842105263158,
+      "grad_norm": 0.654516339302063,
+      "learning_rate": 4.99263250161158e-06,
+      "loss": 0.769,
+      "step": 627
+    },
+    {
+      "epoch": 0.1739612188365651,
+      "grad_norm": 0.5695293545722961,
+      "learning_rate": 4.992604528744705e-06,
+      "loss": 0.7282,
+      "step": 628
+    },
+    {
+      "epoch": 0.17423822714681442,
+      "grad_norm": 0.565676748752594,
+      "learning_rate": 4.992576502953471e-06,
+      "loss": 0.694,
+      "step": 629
+    },
+    {
+      "epoch": 0.1745152354570637,
+      "grad_norm": 0.6509530544281006,
+      "learning_rate": 4.992548424238472e-06,
+      "loss": 0.7377,
+      "step": 630
+    },
+    {
+      "epoch": 0.17479224376731303,
+      "grad_norm": 0.620535135269165,
+      "learning_rate": 4.9925202926003045e-06,
+      "loss": 0.707,
+      "step": 631
+    },
+    {
+      "epoch": 0.17506925207756233,
+      "grad_norm": 0.6193838119506836,
+      "learning_rate": 4.992492108039566e-06,
+      "loss": 0.7609,
+      "step": 632
+    },
+    {
+      "epoch": 0.17534626038781165,
+      "grad_norm": 0.6202815771102905,
+      "learning_rate": 4.992463870556856e-06,
+      "loss": 0.7457,
+      "step": 633
+    },
+    {
+      "epoch": 0.17562326869806094,
+      "grad_norm": 0.5950646996498108,
+      "learning_rate": 4.9924355801527714e-06,
+      "loss": 0.7154,
+      "step": 634
+    },
+    {
+      "epoch": 0.17590027700831026,
+      "grad_norm": 0.6505277156829834,
+      "learning_rate": 4.992407236827916e-06,
+      "loss": 0.7481,
+      "step": 635
+    },
+    {
+      "epoch": 0.17617728531855956,
+      "grad_norm": 0.5843820571899414,
+      "learning_rate": 4.9923788405828884e-06,
+      "loss": 0.7416,
+      "step": 636
+    },
+    {
+      "epoch": 0.17645429362880888,
+      "grad_norm": 0.6132304668426514,
+      "learning_rate": 4.9923503914182945e-06,
+      "loss": 0.7854,
+      "step": 637
+    },
+    {
+      "epoch": 0.17673130193905817,
+      "grad_norm": 0.6458740830421448,
+      "learning_rate": 4.992321889334737e-06,
+      "loss": 0.7823,
+      "step": 638
+    },
+    {
+      "epoch": 0.17700831024930747,
+      "grad_norm": 0.6528242230415344,
+      "learning_rate": 4.992293334332821e-06,
+      "loss": 0.762,
+      "step": 639
+    },
+    {
+      "epoch": 0.1772853185595568,
+      "grad_norm": 0.5972003936767578,
+      "learning_rate": 4.992264726413152e-06,
+      "loss": 0.7252,
+      "step": 640
+    },
+    {
+      "epoch": 0.17756232686980608,
+      "grad_norm": 0.5978172421455383,
+      "learning_rate": 4.992236065576339e-06,
+      "loss": 0.7056,
+      "step": 641
+    },
+    {
+      "epoch": 0.1778393351800554,
+      "grad_norm": 0.5857217907905579,
+      "learning_rate": 4.992207351822989e-06,
+      "loss": 0.7205,
+      "step": 642
+    },
+    {
+      "epoch": 0.1781163434903047,
+      "grad_norm": 0.5811772346496582,
+      "learning_rate": 4.9921785851537136e-06,
+      "loss": 0.7363,
+      "step": 643
+    },
+    {
+      "epoch": 0.17839335180055402,
+      "grad_norm": 0.5959508419036865,
+      "learning_rate": 4.992149765569122e-06,
+      "loss": 0.7724,
+      "step": 644
+    },
+    {
+      "epoch": 0.1786703601108033,
+      "grad_norm": 0.5911051630973816,
+      "learning_rate": 4.992120893069827e-06,
+      "loss": 0.7241,
+      "step": 645
+    },
+    {
+      "epoch": 0.17894736842105263,
+      "grad_norm": 0.6051844954490662,
+      "learning_rate": 4.992091967656441e-06,
+      "loss": 0.7279,
+      "step": 646
+    },
+    {
+      "epoch": 0.17922437673130193,
+      "grad_norm": 0.5811322331428528,
+      "learning_rate": 4.992062989329578e-06,
+      "loss": 0.6759,
+      "step": 647
+    },
+    {
+      "epoch": 0.17950138504155125,
+      "grad_norm": 0.5511055588722229,
+      "learning_rate": 4.9920339580898545e-06,
+      "loss": 0.6911,
+      "step": 648
+    },
+    {
+      "epoch": 0.17977839335180054,
+      "grad_norm": 0.5894351601600647,
+      "learning_rate": 4.992004873937886e-06,
+      "loss": 0.7258,
+      "step": 649
+    },
+    {
+      "epoch": 0.18005540166204986,
+      "grad_norm": 0.6022338271141052,
+      "learning_rate": 4.9919757368742895e-06,
+      "loss": 0.7349,
+      "step": 650
+    },
+    {
+      "epoch": 0.18033240997229916,
+      "grad_norm": 0.630414605140686,
+      "learning_rate": 4.991946546899685e-06,
+      "loss": 0.7336,
+      "step": 651
+    },
+    {
+      "epoch": 0.18060941828254848,
+      "grad_norm": 0.593589186668396,
+      "learning_rate": 4.991917304014691e-06,
+      "loss": 0.7004,
+      "step": 652
+    },
+    {
+      "epoch": 0.18088642659279777,
+      "grad_norm": 0.6065483093261719,
+      "learning_rate": 4.99188800821993e-06,
+      "loss": 0.759,
+      "step": 653
+    },
+    {
+      "epoch": 0.1811634349030471,
+      "grad_norm": 0.6059734225273132,
+      "learning_rate": 4.9918586595160214e-06,
+      "loss": 0.7643,
+      "step": 654
+    },
+    {
+      "epoch": 0.1814404432132964,
+      "grad_norm": 0.5951828360557556,
+      "learning_rate": 4.991829257903591e-06,
+      "loss": 0.7215,
+      "step": 655
+    },
+    {
+      "epoch": 0.1817174515235457,
+      "grad_norm": 0.5957050919532776,
+      "learning_rate": 4.991799803383262e-06,
+      "loss": 0.7198,
+      "step": 656
+    },
+    {
+      "epoch": 0.181994459833795,
+      "grad_norm": 0.6348684430122375,
+      "learning_rate": 4.991770295955659e-06,
+      "loss": 0.7064,
+      "step": 657
+    },
+    {
+      "epoch": 0.18227146814404432,
+      "grad_norm": 0.5806444883346558,
+      "learning_rate": 4.99174073562141e-06,
+      "loss": 0.7122,
+      "step": 658
+    },
+    {
+      "epoch": 0.18254847645429362,
+      "grad_norm": 0.614686131477356,
+      "learning_rate": 4.991711122381142e-06,
+      "loss": 0.7286,
+      "step": 659
+    },
+    {
+      "epoch": 0.18282548476454294,
+      "grad_norm": 0.612411618232727,
+      "learning_rate": 4.991681456235483e-06,
+      "loss": 0.7621,
+      "step": 660
+    },
+    {
+      "epoch": 0.18310249307479223,
+      "grad_norm": 0.5956481099128723,
+      "learning_rate": 4.9916517371850646e-06,
+      "loss": 0.7495,
+      "step": 661
+    },
+    {
+      "epoch": 0.18337950138504155,
+      "grad_norm": 0.5914007425308228,
+      "learning_rate": 4.991621965230516e-06,
+      "loss": 0.7157,
+      "step": 662
+    },
+    {
+      "epoch": 0.18365650969529085,
+      "grad_norm": 0.6067460179328918,
+      "learning_rate": 4.991592140372471e-06,
+      "loss": 0.7401,
+      "step": 663
+    },
+    {
+      "epoch": 0.18393351800554017,
+      "grad_norm": 0.6207283139228821,
+      "learning_rate": 4.991562262611562e-06,
+      "loss": 0.7777,
+      "step": 664
+    },
+    {
+      "epoch": 0.18421052631578946,
+      "grad_norm": 0.6362254619598389,
+      "learning_rate": 4.991532331948423e-06,
+      "loss": 0.756,
+      "step": 665
+    },
+    {
+      "epoch": 0.18448753462603878,
+      "grad_norm": 0.5996198654174805,
+      "learning_rate": 4.99150234838369e-06,
+      "loss": 0.7505,
+      "step": 666
+    },
+    {
+      "epoch": 0.18476454293628808,
+      "grad_norm": 0.6403332352638245,
+      "learning_rate": 4.9914723119179995e-06,
+      "loss": 0.7641,
+      "step": 667
+    },
+    {
+      "epoch": 0.1850415512465374,
+      "grad_norm": 0.5878950357437134,
+      "learning_rate": 4.99144222255199e-06,
+      "loss": 0.7176,
+      "step": 668
+    },
+    {
+      "epoch": 0.1853185595567867,
+      "grad_norm": 0.5872005224227905,
+      "learning_rate": 4.991412080286299e-06,
+      "loss": 0.7196,
+      "step": 669
+    },
+    {
+      "epoch": 0.18559556786703602,
+      "grad_norm": 0.607134222984314,
+      "learning_rate": 4.991381885121567e-06,
+      "loss": 0.6374,
+      "step": 670
+    },
+    {
+      "epoch": 0.1858725761772853,
+      "grad_norm": 0.6430603861808777,
+      "learning_rate": 4.991351637058435e-06,
+      "loss": 0.7413,
+      "step": 671
+    },
+    {
+      "epoch": 0.18614958448753463,
+      "grad_norm": 0.6369723081588745,
+      "learning_rate": 4.991321336097546e-06,
+      "loss": 0.695,
+      "step": 672
+    },
+    {
+      "epoch": 0.18642659279778392,
+      "grad_norm": 0.6280376315116882,
+      "learning_rate": 4.991290982239543e-06,
+      "loss": 0.7364,
+      "step": 673
+    },
+    {
+      "epoch": 0.18670360110803325,
+      "grad_norm": 0.6183046102523804,
+      "learning_rate": 4.99126057548507e-06,
+      "loss": 0.7221,
+      "step": 674
+    },
+    {
+      "epoch": 0.18698060941828254,
+      "grad_norm": 0.6224685311317444,
+      "learning_rate": 4.991230115834773e-06,
+      "loss": 0.7818,
+      "step": 675
+    },
+    {
+      "epoch": 0.18725761772853186,
+      "grad_norm": 0.6200198531150818,
+      "learning_rate": 4.991199603289299e-06,
+      "loss": 0.7769,
+      "step": 676
+    },
+    {
+      "epoch": 0.18753462603878115,
+      "grad_norm": 0.6149685978889465,
+      "learning_rate": 4.991169037849296e-06,
+      "loss": 0.7371,
+      "step": 677
+    },
+    {
+      "epoch": 0.18781163434903048,
+      "grad_norm": 0.6446099877357483,
+      "learning_rate": 4.991138419515412e-06,
+      "loss": 0.7273,
+      "step": 678
+    },
+    {
+      "epoch": 0.18808864265927977,
+      "grad_norm": 0.6008735299110413,
+      "learning_rate": 4.991107748288297e-06,
+      "loss": 0.751,
+      "step": 679
+    },
+    {
+      "epoch": 0.1883656509695291,
+      "grad_norm": 0.6307796239852905,
+      "learning_rate": 4.9910770241686045e-06,
+      "loss": 0.7407,
+      "step": 680
+    },
+    {
+      "epoch": 0.18864265927977839,
+      "grad_norm": 0.5771614909172058,
+      "learning_rate": 4.991046247156984e-06,
+      "loss": 0.7596,
+      "step": 681
+    },
+    {
+      "epoch": 0.1889196675900277,
+      "grad_norm": 0.5759564638137817,
+      "learning_rate": 4.991015417254092e-06,
+      "loss": 0.731,
+      "step": 682
+    },
+    {
+      "epoch": 0.189196675900277,
+      "grad_norm": 0.5892442464828491,
+      "learning_rate": 4.9909845344605786e-06,
+      "loss": 0.712,
+      "step": 683
+    },
+    {
+      "epoch": 0.18947368421052632,
+      "grad_norm": 0.6216954588890076,
+      "learning_rate": 4.990953598777104e-06,
+      "loss": 0.7522,
+      "step": 684
+    },
+    {
+      "epoch": 0.18975069252077562,
+      "grad_norm": 0.6018840670585632,
+      "learning_rate": 4.9909226102043226e-06,
+      "loss": 0.6901,
+      "step": 685
+    },
+    {
+      "epoch": 0.19002770083102494,
+      "grad_norm": 0.609626054763794,
+      "learning_rate": 4.9908915687428935e-06,
+      "loss": 0.7855,
+      "step": 686
+    },
+    {
+      "epoch": 0.19030470914127423,
+      "grad_norm": 0.604314386844635,
+      "learning_rate": 4.990860474393476e-06,
+      "loss": 0.7274,
+      "step": 687
+    },
+    {
+      "epoch": 0.19058171745152355,
+      "grad_norm": 0.6209576725959778,
+      "learning_rate": 4.990829327156729e-06,
+      "loss": 0.7528,
+      "step": 688
+    },
+    {
+      "epoch": 0.19085872576177285,
+      "grad_norm": 0.6131774187088013,
+      "learning_rate": 4.990798127033314e-06,
+      "loss": 0.7726,
+      "step": 689
+    },
+    {
+      "epoch": 0.19113573407202217,
+      "grad_norm": 0.5770470499992371,
+      "learning_rate": 4.990766874023894e-06,
+      "loss": 0.7164,
+      "step": 690
+    },
+    {
+      "epoch": 0.19141274238227146,
+      "grad_norm": 0.6258401274681091,
+      "learning_rate": 4.990735568129133e-06,
+      "loss": 0.7613,
+      "step": 691
+    },
+    {
+      "epoch": 0.19168975069252078,
+      "grad_norm": 0.5941904783248901,
+      "learning_rate": 4.990704209349696e-06,
+      "loss": 0.7429,
+      "step": 692
+    },
+    {
+      "epoch": 0.19196675900277008,
+      "grad_norm": 0.633047878742218,
+      "learning_rate": 4.990672797686247e-06,
+      "loss": 0.707,
+      "step": 693
+    },
+    {
+      "epoch": 0.1922437673130194,
+      "grad_norm": 0.6325700283050537,
+      "learning_rate": 4.990641333139455e-06,
+      "loss": 0.7246,
+      "step": 694
+    },
+    {
+      "epoch": 0.1925207756232687,
+      "grad_norm": 0.6009621024131775,
+      "learning_rate": 4.990609815709985e-06,
+      "loss": 0.7101,
+      "step": 695
+    },
+    {
+      "epoch": 0.192797783933518,
+      "grad_norm": 0.5919104218482971,
+      "learning_rate": 4.990578245398511e-06,
+      "loss": 0.6893,
+      "step": 696
+    },
+    {
+      "epoch": 0.1930747922437673,
+      "grad_norm": 0.6189974546432495,
+      "learning_rate": 4.990546622205698e-06,
+      "loss": 0.756,
+      "step": 697
+    },
+    {
+      "epoch": 0.19335180055401663,
+      "grad_norm": 0.632219135761261,
+      "learning_rate": 4.990514946132222e-06,
+      "loss": 0.7153,
+      "step": 698
+    },
+    {
+      "epoch": 0.19362880886426592,
+      "grad_norm": 0.6028069853782654,
+      "learning_rate": 4.990483217178753e-06,
+      "loss": 0.7601,
+      "step": 699
+    },
+    {
+      "epoch": 0.19390581717451524,
+      "grad_norm": 0.6530799865722656,
+      "learning_rate": 4.9904514353459656e-06,
+      "loss": 0.7119,
+      "step": 700
+    },
+    {
+      "epoch": 0.19418282548476454,
+      "grad_norm": 0.624106764793396,
+      "learning_rate": 4.990419600634534e-06,
+      "loss": 0.7418,
+      "step": 701
+    },
+    {
+      "epoch": 0.19445983379501386,
+      "grad_norm": 0.651628851890564,
+      "learning_rate": 4.990387713045134e-06,
+      "loss": 0.7631,
+      "step": 702
+    },
+    {
+      "epoch": 0.19473684210526315,
+      "grad_norm": 0.6010842323303223,
+      "learning_rate": 4.990355772578444e-06,
+      "loss": 0.7143,
+      "step": 703
+    },
+    {
+      "epoch": 0.19501385041551247,
+      "grad_norm": 0.5809913873672485,
+      "learning_rate": 4.990323779235141e-06,
+      "loss": 0.7532,
+      "step": 704
+    },
+    {
+      "epoch": 0.19529085872576177,
+      "grad_norm": 0.610136866569519,
+      "learning_rate": 4.990291733015904e-06,
+      "loss": 0.7433,
+      "step": 705
+    },
+    {
+      "epoch": 0.1955678670360111,
+      "grad_norm": 0.6137512922286987,
+      "learning_rate": 4.990259633921414e-06,
+      "loss": 0.7503,
+      "step": 706
+    },
+    {
+      "epoch": 0.19584487534626038,
+      "grad_norm": 0.5880718231201172,
+      "learning_rate": 4.990227481952353e-06,
+      "loss": 0.7344,
+      "step": 707
+    },
+    {
+      "epoch": 0.1961218836565097,
+      "grad_norm": 0.6280733346939087,
+      "learning_rate": 4.990195277109403e-06,
+      "loss": 0.7246,
+      "step": 708
+    },
+    {
+      "epoch": 0.196398891966759,
+      "grad_norm": 0.7048695683479309,
+      "learning_rate": 4.9901630193932485e-06,
+      "loss": 0.7129,
+      "step": 709
+    },
+    {
+      "epoch": 0.19667590027700832,
+      "grad_norm": 0.5997457504272461,
+      "learning_rate": 4.990130708804573e-06,
+      "loss": 0.7224,
+      "step": 710
+    },
+    {
+      "epoch": 0.1969529085872576,
+      "grad_norm": 0.5950084328651428,
+      "learning_rate": 4.990098345344064e-06,
+      "loss": 0.7666,
+      "step": 711
+    },
+    {
+      "epoch": 0.19722991689750694,
+      "grad_norm": 0.5888988375663757,
+      "learning_rate": 4.9900659290124085e-06,
+      "loss": 0.7168,
+      "step": 712
+    },
+    {
+      "epoch": 0.19750692520775623,
+      "grad_norm": 0.5948391556739807,
+      "learning_rate": 4.990033459810294e-06,
+      "loss": 0.6632,
+      "step": 713
+    },
+    {
+      "epoch": 0.19778393351800555,
+      "grad_norm": 0.5904232263565063,
+      "learning_rate": 4.9900009377384094e-06,
+      "loss": 0.6901,
+      "step": 714
+    },
+    {
+      "epoch": 0.19806094182825484,
+      "grad_norm": 0.6024726033210754,
+      "learning_rate": 4.9899683627974475e-06,
+      "loss": 0.7396,
+      "step": 715
+    },
+    {
+      "epoch": 0.19833795013850417,
+      "grad_norm": 0.6176775693893433,
+      "learning_rate": 4.989935734988098e-06,
+      "loss": 0.7345,
+      "step": 716
+    },
+    {
+      "epoch": 0.19861495844875346,
+      "grad_norm": 0.6300408840179443,
+      "learning_rate": 4.989903054311055e-06,
+      "loss": 0.7106,
+      "step": 717
+    },
+    {
+      "epoch": 0.19889196675900278,
+      "grad_norm": 0.6302835941314697,
+      "learning_rate": 4.9898703207670106e-06,
+      "loss": 0.7154,
+      "step": 718
+    },
+    {
+      "epoch": 0.19916897506925207,
+      "grad_norm": 0.5916628241539001,
+      "learning_rate": 4.989837534356661e-06,
+      "loss": 0.7096,
+      "step": 719
+    },
+    {
+      "epoch": 0.1994459833795014,
+      "grad_norm": 0.6278592944145203,
+      "learning_rate": 4.989804695080703e-06,
+      "loss": 0.7551,
+      "step": 720
+    },
+    {
+      "epoch": 0.1997229916897507,
+      "grad_norm": 0.6208764314651489,
+      "learning_rate": 4.989771802939831e-06,
+      "loss": 0.7568,
+      "step": 721
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6690845489501953,
+      "learning_rate": 4.989738857934747e-06,
+      "loss": 0.7503,
+      "step": 722
+    },
+    {
+      "epoch": 0.2002770083102493,
+      "grad_norm": 0.6054900288581848,
+      "learning_rate": 4.989705860066149e-06,
+      "loss": 0.709,
+      "step": 723
+    },
+    {
+      "epoch": 0.20055401662049863,
+      "grad_norm": 0.6240448951721191,
+      "learning_rate": 4.989672809334738e-06,
+      "loss": 0.6887,
+      "step": 724
+    },
+    {
+      "epoch": 0.20083102493074792,
+      "grad_norm": 0.5848708748817444,
+      "learning_rate": 4.989639705741214e-06,
+      "loss": 0.733,
+      "step": 725
+    },
+    {
+      "epoch": 0.20110803324099724,
+      "grad_norm": 0.6031309962272644,
+      "learning_rate": 4.989606549286282e-06,
+      "loss": 0.6822,
+      "step": 726
+    },
+    {
+      "epoch": 0.20138504155124654,
+      "grad_norm": 0.6088325381278992,
+      "learning_rate": 4.989573339970645e-06,
+      "loss": 0.7376,
+      "step": 727
+    },
+    {
+      "epoch": 0.20166204986149586,
+      "grad_norm": 0.6401178240776062,
+      "learning_rate": 4.989540077795009e-06,
+      "loss": 0.6911,
+      "step": 728
+    },
+    {
+      "epoch": 0.20193905817174515,
+      "grad_norm": 0.642235517501831,
+      "learning_rate": 4.989506762760079e-06,
+      "loss": 0.7624,
+      "step": 729
+    },
+    {
+      "epoch": 0.20221606648199447,
+      "grad_norm": 0.6320951581001282,
+      "learning_rate": 4.989473394866563e-06,
+      "loss": 0.7246,
+      "step": 730
+    },
+    {
+      "epoch": 0.20249307479224377,
+      "grad_norm": 0.6383892893791199,
+      "learning_rate": 4.989439974115169e-06,
+      "loss": 0.7692,
+      "step": 731
+    },
+    {
+      "epoch": 0.2027700831024931,
+      "grad_norm": 0.5928354263305664,
+      "learning_rate": 4.989406500506607e-06,
+      "loss": 0.7087,
+      "step": 732
+    },
+    {
+      "epoch": 0.20304709141274238,
+      "grad_norm": 0.6310860514640808,
+      "learning_rate": 4.989372974041589e-06,
+      "loss": 0.7506,
+      "step": 733
+    },
+    {
+      "epoch": 0.2033240997229917,
+      "grad_norm": 0.5935929417610168,
+      "learning_rate": 4.989339394720824e-06,
+      "loss": 0.7258,
+      "step": 734
+    },
+    {
+      "epoch": 0.203601108033241,
+      "grad_norm": 0.6174879670143127,
+      "learning_rate": 4.989305762545027e-06,
+      "loss": 0.7278,
+      "step": 735
+    },
+    {
+      "epoch": 0.20387811634349032,
+      "grad_norm": 0.6096144318580627,
+      "learning_rate": 4.989272077514912e-06,
+      "loss": 0.7432,
+      "step": 736
+    },
+    {
+      "epoch": 0.2041551246537396,
+      "grad_norm": 0.6143541932106018,
+      "learning_rate": 4.9892383396311936e-06,
+      "loss": 0.6882,
+      "step": 737
+    },
+    {
+      "epoch": 0.20443213296398893,
+      "grad_norm": 0.6350477337837219,
+      "learning_rate": 4.989204548894589e-06,
+      "loss": 0.7577,
+      "step": 738
+    },
+    {
+      "epoch": 0.20470914127423823,
+      "grad_norm": 0.5977352857589722,
+      "learning_rate": 4.989170705305815e-06,
+      "loss": 0.765,
+      "step": 739
+    },
+    {
+      "epoch": 0.20498614958448755,
+      "grad_norm": 0.613018274307251,
+      "learning_rate": 4.98913680886559e-06,
+      "loss": 0.7169,
+      "step": 740
+    },
+    {
+      "epoch": 0.20526315789473684,
+      "grad_norm": 0.6501178741455078,
+      "learning_rate": 4.989102859574634e-06,
+      "loss": 0.7803,
+      "step": 741
+    },
+    {
+      "epoch": 0.20554016620498616,
+      "grad_norm": 0.6331166625022888,
+      "learning_rate": 4.9890688574336675e-06,
+      "loss": 0.7322,
+      "step": 742
+    },
+    {
+      "epoch": 0.20581717451523546,
+      "grad_norm": 0.6125751733779907,
+      "learning_rate": 4.989034802443413e-06,
+      "loss": 0.7569,
+      "step": 743
+    },
+    {
+      "epoch": 0.20609418282548475,
+      "grad_norm": 0.6181766390800476,
+      "learning_rate": 4.989000694604593e-06,
+      "loss": 0.7721,
+      "step": 744
+    },
+    {
+      "epoch": 0.20637119113573407,
+      "grad_norm": 0.6107795834541321,
+      "learning_rate": 4.988966533917933e-06,
+      "loss": 0.7308,
+      "step": 745
+    },
+    {
+      "epoch": 0.20664819944598337,
+      "grad_norm": 0.6483302116394043,
+      "learning_rate": 4.988932320384157e-06,
+      "loss": 0.78,
+      "step": 746
+    },
+    {
+      "epoch": 0.2069252077562327,
+      "grad_norm": 0.6239367723464966,
+      "learning_rate": 4.988898054003991e-06,
+      "loss": 0.7576,
+      "step": 747
+    },
+    {
+      "epoch": 0.20720221606648198,
+      "grad_norm": 0.5802024602890015,
+      "learning_rate": 4.988863734778165e-06,
+      "loss": 0.6888,
+      "step": 748
+    },
+    {
+      "epoch": 0.2074792243767313,
+      "grad_norm": 0.5955031514167786,
+      "learning_rate": 4.988829362707404e-06,
+      "loss": 0.7015,
+      "step": 749
+    },
+    {
+      "epoch": 0.2077562326869806,
+      "grad_norm": 0.651870846748352,
+      "learning_rate": 4.98879493779244e-06,
+      "loss": 0.724,
+      "step": 750
+    },
+    {
+      "epoch": 0.20803324099722992,
+      "grad_norm": 0.6392204165458679,
+      "learning_rate": 4.9887604600340055e-06,
+      "loss": 0.7452,
+      "step": 751
+    },
+    {
+      "epoch": 0.2083102493074792,
+      "grad_norm": 0.5952602028846741,
+      "learning_rate": 4.98872592943283e-06,
+      "loss": 0.7044,
+      "step": 752
+    },
+    {
+      "epoch": 0.20858725761772853,
+      "grad_norm": 0.6271218061447144,
+      "learning_rate": 4.9886913459896475e-06,
+      "loss": 0.7442,
+      "step": 753
+    },
+    {
+      "epoch": 0.20886426592797783,
+      "grad_norm": 0.6348155736923218,
+      "learning_rate": 4.988656709705192e-06,
+      "loss": 0.6851,
+      "step": 754
+    },
+    {
+      "epoch": 0.20914127423822715,
+      "grad_norm": 0.6025179028511047,
+      "learning_rate": 4.9886220205802e-06,
+      "loss": 0.7291,
+      "step": 755
+    },
+    {
+      "epoch": 0.20941828254847644,
+      "grad_norm": 0.607294499874115,
+      "learning_rate": 4.9885872786154065e-06,
+      "loss": 0.7674,
+      "step": 756
+    },
+    {
+      "epoch": 0.20969529085872576,
+      "grad_norm": 0.5688076615333557,
+      "learning_rate": 4.98855248381155e-06,
+      "loss": 0.6903,
+      "step": 757
+    },
+    {
+      "epoch": 0.20997229916897506,
+      "grad_norm": 0.6307339072227478,
+      "learning_rate": 4.988517636169369e-06,
+      "loss": 0.7723,
+      "step": 758
+    },
+    {
+      "epoch": 0.21024930747922438,
+      "grad_norm": 0.569485068321228,
+      "learning_rate": 4.988482735689605e-06,
+      "loss": 0.6656,
+      "step": 759
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.6402713060379028,
+      "learning_rate": 4.988447782372996e-06,
+      "loss": 0.75,
+      "step": 760
+    },
+    {
+      "epoch": 0.210803324099723,
+      "grad_norm": 0.63001549243927,
+      "learning_rate": 4.988412776220286e-06,
+      "loss": 0.7192,
+      "step": 761
+    },
+    {
+      "epoch": 0.2110803324099723,
+      "grad_norm": 0.6070324778556824,
+      "learning_rate": 4.988377717232219e-06,
+      "loss": 0.7489,
+      "step": 762
+    },
+    {
+      "epoch": 0.2113573407202216,
+      "grad_norm": 0.5573827624320984,
+      "learning_rate": 4.9883426054095375e-06,
+      "loss": 0.6487,
+      "step": 763
+    },
+    {
+      "epoch": 0.2116343490304709,
+      "grad_norm": 0.6198872327804565,
+      "learning_rate": 4.988307440752989e-06,
+      "loss": 0.7175,
+      "step": 764
+    },
+    {
+      "epoch": 0.21191135734072022,
+      "grad_norm": 0.6368589401245117,
+      "learning_rate": 4.988272223263317e-06,
+      "loss": 0.7331,
+      "step": 765
+    },
+    {
+      "epoch": 0.21218836565096952,
+      "grad_norm": 0.6162055730819702,
+      "learning_rate": 4.988236952941273e-06,
+      "loss": 0.7289,
+      "step": 766
+    },
+    {
+      "epoch": 0.21246537396121884,
+      "grad_norm": 0.6234159469604492,
+      "learning_rate": 4.988201629787605e-06,
+      "loss": 0.7229,
+      "step": 767
+    },
+    {
+      "epoch": 0.21274238227146813,
+      "grad_norm": 0.6246781945228577,
+      "learning_rate": 4.98816625380306e-06,
+      "loss": 0.7409,
+      "step": 768
+    },
+    {
+      "epoch": 0.21301939058171745,
+      "grad_norm": 0.5837066173553467,
+      "learning_rate": 4.988130824988393e-06,
+      "loss": 0.6673,
+      "step": 769
+    },
+    {
+      "epoch": 0.21329639889196675,
+      "grad_norm": 0.6310555934906006,
+      "learning_rate": 4.988095343344354e-06,
+      "loss": 0.7375,
+      "step": 770
+    },
+    {
+      "epoch": 0.21357340720221607,
+      "grad_norm": 0.5986537933349609,
+      "learning_rate": 4.988059808871697e-06,
+      "loss": 0.6994,
+      "step": 771
+    },
+    {
+      "epoch": 0.21385041551246536,
+      "grad_norm": 0.6631371974945068,
+      "learning_rate": 4.988024221571177e-06,
+      "loss": 0.746,
+      "step": 772
+    },
+    {
+      "epoch": 0.21412742382271469,
+      "grad_norm": 0.627317488193512,
+      "learning_rate": 4.987988581443548e-06,
+      "loss": 0.7713,
+      "step": 773
+    },
+    {
+      "epoch": 0.21440443213296398,
+      "grad_norm": 0.5771815776824951,
+      "learning_rate": 4.987952888489569e-06,
+      "loss": 0.7191,
+      "step": 774
+    },
+    {
+      "epoch": 0.2146814404432133,
+      "grad_norm": 0.5975073575973511,
+      "learning_rate": 4.9879171427099956e-06,
+      "loss": 0.7169,
+      "step": 775
+    },
+    {
+      "epoch": 0.2149584487534626,
+      "grad_norm": 0.6115068197250366,
+      "learning_rate": 4.987881344105589e-06,
+      "loss": 0.7416,
+      "step": 776
+    },
+    {
+      "epoch": 0.21523545706371192,
+      "grad_norm": 0.633957028388977,
+      "learning_rate": 4.987845492677107e-06,
+      "loss": 0.7591,
+      "step": 777
+    },
+    {
+      "epoch": 0.2155124653739612,
+      "grad_norm": 0.6105467677116394,
+      "learning_rate": 4.987809588425312e-06,
+      "loss": 0.7584,
+      "step": 778
+    },
+    {
+      "epoch": 0.21578947368421053,
+      "grad_norm": 0.6133649945259094,
+      "learning_rate": 4.987773631350966e-06,
+      "loss": 0.7357,
+      "step": 779
+    },
+    {
+      "epoch": 0.21606648199445982,
+      "grad_norm": 0.607979416847229,
+      "learning_rate": 4.9877376214548335e-06,
+      "loss": 0.6934,
+      "step": 780
+    },
+    {
+      "epoch": 0.21634349030470915,
+      "grad_norm": 0.5923281311988831,
+      "learning_rate": 4.987701558737678e-06,
+      "loss": 0.7748,
+      "step": 781
+    },
+    {
+      "epoch": 0.21662049861495844,
+      "grad_norm": 0.6248568296432495,
+      "learning_rate": 4.9876654432002655e-06,
+      "loss": 0.7222,
+      "step": 782
+    },
+    {
+      "epoch": 0.21689750692520776,
+      "grad_norm": 0.624470055103302,
+      "learning_rate": 4.987629274843363e-06,
+      "loss": 0.7217,
+      "step": 783
+    },
+    {
+      "epoch": 0.21717451523545706,
+      "grad_norm": 0.6328449249267578,
+      "learning_rate": 4.9875930536677385e-06,
+      "loss": 0.7603,
+      "step": 784
+    },
+    {
+      "epoch": 0.21745152354570638,
+      "grad_norm": 0.608456552028656,
+      "learning_rate": 4.987556779674161e-06,
+      "loss": 0.7228,
+      "step": 785
+    },
+    {
+      "epoch": 0.21772853185595567,
+      "grad_norm": 0.8374035954475403,
+      "learning_rate": 4.9875204528633995e-06,
+      "loss": 0.7098,
+      "step": 786
+    },
+    {
+      "epoch": 0.218005540166205,
+      "grad_norm": 0.6189671158790588,
+      "learning_rate": 4.987484073236227e-06,
+      "loss": 0.7367,
+      "step": 787
+    },
+    {
+      "epoch": 0.21828254847645429,
+      "grad_norm": 0.6146777272224426,
+      "learning_rate": 4.987447640793415e-06,
+      "loss": 0.7274,
+      "step": 788
+    },
+    {
+      "epoch": 0.2185595567867036,
+      "grad_norm": 0.6332618594169617,
+      "learning_rate": 4.987411155535737e-06,
+      "loss": 0.7184,
+      "step": 789
+    },
+    {
+      "epoch": 0.2188365650969529,
+      "grad_norm": 0.6429280638694763,
+      "learning_rate": 4.9873746174639695e-06,
+      "loss": 0.7441,
+      "step": 790
+    },
+    {
+      "epoch": 0.21911357340720222,
+      "grad_norm": 0.6396970748901367,
+      "learning_rate": 4.987338026578885e-06,
+      "loss": 0.7532,
+      "step": 791
+    },
+    {
+      "epoch": 0.21939058171745152,
+      "grad_norm": 0.609308123588562,
+      "learning_rate": 4.987301382881263e-06,
+      "loss": 0.719,
+      "step": 792
+    },
+    {
+      "epoch": 0.21966759002770084,
+      "grad_norm": 0.6831649541854858,
+      "learning_rate": 4.987264686371881e-06,
+      "loss": 0.7554,
+      "step": 793
+    },
+    {
+      "epoch": 0.21994459833795013,
+      "grad_norm": 0.6445385813713074,
+      "learning_rate": 4.987227937051518e-06,
+      "loss": 0.7446,
+      "step": 794
+    },
+    {
+      "epoch": 0.22022160664819945,
+      "grad_norm": 0.6014829277992249,
+      "learning_rate": 4.987191134920954e-06,
+      "loss": 0.7416,
+      "step": 795
+    },
+    {
+      "epoch": 0.22049861495844875,
+      "grad_norm": 0.6108548045158386,
+      "learning_rate": 4.987154279980971e-06,
+      "loss": 0.7034,
+      "step": 796
+    },
+    {
+      "epoch": 0.22077562326869807,
+      "grad_norm": 0.6140116453170776,
+      "learning_rate": 4.9871173722323504e-06,
+      "loss": 0.6733,
+      "step": 797
+    },
+    {
+      "epoch": 0.22105263157894736,
+      "grad_norm": 0.6687027215957642,
+      "learning_rate": 4.987080411675877e-06,
+      "loss": 0.7755,
+      "step": 798
+    },
+    {
+      "epoch": 0.22132963988919668,
+      "grad_norm": 0.6575332880020142,
+      "learning_rate": 4.987043398312335e-06,
+      "loss": 0.7536,
+      "step": 799
+    },
+    {
+      "epoch": 0.22160664819944598,
+      "grad_norm": 0.6619399785995483,
+      "learning_rate": 4.98700633214251e-06,
+      "loss": 0.7511,
+      "step": 800
+    },
+    {
+      "epoch": 0.2218836565096953,
+      "grad_norm": 0.6481317281723022,
+      "learning_rate": 4.986969213167191e-06,
+      "loss": 0.7763,
+      "step": 801
+    },
+    {
+      "epoch": 0.2221606648199446,
+      "grad_norm": 0.6060312390327454,
+      "learning_rate": 4.986932041387163e-06,
+      "loss": 0.709,
+      "step": 802
+    },
+    {
+      "epoch": 0.2224376731301939,
+      "grad_norm": 0.6025817394256592,
+      "learning_rate": 4.986894816803217e-06,
+      "loss": 0.759,
+      "step": 803
+    },
+    {
+      "epoch": 0.2227146814404432,
+      "grad_norm": 0.6360222697257996,
+      "learning_rate": 4.986857539416144e-06,
+      "loss": 0.7556,
+      "step": 804
+    },
+    {
+      "epoch": 0.22299168975069253,
+      "grad_norm": 0.6374087333679199,
+      "learning_rate": 4.986820209226734e-06,
+      "loss": 0.7051,
+      "step": 805
+    },
+    {
+      "epoch": 0.22326869806094182,
+      "grad_norm": 0.6589286923408508,
+      "learning_rate": 4.9867828262357805e-06,
+      "loss": 0.7401,
+      "step": 806
+    },
+    {
+      "epoch": 0.22354570637119114,
+      "grad_norm": 0.6293102502822876,
+      "learning_rate": 4.986745390444077e-06,
+      "loss": 0.6715,
+      "step": 807
+    },
+    {
+      "epoch": 0.22382271468144044,
+      "grad_norm": 0.6365495324134827,
+      "learning_rate": 4.986707901852419e-06,
+      "loss": 0.7471,
+      "step": 808
+    },
+    {
+      "epoch": 0.22409972299168976,
+      "grad_norm": 0.5922935605049133,
+      "learning_rate": 4.986670360461601e-06,
+      "loss": 0.7118,
+      "step": 809
+    },
+    {
+      "epoch": 0.22437673130193905,
+      "grad_norm": 0.6289665102958679,
+      "learning_rate": 4.986632766272421e-06,
+      "loss": 0.7085,
+      "step": 810
+    },
+    {
+      "epoch": 0.22465373961218837,
+      "grad_norm": 0.6416603922843933,
+      "learning_rate": 4.986595119285678e-06,
+      "loss": 0.6975,
+      "step": 811
+    },
+    {
+      "epoch": 0.22493074792243767,
+      "grad_norm": 0.6333436965942383,
+      "learning_rate": 4.986557419502171e-06,
+      "loss": 0.6943,
+      "step": 812
+    },
+    {
+      "epoch": 0.225207756232687,
+      "grad_norm": 0.6685619354248047,
+      "learning_rate": 4.986519666922698e-06,
+      "loss": 0.7649,
+      "step": 813
+    },
+    {
+      "epoch": 0.22548476454293628,
+      "grad_norm": 0.6346633434295654,
+      "learning_rate": 4.986481861548064e-06,
+      "loss": 0.6692,
+      "step": 814
+    },
+    {
+      "epoch": 0.2257617728531856,
+      "grad_norm": 0.6223672032356262,
+      "learning_rate": 4.98644400337907e-06,
+      "loss": 0.7081,
+      "step": 815
+    },
+    {
+      "epoch": 0.2260387811634349,
+      "grad_norm": 0.6629369258880615,
+      "learning_rate": 4.986406092416521e-06,
+      "loss": 0.7281,
+      "step": 816
+    },
+    {
+      "epoch": 0.22631578947368422,
+      "grad_norm": 0.629992663860321,
+      "learning_rate": 4.98636812866122e-06,
+      "loss": 0.731,
+      "step": 817
+    },
+    {
+      "epoch": 0.22659279778393351,
+      "grad_norm": 0.6937189102172852,
+      "learning_rate": 4.986330112113974e-06,
+      "loss": 0.7315,
+      "step": 818
+    },
+    {
+      "epoch": 0.22686980609418284,
+      "grad_norm": 0.6419219374656677,
+      "learning_rate": 4.9862920427755905e-06,
+      "loss": 0.7175,
+      "step": 819
+    },
+    {
+      "epoch": 0.22714681440443213,
+      "grad_norm": 0.6525835990905762,
+      "learning_rate": 4.986253920646878e-06,
+      "loss": 0.719,
+      "step": 820
+    },
+    {
+      "epoch": 0.22742382271468145,
+      "grad_norm": 0.6850484013557434,
+      "learning_rate": 4.986215745728645e-06,
+      "loss": 0.761,
+      "step": 821
+    },
+    {
+      "epoch": 0.22770083102493074,
+      "grad_norm": 0.6086088418960571,
+      "learning_rate": 4.986177518021703e-06,
+      "loss": 0.7367,
+      "step": 822
+    },
+    {
+      "epoch": 0.22797783933518007,
+      "grad_norm": 0.6392209529876709,
+      "learning_rate": 4.986139237526863e-06,
+      "loss": 0.7346,
+      "step": 823
+    },
+    {
+      "epoch": 0.22825484764542936,
+      "grad_norm": 0.6084146499633789,
+      "learning_rate": 4.9861009042449385e-06,
+      "loss": 0.7371,
+      "step": 824
+    },
+    {
+      "epoch": 0.22853185595567868,
+      "grad_norm": 0.6334027051925659,
+      "learning_rate": 4.986062518176743e-06,
+      "loss": 0.7736,
+      "step": 825
+    },
+    {
+      "epoch": 0.22880886426592797,
+      "grad_norm": 0.610819935798645,
+      "learning_rate": 4.986024079323092e-06,
+      "loss": 0.7199,
+      "step": 826
+    },
+    {
+      "epoch": 0.2290858725761773,
+      "grad_norm": 0.6382309198379517,
+      "learning_rate": 4.9859855876848e-06,
+      "loss": 0.7258,
+      "step": 827
+    },
+    {
+      "epoch": 0.2293628808864266,
+      "grad_norm": 0.6213927268981934,
+      "learning_rate": 4.985947043262686e-06,
+      "loss": 0.7467,
+      "step": 828
+    },
+    {
+      "epoch": 0.2296398891966759,
+      "grad_norm": 0.66412752866745,
+      "learning_rate": 4.985908446057568e-06,
+      "loss": 0.741,
+      "step": 829
+    },
+    {
+      "epoch": 0.2299168975069252,
+      "grad_norm": 0.6357157826423645,
+      "learning_rate": 4.985869796070265e-06,
+      "loss": 0.7309,
+      "step": 830
+    },
+    {
+      "epoch": 0.23019390581717453,
+      "grad_norm": 0.6128206849098206,
+      "learning_rate": 4.985831093301599e-06,
+      "loss": 0.7136,
+      "step": 831
+    },
+    {
+      "epoch": 0.23047091412742382,
+      "grad_norm": 0.6191337704658508,
+      "learning_rate": 4.98579233775239e-06,
+      "loss": 0.707,
+      "step": 832
+    },
+    {
+      "epoch": 0.23074792243767314,
+      "grad_norm": 0.6454660892486572,
+      "learning_rate": 4.9857535294234615e-06,
+      "loss": 0.7063,
+      "step": 833
+    },
+    {
+      "epoch": 0.23102493074792244,
+      "grad_norm": 0.6227606534957886,
+      "learning_rate": 4.985714668315637e-06,
+      "loss": 0.6786,
+      "step": 834
+    },
+    {
+      "epoch": 0.23130193905817176,
+      "grad_norm": 0.6117094159126282,
+      "learning_rate": 4.985675754429744e-06,
+      "loss": 0.6787,
+      "step": 835
+    },
+    {
+      "epoch": 0.23157894736842105,
+      "grad_norm": 0.6687617301940918,
+      "learning_rate": 4.985636787766605e-06,
+      "loss": 0.7481,
+      "step": 836
+    },
+    {
+      "epoch": 0.23185595567867037,
+      "grad_norm": 0.6518902778625488,
+      "learning_rate": 4.985597768327051e-06,
+      "loss": 0.6721,
+      "step": 837
+    },
+    {
+      "epoch": 0.23213296398891967,
+      "grad_norm": 0.6399495005607605,
+      "learning_rate": 4.9855586961119086e-06,
+      "loss": 0.7481,
+      "step": 838
+    },
+    {
+      "epoch": 0.232409972299169,
+      "grad_norm": 0.6600784063339233,
+      "learning_rate": 4.985519571122007e-06,
+      "loss": 0.7392,
+      "step": 839
+    },
+    {
+      "epoch": 0.23268698060941828,
+      "grad_norm": 0.6567347049713135,
+      "learning_rate": 4.985480393358178e-06,
+      "loss": 0.7524,
+      "step": 840
+    },
+    {
+      "epoch": 0.2329639889196676,
+      "grad_norm": 0.6045350432395935,
+      "learning_rate": 4.985441162821253e-06,
+      "loss": 0.6875,
+      "step": 841
+    },
+    {
+      "epoch": 0.2332409972299169,
+      "grad_norm": 0.6198522448539734,
+      "learning_rate": 4.985401879512065e-06,
+      "loss": 0.7168,
+      "step": 842
+    },
+    {
+      "epoch": 0.23351800554016622,
+      "grad_norm": 0.6399564146995544,
+      "learning_rate": 4.985362543431449e-06,
+      "loss": 0.7649,
+      "step": 843
+    },
+    {
+      "epoch": 0.2337950138504155,
+      "grad_norm": 0.6472421884536743,
+      "learning_rate": 4.985323154580237e-06,
+      "loss": 0.7257,
+      "step": 844
+    },
+    {
+      "epoch": 0.23407202216066483,
+      "grad_norm": 0.6288352608680725,
+      "learning_rate": 4.98528371295927e-06,
+      "loss": 0.7581,
+      "step": 845
+    },
+    {
+      "epoch": 0.23434903047091413,
+      "grad_norm": 0.664064884185791,
+      "learning_rate": 4.985244218569382e-06,
+      "loss": 0.7395,
+      "step": 846
+    },
+    {
+      "epoch": 0.23462603878116345,
+      "grad_norm": 0.6480230689048767,
+      "learning_rate": 4.985204671411412e-06,
+      "loss": 0.7486,
+      "step": 847
+    },
+    {
+      "epoch": 0.23490304709141274,
+      "grad_norm": 0.661375105381012,
+      "learning_rate": 4.985165071486201e-06,
+      "loss": 0.7107,
+      "step": 848
+    },
+    {
+      "epoch": 0.23518005540166206,
+      "grad_norm": 0.6101104021072388,
+      "learning_rate": 4.985125418794589e-06,
+      "loss": 0.6472,
+      "step": 849
+    },
+    {
+      "epoch": 0.23545706371191136,
+      "grad_norm": 0.6327786445617676,
+      "learning_rate": 4.985085713337418e-06,
+      "loss": 0.7367,
+      "step": 850
+    },
+    {
+      "epoch": 0.23573407202216065,
+      "grad_norm": 0.645991325378418,
+      "learning_rate": 4.9850459551155316e-06,
+      "loss": 0.7331,
+      "step": 851
+    },
+    {
+      "epoch": 0.23601108033240997,
+      "grad_norm": 0.6416239738464355,
+      "learning_rate": 4.985006144129772e-06,
+      "loss": 0.7287,
+      "step": 852
+    },
+    {
+      "epoch": 0.23628808864265927,
+      "grad_norm": 0.6345062851905823,
+      "learning_rate": 4.984966280380987e-06,
+      "loss": 0.7065,
+      "step": 853
+    },
+    {
+      "epoch": 0.2365650969529086,
+      "grad_norm": 0.627337634563446,
+      "learning_rate": 4.984926363870023e-06,
+      "loss": 0.7425,
+      "step": 854
+    },
+    {
+      "epoch": 0.23684210526315788,
+      "grad_norm": 0.5949876308441162,
+      "learning_rate": 4.984886394597725e-06,
+      "loss": 0.6768,
+      "step": 855
+    },
+    {
+      "epoch": 0.2371191135734072,
+      "grad_norm": 0.6694568395614624,
+      "learning_rate": 4.984846372564943e-06,
+      "loss": 0.6219,
+      "step": 856
+    },
+    {
+      "epoch": 0.2373961218836565,
+      "grad_norm": 0.6464795470237732,
+      "learning_rate": 4.9848062977725275e-06,
+      "loss": 0.7199,
+      "step": 857
+    },
+    {
+      "epoch": 0.23767313019390582,
+      "grad_norm": 0.6215987801551819,
+      "learning_rate": 4.984766170221329e-06,
+      "loss": 0.6829,
+      "step": 858
+    },
+    {
+      "epoch": 0.2379501385041551,
+      "grad_norm": 0.6603294014930725,
+      "learning_rate": 4.9847259899121995e-06,
+      "loss": 0.7445,
+      "step": 859
+    },
+    {
+      "epoch": 0.23822714681440443,
+      "grad_norm": 0.6341590881347656,
+      "learning_rate": 4.984685756845991e-06,
+      "loss": 0.7204,
+      "step": 860
+    },
+    {
+      "epoch": 0.23850415512465373,
+      "grad_norm": 0.6417688727378845,
+      "learning_rate": 4.984645471023559e-06,
+      "loss": 0.7094,
+      "step": 861
+    },
+    {
+      "epoch": 0.23878116343490305,
+      "grad_norm": 0.6469441056251526,
+      "learning_rate": 4.984605132445759e-06,
+      "loss": 0.7574,
+      "step": 862
+    },
+    {
+      "epoch": 0.23905817174515234,
+      "grad_norm": 0.6261152625083923,
+      "learning_rate": 4.984564741113447e-06,
+      "loss": 0.7736,
+      "step": 863
+    },
+    {
+      "epoch": 0.23933518005540166,
+      "grad_norm": 0.6391873955726624,
+      "learning_rate": 4.98452429702748e-06,
+      "loss": 0.7671,
+      "step": 864
+    },
+    {
+      "epoch": 0.23961218836565096,
+      "grad_norm": 0.6160810589790344,
+      "learning_rate": 4.984483800188718e-06,
+      "loss": 0.7399,
+      "step": 865
+    },
+    {
+      "epoch": 0.23988919667590028,
+      "grad_norm": 0.6426335573196411,
+      "learning_rate": 4.98444325059802e-06,
+      "loss": 0.6894,
+      "step": 866
+    },
+    {
+      "epoch": 0.24016620498614957,
+      "grad_norm": 0.6205769777297974,
+      "learning_rate": 4.984402648256247e-06,
+      "loss": 0.73,
+      "step": 867
+    },
+    {
+      "epoch": 0.2404432132963989,
+      "grad_norm": 0.6322343349456787,
+      "learning_rate": 4.9843619931642615e-06,
+      "loss": 0.7144,
+      "step": 868
+    },
+    {
+      "epoch": 0.2407202216066482,
+      "grad_norm": 0.6147611737251282,
+      "learning_rate": 4.984321285322927e-06,
+      "loss": 0.7145,
+      "step": 869
+    },
+    {
+      "epoch": 0.2409972299168975,
+      "grad_norm": 0.6592707633972168,
+      "learning_rate": 4.984280524733107e-06,
+      "loss": 0.7235,
+      "step": 870
+    },
+    {
+      "epoch": 0.2412742382271468,
+      "grad_norm": 0.6175403594970703,
+      "learning_rate": 4.984239711395667e-06,
+      "loss": 0.7058,
+      "step": 871
+    },
+    {
+      "epoch": 0.24155124653739612,
+      "grad_norm": 0.6178988218307495,
+      "learning_rate": 4.984198845311474e-06,
+      "loss": 0.7353,
+      "step": 872
+    },
+    {
+      "epoch": 0.24182825484764542,
+      "grad_norm": 0.6353136897087097,
+      "learning_rate": 4.984157926481396e-06,
+      "loss": 0.6959,
+      "step": 873
+    },
+    {
+      "epoch": 0.24210526315789474,
+      "grad_norm": 0.6403223276138306,
+      "learning_rate": 4.9841169549063e-06,
+      "loss": 0.71,
+      "step": 874
+    },
+    {
+      "epoch": 0.24238227146814403,
+      "grad_norm": 0.7069513201713562,
+      "learning_rate": 4.984075930587059e-06,
+      "loss": 0.7327,
+      "step": 875
+    },
+    {
+      "epoch": 0.24265927977839336,
+      "grad_norm": 0.6697827577590942,
+      "learning_rate": 4.984034853524542e-06,
+      "loss": 0.7323,
+      "step": 876
+    },
+    {
+      "epoch": 0.24293628808864265,
+      "grad_norm": 0.6514555811882019,
+      "learning_rate": 4.983993723719621e-06,
+      "loss": 0.6917,
+      "step": 877
+    },
+    {
+      "epoch": 0.24321329639889197,
+      "grad_norm": 0.6172816157341003,
+      "learning_rate": 4.9839525411731705e-06,
+      "loss": 0.7335,
+      "step": 878
+    },
+    {
+      "epoch": 0.24349030470914126,
+      "grad_norm": 0.6580623388290405,
+      "learning_rate": 4.983911305886063e-06,
+      "loss": 0.7731,
+      "step": 879
+    },
+    {
+      "epoch": 0.24376731301939059,
+      "grad_norm": 0.6681740880012512,
+      "learning_rate": 4.983870017859177e-06,
+      "loss": 0.7812,
+      "step": 880
+    },
+    {
+      "epoch": 0.24404432132963988,
+      "grad_norm": 0.6525234580039978,
+      "learning_rate": 4.983828677093386e-06,
+      "loss": 0.6724,
+      "step": 881
+    },
+    {
+      "epoch": 0.2443213296398892,
+      "grad_norm": 0.6404354572296143,
+      "learning_rate": 4.98378728358957e-06,
+      "loss": 0.7743,
+      "step": 882
+    },
+    {
+      "epoch": 0.2445983379501385,
+      "grad_norm": 0.6499460935592651,
+      "learning_rate": 4.983745837348606e-06,
+      "loss": 0.7337,
+      "step": 883
+    },
+    {
+      "epoch": 0.24487534626038782,
+      "grad_norm": 0.6775996685028076,
+      "learning_rate": 4.983704338371375e-06,
+      "loss": 0.7413,
+      "step": 884
+    },
+    {
+      "epoch": 0.2451523545706371,
+      "grad_norm": 0.6776655912399292,
+      "learning_rate": 4.98366278665876e-06,
+      "loss": 0.7211,
+      "step": 885
+    },
+    {
+      "epoch": 0.24542936288088643,
+      "grad_norm": 0.6310790777206421,
+      "learning_rate": 4.9836211822116405e-06,
+      "loss": 0.6957,
+      "step": 886
+    },
+    {
+      "epoch": 0.24570637119113573,
+      "grad_norm": 0.5974736213684082,
+      "learning_rate": 4.983579525030901e-06,
+      "loss": 0.7169,
+      "step": 887
+    },
+    {
+      "epoch": 0.24598337950138505,
+      "grad_norm": 0.6585884690284729,
+      "learning_rate": 4.983537815117425e-06,
+      "loss": 0.7598,
+      "step": 888
+    },
+    {
+      "epoch": 0.24626038781163434,
+      "grad_norm": 0.6152383685112,
+      "learning_rate": 4.9834960524721e-06,
+      "loss": 0.7222,
+      "step": 889
+    },
+    {
+      "epoch": 0.24653739612188366,
+      "grad_norm": 0.6026103496551514,
+      "learning_rate": 4.98345423709581e-06,
+      "loss": 0.7385,
+      "step": 890
+    },
+    {
+      "epoch": 0.24681440443213296,
+      "grad_norm": 0.5877223610877991,
+      "learning_rate": 4.983412368989447e-06,
+      "loss": 0.6195,
+      "step": 891
+    },
+    {
+      "epoch": 0.24709141274238228,
+      "grad_norm": 0.648893415927887,
+      "learning_rate": 4.983370448153896e-06,
+      "loss": 0.7223,
+      "step": 892
+    },
+    {
+      "epoch": 0.24736842105263157,
+      "grad_norm": 0.6290050745010376,
+      "learning_rate": 4.983328474590048e-06,
+      "loss": 0.6909,
+      "step": 893
+    },
+    {
+      "epoch": 0.2476454293628809,
+      "grad_norm": 0.6397932171821594,
+      "learning_rate": 4.983286448298795e-06,
+      "loss": 0.7623,
+      "step": 894
+    },
+    {
+      "epoch": 0.24792243767313019,
+      "grad_norm": 0.6201431155204773,
+      "learning_rate": 4.98324436928103e-06,
+      "loss": 0.6838,
+      "step": 895
+    },
+    {
+      "epoch": 0.2481994459833795,
+      "grad_norm": 0.636030375957489,
+      "learning_rate": 4.983202237537645e-06,
+      "loss": 0.7118,
+      "step": 896
+    },
+    {
+      "epoch": 0.2484764542936288,
+      "grad_norm": 0.6343379020690918,
+      "learning_rate": 4.9831600530695345e-06,
+      "loss": 0.7253,
+      "step": 897
+    },
+    {
+      "epoch": 0.24875346260387812,
+      "grad_norm": 0.6185975670814514,
+      "learning_rate": 4.983117815877596e-06,
+      "loss": 0.679,
+      "step": 898
+    },
+    {
+      "epoch": 0.24903047091412742,
+      "grad_norm": 0.6295162439346313,
+      "learning_rate": 4.983075525962724e-06,
+      "loss": 0.6748,
+      "step": 899
+    },
+    {
+      "epoch": 0.24930747922437674,
+      "grad_norm": 0.6754584312438965,
+      "learning_rate": 4.983033183325818e-06,
+      "loss": 0.6952,
+      "step": 900
+    },
+    {
+      "epoch": 0.24958448753462603,
+      "grad_norm": 0.6608603000640869,
+      "learning_rate": 4.982990787967777e-06,
+      "loss": 0.6945,
+      "step": 901
+    },
+    {
+      "epoch": 0.24986149584487535,
+      "grad_norm": 0.6505388617515564,
+      "learning_rate": 4.982948339889499e-06,
+      "loss": 0.6935,
+      "step": 902
+    },
+    {
+      "epoch": 0.2501385041551247,
+      "grad_norm": 0.6118568778038025,
+      "learning_rate": 4.9829058390918885e-06,
+      "loss": 0.7013,
+      "step": 903
+    },
+    {
+      "epoch": 0.25041551246537397,
+      "grad_norm": 0.6262096762657166,
+      "learning_rate": 4.982863285575846e-06,
+      "loss": 0.7221,
+      "step": 904
+    },
+    {
+      "epoch": 0.25069252077562326,
+      "grad_norm": 0.6136447787284851,
+      "learning_rate": 4.982820679342276e-06,
+      "loss": 0.6722,
+      "step": 905
+    },
+    {
+      "epoch": 0.25096952908587256,
+      "grad_norm": 0.6317170858383179,
+      "learning_rate": 4.982778020392082e-06,
+      "loss": 0.712,
+      "step": 906
+    },
+    {
+      "epoch": 0.2512465373961219,
+      "grad_norm": 0.6549311876296997,
+      "learning_rate": 4.98273530872617e-06,
+      "loss": 0.7045,
+      "step": 907
+    },
+    {
+      "epoch": 0.2515235457063712,
+      "grad_norm": 0.6151458621025085,
+      "learning_rate": 4.982692544345448e-06,
+      "loss": 0.663,
+      "step": 908
+    },
+    {
+      "epoch": 0.2518005540166205,
+      "grad_norm": 0.6190466284751892,
+      "learning_rate": 4.9826497272508226e-06,
+      "loss": 0.6824,
+      "step": 909
+    },
+    {
+      "epoch": 0.2520775623268698,
+      "grad_norm": 0.6535850763320923,
+      "learning_rate": 4.982606857443203e-06,
+      "loss": 0.6979,
+      "step": 910
+    },
+    {
+      "epoch": 0.25235457063711914,
+      "grad_norm": 0.6364226341247559,
+      "learning_rate": 4.982563934923501e-06,
+      "loss": 0.7438,
+      "step": 911
+    },
+    {
+      "epoch": 0.25263157894736843,
+      "grad_norm": 0.6416041254997253,
+      "learning_rate": 4.982520959692626e-06,
+      "loss": 0.7186,
+      "step": 912
+    },
+    {
+      "epoch": 0.2529085872576177,
+      "grad_norm": 0.6356797814369202,
+      "learning_rate": 4.982477931751492e-06,
+      "loss": 0.7231,
+      "step": 913
+    },
+    {
+      "epoch": 0.253185595567867,
+      "grad_norm": 0.6105824708938599,
+      "learning_rate": 4.9824348511010115e-06,
+      "loss": 0.687,
+      "step": 914
+    },
+    {
+      "epoch": 0.25346260387811637,
+      "grad_norm": 0.6553460955619812,
+      "learning_rate": 4.9823917177421e-06,
+      "loss": 0.749,
+      "step": 915
+    },
+    {
+      "epoch": 0.25373961218836566,
+      "grad_norm": 0.6527493000030518,
+      "learning_rate": 4.982348531675673e-06,
+      "loss": 0.7491,
+      "step": 916
+    },
+    {
+      "epoch": 0.25401662049861495,
+      "grad_norm": 0.6494553685188293,
+      "learning_rate": 4.982305292902647e-06,
+      "loss": 0.7337,
+      "step": 917
+    },
+    {
+      "epoch": 0.25429362880886425,
+      "grad_norm": 0.6618565320968628,
+      "learning_rate": 4.982262001423941e-06,
+      "loss": 0.72,
+      "step": 918
+    },
+    {
+      "epoch": 0.2545706371191136,
+      "grad_norm": 0.6225207448005676,
+      "learning_rate": 4.982218657240473e-06,
+      "loss": 0.7314,
+      "step": 919
+    },
+    {
+      "epoch": 0.2548476454293629,
+      "grad_norm": 0.6372534036636353,
+      "learning_rate": 4.982175260353164e-06,
+      "loss": 0.7407,
+      "step": 920
+    },
+    {
+      "epoch": 0.2551246537396122,
+      "grad_norm": 0.633301317691803,
+      "learning_rate": 4.982131810762937e-06,
+      "loss": 0.6913,
+      "step": 921
+    },
+    {
+      "epoch": 0.2554016620498615,
+      "grad_norm": 0.6848824620246887,
+      "learning_rate": 4.982088308470712e-06,
+      "loss": 0.7883,
+      "step": 922
+    },
+    {
+      "epoch": 0.2556786703601108,
+      "grad_norm": 0.6465626955032349,
+      "learning_rate": 4.982044753477413e-06,
+      "loss": 0.7233,
+      "step": 923
+    },
+    {
+      "epoch": 0.2559556786703601,
+      "grad_norm": 0.6425303220748901,
+      "learning_rate": 4.982001145783966e-06,
+      "loss": 0.7849,
+      "step": 924
+    },
+    {
+      "epoch": 0.2562326869806094,
+      "grad_norm": 0.6424234509468079,
+      "learning_rate": 4.981957485391297e-06,
+      "loss": 0.7498,
+      "step": 925
+    },
+    {
+      "epoch": 0.2565096952908587,
+      "grad_norm": 0.6227364540100098,
+      "learning_rate": 4.981913772300332e-06,
+      "loss": 0.6802,
+      "step": 926
+    },
+    {
+      "epoch": 0.25678670360110806,
+      "grad_norm": 0.6455993056297302,
+      "learning_rate": 4.981870006511999e-06,
+      "loss": 0.7114,
+      "step": 927
+    },
+    {
+      "epoch": 0.25706371191135735,
+      "grad_norm": 0.6288533806800842,
+      "learning_rate": 4.981826188027228e-06,
+      "loss": 0.6904,
+      "step": 928
+    },
+    {
+      "epoch": 0.25734072022160664,
+      "grad_norm": 0.6399247646331787,
+      "learning_rate": 4.981782316846948e-06,
+      "loss": 0.7539,
+      "step": 929
+    },
+    {
+      "epoch": 0.25761772853185594,
+      "grad_norm": 0.6298650503158569,
+      "learning_rate": 4.981738392972093e-06,
+      "loss": 0.7166,
+      "step": 930
+    },
+    {
+      "epoch": 0.2578947368421053,
+      "grad_norm": 0.6303683519363403,
+      "learning_rate": 4.9816944164035944e-06,
+      "loss": 0.6957,
+      "step": 931
+    },
+    {
+      "epoch": 0.2581717451523546,
+      "grad_norm": 0.6419861316680908,
+      "learning_rate": 4.981650387142385e-06,
+      "loss": 0.7509,
+      "step": 932
+    },
+    {
+      "epoch": 0.2584487534626039,
+      "grad_norm": 0.6510540843009949,
+      "learning_rate": 4.981606305189402e-06,
+      "loss": 0.7132,
+      "step": 933
+    },
+    {
+      "epoch": 0.25872576177285317,
+      "grad_norm": 0.6482325792312622,
+      "learning_rate": 4.981562170545578e-06,
+      "loss": 0.6979,
+      "step": 934
+    },
+    {
+      "epoch": 0.2590027700831025,
+      "grad_norm": 0.6792583465576172,
+      "learning_rate": 4.9815179832118525e-06,
+      "loss": 0.7653,
+      "step": 935
+    },
+    {
+      "epoch": 0.2592797783933518,
+      "grad_norm": 0.6865032911300659,
+      "learning_rate": 4.981473743189163e-06,
+      "loss": 0.7413,
+      "step": 936
+    },
+    {
+      "epoch": 0.2595567867036011,
+      "grad_norm": 0.6258604526519775,
+      "learning_rate": 4.9814294504784484e-06,
+      "loss": 0.7152,
+      "step": 937
+    },
+    {
+      "epoch": 0.2598337950138504,
+      "grad_norm": 0.6872163414955139,
+      "learning_rate": 4.98138510508065e-06,
+      "loss": 0.7077,
+      "step": 938
+    },
+    {
+      "epoch": 0.26011080332409975,
+      "grad_norm": 0.6078726649284363,
+      "learning_rate": 4.98134070699671e-06,
+      "loss": 0.662,
+      "step": 939
+    },
+    {
+      "epoch": 0.26038781163434904,
+      "grad_norm": 0.6605315804481506,
+      "learning_rate": 4.98129625622757e-06,
+      "loss": 0.759,
+      "step": 940
+    },
+    {
+      "epoch": 0.26066481994459834,
+      "grad_norm": 0.691207230091095,
+      "learning_rate": 4.981251752774173e-06,
+      "loss": 0.7287,
+      "step": 941
+    },
+    {
+      "epoch": 0.26094182825484763,
+      "grad_norm": 0.6527872681617737,
+      "learning_rate": 4.981207196637465e-06,
+      "loss": 0.696,
+      "step": 942
+    },
+    {
+      "epoch": 0.261218836565097,
+      "grad_norm": 0.6540459990501404,
+      "learning_rate": 4.981162587818393e-06,
+      "loss": 0.7408,
+      "step": 943
+    },
+    {
+      "epoch": 0.2614958448753463,
+      "grad_norm": 0.6336873769760132,
+      "learning_rate": 4.981117926317902e-06,
+      "loss": 0.774,
+      "step": 944
+    },
+    {
+      "epoch": 0.26177285318559557,
+      "grad_norm": 0.6425933837890625,
+      "learning_rate": 4.981073212136942e-06,
+      "loss": 0.7575,
+      "step": 945
+    },
+    {
+      "epoch": 0.26204986149584486,
+      "grad_norm": 0.6415480971336365,
+      "learning_rate": 4.981028445276461e-06,
+      "loss": 0.7147,
+      "step": 946
+    },
+    {
+      "epoch": 0.2623268698060942,
+      "grad_norm": 0.6138240694999695,
+      "learning_rate": 4.980983625737411e-06,
+      "loss": 0.7271,
+      "step": 947
+    },
+    {
+      "epoch": 0.2626038781163435,
+      "grad_norm": 0.6276071667671204,
+      "learning_rate": 4.980938753520742e-06,
+      "loss": 0.7136,
+      "step": 948
+    },
+    {
+      "epoch": 0.2628808864265928,
+      "grad_norm": 0.6506159901618958,
+      "learning_rate": 4.980893828627408e-06,
+      "loss": 0.687,
+      "step": 949
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.6397688388824463,
+      "learning_rate": 4.980848851058363e-06,
+      "loss": 0.7399,
+      "step": 950
+    },
+    {
+      "epoch": 0.26343490304709144,
+      "grad_norm": 0.6355700492858887,
+      "learning_rate": 4.980803820814562e-06,
+      "loss": 0.7082,
+      "step": 951
+    },
+    {
+      "epoch": 0.26371191135734073,
+      "grad_norm": 0.6730945706367493,
+      "learning_rate": 4.980758737896959e-06,
+      "loss": 0.7546,
+      "step": 952
+    },
+    {
+      "epoch": 0.26398891966759,
+      "grad_norm": 0.6312103271484375,
+      "learning_rate": 4.980713602306514e-06,
+      "loss": 0.6864,
+      "step": 953
+    },
+    {
+      "epoch": 0.2642659279778393,
+      "grad_norm": 0.640343189239502,
+      "learning_rate": 4.980668414044183e-06,
+      "loss": 0.6764,
+      "step": 954
+    },
+    {
+      "epoch": 0.26454293628808867,
+      "grad_norm": 0.6530020833015442,
+      "learning_rate": 4.980623173110928e-06,
+      "loss": 0.7476,
+      "step": 955
+    },
+    {
+      "epoch": 0.26481994459833796,
+      "grad_norm": 0.6287079453468323,
+      "learning_rate": 4.9805778795077074e-06,
+      "loss": 0.7068,
+      "step": 956
+    },
+    {
+      "epoch": 0.26509695290858726,
+      "grad_norm": 0.6119797825813293,
+      "learning_rate": 4.980532533235484e-06,
+      "loss": 0.7638,
+      "step": 957
+    },
+    {
+      "epoch": 0.26537396121883655,
+      "grad_norm": 0.6286291480064392,
+      "learning_rate": 4.98048713429522e-06,
+      "loss": 0.7404,
+      "step": 958
+    },
+    {
+      "epoch": 0.26565096952908585,
+      "grad_norm": 0.6268792748451233,
+      "learning_rate": 4.98044168268788e-06,
+      "loss": 0.7177,
+      "step": 959
+    },
+    {
+      "epoch": 0.2659279778393352,
+      "grad_norm": 0.639512300491333,
+      "learning_rate": 4.980396178414429e-06,
+      "loss": 0.7573,
+      "step": 960
+    },
+    {
+      "epoch": 0.2662049861495845,
+      "grad_norm": 0.6351445317268372,
+      "learning_rate": 4.980350621475832e-06,
+      "loss": 0.7487,
+      "step": 961
+    },
+    {
+      "epoch": 0.2664819944598338,
+      "grad_norm": 0.8666406273841858,
+      "learning_rate": 4.980305011873058e-06,
+      "loss": 0.6926,
+      "step": 962
+    },
+    {
+      "epoch": 0.2667590027700831,
+      "grad_norm": 0.6362487077713013,
+      "learning_rate": 4.980259349607074e-06,
+      "loss": 0.7169,
+      "step": 963
+    },
+    {
+      "epoch": 0.2670360110803324,
+      "grad_norm": 0.6595889925956726,
+      "learning_rate": 4.98021363467885e-06,
+      "loss": 0.7173,
+      "step": 964
+    },
+    {
+      "epoch": 0.2673130193905817,
+      "grad_norm": 0.6907031536102295,
+      "learning_rate": 4.9801678670893575e-06,
+      "loss": 0.7666,
+      "step": 965
+    },
+    {
+      "epoch": 0.267590027700831,
+      "grad_norm": 0.6077108979225159,
+      "learning_rate": 4.980122046839567e-06,
+      "loss": 0.6915,
+      "step": 966
+    },
+    {
+      "epoch": 0.2678670360110803,
+      "grad_norm": 0.685006320476532,
+      "learning_rate": 4.980076173930452e-06,
+      "loss": 0.7328,
+      "step": 967
+    },
+    {
+      "epoch": 0.26814404432132966,
+      "grad_norm": 0.65442955493927,
+      "learning_rate": 4.980030248362987e-06,
+      "loss": 0.7229,
+      "step": 968
+    },
+    {
+      "epoch": 0.26842105263157895,
+      "grad_norm": 0.6928679943084717,
+      "learning_rate": 4.979984270138146e-06,
+      "loss": 0.7643,
+      "step": 969
+    },
+    {
+      "epoch": 0.26869806094182824,
+      "grad_norm": 0.6740157008171082,
+      "learning_rate": 4.979938239256905e-06,
+      "loss": 0.6848,
+      "step": 970
+    },
+    {
+      "epoch": 0.26897506925207754,
+      "grad_norm": 0.675374448299408,
+      "learning_rate": 4.979892155720243e-06,
+      "loss": 0.725,
+      "step": 971
+    },
+    {
+      "epoch": 0.2692520775623269,
+      "grad_norm": 0.6164215207099915,
+      "learning_rate": 4.979846019529138e-06,
+      "loss": 0.6557,
+      "step": 972
+    },
+    {
+      "epoch": 0.2695290858725762,
+      "grad_norm": 0.6450408101081848,
+      "learning_rate": 4.979799830684568e-06,
+      "loss": 0.7266,
+      "step": 973
+    },
+    {
+      "epoch": 0.2698060941828255,
+      "grad_norm": 0.6351450085639954,
+      "learning_rate": 4.9797535891875156e-06,
+      "loss": 0.7281,
+      "step": 974
+    },
+    {
+      "epoch": 0.27008310249307477,
+      "grad_norm": 0.6664936542510986,
+      "learning_rate": 4.9797072950389625e-06,
+      "loss": 0.6896,
+      "step": 975
+    },
+    {
+      "epoch": 0.2703601108033241,
+      "grad_norm": 0.6203340888023376,
+      "learning_rate": 4.97966094823989e-06,
+      "loss": 0.6914,
+      "step": 976
+    },
+    {
+      "epoch": 0.2706371191135734,
+      "grad_norm": 0.6529099941253662,
+      "learning_rate": 4.979614548791283e-06,
+      "loss": 0.727,
+      "step": 977
+    },
+    {
+      "epoch": 0.2709141274238227,
+      "grad_norm": 0.6582198739051819,
+      "learning_rate": 4.979568096694128e-06,
+      "loss": 0.7072,
+      "step": 978
+    },
+    {
+      "epoch": 0.271191135734072,
+      "grad_norm": 0.6922197937965393,
+      "learning_rate": 4.979521591949409e-06,
+      "loss": 0.7379,
+      "step": 979
+    },
+    {
+      "epoch": 0.27146814404432135,
+      "grad_norm": 0.5990357398986816,
+      "learning_rate": 4.979475034558115e-06,
+      "loss": 0.6885,
+      "step": 980
+    },
+    {
+      "epoch": 0.27174515235457064,
+      "grad_norm": 0.5973677635192871,
+      "learning_rate": 4.979428424521235e-06,
+      "loss": 0.711,
+      "step": 981
+    },
+    {
+      "epoch": 0.27202216066481993,
+      "grad_norm": 0.6486063599586487,
+      "learning_rate": 4.979381761839757e-06,
+      "loss": 0.722,
+      "step": 982
+    },
+    {
+      "epoch": 0.27229916897506923,
+      "grad_norm": 0.6575403809547424,
+      "learning_rate": 4.979335046514673e-06,
+      "loss": 0.7246,
+      "step": 983
+    },
+    {
+      "epoch": 0.2725761772853186,
+      "grad_norm": 0.6764856576919556,
+      "learning_rate": 4.979288278546974e-06,
+      "loss": 0.7574,
+      "step": 984
+    },
+    {
+      "epoch": 0.27285318559556787,
+      "grad_norm": 0.6495998501777649,
+      "learning_rate": 4.979241457937653e-06,
+      "loss": 0.7221,
+      "step": 985
+    },
+    {
+      "epoch": 0.27313019390581716,
+      "grad_norm": 0.6304432153701782,
+      "learning_rate": 4.979194584687706e-06,
+      "loss": 0.7644,
+      "step": 986
+    },
+    {
+      "epoch": 0.27340720221606646,
+      "grad_norm": 0.7039188146591187,
+      "learning_rate": 4.9791476587981255e-06,
+      "loss": 0.71,
+      "step": 987
+    },
+    {
+      "epoch": 0.2736842105263158,
+      "grad_norm": 0.6315945386886597,
+      "learning_rate": 4.979100680269909e-06,
+      "loss": 0.7015,
+      "step": 988
+    },
+    {
+      "epoch": 0.2739612188365651,
+      "grad_norm": 0.6366348266601562,
+      "learning_rate": 4.979053649104056e-06,
+      "loss": 0.7001,
+      "step": 989
+    },
+    {
+      "epoch": 0.2742382271468144,
+      "grad_norm": 0.5995786786079407,
+      "learning_rate": 4.979006565301561e-06,
+      "loss": 0.6959,
+      "step": 990
+    },
+    {
+      "epoch": 0.2745152354570637,
+      "grad_norm": 0.6954519152641296,
+      "learning_rate": 4.978959428863427e-06,
+      "loss": 0.7767,
+      "step": 991
+    },
+    {
+      "epoch": 0.27479224376731304,
+      "grad_norm": 0.6461866497993469,
+      "learning_rate": 4.978912239790653e-06,
+      "loss": 0.7315,
+      "step": 992
+    },
+    {
+      "epoch": 0.27506925207756233,
+      "grad_norm": 0.6643589735031128,
+      "learning_rate": 4.978864998084243e-06,
+      "loss": 0.7343,
+      "step": 993
+    },
+    {
+      "epoch": 0.2753462603878116,
+      "grad_norm": 0.6773586869239807,
+      "learning_rate": 4.978817703745198e-06,
+      "loss": 0.7251,
+      "step": 994
+    },
+    {
+      "epoch": 0.2756232686980609,
+      "grad_norm": 0.6450729370117188,
+      "learning_rate": 4.978770356774523e-06,
+      "loss": 0.7223,
+      "step": 995
+    },
+    {
+      "epoch": 0.27590027700831027,
+      "grad_norm": 0.6580459475517273,
+      "learning_rate": 4.978722957173222e-06,
+      "loss": 0.7083,
+      "step": 996
+    },
+    {
+      "epoch": 0.27617728531855956,
+      "grad_norm": 0.6939603090286255,
+      "learning_rate": 4.978675504942304e-06,
+      "loss": 0.6858,
+      "step": 997
+    },
+    {
+      "epoch": 0.27645429362880886,
+      "grad_norm": 0.623268187046051,
+      "learning_rate": 4.9786280000827754e-06,
+      "loss": 0.7417,
+      "step": 998
+    },
+    {
+      "epoch": 0.27673130193905815,
+      "grad_norm": 0.6503739356994629,
+      "learning_rate": 4.978580442595644e-06,
+      "loss": 0.7275,
+      "step": 999
+    },
+    {
+      "epoch": 0.2770083102493075,
+      "grad_norm": 0.6480211019515991,
+      "learning_rate": 4.978532832481921e-06,
+      "loss": 0.7379,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2772853185595568,
+      "grad_norm": 0.6500074863433838,
+      "learning_rate": 4.9784851697426155e-06,
+      "loss": 0.6806,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2775623268698061,
+      "grad_norm": 0.6740723848342896,
+      "learning_rate": 4.978437454378741e-06,
+      "loss": 0.7479,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2778393351800554,
+      "grad_norm": 0.6438689827919006,
+      "learning_rate": 4.97838968639131e-06,
+      "loss": 0.6952,
+      "step": 1003
+    },
+    {
+      "epoch": 0.27811634349030473,
+      "grad_norm": 0.6317799687385559,
+      "learning_rate": 4.978341865781336e-06,
+      "loss": 0.7218,
+      "step": 1004
+    },
+    {
+      "epoch": 0.278393351800554,
+      "grad_norm": 0.6593930721282959,
+      "learning_rate": 4.978293992549837e-06,
+      "loss": 0.7173,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2786703601108033,
+      "grad_norm": 0.6122359037399292,
+      "learning_rate": 4.9782460666978265e-06,
+      "loss": 0.7294,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2789473684210526,
+      "grad_norm": 0.6612704992294312,
+      "learning_rate": 4.978198088226324e-06,
+      "loss": 0.7177,
+      "step": 1007
+    },
+    {
+      "epoch": 0.27922437673130196,
+      "grad_norm": 0.6342014670372009,
+      "learning_rate": 4.9781500571363465e-06,
+      "loss": 0.6808,
+      "step": 1008
+    },
+    {
+      "epoch": 0.27950138504155125,
+      "grad_norm": 0.6799290776252747,
+      "learning_rate": 4.978101973428915e-06,
+      "loss": 0.7603,
+      "step": 1009
+    },
+    {
+      "epoch": 0.27977839335180055,
+      "grad_norm": 0.6468144655227661,
+      "learning_rate": 4.97805383710505e-06,
+      "loss": 0.692,
+      "step": 1010
+    },
+    {
+      "epoch": 0.28005540166204984,
+      "grad_norm": 0.6199710369110107,
+      "learning_rate": 4.978005648165774e-06,
+      "loss": 0.7059,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2803324099722992,
+      "grad_norm": 0.6647071242332458,
+      "learning_rate": 4.97795740661211e-06,
+      "loss": 0.7496,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2806094182825485,
+      "grad_norm": 0.6575323939323425,
+      "learning_rate": 4.977909112445083e-06,
+      "loss": 0.7271,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2808864265927978,
+      "grad_norm": 0.6441571116447449,
+      "learning_rate": 4.977860765665717e-06,
+      "loss": 0.7272,
+      "step": 1014
+    },
+    {
+      "epoch": 0.28116343490304707,
+      "grad_norm": 0.6441684365272522,
+      "learning_rate": 4.97781236627504e-06,
+      "loss": 0.7068,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2814404432132964,
+      "grad_norm": 0.6357406377792358,
+      "learning_rate": 4.977763914274077e-06,
+      "loss": 0.7275,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2817174515235457,
+      "grad_norm": 0.6525790691375732,
+      "learning_rate": 4.9777154096638605e-06,
+      "loss": 0.7323,
+      "step": 1017
+    },
+    {
+      "epoch": 0.281994459833795,
+      "grad_norm": 0.6308574676513672,
+      "learning_rate": 4.977666852445418e-06,
+      "loss": 0.6859,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2822714681440443,
+      "grad_norm": 0.6342403292655945,
+      "learning_rate": 4.97761824261978e-06,
+      "loss": 0.7752,
+      "step": 1019
+    },
+    {
+      "epoch": 0.28254847645429365,
+      "grad_norm": 0.6695272922515869,
+      "learning_rate": 4.97756958018798e-06,
+      "loss": 0.6806,
+      "step": 1020
+    },
+    {
+      "epoch": 0.28282548476454294,
+      "grad_norm": 0.6361978054046631,
+      "learning_rate": 4.9775208651510505e-06,
+      "loss": 0.719,
+      "step": 1021
+    },
+    {
+      "epoch": 0.28310249307479224,
+      "grad_norm": 0.6587631106376648,
+      "learning_rate": 4.977472097510027e-06,
+      "loss": 0.7863,
+      "step": 1022
+    },
+    {
+      "epoch": 0.28337950138504153,
+      "grad_norm": 0.6483465433120728,
+      "learning_rate": 4.977423277265943e-06,
+      "loss": 0.7512,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2836565096952909,
+      "grad_norm": 0.6564074754714966,
+      "learning_rate": 4.977374404419838e-06,
+      "loss": 0.7031,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2839335180055402,
+      "grad_norm": 0.6523749828338623,
+      "learning_rate": 4.977325478972746e-06,
+      "loss": 0.68,
+      "step": 1025
+    },
+    {
+      "epoch": 0.28421052631578947,
+      "grad_norm": 0.7020553946495056,
+      "learning_rate": 4.977276500925708e-06,
+      "loss": 0.7362,
+      "step": 1026
+    },
+    {
+      "epoch": 0.28448753462603876,
+      "grad_norm": 0.7095143795013428,
+      "learning_rate": 4.977227470279764e-06,
+      "loss": 0.7392,
+      "step": 1027
+    },
+    {
+      "epoch": 0.2847645429362881,
+      "grad_norm": 0.6133536696434021,
+      "learning_rate": 4.977178387035955e-06,
+      "loss": 0.6867,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2850415512465374,
+      "grad_norm": 0.6639204025268555,
+      "learning_rate": 4.977129251195322e-06,
+      "loss": 0.7362,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2853185595567867,
+      "grad_norm": 0.6284003853797913,
+      "learning_rate": 4.97708006275891e-06,
+      "loss": 0.7444,
+      "step": 1030
+    },
+    {
+      "epoch": 0.285595567867036,
+      "grad_norm": 0.6329095363616943,
+      "learning_rate": 4.977030821727762e-06,
+      "loss": 0.6701,
+      "step": 1031
+    },
+    {
+      "epoch": 0.28587257617728534,
+      "grad_norm": 0.6395769715309143,
+      "learning_rate": 4.976981528102924e-06,
+      "loss": 0.7316,
+      "step": 1032
+    },
+    {
+      "epoch": 0.28614958448753464,
+      "grad_norm": 0.6371200084686279,
+      "learning_rate": 4.976932181885443e-06,
+      "loss": 0.7063,
+      "step": 1033
+    },
+    {
+      "epoch": 0.28642659279778393,
+      "grad_norm": 0.6195498704910278,
+      "learning_rate": 4.976882783076366e-06,
+      "loss": 0.6956,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2867036011080332,
+      "grad_norm": 0.6352466940879822,
+      "learning_rate": 4.976833331676742e-06,
+      "loss": 0.756,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2869806094182826,
+      "grad_norm": 0.6565999388694763,
+      "learning_rate": 4.976783827687622e-06,
+      "loss": 0.7226,
+      "step": 1036
+    },
+    {
+      "epoch": 0.28725761772853187,
+      "grad_norm": 0.6376464366912842,
+      "learning_rate": 4.9767342711100555e-06,
+      "loss": 0.7284,
+      "step": 1037
+    },
+    {
+      "epoch": 0.28753462603878116,
+      "grad_norm": 0.6268242001533508,
+      "learning_rate": 4.976684661945096e-06,
+      "loss": 0.7511,
+      "step": 1038
+    },
+    {
+      "epoch": 0.28781163434903045,
+      "grad_norm": 0.610745370388031,
+      "learning_rate": 4.976635000193796e-06,
+      "loss": 0.6825,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2880886426592798,
+      "grad_norm": 0.6473029851913452,
+      "learning_rate": 4.976585285857211e-06,
+      "loss": 0.7226,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2883656509695291,
+      "grad_norm": 0.6430656313896179,
+      "learning_rate": 4.976535518936395e-06,
+      "loss": 0.7128,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2886426592797784,
+      "grad_norm": 0.6196282505989075,
+      "learning_rate": 4.9764856994324065e-06,
+      "loss": 0.7277,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2889196675900277,
+      "grad_norm": 0.6294348835945129,
+      "learning_rate": 4.976435827346301e-06,
+      "loss": 0.66,
+      "step": 1043
+    },
+    {
+      "epoch": 0.28919667590027703,
+      "grad_norm": 0.640680193901062,
+      "learning_rate": 4.976385902679139e-06,
+      "loss": 0.7064,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2894736842105263,
+      "grad_norm": 0.6399988532066345,
+      "learning_rate": 4.976335925431982e-06,
+      "loss": 0.6775,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2897506925207756,
+      "grad_norm": 0.6854329109191895,
+      "learning_rate": 4.976285895605888e-06,
+      "loss": 0.7241,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2900277008310249,
+      "grad_norm": 0.6851531267166138,
+      "learning_rate": 4.976235813201921e-06,
+      "loss": 0.7943,
+      "step": 1047
+    },
+    {
+      "epoch": 0.29030470914127426,
+      "grad_norm": 0.6644576191902161,
+      "learning_rate": 4.976185678221144e-06,
+      "loss": 0.6906,
+      "step": 1048
+    },
+    {
+      "epoch": 0.29058171745152356,
+      "grad_norm": 0.6867406964302063,
+      "learning_rate": 4.976135490664622e-06,
+      "loss": 0.7291,
+      "step": 1049
+    },
+    {
+      "epoch": 0.29085872576177285,
+      "grad_norm": 0.6741674542427063,
+      "learning_rate": 4.976085250533419e-06,
+      "loss": 0.7314,
+      "step": 1050
+    },
+    {
+      "epoch": 0.29113573407202215,
+      "grad_norm": 0.6466678977012634,
+      "learning_rate": 4.976034957828604e-06,
+      "loss": 0.719,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2914127423822715,
+      "grad_norm": 0.6052623987197876,
+      "learning_rate": 4.975984612551243e-06,
+      "loss": 0.6441,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2916897506925208,
+      "grad_norm": 0.62298583984375,
+      "learning_rate": 4.975934214702407e-06,
+      "loss": 0.6747,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2919667590027701,
+      "grad_norm": 0.6360480785369873,
+      "learning_rate": 4.975883764283164e-06,
+      "loss": 0.7081,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2922437673130194,
+      "grad_norm": 0.674088180065155,
+      "learning_rate": 4.9758332612945855e-06,
+      "loss": 0.7371,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2925207756232687,
+      "grad_norm": 0.6503378748893738,
+      "learning_rate": 4.975782705737745e-06,
+      "loss": 0.7175,
+      "step": 1056
+    },
+    {
+      "epoch": 0.292797783933518,
+      "grad_norm": 0.6464011073112488,
+      "learning_rate": 4.975732097613715e-06,
+      "loss": 0.7175,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2930747922437673,
+      "grad_norm": 0.6450788974761963,
+      "learning_rate": 4.97568143692357e-06,
+      "loss": 0.7429,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2933518005540166,
+      "grad_norm": 0.672749936580658,
+      "learning_rate": 4.975630723668385e-06,
+      "loss": 0.7046,
+      "step": 1059
+    },
+    {
+      "epoch": 0.29362880886426596,
+      "grad_norm": 0.638154923915863,
+      "learning_rate": 4.975579957849239e-06,
+      "loss": 0.6932,
+      "step": 1060
+    },
+    {
+      "epoch": 0.29390581717451525,
+      "grad_norm": 0.6246246099472046,
+      "learning_rate": 4.975529139467208e-06,
+      "loss": 0.7035,
+      "step": 1061
+    },
+    {
+      "epoch": 0.29418282548476454,
+      "grad_norm": 0.6619797348976135,
+      "learning_rate": 4.975478268523372e-06,
+      "loss": 0.7464,
+      "step": 1062
+    },
+    {
+      "epoch": 0.29445983379501384,
+      "grad_norm": 0.6233973503112793,
+      "learning_rate": 4.97542734501881e-06,
+      "loss": 0.7053,
+      "step": 1063
+    },
+    {
+      "epoch": 0.29473684210526313,
+      "grad_norm": 0.6285495162010193,
+      "learning_rate": 4.975376368954605e-06,
+      "loss": 0.7718,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2950138504155125,
+      "grad_norm": 0.6749695539474487,
+      "learning_rate": 4.9753253403318375e-06,
+      "loss": 0.7252,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2952908587257618,
+      "grad_norm": 0.6395245790481567,
+      "learning_rate": 4.975274259151592e-06,
+      "loss": 0.6891,
+      "step": 1066
+    },
+    {
+      "epoch": 0.29556786703601107,
+      "grad_norm": 0.6453225612640381,
+      "learning_rate": 4.975223125414952e-06,
+      "loss": 0.7353,
+      "step": 1067
+    },
+    {
+      "epoch": 0.29584487534626036,
+      "grad_norm": 0.6453205943107605,
+      "learning_rate": 4.9751719391230055e-06,
+      "loss": 0.7296,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2961218836565097,
+      "grad_norm": 0.6330751776695251,
+      "learning_rate": 4.975120700276837e-06,
+      "loss": 0.7129,
+      "step": 1069
+    },
+    {
+      "epoch": 0.296398891966759,
+      "grad_norm": 0.6840337514877319,
+      "learning_rate": 4.975069408877535e-06,
+      "loss": 0.7683,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2966759002770083,
+      "grad_norm": 0.6366419196128845,
+      "learning_rate": 4.975018064926189e-06,
+      "loss": 0.682,
+      "step": 1071
+    },
+    {
+      "epoch": 0.2969529085872576,
+      "grad_norm": 0.6742862462997437,
+      "learning_rate": 4.974966668423889e-06,
+      "loss": 0.6991,
+      "step": 1072
+    },
+    {
+      "epoch": 0.29722991689750694,
+      "grad_norm": 0.6317266225814819,
+      "learning_rate": 4.974915219371726e-06,
+      "loss": 0.7467,
+      "step": 1073
+    },
+    {
+      "epoch": 0.29750692520775623,
+      "grad_norm": 0.6813490390777588,
+      "learning_rate": 4.974863717770793e-06,
+      "loss": 0.7729,
+      "step": 1074
+    },
+    {
+      "epoch": 0.29778393351800553,
+      "grad_norm": 0.6082154512405396,
+      "learning_rate": 4.974812163622183e-06,
+      "loss": 0.7586,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2980609418282548,
+      "grad_norm": 0.6370641589164734,
+      "learning_rate": 4.974760556926991e-06,
+      "loss": 0.7545,
+      "step": 1076
+    },
+    {
+      "epoch": 0.29833795013850417,
+      "grad_norm": 0.6440302729606628,
+      "learning_rate": 4.974708897686312e-06,
+      "loss": 0.7222,
+      "step": 1077
+    },
+    {
+      "epoch": 0.29861495844875346,
+      "grad_norm": 0.6516856551170349,
+      "learning_rate": 4.974657185901245e-06,
+      "loss": 0.694,
+      "step": 1078
+    },
+    {
+      "epoch": 0.29889196675900276,
+      "grad_norm": 0.6376709938049316,
+      "learning_rate": 4.974605421572885e-06,
+      "loss": 0.6705,
+      "step": 1079
+    },
+    {
+      "epoch": 0.29916897506925205,
+      "grad_norm": 0.637450098991394,
+      "learning_rate": 4.974553604702332e-06,
+      "loss": 0.7093,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2994459833795014,
+      "grad_norm": 0.6694522500038147,
+      "learning_rate": 4.974501735290688e-06,
+      "loss": 0.7674,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2997229916897507,
+      "grad_norm": 0.6781630516052246,
+      "learning_rate": 4.9744498133390515e-06,
+      "loss": 0.7496,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.647311270236969,
+      "learning_rate": 4.974397838848527e-06,
+      "loss": 0.7152,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3002770083102493,
+      "grad_norm": 0.6689607501029968,
+      "learning_rate": 4.974345811820218e-06,
+      "loss": 0.6752,
+      "step": 1084
+    },
+    {
+      "epoch": 0.30055401662049863,
+      "grad_norm": 0.697877049446106,
+      "learning_rate": 4.9742937322552285e-06,
+      "loss": 0.7165,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3008310249307479,
+      "grad_norm": 0.6767156720161438,
+      "learning_rate": 4.974241600154664e-06,
+      "loss": 0.7574,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3011080332409972,
+      "grad_norm": 0.6254011988639832,
+      "learning_rate": 4.974189415519632e-06,
+      "loss": 0.6666,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3013850415512465,
+      "grad_norm": 0.6448764204978943,
+      "learning_rate": 4.9741371783512395e-06,
+      "loss": 0.6697,
+      "step": 1088
+    },
+    {
+      "epoch": 0.30166204986149586,
+      "grad_norm": 0.6668729186058044,
+      "learning_rate": 4.974084888650597e-06,
+      "loss": 0.7035,
+      "step": 1089
+    },
+    {
+      "epoch": 0.30193905817174516,
+      "grad_norm": 0.6547879576683044,
+      "learning_rate": 4.974032546418816e-06,
+      "loss": 0.7492,
+      "step": 1090
+    },
+    {
+      "epoch": 0.30221606648199445,
+      "grad_norm": 0.646012544631958,
+      "learning_rate": 4.9739801516570035e-06,
+      "loss": 0.6893,
+      "step": 1091
+    },
+    {
+      "epoch": 0.30249307479224374,
+      "grad_norm": 0.6936621069908142,
+      "learning_rate": 4.973927704366275e-06,
+      "loss": 0.783,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3027700831024931,
+      "grad_norm": 0.5984358787536621,
+      "learning_rate": 4.973875204547744e-06,
+      "loss": 0.6532,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3030470914127424,
+      "grad_norm": 0.6124839782714844,
+      "learning_rate": 4.973822652202526e-06,
+      "loss": 0.6487,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3033240997229917,
+      "grad_norm": 0.6612206101417542,
+      "learning_rate": 4.973770047331734e-06,
+      "loss": 0.7466,
+      "step": 1095
+    },
+    {
+      "epoch": 0.303601108033241,
+      "grad_norm": 0.6440140008926392,
+      "learning_rate": 4.973717389936487e-06,
+      "loss": 0.6941,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3038781163434903,
+      "grad_norm": 0.6640320420265198,
+      "learning_rate": 4.973664680017903e-06,
+      "loss": 0.7211,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3041551246537396,
+      "grad_norm": 0.6480273604393005,
+      "learning_rate": 4.9736119175771005e-06,
+      "loss": 0.7297,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3044321329639889,
+      "grad_norm": 0.6550040245056152,
+      "learning_rate": 4.973559102615199e-06,
+      "loss": 0.6963,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3047091412742382,
+      "grad_norm": 0.6494340896606445,
+      "learning_rate": 4.973506235133323e-06,
+      "loss": 0.7035,
+      "step": 1100
+    },
+    {
+      "epoch": 0.30498614958448755,
+      "grad_norm": 0.6211956143379211,
+      "learning_rate": 4.973453315132592e-06,
+      "loss": 0.6444,
+      "step": 1101
+    },
+    {
+      "epoch": 0.30526315789473685,
+      "grad_norm": 0.6703330874443054,
+      "learning_rate": 4.9734003426141305e-06,
+      "loss": 0.6802,
+      "step": 1102
+    },
+    {
+      "epoch": 0.30554016620498614,
+      "grad_norm": 0.6802040934562683,
+      "learning_rate": 4.973347317579064e-06,
+      "loss": 0.6729,
+      "step": 1103
+    },
+    {
+      "epoch": 0.30581717451523543,
+      "grad_norm": 0.6808921098709106,
+      "learning_rate": 4.973294240028517e-06,
+      "loss": 0.6976,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3060941828254848,
+      "grad_norm": 0.6503647565841675,
+      "learning_rate": 4.973241109963618e-06,
+      "loss": 0.7282,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3063711911357341,
+      "grad_norm": 0.6694557070732117,
+      "learning_rate": 4.973187927385493e-06,
+      "loss": 0.716,
+      "step": 1106
+    },
+    {
+      "epoch": 0.30664819944598337,
+      "grad_norm": 0.6967760920524597,
+      "learning_rate": 4.973134692295273e-06,
+      "loss": 0.7956,
+      "step": 1107
+    },
+    {
+      "epoch": 0.30692520775623267,
+      "grad_norm": 0.6802443861961365,
+      "learning_rate": 4.973081404694088e-06,
+      "loss": 0.7406,
+      "step": 1108
+    },
+    {
+      "epoch": 0.307202216066482,
+      "grad_norm": 0.6713878512382507,
+      "learning_rate": 4.973028064583069e-06,
+      "loss": 0.7273,
+      "step": 1109
+    },
+    {
+      "epoch": 0.3074792243767313,
+      "grad_norm": 0.6317265629768372,
+      "learning_rate": 4.972974671963349e-06,
+      "loss": 0.7583,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3077562326869806,
+      "grad_norm": 0.6492122411727905,
+      "learning_rate": 4.972921226836061e-06,
+      "loss": 0.7243,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3080332409972299,
+      "grad_norm": 0.630206286907196,
+      "learning_rate": 4.9728677292023405e-06,
+      "loss": 0.6688,
+      "step": 1112
+    },
+    {
+      "epoch": 0.30831024930747924,
+      "grad_norm": 0.69725102186203,
+      "learning_rate": 4.972814179063323e-06,
+      "loss": 0.7342,
+      "step": 1113
+    },
+    {
+      "epoch": 0.30858725761772854,
+      "grad_norm": 0.68944251537323,
+      "learning_rate": 4.9727605764201455e-06,
+      "loss": 0.7534,
+      "step": 1114
+    },
+    {
+      "epoch": 0.30886426592797783,
+      "grad_norm": 0.6732880473136902,
+      "learning_rate": 4.972706921273947e-06,
+      "loss": 0.7125,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3091412742382271,
+      "grad_norm": 0.6394022703170776,
+      "learning_rate": 4.972653213625865e-06,
+      "loss": 0.7135,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3094182825484765,
+      "grad_norm": 0.6659450531005859,
+      "learning_rate": 4.972599453477041e-06,
+      "loss": 0.7337,
+      "step": 1117
+    },
+    {
+      "epoch": 0.30969529085872577,
+      "grad_norm": 0.6496633887290955,
+      "learning_rate": 4.972545640828617e-06,
+      "loss": 0.7601,
+      "step": 1118
+    },
+    {
+      "epoch": 0.30997229916897506,
+      "grad_norm": 0.6592468619346619,
+      "learning_rate": 4.972491775681735e-06,
+      "loss": 0.7553,
+      "step": 1119
+    },
+    {
+      "epoch": 0.31024930747922436,
+      "grad_norm": 0.6824886202812195,
+      "learning_rate": 4.972437858037538e-06,
+      "loss": 0.7442,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3105263157894737,
+      "grad_norm": 0.7013871073722839,
+      "learning_rate": 4.972383887897171e-06,
+      "loss": 0.7405,
+      "step": 1121
+    },
+    {
+      "epoch": 0.310803324099723,
+      "grad_norm": 0.7087879180908203,
+      "learning_rate": 4.97232986526178e-06,
+      "loss": 0.7416,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3110803324099723,
+      "grad_norm": 0.6554746031761169,
+      "learning_rate": 4.972275790132513e-06,
+      "loss": 0.7197,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3113573407202216,
+      "grad_norm": 0.6627053022384644,
+      "learning_rate": 4.972221662510517e-06,
+      "loss": 0.687,
+      "step": 1124
+    },
+    {
+      "epoch": 0.31163434903047094,
+      "grad_norm": 0.6275670528411865,
+      "learning_rate": 4.9721674823969425e-06,
+      "loss": 0.7252,
+      "step": 1125
+    },
+    {
+      "epoch": 0.31191135734072023,
+      "grad_norm": 0.6724969744682312,
+      "learning_rate": 4.972113249792938e-06,
+      "loss": 0.6879,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3121883656509695,
+      "grad_norm": 0.6502825021743774,
+      "learning_rate": 4.972058964699658e-06,
+      "loss": 0.7256,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3124653739612188,
+      "grad_norm": 0.6835956573486328,
+      "learning_rate": 4.972004627118251e-06,
+      "loss": 0.6942,
+      "step": 1128
+    },
+    {
+      "epoch": 0.31274238227146817,
+      "grad_norm": 0.6625363230705261,
+      "learning_rate": 4.971950237049874e-06,
+      "loss": 0.7493,
+      "step": 1129
+    },
+    {
+      "epoch": 0.31301939058171746,
+      "grad_norm": 0.6615988612174988,
+      "learning_rate": 4.971895794495681e-06,
+      "loss": 0.6836,
+      "step": 1130
+    },
+    {
+      "epoch": 0.31329639889196675,
+      "grad_norm": 0.6387788653373718,
+      "learning_rate": 4.971841299456827e-06,
+      "loss": 0.7134,
+      "step": 1131
+    },
+    {
+      "epoch": 0.31357340720221605,
+      "grad_norm": 0.628496527671814,
+      "learning_rate": 4.971786751934471e-06,
+      "loss": 0.7054,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3138504155124654,
+      "grad_norm": 0.6888827681541443,
+      "learning_rate": 4.971732151929769e-06,
+      "loss": 0.7168,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3141274238227147,
+      "grad_norm": 0.67254638671875,
+      "learning_rate": 4.971677499443882e-06,
+      "loss": 0.7034,
+      "step": 1134
+    },
+    {
+      "epoch": 0.314404432132964,
+      "grad_norm": 0.6414274573326111,
+      "learning_rate": 4.971622794477969e-06,
+      "loss": 0.6905,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3146814404432133,
+      "grad_norm": 0.6497774720191956,
+      "learning_rate": 4.971568037033193e-06,
+      "loss": 0.6351,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3149584487534626,
+      "grad_norm": 0.6576498746871948,
+      "learning_rate": 4.971513227110714e-06,
+      "loss": 0.72,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3152354570637119,
+      "grad_norm": 0.6286632418632507,
+      "learning_rate": 4.9714583647117e-06,
+      "loss": 0.6696,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3155124653739612,
+      "grad_norm": 0.6436337828636169,
+      "learning_rate": 4.971403449837313e-06,
+      "loss": 0.6442,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.650290310382843,
+      "learning_rate": 4.971348482488719e-06,
+      "loss": 0.7054,
+      "step": 1140
+    },
+    {
+      "epoch": 0.31606648199445986,
+      "grad_norm": 0.6691767573356628,
+      "learning_rate": 4.971293462667085e-06,
+      "loss": 0.7066,
+      "step": 1141
+    },
+    {
+      "epoch": 0.31634349030470915,
+      "grad_norm": 0.6176061034202576,
+      "learning_rate": 4.971238390373581e-06,
+      "loss": 0.6673,
+      "step": 1142
+    },
+    {
+      "epoch": 0.31662049861495845,
+      "grad_norm": 0.6956735253334045,
+      "learning_rate": 4.971183265609374e-06,
+      "loss": 0.7598,
+      "step": 1143
+    },
+    {
+      "epoch": 0.31689750692520774,
+      "grad_norm": 0.6023033261299133,
+      "learning_rate": 4.971128088375636e-06,
+      "loss": 0.6681,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3171745152354571,
+      "grad_norm": 0.6559221744537354,
+      "learning_rate": 4.971072858673539e-06,
+      "loss": 0.6867,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3174515235457064,
+      "grad_norm": 0.6444343328475952,
+      "learning_rate": 4.971017576504254e-06,
+      "loss": 0.7175,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3177285318559557,
+      "grad_norm": 0.7063049077987671,
+      "learning_rate": 4.970962241868955e-06,
+      "loss": 0.6948,
+      "step": 1147
+    },
+    {
+      "epoch": 0.31800554016620497,
+      "grad_norm": 0.6646923422813416,
+      "learning_rate": 4.9709068547688186e-06,
+      "loss": 0.7128,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3182825484764543,
+      "grad_norm": 0.6788060665130615,
+      "learning_rate": 4.970851415205018e-06,
+      "loss": 0.7153,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3185595567867036,
+      "grad_norm": 0.6538786292076111,
+      "learning_rate": 4.970795923178734e-06,
+      "loss": 0.7012,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3188365650969529,
+      "grad_norm": 0.6571323871612549,
+      "learning_rate": 4.970740378691141e-06,
+      "loss": 0.7051,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3191135734072022,
+      "grad_norm": 0.623737096786499,
+      "learning_rate": 4.970684781743422e-06,
+      "loss": 0.6843,
+      "step": 1152
+    },
+    {
+      "epoch": 0.31939058171745155,
+      "grad_norm": 0.6650036573410034,
+      "learning_rate": 4.970629132336755e-06,
+      "loss": 0.7052,
+      "step": 1153
+    },
+    {
+      "epoch": 0.31966759002770084,
+      "grad_norm": 0.6788172721862793,
+      "learning_rate": 4.9705734304723215e-06,
+      "loss": 0.679,
+      "step": 1154
+    },
+    {
+      "epoch": 0.31994459833795014,
+      "grad_norm": 0.7293535470962524,
+      "learning_rate": 4.970517676151305e-06,
+      "loss": 0.7561,
+      "step": 1155
+    },
+    {
+      "epoch": 0.32022160664819943,
+      "grad_norm": 0.7336902022361755,
+      "learning_rate": 4.97046186937489e-06,
+      "loss": 0.7271,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3204986149584488,
+      "grad_norm": 0.676217257976532,
+      "learning_rate": 4.970406010144259e-06,
+      "loss": 0.7176,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3207756232686981,
+      "grad_norm": 0.6698077321052551,
+      "learning_rate": 4.970350098460602e-06,
+      "loss": 0.7252,
+      "step": 1158
+    },
+    {
+      "epoch": 0.32105263157894737,
+      "grad_norm": 0.6944154500961304,
+      "learning_rate": 4.970294134325102e-06,
+      "loss": 0.7368,
+      "step": 1159
+    },
+    {
+      "epoch": 0.32132963988919666,
+      "grad_norm": 0.6670475006103516,
+      "learning_rate": 4.97023811773895e-06,
+      "loss": 0.7018,
+      "step": 1160
+    },
+    {
+      "epoch": 0.321606648199446,
+      "grad_norm": 0.6607210040092468,
+      "learning_rate": 4.970182048703335e-06,
+      "loss": 0.7113,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3218836565096953,
+      "grad_norm": 0.7043123245239258,
+      "learning_rate": 4.970125927219446e-06,
+      "loss": 0.6951,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3221606648199446,
+      "grad_norm": 0.6905112266540527,
+      "learning_rate": 4.970069753288476e-06,
+      "loss": 0.7125,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3224376731301939,
+      "grad_norm": 0.6552289724349976,
+      "learning_rate": 4.970013526911617e-06,
+      "loss": 0.7344,
+      "step": 1164
+    },
+    {
+      "epoch": 0.32271468144044324,
+      "grad_norm": 0.6786649227142334,
+      "learning_rate": 4.969957248090062e-06,
+      "loss": 0.7108,
+      "step": 1165
+    },
+    {
+      "epoch": 0.32299168975069253,
+      "grad_norm": 0.6703157424926758,
+      "learning_rate": 4.969900916825009e-06,
+      "loss": 0.7072,
+      "step": 1166
+    },
+    {
+      "epoch": 0.32326869806094183,
+      "grad_norm": 0.6406513452529907,
+      "learning_rate": 4.969844533117651e-06,
+      "loss": 0.6988,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3235457063711911,
+      "grad_norm": 0.6920865178108215,
+      "learning_rate": 4.969788096969186e-06,
+      "loss": 0.7192,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3238227146814404,
+      "grad_norm": 0.7104857563972473,
+      "learning_rate": 4.969731608380814e-06,
+      "loss": 0.7087,
+      "step": 1169
+    },
+    {
+      "epoch": 0.32409972299168976,
+      "grad_norm": 0.6691415905952454,
+      "learning_rate": 4.9696750673537315e-06,
+      "loss": 0.6976,
+      "step": 1170
+    },
+    {
+      "epoch": 0.32437673130193906,
+      "grad_norm": 0.7121404409408569,
+      "learning_rate": 4.969618473889141e-06,
+      "loss": 0.7591,
+      "step": 1171
+    },
+    {
+      "epoch": 0.32465373961218835,
+      "grad_norm": 0.655210554599762,
+      "learning_rate": 4.969561827988244e-06,
+      "loss": 0.7052,
+      "step": 1172
+    },
+    {
+      "epoch": 0.32493074792243765,
+      "grad_norm": 0.6501975059509277,
+      "learning_rate": 4.969505129652242e-06,
+      "loss": 0.735,
+      "step": 1173
+    },
+    {
+      "epoch": 0.325207756232687,
+      "grad_norm": 0.639132559299469,
+      "learning_rate": 4.96944837888234e-06,
+      "loss": 0.7081,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3254847645429363,
+      "grad_norm": 0.6479787826538086,
+      "learning_rate": 4.969391575679743e-06,
+      "loss": 0.7136,
+      "step": 1175
+    },
+    {
+      "epoch": 0.3257617728531856,
+      "grad_norm": 0.6485690474510193,
+      "learning_rate": 4.969334720045656e-06,
+      "loss": 0.6554,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3260387811634349,
+      "grad_norm": 0.6593156456947327,
+      "learning_rate": 4.969277811981287e-06,
+      "loss": 0.6959,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3263157894736842,
+      "grad_norm": 0.6410382986068726,
+      "learning_rate": 4.9692208514878445e-06,
+      "loss": 0.6804,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3265927977839335,
+      "grad_norm": 0.6194038391113281,
+      "learning_rate": 4.969163838566537e-06,
+      "loss": 0.6745,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3268698060941828,
+      "grad_norm": 0.6443800926208496,
+      "learning_rate": 4.969106773218577e-06,
+      "loss": 0.7266,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3271468144044321,
+      "grad_norm": 0.6541767120361328,
+      "learning_rate": 4.969049655445174e-06,
+      "loss": 0.7135,
+      "step": 1181
+    },
+    {
+      "epoch": 0.32742382271468146,
+      "grad_norm": 0.6695902347564697,
+      "learning_rate": 4.968992485247541e-06,
+      "loss": 0.7551,
+      "step": 1182
+    },
+    {
+      "epoch": 0.32770083102493075,
+      "grad_norm": 0.6857314109802246,
+      "learning_rate": 4.9689352626268935e-06,
+      "loss": 0.7272,
+      "step": 1183
+    },
+    {
+      "epoch": 0.32797783933518004,
+      "grad_norm": 0.6714820861816406,
+      "learning_rate": 4.968877987584445e-06,
+      "loss": 0.7372,
+      "step": 1184
+    },
+    {
+      "epoch": 0.32825484764542934,
+      "grad_norm": 0.7039293646812439,
+      "learning_rate": 4.968820660121412e-06,
+      "loss": 0.7096,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3285318559556787,
+      "grad_norm": 0.658528745174408,
+      "learning_rate": 4.968763280239012e-06,
+      "loss": 0.7096,
+      "step": 1186
+    },
+    {
+      "epoch": 0.328808864265928,
+      "grad_norm": 0.6193217635154724,
+      "learning_rate": 4.968705847938462e-06,
+      "loss": 0.6464,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3290858725761773,
+      "grad_norm": 0.6533862352371216,
+      "learning_rate": 4.968648363220984e-06,
+      "loss": 0.7257,
+      "step": 1188
+    },
+    {
+      "epoch": 0.32936288088642657,
+      "grad_norm": 0.6659977436065674,
+      "learning_rate": 4.9685908260877955e-06,
+      "loss": 0.7223,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3296398891966759,
+      "grad_norm": 0.6677512526512146,
+      "learning_rate": 4.9685332365401205e-06,
+      "loss": 0.7226,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3299168975069252,
+      "grad_norm": 0.6727151274681091,
+      "learning_rate": 4.968475594579181e-06,
+      "loss": 0.6983,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3301939058171745,
+      "grad_norm": 0.6492422223091125,
+      "learning_rate": 4.9684179002062e-06,
+      "loss": 0.6865,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3304709141274238,
+      "grad_norm": 0.6984854936599731,
+      "learning_rate": 4.968360153422404e-06,
+      "loss": 0.7665,
+      "step": 1193
+    },
+    {
+      "epoch": 0.33074792243767315,
+      "grad_norm": 0.6713538765907288,
+      "learning_rate": 4.9683023542290184e-06,
+      "loss": 0.7214,
+      "step": 1194
+    },
+    {
+      "epoch": 0.33102493074792244,
+      "grad_norm": 0.685975968837738,
+      "learning_rate": 4.96824450262727e-06,
+      "loss": 0.7515,
+      "step": 1195
+    },
+    {
+      "epoch": 0.33130193905817173,
+      "grad_norm": 0.6598302125930786,
+      "learning_rate": 4.968186598618388e-06,
+      "loss": 0.6666,
+      "step": 1196
+    },
+    {
+      "epoch": 0.33157894736842103,
+      "grad_norm": 0.6653509736061096,
+      "learning_rate": 4.968128642203602e-06,
+      "loss": 0.7141,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3318559556786704,
+      "grad_norm": 0.649482011795044,
+      "learning_rate": 4.968070633384141e-06,
+      "loss": 0.6893,
+      "step": 1198
+    },
+    {
+      "epoch": 0.33213296398891967,
+      "grad_norm": 0.6657583117485046,
+      "learning_rate": 4.968012572161239e-06,
+      "loss": 0.7252,
+      "step": 1199
+    },
+    {
+      "epoch": 0.33240997229916897,
+      "grad_norm": 0.659673273563385,
+      "learning_rate": 4.967954458536126e-06,
+      "loss": 0.7276,
+      "step": 1200
+    },
+    {
+      "epoch": 0.33268698060941826,
+      "grad_norm": 0.646888792514801,
+      "learning_rate": 4.9678962925100375e-06,
+      "loss": 0.6894,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3329639889196676,
+      "grad_norm": 0.6243681311607361,
+      "learning_rate": 4.967838074084209e-06,
+      "loss": 0.6723,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3332409972299169,
+      "grad_norm": 0.6913838386535645,
+      "learning_rate": 4.9677798032598754e-06,
+      "loss": 0.7173,
+      "step": 1203
+    },
+    {
+      "epoch": 0.3335180055401662,
+      "grad_norm": 0.7205289602279663,
+      "learning_rate": 4.967721480038276e-06,
+      "loss": 0.7796,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3337950138504155,
+      "grad_norm": 0.6805700063705444,
+      "learning_rate": 4.967663104420647e-06,
+      "loss": 0.724,
+      "step": 1205
+    },
+    {
+      "epoch": 0.33407202216066484,
+      "grad_norm": 0.6902198791503906,
+      "learning_rate": 4.967604676408228e-06,
+      "loss": 0.7242,
+      "step": 1206
+    },
+    {
+      "epoch": 0.33434903047091413,
+      "grad_norm": 0.7033138275146484,
+      "learning_rate": 4.96754619600226e-06,
+      "loss": 0.7286,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3346260387811634,
+      "grad_norm": 0.6692147254943848,
+      "learning_rate": 4.967487663203985e-06,
+      "loss": 0.7379,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3349030470914127,
+      "grad_norm": 0.6471096277236938,
+      "learning_rate": 4.967429078014646e-06,
+      "loss": 0.6782,
+      "step": 1209
+    },
+    {
+      "epoch": 0.33518005540166207,
+      "grad_norm": 0.6385185718536377,
+      "learning_rate": 4.967370440435486e-06,
+      "loss": 0.7313,
+      "step": 1210
+    },
+    {
+      "epoch": 0.33545706371191136,
+      "grad_norm": 0.6591436266899109,
+      "learning_rate": 4.967311750467751e-06,
+      "loss": 0.709,
+      "step": 1211
+    },
+    {
+      "epoch": 0.33573407202216066,
+      "grad_norm": 0.6667162179946899,
+      "learning_rate": 4.967253008112688e-06,
+      "loss": 0.7272,
+      "step": 1212
+    },
+    {
+      "epoch": 0.33601108033240995,
+      "grad_norm": 0.690365195274353,
+      "learning_rate": 4.967194213371541e-06,
+      "loss": 0.7758,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3362880886426593,
+      "grad_norm": 0.658421516418457,
+      "learning_rate": 4.967135366245562e-06,
+      "loss": 0.6383,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3365650969529086,
+      "grad_norm": 0.6660083532333374,
+      "learning_rate": 4.967076466735997e-06,
+      "loss": 0.6689,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3368421052631579,
+      "grad_norm": 0.631989598274231,
+      "learning_rate": 4.9670175148440995e-06,
+      "loss": 0.6726,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3371191135734072,
+      "grad_norm": 0.6711675524711609,
+      "learning_rate": 4.966958510571119e-06,
+      "loss": 0.6839,
+      "step": 1217
+    },
+    {
+      "epoch": 0.33739612188365653,
+      "grad_norm": 0.7011405229568481,
+      "learning_rate": 4.966899453918311e-06,
+      "loss": 0.7689,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3376731301939058,
+      "grad_norm": 0.6415798664093018,
+      "learning_rate": 4.966840344886927e-06,
+      "loss": 0.6256,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3379501385041551,
+      "grad_norm": 0.672176718711853,
+      "learning_rate": 4.966781183478223e-06,
+      "loss": 0.7184,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3382271468144044,
+      "grad_norm": 0.6688186526298523,
+      "learning_rate": 4.966721969693455e-06,
+      "loss": 0.7292,
+      "step": 1221
+    },
+    {
+      "epoch": 0.33850415512465376,
+      "grad_norm": 0.6519632935523987,
+      "learning_rate": 4.96666270353388e-06,
+      "loss": 0.6372,
+      "step": 1222
+    },
+    {
+      "epoch": 0.33878116343490305,
+      "grad_norm": 0.6690004467964172,
+      "learning_rate": 4.966603385000756e-06,
+      "loss": 0.6819,
+      "step": 1223
+    },
+    {
+      "epoch": 0.33905817174515235,
+      "grad_norm": 0.674385130405426,
+      "learning_rate": 4.966544014095345e-06,
+      "loss": 0.6891,
+      "step": 1224
+    },
+    {
+      "epoch": 0.33933518005540164,
+      "grad_norm": 0.6642457246780396,
+      "learning_rate": 4.966484590818904e-06,
+      "loss": 0.6747,
+      "step": 1225
+    },
+    {
+      "epoch": 0.339612188365651,
+      "grad_norm": 0.6616514325141907,
+      "learning_rate": 4.9664251151726975e-06,
+      "loss": 0.7419,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3398891966759003,
+      "grad_norm": 0.6353458762168884,
+      "learning_rate": 4.966365587157986e-06,
+      "loss": 0.6848,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3401662049861496,
+      "grad_norm": 0.6132497191429138,
+      "learning_rate": 4.966306006776036e-06,
+      "loss": 0.6811,
+      "step": 1228
+    },
+    {
+      "epoch": 0.34044321329639887,
+      "grad_norm": 0.6806102395057678,
+      "learning_rate": 4.96624637402811e-06,
+      "loss": 0.6822,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3407202216066482,
+      "grad_norm": 0.6873536109924316,
+      "learning_rate": 4.9661866889154766e-06,
+      "loss": 0.7172,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3409972299168975,
+      "grad_norm": 0.6916210055351257,
+      "learning_rate": 4.966126951439401e-06,
+      "loss": 0.7106,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3412742382271468,
+      "grad_norm": 0.6565237641334534,
+      "learning_rate": 4.966067161601152e-06,
+      "loss": 0.7058,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3415512465373961,
+      "grad_norm": 0.6750245690345764,
+      "learning_rate": 4.9660073194020005e-06,
+      "loss": 0.7502,
+      "step": 1233
+    },
+    {
+      "epoch": 0.34182825484764545,
+      "grad_norm": 0.6854074001312256,
+      "learning_rate": 4.965947424843216e-06,
+      "loss": 0.7278,
+      "step": 1234
+    },
+    {
+      "epoch": 0.34210526315789475,
+      "grad_norm": 0.6503132581710815,
+      "learning_rate": 4.96588747792607e-06,
+      "loss": 0.7383,
+      "step": 1235
+    },
+    {
+      "epoch": 0.34238227146814404,
+      "grad_norm": 0.6417048573493958,
+      "learning_rate": 4.965827478651836e-06,
+      "loss": 0.6405,
+      "step": 1236
+    },
+    {
+      "epoch": 0.34265927977839333,
+      "grad_norm": 0.6622879505157471,
+      "learning_rate": 4.965767427021787e-06,
+      "loss": 0.6652,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3429362880886427,
+      "grad_norm": 0.7433401346206665,
+      "learning_rate": 4.965707323037199e-06,
+      "loss": 0.7381,
+      "step": 1238
+    },
+    {
+      "epoch": 0.343213296398892,
+      "grad_norm": 0.7252357602119446,
+      "learning_rate": 4.965647166699348e-06,
+      "loss": 0.7311,
+      "step": 1239
+    },
+    {
+      "epoch": 0.34349030470914127,
+      "grad_norm": 0.7050260901451111,
+      "learning_rate": 4.965586958009511e-06,
+      "loss": 0.7243,
+      "step": 1240
+    },
+    {
+      "epoch": 0.34376731301939056,
+      "grad_norm": 0.6850735545158386,
+      "learning_rate": 4.965526696968966e-06,
+      "loss": 0.6879,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3440443213296399,
+      "grad_norm": 0.673936128616333,
+      "learning_rate": 4.965466383578994e-06,
+      "loss": 0.718,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3443213296398892,
+      "grad_norm": 0.6669742465019226,
+      "learning_rate": 4.965406017840873e-06,
+      "loss": 0.6727,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3445983379501385,
+      "grad_norm": 0.6629636287689209,
+      "learning_rate": 4.965345599755888e-06,
+      "loss": 0.7094,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3448753462603878,
+      "grad_norm": 0.647504448890686,
+      "learning_rate": 4.965285129325319e-06,
+      "loss": 0.6986,
+      "step": 1245
+    },
+    {
+      "epoch": 0.34515235457063714,
+      "grad_norm": 0.6685324311256409,
+      "learning_rate": 4.965224606550452e-06,
+      "loss": 0.6984,
+      "step": 1246
+    },
+    {
+      "epoch": 0.34542936288088644,
+      "grad_norm": 0.6271531581878662,
+      "learning_rate": 4.96516403143257e-06,
+      "loss": 0.6799,
+      "step": 1247
+    },
+    {
+      "epoch": 0.34570637119113573,
+      "grad_norm": 0.6847185492515564,
+      "learning_rate": 4.96510340397296e-06,
+      "loss": 0.7843,
+      "step": 1248
+    },
+    {
+      "epoch": 0.345983379501385,
+      "grad_norm": 0.6534786224365234,
+      "learning_rate": 4.96504272417291e-06,
+      "loss": 0.6948,
+      "step": 1249
+    },
+    {
+      "epoch": 0.3462603878116344,
+      "grad_norm": 0.6473841667175293,
+      "learning_rate": 4.964981992033709e-06,
+      "loss": 0.6932,
+      "step": 1250
+    },
+    {
+      "epoch": 0.34653739612188367,
+      "grad_norm": 0.657150149345398,
+      "learning_rate": 4.964921207556645e-06,
+      "loss": 0.7882,
+      "step": 1251
+    },
+    {
+      "epoch": 0.34681440443213296,
+      "grad_norm": 0.6625207662582397,
+      "learning_rate": 4.9648603707430085e-06,
+      "loss": 0.742,
+      "step": 1252
+    },
+    {
+      "epoch": 0.34709141274238225,
+      "grad_norm": 0.6801496148109436,
+      "learning_rate": 4.9647994815940916e-06,
+      "loss": 0.6557,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3473684210526316,
+      "grad_norm": 0.6741940379142761,
+      "learning_rate": 4.964738540111187e-06,
+      "loss": 0.6849,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3476454293628809,
+      "grad_norm": 0.6735914349555969,
+      "learning_rate": 4.96467754629559e-06,
+      "loss": 0.6854,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3479224376731302,
+      "grad_norm": 0.6577902436256409,
+      "learning_rate": 4.964616500148594e-06,
+      "loss": 0.7293,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3481994459833795,
+      "grad_norm": 0.6584300398826599,
+      "learning_rate": 4.964555401671496e-06,
+      "loss": 0.6902,
+      "step": 1257
+    },
+    {
+      "epoch": 0.34847645429362883,
+      "grad_norm": 0.613704264163971,
+      "learning_rate": 4.964494250865593e-06,
+      "loss": 0.6745,
+      "step": 1258
+    },
+    {
+      "epoch": 0.34875346260387813,
+      "grad_norm": 0.6942754983901978,
+      "learning_rate": 4.964433047732183e-06,
+      "loss": 0.7453,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3490304709141274,
+      "grad_norm": 0.6607272624969482,
+      "learning_rate": 4.964371792272566e-06,
+      "loss": 0.7024,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3493074792243767,
+      "grad_norm": 0.6645615100860596,
+      "learning_rate": 4.964310484488044e-06,
+      "loss": 0.7461,
+      "step": 1261
+    },
+    {
+      "epoch": 0.34958448753462606,
+      "grad_norm": 0.6562692523002625,
+      "learning_rate": 4.964249124379915e-06,
+      "loss": 0.6633,
+      "step": 1262
+    },
+    {
+      "epoch": 0.34986149584487536,
+      "grad_norm": 0.6694803833961487,
+      "learning_rate": 4.964187711949486e-06,
+      "loss": 0.6676,
+      "step": 1263
+    },
+    {
+      "epoch": 0.35013850415512465,
+      "grad_norm": 0.6911068558692932,
+      "learning_rate": 4.964126247198058e-06,
+      "loss": 0.7291,
+      "step": 1264
+    },
+    {
+      "epoch": 0.35041551246537395,
+      "grad_norm": 0.6374126076698303,
+      "learning_rate": 4.964064730126937e-06,
+      "loss": 0.766,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3506925207756233,
+      "grad_norm": 0.6650824546813965,
+      "learning_rate": 4.964003160737429e-06,
+      "loss": 0.7294,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3509695290858726,
+      "grad_norm": 0.6754046678543091,
+      "learning_rate": 4.963941539030842e-06,
+      "loss": 0.6993,
+      "step": 1267
+    },
+    {
+      "epoch": 0.3512465373961219,
+      "grad_norm": 0.6406125426292419,
+      "learning_rate": 4.963879865008484e-06,
+      "loss": 0.6803,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3515235457063712,
+      "grad_norm": 0.6519691944122314,
+      "learning_rate": 4.963818138671664e-06,
+      "loss": 0.7288,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3518005540166205,
+      "grad_norm": 0.6560389399528503,
+      "learning_rate": 4.963756360021694e-06,
+      "loss": 0.7381,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3520775623268698,
+      "grad_norm": 0.6318975687026978,
+      "learning_rate": 4.963694529059884e-06,
+      "loss": 0.7164,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3523545706371191,
+      "grad_norm": 0.6151561737060547,
+      "learning_rate": 4.9636326457875475e-06,
+      "loss": 0.6685,
+      "step": 1272
+    },
+    {
+      "epoch": 0.3526315789473684,
+      "grad_norm": 0.6672805547714233,
+      "learning_rate": 4.9635707102059985e-06,
+      "loss": 0.7358,
+      "step": 1273
+    },
+    {
+      "epoch": 0.35290858725761776,
+      "grad_norm": 0.6632043123245239,
+      "learning_rate": 4.963508722316552e-06,
+      "loss": 0.6698,
+      "step": 1274
+    },
+    {
+      "epoch": 0.35318559556786705,
+      "grad_norm": 0.6473809480667114,
+      "learning_rate": 4.963446682120525e-06,
+      "loss": 0.6826,
+      "step": 1275
+    },
+    {
+      "epoch": 0.35346260387811634,
+      "grad_norm": 0.6828525066375732,
+      "learning_rate": 4.963384589619233e-06,
+      "loss": 0.7255,
+      "step": 1276
+    },
+    {
+      "epoch": 0.35373961218836564,
+      "grad_norm": 0.6744553446769714,
+      "learning_rate": 4.963322444813995e-06,
+      "loss": 0.6756,
+      "step": 1277
+    },
+    {
+      "epoch": 0.35401662049861493,
+      "grad_norm": 0.6551529765129089,
+      "learning_rate": 4.963260247706132e-06,
+      "loss": 0.6939,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3542936288088643,
+      "grad_norm": 0.6692739129066467,
+      "learning_rate": 4.963197998296964e-06,
+      "loss": 0.7031,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3545706371191136,
+      "grad_norm": 0.6385998129844666,
+      "learning_rate": 4.96313569658781e-06,
+      "loss": 0.7017,
+      "step": 1280
+    },
+    {
+      "epoch": 0.35484764542936287,
+      "grad_norm": 0.6609788537025452,
+      "learning_rate": 4.9630733425799965e-06,
+      "loss": 0.6635,
+      "step": 1281
+    },
+    {
+      "epoch": 0.35512465373961216,
+      "grad_norm": 0.6580247282981873,
+      "learning_rate": 4.963010936274846e-06,
+      "loss": 0.722,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3554016620498615,
+      "grad_norm": 0.6596221923828125,
+      "learning_rate": 4.9629484776736825e-06,
+      "loss": 0.6768,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3556786703601108,
+      "grad_norm": 0.6421349048614502,
+      "learning_rate": 4.962885966777834e-06,
+      "loss": 0.6835,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3559556786703601,
+      "grad_norm": 0.6218045949935913,
+      "learning_rate": 4.9628234035886265e-06,
+      "loss": 0.6314,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3562326869806094,
+      "grad_norm": 0.6451228260993958,
+      "learning_rate": 4.962760788107389e-06,
+      "loss": 0.6934,
+      "step": 1286
+    },
+    {
+      "epoch": 0.35650969529085874,
+      "grad_norm": 0.6965139508247375,
+      "learning_rate": 4.96269812033545e-06,
+      "loss": 0.7118,
+      "step": 1287
+    },
+    {
+      "epoch": 0.35678670360110804,
+      "grad_norm": 0.6180694699287415,
+      "learning_rate": 4.9626354002741424e-06,
+      "loss": 0.6706,
+      "step": 1288
+    },
+    {
+      "epoch": 0.35706371191135733,
+      "grad_norm": 0.6663467884063721,
+      "learning_rate": 4.962572627924795e-06,
+      "loss": 0.6871,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3573407202216066,
+      "grad_norm": 0.672592043876648,
+      "learning_rate": 4.962509803288743e-06,
+      "loss": 0.7238,
+      "step": 1290
+    },
+    {
+      "epoch": 0.35761772853185597,
+      "grad_norm": 0.6334097385406494,
+      "learning_rate": 4.962446926367319e-06,
+      "loss": 0.7161,
+      "step": 1291
+    },
+    {
+      "epoch": 0.35789473684210527,
+      "grad_norm": 0.6492845416069031,
+      "learning_rate": 4.962383997161858e-06,
+      "loss": 0.6651,
+      "step": 1292
+    },
+    {
+      "epoch": 0.35817174515235456,
+      "grad_norm": 0.6691606640815735,
+      "learning_rate": 4.9623210156736966e-06,
+      "loss": 0.6966,
+      "step": 1293
+    },
+    {
+      "epoch": 0.35844875346260385,
+      "grad_norm": 0.6494303941726685,
+      "learning_rate": 4.962257981904172e-06,
+      "loss": 0.6527,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3587257617728532,
+      "grad_norm": 0.6685274839401245,
+      "learning_rate": 4.962194895854623e-06,
+      "loss": 0.7082,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3590027700831025,
+      "grad_norm": 0.6498396992683411,
+      "learning_rate": 4.962131757526387e-06,
+      "loss": 0.6691,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3592797783933518,
+      "grad_norm": 0.6530342698097229,
+      "learning_rate": 4.962068566920808e-06,
+      "loss": 0.7173,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3595567867036011,
+      "grad_norm": 0.6080851554870605,
+      "learning_rate": 4.962005324039226e-06,
+      "loss": 0.6196,
+      "step": 1298
+    },
+    {
+      "epoch": 0.35983379501385043,
+      "grad_norm": 0.6345664858818054,
+      "learning_rate": 4.961942028882981e-06,
+      "loss": 0.6687,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3601108033240997,
+      "grad_norm": 0.6660663485527039,
+      "learning_rate": 4.961878681453423e-06,
+      "loss": 0.6624,
+      "step": 1300
+    },
+    {
+      "epoch": 0.360387811634349,
+      "grad_norm": 0.6534367799758911,
+      "learning_rate": 4.961815281751892e-06,
+      "loss": 0.7117,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3606648199445983,
+      "grad_norm": 0.6991186738014221,
+      "learning_rate": 4.9617518297797355e-06,
+      "loss": 0.7329,
+      "step": 1302
+    },
+    {
+      "epoch": 0.36094182825484766,
+      "grad_norm": 0.6646499037742615,
+      "learning_rate": 4.9616883255383015e-06,
+      "loss": 0.6842,
+      "step": 1303
+    },
+    {
+      "epoch": 0.36121883656509696,
+      "grad_norm": 0.6496413350105286,
+      "learning_rate": 4.961624769028938e-06,
+      "loss": 0.7275,
+      "step": 1304
+    },
+    {
+      "epoch": 0.36149584487534625,
+      "grad_norm": 0.6743121147155762,
+      "learning_rate": 4.961561160252994e-06,
+      "loss": 0.7148,
+      "step": 1305
+    },
+    {
+      "epoch": 0.36177285318559554,
+      "grad_norm": 0.6612362265586853,
+      "learning_rate": 4.96149749921182e-06,
+      "loss": 0.6799,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3620498614958449,
+      "grad_norm": 0.6754231452941895,
+      "learning_rate": 4.961433785906769e-06,
+      "loss": 0.7226,
+      "step": 1307
+    },
+    {
+      "epoch": 0.3623268698060942,
+      "grad_norm": 0.6662765145301819,
+      "learning_rate": 4.961370020339193e-06,
+      "loss": 0.7089,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3626038781163435,
+      "grad_norm": 0.6733217239379883,
+      "learning_rate": 4.9613062025104454e-06,
+      "loss": 0.6817,
+      "step": 1309
+    },
+    {
+      "epoch": 0.3628808864265928,
+      "grad_norm": 0.671064555644989,
+      "learning_rate": 4.9612423324218816e-06,
+      "loss": 0.6983,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3631578947368421,
+      "grad_norm": 0.6356818079948425,
+      "learning_rate": 4.961178410074858e-06,
+      "loss": 0.6906,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3634349030470914,
+      "grad_norm": 0.6797766089439392,
+      "learning_rate": 4.961114435470731e-06,
+      "loss": 0.7292,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3637119113573407,
+      "grad_norm": 0.6782345771789551,
+      "learning_rate": 4.96105040861086e-06,
+      "loss": 0.7163,
+      "step": 1313
+    },
+    {
+      "epoch": 0.36398891966759,
+      "grad_norm": 0.6355205774307251,
+      "learning_rate": 4.960986329496605e-06,
+      "loss": 0.6805,
+      "step": 1314
+    },
+    {
+      "epoch": 0.36426592797783935,
+      "grad_norm": 0.6918122172355652,
+      "learning_rate": 4.960922198129323e-06,
+      "loss": 0.6954,
+      "step": 1315
+    },
+    {
+      "epoch": 0.36454293628808865,
+      "grad_norm": 0.6936689615249634,
+      "learning_rate": 4.96085801451038e-06,
+      "loss": 0.675,
+      "step": 1316
+    },
+    {
+      "epoch": 0.36481994459833794,
+      "grad_norm": 0.6387231349945068,
+      "learning_rate": 4.960793778641137e-06,
+      "loss": 0.6831,
+      "step": 1317
+    },
+    {
+      "epoch": 0.36509695290858724,
+      "grad_norm": 0.6836456060409546,
+      "learning_rate": 4.960729490522957e-06,
+      "loss": 0.7293,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3653739612188366,
+      "grad_norm": 0.660629153251648,
+      "learning_rate": 4.9606651501572066e-06,
+      "loss": 0.6764,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3656509695290859,
+      "grad_norm": 0.6675608158111572,
+      "learning_rate": 4.960600757545252e-06,
+      "loss": 0.6952,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3659279778393352,
+      "grad_norm": 0.6543275117874146,
+      "learning_rate": 4.960536312688458e-06,
+      "loss": 0.7,
+      "step": 1321
+    },
+    {
+      "epoch": 0.36620498614958447,
+      "grad_norm": 0.6812704801559448,
+      "learning_rate": 4.960471815588196e-06,
+      "loss": 0.6978,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3664819944598338,
+      "grad_norm": 0.6616919040679932,
+      "learning_rate": 4.9604072662458325e-06,
+      "loss": 0.7082,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3667590027700831,
+      "grad_norm": 0.6452640295028687,
+      "learning_rate": 4.960342664662741e-06,
+      "loss": 0.7089,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3670360110803324,
+      "grad_norm": 0.6687643527984619,
+      "learning_rate": 4.96027801084029e-06,
+      "loss": 0.6747,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3673130193905817,
+      "grad_norm": 0.6758833527565002,
+      "learning_rate": 4.960213304779856e-06,
+      "loss": 0.7496,
+      "step": 1326
+    },
+    {
+      "epoch": 0.36759002770083105,
+      "grad_norm": 0.6907699108123779,
+      "learning_rate": 4.9601485464828095e-06,
+      "loss": 0.7585,
+      "step": 1327
+    },
+    {
+      "epoch": 0.36786703601108034,
+      "grad_norm": 0.6245575547218323,
+      "learning_rate": 4.960083735950527e-06,
+      "loss": 0.6814,
+      "step": 1328
+    },
+    {
+      "epoch": 0.36814404432132963,
+      "grad_norm": 0.6503439545631409,
+      "learning_rate": 4.960018873184385e-06,
+      "loss": 0.7128,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.6851927638053894,
+      "learning_rate": 4.959953958185759e-06,
+      "loss": 0.6877,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3686980609418283,
+      "grad_norm": 0.653220534324646,
+      "learning_rate": 4.95988899095603e-06,
+      "loss": 0.6847,
+      "step": 1331
+    },
+    {
+      "epoch": 0.36897506925207757,
+      "grad_norm": 0.700905978679657,
+      "learning_rate": 4.959823971496575e-06,
+      "loss": 0.6939,
+      "step": 1332
+    },
+    {
+      "epoch": 0.36925207756232686,
+      "grad_norm": 0.6579366326332092,
+      "learning_rate": 4.959758899808775e-06,
+      "loss": 0.7145,
+      "step": 1333
+    },
+    {
+      "epoch": 0.36952908587257616,
+      "grad_norm": 0.6845544576644897,
+      "learning_rate": 4.959693775894012e-06,
+      "loss": 0.7305,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3698060941828255,
+      "grad_norm": 0.6898097991943359,
+      "learning_rate": 4.95962859975367e-06,
+      "loss": 0.7307,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3700831024930748,
+      "grad_norm": 0.6651111245155334,
+      "learning_rate": 4.95956337138913e-06,
+      "loss": 0.7074,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3703601108033241,
+      "grad_norm": 0.7062095999717712,
+      "learning_rate": 4.9594980908017795e-06,
+      "loss": 0.7392,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3706371191135734,
+      "grad_norm": 0.6984573006629944,
+      "learning_rate": 4.9594327579930034e-06,
+      "loss": 0.7396,
+      "step": 1338
+    },
+    {
+      "epoch": 0.37091412742382274,
+      "grad_norm": 0.7115591764450073,
+      "learning_rate": 4.95936737296419e-06,
+      "loss": 0.7084,
+      "step": 1339
+    },
+    {
+      "epoch": 0.37119113573407203,
+      "grad_norm": 0.6430774331092834,
+      "learning_rate": 4.959301935716725e-06,
+      "loss": 0.6557,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3714681440443213,
+      "grad_norm": 0.6756094694137573,
+      "learning_rate": 4.959236446252001e-06,
+      "loss": 0.6998,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3717451523545706,
+      "grad_norm": 0.6710341572761536,
+      "learning_rate": 4.959170904571406e-06,
+      "loss": 0.7517,
+      "step": 1342
+    },
+    {
+      "epoch": 0.37202216066481997,
+      "grad_norm": 0.7036310434341431,
+      "learning_rate": 4.9591053106763334e-06,
+      "loss": 0.7403,
+      "step": 1343
+    },
+    {
+      "epoch": 0.37229916897506926,
+      "grad_norm": 0.6434355974197388,
+      "learning_rate": 4.959039664568175e-06,
+      "loss": 0.6965,
+      "step": 1344
+    },
+    {
+      "epoch": 0.37257617728531855,
+      "grad_norm": 0.6460430026054382,
+      "learning_rate": 4.958973966248325e-06,
+      "loss": 0.6901,
+      "step": 1345
+    },
+    {
+      "epoch": 0.37285318559556785,
+      "grad_norm": 0.6600189208984375,
+      "learning_rate": 4.958908215718178e-06,
+      "loss": 0.7247,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3731301939058172,
+      "grad_norm": 0.6722584962844849,
+      "learning_rate": 4.95884241297913e-06,
+      "loss": 0.6923,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3734072022160665,
+      "grad_norm": 0.6506482362747192,
+      "learning_rate": 4.958776558032578e-06,
+      "loss": 0.7135,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3736842105263158,
+      "grad_norm": 0.6147697567939758,
+      "learning_rate": 4.958710650879922e-06,
+      "loss": 0.6569,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3739612188365651,
+      "grad_norm": 0.6361560821533203,
+      "learning_rate": 4.958644691522559e-06,
+      "loss": 0.7233,
+      "step": 1350
+    },
+    {
+      "epoch": 0.37423822714681443,
+      "grad_norm": 0.6788303852081299,
+      "learning_rate": 4.95857867996189e-06,
+      "loss": 0.7168,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3745152354570637,
+      "grad_norm": 0.6241387724876404,
+      "learning_rate": 4.958512616199319e-06,
+      "loss": 0.712,
+      "step": 1352
+    },
+    {
+      "epoch": 0.374792243767313,
+      "grad_norm": 0.6549175977706909,
+      "learning_rate": 4.9584465002362455e-06,
+      "loss": 0.6646,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3750692520775623,
+      "grad_norm": 0.6630549430847168,
+      "learning_rate": 4.958380332074074e-06,
+      "loss": 0.6819,
+      "step": 1354
+    },
+    {
+      "epoch": 0.37534626038781166,
+      "grad_norm": 0.6520205140113831,
+      "learning_rate": 4.958314111714211e-06,
+      "loss": 0.7394,
+      "step": 1355
+    },
+    {
+      "epoch": 0.37562326869806095,
+      "grad_norm": 0.6434879302978516,
+      "learning_rate": 4.958247839158061e-06,
+      "loss": 0.6922,
+      "step": 1356
+    },
+    {
+      "epoch": 0.37590027700831025,
+      "grad_norm": 0.7042064070701599,
+      "learning_rate": 4.958181514407033e-06,
+      "loss": 0.6784,
+      "step": 1357
+    },
+    {
+      "epoch": 0.37617728531855954,
+      "grad_norm": 0.6494789719581604,
+      "learning_rate": 4.958115137462533e-06,
+      "loss": 0.7081,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3764542936288089,
+      "grad_norm": 0.6651448607444763,
+      "learning_rate": 4.958048708325971e-06,
+      "loss": 0.6476,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3767313019390582,
+      "grad_norm": 0.6566942930221558,
+      "learning_rate": 4.957982226998758e-06,
+      "loss": 0.7013,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3770083102493075,
+      "grad_norm": 0.6861854195594788,
+      "learning_rate": 4.957915693482305e-06,
+      "loss": 0.7128,
+      "step": 1361
+    },
+    {
+      "epoch": 0.37728531855955677,
+      "grad_norm": 0.6896612644195557,
+      "learning_rate": 4.957849107778026e-06,
+      "loss": 0.7239,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3775623268698061,
+      "grad_norm": 0.6557917594909668,
+      "learning_rate": 4.957782469887332e-06,
+      "loss": 0.7025,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3778393351800554,
+      "grad_norm": 0.6517524123191833,
+      "learning_rate": 4.957715779811642e-06,
+      "loss": 0.6948,
+      "step": 1364
+    },
+    {
+      "epoch": 0.3781163434903047,
+      "grad_norm": 0.7278082966804504,
+      "learning_rate": 4.957649037552368e-06,
+      "loss": 0.7248,
+      "step": 1365
+    },
+    {
+      "epoch": 0.378393351800554,
+      "grad_norm": 0.6831822395324707,
+      "learning_rate": 4.957582243110929e-06,
+      "loss": 0.7396,
+      "step": 1366
+    },
+    {
+      "epoch": 0.37867036011080335,
+      "grad_norm": 0.6876621842384338,
+      "learning_rate": 4.957515396488744e-06,
+      "loss": 0.7126,
+      "step": 1367
+    },
+    {
+      "epoch": 0.37894736842105264,
+      "grad_norm": 0.679372251033783,
+      "learning_rate": 4.957448497687229e-06,
+      "loss": 0.7149,
+      "step": 1368
+    },
+    {
+      "epoch": 0.37922437673130194,
+      "grad_norm": 0.662017285823822,
+      "learning_rate": 4.957381546707809e-06,
+      "loss": 0.7438,
+      "step": 1369
+    },
+    {
+      "epoch": 0.37950138504155123,
+      "grad_norm": 0.6179419159889221,
+      "learning_rate": 4.957314543551903e-06,
+      "loss": 0.6347,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3797783933518006,
+      "grad_norm": 0.6868166923522949,
+      "learning_rate": 4.957247488220933e-06,
+      "loss": 0.696,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3800554016620499,
+      "grad_norm": 0.7437651753425598,
+      "learning_rate": 4.957180380716324e-06,
+      "loss": 0.6938,
+      "step": 1372
+    },
+    {
+      "epoch": 0.38033240997229917,
+      "grad_norm": 0.6506156325340271,
+      "learning_rate": 4.957113221039501e-06,
+      "loss": 0.7085,
+      "step": 1373
+    },
+    {
+      "epoch": 0.38060941828254846,
+      "grad_norm": 0.6750197410583496,
+      "learning_rate": 4.957046009191889e-06,
+      "loss": 0.7394,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3808864265927978,
+      "grad_norm": 0.6268515586853027,
+      "learning_rate": 4.956978745174916e-06,
+      "loss": 0.6687,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3811634349030471,
+      "grad_norm": 0.6731964945793152,
+      "learning_rate": 4.95691142899001e-06,
+      "loss": 0.7027,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3814404432132964,
+      "grad_norm": 0.6968632936477661,
+      "learning_rate": 4.9568440606386e-06,
+      "loss": 0.7307,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3817174515235457,
+      "grad_norm": 0.6785265207290649,
+      "learning_rate": 4.956776640122116e-06,
+      "loss": 0.7197,
+      "step": 1378
+    },
+    {
+      "epoch": 0.38199445983379504,
+      "grad_norm": 0.6669479608535767,
+      "learning_rate": 4.956709167441991e-06,
+      "loss": 0.6942,
+      "step": 1379
+    },
+    {
+      "epoch": 0.38227146814404434,
+      "grad_norm": 0.7017685770988464,
+      "learning_rate": 4.9566416425996555e-06,
+      "loss": 0.7041,
+      "step": 1380
+    },
+    {
+      "epoch": 0.38254847645429363,
+      "grad_norm": 0.6782371401786804,
+      "learning_rate": 4.956574065596545e-06,
+      "loss": 0.7378,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3828254847645429,
+      "grad_norm": 0.6499651074409485,
+      "learning_rate": 4.956506436434093e-06,
+      "loss": 0.6821,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3831024930747922,
+      "grad_norm": 0.653150737285614,
+      "learning_rate": 4.956438755113737e-06,
+      "loss": 0.6934,
+      "step": 1383
+    },
+    {
+      "epoch": 0.38337950138504157,
+      "grad_norm": 0.6766655445098877,
+      "learning_rate": 4.956371021636912e-06,
+      "loss": 0.6898,
+      "step": 1384
+    },
+    {
+      "epoch": 0.38365650969529086,
+      "grad_norm": 0.6720975041389465,
+      "learning_rate": 4.956303236005059e-06,
+      "loss": 0.7228,
+      "step": 1385
+    },
+    {
+      "epoch": 0.38393351800554015,
+      "grad_norm": 0.6506999731063843,
+      "learning_rate": 4.956235398219614e-06,
+      "loss": 0.7127,
+      "step": 1386
+    },
+    {
+      "epoch": 0.38421052631578945,
+      "grad_norm": 0.6865505576133728,
+      "learning_rate": 4.956167508282021e-06,
+      "loss": 0.7291,
+      "step": 1387
+    },
+    {
+      "epoch": 0.3844875346260388,
+      "grad_norm": 0.6950021982192993,
+      "learning_rate": 4.9560995661937175e-06,
+      "loss": 0.6999,
+      "step": 1388
+    },
+    {
+      "epoch": 0.3847645429362881,
+      "grad_norm": 0.666877269744873,
+      "learning_rate": 4.956031571956148e-06,
+      "loss": 0.6909,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3850415512465374,
+      "grad_norm": 0.6657650470733643,
+      "learning_rate": 4.955963525570757e-06,
+      "loss": 0.6762,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3853185595567867,
+      "grad_norm": 0.6917940378189087,
+      "learning_rate": 4.955895427038988e-06,
+      "loss": 0.6762,
+      "step": 1391
+    },
+    {
+      "epoch": 0.385595567867036,
+      "grad_norm": 0.6938834190368652,
+      "learning_rate": 4.955827276362288e-06,
+      "loss": 0.734,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3858725761772853,
+      "grad_norm": 0.7303954362869263,
+      "learning_rate": 4.955759073542104e-06,
+      "loss": 0.7666,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3861495844875346,
+      "grad_norm": 0.6774318814277649,
+      "learning_rate": 4.9556908185798815e-06,
+      "loss": 0.7199,
+      "step": 1394
+    },
+    {
+      "epoch": 0.3864265927977839,
+      "grad_norm": 0.6571752429008484,
+      "learning_rate": 4.955622511477073e-06,
+      "loss": 0.7084,
+      "step": 1395
+    },
+    {
+      "epoch": 0.38670360110803326,
+      "grad_norm": 0.646608293056488,
+      "learning_rate": 4.955554152235128e-06,
+      "loss": 0.6636,
+      "step": 1396
+    },
+    {
+      "epoch": 0.38698060941828255,
+      "grad_norm": 0.6419099569320679,
+      "learning_rate": 4.955485740855497e-06,
+      "loss": 0.6862,
+      "step": 1397
+    },
+    {
+      "epoch": 0.38725761772853184,
+      "grad_norm": 0.6539189219474792,
+      "learning_rate": 4.955417277339633e-06,
+      "loss": 0.6339,
+      "step": 1398
+    },
+    {
+      "epoch": 0.38753462603878114,
+      "grad_norm": 0.6766827702522278,
+      "learning_rate": 4.95534876168899e-06,
+      "loss": 0.6939,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3878116343490305,
+      "grad_norm": 0.6848649382591248,
+      "learning_rate": 4.955280193905023e-06,
+      "loss": 0.7284,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3880886426592798,
+      "grad_norm": 0.6462283134460449,
+      "learning_rate": 4.955211573989186e-06,
+      "loss": 0.7049,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3883656509695291,
+      "grad_norm": 0.6433562636375427,
+      "learning_rate": 4.955142901942939e-06,
+      "loss": 0.6847,
+      "step": 1402
+    },
+    {
+      "epoch": 0.38864265927977837,
+      "grad_norm": 0.7061001062393188,
+      "learning_rate": 4.955074177767737e-06,
+      "loss": 0.7501,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3889196675900277,
+      "grad_norm": 0.6696621775627136,
+      "learning_rate": 4.955005401465042e-06,
+      "loss": 0.7088,
+      "step": 1404
+    },
+    {
+      "epoch": 0.389196675900277,
+      "grad_norm": 0.6648236513137817,
+      "learning_rate": 4.954936573036312e-06,
+      "loss": 0.7133,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3894736842105263,
+      "grad_norm": 0.6691426038742065,
+      "learning_rate": 4.954867692483009e-06,
+      "loss": 0.7026,
+      "step": 1406
+    },
+    {
+      "epoch": 0.3897506925207756,
+      "grad_norm": 0.6885944604873657,
+      "learning_rate": 4.954798759806596e-06,
+      "loss": 0.6919,
+      "step": 1407
+    },
+    {
+      "epoch": 0.39002770083102495,
+      "grad_norm": 0.6416887640953064,
+      "learning_rate": 4.9547297750085375e-06,
+      "loss": 0.6601,
+      "step": 1408
+    },
+    {
+      "epoch": 0.39030470914127424,
+      "grad_norm": 0.6520400643348694,
+      "learning_rate": 4.954660738090297e-06,
+      "loss": 0.6856,
+      "step": 1409
+    },
+    {
+      "epoch": 0.39058171745152354,
+      "grad_norm": 0.6656164526939392,
+      "learning_rate": 4.95459164905334e-06,
+      "loss": 0.6803,
+      "step": 1410
+    },
+    {
+      "epoch": 0.39085872576177283,
+      "grad_norm": 0.7128254771232605,
+      "learning_rate": 4.954522507899133e-06,
+      "loss": 0.6548,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3911357340720222,
+      "grad_norm": 0.6742157340049744,
+      "learning_rate": 4.954453314629146e-06,
+      "loss": 0.7053,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3914127423822715,
+      "grad_norm": 0.6287198066711426,
+      "learning_rate": 4.954384069244848e-06,
+      "loss": 0.6742,
+      "step": 1413
+    },
+    {
+      "epoch": 0.39168975069252077,
+      "grad_norm": 0.6757739782333374,
+      "learning_rate": 4.954314771747707e-06,
+      "loss": 0.7236,
+      "step": 1414
+    },
+    {
+      "epoch": 0.39196675900277006,
+      "grad_norm": 0.68620365858078,
+      "learning_rate": 4.954245422139196e-06,
+      "loss": 0.6842,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3922437673130194,
+      "grad_norm": 0.6858413815498352,
+      "learning_rate": 4.954176020420788e-06,
+      "loss": 0.7019,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3925207756232687,
+      "grad_norm": 0.6674286723136902,
+      "learning_rate": 4.954106566593955e-06,
+      "loss": 0.7249,
+      "step": 1417
+    },
+    {
+      "epoch": 0.392797783933518,
+      "grad_norm": 0.6651579737663269,
+      "learning_rate": 4.954037060660173e-06,
+      "loss": 0.7113,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3930747922437673,
+      "grad_norm": 0.6830371618270874,
+      "learning_rate": 4.953967502620916e-06,
+      "loss": 0.727,
+      "step": 1419
+    },
+    {
+      "epoch": 0.39335180055401664,
+      "grad_norm": 0.6857573986053467,
+      "learning_rate": 4.953897892477664e-06,
+      "loss": 0.7197,
+      "step": 1420
+    },
+    {
+      "epoch": 0.39362880886426593,
+      "grad_norm": 0.6651442646980286,
+      "learning_rate": 4.953828230231892e-06,
+      "loss": 0.6771,
+      "step": 1421
+    },
+    {
+      "epoch": 0.3939058171745152,
+      "grad_norm": 0.7004607319831848,
+      "learning_rate": 4.953758515885081e-06,
+      "loss": 0.7428,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3941828254847645,
+      "grad_norm": 0.6818897724151611,
+      "learning_rate": 4.95368874943871e-06,
+      "loss": 0.7024,
+      "step": 1423
+    },
+    {
+      "epoch": 0.39445983379501387,
+      "grad_norm": 0.6543837189674377,
+      "learning_rate": 4.95361893089426e-06,
+      "loss": 0.7267,
+      "step": 1424
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 0.6330406665802002,
+      "learning_rate": 4.953549060253216e-06,
+      "loss": 0.7249,
+      "step": 1425
+    },
+    {
+      "epoch": 0.39501385041551246,
+      "grad_norm": 0.6749814748764038,
+      "learning_rate": 4.953479137517059e-06,
+      "loss": 0.7158,
+      "step": 1426
+    },
+    {
+      "epoch": 0.39529085872576175,
+      "grad_norm": 0.6552304029464722,
+      "learning_rate": 4.953409162687274e-06,
+      "loss": 0.7637,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3955678670360111,
+      "grad_norm": 0.6880520582199097,
+      "learning_rate": 4.953339135765348e-06,
+      "loss": 0.7158,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3958448753462604,
+      "grad_norm": 0.6481505036354065,
+      "learning_rate": 4.953269056752767e-06,
+      "loss": 0.7075,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3961218836565097,
+      "grad_norm": 0.6551262736320496,
+      "learning_rate": 4.953198925651018e-06,
+      "loss": 0.6933,
+      "step": 1430
+    },
+    {
+      "epoch": 0.396398891966759,
+      "grad_norm": 0.6902202367782593,
+      "learning_rate": 4.953128742461592e-06,
+      "loss": 0.7142,
+      "step": 1431
+    },
+    {
+      "epoch": 0.39667590027700833,
+      "grad_norm": 0.6663317084312439,
+      "learning_rate": 4.9530585071859775e-06,
+      "loss": 0.7366,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3969529085872576,
+      "grad_norm": 0.6709367036819458,
+      "learning_rate": 4.952988219825666e-06,
+      "loss": 0.7188,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3972299168975069,
+      "grad_norm": 0.6727588176727295,
+      "learning_rate": 4.952917880382151e-06,
+      "loss": 0.7175,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3975069252077562,
+      "grad_norm": 0.6796942949295044,
+      "learning_rate": 4.952847488856925e-06,
+      "loss": 0.6983,
+      "step": 1435
+    },
+    {
+      "epoch": 0.39778393351800556,
+      "grad_norm": 0.6838971376419067,
+      "learning_rate": 4.952777045251484e-06,
+      "loss": 0.65,
+      "step": 1436
+    },
+    {
+      "epoch": 0.39806094182825486,
+      "grad_norm": 0.6644639372825623,
+      "learning_rate": 4.952706549567322e-06,
+      "loss": 0.6994,
+      "step": 1437
+    },
+    {
+      "epoch": 0.39833795013850415,
+      "grad_norm": 0.6655013561248779,
+      "learning_rate": 4.952636001805936e-06,
+      "loss": 0.6803,
+      "step": 1438
+    },
+    {
+      "epoch": 0.39861495844875344,
+      "grad_norm": 0.6782181262969971,
+      "learning_rate": 4.952565401968824e-06,
+      "loss": 0.7076,
+      "step": 1439
+    },
+    {
+      "epoch": 0.3988919667590028,
+      "grad_norm": 0.8163707852363586,
+      "learning_rate": 4.952494750057487e-06,
+      "loss": 0.731,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3991689750692521,
+      "grad_norm": 0.6641796231269836,
+      "learning_rate": 4.952424046073422e-06,
+      "loss": 0.7001,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3994459833795014,
+      "grad_norm": 0.6930070519447327,
+      "learning_rate": 4.952353290018132e-06,
+      "loss": 0.6668,
+      "step": 1442
+    },
+    {
+      "epoch": 0.3997229916897507,
+      "grad_norm": 0.6547805070877075,
+      "learning_rate": 4.9522824818931195e-06,
+      "loss": 0.725,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6426491737365723,
+      "learning_rate": 4.9522116216998876e-06,
+      "loss": 0.6825,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4002770083102493,
+      "grad_norm": 0.6577370762825012,
+      "learning_rate": 4.95214070943994e-06,
+      "loss": 0.6654,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4005540166204986,
+      "grad_norm": 0.6261408925056458,
+      "learning_rate": 4.9520697451147825e-06,
+      "loss": 0.7009,
+      "step": 1446
+    },
+    {
+      "epoch": 0.4008310249307479,
+      "grad_norm": 0.6702444553375244,
+      "learning_rate": 4.951998728725924e-06,
+      "loss": 0.7356,
+      "step": 1447
+    },
+    {
+      "epoch": 0.40110803324099725,
+      "grad_norm": 0.6387395858764648,
+      "learning_rate": 4.95192766027487e-06,
+      "loss": 0.6711,
+      "step": 1448
+    },
+    {
+      "epoch": 0.40138504155124655,
+      "grad_norm": 0.685088038444519,
+      "learning_rate": 4.951856539763129e-06,
+      "loss": 0.7028,
+      "step": 1449
+    },
+    {
+      "epoch": 0.40166204986149584,
+      "grad_norm": 0.6737648248672485,
+      "learning_rate": 4.951785367192214e-06,
+      "loss": 0.7156,
+      "step": 1450
+    },
+    {
+      "epoch": 0.40193905817174513,
+      "grad_norm": 0.7130826115608215,
+      "learning_rate": 4.951714142563634e-06,
+      "loss": 0.7014,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4022160664819945,
+      "grad_norm": 0.6648553013801575,
+      "learning_rate": 4.951642865878902e-06,
+      "loss": 0.7137,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4024930747922438,
+      "grad_norm": 0.6709924340248108,
+      "learning_rate": 4.951571537139531e-06,
+      "loss": 0.7026,
+      "step": 1453
+    },
+    {
+      "epoch": 0.40277008310249307,
+      "grad_norm": 0.6769849061965942,
+      "learning_rate": 4.951500156347035e-06,
+      "loss": 0.7092,
+      "step": 1454
+    },
+    {
+      "epoch": 0.40304709141274236,
+      "grad_norm": 0.6583406329154968,
+      "learning_rate": 4.951428723502932e-06,
+      "loss": 0.7081,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4033240997229917,
+      "grad_norm": 0.6608618497848511,
+      "learning_rate": 4.951357238608735e-06,
+      "loss": 0.6764,
+      "step": 1456
+    },
+    {
+      "epoch": 0.403601108033241,
+      "grad_norm": 0.6477077603340149,
+      "learning_rate": 4.9512857016659645e-06,
+      "loss": 0.7206,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4038781163434903,
+      "grad_norm": 0.6960606575012207,
+      "learning_rate": 4.951214112676138e-06,
+      "loss": 0.6939,
+      "step": 1458
+    },
+    {
+      "epoch": 0.4041551246537396,
+      "grad_norm": 0.6560659408569336,
+      "learning_rate": 4.9511424716407765e-06,
+      "loss": 0.6741,
+      "step": 1459
+    },
+    {
+      "epoch": 0.40443213296398894,
+      "grad_norm": 0.7191847562789917,
+      "learning_rate": 4.951070778561401e-06,
+      "loss": 0.6589,
+      "step": 1460
+    },
+    {
+      "epoch": 0.40470914127423824,
+      "grad_norm": 0.6729449033737183,
+      "learning_rate": 4.950999033439534e-06,
+      "loss": 0.7091,
+      "step": 1461
+    },
+    {
+      "epoch": 0.40498614958448753,
+      "grad_norm": 0.6878858804702759,
+      "learning_rate": 4.950927236276698e-06,
+      "loss": 0.6943,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4052631578947368,
+      "grad_norm": 0.6467249393463135,
+      "learning_rate": 4.950855387074417e-06,
+      "loss": 0.7039,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4055401662049862,
+      "grad_norm": 0.6764515042304993,
+      "learning_rate": 4.950783485834218e-06,
+      "loss": 0.7004,
+      "step": 1464
+    },
+    {
+      "epoch": 0.40581717451523547,
+      "grad_norm": 0.6590722799301147,
+      "learning_rate": 4.950711532557627e-06,
+      "loss": 0.6954,
+      "step": 1465
+    },
+    {
+      "epoch": 0.40609418282548476,
+      "grad_norm": 0.6726288199424744,
+      "learning_rate": 4.950639527246172e-06,
+      "loss": 0.7311,
+      "step": 1466
+    },
+    {
+      "epoch": 0.40637119113573406,
+      "grad_norm": 0.6541720032691956,
+      "learning_rate": 4.950567469901382e-06,
+      "loss": 0.6552,
+      "step": 1467
+    },
+    {
+      "epoch": 0.4066481994459834,
+      "grad_norm": 0.6888921856880188,
+      "learning_rate": 4.950495360524786e-06,
+      "loss": 0.6341,
+      "step": 1468
+    },
+    {
+      "epoch": 0.4069252077562327,
+      "grad_norm": 0.7008625268936157,
+      "learning_rate": 4.950423199117915e-06,
+      "loss": 0.7526,
+      "step": 1469
+    },
+    {
+      "epoch": 0.407202216066482,
+      "grad_norm": 0.6698315143585205,
+      "learning_rate": 4.950350985682302e-06,
+      "loss": 0.6792,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4074792243767313,
+      "grad_norm": 0.615836501121521,
+      "learning_rate": 4.95027872021948e-06,
+      "loss": 0.6562,
+      "step": 1471
+    },
+    {
+      "epoch": 0.40775623268698064,
+      "grad_norm": 0.6405780911445618,
+      "learning_rate": 4.950206402730984e-06,
+      "loss": 0.6895,
+      "step": 1472
+    },
+    {
+      "epoch": 0.40803324099722993,
+      "grad_norm": 0.6810288429260254,
+      "learning_rate": 4.950134033218349e-06,
+      "loss": 0.7323,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4083102493074792,
+      "grad_norm": 0.6716300249099731,
+      "learning_rate": 4.950061611683111e-06,
+      "loss": 0.6745,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4085872576177285,
+      "grad_norm": 0.6801728010177612,
+      "learning_rate": 4.949989138126809e-06,
+      "loss": 0.7043,
+      "step": 1475
+    },
+    {
+      "epoch": 0.40886426592797787,
+      "grad_norm": 0.7009851336479187,
+      "learning_rate": 4.949916612550979e-06,
+      "loss": 0.7346,
+      "step": 1476
+    },
+    {
+      "epoch": 0.40914127423822716,
+      "grad_norm": 0.7393293380737305,
+      "learning_rate": 4.949844034957165e-06,
+      "loss": 0.7276,
+      "step": 1477
+    },
+    {
+      "epoch": 0.40941828254847645,
+      "grad_norm": 0.6590290069580078,
+      "learning_rate": 4.949771405346905e-06,
+      "loss": 0.7004,
+      "step": 1478
+    },
+    {
+      "epoch": 0.40969529085872575,
+      "grad_norm": 0.6548259258270264,
+      "learning_rate": 4.949698723721742e-06,
+      "loss": 0.7383,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4099722991689751,
+      "grad_norm": 0.7192888855934143,
+      "learning_rate": 4.94962599008322e-06,
+      "loss": 0.7471,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4102493074792244,
+      "grad_norm": 0.6807044148445129,
+      "learning_rate": 4.949553204432882e-06,
+      "loss": 0.7549,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4105263157894737,
+      "grad_norm": 0.6511745452880859,
+      "learning_rate": 4.949480366772274e-06,
+      "loss": 0.6995,
+      "step": 1482
+    },
+    {
+      "epoch": 0.410803324099723,
+      "grad_norm": 0.6546177268028259,
+      "learning_rate": 4.949407477102943e-06,
+      "loss": 0.6919,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4110803324099723,
+      "grad_norm": 0.6770357489585876,
+      "learning_rate": 4.949334535426435e-06,
+      "loss": 0.7243,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4113573407202216,
+      "grad_norm": 0.6877369284629822,
+      "learning_rate": 4.949261541744301e-06,
+      "loss": 0.6891,
+      "step": 1485
+    },
+    {
+      "epoch": 0.4116343490304709,
+      "grad_norm": 0.6853678822517395,
+      "learning_rate": 4.949188496058089e-06,
+      "loss": 0.7018,
+      "step": 1486
+    },
+    {
+      "epoch": 0.4119113573407202,
+      "grad_norm": 0.6893583536148071,
+      "learning_rate": 4.949115398369351e-06,
+      "loss": 0.6695,
+      "step": 1487
+    },
+    {
+      "epoch": 0.4121883656509695,
+      "grad_norm": 0.7509121894836426,
+      "learning_rate": 4.949042248679639e-06,
+      "loss": 0.7029,
+      "step": 1488
+    },
+    {
+      "epoch": 0.41246537396121885,
+      "grad_norm": 0.7181180715560913,
+      "learning_rate": 4.948969046990506e-06,
+      "loss": 0.6863,
+      "step": 1489
+    },
+    {
+      "epoch": 0.41274238227146814,
+      "grad_norm": 0.6763536930084229,
+      "learning_rate": 4.948895793303505e-06,
+      "loss": 0.7526,
+      "step": 1490
+    },
+    {
+      "epoch": 0.41301939058171744,
+      "grad_norm": 0.6821795701980591,
+      "learning_rate": 4.948822487620194e-06,
+      "loss": 0.6841,
+      "step": 1491
+    },
+    {
+      "epoch": 0.41329639889196673,
+      "grad_norm": 0.6655170321464539,
+      "learning_rate": 4.948749129942127e-06,
+      "loss": 0.6658,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4135734072022161,
+      "grad_norm": 0.6946587562561035,
+      "learning_rate": 4.948675720270863e-06,
+      "loss": 0.7428,
+      "step": 1493
+    },
+    {
+      "epoch": 0.4138504155124654,
+      "grad_norm": 0.6431707143783569,
+      "learning_rate": 4.94860225860796e-06,
+      "loss": 0.6918,
+      "step": 1494
+    },
+    {
+      "epoch": 0.41412742382271467,
+      "grad_norm": 0.6807214021682739,
+      "learning_rate": 4.948528744954978e-06,
+      "loss": 0.6854,
+      "step": 1495
+    },
+    {
+      "epoch": 0.41440443213296396,
+      "grad_norm": 0.6434029936790466,
+      "learning_rate": 4.948455179313479e-06,
+      "loss": 0.7075,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4146814404432133,
+      "grad_norm": 0.6675246953964233,
+      "learning_rate": 4.948381561685022e-06,
+      "loss": 0.6956,
+      "step": 1497
+    },
+    {
+      "epoch": 0.4149584487534626,
+      "grad_norm": 0.6771344542503357,
+      "learning_rate": 4.948307892071173e-06,
+      "loss": 0.702,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4152354570637119,
+      "grad_norm": 0.7052770256996155,
+      "learning_rate": 4.948234170473495e-06,
+      "loss": 0.7251,
+      "step": 1499
+    },
+    {
+      "epoch": 0.4155124653739612,
+      "grad_norm": 0.6624787449836731,
+      "learning_rate": 4.948160396893553e-06,
+      "loss": 0.7037,
+      "step": 1500
+    },
+    {
+      "epoch": 0.41578947368421054,
+      "grad_norm": 0.6854806542396545,
+      "learning_rate": 4.948086571332914e-06,
+      "loss": 0.746,
+      "step": 1501
+    },
+    {
+      "epoch": 0.41606648199445984,
+      "grad_norm": 0.7364691495895386,
+      "learning_rate": 4.9480126937931465e-06,
+      "loss": 0.7121,
+      "step": 1502
+    },
+    {
+      "epoch": 0.41634349030470913,
+      "grad_norm": 0.6633167266845703,
+      "learning_rate": 4.947938764275817e-06,
+      "loss": 0.6913,
+      "step": 1503
+    },
+    {
+      "epoch": 0.4166204986149584,
+      "grad_norm": 0.686250627040863,
+      "learning_rate": 4.947864782782496e-06,
+      "loss": 0.7368,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4168975069252078,
+      "grad_norm": 0.6367915272712708,
+      "learning_rate": 4.9477907493147546e-06,
+      "loss": 0.6641,
+      "step": 1505
+    },
+    {
+      "epoch": 0.41717451523545707,
+      "grad_norm": 0.6631414294242859,
+      "learning_rate": 4.947716663874164e-06,
+      "loss": 0.7006,
+      "step": 1506
+    },
+    {
+      "epoch": 0.41745152354570636,
+      "grad_norm": 0.6712923645973206,
+      "learning_rate": 4.9476425264623e-06,
+      "loss": 0.683,
+      "step": 1507
+    },
+    {
+      "epoch": 0.41772853185595565,
+      "grad_norm": 0.6794381141662598,
+      "learning_rate": 4.947568337080733e-06,
+      "loss": 0.6954,
+      "step": 1508
+    },
+    {
+      "epoch": 0.418005540166205,
+      "grad_norm": 0.6655251979827881,
+      "learning_rate": 4.947494095731039e-06,
+      "loss": 0.703,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4182825484764543,
+      "grad_norm": 0.7107792496681213,
+      "learning_rate": 4.947419802414797e-06,
+      "loss": 0.7476,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4185595567867036,
+      "grad_norm": 0.6949066519737244,
+      "learning_rate": 4.947345457133581e-06,
+      "loss": 0.7008,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4188365650969529,
+      "grad_norm": 0.6868277192115784,
+      "learning_rate": 4.947271059888972e-06,
+      "loss": 0.7149,
+      "step": 1512
+    },
+    {
+      "epoch": 0.41911357340720223,
+      "grad_norm": 0.6967531442642212,
+      "learning_rate": 4.9471966106825485e-06,
+      "loss": 0.7146,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4193905817174515,
+      "grad_norm": 0.6734687685966492,
+      "learning_rate": 4.9471221095158914e-06,
+      "loss": 0.6087,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4196675900277008,
+      "grad_norm": 0.6532918214797974,
+      "learning_rate": 4.9470475563905825e-06,
+      "loss": 0.6585,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4199445983379501,
+      "grad_norm": 0.6667916774749756,
+      "learning_rate": 4.946972951308206e-06,
+      "loss": 0.7538,
+      "step": 1516
+    },
+    {
+      "epoch": 0.42022160664819946,
+      "grad_norm": 0.6874780654907227,
+      "learning_rate": 4.946898294270343e-06,
+      "loss": 0.6466,
+      "step": 1517
+    },
+    {
+      "epoch": 0.42049861495844876,
+      "grad_norm": 0.6987433433532715,
+      "learning_rate": 4.9468235852785825e-06,
+      "loss": 0.7162,
+      "step": 1518
+    },
+    {
+      "epoch": 0.42077562326869805,
+      "grad_norm": 0.7233112454414368,
+      "learning_rate": 4.946748824334507e-06,
+      "loss": 0.7059,
+      "step": 1519
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.6833441257476807,
+      "learning_rate": 4.946674011439707e-06,
+      "loss": 0.6819,
+      "step": 1520
+    },
+    {
+      "epoch": 0.4213296398891967,
+      "grad_norm": 0.677476167678833,
+      "learning_rate": 4.946599146595769e-06,
+      "loss": 0.7347,
+      "step": 1521
+    },
+    {
+      "epoch": 0.421606648199446,
+      "grad_norm": 0.670714259147644,
+      "learning_rate": 4.946524229804284e-06,
+      "loss": 0.7048,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4218836565096953,
+      "grad_norm": 0.676911473274231,
+      "learning_rate": 4.94644926106684e-06,
+      "loss": 0.7101,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4221606648199446,
+      "grad_norm": 0.6721144914627075,
+      "learning_rate": 4.946374240385033e-06,
+      "loss": 0.6177,
+      "step": 1524
+    },
+    {
+      "epoch": 0.4224376731301939,
+      "grad_norm": 0.6739724278450012,
+      "learning_rate": 4.946299167760451e-06,
+      "loss": 0.6971,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4227146814404432,
+      "grad_norm": 0.680992066860199,
+      "learning_rate": 4.9462240431946925e-06,
+      "loss": 0.7097,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4229916897506925,
+      "grad_norm": 0.6601333618164062,
+      "learning_rate": 4.94614886668935e-06,
+      "loss": 0.6652,
+      "step": 1527
+    },
+    {
+      "epoch": 0.4232686980609418,
+      "grad_norm": 0.6684186458587646,
+      "learning_rate": 4.946073638246019e-06,
+      "loss": 0.7136,
+      "step": 1528
+    },
+    {
+      "epoch": 0.42354570637119116,
+      "grad_norm": 0.6920457482337952,
+      "learning_rate": 4.9459983578663e-06,
+      "loss": 0.7124,
+      "step": 1529
+    },
+    {
+      "epoch": 0.42382271468144045,
+      "grad_norm": 0.7031549215316772,
+      "learning_rate": 4.945923025551789e-06,
+      "loss": 0.7284,
+      "step": 1530
+    },
+    {
+      "epoch": 0.42409972299168974,
+      "grad_norm": 0.6548120975494385,
+      "learning_rate": 4.945847641304085e-06,
+      "loss": 0.7091,
+      "step": 1531
+    },
+    {
+      "epoch": 0.42437673130193904,
+      "grad_norm": 0.6406707763671875,
+      "learning_rate": 4.94577220512479e-06,
+      "loss": 0.6734,
+      "step": 1532
+    },
+    {
+      "epoch": 0.4246537396121884,
+      "grad_norm": 0.6942365169525146,
+      "learning_rate": 4.9456967170155056e-06,
+      "loss": 0.6622,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4249307479224377,
+      "grad_norm": 0.643136739730835,
+      "learning_rate": 4.945621176977835e-06,
+      "loss": 0.6703,
+      "step": 1534
+    },
+    {
+      "epoch": 0.425207756232687,
+      "grad_norm": 0.677213728427887,
+      "learning_rate": 4.94554558501338e-06,
+      "loss": 0.6994,
+      "step": 1535
+    },
+    {
+      "epoch": 0.42548476454293627,
+      "grad_norm": 0.6650383472442627,
+      "learning_rate": 4.945469941123747e-06,
+      "loss": 0.692,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4257617728531856,
+      "grad_norm": 0.6682860255241394,
+      "learning_rate": 4.945394245310543e-06,
+      "loss": 0.7249,
+      "step": 1537
+    },
+    {
+      "epoch": 0.4260387811634349,
+      "grad_norm": 0.7072668671607971,
+      "learning_rate": 4.945318497575375e-06,
+      "loss": 0.7446,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4263157894736842,
+      "grad_norm": 0.654718816280365,
+      "learning_rate": 4.945242697919849e-06,
+      "loss": 0.7082,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4265927977839335,
+      "grad_norm": 0.6829707622528076,
+      "learning_rate": 4.945166846345576e-06,
+      "loss": 0.6848,
+      "step": 1540
+    },
+    {
+      "epoch": 0.42686980609418285,
+      "grad_norm": 0.6674022674560547,
+      "learning_rate": 4.945090942854167e-06,
+      "loss": 0.6784,
+      "step": 1541
+    },
+    {
+      "epoch": 0.42714681440443214,
+      "grad_norm": 0.6758620142936707,
+      "learning_rate": 4.945014987447234e-06,
+      "loss": 0.7442,
+      "step": 1542
+    },
+    {
+      "epoch": 0.42742382271468143,
+      "grad_norm": 0.6620382070541382,
+      "learning_rate": 4.944938980126388e-06,
+      "loss": 0.6503,
+      "step": 1543
+    },
+    {
+      "epoch": 0.4277008310249307,
+      "grad_norm": 0.6612845659255981,
+      "learning_rate": 4.9448629208932446e-06,
+      "loss": 0.7045,
+      "step": 1544
+    },
+    {
+      "epoch": 0.4279778393351801,
+      "grad_norm": 0.7068538665771484,
+      "learning_rate": 4.944786809749417e-06,
+      "loss": 0.7549,
+      "step": 1545
+    },
+    {
+      "epoch": 0.42825484764542937,
+      "grad_norm": 0.6271569728851318,
+      "learning_rate": 4.944710646696522e-06,
+      "loss": 0.6106,
+      "step": 1546
+    },
+    {
+      "epoch": 0.42853185595567866,
+      "grad_norm": 0.6843795776367188,
+      "learning_rate": 4.944634431736177e-06,
+      "loss": 0.7204,
+      "step": 1547
+    },
+    {
+      "epoch": 0.42880886426592796,
+      "grad_norm": 0.6622493863105774,
+      "learning_rate": 4.9445581648700014e-06,
+      "loss": 0.6639,
+      "step": 1548
+    },
+    {
+      "epoch": 0.4290858725761773,
+      "grad_norm": 0.6531876921653748,
+      "learning_rate": 4.944481846099612e-06,
+      "loss": 0.6987,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4293628808864266,
+      "grad_norm": 0.6773236989974976,
+      "learning_rate": 4.944405475426631e-06,
+      "loss": 0.6525,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4296398891966759,
+      "grad_norm": 0.6729045510292053,
+      "learning_rate": 4.944329052852679e-06,
+      "loss": 0.6935,
+      "step": 1551
+    },
+    {
+      "epoch": 0.4299168975069252,
+      "grad_norm": 0.7236620783805847,
+      "learning_rate": 4.944252578379379e-06,
+      "loss": 0.7528,
+      "step": 1552
+    },
+    {
+      "epoch": 0.43019390581717454,
+      "grad_norm": 0.711037278175354,
+      "learning_rate": 4.944176052008355e-06,
+      "loss": 0.6764,
+      "step": 1553
+    },
+    {
+      "epoch": 0.43047091412742383,
+      "grad_norm": 0.6864672899246216,
+      "learning_rate": 4.944099473741232e-06,
+      "loss": 0.711,
+      "step": 1554
+    },
+    {
+      "epoch": 0.4307479224376731,
+      "grad_norm": 0.6570982933044434,
+      "learning_rate": 4.944022843579635e-06,
+      "loss": 0.6807,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4310249307479224,
+      "grad_norm": 0.6892886757850647,
+      "learning_rate": 4.943946161525192e-06,
+      "loss": 0.6612,
+      "step": 1556
+    },
+    {
+      "epoch": 0.43130193905817177,
+      "grad_norm": 0.6675722002983093,
+      "learning_rate": 4.943869427579531e-06,
+      "loss": 0.7212,
+      "step": 1557
+    },
+    {
+      "epoch": 0.43157894736842106,
+      "grad_norm": 0.6782935261726379,
+      "learning_rate": 4.943792641744281e-06,
+      "loss": 0.6728,
+      "step": 1558
+    },
+    {
+      "epoch": 0.43185595567867036,
+      "grad_norm": 0.6436980366706848,
+      "learning_rate": 4.943715804021072e-06,
+      "loss": 0.6851,
+      "step": 1559
+    },
+    {
+      "epoch": 0.43213296398891965,
+      "grad_norm": 0.7117411494255066,
+      "learning_rate": 4.943638914411537e-06,
+      "loss": 0.7315,
+      "step": 1560
+    },
+    {
+      "epoch": 0.432409972299169,
+      "grad_norm": 0.6860139966011047,
+      "learning_rate": 4.943561972917307e-06,
+      "loss": 0.736,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4326869806094183,
+      "grad_norm": 0.7034031748771667,
+      "learning_rate": 4.943484979540015e-06,
+      "loss": 0.7268,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4329639889196676,
+      "grad_norm": 0.6490092873573303,
+      "learning_rate": 4.943407934281298e-06,
+      "loss": 0.6534,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4332409972299169,
+      "grad_norm": 0.6792609691619873,
+      "learning_rate": 4.943330837142792e-06,
+      "loss": 0.7067,
+      "step": 1564
+    },
+    {
+      "epoch": 0.43351800554016623,
+      "grad_norm": 0.6546031832695007,
+      "learning_rate": 4.94325368812613e-06,
+      "loss": 0.6907,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4337950138504155,
+      "grad_norm": 0.688127875328064,
+      "learning_rate": 4.943176487232955e-06,
+      "loss": 0.6616,
+      "step": 1566
+    },
+    {
+      "epoch": 0.4340720221606648,
+      "grad_norm": 0.6924400329589844,
+      "learning_rate": 4.943099234464902e-06,
+      "loss": 0.6945,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4343490304709141,
+      "grad_norm": 0.7010718584060669,
+      "learning_rate": 4.943021929823615e-06,
+      "loss": 0.73,
+      "step": 1568
+    },
+    {
+      "epoch": 0.43462603878116346,
+      "grad_norm": 0.6934574246406555,
+      "learning_rate": 4.942944573310733e-06,
+      "loss": 0.7208,
+      "step": 1569
+    },
+    {
+      "epoch": 0.43490304709141275,
+      "grad_norm": 0.6866487860679626,
+      "learning_rate": 4.942867164927899e-06,
+      "loss": 0.6729,
+      "step": 1570
+    },
+    {
+      "epoch": 0.43518005540166205,
+      "grad_norm": 0.6813799142837524,
+      "learning_rate": 4.942789704676757e-06,
+      "loss": 0.6887,
+      "step": 1571
+    },
+    {
+      "epoch": 0.43545706371191134,
+      "grad_norm": 0.7103790640830994,
+      "learning_rate": 4.942712192558951e-06,
+      "loss": 0.684,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4357340720221607,
+      "grad_norm": 0.7329322099685669,
+      "learning_rate": 4.942634628576127e-06,
+      "loss": 0.7417,
+      "step": 1573
+    },
+    {
+      "epoch": 0.43601108033241,
+      "grad_norm": 0.6444783806800842,
+      "learning_rate": 4.942557012729933e-06,
+      "loss": 0.7252,
+      "step": 1574
+    },
+    {
+      "epoch": 0.4362880886426593,
+      "grad_norm": 0.6684467196464539,
+      "learning_rate": 4.942479345022016e-06,
+      "loss": 0.6494,
+      "step": 1575
+    },
+    {
+      "epoch": 0.43656509695290857,
+      "grad_norm": 0.6751431822776794,
+      "learning_rate": 4.942401625454024e-06,
+      "loss": 0.6628,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4368421052631579,
+      "grad_norm": 0.7155023217201233,
+      "learning_rate": 4.942323854027609e-06,
+      "loss": 0.7036,
+      "step": 1577
+    },
+    {
+      "epoch": 0.4371191135734072,
+      "grad_norm": 1.0883293151855469,
+      "learning_rate": 4.942246030744421e-06,
+      "loss": 0.6697,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4373961218836565,
+      "grad_norm": 0.6593711972236633,
+      "learning_rate": 4.942168155606113e-06,
+      "loss": 0.6732,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4376731301939058,
+      "grad_norm": 0.6591055989265442,
+      "learning_rate": 4.942090228614339e-06,
+      "loss": 0.7144,
+      "step": 1580
+    },
+    {
+      "epoch": 0.43795013850415515,
+      "grad_norm": 0.6900238990783691,
+      "learning_rate": 4.942012249770753e-06,
+      "loss": 0.7171,
+      "step": 1581
+    },
+    {
+      "epoch": 0.43822714681440444,
+      "grad_norm": 0.6809273958206177,
+      "learning_rate": 4.94193421907701e-06,
+      "loss": 0.7234,
+      "step": 1582
+    },
+    {
+      "epoch": 0.43850415512465374,
+      "grad_norm": 0.6896465420722961,
+      "learning_rate": 4.941856136534768e-06,
+      "loss": 0.7384,
+      "step": 1583
+    },
+    {
+      "epoch": 0.43878116343490303,
+      "grad_norm": 0.6935831904411316,
+      "learning_rate": 4.941778002145684e-06,
+      "loss": 0.6968,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4390581717451524,
+      "grad_norm": 0.6583067178726196,
+      "learning_rate": 4.941699815911418e-06,
+      "loss": 0.6905,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4393351800554017,
+      "grad_norm": 0.6565580368041992,
+      "learning_rate": 4.941621577833629e-06,
+      "loss": 0.6667,
+      "step": 1586
+    },
+    {
+      "epoch": 0.43961218836565097,
+      "grad_norm": 0.6886636018753052,
+      "learning_rate": 4.941543287913979e-06,
+      "loss": 0.7294,
+      "step": 1587
+    },
+    {
+      "epoch": 0.43988919667590026,
+      "grad_norm": 0.6666744351387024,
+      "learning_rate": 4.94146494615413e-06,
+      "loss": 0.665,
+      "step": 1588
+    },
+    {
+      "epoch": 0.4401662049861496,
+      "grad_norm": 0.6424551606178284,
+      "learning_rate": 4.941386552555745e-06,
+      "loss": 0.6942,
+      "step": 1589
+    },
+    {
+      "epoch": 0.4404432132963989,
+      "grad_norm": 0.6820762157440186,
+      "learning_rate": 4.941308107120489e-06,
+      "loss": 0.7457,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4407202216066482,
+      "grad_norm": 0.6450785994529724,
+      "learning_rate": 4.941229609850028e-06,
+      "loss": 0.6497,
+      "step": 1591
+    },
+    {
+      "epoch": 0.4409972299168975,
+      "grad_norm": 0.6925249099731445,
+      "learning_rate": 4.941151060746028e-06,
+      "loss": 0.7306,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4412742382271468,
+      "grad_norm": 0.6624888181686401,
+      "learning_rate": 4.941072459810157e-06,
+      "loss": 0.7172,
+      "step": 1593
+    },
+    {
+      "epoch": 0.44155124653739614,
+      "grad_norm": 0.699098527431488,
+      "learning_rate": 4.940993807044083e-06,
+      "loss": 0.6957,
+      "step": 1594
+    },
+    {
+      "epoch": 0.44182825484764543,
+      "grad_norm": 0.6670101881027222,
+      "learning_rate": 4.9409151024494775e-06,
+      "loss": 0.7194,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4421052631578947,
+      "grad_norm": 0.6752918362617493,
+      "learning_rate": 4.940836346028011e-06,
+      "loss": 0.7369,
+      "step": 1596
+    },
+    {
+      "epoch": 0.442382271468144,
+      "grad_norm": 0.6847531199455261,
+      "learning_rate": 4.9407575377813555e-06,
+      "loss": 0.6431,
+      "step": 1597
+    },
+    {
+      "epoch": 0.44265927977839337,
+      "grad_norm": 0.6920828223228455,
+      "learning_rate": 4.9406786777111846e-06,
+      "loss": 0.703,
+      "step": 1598
+    },
+    {
+      "epoch": 0.44293628808864266,
+      "grad_norm": 0.6975665092468262,
+      "learning_rate": 4.940599765819172e-06,
+      "loss": 0.6772,
+      "step": 1599
+    },
+    {
+      "epoch": 0.44321329639889195,
+      "grad_norm": 0.7083530426025391,
+      "learning_rate": 4.940520802106994e-06,
+      "loss": 0.7113,
+      "step": 1600
+    },
+    {
+      "epoch": 0.44349030470914125,
+      "grad_norm": 0.6941363215446472,
+      "learning_rate": 4.940441786576328e-06,
+      "loss": 0.7125,
+      "step": 1601
+    },
+    {
+      "epoch": 0.4437673130193906,
+      "grad_norm": 0.6918179392814636,
+      "learning_rate": 4.94036271922885e-06,
+      "loss": 0.7073,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4440443213296399,
+      "grad_norm": 0.7111680507659912,
+      "learning_rate": 4.940283600066239e-06,
+      "loss": 0.7133,
+      "step": 1603
+    },
+    {
+      "epoch": 0.4443213296398892,
+      "grad_norm": 0.6951150298118591,
+      "learning_rate": 4.9402044290901756e-06,
+      "loss": 0.733,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4445983379501385,
+      "grad_norm": 0.7251552939414978,
+      "learning_rate": 4.9401252063023405e-06,
+      "loss": 0.6939,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4448753462603878,
+      "grad_norm": 0.6868443489074707,
+      "learning_rate": 4.9400459317044155e-06,
+      "loss": 0.7334,
+      "step": 1606
+    },
+    {
+      "epoch": 0.4451523545706371,
+      "grad_norm": 0.6709305644035339,
+      "learning_rate": 4.939966605298085e-06,
+      "loss": 0.6766,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4454293628808864,
+      "grad_norm": 0.7305266261100769,
+      "learning_rate": 4.939887227085032e-06,
+      "loss": 0.7549,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4457063711911357,
+      "grad_norm": 0.6449017524719238,
+      "learning_rate": 4.939807797066943e-06,
+      "loss": 0.6886,
+      "step": 1609
+    },
+    {
+      "epoch": 0.44598337950138506,
+      "grad_norm": 0.6638166308403015,
+      "learning_rate": 4.939728315245503e-06,
+      "loss": 0.7256,
+      "step": 1610
+    },
+    {
+      "epoch": 0.44626038781163435,
+      "grad_norm": 0.6473711133003235,
+      "learning_rate": 4.9396487816224014e-06,
+      "loss": 0.684,
+      "step": 1611
+    },
+    {
+      "epoch": 0.44653739612188365,
+      "grad_norm": 0.6618096232414246,
+      "learning_rate": 4.939569196199325e-06,
+      "loss": 0.6679,
+      "step": 1612
+    },
+    {
+      "epoch": 0.44681440443213294,
+      "grad_norm": 0.6737478375434875,
+      "learning_rate": 4.939489558977965e-06,
+      "loss": 0.6957,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4470914127423823,
+      "grad_norm": 0.7004351615905762,
+      "learning_rate": 4.9394098699600125e-06,
+      "loss": 0.7231,
+      "step": 1614
+    },
+    {
+      "epoch": 0.4473684210526316,
+      "grad_norm": 0.658447802066803,
+      "learning_rate": 4.939330129147158e-06,
+      "loss": 0.6537,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4476454293628809,
+      "grad_norm": 0.6831145286560059,
+      "learning_rate": 4.939250336541096e-06,
+      "loss": 0.7153,
+      "step": 1616
+    },
+    {
+      "epoch": 0.44792243767313017,
+      "grad_norm": 0.6539595127105713,
+      "learning_rate": 4.939170492143521e-06,
+      "loss": 0.7445,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4481994459833795,
+      "grad_norm": 0.682756781578064,
+      "learning_rate": 4.9390905959561254e-06,
+      "loss": 0.7261,
+      "step": 1618
+    },
+    {
+      "epoch": 0.4484764542936288,
+      "grad_norm": 0.6738179326057434,
+      "learning_rate": 4.939010647980609e-06,
+      "loss": 0.6729,
+      "step": 1619
+    },
+    {
+      "epoch": 0.4487534626038781,
+      "grad_norm": 0.6528425216674805,
+      "learning_rate": 4.938930648218667e-06,
+      "loss": 0.7388,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4490304709141274,
+      "grad_norm": 0.7098863124847412,
+      "learning_rate": 4.9388505966720006e-06,
+      "loss": 0.6978,
+      "step": 1621
+    },
+    {
+      "epoch": 0.44930747922437675,
+      "grad_norm": 0.6853944659233093,
+      "learning_rate": 4.938770493342307e-06,
+      "loss": 0.7558,
+      "step": 1622
+    },
+    {
+      "epoch": 0.44958448753462604,
+      "grad_norm": 0.7014193534851074,
+      "learning_rate": 4.9386903382312876e-06,
+      "loss": 0.7404,
+      "step": 1623
+    },
+    {
+      "epoch": 0.44986149584487534,
+      "grad_norm": 0.6289151310920715,
+      "learning_rate": 4.9386101313406445e-06,
+      "loss": 0.6741,
+      "step": 1624
+    },
+    {
+      "epoch": 0.45013850415512463,
+      "grad_norm": 0.6713910698890686,
+      "learning_rate": 4.938529872672082e-06,
+      "loss": 0.6464,
+      "step": 1625
+    },
+    {
+      "epoch": 0.450415512465374,
+      "grad_norm": 0.6423241496086121,
+      "learning_rate": 4.938449562227302e-06,
+      "loss": 0.717,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4506925207756233,
+      "grad_norm": 0.6668711304664612,
+      "learning_rate": 4.9383692000080105e-06,
+      "loss": 0.7101,
+      "step": 1627
+    },
+    {
+      "epoch": 0.45096952908587257,
+      "grad_norm": 0.6786472201347351,
+      "learning_rate": 4.9382887860159145e-06,
+      "loss": 0.7295,
+      "step": 1628
+    },
+    {
+      "epoch": 0.45124653739612186,
+      "grad_norm": 0.6926882863044739,
+      "learning_rate": 4.938208320252721e-06,
+      "loss": 0.7362,
+      "step": 1629
+    },
+    {
+      "epoch": 0.4515235457063712,
+      "grad_norm": 0.7050982713699341,
+      "learning_rate": 4.938127802720137e-06,
+      "loss": 0.6786,
+      "step": 1630
+    },
+    {
+      "epoch": 0.4518005540166205,
+      "grad_norm": 0.663986325263977,
+      "learning_rate": 4.938047233419875e-06,
+      "loss": 0.7133,
+      "step": 1631
+    },
+    {
+      "epoch": 0.4520775623268698,
+      "grad_norm": 0.6806838512420654,
+      "learning_rate": 4.937966612353644e-06,
+      "loss": 0.675,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4523545706371191,
+      "grad_norm": 0.6535438895225525,
+      "learning_rate": 4.937885939523155e-06,
+      "loss": 0.6823,
+      "step": 1633
+    },
+    {
+      "epoch": 0.45263157894736844,
+      "grad_norm": 0.6611151695251465,
+      "learning_rate": 4.937805214930123e-06,
+      "loss": 0.6661,
+      "step": 1634
+    },
+    {
+      "epoch": 0.45290858725761773,
+      "grad_norm": 0.6430408358573914,
+      "learning_rate": 4.93772443857626e-06,
+      "loss": 0.6687,
+      "step": 1635
+    },
+    {
+      "epoch": 0.45318559556786703,
+      "grad_norm": 0.6679038405418396,
+      "learning_rate": 4.937643610463282e-06,
+      "loss": 0.7235,
+      "step": 1636
+    },
+    {
+      "epoch": 0.4534626038781163,
+      "grad_norm": 0.6662124395370483,
+      "learning_rate": 4.937562730592905e-06,
+      "loss": 0.6389,
+      "step": 1637
+    },
+    {
+      "epoch": 0.45373961218836567,
+      "grad_norm": 0.6467198729515076,
+      "learning_rate": 4.937481798966847e-06,
+      "loss": 0.6662,
+      "step": 1638
+    },
+    {
+      "epoch": 0.45401662049861496,
+      "grad_norm": 0.6903800964355469,
+      "learning_rate": 4.937400815586825e-06,
+      "loss": 0.7161,
+      "step": 1639
+    },
+    {
+      "epoch": 0.45429362880886426,
+      "grad_norm": 0.6393138766288757,
+      "learning_rate": 4.937319780454559e-06,
+      "loss": 0.6788,
+      "step": 1640
+    },
+    {
+      "epoch": 0.45457063711911355,
+      "grad_norm": 0.6818089485168457,
+      "learning_rate": 4.9372386935717705e-06,
+      "loss": 0.7274,
+      "step": 1641
+    },
+    {
+      "epoch": 0.4548476454293629,
+      "grad_norm": 0.6777812242507935,
+      "learning_rate": 4.937157554940181e-06,
+      "loss": 0.7134,
+      "step": 1642
+    },
+    {
+      "epoch": 0.4551246537396122,
+      "grad_norm": 0.6755929589271545,
+      "learning_rate": 4.937076364561512e-06,
+      "loss": 0.6949,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4554016620498615,
+      "grad_norm": 0.6555960774421692,
+      "learning_rate": 4.936995122437488e-06,
+      "loss": 0.6567,
+      "step": 1644
+    },
+    {
+      "epoch": 0.4556786703601108,
+      "grad_norm": 0.6834290027618408,
+      "learning_rate": 4.9369138285698345e-06,
+      "loss": 0.6743,
+      "step": 1645
+    },
+    {
+      "epoch": 0.45595567867036013,
+      "grad_norm": 0.640485405921936,
+      "learning_rate": 4.936832482960278e-06,
+      "loss": 0.663,
+      "step": 1646
+    },
+    {
+      "epoch": 0.4562326869806094,
+      "grad_norm": 0.6723644137382507,
+      "learning_rate": 4.936751085610543e-06,
+      "loss": 0.6687,
+      "step": 1647
+    },
+    {
+      "epoch": 0.4565096952908587,
+      "grad_norm": 0.7141152620315552,
+      "learning_rate": 4.936669636522361e-06,
+      "loss": 0.6685,
+      "step": 1648
+    },
+    {
+      "epoch": 0.456786703601108,
+      "grad_norm": 0.6581059694290161,
+      "learning_rate": 4.93658813569746e-06,
+      "loss": 0.7371,
+      "step": 1649
+    },
+    {
+      "epoch": 0.45706371191135736,
+      "grad_norm": 0.6734123826026917,
+      "learning_rate": 4.93650658313757e-06,
+      "loss": 0.6696,
+      "step": 1650
+    },
+    {
+      "epoch": 0.45734072022160666,
+      "grad_norm": 0.6516708135604858,
+      "learning_rate": 4.936424978844424e-06,
+      "loss": 0.6508,
+      "step": 1651
+    },
+    {
+      "epoch": 0.45761772853185595,
+      "grad_norm": 0.68487149477005,
+      "learning_rate": 4.9363433228197524e-06,
+      "loss": 0.7158,
+      "step": 1652
+    },
+    {
+      "epoch": 0.45789473684210524,
+      "grad_norm": 0.667688250541687,
+      "learning_rate": 4.936261615065292e-06,
+      "loss": 0.679,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4581717451523546,
+      "grad_norm": 0.7019890546798706,
+      "learning_rate": 4.9361798555827745e-06,
+      "loss": 0.6734,
+      "step": 1654
+    },
+    {
+      "epoch": 0.4584487534626039,
+      "grad_norm": 0.6553559303283691,
+      "learning_rate": 4.936098044373938e-06,
+      "loss": 0.7148,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4587257617728532,
+      "grad_norm": 0.680182933807373,
+      "learning_rate": 4.936016181440518e-06,
+      "loss": 0.6797,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4590027700831025,
+      "grad_norm": 0.6735982298851013,
+      "learning_rate": 4.935934266784254e-06,
+      "loss": 0.7265,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4592797783933518,
+      "grad_norm": 0.6884268522262573,
+      "learning_rate": 4.935852300406886e-06,
+      "loss": 0.7491,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4595567867036011,
+      "grad_norm": 0.7029476761817932,
+      "learning_rate": 4.935770282310152e-06,
+      "loss": 0.6959,
+      "step": 1659
+    },
+    {
+      "epoch": 0.4598337950138504,
+      "grad_norm": 0.6848331093788147,
+      "learning_rate": 4.935688212495795e-06,
+      "loss": 0.7447,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4601108033240997,
+      "grad_norm": 0.6873815655708313,
+      "learning_rate": 4.935606090965557e-06,
+      "loss": 0.7022,
+      "step": 1661
+    },
+    {
+      "epoch": 0.46038781163434905,
+      "grad_norm": 0.6748994588851929,
+      "learning_rate": 4.935523917721182e-06,
+      "loss": 0.732,
+      "step": 1662
+    },
+    {
+      "epoch": 0.46066481994459835,
+      "grad_norm": 0.6672415733337402,
+      "learning_rate": 4.935441692764415e-06,
+      "loss": 0.7035,
+      "step": 1663
+    },
+    {
+      "epoch": 0.46094182825484764,
+      "grad_norm": 0.7236402034759521,
+      "learning_rate": 4.935359416097001e-06,
+      "loss": 0.7004,
+      "step": 1664
+    },
+    {
+      "epoch": 0.46121883656509693,
+      "grad_norm": 0.6981213688850403,
+      "learning_rate": 4.9352770877206865e-06,
+      "loss": 0.686,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4614958448753463,
+      "grad_norm": 0.7025541067123413,
+      "learning_rate": 4.935194707637222e-06,
+      "loss": 0.7322,
+      "step": 1666
+    },
+    {
+      "epoch": 0.4617728531855956,
+      "grad_norm": 0.6887792348861694,
+      "learning_rate": 4.935112275848354e-06,
+      "loss": 0.6834,
+      "step": 1667
+    },
+    {
+      "epoch": 0.46204986149584487,
+      "grad_norm": 0.6790734529495239,
+      "learning_rate": 4.935029792355834e-06,
+      "loss": 0.6889,
+      "step": 1668
+    },
+    {
+      "epoch": 0.46232686980609417,
+      "grad_norm": 0.6590743064880371,
+      "learning_rate": 4.934947257161413e-06,
+      "loss": 0.6941,
+      "step": 1669
+    },
+    {
+      "epoch": 0.4626038781163435,
+      "grad_norm": 0.7474731802940369,
+      "learning_rate": 4.934864670266844e-06,
+      "loss": 0.7576,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4628808864265928,
+      "grad_norm": 0.6757420897483826,
+      "learning_rate": 4.934782031673879e-06,
+      "loss": 0.7377,
+      "step": 1671
+    },
+    {
+      "epoch": 0.4631578947368421,
+      "grad_norm": 0.668341338634491,
+      "learning_rate": 4.9346993413842755e-06,
+      "loss": 0.7288,
+      "step": 1672
+    },
+    {
+      "epoch": 0.4634349030470914,
+      "grad_norm": 0.6508594751358032,
+      "learning_rate": 4.934616599399786e-06,
+      "loss": 0.6765,
+      "step": 1673
+    },
+    {
+      "epoch": 0.46371191135734074,
+      "grad_norm": 0.6187620759010315,
+      "learning_rate": 4.934533805722169e-06,
+      "loss": 0.6626,
+      "step": 1674
+    },
+    {
+      "epoch": 0.46398891966759004,
+      "grad_norm": 0.6785850524902344,
+      "learning_rate": 4.934450960353183e-06,
+      "loss": 0.6788,
+      "step": 1675
+    },
+    {
+      "epoch": 0.46426592797783933,
+      "grad_norm": 0.6822246313095093,
+      "learning_rate": 4.934368063294586e-06,
+      "loss": 0.7357,
+      "step": 1676
+    },
+    {
+      "epoch": 0.4645429362880886,
+      "grad_norm": 0.6765061616897583,
+      "learning_rate": 4.934285114548137e-06,
+      "loss": 0.6875,
+      "step": 1677
+    },
+    {
+      "epoch": 0.464819944598338,
+      "grad_norm": 0.713993489742279,
+      "learning_rate": 4.9342021141156e-06,
+      "loss": 0.7364,
+      "step": 1678
+    },
+    {
+      "epoch": 0.46509695290858727,
+      "grad_norm": 0.7380864024162292,
+      "learning_rate": 4.934119061998735e-06,
+      "loss": 0.6715,
+      "step": 1679
+    },
+    {
+      "epoch": 0.46537396121883656,
+      "grad_norm": 0.6553770303726196,
+      "learning_rate": 4.934035958199307e-06,
+      "loss": 0.6442,
+      "step": 1680
+    },
+    {
+      "epoch": 0.46565096952908586,
+      "grad_norm": 0.6353982090950012,
+      "learning_rate": 4.9339528027190795e-06,
+      "loss": 0.7155,
+      "step": 1681
+    },
+    {
+      "epoch": 0.4659279778393352,
+      "grad_norm": 0.669859766960144,
+      "learning_rate": 4.933869595559818e-06,
+      "loss": 0.6769,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4662049861495845,
+      "grad_norm": 0.6884238719940186,
+      "learning_rate": 4.9337863367232896e-06,
+      "loss": 0.6827,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4664819944598338,
+      "grad_norm": 0.6490418314933777,
+      "learning_rate": 4.933703026211262e-06,
+      "loss": 0.6591,
+      "step": 1684
+    },
+    {
+      "epoch": 0.4667590027700831,
+      "grad_norm": 0.6696502566337585,
+      "learning_rate": 4.933619664025505e-06,
+      "loss": 0.6958,
+      "step": 1685
+    },
+    {
+      "epoch": 0.46703601108033244,
+      "grad_norm": 0.6910261511802673,
+      "learning_rate": 4.9335362501677865e-06,
+      "loss": 0.7157,
+      "step": 1686
+    },
+    {
+      "epoch": 0.46731301939058173,
+      "grad_norm": 0.6383054256439209,
+      "learning_rate": 4.933452784639879e-06,
+      "loss": 0.6681,
+      "step": 1687
+    },
+    {
+      "epoch": 0.467590027700831,
+      "grad_norm": 0.7032589316368103,
+      "learning_rate": 4.933369267443555e-06,
+      "loss": 0.7133,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4678670360110803,
+      "grad_norm": 0.6746159791946411,
+      "learning_rate": 4.933285698580588e-06,
+      "loss": 0.7086,
+      "step": 1689
+    },
+    {
+      "epoch": 0.46814404432132967,
+      "grad_norm": 0.6999416351318359,
+      "learning_rate": 4.93320207805275e-06,
+      "loss": 0.7308,
+      "step": 1690
+    },
+    {
+      "epoch": 0.46842105263157896,
+      "grad_norm": 0.6476268172264099,
+      "learning_rate": 4.93311840586182e-06,
+      "loss": 0.6892,
+      "step": 1691
+    },
+    {
+      "epoch": 0.46869806094182825,
+      "grad_norm": 0.6434255242347717,
+      "learning_rate": 4.933034682009572e-06,
+      "loss": 0.7122,
+      "step": 1692
+    },
+    {
+      "epoch": 0.46897506925207755,
+      "grad_norm": 0.6608331203460693,
+      "learning_rate": 4.932950906497784e-06,
+      "loss": 0.6961,
+      "step": 1693
+    },
+    {
+      "epoch": 0.4692520775623269,
+      "grad_norm": 0.6635152101516724,
+      "learning_rate": 4.932867079328235e-06,
+      "loss": 0.7058,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4695290858725762,
+      "grad_norm": 0.6834961771965027,
+      "learning_rate": 4.932783200502705e-06,
+      "loss": 0.6907,
+      "step": 1695
+    },
+    {
+      "epoch": 0.4698060941828255,
+      "grad_norm": 0.6840958595275879,
+      "learning_rate": 4.932699270022976e-06,
+      "loss": 0.7066,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4700831024930748,
+      "grad_norm": 0.659977912902832,
+      "learning_rate": 4.932615287890828e-06,
+      "loss": 0.6874,
+      "step": 1697
+    },
+    {
+      "epoch": 0.4703601108033241,
+      "grad_norm": 0.6732885837554932,
+      "learning_rate": 4.932531254108046e-06,
+      "loss": 0.7228,
+      "step": 1698
+    },
+    {
+      "epoch": 0.4706371191135734,
+      "grad_norm": 0.7419905662536621,
+      "learning_rate": 4.932447168676414e-06,
+      "loss": 0.6943,
+      "step": 1699
+    },
+    {
+      "epoch": 0.4709141274238227,
+      "grad_norm": 0.6802665591239929,
+      "learning_rate": 4.932363031597715e-06,
+      "loss": 0.7504,
+      "step": 1700
+    },
+    {
+      "epoch": 0.471191135734072,
+      "grad_norm": 0.6577280759811401,
+      "learning_rate": 4.9322788428737385e-06,
+      "loss": 0.7208,
+      "step": 1701
+    },
+    {
+      "epoch": 0.4714681440443213,
+      "grad_norm": 0.6920000910758972,
+      "learning_rate": 4.932194602506271e-06,
+      "loss": 0.7027,
+      "step": 1702
+    },
+    {
+      "epoch": 0.47174515235457065,
+      "grad_norm": 0.6805425882339478,
+      "learning_rate": 4.9321103104971e-06,
+      "loss": 0.7249,
+      "step": 1703
+    },
+    {
+      "epoch": 0.47202216066481995,
+      "grad_norm": 0.6763489842414856,
+      "learning_rate": 4.932025966848017e-06,
+      "loss": 0.7165,
+      "step": 1704
+    },
+    {
+      "epoch": 0.47229916897506924,
+      "grad_norm": 0.6553173661231995,
+      "learning_rate": 4.931941571560812e-06,
+      "loss": 0.6503,
+      "step": 1705
+    },
+    {
+      "epoch": 0.47257617728531853,
+      "grad_norm": 0.6607550382614136,
+      "learning_rate": 4.931857124637276e-06,
+      "loss": 0.6886,
+      "step": 1706
+    },
+    {
+      "epoch": 0.4728531855955679,
+      "grad_norm": 0.6413080096244812,
+      "learning_rate": 4.931772626079204e-06,
+      "loss": 0.6842,
+      "step": 1707
+    },
+    {
+      "epoch": 0.4731301939058172,
+      "grad_norm": 0.6337507367134094,
+      "learning_rate": 4.931688075888389e-06,
+      "loss": 0.6366,
+      "step": 1708
+    },
+    {
+      "epoch": 0.47340720221606647,
+      "grad_norm": 0.6673510074615479,
+      "learning_rate": 4.931603474066626e-06,
+      "loss": 0.6353,
+      "step": 1709
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.651106595993042,
+      "learning_rate": 4.931518820615712e-06,
+      "loss": 0.715,
+      "step": 1710
+    },
+    {
+      "epoch": 0.4739612188365651,
+      "grad_norm": 0.6641266942024231,
+      "learning_rate": 4.9314341155374425e-06,
+      "loss": 0.7022,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4742382271468144,
+      "grad_norm": 0.6641416549682617,
+      "learning_rate": 4.9313493588336196e-06,
+      "loss": 0.6821,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4745152354570637,
+      "grad_norm": 0.6898216009140015,
+      "learning_rate": 4.9312645505060396e-06,
+      "loss": 0.6994,
+      "step": 1713
+    },
+    {
+      "epoch": 0.474792243767313,
+      "grad_norm": 0.66060471534729,
+      "learning_rate": 4.931179690556504e-06,
+      "loss": 0.6867,
+      "step": 1714
+    },
+    {
+      "epoch": 0.47506925207756234,
+      "grad_norm": 0.673663318157196,
+      "learning_rate": 4.931094778986816e-06,
+      "loss": 0.737,
+      "step": 1715
+    },
+    {
+      "epoch": 0.47534626038781164,
+      "grad_norm": 0.695510745048523,
+      "learning_rate": 4.931009815798777e-06,
+      "loss": 0.7038,
+      "step": 1716
+    },
+    {
+      "epoch": 0.47562326869806093,
+      "grad_norm": 0.6945306062698364,
+      "learning_rate": 4.930924800994192e-06,
+      "loss": 0.7442,
+      "step": 1717
+    },
+    {
+      "epoch": 0.4759002770083102,
+      "grad_norm": 0.6963913440704346,
+      "learning_rate": 4.930839734574865e-06,
+      "loss": 0.6816,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4761772853185596,
+      "grad_norm": 0.7170199751853943,
+      "learning_rate": 4.930754616542603e-06,
+      "loss": 0.7057,
+      "step": 1719
+    },
+    {
+      "epoch": 0.47645429362880887,
+      "grad_norm": 0.6967073082923889,
+      "learning_rate": 4.930669446899213e-06,
+      "loss": 0.7028,
+      "step": 1720
+    },
+    {
+      "epoch": 0.47673130193905816,
+      "grad_norm": 0.7398955225944519,
+      "learning_rate": 4.9305842256465035e-06,
+      "loss": 0.7457,
+      "step": 1721
+    },
+    {
+      "epoch": 0.47700831024930745,
+      "grad_norm": 0.6501532196998596,
+      "learning_rate": 4.930498952786284e-06,
+      "loss": 0.6815,
+      "step": 1722
+    },
+    {
+      "epoch": 0.4772853185595568,
+      "grad_norm": 0.6739403605461121,
+      "learning_rate": 4.930413628320364e-06,
+      "loss": 0.7195,
+      "step": 1723
+    },
+    {
+      "epoch": 0.4775623268698061,
+      "grad_norm": 0.7090854048728943,
+      "learning_rate": 4.930328252250557e-06,
+      "loss": 0.7021,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4778393351800554,
+      "grad_norm": 0.6927851438522339,
+      "learning_rate": 4.930242824578675e-06,
+      "loss": 0.6947,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4781163434903047,
+      "grad_norm": 0.6775885224342346,
+      "learning_rate": 4.930157345306531e-06,
+      "loss": 0.6831,
+      "step": 1726
+    },
+    {
+      "epoch": 0.47839335180055403,
+      "grad_norm": 0.6729276776313782,
+      "learning_rate": 4.930071814435941e-06,
+      "loss": 0.7166,
+      "step": 1727
+    },
+    {
+      "epoch": 0.47867036011080333,
+      "grad_norm": 0.6515985727310181,
+      "learning_rate": 4.92998623196872e-06,
+      "loss": 0.6913,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4789473684210526,
+      "grad_norm": 0.6873002052307129,
+      "learning_rate": 4.9299005979066865e-06,
+      "loss": 0.7161,
+      "step": 1729
+    },
+    {
+      "epoch": 0.4792243767313019,
+      "grad_norm": 0.6676026582717896,
+      "learning_rate": 4.929814912251658e-06,
+      "loss": 0.6837,
+      "step": 1730
+    },
+    {
+      "epoch": 0.47950138504155126,
+      "grad_norm": 0.6411203742027283,
+      "learning_rate": 4.929729175005453e-06,
+      "loss": 0.6805,
+      "step": 1731
+    },
+    {
+      "epoch": 0.47977839335180056,
+      "grad_norm": 0.6711311936378479,
+      "learning_rate": 4.929643386169894e-06,
+      "loss": 0.7175,
+      "step": 1732
+    },
+    {
+      "epoch": 0.48005540166204985,
+      "grad_norm": 0.6989825963973999,
+      "learning_rate": 4.9295575457468e-06,
+      "loss": 0.7289,
+      "step": 1733
+    },
+    {
+      "epoch": 0.48033240997229915,
+      "grad_norm": 0.683743953704834,
+      "learning_rate": 4.9294716537379965e-06,
+      "loss": 0.7232,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4806094182825485,
+      "grad_norm": 0.6814141273498535,
+      "learning_rate": 4.929385710145304e-06,
+      "loss": 0.7092,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4808864265927978,
+      "grad_norm": 0.6862533688545227,
+      "learning_rate": 4.92929971497055e-06,
+      "loss": 0.756,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4811634349030471,
+      "grad_norm": 0.7218449115753174,
+      "learning_rate": 4.929213668215559e-06,
+      "loss": 0.7135,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4814404432132964,
+      "grad_norm": 0.683336615562439,
+      "learning_rate": 4.929127569882158e-06,
+      "loss": 0.7294,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4817174515235457,
+      "grad_norm": 0.6259140968322754,
+      "learning_rate": 4.9290414199721755e-06,
+      "loss": 0.6527,
+      "step": 1739
+    },
+    {
+      "epoch": 0.481994459833795,
+      "grad_norm": 0.6623820662498474,
+      "learning_rate": 4.92895521848744e-06,
+      "loss": 0.6523,
+      "step": 1740
+    },
+    {
+      "epoch": 0.4822714681440443,
+      "grad_norm": 0.6865464448928833,
+      "learning_rate": 4.928868965429783e-06,
+      "loss": 0.7055,
+      "step": 1741
+    },
+    {
+      "epoch": 0.4825484764542936,
+      "grad_norm": 0.7001657485961914,
+      "learning_rate": 4.928782660801035e-06,
+      "loss": 0.7136,
+      "step": 1742
+    },
+    {
+      "epoch": 0.48282548476454296,
+      "grad_norm": 0.6874849200248718,
+      "learning_rate": 4.928696304603029e-06,
+      "loss": 0.7043,
+      "step": 1743
+    },
+    {
+      "epoch": 0.48310249307479225,
+      "grad_norm": 0.6774317026138306,
+      "learning_rate": 4.928609896837599e-06,
+      "loss": 0.6533,
+      "step": 1744
+    },
+    {
+      "epoch": 0.48337950138504154,
+      "grad_norm": 0.689127504825592,
+      "learning_rate": 4.928523437506577e-06,
+      "loss": 0.7142,
+      "step": 1745
+    },
+    {
+      "epoch": 0.48365650969529084,
+      "grad_norm": 0.6922279596328735,
+      "learning_rate": 4.928436926611801e-06,
+      "loss": 0.7517,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4839335180055402,
+      "grad_norm": 0.7059768438339233,
+      "learning_rate": 4.928350364155109e-06,
+      "loss": 0.6873,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4842105263157895,
+      "grad_norm": 0.6696379780769348,
+      "learning_rate": 4.9282637501383365e-06,
+      "loss": 0.6752,
+      "step": 1748
+    },
+    {
+      "epoch": 0.4844875346260388,
+      "grad_norm": 0.6618900299072266,
+      "learning_rate": 4.9281770845633235e-06,
+      "loss": 0.6876,
+      "step": 1749
+    },
+    {
+      "epoch": 0.48476454293628807,
+      "grad_norm": 0.6845422387123108,
+      "learning_rate": 4.92809036743191e-06,
+      "loss": 0.7124,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4850415512465374,
+      "grad_norm": 0.6621571779251099,
+      "learning_rate": 4.928003598745938e-06,
+      "loss": 0.6814,
+      "step": 1751
+    },
+    {
+      "epoch": 0.4853185595567867,
+      "grad_norm": 0.6772110462188721,
+      "learning_rate": 4.927916778507248e-06,
+      "loss": 0.6969,
+      "step": 1752
+    },
+    {
+      "epoch": 0.485595567867036,
+      "grad_norm": 0.6477333307266235,
+      "learning_rate": 4.927829906717685e-06,
+      "loss": 0.6031,
+      "step": 1753
+    },
+    {
+      "epoch": 0.4858725761772853,
+      "grad_norm": 0.6633053421974182,
+      "learning_rate": 4.927742983379094e-06,
+      "loss": 0.7203,
+      "step": 1754
+    },
+    {
+      "epoch": 0.48614958448753465,
+      "grad_norm": 0.6834869980812073,
+      "learning_rate": 4.927656008493318e-06,
+      "loss": 0.6837,
+      "step": 1755
+    },
+    {
+      "epoch": 0.48642659279778394,
+      "grad_norm": 0.6382920742034912,
+      "learning_rate": 4.9275689820622065e-06,
+      "loss": 0.7213,
+      "step": 1756
+    },
+    {
+      "epoch": 0.48670360110803323,
+      "grad_norm": 0.6626554727554321,
+      "learning_rate": 4.927481904087606e-06,
+      "loss": 0.7088,
+      "step": 1757
+    },
+    {
+      "epoch": 0.48698060941828253,
+      "grad_norm": 0.6923210620880127,
+      "learning_rate": 4.927394774571366e-06,
+      "loss": 0.6638,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4872576177285319,
+      "grad_norm": 0.6683520674705505,
+      "learning_rate": 4.927307593515335e-06,
+      "loss": 0.6765,
+      "step": 1759
+    },
+    {
+      "epoch": 0.48753462603878117,
+      "grad_norm": 0.6730638146400452,
+      "learning_rate": 4.927220360921366e-06,
+      "loss": 0.7028,
+      "step": 1760
+    },
+    {
+      "epoch": 0.48781163434903047,
+      "grad_norm": 0.6665538549423218,
+      "learning_rate": 4.927133076791309e-06,
+      "loss": 0.6821,
+      "step": 1761
+    },
+    {
+      "epoch": 0.48808864265927976,
+      "grad_norm": 0.6897082924842834,
+      "learning_rate": 4.9270457411270204e-06,
+      "loss": 0.6895,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4883656509695291,
+      "grad_norm": 0.6820263862609863,
+      "learning_rate": 4.926958353930352e-06,
+      "loss": 0.6818,
+      "step": 1763
+    },
+    {
+      "epoch": 0.4886426592797784,
+      "grad_norm": 0.7254486083984375,
+      "learning_rate": 4.92687091520316e-06,
+      "loss": 0.7015,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4889196675900277,
+      "grad_norm": 0.705342710018158,
+      "learning_rate": 4.9267834249473005e-06,
+      "loss": 0.6557,
+      "step": 1765
+    },
+    {
+      "epoch": 0.489196675900277,
+      "grad_norm": 0.6940473914146423,
+      "learning_rate": 4.926695883164632e-06,
+      "loss": 0.718,
+      "step": 1766
+    },
+    {
+      "epoch": 0.48947368421052634,
+      "grad_norm": 0.6987528800964355,
+      "learning_rate": 4.926608289857012e-06,
+      "loss": 0.688,
+      "step": 1767
+    },
+    {
+      "epoch": 0.48975069252077563,
+      "grad_norm": 0.7007743716239929,
+      "learning_rate": 4.926520645026302e-06,
+      "loss": 0.6859,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4900277008310249,
+      "grad_norm": 0.6608262062072754,
+      "learning_rate": 4.9264329486743615e-06,
+      "loss": 0.6685,
+      "step": 1769
+    },
+    {
+      "epoch": 0.4903047091412742,
+      "grad_norm": 0.6854458451271057,
+      "learning_rate": 4.9263452008030535e-06,
+      "loss": 0.7118,
+      "step": 1770
+    },
+    {
+      "epoch": 0.49058171745152357,
+      "grad_norm": 0.6807644963264465,
+      "learning_rate": 4.92625740141424e-06,
+      "loss": 0.6473,
+      "step": 1771
+    },
+    {
+      "epoch": 0.49085872576177286,
+      "grad_norm": 0.6337978839874268,
+      "learning_rate": 4.926169550509787e-06,
+      "loss": 0.6851,
+      "step": 1772
+    },
+    {
+      "epoch": 0.49113573407202216,
+      "grad_norm": 0.672011137008667,
+      "learning_rate": 4.926081648091559e-06,
+      "loss": 0.6875,
+      "step": 1773
+    },
+    {
+      "epoch": 0.49141274238227145,
+      "grad_norm": 0.7125269770622253,
+      "learning_rate": 4.925993694161421e-06,
+      "loss": 0.6748,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4916897506925208,
+      "grad_norm": 0.6711752414703369,
+      "learning_rate": 4.925905688721241e-06,
+      "loss": 0.6989,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4919667590027701,
+      "grad_norm": 0.6642163991928101,
+      "learning_rate": 4.925817631772889e-06,
+      "loss": 0.6894,
+      "step": 1776
+    },
+    {
+      "epoch": 0.4922437673130194,
+      "grad_norm": 0.6630699634552002,
+      "learning_rate": 4.925729523318233e-06,
+      "loss": 0.712,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4925207756232687,
+      "grad_norm": 0.6922430992126465,
+      "learning_rate": 4.925641363359146e-06,
+      "loss": 0.6625,
+      "step": 1778
+    },
+    {
+      "epoch": 0.49279778393351803,
+      "grad_norm": 0.6554149389266968,
+      "learning_rate": 4.925553151897497e-06,
+      "loss": 0.6949,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4930747922437673,
+      "grad_norm": 0.6485135555267334,
+      "learning_rate": 4.925464888935162e-06,
+      "loss": 0.6487,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4933518005540166,
+      "grad_norm": 0.7048740386962891,
+      "learning_rate": 4.925376574474013e-06,
+      "loss": 0.6684,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4936288088642659,
+      "grad_norm": 0.6761953234672546,
+      "learning_rate": 4.925288208515924e-06,
+      "loss": 0.6866,
+      "step": 1782
+    },
+    {
+      "epoch": 0.49390581717451526,
+      "grad_norm": 0.7035673260688782,
+      "learning_rate": 4.925199791062774e-06,
+      "loss": 0.7412,
+      "step": 1783
+    },
+    {
+      "epoch": 0.49418282548476455,
+      "grad_norm": 0.6940863728523254,
+      "learning_rate": 4.925111322116439e-06,
+      "loss": 0.6624,
+      "step": 1784
+    },
+    {
+      "epoch": 0.49445983379501385,
+      "grad_norm": 0.6827403903007507,
+      "learning_rate": 4.925022801678798e-06,
+      "loss": 0.6649,
+      "step": 1785
+    },
+    {
+      "epoch": 0.49473684210526314,
+      "grad_norm": 0.7142612338066101,
+      "learning_rate": 4.92493422975173e-06,
+      "loss": 0.6943,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4950138504155125,
+      "grad_norm": 0.6687315702438354,
+      "learning_rate": 4.924845606337115e-06,
+      "loss": 0.6855,
+      "step": 1787
+    },
+    {
+      "epoch": 0.4952908587257618,
+      "grad_norm": 0.6995869874954224,
+      "learning_rate": 4.924756931436836e-06,
+      "loss": 0.6714,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4955678670360111,
+      "grad_norm": 0.7208627462387085,
+      "learning_rate": 4.924668205052775e-06,
+      "loss": 0.6602,
+      "step": 1789
+    },
+    {
+      "epoch": 0.49584487534626037,
+      "grad_norm": 0.7047656774520874,
+      "learning_rate": 4.924579427186817e-06,
+      "loss": 0.7161,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4961218836565097,
+      "grad_norm": 0.6512801051139832,
+      "learning_rate": 4.924490597840844e-06,
+      "loss": 0.6679,
+      "step": 1791
+    },
+    {
+      "epoch": 0.496398891966759,
+      "grad_norm": 0.6830857992172241,
+      "learning_rate": 4.924401717016746e-06,
+      "loss": 0.6823,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4966759002770083,
+      "grad_norm": 0.6951817870140076,
+      "learning_rate": 4.924312784716407e-06,
+      "loss": 0.6831,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4969529085872576,
+      "grad_norm": 0.6855754256248474,
+      "learning_rate": 4.924223800941718e-06,
+      "loss": 0.6707,
+      "step": 1794
+    },
+    {
+      "epoch": 0.49722991689750695,
+      "grad_norm": 0.7017086744308472,
+      "learning_rate": 4.924134765694566e-06,
+      "loss": 0.658,
+      "step": 1795
+    },
+    {
+      "epoch": 0.49750692520775625,
+      "grad_norm": 0.7068955302238464,
+      "learning_rate": 4.924045678976842e-06,
+      "loss": 0.7202,
+      "step": 1796
+    },
+    {
+      "epoch": 0.49778393351800554,
+      "grad_norm": 0.6873322129249573,
+      "learning_rate": 4.9239565407904385e-06,
+      "loss": 0.7351,
+      "step": 1797
+    },
+    {
+      "epoch": 0.49806094182825483,
+      "grad_norm": 0.6638666987419128,
+      "learning_rate": 4.923867351137247e-06,
+      "loss": 0.7046,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4983379501385042,
+      "grad_norm": 0.680953323841095,
+      "learning_rate": 4.923778110019163e-06,
+      "loss": 0.6932,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4986149584487535,
+      "grad_norm": 0.715457022190094,
+      "learning_rate": 4.923688817438079e-06,
+      "loss": 0.7159,
+      "step": 1800
+    },
+    {
+      "epoch": 0.49889196675900277,
+      "grad_norm": 0.6973574161529541,
+      "learning_rate": 4.923599473395892e-06,
+      "loss": 0.6758,
+      "step": 1801
+    },
+    {
+      "epoch": 0.49916897506925206,
+      "grad_norm": 0.6752958297729492,
+      "learning_rate": 4.9235100778944985e-06,
+      "loss": 0.7031,
+      "step": 1802
+    },
+    {
+      "epoch": 0.4994459833795014,
+      "grad_norm": 0.6299696564674377,
+      "learning_rate": 4.923420630935798e-06,
+      "loss": 0.6299,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4997229916897507,
+      "grad_norm": 0.6795063018798828,
+      "learning_rate": 4.923331132521689e-06,
+      "loss": 0.6879,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6836988925933838,
+      "learning_rate": 4.923241582654071e-06,
+      "loss": 0.6935,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5002770083102493,
+      "grad_norm": 0.6711650490760803,
+      "learning_rate": 4.9231519813348464e-06,
+      "loss": 0.6667,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5005540166204986,
+      "grad_norm": 0.6842010617256165,
+      "learning_rate": 4.923062328565916e-06,
+      "loss": 0.667,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5008310249307479,
+      "grad_norm": 0.6528140902519226,
+      "learning_rate": 4.922972624349185e-06,
+      "loss": 0.6881,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5011080332409972,
+      "grad_norm": 0.7012379765510559,
+      "learning_rate": 4.922882868686558e-06,
+      "loss": 0.7278,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5013850415512465,
+      "grad_norm": 0.6806972622871399,
+      "learning_rate": 4.92279306157994e-06,
+      "loss": 0.7221,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5016620498614959,
+      "grad_norm": 0.7014011740684509,
+      "learning_rate": 4.922703203031239e-06,
+      "loss": 0.6982,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5019390581717451,
+      "grad_norm": 0.6584193706512451,
+      "learning_rate": 4.922613293042361e-06,
+      "loss": 0.6485,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5022160664819945,
+      "grad_norm": 0.6598594188690186,
+      "learning_rate": 4.922523331615217e-06,
+      "loss": 0.7386,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5024930747922438,
+      "grad_norm": 0.7168267965316772,
+      "learning_rate": 4.922433318751716e-06,
+      "loss": 0.7082,
+      "step": 1814
+    },
+    {
+      "epoch": 0.502770083102493,
+      "grad_norm": 0.6848852038383484,
+      "learning_rate": 4.922343254453769e-06,
+      "loss": 0.7277,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5030470914127424,
+      "grad_norm": 0.683768630027771,
+      "learning_rate": 4.9222531387232885e-06,
+      "loss": 0.6798,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5033240997229916,
+      "grad_norm": 0.6405003070831299,
+      "learning_rate": 4.922162971562189e-06,
+      "loss": 0.6653,
+      "step": 1817
+    },
+    {
+      "epoch": 0.503601108033241,
+      "grad_norm": 0.6444788575172424,
+      "learning_rate": 4.922072752972383e-06,
+      "loss": 0.6698,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5038781163434903,
+      "grad_norm": 0.6805223822593689,
+      "learning_rate": 4.921982482955788e-06,
+      "loss": 0.7485,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5041551246537396,
+      "grad_norm": 0.6668921709060669,
+      "learning_rate": 4.921892161514319e-06,
+      "loss": 0.6912,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5044321329639889,
+      "grad_norm": 0.663844883441925,
+      "learning_rate": 4.9218017886498955e-06,
+      "loss": 0.681,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5047091412742383,
+      "grad_norm": 0.6882054805755615,
+      "learning_rate": 4.921711364364435e-06,
+      "loss": 0.6901,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5049861495844875,
+      "grad_norm": 0.6432827115058899,
+      "learning_rate": 4.921620888659858e-06,
+      "loss": 0.6772,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5052631578947369,
+      "grad_norm": 0.6826911568641663,
+      "learning_rate": 4.921530361538086e-06,
+      "loss": 0.6965,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5055401662049861,
+      "grad_norm": 0.6841583847999573,
+      "learning_rate": 4.921439783001039e-06,
+      "loss": 0.7128,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5058171745152354,
+      "grad_norm": 0.6850943565368652,
+      "learning_rate": 4.921349153050643e-06,
+      "loss": 0.6691,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5060941828254848,
+      "grad_norm": 0.679665744304657,
+      "learning_rate": 4.92125847168882e-06,
+      "loss": 0.7019,
+      "step": 1827
+    },
+    {
+      "epoch": 0.506371191135734,
+      "grad_norm": 0.6384556293487549,
+      "learning_rate": 4.921167738917497e-06,
+      "loss": 0.6531,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5066481994459834,
+      "grad_norm": 0.6422626972198486,
+      "learning_rate": 4.9210769547386e-06,
+      "loss": 0.6471,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5069252077562327,
+      "grad_norm": 0.6919006109237671,
+      "learning_rate": 4.920986119154057e-06,
+      "loss": 0.6955,
+      "step": 1830
+    },
+    {
+      "epoch": 0.507202216066482,
+      "grad_norm": 0.6437128782272339,
+      "learning_rate": 4.920895232165795e-06,
+      "loss": 0.6925,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5074792243767313,
+      "grad_norm": 0.6609724164009094,
+      "learning_rate": 4.920804293775746e-06,
+      "loss": 0.691,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5077562326869806,
+      "grad_norm": 0.6840398907661438,
+      "learning_rate": 4.920713303985839e-06,
+      "loss": 0.6822,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5080332409972299,
+      "grad_norm": 0.6742506623268127,
+      "learning_rate": 4.920622262798007e-06,
+      "loss": 0.7082,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5083102493074793,
+      "grad_norm": 0.6945616602897644,
+      "learning_rate": 4.920531170214183e-06,
+      "loss": 0.6722,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5085872576177285,
+      "grad_norm": 1.1759380102157593,
+      "learning_rate": 4.920440026236301e-06,
+      "loss": 0.6814,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5088642659279778,
+      "grad_norm": 0.72176593542099,
+      "learning_rate": 4.920348830866296e-06,
+      "loss": 0.6785,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5091412742382272,
+      "grad_norm": 0.6691416501998901,
+      "learning_rate": 4.920257584106104e-06,
+      "loss": 0.7269,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5094182825484764,
+      "grad_norm": 0.7100477814674377,
+      "learning_rate": 4.9201662859576626e-06,
+      "loss": 0.6908,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5096952908587258,
+      "grad_norm": 0.658385694026947,
+      "learning_rate": 4.920074936422911e-06,
+      "loss": 0.6894,
+      "step": 1840
+    },
+    {
+      "epoch": 0.509972299168975,
+      "grad_norm": 0.6802435517311096,
+      "learning_rate": 4.9199835355037885e-06,
+      "loss": 0.7067,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5102493074792244,
+      "grad_norm": 0.6722739338874817,
+      "learning_rate": 4.919892083202235e-06,
+      "loss": 0.689,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5105263157894737,
+      "grad_norm": 0.671492338180542,
+      "learning_rate": 4.919800579520193e-06,
+      "loss": 0.6686,
+      "step": 1843
+    },
+    {
+      "epoch": 0.510803324099723,
+      "grad_norm": 0.6547128558158875,
+      "learning_rate": 4.919709024459605e-06,
+      "loss": 0.703,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5110803324099723,
+      "grad_norm": 0.6681481003761292,
+      "learning_rate": 4.9196174180224165e-06,
+      "loss": 0.6717,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5113573407202217,
+      "grad_norm": 0.6683051586151123,
+      "learning_rate": 4.91952576021057e-06,
+      "loss": 0.6832,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5116343490304709,
+      "grad_norm": 0.6824548840522766,
+      "learning_rate": 4.919434051026013e-06,
+      "loss": 0.6598,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5119113573407202,
+      "grad_norm": 0.6966956257820129,
+      "learning_rate": 4.919342290470692e-06,
+      "loss": 0.7482,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5121883656509695,
+      "grad_norm": 0.7106775641441345,
+      "learning_rate": 4.919250478546556e-06,
+      "loss": 0.6644,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5124653739612188,
+      "grad_norm": 0.6864263415336609,
+      "learning_rate": 4.9191586152555555e-06,
+      "loss": 0.6907,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5127423822714682,
+      "grad_norm": 0.6232467889785767,
+      "learning_rate": 4.919066700599639e-06,
+      "loss": 0.6006,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5130193905817174,
+      "grad_norm": 0.6870824098587036,
+      "learning_rate": 4.918974734580759e-06,
+      "loss": 0.6423,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5132963988919668,
+      "grad_norm": 0.6564236879348755,
+      "learning_rate": 4.9188827172008674e-06,
+      "loss": 0.7023,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5135734072022161,
+      "grad_norm": 0.6429625153541565,
+      "learning_rate": 4.918790648461919e-06,
+      "loss": 0.6377,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5138504155124654,
+      "grad_norm": 0.6833292841911316,
+      "learning_rate": 4.918698528365868e-06,
+      "loss": 0.7047,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5141274238227147,
+      "grad_norm": 0.6977648138999939,
+      "learning_rate": 4.918606356914671e-06,
+      "loss": 0.6748,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5144044321329639,
+      "grad_norm": 0.6762489676475525,
+      "learning_rate": 4.918514134110284e-06,
+      "loss": 0.7136,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5146814404432133,
+      "grad_norm": 0.719980001449585,
+      "learning_rate": 4.918421859954666e-06,
+      "loss": 0.678,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5149584487534626,
+      "grad_norm": 0.6446436047554016,
+      "learning_rate": 4.918329534449776e-06,
+      "loss": 0.682,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5152354570637119,
+      "grad_norm": 0.6879029273986816,
+      "learning_rate": 4.918237157597574e-06,
+      "loss": 0.6936,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5155124653739612,
+      "grad_norm": 0.6882273554801941,
+      "learning_rate": 4.918144729400022e-06,
+      "loss": 0.6932,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5157894736842106,
+      "grad_norm": 0.660173237323761,
+      "learning_rate": 4.9180522498590814e-06,
+      "loss": 0.6947,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5160664819944598,
+      "grad_norm": 0.682055652141571,
+      "learning_rate": 4.917959718976717e-06,
+      "loss": 0.7055,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5163434903047092,
+      "grad_norm": 0.7079998850822449,
+      "learning_rate": 4.917867136754894e-06,
+      "loss": 0.6849,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5166204986149584,
+      "grad_norm": 0.6993677616119385,
+      "learning_rate": 4.917774503195576e-06,
+      "loss": 0.7021,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5168975069252078,
+      "grad_norm": 0.6869961619377136,
+      "learning_rate": 4.91768181830073e-06,
+      "loss": 0.7318,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5171745152354571,
+      "grad_norm": 0.6560007929801941,
+      "learning_rate": 4.917589082072327e-06,
+      "loss": 0.6704,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5174515235457063,
+      "grad_norm": 0.6822911500930786,
+      "learning_rate": 4.917496294512333e-06,
+      "loss": 0.7208,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5177285318559557,
+      "grad_norm": 0.6258894801139832,
+      "learning_rate": 4.917403455622718e-06,
+      "loss": 0.6708,
+      "step": 1869
+    },
+    {
+      "epoch": 0.518005540166205,
+      "grad_norm": 0.6495811939239502,
+      "learning_rate": 4.917310565405456e-06,
+      "loss": 0.6872,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5182825484764543,
+      "grad_norm": 0.6962676644325256,
+      "learning_rate": 4.917217623862516e-06,
+      "loss": 0.7156,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5185595567867036,
+      "grad_norm": 0.6337746977806091,
+      "learning_rate": 4.917124630995874e-06,
+      "loss": 0.6518,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5188365650969529,
+      "grad_norm": 0.6667745113372803,
+      "learning_rate": 4.917031586807503e-06,
+      "loss": 0.7224,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5191135734072022,
+      "grad_norm": 0.6814945936203003,
+      "learning_rate": 4.91693849129938e-06,
+      "loss": 0.6924,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5193905817174516,
+      "grad_norm": 0.7064763903617859,
+      "learning_rate": 4.916845344473479e-06,
+      "loss": 0.6947,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5196675900277008,
+      "grad_norm": 0.7004358768463135,
+      "learning_rate": 4.91675214633178e-06,
+      "loss": 0.6945,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5199445983379501,
+      "grad_norm": 0.6734568476676941,
+      "learning_rate": 4.916658896876262e-06,
+      "loss": 0.7157,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5202216066481995,
+      "grad_norm": 0.6627720594406128,
+      "learning_rate": 4.916565596108903e-06,
+      "loss": 0.6442,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5204986149584487,
+      "grad_norm": 0.666963517665863,
+      "learning_rate": 4.916472244031686e-06,
+      "loss": 0.6916,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5207756232686981,
+      "grad_norm": 0.7252178192138672,
+      "learning_rate": 4.916378840646592e-06,
+      "loss": 0.6777,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5210526315789473,
+      "grad_norm": 0.6848912835121155,
+      "learning_rate": 4.9162853859556044e-06,
+      "loss": 0.6795,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5213296398891967,
+      "grad_norm": 0.6617556214332581,
+      "learning_rate": 4.916191879960708e-06,
+      "loss": 0.6869,
+      "step": 1882
+    },
+    {
+      "epoch": 0.521606648199446,
+      "grad_norm": 0.6472238898277283,
+      "learning_rate": 4.916098322663887e-06,
+      "loss": 0.6688,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5218836565096953,
+      "grad_norm": 0.6495205163955688,
+      "learning_rate": 4.916004714067128e-06,
+      "loss": 0.6866,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5221606648199446,
+      "grad_norm": 0.7360695004463196,
+      "learning_rate": 4.915911054172421e-06,
+      "loss": 0.6491,
+      "step": 1885
+    },
+    {
+      "epoch": 0.522437673130194,
+      "grad_norm": 0.6612504720687866,
+      "learning_rate": 4.915817342981751e-06,
+      "loss": 0.7256,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5227146814404432,
+      "grad_norm": 0.6689541935920715,
+      "learning_rate": 4.915723580497111e-06,
+      "loss": 0.6962,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5229916897506925,
+      "grad_norm": 0.7215927243232727,
+      "learning_rate": 4.915629766720489e-06,
+      "loss": 0.684,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5232686980609418,
+      "grad_norm": 0.6565717458724976,
+      "learning_rate": 4.915535901653879e-06,
+      "loss": 0.7088,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5235457063711911,
+      "grad_norm": 0.6787716150283813,
+      "learning_rate": 4.9154419852992725e-06,
+      "loss": 0.6987,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5238227146814405,
+      "grad_norm": 0.6738946437835693,
+      "learning_rate": 4.915348017658666e-06,
+      "loss": 0.7169,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5240997229916897,
+      "grad_norm": 0.6366645693778992,
+      "learning_rate": 4.915253998734051e-06,
+      "loss": 0.7423,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5243767313019391,
+      "grad_norm": 0.6482120752334595,
+      "learning_rate": 4.9151599285274265e-06,
+      "loss": 0.7022,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5246537396121884,
+      "grad_norm": 0.625254213809967,
+      "learning_rate": 4.915065807040789e-06,
+      "loss": 0.6335,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5249307479224377,
+      "grad_norm": 0.6547906398773193,
+      "learning_rate": 4.914971634276137e-06,
+      "loss": 0.6519,
+      "step": 1895
+    },
+    {
+      "epoch": 0.525207756232687,
+      "grad_norm": 0.6612831354141235,
+      "learning_rate": 4.914877410235469e-06,
+      "loss": 0.713,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5254847645429362,
+      "grad_norm": 0.6797193884849548,
+      "learning_rate": 4.914783134920788e-06,
+      "loss": 0.6793,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5257617728531856,
+      "grad_norm": 0.665197491645813,
+      "learning_rate": 4.914688808334094e-06,
+      "loss": 0.6594,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5260387811634349,
+      "grad_norm": 0.6539757251739502,
+      "learning_rate": 4.914594430477391e-06,
+      "loss": 0.7063,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.6706299185752869,
+      "learning_rate": 4.91450000135268e-06,
+      "loss": 0.6764,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5265927977839335,
+      "grad_norm": 0.6564507484436035,
+      "learning_rate": 4.91440552096197e-06,
+      "loss": 0.6905,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5268698060941829,
+      "grad_norm": 0.6501426696777344,
+      "learning_rate": 4.914310989307264e-06,
+      "loss": 0.7434,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5271468144044321,
+      "grad_norm": 0.6355409026145935,
+      "learning_rate": 4.914216406390571e-06,
+      "loss": 0.7126,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5274238227146815,
+      "grad_norm": 0.7012143731117249,
+      "learning_rate": 4.914121772213898e-06,
+      "loss": 0.7368,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5277008310249307,
+      "grad_norm": 0.678661584854126,
+      "learning_rate": 4.914027086779254e-06,
+      "loss": 0.6804,
+      "step": 1905
+    },
+    {
+      "epoch": 0.52797783933518,
+      "grad_norm": 0.6787761449813843,
+      "learning_rate": 4.913932350088652e-06,
+      "loss": 0.7025,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5282548476454294,
+      "grad_norm": 0.6858866810798645,
+      "learning_rate": 4.9138375621441e-06,
+      "loss": 0.7017,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5285318559556786,
+      "grad_norm": 0.7146446704864502,
+      "learning_rate": 4.913742722947613e-06,
+      "loss": 0.7135,
+      "step": 1908
+    },
+    {
+      "epoch": 0.528808864265928,
+      "grad_norm": 0.6727023720741272,
+      "learning_rate": 4.9136478325012035e-06,
+      "loss": 0.6683,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5290858725761773,
+      "grad_norm": 0.648564338684082,
+      "learning_rate": 4.913552890806888e-06,
+      "loss": 0.7056,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5293628808864266,
+      "grad_norm": 0.678922712802887,
+      "learning_rate": 4.9134578978666796e-06,
+      "loss": 0.6711,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5296398891966759,
+      "grad_norm": 0.6676106452941895,
+      "learning_rate": 4.913362853682597e-06,
+      "loss": 0.6928,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5299168975069252,
+      "grad_norm": 0.68855220079422,
+      "learning_rate": 4.913267758256659e-06,
+      "loss": 0.713,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5301939058171745,
+      "grad_norm": 0.6664445996284485,
+      "learning_rate": 4.913172611590882e-06,
+      "loss": 0.6194,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5304709141274239,
+      "grad_norm": 0.7783384919166565,
+      "learning_rate": 4.913077413687289e-06,
+      "loss": 0.7466,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5307479224376731,
+      "grad_norm": 0.6518760323524475,
+      "learning_rate": 4.9129821645479e-06,
+      "loss": 0.6592,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5310249307479225,
+      "grad_norm": 0.6837249398231506,
+      "learning_rate": 4.912886864174738e-06,
+      "loss": 0.6914,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5313019390581717,
+      "grad_norm": 0.6892364621162415,
+      "learning_rate": 4.912791512569826e-06,
+      "loss": 0.6914,
+      "step": 1918
+    },
+    {
+      "epoch": 0.531578947368421,
+      "grad_norm": 0.6553475856781006,
+      "learning_rate": 4.912696109735188e-06,
+      "loss": 0.6355,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5318559556786704,
+      "grad_norm": 0.6912408471107483,
+      "learning_rate": 4.91260065567285e-06,
+      "loss": 0.6365,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5321329639889196,
+      "grad_norm": 0.6865231394767761,
+      "learning_rate": 4.9125051503848395e-06,
+      "loss": 0.7024,
+      "step": 1921
+    },
+    {
+      "epoch": 0.532409972299169,
+      "grad_norm": 0.6905850172042847,
+      "learning_rate": 4.912409593873184e-06,
+      "loss": 0.7336,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5326869806094183,
+      "grad_norm": 0.6499538421630859,
+      "learning_rate": 4.912313986139911e-06,
+      "loss": 0.7127,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5329639889196676,
+      "grad_norm": 0.6666481494903564,
+      "learning_rate": 4.912218327187053e-06,
+      "loss": 0.6548,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5332409972299169,
+      "grad_norm": 0.7385328412055969,
+      "learning_rate": 4.91212261701664e-06,
+      "loss": 0.7113,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5335180055401662,
+      "grad_norm": 0.710387110710144,
+      "learning_rate": 4.912026855630703e-06,
+      "loss": 0.7185,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5337950138504155,
+      "grad_norm": 0.6992027163505554,
+      "learning_rate": 4.911931043031276e-06,
+      "loss": 0.6737,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5340720221606648,
+      "grad_norm": 0.6654460430145264,
+      "learning_rate": 4.9118351792203945e-06,
+      "loss": 0.6781,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5343490304709141,
+      "grad_norm": 0.6434751749038696,
+      "learning_rate": 4.911739264200093e-06,
+      "loss": 0.7154,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5346260387811634,
+      "grad_norm": 0.6851822137832642,
+      "learning_rate": 4.911643297972407e-06,
+      "loss": 0.6425,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5349030470914128,
+      "grad_norm": 0.6717454195022583,
+      "learning_rate": 4.911547280539376e-06,
+      "loss": 0.7232,
+      "step": 1931
+    },
+    {
+      "epoch": 0.535180055401662,
+      "grad_norm": 0.6625047326087952,
+      "learning_rate": 4.9114512119030376e-06,
+      "loss": 0.6991,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5354570637119114,
+      "grad_norm": 0.6835640072822571,
+      "learning_rate": 4.911355092065432e-06,
+      "loss": 0.6726,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5357340720221606,
+      "grad_norm": 0.6565011143684387,
+      "learning_rate": 4.9112589210285995e-06,
+      "loss": 0.705,
+      "step": 1934
+    },
+    {
+      "epoch": 0.53601108033241,
+      "grad_norm": 0.6798642873764038,
+      "learning_rate": 4.911162698794583e-06,
+      "loss": 0.7011,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5362880886426593,
+      "grad_norm": 0.7038365602493286,
+      "learning_rate": 4.9110664253654235e-06,
+      "loss": 0.6952,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5365650969529085,
+      "grad_norm": 0.691293478012085,
+      "learning_rate": 4.910970100743168e-06,
+      "loss": 0.7138,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5368421052631579,
+      "grad_norm": 0.6723795533180237,
+      "learning_rate": 4.91087372492986e-06,
+      "loss": 0.676,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5371191135734072,
+      "grad_norm": 0.6425924897193909,
+      "learning_rate": 4.910777297927546e-06,
+      "loss": 0.6654,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5373961218836565,
+      "grad_norm": 0.6891401410102844,
+      "learning_rate": 4.910680819738274e-06,
+      "loss": 0.7162,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5376731301939058,
+      "grad_norm": 0.6883473992347717,
+      "learning_rate": 4.910584290364091e-06,
+      "loss": 0.6866,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5379501385041551,
+      "grad_norm": 0.6681075096130371,
+      "learning_rate": 4.910487709807048e-06,
+      "loss": 0.6435,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5382271468144044,
+      "grad_norm": 0.6832405924797058,
+      "learning_rate": 4.9103910780691955e-06,
+      "loss": 0.6804,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5385041551246538,
+      "grad_norm": 0.6717491745948792,
+      "learning_rate": 4.910294395152585e-06,
+      "loss": 0.6988,
+      "step": 1944
+    },
+    {
+      "epoch": 0.538781163434903,
+      "grad_norm": 0.7094027400016785,
+      "learning_rate": 4.9101976610592685e-06,
+      "loss": 0.6964,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5390581717451524,
+      "grad_norm": 0.6554908752441406,
+      "learning_rate": 4.910100875791301e-06,
+      "loss": 0.6476,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5393351800554017,
+      "grad_norm": 0.6950293779373169,
+      "learning_rate": 4.910004039350737e-06,
+      "loss": 0.6839,
+      "step": 1947
+    },
+    {
+      "epoch": 0.539612188365651,
+      "grad_norm": 0.6729973554611206,
+      "learning_rate": 4.909907151739634e-06,
+      "loss": 0.6865,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5398891966759003,
+      "grad_norm": 0.6982777714729309,
+      "learning_rate": 4.909810212960047e-06,
+      "loss": 0.7073,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5401662049861495,
+      "grad_norm": 0.6632289886474609,
+      "learning_rate": 4.9097132230140345e-06,
+      "loss": 0.6576,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5404432132963989,
+      "grad_norm": 0.7004045844078064,
+      "learning_rate": 4.9096161819036585e-06,
+      "loss": 0.6677,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5407202216066482,
+      "grad_norm": 0.6294344663619995,
+      "learning_rate": 4.9095190896309755e-06,
+      "loss": 0.6924,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5409972299168975,
+      "grad_norm": 0.6689262390136719,
+      "learning_rate": 4.90942194619805e-06,
+      "loss": 0.726,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5412742382271468,
+      "grad_norm": 0.6792681813240051,
+      "learning_rate": 4.909324751606943e-06,
+      "loss": 0.6986,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5415512465373962,
+      "grad_norm": 0.6492560505867004,
+      "learning_rate": 4.9092275058597195e-06,
+      "loss": 0.6953,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5418282548476454,
+      "grad_norm": 0.6609320044517517,
+      "learning_rate": 4.909130208958443e-06,
+      "loss": 0.6908,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5421052631578948,
+      "grad_norm": 0.6573171615600586,
+      "learning_rate": 4.909032860905179e-06,
+      "loss": 0.635,
+      "step": 1957
+    },
+    {
+      "epoch": 0.542382271468144,
+      "grad_norm": 0.6838486194610596,
+      "learning_rate": 4.9089354617019975e-06,
+      "loss": 0.7012,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5426592797783933,
+      "grad_norm": 0.6595408916473389,
+      "learning_rate": 4.908838011350963e-06,
+      "loss": 0.6127,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5429362880886427,
+      "grad_norm": 0.6571938395500183,
+      "learning_rate": 4.908740509854146e-06,
+      "loss": 0.6551,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5432132963988919,
+      "grad_norm": 0.6965537071228027,
+      "learning_rate": 4.908642957213617e-06,
+      "loss": 0.6995,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5434903047091413,
+      "grad_norm": 0.6637973189353943,
+      "learning_rate": 4.9085453534314474e-06,
+      "loss": 0.6816,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5437673130193906,
+      "grad_norm": 0.6605237722396851,
+      "learning_rate": 4.908447698509709e-06,
+      "loss": 0.6722,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5440443213296399,
+      "grad_norm": 0.6632689833641052,
+      "learning_rate": 4.908349992450475e-06,
+      "loss": 0.6715,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5443213296398892,
+      "grad_norm": 0.6106252670288086,
+      "learning_rate": 4.9082522352558215e-06,
+      "loss": 0.6358,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5445983379501385,
+      "grad_norm": 0.6894325613975525,
+      "learning_rate": 4.9081544269278225e-06,
+      "loss": 0.7247,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5448753462603878,
+      "grad_norm": 0.6844300627708435,
+      "learning_rate": 4.908056567468554e-06,
+      "loss": 0.6752,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5451523545706372,
+      "grad_norm": 0.671852171421051,
+      "learning_rate": 4.907958656880097e-06,
+      "loss": 0.6932,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5454293628808864,
+      "grad_norm": 0.6739179491996765,
+      "learning_rate": 4.907860695164527e-06,
+      "loss": 0.6638,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5457063711911357,
+      "grad_norm": 0.6991792321205139,
+      "learning_rate": 4.907762682323926e-06,
+      "loss": 0.6845,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5459833795013851,
+      "grad_norm": 0.6290937662124634,
+      "learning_rate": 4.907664618360375e-06,
+      "loss": 0.6629,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5462603878116343,
+      "grad_norm": 0.7173877954483032,
+      "learning_rate": 4.907566503275955e-06,
+      "loss": 0.6929,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5465373961218837,
+      "grad_norm": 0.667913556098938,
+      "learning_rate": 4.90746833707275e-06,
+      "loss": 0.6246,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5468144044321329,
+      "grad_norm": 0.6834477782249451,
+      "learning_rate": 4.907370119752844e-06,
+      "loss": 0.7205,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5470914127423823,
+      "grad_norm": 0.7111289501190186,
+      "learning_rate": 4.9072718513183235e-06,
+      "loss": 0.6614,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5473684210526316,
+      "grad_norm": 0.7072902917861938,
+      "learning_rate": 4.9071735317712735e-06,
+      "loss": 0.6774,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5476454293628809,
+      "grad_norm": 0.6918482184410095,
+      "learning_rate": 4.907075161113782e-06,
+      "loss": 0.6708,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5479224376731302,
+      "grad_norm": 0.6814166307449341,
+      "learning_rate": 4.906976739347938e-06,
+      "loss": 0.7135,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5481994459833796,
+      "grad_norm": 0.6976650357246399,
+      "learning_rate": 4.906878266475831e-06,
+      "loss": 0.7145,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5484764542936288,
+      "grad_norm": 0.6577694416046143,
+      "learning_rate": 4.906779742499551e-06,
+      "loss": 0.7047,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5487534626038781,
+      "grad_norm": 0.642149806022644,
+      "learning_rate": 4.906681167421192e-06,
+      "loss": 0.6948,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5490304709141274,
+      "grad_norm": 0.6713173985481262,
+      "learning_rate": 4.906582541242846e-06,
+      "loss": 0.6845,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5493074792243767,
+      "grad_norm": 0.6460973024368286,
+      "learning_rate": 4.906483863966606e-06,
+      "loss": 0.7429,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5495844875346261,
+      "grad_norm": 0.6690168380737305,
+      "learning_rate": 4.906385135594569e-06,
+      "loss": 0.6535,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5498614958448753,
+      "grad_norm": 0.6556865572929382,
+      "learning_rate": 4.9062863561288295e-06,
+      "loss": 0.6635,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5501385041551247,
+      "grad_norm": 0.675041675567627,
+      "learning_rate": 4.9061875255714864e-06,
+      "loss": 0.6961,
+      "step": 1986
+    },
+    {
+      "epoch": 0.550415512465374,
+      "grad_norm": 0.6664207577705383,
+      "learning_rate": 4.906088643924637e-06,
+      "loss": 0.6739,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5506925207756233,
+      "grad_norm": 0.6541404128074646,
+      "learning_rate": 4.9059897111903806e-06,
+      "loss": 0.6884,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5509695290858726,
+      "grad_norm": 0.7333690524101257,
+      "learning_rate": 4.90589072737082e-06,
+      "loss": 0.6715,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5512465373961218,
+      "grad_norm": 0.6783778667449951,
+      "learning_rate": 4.905791692468054e-06,
+      "loss": 0.638,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5515235457063712,
+      "grad_norm": 0.6893191933631897,
+      "learning_rate": 4.905692606484187e-06,
+      "loss": 0.6847,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5518005540166205,
+      "grad_norm": 0.654119074344635,
+      "learning_rate": 4.905593469421323e-06,
+      "loss": 0.6741,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5520775623268698,
+      "grad_norm": 0.7027085423469543,
+      "learning_rate": 4.905494281281565e-06,
+      "loss": 0.7323,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5523545706371191,
+      "grad_norm": 0.6599571704864502,
+      "learning_rate": 4.9053950420670216e-06,
+      "loss": 0.7109,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5526315789473685,
+      "grad_norm": 0.6476833820343018,
+      "learning_rate": 4.905295751779799e-06,
+      "loss": 0.6835,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5529085872576177,
+      "grad_norm": 0.6545064449310303,
+      "learning_rate": 4.905196410422004e-06,
+      "loss": 0.6604,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5531855955678671,
+      "grad_norm": 0.6796197295188904,
+      "learning_rate": 4.905097017995748e-06,
+      "loss": 0.6958,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5534626038781163,
+      "grad_norm": 0.7135651111602783,
+      "learning_rate": 4.90499757450314e-06,
+      "loss": 0.6845,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5537396121883656,
+      "grad_norm": 0.6246007680892944,
+      "learning_rate": 4.904898079946292e-06,
+      "loss": 0.65,
+      "step": 1999
+    },
+    {
+      "epoch": 0.554016620498615,
+      "grad_norm": 0.6683261394500732,
+      "learning_rate": 4.9047985343273156e-06,
+      "loss": 0.7332,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5542936288088642,
+      "grad_norm": 0.7060092687606812,
+      "learning_rate": 4.9046989376483255e-06,
+      "loss": 0.6491,
+      "step": 2001
+    },
+    {
+      "epoch": 0.5545706371191136,
+      "grad_norm": 0.6899551153182983,
+      "learning_rate": 4.904599289911437e-06,
+      "loss": 0.7351,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5548476454293629,
+      "grad_norm": 0.7160254120826721,
+      "learning_rate": 4.9044995911187635e-06,
+      "loss": 0.7185,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5551246537396122,
+      "grad_norm": 0.6531845927238464,
+      "learning_rate": 4.904399841272423e-06,
+      "loss": 0.6797,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5554016620498615,
+      "grad_norm": 0.7408345341682434,
+      "learning_rate": 4.904300040374536e-06,
+      "loss": 0.7409,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5556786703601108,
+      "grad_norm": 0.7078654170036316,
+      "learning_rate": 4.904200188427217e-06,
+      "loss": 0.6958,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5559556786703601,
+      "grad_norm": 0.6584929823875427,
+      "learning_rate": 4.904100285432589e-06,
+      "loss": 0.7117,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5562326869806095,
+      "grad_norm": 0.7188738584518433,
+      "learning_rate": 4.904000331392773e-06,
+      "loss": 0.6743,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5565096952908587,
+      "grad_norm": 0.7072994709014893,
+      "learning_rate": 4.903900326309891e-06,
+      "loss": 0.6766,
+      "step": 2009
+    },
+    {
+      "epoch": 0.556786703601108,
+      "grad_norm": 0.6705852150917053,
+      "learning_rate": 4.903800270186065e-06,
+      "loss": 0.718,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5570637119113574,
+      "grad_norm": 0.6577473282814026,
+      "learning_rate": 4.903700163023421e-06,
+      "loss": 0.6679,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5573407202216066,
+      "grad_norm": 0.6974996328353882,
+      "learning_rate": 4.903600004824085e-06,
+      "loss": 0.7009,
+      "step": 2012
+    },
+    {
+      "epoch": 0.557617728531856,
+      "grad_norm": 0.6663270592689514,
+      "learning_rate": 4.903499795590182e-06,
+      "loss": 0.6532,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5578947368421052,
+      "grad_norm": 0.689300537109375,
+      "learning_rate": 4.90339953532384e-06,
+      "loss": 0.6609,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5581717451523546,
+      "grad_norm": 0.693435788154602,
+      "learning_rate": 4.90329922402719e-06,
+      "loss": 0.6501,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5584487534626039,
+      "grad_norm": 0.662970781326294,
+      "learning_rate": 4.903198861702359e-06,
+      "loss": 0.7003,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5587257617728532,
+      "grad_norm": 0.6968392133712769,
+      "learning_rate": 4.903098448351479e-06,
+      "loss": 0.7019,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5590027700831025,
+      "grad_norm": 0.6442001461982727,
+      "learning_rate": 4.902997983976682e-06,
+      "loss": 0.6748,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5592797783933517,
+      "grad_norm": 0.6815208196640015,
+      "learning_rate": 4.902897468580101e-06,
+      "loss": 0.6804,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5595567867036011,
+      "grad_norm": 0.6363765597343445,
+      "learning_rate": 4.902796902163871e-06,
+      "loss": 0.6565,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5598337950138504,
+      "grad_norm": 0.6487285494804382,
+      "learning_rate": 4.9026962847301265e-06,
+      "loss": 0.6293,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5601108033240997,
+      "grad_norm": 0.6575140357017517,
+      "learning_rate": 4.902595616281004e-06,
+      "loss": 0.6957,
+      "step": 2022
+    },
+    {
+      "epoch": 0.560387811634349,
+      "grad_norm": 0.7005342245101929,
+      "learning_rate": 4.9024948968186414e-06,
+      "loss": 0.7377,
+      "step": 2023
+    },
+    {
+      "epoch": 0.5606648199445984,
+      "grad_norm": 0.6797147989273071,
+      "learning_rate": 4.902394126345177e-06,
+      "loss": 0.7089,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5609418282548476,
+      "grad_norm": 0.6822004318237305,
+      "learning_rate": 4.9022933048627496e-06,
+      "loss": 0.7156,
+      "step": 2025
+    },
+    {
+      "epoch": 0.561218836565097,
+      "grad_norm": 0.6667379140853882,
+      "learning_rate": 4.902192432373501e-06,
+      "loss": 0.695,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5614958448753462,
+      "grad_norm": 0.6670651435852051,
+      "learning_rate": 4.902091508879572e-06,
+      "loss": 0.6792,
+      "step": 2027
+    },
+    {
+      "epoch": 0.5617728531855956,
+      "grad_norm": 0.6695021390914917,
+      "learning_rate": 4.901990534383106e-06,
+      "loss": 0.7117,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5620498614958449,
+      "grad_norm": 0.6924020051956177,
+      "learning_rate": 4.901889508886247e-06,
+      "loss": 0.6988,
+      "step": 2029
+    },
+    {
+      "epoch": 0.5623268698060941,
+      "grad_norm": 0.6992893218994141,
+      "learning_rate": 4.90178843239114e-06,
+      "loss": 0.6979,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5626038781163435,
+      "grad_norm": 0.6664560437202454,
+      "learning_rate": 4.9016873048999315e-06,
+      "loss": 0.6932,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5628808864265928,
+      "grad_norm": 0.6536049246788025,
+      "learning_rate": 4.901586126414768e-06,
+      "loss": 0.6256,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5631578947368421,
+      "grad_norm": 0.7021706104278564,
+      "learning_rate": 4.901484896937798e-06,
+      "loss": 0.6799,
+      "step": 2033
+    },
+    {
+      "epoch": 0.5634349030470914,
+      "grad_norm": 0.6776540279388428,
+      "learning_rate": 4.901383616471171e-06,
+      "loss": 0.6923,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5637119113573407,
+      "grad_norm": 0.6760473251342773,
+      "learning_rate": 4.901282285017037e-06,
+      "loss": 0.6688,
+      "step": 2035
+    },
+    {
+      "epoch": 0.56398891966759,
+      "grad_norm": 0.6964700818061829,
+      "learning_rate": 4.901180902577549e-06,
+      "loss": 0.6396,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5642659279778394,
+      "grad_norm": 0.6710968613624573,
+      "learning_rate": 4.901079469154857e-06,
+      "loss": 0.6854,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5645429362880886,
+      "grad_norm": 0.7143495678901672,
+      "learning_rate": 4.900977984751117e-06,
+      "loss": 0.7059,
+      "step": 2038
+    },
+    {
+      "epoch": 0.564819944598338,
+      "grad_norm": 0.6527528166770935,
+      "learning_rate": 4.900876449368483e-06,
+      "loss": 0.6767,
+      "step": 2039
+    },
+    {
+      "epoch": 0.5650969529085873,
+      "grad_norm": 0.677239716053009,
+      "learning_rate": 4.90077486300911e-06,
+      "loss": 0.6588,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5653739612188365,
+      "grad_norm": 0.728654682636261,
+      "learning_rate": 4.900673225675157e-06,
+      "loss": 0.6777,
+      "step": 2041
+    },
+    {
+      "epoch": 0.5656509695290859,
+      "grad_norm": 0.6956666111946106,
+      "learning_rate": 4.90057153736878e-06,
+      "loss": 0.6898,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5659279778393351,
+      "grad_norm": 0.6565061807632446,
+      "learning_rate": 4.900469798092139e-06,
+      "loss": 0.6801,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5662049861495845,
+      "grad_norm": 0.6923288702964783,
+      "learning_rate": 4.900368007847394e-06,
+      "loss": 0.7155,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5664819944598338,
+      "grad_norm": 0.6817392706871033,
+      "learning_rate": 4.900266166636707e-06,
+      "loss": 0.7134,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5667590027700831,
+      "grad_norm": 0.6653991341590881,
+      "learning_rate": 4.900164274462239e-06,
+      "loss": 0.6989,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5670360110803324,
+      "grad_norm": 0.671635091304779,
+      "learning_rate": 4.900062331326154e-06,
+      "loss": 0.6963,
+      "step": 2047
+    },
+    {
+      "epoch": 0.5673130193905818,
+      "grad_norm": 0.7176443934440613,
+      "learning_rate": 4.899960337230617e-06,
+      "loss": 0.6653,
+      "step": 2048
+    },
+    {
+      "epoch": 0.567590027700831,
+      "grad_norm": 0.6723015904426575,
+      "learning_rate": 4.899858292177792e-06,
+      "loss": 0.6992,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5678670360110804,
+      "grad_norm": 0.6572104096412659,
+      "learning_rate": 4.899756196169849e-06,
+      "loss": 0.6819,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5681440443213296,
+      "grad_norm": 0.6798298954963684,
+      "learning_rate": 4.899654049208952e-06,
+      "loss": 0.6943,
+      "step": 2051
+    },
+    {
+      "epoch": 0.5684210526315789,
+      "grad_norm": 0.6697311401367188,
+      "learning_rate": 4.8995518512972726e-06,
+      "loss": 0.694,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5686980609418283,
+      "grad_norm": 0.6293474435806274,
+      "learning_rate": 4.8994496024369786e-06,
+      "loss": 0.6486,
+      "step": 2053
+    },
+    {
+      "epoch": 0.5689750692520775,
+      "grad_norm": 0.6766144633293152,
+      "learning_rate": 4.899347302630242e-06,
+      "loss": 0.7201,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5692520775623269,
+      "grad_norm": 0.6531354784965515,
+      "learning_rate": 4.899244951879236e-06,
+      "loss": 0.6748,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5695290858725762,
+      "grad_norm": 0.6813303828239441,
+      "learning_rate": 4.899142550186132e-06,
+      "loss": 0.6287,
+      "step": 2056
+    },
+    {
+      "epoch": 0.5698060941828255,
+      "grad_norm": 0.6930283904075623,
+      "learning_rate": 4.899040097553105e-06,
+      "loss": 0.7111,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5700831024930748,
+      "grad_norm": 0.6502225399017334,
+      "learning_rate": 4.8989375939823305e-06,
+      "loss": 0.6785,
+      "step": 2058
+    },
+    {
+      "epoch": 0.570360110803324,
+      "grad_norm": 0.6486106514930725,
+      "learning_rate": 4.898835039475984e-06,
+      "loss": 0.6774,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5706371191135734,
+      "grad_norm": 0.676793098449707,
+      "learning_rate": 4.8987324340362445e-06,
+      "loss": 0.697,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5709141274238227,
+      "grad_norm": 0.6627329587936401,
+      "learning_rate": 4.898629777665288e-06,
+      "loss": 0.6715,
+      "step": 2061
+    },
+    {
+      "epoch": 0.571191135734072,
+      "grad_norm": 0.7187315225601196,
+      "learning_rate": 4.898527070365297e-06,
+      "loss": 0.6544,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5714681440443213,
+      "grad_norm": 0.6739577054977417,
+      "learning_rate": 4.898424312138452e-06,
+      "loss": 0.629,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5717451523545707,
+      "grad_norm": 0.6906730532646179,
+      "learning_rate": 4.898321502986934e-06,
+      "loss": 0.706,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5720221606648199,
+      "grad_norm": 0.6810632348060608,
+      "learning_rate": 4.898218642912924e-06,
+      "loss": 0.6717,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5722991689750693,
+      "grad_norm": 0.6437877416610718,
+      "learning_rate": 4.898115731918609e-06,
+      "loss": 0.6883,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5725761772853185,
+      "grad_norm": 0.6814095377922058,
+      "learning_rate": 4.898012770006173e-06,
+      "loss": 0.7277,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5728531855955679,
+      "grad_norm": 0.6995571255683899,
+      "learning_rate": 4.897909757177801e-06,
+      "loss": 0.6707,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5731301939058172,
+      "grad_norm": 0.709454357624054,
+      "learning_rate": 4.897806693435683e-06,
+      "loss": 0.6806,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5734072022160664,
+      "grad_norm": 0.6949672698974609,
+      "learning_rate": 4.897703578782004e-06,
+      "loss": 0.7012,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5736842105263158,
+      "grad_norm": 0.6759223341941833,
+      "learning_rate": 4.8976004132189555e-06,
+      "loss": 0.6652,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5739612188365651,
+      "grad_norm": 0.710740327835083,
+      "learning_rate": 4.897497196748728e-06,
+      "loss": 0.7139,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5742382271468144,
+      "grad_norm": 0.6668627858161926,
+      "learning_rate": 4.897393929373512e-06,
+      "loss": 0.7286,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5745152354570637,
+      "grad_norm": 0.6559934616088867,
+      "learning_rate": 4.8972906110955e-06,
+      "loss": 0.6539,
+      "step": 2074
+    },
+    {
+      "epoch": 0.574792243767313,
+      "grad_norm": 0.6624038219451904,
+      "learning_rate": 4.897187241916888e-06,
+      "loss": 0.6742,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5750692520775623,
+      "grad_norm": 0.6761890649795532,
+      "learning_rate": 4.897083821839868e-06,
+      "loss": 0.6862,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5753462603878117,
+      "grad_norm": 0.7120620012283325,
+      "learning_rate": 4.896980350866637e-06,
+      "loss": 0.7066,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5756232686980609,
+      "grad_norm": 0.6602346301078796,
+      "learning_rate": 4.896876828999391e-06,
+      "loss": 0.6686,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5759002770083103,
+      "grad_norm": 0.6451171040534973,
+      "learning_rate": 4.89677325624033e-06,
+      "loss": 0.6283,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5761772853185596,
+      "grad_norm": 0.6720165014266968,
+      "learning_rate": 4.896669632591652e-06,
+      "loss": 0.6713,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5764542936288088,
+      "grad_norm": 0.6697086691856384,
+      "learning_rate": 4.896565958055557e-06,
+      "loss": 0.6797,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5767313019390582,
+      "grad_norm": 0.7319167256355286,
+      "learning_rate": 4.896462232634247e-06,
+      "loss": 0.6524,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5770083102493074,
+      "grad_norm": 0.6829770803451538,
+      "learning_rate": 4.896358456329924e-06,
+      "loss": 0.6859,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5772853185595568,
+      "grad_norm": 0.6501836180686951,
+      "learning_rate": 4.896254629144791e-06,
+      "loss": 0.6848,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5775623268698061,
+      "grad_norm": 0.6823615431785583,
+      "learning_rate": 4.896150751081052e-06,
+      "loss": 0.689,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5778393351800554,
+      "grad_norm": 0.6782605648040771,
+      "learning_rate": 4.896046822140915e-06,
+      "loss": 0.6689,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5781163434903047,
+      "grad_norm": 0.6973512768745422,
+      "learning_rate": 4.895942842326584e-06,
+      "loss": 0.6894,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5783933518005541,
+      "grad_norm": 0.7102792859077454,
+      "learning_rate": 4.895838811640269e-06,
+      "loss": 0.6893,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5786703601108033,
+      "grad_norm": 0.6860445141792297,
+      "learning_rate": 4.895734730084176e-06,
+      "loss": 0.7189,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.6870277523994446,
+      "learning_rate": 4.895630597660518e-06,
+      "loss": 0.7391,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5792243767313019,
+      "grad_norm": 0.8472511768341064,
+      "learning_rate": 4.895526414371505e-06,
+      "loss": 0.6462,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5795013850415512,
+      "grad_norm": 0.6843624114990234,
+      "learning_rate": 4.895422180219348e-06,
+      "loss": 0.6971,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5797783933518006,
+      "grad_norm": 0.6715996861457825,
+      "learning_rate": 4.895317895206261e-06,
+      "loss": 0.6273,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5800554016620498,
+      "grad_norm": 0.6269164085388184,
+      "learning_rate": 4.895213559334458e-06,
+      "loss": 0.7139,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5803324099722992,
+      "grad_norm": 0.7106332182884216,
+      "learning_rate": 4.895109172606154e-06,
+      "loss": 0.7015,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5806094182825485,
+      "grad_norm": 0.6532184481620789,
+      "learning_rate": 4.895004735023567e-06,
+      "loss": 0.6837,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5808864265927978,
+      "grad_norm": 0.6516802310943604,
+      "learning_rate": 4.894900246588912e-06,
+      "loss": 0.6981,
+      "step": 2097
+    },
+    {
+      "epoch": 0.5811634349030471,
+      "grad_norm": 0.6725658774375916,
+      "learning_rate": 4.894795707304409e-06,
+      "loss": 0.6277,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5814404432132964,
+      "grad_norm": 0.6376446485519409,
+      "learning_rate": 4.894691117172279e-06,
+      "loss": 0.6467,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5817174515235457,
+      "grad_norm": 0.663567304611206,
+      "learning_rate": 4.894586476194739e-06,
+      "loss": 0.7279,
+      "step": 2100
+    },
+    {
+      "epoch": 0.581994459833795,
+      "grad_norm": 0.6665627360343933,
+      "learning_rate": 4.8944817843740145e-06,
+      "loss": 0.6799,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5822714681440443,
+      "grad_norm": 0.6752210855484009,
+      "learning_rate": 4.894377041712327e-06,
+      "loss": 0.6438,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5825484764542936,
+      "grad_norm": 0.6799635887145996,
+      "learning_rate": 4.8942722482119e-06,
+      "loss": 0.7247,
+      "step": 2103
+    },
+    {
+      "epoch": 0.582825484764543,
+      "grad_norm": 0.6850577592849731,
+      "learning_rate": 4.894167403874958e-06,
+      "loss": 0.6415,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5831024930747922,
+      "grad_norm": 0.684273362159729,
+      "learning_rate": 4.894062508703729e-06,
+      "loss": 0.702,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5833795013850416,
+      "grad_norm": 0.6766973733901978,
+      "learning_rate": 4.893957562700439e-06,
+      "loss": 0.7174,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5836565096952908,
+      "grad_norm": 0.7171596884727478,
+      "learning_rate": 4.8938525658673165e-06,
+      "loss": 0.6696,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5839335180055402,
+      "grad_norm": 0.6781456470489502,
+      "learning_rate": 4.8937475182065905e-06,
+      "loss": 0.731,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5842105263157895,
+      "grad_norm": 0.6551393270492554,
+      "learning_rate": 4.893642419720491e-06,
+      "loss": 0.6546,
+      "step": 2109
+    },
+    {
+      "epoch": 0.5844875346260388,
+      "grad_norm": 0.6730072498321533,
+      "learning_rate": 4.893537270411252e-06,
+      "loss": 0.6606,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5847645429362881,
+      "grad_norm": 0.7042357921600342,
+      "learning_rate": 4.893432070281102e-06,
+      "loss": 0.7056,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5850415512465375,
+      "grad_norm": 0.6876088380813599,
+      "learning_rate": 4.893326819332279e-06,
+      "loss": 0.73,
+      "step": 2112
+    },
+    {
+      "epoch": 0.5853185595567867,
+      "grad_norm": 0.700372576713562,
+      "learning_rate": 4.893221517567015e-06,
+      "loss": 0.6916,
+      "step": 2113
+    },
+    {
+      "epoch": 0.585595567867036,
+      "grad_norm": 0.6858752965927124,
+      "learning_rate": 4.893116164987547e-06,
+      "loss": 0.6089,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5858725761772853,
+      "grad_norm": 0.6903958916664124,
+      "learning_rate": 4.893010761596111e-06,
+      "loss": 0.6903,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5861495844875346,
+      "grad_norm": 0.7026486396789551,
+      "learning_rate": 4.892905307394945e-06,
+      "loss": 0.7043,
+      "step": 2116
+    },
+    {
+      "epoch": 0.586426592797784,
+      "grad_norm": 0.6673232913017273,
+      "learning_rate": 4.89279980238629e-06,
+      "loss": 0.6843,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5867036011080332,
+      "grad_norm": 0.6887499094009399,
+      "learning_rate": 4.892694246572384e-06,
+      "loss": 0.6362,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5869806094182826,
+      "grad_norm": 0.6565698385238647,
+      "learning_rate": 4.892588639955468e-06,
+      "loss": 0.6743,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5872576177285319,
+      "grad_norm": 0.6758242249488831,
+      "learning_rate": 4.892482982537786e-06,
+      "loss": 0.7125,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5875346260387811,
+      "grad_norm": 0.7436144948005676,
+      "learning_rate": 4.892377274321581e-06,
+      "loss": 0.7179,
+      "step": 2121
+    },
+    {
+      "epoch": 0.5878116343490305,
+      "grad_norm": 0.6895393133163452,
+      "learning_rate": 4.892271515309096e-06,
+      "loss": 0.7159,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5880886426592797,
+      "grad_norm": 0.6338022947311401,
+      "learning_rate": 4.892165705502578e-06,
+      "loss": 0.6417,
+      "step": 2123
+    },
+    {
+      "epoch": 0.5883656509695291,
+      "grad_norm": 0.6784072518348694,
+      "learning_rate": 4.892059844904273e-06,
+      "loss": 0.6617,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5886426592797784,
+      "grad_norm": 0.6545912623405457,
+      "learning_rate": 4.891953933516429e-06,
+      "loss": 0.6912,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5889196675900277,
+      "grad_norm": 0.6839507222175598,
+      "learning_rate": 4.891847971341293e-06,
+      "loss": 0.6548,
+      "step": 2126
+    },
+    {
+      "epoch": 0.589196675900277,
+      "grad_norm": 0.690337061882019,
+      "learning_rate": 4.891741958381119e-06,
+      "loss": 0.7042,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5894736842105263,
+      "grad_norm": 0.6950432062149048,
+      "learning_rate": 4.891635894638154e-06,
+      "loss": 0.6529,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5897506925207756,
+      "grad_norm": 0.6476900577545166,
+      "learning_rate": 4.8915297801146515e-06,
+      "loss": 0.6045,
+      "step": 2129
+    },
+    {
+      "epoch": 0.590027700831025,
+      "grad_norm": 0.7296699285507202,
+      "learning_rate": 4.891423614812864e-06,
+      "loss": 0.7426,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5903047091412742,
+      "grad_norm": 0.6767834424972534,
+      "learning_rate": 4.8913173987350474e-06,
+      "loss": 0.6965,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5905817174515235,
+      "grad_norm": 0.7208825349807739,
+      "learning_rate": 4.891211131883455e-06,
+      "loss": 0.7025,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5908587257617729,
+      "grad_norm": 0.6839019656181335,
+      "learning_rate": 4.891104814260344e-06,
+      "loss": 0.6779,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5911357340720221,
+      "grad_norm": 0.7386277318000793,
+      "learning_rate": 4.890998445867971e-06,
+      "loss": 0.7333,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5914127423822715,
+      "grad_norm": 0.7242138385772705,
+      "learning_rate": 4.890892026708596e-06,
+      "loss": 0.693,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5916897506925207,
+      "grad_norm": 0.7165040969848633,
+      "learning_rate": 4.8907855567844765e-06,
+      "loss": 0.7229,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5919667590027701,
+      "grad_norm": 0.6969289183616638,
+      "learning_rate": 4.890679036097875e-06,
+      "loss": 0.7051,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5922437673130194,
+      "grad_norm": 0.6765584349632263,
+      "learning_rate": 4.890572464651052e-06,
+      "loss": 0.6803,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5925207756232687,
+      "grad_norm": 0.6918100714683533,
+      "learning_rate": 4.890465842446272e-06,
+      "loss": 0.6563,
+      "step": 2139
+    },
+    {
+      "epoch": 0.592797783933518,
+      "grad_norm": 0.6466085314750671,
+      "learning_rate": 4.890359169485796e-06,
+      "loss": 0.686,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5930747922437674,
+      "grad_norm": 0.6893829107284546,
+      "learning_rate": 4.89025244577189e-06,
+      "loss": 0.7093,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5933518005540166,
+      "grad_norm": 0.7149084210395813,
+      "learning_rate": 4.890145671306822e-06,
+      "loss": 0.7221,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5936288088642659,
+      "grad_norm": 0.6596125960350037,
+      "learning_rate": 4.890038846092857e-06,
+      "loss": 0.677,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5939058171745152,
+      "grad_norm": 0.6518341302871704,
+      "learning_rate": 4.889931970132264e-06,
+      "loss": 0.6836,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5941828254847645,
+      "grad_norm": 0.6978798508644104,
+      "learning_rate": 4.8898250434273125e-06,
+      "loss": 0.6593,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5944598337950139,
+      "grad_norm": 0.7356164455413818,
+      "learning_rate": 4.889718065980272e-06,
+      "loss": 0.7028,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5947368421052631,
+      "grad_norm": 0.670665979385376,
+      "learning_rate": 4.889611037793414e-06,
+      "loss": 0.7081,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5950138504155125,
+      "grad_norm": 0.6960236430168152,
+      "learning_rate": 4.8895039588690114e-06,
+      "loss": 0.7024,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5952908587257618,
+      "grad_norm": 0.6881184577941895,
+      "learning_rate": 4.889396829209338e-06,
+      "loss": 0.7227,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5955678670360111,
+      "grad_norm": 0.6917644739151001,
+      "learning_rate": 4.889289648816667e-06,
+      "loss": 0.6668,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5958448753462604,
+      "grad_norm": 0.6641549468040466,
+      "learning_rate": 4.889182417693277e-06,
+      "loss": 0.7113,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5961218836565096,
+      "grad_norm": 0.7198570966720581,
+      "learning_rate": 4.889075135841441e-06,
+      "loss": 0.7505,
+      "step": 2152
+    },
+    {
+      "epoch": 0.596398891966759,
+      "grad_norm": 0.6820452213287354,
+      "learning_rate": 4.88896780326344e-06,
+      "loss": 0.6594,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5966759002770083,
+      "grad_norm": 0.7384587526321411,
+      "learning_rate": 4.888860419961552e-06,
+      "loss": 0.7326,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5969529085872576,
+      "grad_norm": 0.6759530901908875,
+      "learning_rate": 4.888752985938056e-06,
+      "loss": 0.6678,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5972299168975069,
+      "grad_norm": 0.6574460864067078,
+      "learning_rate": 4.888645501195234e-06,
+      "loss": 0.6978,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5975069252077563,
+      "grad_norm": 0.6380599737167358,
+      "learning_rate": 4.888537965735368e-06,
+      "loss": 0.6162,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5977839335180055,
+      "grad_norm": 0.6682248711585999,
+      "learning_rate": 4.8884303795607424e-06,
+      "loss": 0.7114,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5980609418282549,
+      "grad_norm": 0.7036237120628357,
+      "learning_rate": 4.888322742673639e-06,
+      "loss": 0.745,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5983379501385041,
+      "grad_norm": 0.6678205132484436,
+      "learning_rate": 4.888215055076346e-06,
+      "loss": 0.7086,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5986149584487535,
+      "grad_norm": 0.6644747853279114,
+      "learning_rate": 4.888107316771148e-06,
+      "loss": 0.6699,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5988919667590028,
+      "grad_norm": 0.6806349158287048,
+      "learning_rate": 4.8879995277603325e-06,
+      "loss": 0.7001,
+      "step": 2162
+    },
+    {
+      "epoch": 0.599168975069252,
+      "grad_norm": 0.6920258402824402,
+      "learning_rate": 4.887891688046189e-06,
+      "loss": 0.6539,
+      "step": 2163
+    },
+    {
+      "epoch": 0.5994459833795014,
+      "grad_norm": 0.6764971017837524,
+      "learning_rate": 4.887783797631007e-06,
+      "loss": 0.7057,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5997229916897507,
+      "grad_norm": 0.6854003071784973,
+      "learning_rate": 4.8876758565170775e-06,
+      "loss": 0.7155,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6682254076004028,
+      "learning_rate": 4.887567864706693e-06,
+      "loss": 0.7131,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6002770083102493,
+      "grad_norm": 0.7079361081123352,
+      "learning_rate": 4.887459822202144e-06,
+      "loss": 0.7172,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6005540166204986,
+      "grad_norm": 0.6760583519935608,
+      "learning_rate": 4.8873517290057265e-06,
+      "loss": 0.6874,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6008310249307479,
+      "grad_norm": 0.6594741940498352,
+      "learning_rate": 4.887243585119735e-06,
+      "loss": 0.7076,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6011080332409973,
+      "grad_norm": 0.6586843729019165,
+      "learning_rate": 4.887135390546467e-06,
+      "loss": 0.7064,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6013850415512465,
+      "grad_norm": 0.6381613612174988,
+      "learning_rate": 4.887027145288217e-06,
+      "loss": 0.6532,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6016620498614959,
+      "grad_norm": 0.7136830687522888,
+      "learning_rate": 4.886918849347286e-06,
+      "loss": 0.698,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6019390581717452,
+      "grad_norm": 0.6528488993644714,
+      "learning_rate": 4.886810502725971e-06,
+      "loss": 0.6945,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6022160664819944,
+      "grad_norm": 0.6757421493530273,
+      "learning_rate": 4.886702105426575e-06,
+      "loss": 0.6753,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6024930747922438,
+      "grad_norm": 0.7247016429901123,
+      "learning_rate": 4.886593657451398e-06,
+      "loss": 0.6999,
+      "step": 2175
+    },
+    {
+      "epoch": 0.602770083102493,
+      "grad_norm": 0.7493066191673279,
+      "learning_rate": 4.886485158802744e-06,
+      "loss": 0.6965,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6030470914127424,
+      "grad_norm": 0.7063272595405579,
+      "learning_rate": 4.886376609482914e-06,
+      "loss": 0.6789,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6033240997229917,
+      "grad_norm": 0.6630057096481323,
+      "learning_rate": 4.886268009494215e-06,
+      "loss": 0.6746,
+      "step": 2178
+    },
+    {
+      "epoch": 0.603601108033241,
+      "grad_norm": 0.7254530787467957,
+      "learning_rate": 4.886159358838952e-06,
+      "loss": 0.7137,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6038781163434903,
+      "grad_norm": 0.7073795199394226,
+      "learning_rate": 4.886050657519433e-06,
+      "loss": 0.6509,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6041551246537397,
+      "grad_norm": 0.7252312898635864,
+      "learning_rate": 4.885941905537965e-06,
+      "loss": 0.6888,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6044321329639889,
+      "grad_norm": 0.6500841379165649,
+      "learning_rate": 4.885833102896857e-06,
+      "loss": 0.7004,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6047091412742382,
+      "grad_norm": 0.6450377702713013,
+      "learning_rate": 4.885724249598419e-06,
+      "loss": 0.6183,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6049861495844875,
+      "grad_norm": 0.7047394514083862,
+      "learning_rate": 4.885615345644964e-06,
+      "loss": 0.6698,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6052631578947368,
+      "grad_norm": 0.7013279795646667,
+      "learning_rate": 4.885506391038803e-06,
+      "loss": 0.6704,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6055401662049862,
+      "grad_norm": 0.7043224573135376,
+      "learning_rate": 4.885397385782248e-06,
+      "loss": 0.6892,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6058171745152354,
+      "grad_norm": 0.6639013886451721,
+      "learning_rate": 4.885288329877616e-06,
+      "loss": 0.6849,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6060941828254848,
+      "grad_norm": 0.6464067101478577,
+      "learning_rate": 4.885179223327221e-06,
+      "loss": 0.691,
+      "step": 2188
+    },
+    {
+      "epoch": 0.6063711911357341,
+      "grad_norm": 0.6812418103218079,
+      "learning_rate": 4.885070066133379e-06,
+      "loss": 0.6825,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6066481994459834,
+      "grad_norm": 0.677861750125885,
+      "learning_rate": 4.88496085829841e-06,
+      "loss": 0.6562,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6069252077562327,
+      "grad_norm": 0.6878042221069336,
+      "learning_rate": 4.884851599824631e-06,
+      "loss": 0.6873,
+      "step": 2191
+    },
+    {
+      "epoch": 0.607202216066482,
+      "grad_norm": 0.6419922113418579,
+      "learning_rate": 4.884742290714363e-06,
+      "loss": 0.6314,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6074792243767313,
+      "grad_norm": 0.669269323348999,
+      "learning_rate": 4.884632930969925e-06,
+      "loss": 0.6585,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6077562326869806,
+      "grad_norm": 0.7030165791511536,
+      "learning_rate": 4.8845235205936415e-06,
+      "loss": 0.7232,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6080332409972299,
+      "grad_norm": 0.7237428426742554,
+      "learning_rate": 4.884414059587833e-06,
+      "loss": 0.6807,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6083102493074792,
+      "grad_norm": 0.6722372770309448,
+      "learning_rate": 4.884304547954826e-06,
+      "loss": 0.6833,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6085872576177286,
+      "grad_norm": 0.6264526844024658,
+      "learning_rate": 4.884194985696944e-06,
+      "loss": 0.6354,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6088642659279778,
+      "grad_norm": 0.6962215900421143,
+      "learning_rate": 4.884085372816514e-06,
+      "loss": 0.6909,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6091412742382272,
+      "grad_norm": 0.7049713134765625,
+      "learning_rate": 4.883975709315862e-06,
+      "loss": 0.6686,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6094182825484764,
+      "grad_norm": 0.6710338592529297,
+      "learning_rate": 4.883865995197319e-06,
+      "loss": 0.7199,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6096952908587258,
+      "grad_norm": 0.6907806396484375,
+      "learning_rate": 4.883756230463214e-06,
+      "loss": 0.7232,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6099722991689751,
+      "grad_norm": 0.6837185025215149,
+      "learning_rate": 4.8836464151158754e-06,
+      "loss": 0.6801,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6102493074792243,
+      "grad_norm": 0.6782760620117188,
+      "learning_rate": 4.883536549157637e-06,
+      "loss": 0.652,
+      "step": 2203
+    },
+    {
+      "epoch": 0.6105263157894737,
+      "grad_norm": 0.6475855708122253,
+      "learning_rate": 4.88342663259083e-06,
+      "loss": 0.6828,
+      "step": 2204
+    },
+    {
+      "epoch": 0.610803324099723,
+      "grad_norm": 0.6402779817581177,
+      "learning_rate": 4.88331666541779e-06,
+      "loss": 0.6555,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6110803324099723,
+      "grad_norm": 0.6540744304656982,
+      "learning_rate": 4.883206647640849e-06,
+      "loss": 0.6333,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6113573407202216,
+      "grad_norm": 0.6887601613998413,
+      "learning_rate": 4.883096579262347e-06,
+      "loss": 0.7423,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6116343490304709,
+      "grad_norm": 0.6777442693710327,
+      "learning_rate": 4.8829864602846176e-06,
+      "loss": 0.6991,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6119113573407202,
+      "grad_norm": 0.7278162240982056,
+      "learning_rate": 4.8828762907100004e-06,
+      "loss": 0.766,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6121883656509696,
+      "grad_norm": 0.6916058659553528,
+      "learning_rate": 4.882766070540835e-06,
+      "loss": 0.662,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6124653739612188,
+      "grad_norm": 0.6415777802467346,
+      "learning_rate": 4.88265579977946e-06,
+      "loss": 0.6727,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6127423822714682,
+      "grad_norm": 0.6946539878845215,
+      "learning_rate": 4.882545478428219e-06,
+      "loss": 0.7491,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6130193905817175,
+      "grad_norm": 0.7359540462493896,
+      "learning_rate": 4.882435106489451e-06,
+      "loss": 0.6904,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6132963988919667,
+      "grad_norm": 0.6972795724868774,
+      "learning_rate": 4.882324683965504e-06,
+      "loss": 0.6665,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6135734072022161,
+      "grad_norm": 0.683138906955719,
+      "learning_rate": 4.882214210858719e-06,
+      "loss": 0.7161,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6138504155124653,
+      "grad_norm": 0.6765393018722534,
+      "learning_rate": 4.882103687171443e-06,
+      "loss": 0.6748,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6141274238227147,
+      "grad_norm": 0.6710306406021118,
+      "learning_rate": 4.881993112906022e-06,
+      "loss": 0.6821,
+      "step": 2217
+    },
+    {
+      "epoch": 0.614404432132964,
+      "grad_norm": 0.6652889251708984,
+      "learning_rate": 4.881882488064805e-06,
+      "loss": 0.6898,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6146814404432133,
+      "grad_norm": 0.6607372164726257,
+      "learning_rate": 4.8817718126501395e-06,
+      "loss": 0.6551,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6149584487534626,
+      "grad_norm": 0.6769833564758301,
+      "learning_rate": 4.881661086664377e-06,
+      "loss": 0.6738,
+      "step": 2220
+    },
+    {
+      "epoch": 0.615235457063712,
+      "grad_norm": 0.7280565500259399,
+      "learning_rate": 4.881550310109866e-06,
+      "loss": 0.6383,
+      "step": 2221
+    },
+    {
+      "epoch": 0.6155124653739612,
+      "grad_norm": 0.7089138627052307,
+      "learning_rate": 4.881439482988961e-06,
+      "loss": 0.6745,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6157894736842106,
+      "grad_norm": 0.6932617425918579,
+      "learning_rate": 4.881328605304014e-06,
+      "loss": 0.6704,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6160664819944598,
+      "grad_norm": 0.6502192616462708,
+      "learning_rate": 4.881217677057379e-06,
+      "loss": 0.7143,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6163434903047091,
+      "grad_norm": 0.7033457159996033,
+      "learning_rate": 4.881106698251413e-06,
+      "loss": 0.7383,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6166204986149585,
+      "grad_norm": 0.704187273979187,
+      "learning_rate": 4.880995668888471e-06,
+      "loss": 0.6947,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6168975069252077,
+      "grad_norm": 0.7004435658454895,
+      "learning_rate": 4.88088458897091e-06,
+      "loss": 0.6534,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6171745152354571,
+      "grad_norm": 0.6873608827590942,
+      "learning_rate": 4.8807734585010894e-06,
+      "loss": 0.695,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6174515235457064,
+      "grad_norm": 0.6766569018363953,
+      "learning_rate": 4.880662277481367e-06,
+      "loss": 0.6757,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6177285318559557,
+      "grad_norm": 0.6786469221115112,
+      "learning_rate": 4.880551045914107e-06,
+      "loss": 0.6997,
+      "step": 2230
+    },
+    {
+      "epoch": 0.618005540166205,
+      "grad_norm": 0.678867757320404,
+      "learning_rate": 4.880439763801668e-06,
+      "loss": 0.6518,
+      "step": 2231
+    },
+    {
+      "epoch": 0.6182825484764543,
+      "grad_norm": 0.7730464339256287,
+      "learning_rate": 4.880328431146414e-06,
+      "loss": 0.7593,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6185595567867036,
+      "grad_norm": 0.6655864715576172,
+      "learning_rate": 4.880217047950708e-06,
+      "loss": 0.6665,
+      "step": 2233
+    },
+    {
+      "epoch": 0.618836565096953,
+      "grad_norm": 0.6689289212226868,
+      "learning_rate": 4.880105614216917e-06,
+      "loss": 0.7367,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6191135734072022,
+      "grad_norm": 0.6539269089698792,
+      "learning_rate": 4.879994129947404e-06,
+      "loss": 0.664,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6193905817174515,
+      "grad_norm": 0.6679507493972778,
+      "learning_rate": 4.8798825951445395e-06,
+      "loss": 0.6525,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6196675900277008,
+      "grad_norm": 0.6500898599624634,
+      "learning_rate": 4.879771009810689e-06,
+      "loss": 0.6509,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6199445983379501,
+      "grad_norm": 0.7124160528182983,
+      "learning_rate": 4.879659373948223e-06,
+      "loss": 0.7638,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6202216066481995,
+      "grad_norm": 0.6627144813537598,
+      "learning_rate": 4.879547687559511e-06,
+      "loss": 0.6739,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6204986149584487,
+      "grad_norm": 0.705913782119751,
+      "learning_rate": 4.879435950646926e-06,
+      "loss": 0.6733,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6207756232686981,
+      "grad_norm": 0.7119175791740417,
+      "learning_rate": 4.879324163212838e-06,
+      "loss": 0.6927,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6210526315789474,
+      "grad_norm": 0.739231526851654,
+      "learning_rate": 4.879212325259623e-06,
+      "loss": 0.7096,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6213296398891966,
+      "grad_norm": 0.7063774466514587,
+      "learning_rate": 4.879100436789653e-06,
+      "loss": 0.7557,
+      "step": 2243
+    },
+    {
+      "epoch": 0.621606648199446,
+      "grad_norm": 0.6932379603385925,
+      "learning_rate": 4.878988497805307e-06,
+      "loss": 0.6585,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6218836565096952,
+      "grad_norm": 0.6909125447273254,
+      "learning_rate": 4.8788765083089586e-06,
+      "loss": 0.7249,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6221606648199446,
+      "grad_norm": 0.7003384232521057,
+      "learning_rate": 4.878764468302988e-06,
+      "loss": 0.673,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6224376731301939,
+      "grad_norm": 0.6720861792564392,
+      "learning_rate": 4.878652377789772e-06,
+      "loss": 0.6629,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6227146814404432,
+      "grad_norm": 0.6857220530509949,
+      "learning_rate": 4.8785402367716926e-06,
+      "loss": 0.7017,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6229916897506925,
+      "grad_norm": 0.674379289150238,
+      "learning_rate": 4.878428045251129e-06,
+      "loss": 0.65,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6232686980609419,
+      "grad_norm": 0.6857823729515076,
+      "learning_rate": 4.878315803230464e-06,
+      "loss": 0.6992,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6235457063711911,
+      "grad_norm": 0.6758940815925598,
+      "learning_rate": 4.8782035107120815e-06,
+      "loss": 0.6714,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6238227146814405,
+      "grad_norm": 0.7062450051307678,
+      "learning_rate": 4.878091167698366e-06,
+      "loss": 0.7084,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6240997229916897,
+      "grad_norm": 0.7142696380615234,
+      "learning_rate": 4.8779787741917005e-06,
+      "loss": 0.6842,
+      "step": 2253
+    },
+    {
+      "epoch": 0.624376731301939,
+      "grad_norm": 0.6867758631706238,
+      "learning_rate": 4.877866330194474e-06,
+      "loss": 0.7025,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6246537396121884,
+      "grad_norm": 0.6744982004165649,
+      "learning_rate": 4.877753835709072e-06,
+      "loss": 0.6642,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6249307479224376,
+      "grad_norm": 0.700770914554596,
+      "learning_rate": 4.8776412907378845e-06,
+      "loss": 0.6839,
+      "step": 2256
+    },
+    {
+      "epoch": 0.625207756232687,
+      "grad_norm": 0.6821143627166748,
+      "learning_rate": 4.8775286952833e-06,
+      "loss": 0.6533,
+      "step": 2257
+    },
+    {
+      "epoch": 0.6254847645429363,
+      "grad_norm": 0.7303215265274048,
+      "learning_rate": 4.877416049347711e-06,
+      "loss": 0.6424,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6257617728531856,
+      "grad_norm": 0.669938862323761,
+      "learning_rate": 4.877303352933507e-06,
+      "loss": 0.6531,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6260387811634349,
+      "grad_norm": 0.753316342830658,
+      "learning_rate": 4.8771906060430815e-06,
+      "loss": 0.6983,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6263157894736842,
+      "grad_norm": 0.7299981713294983,
+      "learning_rate": 4.877077808678829e-06,
+      "loss": 0.6826,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6265927977839335,
+      "grad_norm": 0.6956072449684143,
+      "learning_rate": 4.876964960843145e-06,
+      "loss": 0.704,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6268698060941829,
+      "grad_norm": 0.6745579242706299,
+      "learning_rate": 4.876852062538424e-06,
+      "loss": 0.675,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6271468144044321,
+      "grad_norm": 0.649700403213501,
+      "learning_rate": 4.876739113767065e-06,
+      "loss": 0.6664,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6274238227146814,
+      "grad_norm": 0.711959719657898,
+      "learning_rate": 4.8766261145314646e-06,
+      "loss": 0.6927,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6277008310249308,
+      "grad_norm": 0.6987378597259521,
+      "learning_rate": 4.876513064834022e-06,
+      "loss": 0.7043,
+      "step": 2266
+    },
+    {
+      "epoch": 0.62797783933518,
+      "grad_norm": 0.7144526839256287,
+      "learning_rate": 4.876399964677138e-06,
+      "loss": 0.6768,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6282548476454294,
+      "grad_norm": 0.6852195262908936,
+      "learning_rate": 4.876286814063215e-06,
+      "loss": 0.7018,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6285318559556786,
+      "grad_norm": 0.6548054814338684,
+      "learning_rate": 4.876173612994653e-06,
+      "loss": 0.6419,
+      "step": 2269
+    },
+    {
+      "epoch": 0.628808864265928,
+      "grad_norm": 0.6945639848709106,
+      "learning_rate": 4.8760603614738585e-06,
+      "loss": 0.6724,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6290858725761773,
+      "grad_norm": 0.7267112135887146,
+      "learning_rate": 4.875947059503234e-06,
+      "loss": 0.6318,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6293628808864266,
+      "grad_norm": 0.6593161225318909,
+      "learning_rate": 4.8758337070851865e-06,
+      "loss": 0.6544,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6296398891966759,
+      "grad_norm": 0.6698119044303894,
+      "learning_rate": 4.875720304222122e-06,
+      "loss": 0.6867,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6299168975069253,
+      "grad_norm": 0.7131072878837585,
+      "learning_rate": 4.875606850916449e-06,
+      "loss": 0.6929,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6301939058171745,
+      "grad_norm": 0.6932727098464966,
+      "learning_rate": 4.8754933471705745e-06,
+      "loss": 0.6804,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6304709141274238,
+      "grad_norm": 0.6713318228721619,
+      "learning_rate": 4.8753797929869115e-06,
+      "loss": 0.6545,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6307479224376731,
+      "grad_norm": 0.6698991656303406,
+      "learning_rate": 4.875266188367868e-06,
+      "loss": 0.7334,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6310249307479224,
+      "grad_norm": 0.6613983511924744,
+      "learning_rate": 4.875152533315859e-06,
+      "loss": 0.6761,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6313019390581718,
+      "grad_norm": 0.7016041874885559,
+      "learning_rate": 4.875038827833295e-06,
+      "loss": 0.7175,
+      "step": 2279
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.7024201154708862,
+      "learning_rate": 4.874925071922591e-06,
+      "loss": 0.6813,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6318559556786704,
+      "grad_norm": 0.6844956278800964,
+      "learning_rate": 4.874811265586165e-06,
+      "loss": 0.6588,
+      "step": 2281
+    },
+    {
+      "epoch": 0.6321329639889197,
+      "grad_norm": 0.7185789942741394,
+      "learning_rate": 4.87469740882643e-06,
+      "loss": 0.6906,
+      "step": 2282
+    },
+    {
+      "epoch": 0.632409972299169,
+      "grad_norm": 0.6658942699432373,
+      "learning_rate": 4.874583501645804e-06,
+      "loss": 0.6628,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6326869806094183,
+      "grad_norm": 0.7003322839736938,
+      "learning_rate": 4.874469544046707e-06,
+      "loss": 0.6811,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6329639889196675,
+      "grad_norm": 0.6516456007957458,
+      "learning_rate": 4.874355536031558e-06,
+      "loss": 0.6617,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6332409972299169,
+      "grad_norm": 0.6722559928894043,
+      "learning_rate": 4.874241477602778e-06,
+      "loss": 0.6499,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6335180055401662,
+      "grad_norm": 0.6841036677360535,
+      "learning_rate": 4.874127368762787e-06,
+      "loss": 0.6857,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6337950138504155,
+      "grad_norm": 0.7111048102378845,
+      "learning_rate": 4.8740132095140095e-06,
+      "loss": 0.647,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6340720221606648,
+      "grad_norm": 0.7113420963287354,
+      "learning_rate": 4.873898999858869e-06,
+      "loss": 0.6205,
+      "step": 2289
+    },
+    {
+      "epoch": 0.6343490304709142,
+      "grad_norm": 0.709007978439331,
+      "learning_rate": 4.873784739799791e-06,
+      "loss": 0.7045,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6346260387811634,
+      "grad_norm": 0.7362955808639526,
+      "learning_rate": 4.873670429339201e-06,
+      "loss": 0.6675,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6349030470914128,
+      "grad_norm": 0.6905544400215149,
+      "learning_rate": 4.873556068479525e-06,
+      "loss": 0.6069,
+      "step": 2292
+    },
+    {
+      "epoch": 0.635180055401662,
+      "grad_norm": 0.6953214406967163,
+      "learning_rate": 4.873441657223194e-06,
+      "loss": 0.6839,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6354570637119114,
+      "grad_norm": 0.6331837773323059,
+      "learning_rate": 4.873327195572633e-06,
+      "loss": 0.6406,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6357340720221607,
+      "grad_norm": 0.7017216086387634,
+      "learning_rate": 4.873212683530277e-06,
+      "loss": 0.6999,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6360110803324099,
+      "grad_norm": 0.6879724860191345,
+      "learning_rate": 4.873098121098555e-06,
+      "loss": 0.711,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6362880886426593,
+      "grad_norm": 0.6714237332344055,
+      "learning_rate": 4.872983508279899e-06,
+      "loss": 0.7054,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6365650969529086,
+      "grad_norm": 0.7007396817207336,
+      "learning_rate": 4.872868845076744e-06,
+      "loss": 0.6493,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6368421052631579,
+      "grad_norm": 0.67763352394104,
+      "learning_rate": 4.872754131491523e-06,
+      "loss": 0.6845,
+      "step": 2299
+    },
+    {
+      "epoch": 0.6371191135734072,
+      "grad_norm": 0.6635951995849609,
+      "learning_rate": 4.872639367526672e-06,
+      "loss": 0.7008,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6373961218836565,
+      "grad_norm": 0.6842290163040161,
+      "learning_rate": 4.8725245531846286e-06,
+      "loss": 0.6962,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6376731301939058,
+      "grad_norm": 0.7155497670173645,
+      "learning_rate": 4.87240968846783e-06,
+      "loss": 0.7084,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6379501385041552,
+      "grad_norm": 0.6613172888755798,
+      "learning_rate": 4.872294773378716e-06,
+      "loss": 0.6343,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6382271468144044,
+      "grad_norm": 0.6752813458442688,
+      "learning_rate": 4.872179807919726e-06,
+      "loss": 0.7075,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6385041551246537,
+      "grad_norm": 0.6734436750411987,
+      "learning_rate": 4.8720647920932995e-06,
+      "loss": 0.7254,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6387811634349031,
+      "grad_norm": 0.6765357255935669,
+      "learning_rate": 4.871949725901881e-06,
+      "loss": 0.6971,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6390581717451523,
+      "grad_norm": 0.6636792421340942,
+      "learning_rate": 4.871834609347912e-06,
+      "loss": 0.6661,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6393351800554017,
+      "grad_norm": 0.7011874914169312,
+      "learning_rate": 4.871719442433836e-06,
+      "loss": 0.7426,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6396121883656509,
+      "grad_norm": 0.6820942759513855,
+      "learning_rate": 4.871604225162101e-06,
+      "loss": 0.6943,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6398891966759003,
+      "grad_norm": 0.6526979804039001,
+      "learning_rate": 4.871488957535152e-06,
+      "loss": 0.6081,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6401662049861496,
+      "grad_norm": 0.6738025546073914,
+      "learning_rate": 4.871373639555436e-06,
+      "loss": 0.7388,
+      "step": 2311
+    },
+    {
+      "epoch": 0.6404432132963989,
+      "grad_norm": 0.7214510440826416,
+      "learning_rate": 4.871258271225402e-06,
+      "loss": 0.6754,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6407202216066482,
+      "grad_norm": 0.6859409213066101,
+      "learning_rate": 4.871142852547499e-06,
+      "loss": 0.6421,
+      "step": 2313
+    },
+    {
+      "epoch": 0.6409972299168976,
+      "grad_norm": 0.6700871586799622,
+      "learning_rate": 4.871027383524178e-06,
+      "loss": 0.61,
+      "step": 2314
+    },
+    {
+      "epoch": 0.6412742382271468,
+      "grad_norm": 0.6625704765319824,
+      "learning_rate": 4.870911864157891e-06,
+      "loss": 0.7213,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6415512465373961,
+      "grad_norm": 0.697500467300415,
+      "learning_rate": 4.87079629445109e-06,
+      "loss": 0.7131,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6418282548476454,
+      "grad_norm": 0.6746952533721924,
+      "learning_rate": 4.87068067440623e-06,
+      "loss": 0.6821,
+      "step": 2317
+    },
+    {
+      "epoch": 0.6421052631578947,
+      "grad_norm": 0.6551344990730286,
+      "learning_rate": 4.870565004025764e-06,
+      "loss": 0.6542,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6423822714681441,
+      "grad_norm": 0.6762755513191223,
+      "learning_rate": 4.87044928331215e-06,
+      "loss": 0.7175,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6426592797783933,
+      "grad_norm": 0.664469838142395,
+      "learning_rate": 4.870333512267846e-06,
+      "loss": 0.6826,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6429362880886427,
+      "grad_norm": 0.6875748038291931,
+      "learning_rate": 4.870217690895306e-06,
+      "loss": 0.6722,
+      "step": 2321
+    },
+    {
+      "epoch": 0.643213296398892,
+      "grad_norm": 0.6557368040084839,
+      "learning_rate": 4.870101819196992e-06,
+      "loss": 0.689,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6434903047091413,
+      "grad_norm": 0.6870899796485901,
+      "learning_rate": 4.8699858971753646e-06,
+      "loss": 0.6555,
+      "step": 2323
+    },
+    {
+      "epoch": 0.6437673130193906,
+      "grad_norm": 0.7284519672393799,
+      "learning_rate": 4.869869924832883e-06,
+      "loss": 0.7128,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6440443213296398,
+      "grad_norm": 0.7355677485466003,
+      "learning_rate": 4.869753902172012e-06,
+      "loss": 0.744,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6443213296398892,
+      "grad_norm": 0.6616363525390625,
+      "learning_rate": 4.869637829195214e-06,
+      "loss": 0.6767,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6445983379501385,
+      "grad_norm": 0.6916202306747437,
+      "learning_rate": 4.869521705904953e-06,
+      "loss": 0.7144,
+      "step": 2327
+    },
+    {
+      "epoch": 0.6448753462603878,
+      "grad_norm": 0.7016916275024414,
+      "learning_rate": 4.869405532303695e-06,
+      "loss": 0.7178,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6451523545706371,
+      "grad_norm": 0.7312451601028442,
+      "learning_rate": 4.869289308393907e-06,
+      "loss": 0.6602,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6454293628808865,
+      "grad_norm": 0.6760799884796143,
+      "learning_rate": 4.869173034178057e-06,
+      "loss": 0.6401,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6457063711911357,
+      "grad_norm": 0.6891919374465942,
+      "learning_rate": 4.869056709658611e-06,
+      "loss": 0.6833,
+      "step": 2331
+    },
+    {
+      "epoch": 0.6459833795013851,
+      "grad_norm": 0.694268524646759,
+      "learning_rate": 4.8689403348380435e-06,
+      "loss": 0.7102,
+      "step": 2332
+    },
+    {
+      "epoch": 0.6462603878116343,
+      "grad_norm": 0.6928057074546814,
+      "learning_rate": 4.868823909718823e-06,
+      "loss": 0.7166,
+      "step": 2333
+    },
+    {
+      "epoch": 0.6465373961218837,
+      "grad_norm": 0.6451415419578552,
+      "learning_rate": 4.86870743430342e-06,
+      "loss": 0.7169,
+      "step": 2334
+    },
+    {
+      "epoch": 0.646814404432133,
+      "grad_norm": 0.703238308429718,
+      "learning_rate": 4.86859090859431e-06,
+      "loss": 0.6977,
+      "step": 2335
+    },
+    {
+      "epoch": 0.6470914127423822,
+      "grad_norm": 0.6713940501213074,
+      "learning_rate": 4.868474332593966e-06,
+      "loss": 0.7237,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6473684210526316,
+      "grad_norm": 0.6840459108352661,
+      "learning_rate": 4.8683577063048645e-06,
+      "loss": 0.6935,
+      "step": 2337
+    },
+    {
+      "epoch": 0.6476454293628808,
+      "grad_norm": 0.7354936003684998,
+      "learning_rate": 4.86824102972948e-06,
+      "loss": 0.7613,
+      "step": 2338
+    },
+    {
+      "epoch": 0.6479224376731302,
+      "grad_norm": 0.705204427242279,
+      "learning_rate": 4.868124302870291e-06,
+      "loss": 0.7165,
+      "step": 2339
+    },
+    {
+      "epoch": 0.6481994459833795,
+      "grad_norm": 0.7116757035255432,
+      "learning_rate": 4.8680075257297755e-06,
+      "loss": 0.6692,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6484764542936288,
+      "grad_norm": 0.6490796804428101,
+      "learning_rate": 4.8678906983104125e-06,
+      "loss": 0.6556,
+      "step": 2341
+    },
+    {
+      "epoch": 0.6487534626038781,
+      "grad_norm": 0.66712886095047,
+      "learning_rate": 4.867773820614684e-06,
+      "loss": 0.6004,
+      "step": 2342
+    },
+    {
+      "epoch": 0.6490304709141275,
+      "grad_norm": 0.6557923555374146,
+      "learning_rate": 4.86765689264507e-06,
+      "loss": 0.6802,
+      "step": 2343
+    },
+    {
+      "epoch": 0.6493074792243767,
+      "grad_norm": 0.6623290777206421,
+      "learning_rate": 4.8675399144040535e-06,
+      "loss": 0.6551,
+      "step": 2344
+    },
+    {
+      "epoch": 0.649584487534626,
+      "grad_norm": 0.6883432865142822,
+      "learning_rate": 4.867422885894119e-06,
+      "loss": 0.7095,
+      "step": 2345
+    },
+    {
+      "epoch": 0.6498614958448753,
+      "grad_norm": 0.7088051438331604,
+      "learning_rate": 4.867305807117751e-06,
+      "loss": 0.7272,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6501385041551246,
+      "grad_norm": 0.6738824248313904,
+      "learning_rate": 4.8671886780774345e-06,
+      "loss": 0.6953,
+      "step": 2347
+    },
+    {
+      "epoch": 0.650415512465374,
+      "grad_norm": 0.7325379848480225,
+      "learning_rate": 4.867071498775659e-06,
+      "loss": 0.6544,
+      "step": 2348
+    },
+    {
+      "epoch": 0.6506925207756232,
+      "grad_norm": 0.6509625911712646,
+      "learning_rate": 4.8669542692149085e-06,
+      "loss": 0.6904,
+      "step": 2349
+    },
+    {
+      "epoch": 0.6509695290858726,
+      "grad_norm": 0.647982120513916,
+      "learning_rate": 4.866836989397675e-06,
+      "loss": 0.6693,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6512465373961219,
+      "grad_norm": 0.6922804713249207,
+      "learning_rate": 4.866719659326449e-06,
+      "loss": 0.6671,
+      "step": 2351
+    },
+    {
+      "epoch": 0.6515235457063712,
+      "grad_norm": 0.6838349103927612,
+      "learning_rate": 4.86660227900372e-06,
+      "loss": 0.7061,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6518005540166205,
+      "grad_norm": 0.7052571773529053,
+      "learning_rate": 4.866484848431981e-06,
+      "loss": 0.7264,
+      "step": 2353
+    },
+    {
+      "epoch": 0.6520775623268698,
+      "grad_norm": 0.6909821033477783,
+      "learning_rate": 4.866367367613725e-06,
+      "loss": 0.7305,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6523545706371191,
+      "grad_norm": 0.6939254403114319,
+      "learning_rate": 4.866249836551447e-06,
+      "loss": 0.6665,
+      "step": 2355
+    },
+    {
+      "epoch": 0.6526315789473685,
+      "grad_norm": 0.6795610189437866,
+      "learning_rate": 4.8661322552476425e-06,
+      "loss": 0.6743,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6529085872576177,
+      "grad_norm": 0.6629143357276917,
+      "learning_rate": 4.8660146237048075e-06,
+      "loss": 0.6936,
+      "step": 2357
+    },
+    {
+      "epoch": 0.653185595567867,
+      "grad_norm": 0.7031983733177185,
+      "learning_rate": 4.8658969419254395e-06,
+      "loss": 0.6712,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6534626038781164,
+      "grad_norm": 0.6823177337646484,
+      "learning_rate": 4.865779209912039e-06,
+      "loss": 0.6988,
+      "step": 2359
+    },
+    {
+      "epoch": 0.6537396121883656,
+      "grad_norm": 0.7229796648025513,
+      "learning_rate": 4.8656614276671035e-06,
+      "loss": 0.6827,
+      "step": 2360
+    },
+    {
+      "epoch": 0.654016620498615,
+      "grad_norm": 0.6489887237548828,
+      "learning_rate": 4.865543595193134e-06,
+      "loss": 0.6489,
+      "step": 2361
+    },
+    {
+      "epoch": 0.6542936288088642,
+      "grad_norm": 0.6998574733734131,
+      "learning_rate": 4.865425712492634e-06,
+      "loss": 0.718,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6545706371191136,
+      "grad_norm": 0.6842721700668335,
+      "learning_rate": 4.865307779568104e-06,
+      "loss": 0.6843,
+      "step": 2363
+    },
+    {
+      "epoch": 0.6548476454293629,
+      "grad_norm": 0.6763157248497009,
+      "learning_rate": 4.865189796422051e-06,
+      "loss": 0.6713,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6551246537396122,
+      "grad_norm": 0.7252478003501892,
+      "learning_rate": 4.865071763056979e-06,
+      "loss": 0.6857,
+      "step": 2365
+    },
+    {
+      "epoch": 0.6554016620498615,
+      "grad_norm": 0.6811568140983582,
+      "learning_rate": 4.864953679475392e-06,
+      "loss": 0.6572,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6556786703601108,
+      "grad_norm": 0.7102620005607605,
+      "learning_rate": 4.8648355456798e-06,
+      "loss": 0.6759,
+      "step": 2367
+    },
+    {
+      "epoch": 0.6559556786703601,
+      "grad_norm": 0.6796259880065918,
+      "learning_rate": 4.86471736167271e-06,
+      "loss": 0.6906,
+      "step": 2368
+    },
+    {
+      "epoch": 0.6562326869806094,
+      "grad_norm": 0.6587184071540833,
+      "learning_rate": 4.864599127456632e-06,
+      "loss": 0.6048,
+      "step": 2369
+    },
+    {
+      "epoch": 0.6565096952908587,
+      "grad_norm": 0.6844584345817566,
+      "learning_rate": 4.864480843034075e-06,
+      "loss": 0.7046,
+      "step": 2370
+    },
+    {
+      "epoch": 0.656786703601108,
+      "grad_norm": 0.721359133720398,
+      "learning_rate": 4.864362508407553e-06,
+      "loss": 0.6874,
+      "step": 2371
+    },
+    {
+      "epoch": 0.6570637119113574,
+      "grad_norm": 0.6863132119178772,
+      "learning_rate": 4.8642441235795755e-06,
+      "loss": 0.7056,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6573407202216066,
+      "grad_norm": 0.6410539746284485,
+      "learning_rate": 4.864125688552659e-06,
+      "loss": 0.6306,
+      "step": 2373
+    },
+    {
+      "epoch": 0.657617728531856,
+      "grad_norm": 0.6737743616104126,
+      "learning_rate": 4.864007203329316e-06,
+      "loss": 0.6999,
+      "step": 2374
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 0.699963390827179,
+      "learning_rate": 4.863888667912062e-06,
+      "loss": 0.6792,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6581717451523545,
+      "grad_norm": 0.7324498891830444,
+      "learning_rate": 4.863770082303417e-06,
+      "loss": 0.6925,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6584487534626039,
+      "grad_norm": 0.6902961730957031,
+      "learning_rate": 4.8636514465058955e-06,
+      "loss": 0.6798,
+      "step": 2377
+    },
+    {
+      "epoch": 0.6587257617728531,
+      "grad_norm": 0.6362594962120056,
+      "learning_rate": 4.8635327605220174e-06,
+      "loss": 0.6752,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6590027700831025,
+      "grad_norm": 0.6623821258544922,
+      "learning_rate": 4.863414024354304e-06,
+      "loss": 0.6557,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6592797783933518,
+      "grad_norm": 0.6951994299888611,
+      "learning_rate": 4.863295238005275e-06,
+      "loss": 0.7541,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6595567867036011,
+      "grad_norm": 0.714439332485199,
+      "learning_rate": 4.863176401477453e-06,
+      "loss": 0.7237,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6598337950138504,
+      "grad_norm": 0.6886929869651794,
+      "learning_rate": 4.8630575147733605e-06,
+      "loss": 0.6761,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6601108033240998,
+      "grad_norm": 0.6800076365470886,
+      "learning_rate": 4.862938577895523e-06,
+      "loss": 0.6398,
+      "step": 2383
+    },
+    {
+      "epoch": 0.660387811634349,
+      "grad_norm": 0.6788169145584106,
+      "learning_rate": 4.862819590846465e-06,
+      "loss": 0.682,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6606648199445984,
+      "grad_norm": 0.6839731335639954,
+      "learning_rate": 4.862700553628713e-06,
+      "loss": 0.6691,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6609418282548476,
+      "grad_norm": 0.7404640913009644,
+      "learning_rate": 4.862581466244794e-06,
+      "loss": 0.6836,
+      "step": 2386
+    },
+    {
+      "epoch": 0.661218836565097,
+      "grad_norm": 0.6789003610610962,
+      "learning_rate": 4.862462328697239e-06,
+      "loss": 0.6926,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6614958448753463,
+      "grad_norm": 0.6176121830940247,
+      "learning_rate": 4.862343140988573e-06,
+      "loss": 0.6704,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6617728531855955,
+      "grad_norm": 0.699988067150116,
+      "learning_rate": 4.862223903121331e-06,
+      "loss": 0.6668,
+      "step": 2389
+    },
+    {
+      "epoch": 0.6620498614958449,
+      "grad_norm": 0.7106815576553345,
+      "learning_rate": 4.862104615098043e-06,
+      "loss": 0.7187,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6623268698060942,
+      "grad_norm": 0.758069634437561,
+      "learning_rate": 4.861985276921242e-06,
+      "loss": 0.6749,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6626038781163435,
+      "grad_norm": 0.6663122177124023,
+      "learning_rate": 4.861865888593461e-06,
+      "loss": 0.691,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6628808864265928,
+      "grad_norm": 0.697738766670227,
+      "learning_rate": 4.861746450117236e-06,
+      "loss": 0.7034,
+      "step": 2393
+    },
+    {
+      "epoch": 0.6631578947368421,
+      "grad_norm": 0.6918672323226929,
+      "learning_rate": 4.861626961495103e-06,
+      "loss": 0.694,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6634349030470914,
+      "grad_norm": 0.6631799340248108,
+      "learning_rate": 4.861507422729599e-06,
+      "loss": 0.7189,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6637119113573408,
+      "grad_norm": 0.6965389847755432,
+      "learning_rate": 4.86138783382326e-06,
+      "loss": 0.6919,
+      "step": 2396
+    },
+    {
+      "epoch": 0.66398891966759,
+      "grad_norm": 0.6833652257919312,
+      "learning_rate": 4.861268194778629e-06,
+      "loss": 0.6663,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6642659279778393,
+      "grad_norm": 0.6862128973007202,
+      "learning_rate": 4.8611485055982424e-06,
+      "loss": 0.686,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6645429362880887,
+      "grad_norm": 0.6753442883491516,
+      "learning_rate": 4.861028766284645e-06,
+      "loss": 0.6683,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6648199445983379,
+      "grad_norm": 0.7209076285362244,
+      "learning_rate": 4.860908976840376e-06,
+      "loss": 0.6523,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6650969529085873,
+      "grad_norm": 0.6486983895301819,
+      "learning_rate": 4.860789137267982e-06,
+      "loss": 0.6799,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6653739612188365,
+      "grad_norm": 0.6575525999069214,
+      "learning_rate": 4.860669247570004e-06,
+      "loss": 0.695,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6656509695290859,
+      "grad_norm": 0.7097014784812927,
+      "learning_rate": 4.86054930774899e-06,
+      "loss": 0.6985,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6659279778393352,
+      "grad_norm": 0.7121009826660156,
+      "learning_rate": 4.860429317807487e-06,
+      "loss": 0.6775,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6662049861495845,
+      "grad_norm": 0.6753017902374268,
+      "learning_rate": 4.860309277748041e-06,
+      "loss": 0.6775,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6664819944598338,
+      "grad_norm": 0.7212374806404114,
+      "learning_rate": 4.860189187573201e-06,
+      "loss": 0.7011,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6667590027700832,
+      "grad_norm": 0.6950245499610901,
+      "learning_rate": 4.860069047285517e-06,
+      "loss": 0.7065,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6670360110803324,
+      "grad_norm": 0.6570954918861389,
+      "learning_rate": 4.8599488568875415e-06,
+      "loss": 0.6897,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6673130193905817,
+      "grad_norm": 0.7085040807723999,
+      "learning_rate": 4.859828616381824e-06,
+      "loss": 0.7055,
+      "step": 2409
+    },
+    {
+      "epoch": 0.667590027700831,
+      "grad_norm": 0.6904665231704712,
+      "learning_rate": 4.859708325770919e-06,
+      "loss": 0.6502,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6678670360110803,
+      "grad_norm": 0.6556096076965332,
+      "learning_rate": 4.8595879850573805e-06,
+      "loss": 0.7084,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6681440443213297,
+      "grad_norm": 0.6814101934432983,
+      "learning_rate": 4.859467594243763e-06,
+      "loss": 0.6719,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6684210526315789,
+      "grad_norm": 0.6942593455314636,
+      "learning_rate": 4.859347153332623e-06,
+      "loss": 0.7164,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6686980609418283,
+      "grad_norm": 0.7435287237167358,
+      "learning_rate": 4.859226662326518e-06,
+      "loss": 0.6917,
+      "step": 2414
+    },
+    {
+      "epoch": 0.6689750692520776,
+      "grad_norm": 0.6552598476409912,
+      "learning_rate": 4.859106121228006e-06,
+      "loss": 0.6754,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6692520775623269,
+      "grad_norm": 0.748866856098175,
+      "learning_rate": 4.858985530039647e-06,
+      "loss": 0.7039,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6695290858725762,
+      "grad_norm": 0.6530470848083496,
+      "learning_rate": 4.858864888764001e-06,
+      "loss": 0.7223,
+      "step": 2417
+    },
+    {
+      "epoch": 0.6698060941828254,
+      "grad_norm": 0.6766080856323242,
+      "learning_rate": 4.858744197403629e-06,
+      "loss": 0.6461,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6700831024930748,
+      "grad_norm": 0.6829447746276855,
+      "learning_rate": 4.858623455961095e-06,
+      "loss": 0.7021,
+      "step": 2419
+    },
+    {
+      "epoch": 0.6703601108033241,
+      "grad_norm": 0.6978498101234436,
+      "learning_rate": 4.85850266443896e-06,
+      "loss": 0.7025,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6706371191135734,
+      "grad_norm": 0.6810044050216675,
+      "learning_rate": 4.858381822839792e-06,
+      "loss": 0.7148,
+      "step": 2421
+    },
+    {
+      "epoch": 0.6709141274238227,
+      "grad_norm": 0.6638117432594299,
+      "learning_rate": 4.858260931166155e-06,
+      "loss": 0.6814,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6711911357340721,
+      "grad_norm": 0.6710538864135742,
+      "learning_rate": 4.858139989420616e-06,
+      "loss": 0.6737,
+      "step": 2423
+    },
+    {
+      "epoch": 0.6714681440443213,
+      "grad_norm": 0.6491722464561462,
+      "learning_rate": 4.858018997605742e-06,
+      "loss": 0.6752,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6717451523545707,
+      "grad_norm": 0.6852057576179504,
+      "learning_rate": 4.857897955724105e-06,
+      "loss": 0.6988,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6720221606648199,
+      "grad_norm": 0.6496206521987915,
+      "learning_rate": 4.857776863778272e-06,
+      "loss": 0.6858,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6722991689750693,
+      "grad_norm": 0.6961385011672974,
+      "learning_rate": 4.857655721770814e-06,
+      "loss": 0.6716,
+      "step": 2427
+    },
+    {
+      "epoch": 0.6725761772853186,
+      "grad_norm": 0.6473982930183411,
+      "learning_rate": 4.857534529704305e-06,
+      "loss": 0.6943,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6728531855955678,
+      "grad_norm": 0.6823958158493042,
+      "learning_rate": 4.857413287581319e-06,
+      "loss": 0.6508,
+      "step": 2429
+    },
+    {
+      "epoch": 0.6731301939058172,
+      "grad_norm": 0.645851194858551,
+      "learning_rate": 4.857291995404426e-06,
+      "loss": 0.647,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6734072022160665,
+      "grad_norm": 0.6593954563140869,
+      "learning_rate": 4.857170653176206e-06,
+      "loss": 0.6857,
+      "step": 2431
+    },
+    {
+      "epoch": 0.6736842105263158,
+      "grad_norm": 0.7132719159126282,
+      "learning_rate": 4.857049260899233e-06,
+      "loss": 0.7005,
+      "step": 2432
+    },
+    {
+      "epoch": 0.6739612188365651,
+      "grad_norm": 0.6565352082252502,
+      "learning_rate": 4.856927818576084e-06,
+      "loss": 0.6505,
+      "step": 2433
+    },
+    {
+      "epoch": 0.6742382271468144,
+      "grad_norm": 0.7270442843437195,
+      "learning_rate": 4.85680632620934e-06,
+      "loss": 0.7416,
+      "step": 2434
+    },
+    {
+      "epoch": 0.6745152354570637,
+      "grad_norm": 0.6352047920227051,
+      "learning_rate": 4.856684783801579e-06,
+      "loss": 0.656,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6747922437673131,
+      "grad_norm": 0.6524407863616943,
+      "learning_rate": 4.8565631913553805e-06,
+      "loss": 0.6412,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6750692520775623,
+      "grad_norm": 0.6914234161376953,
+      "learning_rate": 4.8564415488733275e-06,
+      "loss": 0.7128,
+      "step": 2437
+    },
+    {
+      "epoch": 0.6753462603878116,
+      "grad_norm": 0.6721922755241394,
+      "learning_rate": 4.856319856358004e-06,
+      "loss": 0.6491,
+      "step": 2438
+    },
+    {
+      "epoch": 0.675623268698061,
+      "grad_norm": 0.6697131395339966,
+      "learning_rate": 4.856198113811992e-06,
+      "loss": 0.7048,
+      "step": 2439
+    },
+    {
+      "epoch": 0.6759002770083102,
+      "grad_norm": 0.6895703077316284,
+      "learning_rate": 4.856076321237876e-06,
+      "loss": 0.6941,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6761772853185596,
+      "grad_norm": 0.7259820103645325,
+      "learning_rate": 4.8559544786382436e-06,
+      "loss": 0.6767,
+      "step": 2441
+    },
+    {
+      "epoch": 0.6764542936288088,
+      "grad_norm": 0.6969472169876099,
+      "learning_rate": 4.8558325860156815e-06,
+      "loss": 0.7194,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6767313019390582,
+      "grad_norm": 0.7093849778175354,
+      "learning_rate": 4.855710643372777e-06,
+      "loss": 0.7205,
+      "step": 2443
+    },
+    {
+      "epoch": 0.6770083102493075,
+      "grad_norm": 0.6748213768005371,
+      "learning_rate": 4.85558865071212e-06,
+      "loss": 0.6662,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6772853185595568,
+      "grad_norm": 0.7034565806388855,
+      "learning_rate": 4.855466608036301e-06,
+      "loss": 0.7279,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6775623268698061,
+      "grad_norm": 0.6898751854896545,
+      "learning_rate": 4.855344515347909e-06,
+      "loss": 0.6723,
+      "step": 2446
+    },
+    {
+      "epoch": 0.6778393351800553,
+      "grad_norm": 0.6853146553039551,
+      "learning_rate": 4.8552223726495395e-06,
+      "loss": 0.6965,
+      "step": 2447
+    },
+    {
+      "epoch": 0.6781163434903047,
+      "grad_norm": 0.6978676319122314,
+      "learning_rate": 4.855100179943784e-06,
+      "loss": 0.6788,
+      "step": 2448
+    },
+    {
+      "epoch": 0.678393351800554,
+      "grad_norm": 0.7059216499328613,
+      "learning_rate": 4.854977937233238e-06,
+      "loss": 0.6664,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6786703601108033,
+      "grad_norm": 0.735381007194519,
+      "learning_rate": 4.854855644520495e-06,
+      "loss": 0.675,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6789473684210526,
+      "grad_norm": 0.6741946339607239,
+      "learning_rate": 4.854733301808154e-06,
+      "loss": 0.7043,
+      "step": 2451
+    },
+    {
+      "epoch": 0.679224376731302,
+      "grad_norm": 0.6989836096763611,
+      "learning_rate": 4.854610909098813e-06,
+      "loss": 0.6898,
+      "step": 2452
+    },
+    {
+      "epoch": 0.6795013850415512,
+      "grad_norm": 0.7232776284217834,
+      "learning_rate": 4.854488466395067e-06,
+      "loss": 0.7202,
+      "step": 2453
+    },
+    {
+      "epoch": 0.6797783933518006,
+      "grad_norm": 0.6528100967407227,
+      "learning_rate": 4.854365973699519e-06,
+      "loss": 0.6509,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6800554016620498,
+      "grad_norm": 0.703453540802002,
+      "learning_rate": 4.8542434310147694e-06,
+      "loss": 0.7101,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6803324099722992,
+      "grad_norm": 0.6413300037384033,
+      "learning_rate": 4.854120838343419e-06,
+      "loss": 0.6905,
+      "step": 2456
+    },
+    {
+      "epoch": 0.6806094182825485,
+      "grad_norm": 0.6912885904312134,
+      "learning_rate": 4.853998195688072e-06,
+      "loss": 0.6881,
+      "step": 2457
+    },
+    {
+      "epoch": 0.6808864265927977,
+      "grad_norm": 0.6981949210166931,
+      "learning_rate": 4.853875503051332e-06,
+      "loss": 0.674,
+      "step": 2458
+    },
+    {
+      "epoch": 0.6811634349030471,
+      "grad_norm": 0.6651879549026489,
+      "learning_rate": 4.853752760435803e-06,
+      "loss": 0.6696,
+      "step": 2459
+    },
+    {
+      "epoch": 0.6814404432132964,
+      "grad_norm": 0.6497887969017029,
+      "learning_rate": 4.853629967844093e-06,
+      "loss": 0.639,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6817174515235457,
+      "grad_norm": 0.6909531950950623,
+      "learning_rate": 4.853507125278808e-06,
+      "loss": 0.7339,
+      "step": 2461
+    },
+    {
+      "epoch": 0.681994459833795,
+      "grad_norm": 0.6893798112869263,
+      "learning_rate": 4.853384232742556e-06,
+      "loss": 0.677,
+      "step": 2462
+    },
+    {
+      "epoch": 0.6822714681440443,
+      "grad_norm": 0.6748727560043335,
+      "learning_rate": 4.853261290237947e-06,
+      "loss": 0.6947,
+      "step": 2463
+    },
+    {
+      "epoch": 0.6825484764542936,
+      "grad_norm": 0.6517544984817505,
+      "learning_rate": 4.853138297767592e-06,
+      "loss": 0.652,
+      "step": 2464
+    },
+    {
+      "epoch": 0.682825484764543,
+      "grad_norm": 0.6731075644493103,
+      "learning_rate": 4.853015255334101e-06,
+      "loss": 0.7087,
+      "step": 2465
+    },
+    {
+      "epoch": 0.6831024930747922,
+      "grad_norm": 0.6525775194168091,
+      "learning_rate": 4.852892162940088e-06,
+      "loss": 0.6077,
+      "step": 2466
+    },
+    {
+      "epoch": 0.6833795013850416,
+      "grad_norm": 0.6631868481636047,
+      "learning_rate": 4.852769020588165e-06,
+      "loss": 0.7125,
+      "step": 2467
+    },
+    {
+      "epoch": 0.6836565096952909,
+      "grad_norm": 0.6692011952400208,
+      "learning_rate": 4.852645828280948e-06,
+      "loss": 0.6796,
+      "step": 2468
+    },
+    {
+      "epoch": 0.6839335180055401,
+      "grad_norm": 0.6942884922027588,
+      "learning_rate": 4.852522586021051e-06,
+      "loss": 0.6865,
+      "step": 2469
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.6816681027412415,
+      "learning_rate": 4.852399293811093e-06,
+      "loss": 0.6727,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6844875346260387,
+      "grad_norm": 0.7101230025291443,
+      "learning_rate": 4.85227595165369e-06,
+      "loss": 0.7104,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6847645429362881,
+      "grad_norm": 0.6951206922531128,
+      "learning_rate": 4.852152559551461e-06,
+      "loss": 0.6908,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6850415512465374,
+      "grad_norm": 0.7145783305168152,
+      "learning_rate": 4.852029117507027e-06,
+      "loss": 0.6956,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6853185595567867,
+      "grad_norm": 0.6997228860855103,
+      "learning_rate": 4.851905625523009e-06,
+      "loss": 0.6977,
+      "step": 2474
+    },
+    {
+      "epoch": 0.685595567867036,
+      "grad_norm": 0.6776831150054932,
+      "learning_rate": 4.851782083602027e-06,
+      "loss": 0.6397,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6858725761772854,
+      "grad_norm": 0.6669095754623413,
+      "learning_rate": 4.851658491746707e-06,
+      "loss": 0.6706,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6861495844875346,
+      "grad_norm": 0.7028627991676331,
+      "learning_rate": 4.851534849959671e-06,
+      "loss": 0.6527,
+      "step": 2477
+    },
+    {
+      "epoch": 0.686426592797784,
+      "grad_norm": 0.6905294060707092,
+      "learning_rate": 4.8514111582435456e-06,
+      "loss": 0.7172,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6867036011080332,
+      "grad_norm": 0.6870238184928894,
+      "learning_rate": 4.851287416600955e-06,
+      "loss": 0.6673,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6869806094182825,
+      "grad_norm": 0.6995295286178589,
+      "learning_rate": 4.851163625034529e-06,
+      "loss": 0.6624,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6872576177285319,
+      "grad_norm": 0.6662357449531555,
+      "learning_rate": 4.851039783546895e-06,
+      "loss": 0.6174,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6875346260387811,
+      "grad_norm": 0.620506763458252,
+      "learning_rate": 4.850915892140683e-06,
+      "loss": 0.6799,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6878116343490305,
+      "grad_norm": 0.6330281496047974,
+      "learning_rate": 4.850791950818522e-06,
+      "loss": 0.6338,
+      "step": 2483
+    },
+    {
+      "epoch": 0.6880886426592798,
+      "grad_norm": 0.6967810988426208,
+      "learning_rate": 4.850667959583045e-06,
+      "loss": 0.7018,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6883656509695291,
+      "grad_norm": 0.6551015377044678,
+      "learning_rate": 4.850543918436885e-06,
+      "loss": 0.6489,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6886426592797784,
+      "grad_norm": 0.6983478665351868,
+      "learning_rate": 4.8504198273826745e-06,
+      "loss": 0.7017,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6889196675900277,
+      "grad_norm": 0.6824020743370056,
+      "learning_rate": 4.850295686423048e-06,
+      "loss": 0.7076,
+      "step": 2487
+    },
+    {
+      "epoch": 0.689196675900277,
+      "grad_norm": 0.6972485780715942,
+      "learning_rate": 4.850171495560643e-06,
+      "loss": 0.683,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6894736842105263,
+      "grad_norm": 0.684066653251648,
+      "learning_rate": 4.850047254798095e-06,
+      "loss": 0.6491,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6897506925207756,
+      "grad_norm": 0.6561121940612793,
+      "learning_rate": 4.8499229641380425e-06,
+      "loss": 0.644,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6900277008310249,
+      "grad_norm": 0.6939969062805176,
+      "learning_rate": 4.849798623583124e-06,
+      "loss": 0.6638,
+      "step": 2491
+    },
+    {
+      "epoch": 0.6903047091412743,
+      "grad_norm": 0.6910930871963501,
+      "learning_rate": 4.8496742331359805e-06,
+      "loss": 0.6759,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6905817174515235,
+      "grad_norm": 0.7051935791969299,
+      "learning_rate": 4.849549792799253e-06,
+      "loss": 0.6873,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6908587257617729,
+      "grad_norm": 0.678946316242218,
+      "learning_rate": 4.849425302575583e-06,
+      "loss": 0.6753,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6911357340720221,
+      "grad_norm": 0.6918194890022278,
+      "learning_rate": 4.849300762467613e-06,
+      "loss": 0.7027,
+      "step": 2495
+    },
+    {
+      "epoch": 0.6914127423822715,
+      "grad_norm": 0.7290768623352051,
+      "learning_rate": 4.8491761724779905e-06,
+      "loss": 0.7097,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6916897506925208,
+      "grad_norm": 0.6682461500167847,
+      "learning_rate": 4.849051532609357e-06,
+      "loss": 0.6562,
+      "step": 2497
+    },
+    {
+      "epoch": 0.69196675900277,
+      "grad_norm": 0.7157434225082397,
+      "learning_rate": 4.848926842864361e-06,
+      "loss": 0.71,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6922437673130194,
+      "grad_norm": 0.6673200130462646,
+      "learning_rate": 4.84880210324565e-06,
+      "loss": 0.6792,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6925207756232687,
+      "grad_norm": 0.7229270935058594,
+      "learning_rate": 4.848677313755872e-06,
+      "loss": 0.6921,
+      "step": 2500
+    },
+    {
+      "epoch": 0.692797783933518,
+      "grad_norm": 0.6955553889274597,
+      "learning_rate": 4.848552474397677e-06,
+      "loss": 0.7078,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6930747922437673,
+      "grad_norm": 0.6488661170005798,
+      "learning_rate": 4.848427585173715e-06,
+      "loss": 0.6502,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6933518005540166,
+      "grad_norm": 0.6416786313056946,
+      "learning_rate": 4.848302646086638e-06,
+      "loss": 0.641,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6936288088642659,
+      "grad_norm": 0.6498820781707764,
+      "learning_rate": 4.848177657139099e-06,
+      "loss": 0.6882,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6939058171745153,
+      "grad_norm": 0.6757896542549133,
+      "learning_rate": 4.848052618333751e-06,
+      "loss": 0.6767,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6941828254847645,
+      "grad_norm": 0.669786274433136,
+      "learning_rate": 4.8479275296732514e-06,
+      "loss": 0.6998,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6944598337950139,
+      "grad_norm": 0.6754154562950134,
+      "learning_rate": 4.847802391160253e-06,
+      "loss": 0.6024,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6947368421052632,
+      "grad_norm": 0.7129898071289062,
+      "learning_rate": 4.847677202797415e-06,
+      "loss": 0.6552,
+      "step": 2508
+    },
+    {
+      "epoch": 0.6950138504155124,
+      "grad_norm": 0.6702117323875427,
+      "learning_rate": 4.847551964587395e-06,
+      "loss": 0.6768,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6952908587257618,
+      "grad_norm": 0.6439821124076843,
+      "learning_rate": 4.847426676532851e-06,
+      "loss": 0.6517,
+      "step": 2510
+    },
+    {
+      "epoch": 0.695567867036011,
+      "grad_norm": 0.6356789469718933,
+      "learning_rate": 4.847301338636444e-06,
+      "loss": 0.6824,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6958448753462604,
+      "grad_norm": 0.7396525144577026,
+      "learning_rate": 4.847175950900836e-06,
+      "loss": 0.6814,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6961218836565097,
+      "grad_norm": 0.7004340887069702,
+      "learning_rate": 4.847050513328687e-06,
+      "loss": 0.6993,
+      "step": 2513
+    },
+    {
+      "epoch": 0.696398891966759,
+      "grad_norm": 0.6686164736747742,
+      "learning_rate": 4.846925025922663e-06,
+      "loss": 0.6608,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6966759002770083,
+      "grad_norm": 0.6834436058998108,
+      "learning_rate": 4.846799488685427e-06,
+      "loss": 0.6856,
+      "step": 2515
+    },
+    {
+      "epoch": 0.6969529085872577,
+      "grad_norm": 0.6966507434844971,
+      "learning_rate": 4.846673901619645e-06,
+      "loss": 0.6928,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6972299168975069,
+      "grad_norm": 0.6435982584953308,
+      "learning_rate": 4.846548264727984e-06,
+      "loss": 0.6524,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6975069252077563,
+      "grad_norm": 0.6690963506698608,
+      "learning_rate": 4.84642257801311e-06,
+      "loss": 0.6787,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6977839335180055,
+      "grad_norm": 0.6870619058609009,
+      "learning_rate": 4.846296841477692e-06,
+      "loss": 0.6838,
+      "step": 2519
+    },
+    {
+      "epoch": 0.6980609418282548,
+      "grad_norm": 0.6921945810317993,
+      "learning_rate": 4.846171055124401e-06,
+      "loss": 0.6786,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6983379501385042,
+      "grad_norm": 0.6810411810874939,
+      "learning_rate": 4.846045218955907e-06,
+      "loss": 0.6997,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6986149584487534,
+      "grad_norm": 0.7256765961647034,
+      "learning_rate": 4.845919332974882e-06,
+      "loss": 0.6654,
+      "step": 2522
+    },
+    {
+      "epoch": 0.6988919667590028,
+      "grad_norm": 0.706616222858429,
+      "learning_rate": 4.845793397183999e-06,
+      "loss": 0.6626,
+      "step": 2523
+    },
+    {
+      "epoch": 0.6991689750692521,
+      "grad_norm": 0.68105548620224,
+      "learning_rate": 4.845667411585931e-06,
+      "loss": 0.6739,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6994459833795014,
+      "grad_norm": 0.6390605568885803,
+      "learning_rate": 4.845541376183354e-06,
+      "loss": 0.6434,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6997229916897507,
+      "grad_norm": 0.6960382461547852,
+      "learning_rate": 4.8454152909789435e-06,
+      "loss": 0.6883,
+      "step": 2526
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6883596181869507,
+      "learning_rate": 4.845289155975378e-06,
+      "loss": 0.7118,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7002770083102493,
+      "grad_norm": 0.6748420596122742,
+      "learning_rate": 4.845162971175333e-06,
+      "loss": 0.6885,
+      "step": 2528
+    },
+    {
+      "epoch": 0.7005540166204987,
+      "grad_norm": 0.6539015173912048,
+      "learning_rate": 4.8450367365814894e-06,
+      "loss": 0.661,
+      "step": 2529
+    },
+    {
+      "epoch": 0.7008310249307479,
+      "grad_norm": 0.6839054822921753,
+      "learning_rate": 4.844910452196527e-06,
+      "loss": 0.683,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7011080332409972,
+      "grad_norm": 0.6770256757736206,
+      "learning_rate": 4.844784118023129e-06,
+      "loss": 0.6966,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7013850415512466,
+      "grad_norm": 0.6649383902549744,
+      "learning_rate": 4.844657734063975e-06,
+      "loss": 0.7109,
+      "step": 2532
+    },
+    {
+      "epoch": 0.7016620498614958,
+      "grad_norm": 0.7795406579971313,
+      "learning_rate": 4.84453130032175e-06,
+      "loss": 0.7088,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7019390581717452,
+      "grad_norm": 0.65948486328125,
+      "learning_rate": 4.8444048167991386e-06,
+      "loss": 0.6543,
+      "step": 2534
+    },
+    {
+      "epoch": 0.7022160664819944,
+      "grad_norm": 0.7084441781044006,
+      "learning_rate": 4.844278283498826e-06,
+      "loss": 0.6912,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7024930747922438,
+      "grad_norm": 0.6589716672897339,
+      "learning_rate": 4.844151700423498e-06,
+      "loss": 0.6528,
+      "step": 2536
+    },
+    {
+      "epoch": 0.7027700831024931,
+      "grad_norm": 0.6785886883735657,
+      "learning_rate": 4.844025067575844e-06,
+      "loss": 0.6702,
+      "step": 2537
+    },
+    {
+      "epoch": 0.7030470914127424,
+      "grad_norm": 0.6644973754882812,
+      "learning_rate": 4.843898384958551e-06,
+      "loss": 0.6606,
+      "step": 2538
+    },
+    {
+      "epoch": 0.7033240997229917,
+      "grad_norm": 0.6794224381446838,
+      "learning_rate": 4.84377165257431e-06,
+      "loss": 0.6894,
+      "step": 2539
+    },
+    {
+      "epoch": 0.703601108033241,
+      "grad_norm": 0.6537707448005676,
+      "learning_rate": 4.843644870425811e-06,
+      "loss": 0.646,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7038781163434903,
+      "grad_norm": 0.662074625492096,
+      "learning_rate": 4.843518038515747e-06,
+      "loss": 0.6465,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7041551246537396,
+      "grad_norm": 0.6878114342689514,
+      "learning_rate": 4.843391156846811e-06,
+      "loss": 0.6566,
+      "step": 2542
+    },
+    {
+      "epoch": 0.7044321329639889,
+      "grad_norm": 0.666167676448822,
+      "learning_rate": 4.8432642254216945e-06,
+      "loss": 0.6513,
+      "step": 2543
+    },
+    {
+      "epoch": 0.7047091412742382,
+      "grad_norm": 0.6801060438156128,
+      "learning_rate": 4.843137244243095e-06,
+      "loss": 0.6763,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7049861495844876,
+      "grad_norm": 0.7075893878936768,
+      "learning_rate": 4.843010213313709e-06,
+      "loss": 0.7008,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7052631578947368,
+      "grad_norm": 0.6655442714691162,
+      "learning_rate": 4.842883132636233e-06,
+      "loss": 0.6849,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7055401662049862,
+      "grad_norm": 0.6841294765472412,
+      "learning_rate": 4.842756002213364e-06,
+      "loss": 0.6742,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7058171745152355,
+      "grad_norm": 0.6728150248527527,
+      "learning_rate": 4.842628822047803e-06,
+      "loss": 0.7086,
+      "step": 2548
+    },
+    {
+      "epoch": 0.7060941828254848,
+      "grad_norm": 0.7174102663993835,
+      "learning_rate": 4.8425015921422495e-06,
+      "loss": 0.7403,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7063711911357341,
+      "grad_norm": 0.6728398203849792,
+      "learning_rate": 4.842374312499405e-06,
+      "loss": 0.6833,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7066481994459833,
+      "grad_norm": 0.6851624250411987,
+      "learning_rate": 4.842246983121972e-06,
+      "loss": 0.6733,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7069252077562327,
+      "grad_norm": 0.7019652724266052,
+      "learning_rate": 4.842119604012654e-06,
+      "loss": 0.618,
+      "step": 2552
+    },
+    {
+      "epoch": 0.707202216066482,
+      "grad_norm": 0.6876543164253235,
+      "learning_rate": 4.841992175174156e-06,
+      "loss": 0.6982,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7074792243767313,
+      "grad_norm": 0.6728835701942444,
+      "learning_rate": 4.841864696609183e-06,
+      "loss": 0.6736,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7077562326869806,
+      "grad_norm": 0.7015979886054993,
+      "learning_rate": 4.841737168320443e-06,
+      "loss": 0.6532,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7080332409972299,
+      "grad_norm": 0.6520795226097107,
+      "learning_rate": 4.8416095903106415e-06,
+      "loss": 0.6563,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7083102493074792,
+      "grad_norm": 0.6712133884429932,
+      "learning_rate": 4.841481962582489e-06,
+      "loss": 0.712,
+      "step": 2557
+    },
+    {
+      "epoch": 0.7085872576177286,
+      "grad_norm": 0.7220582365989685,
+      "learning_rate": 4.841354285138695e-06,
+      "loss": 0.6635,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7088642659279778,
+      "grad_norm": 0.6810282468795776,
+      "learning_rate": 4.84122655798197e-06,
+      "loss": 0.6837,
+      "step": 2559
+    },
+    {
+      "epoch": 0.7091412742382271,
+      "grad_norm": 0.6945074200630188,
+      "learning_rate": 4.841098781115027e-06,
+      "loss": 0.6894,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7094182825484765,
+      "grad_norm": 0.6712693572044373,
+      "learning_rate": 4.840970954540578e-06,
+      "loss": 0.6993,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7096952908587257,
+      "grad_norm": 0.6899521946907043,
+      "learning_rate": 4.840843078261338e-06,
+      "loss": 0.6969,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7099722991689751,
+      "grad_norm": 0.6725625991821289,
+      "learning_rate": 4.840715152280021e-06,
+      "loss": 0.7176,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7102493074792243,
+      "grad_norm": 0.7018441557884216,
+      "learning_rate": 4.8405871765993435e-06,
+      "loss": 0.7118,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7105263157894737,
+      "grad_norm": 0.6593537330627441,
+      "learning_rate": 4.840459151222024e-06,
+      "loss": 0.6727,
+      "step": 2565
+    },
+    {
+      "epoch": 0.710803324099723,
+      "grad_norm": 0.6770919561386108,
+      "learning_rate": 4.840331076150778e-06,
+      "loss": 0.68,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7110803324099723,
+      "grad_norm": 0.6419566869735718,
+      "learning_rate": 4.840202951388328e-06,
+      "loss": 0.6911,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7113573407202216,
+      "grad_norm": 0.6920489072799683,
+      "learning_rate": 4.840074776937393e-06,
+      "loss": 0.6668,
+      "step": 2568
+    },
+    {
+      "epoch": 0.711634349030471,
+      "grad_norm": 0.6773355603218079,
+      "learning_rate": 4.839946552800695e-06,
+      "loss": 0.6783,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7119113573407202,
+      "grad_norm": 0.6632847189903259,
+      "learning_rate": 4.839818278980956e-06,
+      "loss": 0.7495,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7121883656509695,
+      "grad_norm": 0.697297215461731,
+      "learning_rate": 4.839689955480899e-06,
+      "loss": 0.6777,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7124653739612188,
+      "grad_norm": 0.6623373627662659,
+      "learning_rate": 4.8395615823032505e-06,
+      "loss": 0.6826,
+      "step": 2572
+    },
+    {
+      "epoch": 0.7127423822714681,
+      "grad_norm": 0.6580797433853149,
+      "learning_rate": 4.839433159450734e-06,
+      "loss": 0.6558,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7130193905817175,
+      "grad_norm": 0.6638851761817932,
+      "learning_rate": 4.839304686926077e-06,
+      "loss": 0.6596,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7132963988919667,
+      "grad_norm": 0.7214973568916321,
+      "learning_rate": 4.839176164732009e-06,
+      "loss": 0.6816,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7135734072022161,
+      "grad_norm": 0.6427276730537415,
+      "learning_rate": 4.839047592871257e-06,
+      "loss": 0.6325,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7138504155124654,
+      "grad_norm": 0.7028458714485168,
+      "learning_rate": 4.83891897134655e-06,
+      "loss": 0.6696,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7141274238227147,
+      "grad_norm": 0.6528476476669312,
+      "learning_rate": 4.838790300160621e-06,
+      "loss": 0.6547,
+      "step": 2578
+    },
+    {
+      "epoch": 0.714404432132964,
+      "grad_norm": 0.7301563024520874,
+      "learning_rate": 4.838661579316202e-06,
+      "loss": 0.7337,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7146814404432132,
+      "grad_norm": 0.6242020726203918,
+      "learning_rate": 4.838532808816025e-06,
+      "loss": 0.5942,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7149584487534626,
+      "grad_norm": 0.6679527759552002,
+      "learning_rate": 4.838403988662824e-06,
+      "loss": 0.6418,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7152354570637119,
+      "grad_norm": 0.6938562393188477,
+      "learning_rate": 4.838275118859335e-06,
+      "loss": 0.6301,
+      "step": 2582
+    },
+    {
+      "epoch": 0.7155124653739612,
+      "grad_norm": 0.6652330756187439,
+      "learning_rate": 4.838146199408294e-06,
+      "loss": 0.6818,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7157894736842105,
+      "grad_norm": 0.6956022381782532,
+      "learning_rate": 4.838017230312438e-06,
+      "loss": 0.6836,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7160664819944599,
+      "grad_norm": 0.714000403881073,
+      "learning_rate": 4.8378882115745064e-06,
+      "loss": 0.7088,
+      "step": 2585
+    },
+    {
+      "epoch": 0.7163434903047091,
+      "grad_norm": 0.7130530476570129,
+      "learning_rate": 4.837759143197237e-06,
+      "loss": 0.7168,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7166204986149585,
+      "grad_norm": 0.6532400846481323,
+      "learning_rate": 4.837630025183371e-06,
+      "loss": 0.6817,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7168975069252077,
+      "grad_norm": 0.6796839833259583,
+      "learning_rate": 4.83750085753565e-06,
+      "loss": 0.7166,
+      "step": 2588
+    },
+    {
+      "epoch": 0.717174515235457,
+      "grad_norm": 0.6681724786758423,
+      "learning_rate": 4.837371640256816e-06,
+      "loss": 0.6278,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7174515235457064,
+      "grad_norm": 0.6967169642448425,
+      "learning_rate": 4.837242373349614e-06,
+      "loss": 0.6693,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7177285318559556,
+      "grad_norm": 0.6756342053413391,
+      "learning_rate": 4.837113056816787e-06,
+      "loss": 0.7269,
+      "step": 2591
+    },
+    {
+      "epoch": 0.718005540166205,
+      "grad_norm": 0.6517120599746704,
+      "learning_rate": 4.836983690661082e-06,
+      "loss": 0.6512,
+      "step": 2592
+    },
+    {
+      "epoch": 0.7182825484764543,
+      "grad_norm": 0.6642926931381226,
+      "learning_rate": 4.8368542748852445e-06,
+      "loss": 0.6615,
+      "step": 2593
+    },
+    {
+      "epoch": 0.7185595567867036,
+      "grad_norm": 0.7000516653060913,
+      "learning_rate": 4.836724809492024e-06,
+      "loss": 0.6891,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7188365650969529,
+      "grad_norm": 0.6529144644737244,
+      "learning_rate": 4.836595294484168e-06,
+      "loss": 0.6853,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7191135734072022,
+      "grad_norm": 0.7132996320724487,
+      "learning_rate": 4.836465729864427e-06,
+      "loss": 0.6745,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7193905817174515,
+      "grad_norm": 0.6974446177482605,
+      "learning_rate": 4.836336115635551e-06,
+      "loss": 0.6257,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7196675900277009,
+      "grad_norm": 0.6566895842552185,
+      "learning_rate": 4.836206451800294e-06,
+      "loss": 0.6664,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7199445983379501,
+      "grad_norm": 0.7557151913642883,
+      "learning_rate": 4.836076738361408e-06,
+      "loss": 0.7274,
+      "step": 2599
+    },
+    {
+      "epoch": 0.7202216066481995,
+      "grad_norm": 0.6727636456489563,
+      "learning_rate": 4.835946975321647e-06,
+      "loss": 0.6644,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7204986149584488,
+      "grad_norm": 0.6884294748306274,
+      "learning_rate": 4.835817162683765e-06,
+      "loss": 0.6728,
+      "step": 2601
+    },
+    {
+      "epoch": 0.720775623268698,
+      "grad_norm": 0.6941688060760498,
+      "learning_rate": 4.835687300450522e-06,
+      "loss": 0.6552,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7210526315789474,
+      "grad_norm": 0.6695191860198975,
+      "learning_rate": 4.8355573886246715e-06,
+      "loss": 0.7092,
+      "step": 2603
+    },
+    {
+      "epoch": 0.7213296398891966,
+      "grad_norm": 0.7561391592025757,
+      "learning_rate": 4.835427427208973e-06,
+      "loss": 0.6844,
+      "step": 2604
+    },
+    {
+      "epoch": 0.721606648199446,
+      "grad_norm": 0.7127019166946411,
+      "learning_rate": 4.835297416206187e-06,
+      "loss": 0.6553,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7218836565096953,
+      "grad_norm": 0.6892712116241455,
+      "learning_rate": 4.835167355619073e-06,
+      "loss": 0.7134,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7221606648199446,
+      "grad_norm": 0.6294607520103455,
+      "learning_rate": 4.835037245450393e-06,
+      "loss": 0.6248,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7224376731301939,
+      "grad_norm": 0.7269555926322937,
+      "learning_rate": 4.834907085702909e-06,
+      "loss": 0.6899,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7227146814404433,
+      "grad_norm": 0.6997284889221191,
+      "learning_rate": 4.834776876379384e-06,
+      "loss": 0.6865,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7229916897506925,
+      "grad_norm": 0.6564635634422302,
+      "learning_rate": 4.834646617482585e-06,
+      "loss": 0.6942,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7232686980609419,
+      "grad_norm": 0.6637692451477051,
+      "learning_rate": 4.834516309015274e-06,
+      "loss": 0.6534,
+      "step": 2611
+    },
+    {
+      "epoch": 0.7235457063711911,
+      "grad_norm": 0.6564186215400696,
+      "learning_rate": 4.834385950980222e-06,
+      "loss": 0.6673,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7238227146814404,
+      "grad_norm": 0.669204592704773,
+      "learning_rate": 4.834255543380195e-06,
+      "loss": 0.6218,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7240997229916898,
+      "grad_norm": 0.6596257090568542,
+      "learning_rate": 4.834125086217961e-06,
+      "loss": 0.6568,
+      "step": 2614
+    },
+    {
+      "epoch": 0.724376731301939,
+      "grad_norm": 0.6701518893241882,
+      "learning_rate": 4.833994579496291e-06,
+      "loss": 0.7237,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7246537396121884,
+      "grad_norm": 0.6752590537071228,
+      "learning_rate": 4.833864023217955e-06,
+      "loss": 0.7239,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7249307479224377,
+      "grad_norm": 0.7187966108322144,
+      "learning_rate": 4.8337334173857265e-06,
+      "loss": 0.7146,
+      "step": 2617
+    },
+    {
+      "epoch": 0.725207756232687,
+      "grad_norm": 0.740940511226654,
+      "learning_rate": 4.833602762002378e-06,
+      "loss": 0.735,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7254847645429363,
+      "grad_norm": 0.7270171642303467,
+      "learning_rate": 4.8334720570706825e-06,
+      "loss": 0.6751,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7257617728531855,
+      "grad_norm": 0.6728403568267822,
+      "learning_rate": 4.833341302593417e-06,
+      "loss": 0.6642,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7260387811634349,
+      "grad_norm": 0.6736679077148438,
+      "learning_rate": 4.833210498573358e-06,
+      "loss": 0.7014,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7263157894736842,
+      "grad_norm": 0.6557801365852356,
+      "learning_rate": 4.83307964501328e-06,
+      "loss": 0.6956,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7265927977839335,
+      "grad_norm": 0.6692063212394714,
+      "learning_rate": 4.832948741915963e-06,
+      "loss": 0.7072,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7268698060941828,
+      "grad_norm": 0.6571874022483826,
+      "learning_rate": 4.832817789284188e-06,
+      "loss": 0.6842,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7271468144044322,
+      "grad_norm": 0.6895356178283691,
+      "learning_rate": 4.832686787120733e-06,
+      "loss": 0.7092,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7274238227146814,
+      "grad_norm": 0.6702733635902405,
+      "learning_rate": 4.83255573542838e-06,
+      "loss": 0.6036,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7277008310249308,
+      "grad_norm": 0.7040365934371948,
+      "learning_rate": 4.832424634209913e-06,
+      "loss": 0.6881,
+      "step": 2627
+    },
+    {
+      "epoch": 0.72797783933518,
+      "grad_norm": 0.6673402786254883,
+      "learning_rate": 4.832293483468115e-06,
+      "loss": 0.6887,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7282548476454294,
+      "grad_norm": 0.6921663284301758,
+      "learning_rate": 4.8321622832057695e-06,
+      "loss": 0.6798,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7285318559556787,
+      "grad_norm": 0.6102813482284546,
+      "learning_rate": 4.832031033425663e-06,
+      "loss": 0.622,
+      "step": 2630
+    },
+    {
+      "epoch": 0.728808864265928,
+      "grad_norm": 0.6588747501373291,
+      "learning_rate": 4.8318997341305815e-06,
+      "loss": 0.6443,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7290858725761773,
+      "grad_norm": 0.6669639945030212,
+      "learning_rate": 4.831768385323315e-06,
+      "loss": 0.6138,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7293628808864266,
+      "grad_norm": 0.6380589008331299,
+      "learning_rate": 4.83163698700665e-06,
+      "loss": 0.6112,
+      "step": 2633
+    },
+    {
+      "epoch": 0.7296398891966759,
+      "grad_norm": 0.7220094203948975,
+      "learning_rate": 4.8315055391833785e-06,
+      "loss": 0.6781,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7299168975069252,
+      "grad_norm": 0.6523056030273438,
+      "learning_rate": 4.8313740418562895e-06,
+      "loss": 0.656,
+      "step": 2635
+    },
+    {
+      "epoch": 0.7301939058171745,
+      "grad_norm": 0.679682195186615,
+      "learning_rate": 4.831242495028177e-06,
+      "loss": 0.6812,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7304709141274238,
+      "grad_norm": 0.6778846979141235,
+      "learning_rate": 4.831110898701832e-06,
+      "loss": 0.658,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7307479224376732,
+      "grad_norm": 0.6646358370780945,
+      "learning_rate": 4.83097925288005e-06,
+      "loss": 0.7349,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7310249307479224,
+      "grad_norm": 0.6693096160888672,
+      "learning_rate": 4.830847557565626e-06,
+      "loss": 0.686,
+      "step": 2639
+    },
+    {
+      "epoch": 0.7313019390581718,
+      "grad_norm": 0.6840837597846985,
+      "learning_rate": 4.830715812761355e-06,
+      "loss": 0.6764,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7315789473684211,
+      "grad_norm": 0.6790053844451904,
+      "learning_rate": 4.830584018470036e-06,
+      "loss": 0.6649,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7318559556786703,
+      "grad_norm": 0.7182514667510986,
+      "learning_rate": 4.830452174694466e-06,
+      "loss": 0.6707,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7321329639889197,
+      "grad_norm": 0.6911314129829407,
+      "learning_rate": 4.830320281437446e-06,
+      "loss": 0.6779,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7324099722991689,
+      "grad_norm": 0.6859692335128784,
+      "learning_rate": 4.830188338701776e-06,
+      "loss": 0.7284,
+      "step": 2644
+    },
+    {
+      "epoch": 0.7326869806094183,
+      "grad_norm": 0.7071799635887146,
+      "learning_rate": 4.830056346490255e-06,
+      "loss": 0.7353,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7329639889196676,
+      "grad_norm": 0.6324630975723267,
+      "learning_rate": 4.829924304805689e-06,
+      "loss": 0.6674,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7332409972299169,
+      "grad_norm": 0.654015839099884,
+      "learning_rate": 4.829792213650879e-06,
+      "loss": 0.6058,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7335180055401662,
+      "grad_norm": 0.708918035030365,
+      "learning_rate": 4.829660073028631e-06,
+      "loss": 0.677,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7337950138504156,
+      "grad_norm": 0.6400759220123291,
+      "learning_rate": 4.82952788294175e-06,
+      "loss": 0.6864,
+      "step": 2649
+    },
+    {
+      "epoch": 0.7340720221606648,
+      "grad_norm": 0.6708654761314392,
+      "learning_rate": 4.829395643393045e-06,
+      "loss": 0.7214,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7343490304709142,
+      "grad_norm": 0.7093385457992554,
+      "learning_rate": 4.82926335438532e-06,
+      "loss": 0.7227,
+      "step": 2651
+    },
+    {
+      "epoch": 0.7346260387811634,
+      "grad_norm": 0.688284695148468,
+      "learning_rate": 4.829131015921386e-06,
+      "loss": 0.6762,
+      "step": 2652
+    },
+    {
+      "epoch": 0.7349030470914127,
+      "grad_norm": 0.6453286409378052,
+      "learning_rate": 4.828998628004053e-06,
+      "loss": 0.6432,
+      "step": 2653
+    },
+    {
+      "epoch": 0.7351800554016621,
+      "grad_norm": 0.7219470143318176,
+      "learning_rate": 4.82886619063613e-06,
+      "loss": 0.6863,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7354570637119113,
+      "grad_norm": 0.7020809054374695,
+      "learning_rate": 4.828733703820432e-06,
+      "loss": 0.6498,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7357340720221607,
+      "grad_norm": 0.6752793788909912,
+      "learning_rate": 4.8286011675597705e-06,
+      "loss": 0.6503,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7360110803324099,
+      "grad_norm": 0.6839720010757446,
+      "learning_rate": 4.828468581856959e-06,
+      "loss": 0.6151,
+      "step": 2657
+    },
+    {
+      "epoch": 0.7362880886426593,
+      "grad_norm": 0.6902003288269043,
+      "learning_rate": 4.828335946714814e-06,
+      "loss": 0.6585,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7365650969529086,
+      "grad_norm": 0.6736903190612793,
+      "learning_rate": 4.82820326213615e-06,
+      "loss": 0.6276,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.654117226600647,
+      "learning_rate": 4.8280705281237865e-06,
+      "loss": 0.6586,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7371191135734072,
+      "grad_norm": 0.6765998601913452,
+      "learning_rate": 4.827937744680539e-06,
+      "loss": 0.668,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7373961218836566,
+      "grad_norm": 0.6439967155456543,
+      "learning_rate": 4.8278049118092285e-06,
+      "loss": 0.6644,
+      "step": 2662
+    },
+    {
+      "epoch": 0.7376731301939058,
+      "grad_norm": 0.6422175168991089,
+      "learning_rate": 4.827672029512676e-06,
+      "loss": 0.6662,
+      "step": 2663
+    },
+    {
+      "epoch": 0.7379501385041551,
+      "grad_norm": 0.636263370513916,
+      "learning_rate": 4.827539097793702e-06,
+      "loss": 0.6252,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7382271468144044,
+      "grad_norm": 0.695054829120636,
+      "learning_rate": 4.827406116655128e-06,
+      "loss": 0.6781,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7385041551246537,
+      "grad_norm": 0.6738481521606445,
+      "learning_rate": 4.82727308609978e-06,
+      "loss": 0.6835,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7387811634349031,
+      "grad_norm": 0.656569242477417,
+      "learning_rate": 4.82714000613048e-06,
+      "loss": 0.6515,
+      "step": 2667
+    },
+    {
+      "epoch": 0.7390581717451523,
+      "grad_norm": 0.7025095820426941,
+      "learning_rate": 4.827006876750056e-06,
+      "loss": 0.7286,
+      "step": 2668
+    },
+    {
+      "epoch": 0.7393351800554017,
+      "grad_norm": 0.6530948281288147,
+      "learning_rate": 4.826873697961332e-06,
+      "loss": 0.6761,
+      "step": 2669
+    },
+    {
+      "epoch": 0.739612188365651,
+      "grad_norm": 0.6586973071098328,
+      "learning_rate": 4.826740469767138e-06,
+      "loss": 0.6785,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7398891966759003,
+      "grad_norm": 0.661683976650238,
+      "learning_rate": 4.826607192170301e-06,
+      "loss": 0.7015,
+      "step": 2671
+    },
+    {
+      "epoch": 0.7401662049861496,
+      "grad_norm": 0.6833496689796448,
+      "learning_rate": 4.8264738651736534e-06,
+      "loss": 0.6494,
+      "step": 2672
+    },
+    {
+      "epoch": 0.7404432132963988,
+      "grad_norm": 0.7184470891952515,
+      "learning_rate": 4.8263404887800235e-06,
+      "loss": 0.7068,
+      "step": 2673
+    },
+    {
+      "epoch": 0.7407202216066482,
+      "grad_norm": 0.7301485538482666,
+      "learning_rate": 4.826207062992245e-06,
+      "loss": 0.7386,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7409972299168975,
+      "grad_norm": 0.6598369479179382,
+      "learning_rate": 4.826073587813149e-06,
+      "loss": 0.6654,
+      "step": 2675
+    },
+    {
+      "epoch": 0.7412742382271468,
+      "grad_norm": 0.6928764581680298,
+      "learning_rate": 4.825940063245572e-06,
+      "loss": 0.6326,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7415512465373961,
+      "grad_norm": 0.668339192867279,
+      "learning_rate": 4.825806489292346e-06,
+      "loss": 0.643,
+      "step": 2677
+    },
+    {
+      "epoch": 0.7418282548476455,
+      "grad_norm": 0.6804963946342468,
+      "learning_rate": 4.82567286595631e-06,
+      "loss": 0.7222,
+      "step": 2678
+    },
+    {
+      "epoch": 0.7421052631578947,
+      "grad_norm": 0.6823146939277649,
+      "learning_rate": 4.8255391932403e-06,
+      "loss": 0.6546,
+      "step": 2679
+    },
+    {
+      "epoch": 0.7423822714681441,
+      "grad_norm": 0.7106555700302124,
+      "learning_rate": 4.825405471147153e-06,
+      "loss": 0.7349,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7426592797783933,
+      "grad_norm": 0.6523182988166809,
+      "learning_rate": 4.82527169967971e-06,
+      "loss": 0.6886,
+      "step": 2681
+    },
+    {
+      "epoch": 0.7429362880886426,
+      "grad_norm": 0.6475448608398438,
+      "learning_rate": 4.825137878840811e-06,
+      "loss": 0.693,
+      "step": 2682
+    },
+    {
+      "epoch": 0.743213296398892,
+      "grad_norm": 0.7019377946853638,
+      "learning_rate": 4.825004008633297e-06,
+      "loss": 0.6831,
+      "step": 2683
+    },
+    {
+      "epoch": 0.7434903047091412,
+      "grad_norm": 0.6636565327644348,
+      "learning_rate": 4.82487008906001e-06,
+      "loss": 0.6647,
+      "step": 2684
+    },
+    {
+      "epoch": 0.7437673130193906,
+      "grad_norm": 0.6838463544845581,
+      "learning_rate": 4.824736120123794e-06,
+      "loss": 0.7015,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7440443213296399,
+      "grad_norm": 0.6885688900947571,
+      "learning_rate": 4.824602101827494e-06,
+      "loss": 0.6308,
+      "step": 2686
+    },
+    {
+      "epoch": 0.7443213296398892,
+      "grad_norm": 0.7183178663253784,
+      "learning_rate": 4.824468034173954e-06,
+      "loss": 0.686,
+      "step": 2687
+    },
+    {
+      "epoch": 0.7445983379501385,
+      "grad_norm": 0.6711190938949585,
+      "learning_rate": 4.824333917166023e-06,
+      "loss": 0.6677,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7448753462603878,
+      "grad_norm": 0.6922284364700317,
+      "learning_rate": 4.824199750806547e-06,
+      "loss": 0.6963,
+      "step": 2689
+    },
+    {
+      "epoch": 0.7451523545706371,
+      "grad_norm": 0.6533357501029968,
+      "learning_rate": 4.8240655350983745e-06,
+      "loss": 0.627,
+      "step": 2690
+    },
+    {
+      "epoch": 0.7454293628808865,
+      "grad_norm": 0.7120330929756165,
+      "learning_rate": 4.8239312700443555e-06,
+      "loss": 0.6529,
+      "step": 2691
+    },
+    {
+      "epoch": 0.7457063711911357,
+      "grad_norm": 0.6339964866638184,
+      "learning_rate": 4.823796955647341e-06,
+      "loss": 0.6094,
+      "step": 2692
+    },
+    {
+      "epoch": 0.745983379501385,
+      "grad_norm": 0.6632034778594971,
+      "learning_rate": 4.823662591910183e-06,
+      "loss": 0.7077,
+      "step": 2693
+    },
+    {
+      "epoch": 0.7462603878116344,
+      "grad_norm": 0.7076951265335083,
+      "learning_rate": 4.823528178835734e-06,
+      "loss": 0.6762,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7465373961218836,
+      "grad_norm": 0.6844574809074402,
+      "learning_rate": 4.823393716426849e-06,
+      "loss": 0.7062,
+      "step": 2695
+    },
+    {
+      "epoch": 0.746814404432133,
+      "grad_norm": 0.6797798871994019,
+      "learning_rate": 4.82325920468638e-06,
+      "loss": 0.6544,
+      "step": 2696
+    },
+    {
+      "epoch": 0.7470914127423822,
+      "grad_norm": 0.6964753866195679,
+      "learning_rate": 4.823124643617187e-06,
+      "loss": 0.7259,
+      "step": 2697
+    },
+    {
+      "epoch": 0.7473684210526316,
+      "grad_norm": 0.6596205830574036,
+      "learning_rate": 4.822990033222125e-06,
+      "loss": 0.6765,
+      "step": 2698
+    },
+    {
+      "epoch": 0.7476454293628809,
+      "grad_norm": 0.7021011710166931,
+      "learning_rate": 4.822855373504051e-06,
+      "loss": 0.6504,
+      "step": 2699
+    },
+    {
+      "epoch": 0.7479224376731302,
+      "grad_norm": 0.6649656295776367,
+      "learning_rate": 4.8227206644658275e-06,
+      "loss": 0.7189,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7481994459833795,
+      "grad_norm": 0.6813759803771973,
+      "learning_rate": 4.822585906110311e-06,
+      "loss": 0.6507,
+      "step": 2701
+    },
+    {
+      "epoch": 0.7484764542936289,
+      "grad_norm": 0.6968197822570801,
+      "learning_rate": 4.822451098440366e-06,
+      "loss": 0.6694,
+      "step": 2702
+    },
+    {
+      "epoch": 0.7487534626038781,
+      "grad_norm": 0.6730479598045349,
+      "learning_rate": 4.822316241458852e-06,
+      "loss": 0.6031,
+      "step": 2703
+    },
+    {
+      "epoch": 0.7490304709141274,
+      "grad_norm": 0.6745450496673584,
+      "learning_rate": 4.822181335168634e-06,
+      "loss": 0.6585,
+      "step": 2704
+    },
+    {
+      "epoch": 0.7493074792243767,
+      "grad_norm": 0.7031593322753906,
+      "learning_rate": 4.822046379572576e-06,
+      "loss": 0.7156,
+      "step": 2705
+    },
+    {
+      "epoch": 0.749584487534626,
+      "grad_norm": 0.6351406574249268,
+      "learning_rate": 4.821911374673543e-06,
+      "loss": 0.6556,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7498614958448754,
+      "grad_norm": 0.7087862491607666,
+      "learning_rate": 4.8217763204744035e-06,
+      "loss": 0.7392,
+      "step": 2707
+    },
+    {
+      "epoch": 0.7501385041551246,
+      "grad_norm": 0.7211248278617859,
+      "learning_rate": 4.8216412169780225e-06,
+      "loss": 0.6403,
+      "step": 2708
+    },
+    {
+      "epoch": 0.750415512465374,
+      "grad_norm": 0.6689242124557495,
+      "learning_rate": 4.82150606418727e-06,
+      "loss": 0.6803,
+      "step": 2709
+    },
+    {
+      "epoch": 0.7506925207756233,
+      "grad_norm": 0.6601364016532898,
+      "learning_rate": 4.821370862105015e-06,
+      "loss": 0.6967,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7509695290858726,
+      "grad_norm": 0.7165829539299011,
+      "learning_rate": 4.8212356107341284e-06,
+      "loss": 0.7043,
+      "step": 2711
+    },
+    {
+      "epoch": 0.7512465373961219,
+      "grad_norm": 0.6883214712142944,
+      "learning_rate": 4.821100310077482e-06,
+      "loss": 0.6122,
+      "step": 2712
+    },
+    {
+      "epoch": 0.7515235457063711,
+      "grad_norm": 0.6225855946540833,
+      "learning_rate": 4.8209649601379495e-06,
+      "loss": 0.6616,
+      "step": 2713
+    },
+    {
+      "epoch": 0.7518005540166205,
+      "grad_norm": 0.6740206480026245,
+      "learning_rate": 4.820829560918402e-06,
+      "loss": 0.6976,
+      "step": 2714
+    },
+    {
+      "epoch": 0.7520775623268698,
+      "grad_norm": 0.7862870693206787,
+      "learning_rate": 4.8206941124217175e-06,
+      "loss": 0.6914,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7523545706371191,
+      "grad_norm": 0.6579274535179138,
+      "learning_rate": 4.82055861465077e-06,
+      "loss": 0.6876,
+      "step": 2716
+    },
+    {
+      "epoch": 0.7526315789473684,
+      "grad_norm": 0.6604716181755066,
+      "learning_rate": 4.820423067608438e-06,
+      "loss": 0.6536,
+      "step": 2717
+    },
+    {
+      "epoch": 0.7529085872576178,
+      "grad_norm": 0.6885457634925842,
+      "learning_rate": 4.820287471297598e-06,
+      "loss": 0.6206,
+      "step": 2718
+    },
+    {
+      "epoch": 0.753185595567867,
+      "grad_norm": 0.6242653131484985,
+      "learning_rate": 4.820151825721129e-06,
+      "loss": 0.6144,
+      "step": 2719
+    },
+    {
+      "epoch": 0.7534626038781164,
+      "grad_norm": 0.6885089874267578,
+      "learning_rate": 4.820016130881912e-06,
+      "loss": 0.6358,
+      "step": 2720
+    },
+    {
+      "epoch": 0.7537396121883656,
+      "grad_norm": 0.7248243093490601,
+      "learning_rate": 4.819880386782829e-06,
+      "loss": 0.7197,
+      "step": 2721
+    },
+    {
+      "epoch": 0.754016620498615,
+      "grad_norm": 0.6652242541313171,
+      "learning_rate": 4.81974459342676e-06,
+      "loss": 0.6642,
+      "step": 2722
+    },
+    {
+      "epoch": 0.7542936288088643,
+      "grad_norm": 0.688584566116333,
+      "learning_rate": 4.819608750816589e-06,
+      "loss": 0.6648,
+      "step": 2723
+    },
+    {
+      "epoch": 0.7545706371191135,
+      "grad_norm": 0.6547428965568542,
+      "learning_rate": 4.8194728589552005e-06,
+      "loss": 0.6109,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7548476454293629,
+      "grad_norm": 0.6694716811180115,
+      "learning_rate": 4.819336917845481e-06,
+      "loss": 0.665,
+      "step": 2725
+    },
+    {
+      "epoch": 0.7551246537396122,
+      "grad_norm": 0.6976423263549805,
+      "learning_rate": 4.819200927490315e-06,
+      "loss": 0.735,
+      "step": 2726
+    },
+    {
+      "epoch": 0.7554016620498615,
+      "grad_norm": 0.6669568419456482,
+      "learning_rate": 4.819064887892592e-06,
+      "loss": 0.6577,
+      "step": 2727
+    },
+    {
+      "epoch": 0.7556786703601108,
+      "grad_norm": 0.6943166255950928,
+      "learning_rate": 4.8189287990551974e-06,
+      "loss": 0.6681,
+      "step": 2728
+    },
+    {
+      "epoch": 0.7559556786703601,
+      "grad_norm": 0.6642052531242371,
+      "learning_rate": 4.818792660981023e-06,
+      "loss": 0.68,
+      "step": 2729
+    },
+    {
+      "epoch": 0.7562326869806094,
+      "grad_norm": 0.7273021936416626,
+      "learning_rate": 4.818656473672959e-06,
+      "loss": 0.7097,
+      "step": 2730
+    },
+    {
+      "epoch": 0.7565096952908588,
+      "grad_norm": 0.6770200729370117,
+      "learning_rate": 4.818520237133897e-06,
+      "loss": 0.6669,
+      "step": 2731
+    },
+    {
+      "epoch": 0.756786703601108,
+      "grad_norm": 0.6642410755157471,
+      "learning_rate": 4.81838395136673e-06,
+      "loss": 0.667,
+      "step": 2732
+    },
+    {
+      "epoch": 0.7570637119113574,
+      "grad_norm": 0.7103769183158875,
+      "learning_rate": 4.818247616374351e-06,
+      "loss": 0.6894,
+      "step": 2733
+    },
+    {
+      "epoch": 0.7573407202216067,
+      "grad_norm": 0.652321994304657,
+      "learning_rate": 4.8181112321596544e-06,
+      "loss": 0.6889,
+      "step": 2734
+    },
+    {
+      "epoch": 0.7576177285318559,
+      "grad_norm": 0.684417724609375,
+      "learning_rate": 4.817974798725537e-06,
+      "loss": 0.686,
+      "step": 2735
+    },
+    {
+      "epoch": 0.7578947368421053,
+      "grad_norm": 0.6455621123313904,
+      "learning_rate": 4.817838316074894e-06,
+      "loss": 0.6586,
+      "step": 2736
+    },
+    {
+      "epoch": 0.7581717451523545,
+      "grad_norm": 0.7123239636421204,
+      "learning_rate": 4.817701784210626e-06,
+      "loss": 0.6715,
+      "step": 2737
+    },
+    {
+      "epoch": 0.7584487534626039,
+      "grad_norm": 0.6630301475524902,
+      "learning_rate": 4.817565203135629e-06,
+      "loss": 0.6998,
+      "step": 2738
+    },
+    {
+      "epoch": 0.7587257617728532,
+      "grad_norm": 0.6584340333938599,
+      "learning_rate": 4.817428572852806e-06,
+      "loss": 0.6578,
+      "step": 2739
+    },
+    {
+      "epoch": 0.7590027700831025,
+      "grad_norm": 0.6827883720397949,
+      "learning_rate": 4.817291893365055e-06,
+      "loss": 0.6981,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7592797783933518,
+      "grad_norm": 0.7373248338699341,
+      "learning_rate": 4.81715516467528e-06,
+      "loss": 0.6469,
+      "step": 2741
+    },
+    {
+      "epoch": 0.7595567867036012,
+      "grad_norm": 0.6962078213691711,
+      "learning_rate": 4.817018386786383e-06,
+      "loss": 0.6335,
+      "step": 2742
+    },
+    {
+      "epoch": 0.7598337950138504,
+      "grad_norm": 0.6791117787361145,
+      "learning_rate": 4.816881559701269e-06,
+      "loss": 0.6387,
+      "step": 2743
+    },
+    {
+      "epoch": 0.7601108033240997,
+      "grad_norm": 0.732049822807312,
+      "learning_rate": 4.816744683422843e-06,
+      "loss": 0.6952,
+      "step": 2744
+    },
+    {
+      "epoch": 0.760387811634349,
+      "grad_norm": 0.6718223690986633,
+      "learning_rate": 4.81660775795401e-06,
+      "loss": 0.6392,
+      "step": 2745
+    },
+    {
+      "epoch": 0.7606648199445983,
+      "grad_norm": 0.7064507007598877,
+      "learning_rate": 4.816470783297679e-06,
+      "loss": 0.7156,
+      "step": 2746
+    },
+    {
+      "epoch": 0.7609418282548477,
+      "grad_norm": 0.6774894595146179,
+      "learning_rate": 4.8163337594567576e-06,
+      "loss": 0.6582,
+      "step": 2747
+    },
+    {
+      "epoch": 0.7612188365650969,
+      "grad_norm": 0.6640307307243347,
+      "learning_rate": 4.816196686434156e-06,
+      "loss": 0.6511,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7614958448753463,
+      "grad_norm": 0.6646149754524231,
+      "learning_rate": 4.816059564232782e-06,
+      "loss": 0.6468,
+      "step": 2749
+    },
+    {
+      "epoch": 0.7617728531855956,
+      "grad_norm": 0.6548853516578674,
+      "learning_rate": 4.815922392855551e-06,
+      "loss": 0.6684,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7620498614958449,
+      "grad_norm": 0.6696813702583313,
+      "learning_rate": 4.815785172305372e-06,
+      "loss": 0.6903,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7623268698060942,
+      "grad_norm": 0.6660913825035095,
+      "learning_rate": 4.8156479025851604e-06,
+      "loss": 0.6537,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7626038781163434,
+      "grad_norm": 0.6767622828483582,
+      "learning_rate": 4.81551058369783e-06,
+      "loss": 0.6948,
+      "step": 2753
+    },
+    {
+      "epoch": 0.7628808864265928,
+      "grad_norm": 0.6668721437454224,
+      "learning_rate": 4.815373215646298e-06,
+      "loss": 0.675,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7631578947368421,
+      "grad_norm": 0.7125071883201599,
+      "learning_rate": 4.815235798433478e-06,
+      "loss": 0.6843,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7634349030470914,
+      "grad_norm": 0.6485822796821594,
+      "learning_rate": 4.81509833206229e-06,
+      "loss": 0.6778,
+      "step": 2756
+    },
+    {
+      "epoch": 0.7637119113573407,
+      "grad_norm": 0.6997203230857849,
+      "learning_rate": 4.814960816535653e-06,
+      "loss": 0.7199,
+      "step": 2757
+    },
+    {
+      "epoch": 0.7639889196675901,
+      "grad_norm": 0.6658564805984497,
+      "learning_rate": 4.814823251856486e-06,
+      "loss": 0.6775,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7642659279778393,
+      "grad_norm": 0.6849539875984192,
+      "learning_rate": 4.8146856380277094e-06,
+      "loss": 0.6719,
+      "step": 2759
+    },
+    {
+      "epoch": 0.7645429362880887,
+      "grad_norm": 0.6896138191223145,
+      "learning_rate": 4.814547975052245e-06,
+      "loss": 0.7024,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7648199445983379,
+      "grad_norm": 0.6258552670478821,
+      "learning_rate": 4.814410262933018e-06,
+      "loss": 0.6067,
+      "step": 2761
+    },
+    {
+      "epoch": 0.7650969529085873,
+      "grad_norm": 0.6760597229003906,
+      "learning_rate": 4.81427250167295e-06,
+      "loss": 0.6688,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7653739612188366,
+      "grad_norm": 0.651985228061676,
+      "learning_rate": 4.814134691274966e-06,
+      "loss": 0.6739,
+      "step": 2763
+    },
+    {
+      "epoch": 0.7656509695290858,
+      "grad_norm": 0.7685214281082153,
+      "learning_rate": 4.813996831741994e-06,
+      "loss": 0.6574,
+      "step": 2764
+    },
+    {
+      "epoch": 0.7659279778393352,
+      "grad_norm": 0.7026417851448059,
+      "learning_rate": 4.81385892307696e-06,
+      "loss": 0.6929,
+      "step": 2765
+    },
+    {
+      "epoch": 0.7662049861495844,
+      "grad_norm": 0.6741276383399963,
+      "learning_rate": 4.813720965282792e-06,
+      "loss": 0.6752,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7664819944598338,
+      "grad_norm": 0.6811288595199585,
+      "learning_rate": 4.8135829583624185e-06,
+      "loss": 0.6843,
+      "step": 2767
+    },
+    {
+      "epoch": 0.7667590027700831,
+      "grad_norm": 0.6758236289024353,
+      "learning_rate": 4.8134449023187715e-06,
+      "loss": 0.6943,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7670360110803324,
+      "grad_norm": 0.6365170478820801,
+      "learning_rate": 4.813306797154781e-06,
+      "loss": 0.619,
+      "step": 2769
+    },
+    {
+      "epoch": 0.7673130193905817,
+      "grad_norm": 0.6581631898880005,
+      "learning_rate": 4.813168642873379e-06,
+      "loss": 0.6761,
+      "step": 2770
+    },
+    {
+      "epoch": 0.7675900277008311,
+      "grad_norm": 0.6589594483375549,
+      "learning_rate": 4.813030439477501e-06,
+      "loss": 0.7126,
+      "step": 2771
+    },
+    {
+      "epoch": 0.7678670360110803,
+      "grad_norm": 0.6597155332565308,
+      "learning_rate": 4.812892186970078e-06,
+      "loss": 0.7007,
+      "step": 2772
+    },
+    {
+      "epoch": 0.7681440443213297,
+      "grad_norm": 0.677655816078186,
+      "learning_rate": 4.812753885354049e-06,
+      "loss": 0.6439,
+      "step": 2773
+    },
+    {
+      "epoch": 0.7684210526315789,
+      "grad_norm": 0.6718606948852539,
+      "learning_rate": 4.8126155346323475e-06,
+      "loss": 0.7329,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7686980609418282,
+      "grad_norm": 0.6654607057571411,
+      "learning_rate": 4.812477134807914e-06,
+      "loss": 0.6597,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7689750692520776,
+      "grad_norm": 0.6524559855461121,
+      "learning_rate": 4.812338685883684e-06,
+      "loss": 0.6147,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7692520775623268,
+      "grad_norm": 0.6646273732185364,
+      "learning_rate": 4.8122001878626e-06,
+      "loss": 0.669,
+      "step": 2777
+    },
+    {
+      "epoch": 0.7695290858725762,
+      "grad_norm": 0.6726834774017334,
+      "learning_rate": 4.8120616407476e-06,
+      "loss": 0.672,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7698060941828255,
+      "grad_norm": 0.652396559715271,
+      "learning_rate": 4.811923044541628e-06,
+      "loss": 0.6632,
+      "step": 2779
+    },
+    {
+      "epoch": 0.7700831024930748,
+      "grad_norm": 0.6736995577812195,
+      "learning_rate": 4.811784399247625e-06,
+      "loss": 0.629,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7703601108033241,
+      "grad_norm": 0.6741900444030762,
+      "learning_rate": 4.811645704868537e-06,
+      "loss": 0.6602,
+      "step": 2781
+    },
+    {
+      "epoch": 0.7706371191135734,
+      "grad_norm": 0.6603039503097534,
+      "learning_rate": 4.8115069614073065e-06,
+      "loss": 0.7292,
+      "step": 2782
+    },
+    {
+      "epoch": 0.7709141274238227,
+      "grad_norm": 0.6762208938598633,
+      "learning_rate": 4.81136816886688e-06,
+      "loss": 0.6798,
+      "step": 2783
+    },
+    {
+      "epoch": 0.771191135734072,
+      "grad_norm": 0.668865978717804,
+      "learning_rate": 4.811229327250204e-06,
+      "loss": 0.6508,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7714681440443213,
+      "grad_norm": 0.7083577513694763,
+      "learning_rate": 4.8110904365602285e-06,
+      "loss": 0.6987,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7717451523545706,
+      "grad_norm": 0.67978835105896,
+      "learning_rate": 4.8109514967999e-06,
+      "loss": 0.6525,
+      "step": 2786
+    },
+    {
+      "epoch": 0.77202216066482,
+      "grad_norm": 0.6726179718971252,
+      "learning_rate": 4.8108125079721705e-06,
+      "loss": 0.6151,
+      "step": 2787
+    },
+    {
+      "epoch": 0.7722991689750692,
+      "grad_norm": 0.6633230447769165,
+      "learning_rate": 4.8106734700799905e-06,
+      "loss": 0.7054,
+      "step": 2788
+    },
+    {
+      "epoch": 0.7725761772853186,
+      "grad_norm": 0.6689402461051941,
+      "learning_rate": 4.810534383126311e-06,
+      "loss": 0.6577,
+      "step": 2789
+    },
+    {
+      "epoch": 0.7728531855955678,
+      "grad_norm": 0.706071674823761,
+      "learning_rate": 4.810395247114087e-06,
+      "loss": 0.7239,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7731301939058172,
+      "grad_norm": 0.7108891606330872,
+      "learning_rate": 4.810256062046271e-06,
+      "loss": 0.6916,
+      "step": 2791
+    },
+    {
+      "epoch": 0.7734072022160665,
+      "grad_norm": 0.6585329174995422,
+      "learning_rate": 4.810116827925819e-06,
+      "loss": 0.6815,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7736842105263158,
+      "grad_norm": 0.6893103718757629,
+      "learning_rate": 4.8099775447556885e-06,
+      "loss": 0.7041,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7739612188365651,
+      "grad_norm": 0.604040801525116,
+      "learning_rate": 4.809838212538835e-06,
+      "loss": 0.5979,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7742382271468145,
+      "grad_norm": 0.7108220458030701,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 0.6336,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7745152354570637,
+      "grad_norm": 0.6532512903213501,
+      "learning_rate": 4.809559400976796e-06,
+      "loss": 0.6453,
+      "step": 2796
+    },
+    {
+      "epoch": 0.774792243767313,
+      "grad_norm": 0.6732306480407715,
+      "learning_rate": 4.809419921637529e-06,
+      "loss": 0.697,
+      "step": 2797
+    },
+    {
+      "epoch": 0.7750692520775623,
+      "grad_norm": 0.7318335175514221,
+      "learning_rate": 4.80928039326338e-06,
+      "loss": 0.6928,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7753462603878116,
+      "grad_norm": 0.6802420020103455,
+      "learning_rate": 4.8091408158573115e-06,
+      "loss": 0.6836,
+      "step": 2799
+    },
+    {
+      "epoch": 0.775623268698061,
+      "grad_norm": 0.7560245394706726,
+      "learning_rate": 4.8090011894222874e-06,
+      "loss": 0.7218,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7759002770083102,
+      "grad_norm": 0.6963402628898621,
+      "learning_rate": 4.80886151396127e-06,
+      "loss": 0.6777,
+      "step": 2801
+    },
+    {
+      "epoch": 0.7761772853185596,
+      "grad_norm": 0.6747308969497681,
+      "learning_rate": 4.808721789477227e-06,
+      "loss": 0.6525,
+      "step": 2802
+    },
+    {
+      "epoch": 0.7764542936288089,
+      "grad_norm": 0.6558110117912292,
+      "learning_rate": 4.8085820159731245e-06,
+      "loss": 0.6419,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7767313019390581,
+      "grad_norm": 0.6836314797401428,
+      "learning_rate": 4.808442193451931e-06,
+      "loss": 0.6685,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7770083102493075,
+      "grad_norm": 0.7180214524269104,
+      "learning_rate": 4.808302321916614e-06,
+      "loss": 0.7183,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7772853185595567,
+      "grad_norm": 0.6658193469047546,
+      "learning_rate": 4.8081624013701435e-06,
+      "loss": 0.6348,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7775623268698061,
+      "grad_norm": 0.7152186036109924,
+      "learning_rate": 4.808022431815491e-06,
+      "loss": 0.7257,
+      "step": 2807
+    },
+    {
+      "epoch": 0.7778393351800554,
+      "grad_norm": 0.7137231230735779,
+      "learning_rate": 4.807882413255629e-06,
+      "loss": 0.7042,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7781163434903047,
+      "grad_norm": 0.6712000370025635,
+      "learning_rate": 4.807742345693529e-06,
+      "loss": 0.7017,
+      "step": 2809
+    },
+    {
+      "epoch": 0.778393351800554,
+      "grad_norm": 0.6861433386802673,
+      "learning_rate": 4.807602229132166e-06,
+      "loss": 0.6771,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7786703601108034,
+      "grad_norm": 0.6944916248321533,
+      "learning_rate": 4.807462063574515e-06,
+      "loss": 0.6572,
+      "step": 2811
+    },
+    {
+      "epoch": 0.7789473684210526,
+      "grad_norm": 0.6742745041847229,
+      "learning_rate": 4.8073218490235506e-06,
+      "loss": 0.6802,
+      "step": 2812
+    },
+    {
+      "epoch": 0.779224376731302,
+      "grad_norm": 0.672527551651001,
+      "learning_rate": 4.807181585482252e-06,
+      "loss": 0.6926,
+      "step": 2813
+    },
+    {
+      "epoch": 0.7795013850415512,
+      "grad_norm": 0.7015137672424316,
+      "learning_rate": 4.807041272953596e-06,
+      "loss": 0.7219,
+      "step": 2814
+    },
+    {
+      "epoch": 0.7797783933518005,
+      "grad_norm": 0.6598617434501648,
+      "learning_rate": 4.806900911440563e-06,
+      "loss": 0.695,
+      "step": 2815
+    },
+    {
+      "epoch": 0.7800554016620499,
+      "grad_norm": 0.6474493145942688,
+      "learning_rate": 4.806760500946132e-06,
+      "loss": 0.6592,
+      "step": 2816
+    },
+    {
+      "epoch": 0.7803324099722991,
+      "grad_norm": 0.673628032207489,
+      "learning_rate": 4.806620041473285e-06,
+      "loss": 0.6625,
+      "step": 2817
+    },
+    {
+      "epoch": 0.7806094182825485,
+      "grad_norm": 0.6895366311073303,
+      "learning_rate": 4.806479533025003e-06,
+      "loss": 0.7107,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7808864265927978,
+      "grad_norm": 0.6482427716255188,
+      "learning_rate": 4.806338975604271e-06,
+      "loss": 0.6109,
+      "step": 2819
+    },
+    {
+      "epoch": 0.7811634349030471,
+      "grad_norm": 0.6845487356185913,
+      "learning_rate": 4.806198369214073e-06,
+      "loss": 0.7063,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7814404432132964,
+      "grad_norm": 0.6725820302963257,
+      "learning_rate": 4.8060577138573945e-06,
+      "loss": 0.6446,
+      "step": 2821
+    },
+    {
+      "epoch": 0.7817174515235457,
+      "grad_norm": 0.6608311533927917,
+      "learning_rate": 4.805917009537222e-06,
+      "loss": 0.6662,
+      "step": 2822
+    },
+    {
+      "epoch": 0.781994459833795,
+      "grad_norm": 0.677912175655365,
+      "learning_rate": 4.805776256256543e-06,
+      "loss": 0.6848,
+      "step": 2823
+    },
+    {
+      "epoch": 0.7822714681440444,
+      "grad_norm": 0.6723567843437195,
+      "learning_rate": 4.805635454018344e-06,
+      "loss": 0.7034,
+      "step": 2824
+    },
+    {
+      "epoch": 0.7825484764542936,
+      "grad_norm": 0.7071090936660767,
+      "learning_rate": 4.805494602825617e-06,
+      "loss": 0.6575,
+      "step": 2825
+    },
+    {
+      "epoch": 0.782825484764543,
+      "grad_norm": 0.6822702288627625,
+      "learning_rate": 4.805353702681352e-06,
+      "loss": 0.6527,
+      "step": 2826
+    },
+    {
+      "epoch": 0.7831024930747923,
+      "grad_norm": 0.7353903651237488,
+      "learning_rate": 4.8052127535885415e-06,
+      "loss": 0.6578,
+      "step": 2827
+    },
+    {
+      "epoch": 0.7833795013850415,
+      "grad_norm": 0.6553813815116882,
+      "learning_rate": 4.805071755550177e-06,
+      "loss": 0.6559,
+      "step": 2828
+    },
+    {
+      "epoch": 0.7836565096952909,
+      "grad_norm": 0.6671435236930847,
+      "learning_rate": 4.804930708569252e-06,
+      "loss": 0.7025,
+      "step": 2829
+    },
+    {
+      "epoch": 0.7839335180055401,
+      "grad_norm": 0.6840083003044128,
+      "learning_rate": 4.804789612648762e-06,
+      "loss": 0.7181,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7842105263157895,
+      "grad_norm": 0.6222642660140991,
+      "learning_rate": 4.804648467791704e-06,
+      "loss": 0.5592,
+      "step": 2831
+    },
+    {
+      "epoch": 0.7844875346260388,
+      "grad_norm": 0.6630566120147705,
+      "learning_rate": 4.804507274001073e-06,
+      "loss": 0.68,
+      "step": 2832
+    },
+    {
+      "epoch": 0.7847645429362881,
+      "grad_norm": 0.7513610124588013,
+      "learning_rate": 4.804366031279866e-06,
+      "loss": 0.7193,
+      "step": 2833
+    },
+    {
+      "epoch": 0.7850415512465374,
+      "grad_norm": 0.6853703856468201,
+      "learning_rate": 4.804224739631085e-06,
+      "loss": 0.665,
+      "step": 2834
+    },
+    {
+      "epoch": 0.7853185595567868,
+      "grad_norm": 0.6761021614074707,
+      "learning_rate": 4.804083399057728e-06,
+      "loss": 0.6699,
+      "step": 2835
+    },
+    {
+      "epoch": 0.785595567867036,
+      "grad_norm": 0.6687069535255432,
+      "learning_rate": 4.803942009562797e-06,
+      "loss": 0.6945,
+      "step": 2836
+    },
+    {
+      "epoch": 0.7858725761772853,
+      "grad_norm": 0.6809954643249512,
+      "learning_rate": 4.803800571149292e-06,
+      "loss": 0.6902,
+      "step": 2837
+    },
+    {
+      "epoch": 0.7861495844875346,
+      "grad_norm": 0.7132784128189087,
+      "learning_rate": 4.8036590838202195e-06,
+      "loss": 0.7272,
+      "step": 2838
+    },
+    {
+      "epoch": 0.7864265927977839,
+      "grad_norm": 0.6614826321601868,
+      "learning_rate": 4.80351754757858e-06,
+      "loss": 0.6934,
+      "step": 2839
+    },
+    {
+      "epoch": 0.7867036011080333,
+      "grad_norm": 0.6766554713249207,
+      "learning_rate": 4.803375962427381e-06,
+      "loss": 0.6621,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7869806094182825,
+      "grad_norm": 0.6399201154708862,
+      "learning_rate": 4.803234328369629e-06,
+      "loss": 0.715,
+      "step": 2841
+    },
+    {
+      "epoch": 0.7872576177285319,
+      "grad_norm": 0.68129962682724,
+      "learning_rate": 4.80309264540833e-06,
+      "loss": 0.6761,
+      "step": 2842
+    },
+    {
+      "epoch": 0.7875346260387812,
+      "grad_norm": 0.6482844352722168,
+      "learning_rate": 4.802950913546493e-06,
+      "loss": 0.6969,
+      "step": 2843
+    },
+    {
+      "epoch": 0.7878116343490305,
+      "grad_norm": 0.6589528322219849,
+      "learning_rate": 4.802809132787125e-06,
+      "loss": 0.6856,
+      "step": 2844
+    },
+    {
+      "epoch": 0.7880886426592798,
+      "grad_norm": 0.6505930423736572,
+      "learning_rate": 4.80266730313324e-06,
+      "loss": 0.6447,
+      "step": 2845
+    },
+    {
+      "epoch": 0.788365650969529,
+      "grad_norm": 0.6826377511024475,
+      "learning_rate": 4.8025254245878486e-06,
+      "loss": 0.73,
+      "step": 2846
+    },
+    {
+      "epoch": 0.7886426592797784,
+      "grad_norm": 0.6710218787193298,
+      "learning_rate": 4.802383497153961e-06,
+      "loss": 0.6912,
+      "step": 2847
+    },
+    {
+      "epoch": 0.7889196675900277,
+      "grad_norm": 0.6851953268051147,
+      "learning_rate": 4.802241520834593e-06,
+      "loss": 0.6747,
+      "step": 2848
+    },
+    {
+      "epoch": 0.789196675900277,
+      "grad_norm": 0.6651461124420166,
+      "learning_rate": 4.802099495632757e-06,
+      "loss": 0.6338,
+      "step": 2849
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.6330905556678772,
+      "learning_rate": 4.8019574215514705e-06,
+      "loss": 0.6573,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7897506925207757,
+      "grad_norm": 0.7091717720031738,
+      "learning_rate": 4.80181529859375e-06,
+      "loss": 0.678,
+      "step": 2851
+    },
+    {
+      "epoch": 0.7900277008310249,
+      "grad_norm": 0.6825574040412903,
+      "learning_rate": 4.801673126762612e-06,
+      "loss": 0.6308,
+      "step": 2852
+    },
+    {
+      "epoch": 0.7903047091412743,
+      "grad_norm": 0.6608375310897827,
+      "learning_rate": 4.801530906061076e-06,
+      "loss": 0.6497,
+      "step": 2853
+    },
+    {
+      "epoch": 0.7905817174515235,
+      "grad_norm": 0.641620397567749,
+      "learning_rate": 4.801388636492161e-06,
+      "loss": 0.678,
+      "step": 2854
+    },
+    {
+      "epoch": 0.7908587257617729,
+      "grad_norm": 0.6613548994064331,
+      "learning_rate": 4.801246318058888e-06,
+      "loss": 0.668,
+      "step": 2855
+    },
+    {
+      "epoch": 0.7911357340720222,
+      "grad_norm": 0.6885948181152344,
+      "learning_rate": 4.8011039507642785e-06,
+      "loss": 0.6979,
+      "step": 2856
+    },
+    {
+      "epoch": 0.7914127423822714,
+      "grad_norm": 0.67684006690979,
+      "learning_rate": 4.800961534611357e-06,
+      "loss": 0.7128,
+      "step": 2857
+    },
+    {
+      "epoch": 0.7916897506925208,
+      "grad_norm": 0.6780455112457275,
+      "learning_rate": 4.800819069603145e-06,
+      "loss": 0.6661,
+      "step": 2858
+    },
+    {
+      "epoch": 0.7919667590027701,
+      "grad_norm": 0.6764765977859497,
+      "learning_rate": 4.800676555742669e-06,
+      "loss": 0.6854,
+      "step": 2859
+    },
+    {
+      "epoch": 0.7922437673130194,
+      "grad_norm": 0.6664876341819763,
+      "learning_rate": 4.800533993032955e-06,
+      "loss": 0.6581,
+      "step": 2860
+    },
+    {
+      "epoch": 0.7925207756232687,
+      "grad_norm": 0.6832924485206604,
+      "learning_rate": 4.800391381477028e-06,
+      "loss": 0.6748,
+      "step": 2861
+    },
+    {
+      "epoch": 0.792797783933518,
+      "grad_norm": 0.6635904312133789,
+      "learning_rate": 4.800248721077919e-06,
+      "loss": 0.6682,
+      "step": 2862
+    },
+    {
+      "epoch": 0.7930747922437673,
+      "grad_norm": 0.6542471647262573,
+      "learning_rate": 4.800106011838655e-06,
+      "loss": 0.6875,
+      "step": 2863
+    },
+    {
+      "epoch": 0.7933518005540167,
+      "grad_norm": 0.6659771203994751,
+      "learning_rate": 4.7999632537622655e-06,
+      "loss": 0.6572,
+      "step": 2864
+    },
+    {
+      "epoch": 0.7936288088642659,
+      "grad_norm": 0.6873894929885864,
+      "learning_rate": 4.799820446851784e-06,
+      "loss": 0.6411,
+      "step": 2865
+    },
+    {
+      "epoch": 0.7939058171745152,
+      "grad_norm": 0.6617361903190613,
+      "learning_rate": 4.79967759111024e-06,
+      "loss": 0.6834,
+      "step": 2866
+    },
+    {
+      "epoch": 0.7941828254847645,
+      "grad_norm": 0.6959488391876221,
+      "learning_rate": 4.799534686540669e-06,
+      "loss": 0.7283,
+      "step": 2867
+    },
+    {
+      "epoch": 0.7944598337950138,
+      "grad_norm": 0.7266706824302673,
+      "learning_rate": 4.799391733146104e-06,
+      "loss": 0.6708,
+      "step": 2868
+    },
+    {
+      "epoch": 0.7947368421052632,
+      "grad_norm": 0.6400945782661438,
+      "learning_rate": 4.799248730929581e-06,
+      "loss": 0.6214,
+      "step": 2869
+    },
+    {
+      "epoch": 0.7950138504155124,
+      "grad_norm": 0.6700130105018616,
+      "learning_rate": 4.799105679894135e-06,
+      "loss": 0.6563,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7952908587257618,
+      "grad_norm": 0.6552063226699829,
+      "learning_rate": 4.798962580042804e-06,
+      "loss": 0.6555,
+      "step": 2871
+    },
+    {
+      "epoch": 0.7955678670360111,
+      "grad_norm": 0.6592832803726196,
+      "learning_rate": 4.7988194313786275e-06,
+      "loss": 0.6347,
+      "step": 2872
+    },
+    {
+      "epoch": 0.7958448753462604,
+      "grad_norm": 0.7056299448013306,
+      "learning_rate": 4.798676233904643e-06,
+      "loss": 0.7304,
+      "step": 2873
+    },
+    {
+      "epoch": 0.7961218836565097,
+      "grad_norm": 0.6806883811950684,
+      "learning_rate": 4.798532987623892e-06,
+      "loss": 0.7033,
+      "step": 2874
+    },
+    {
+      "epoch": 0.796398891966759,
+      "grad_norm": 0.6627018451690674,
+      "learning_rate": 4.798389692539416e-06,
+      "loss": 0.7396,
+      "step": 2875
+    },
+    {
+      "epoch": 0.7966759002770083,
+      "grad_norm": 0.6570676565170288,
+      "learning_rate": 4.798246348654258e-06,
+      "loss": 0.6506,
+      "step": 2876
+    },
+    {
+      "epoch": 0.7969529085872576,
+      "grad_norm": 0.6514227390289307,
+      "learning_rate": 4.79810295597146e-06,
+      "loss": 0.6356,
+      "step": 2877
+    },
+    {
+      "epoch": 0.7972299168975069,
+      "grad_norm": 0.6728456020355225,
+      "learning_rate": 4.7979595144940675e-06,
+      "loss": 0.6399,
+      "step": 2878
+    },
+    {
+      "epoch": 0.7975069252077562,
+      "grad_norm": 0.6666335463523865,
+      "learning_rate": 4.797816024225127e-06,
+      "loss": 0.6662,
+      "step": 2879
+    },
+    {
+      "epoch": 0.7977839335180056,
+      "grad_norm": 0.6706849932670593,
+      "learning_rate": 4.797672485167684e-06,
+      "loss": 0.6629,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7980609418282548,
+      "grad_norm": 0.6484897136688232,
+      "learning_rate": 4.797528897324786e-06,
+      "loss": 0.6704,
+      "step": 2881
+    },
+    {
+      "epoch": 0.7983379501385042,
+      "grad_norm": 0.6639708876609802,
+      "learning_rate": 4.797385260699483e-06,
+      "loss": 0.708,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7986149584487534,
+      "grad_norm": 0.7080711722373962,
+      "learning_rate": 4.7972415752948235e-06,
+      "loss": 0.7269,
+      "step": 2883
+    },
+    {
+      "epoch": 0.7988919667590028,
+      "grad_norm": 0.6324862837791443,
+      "learning_rate": 4.797097841113859e-06,
+      "loss": 0.6438,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7991689750692521,
+      "grad_norm": 0.6741070747375488,
+      "learning_rate": 4.796954058159641e-06,
+      "loss": 0.6608,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7994459833795013,
+      "grad_norm": 0.65198814868927,
+      "learning_rate": 4.796810226435223e-06,
+      "loss": 0.6623,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7997229916897507,
+      "grad_norm": 0.6975104212760925,
+      "learning_rate": 4.796666345943658e-06,
+      "loss": 0.7031,
+      "step": 2887
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7220117449760437,
+      "learning_rate": 4.796522416688002e-06,
+      "loss": 0.7107,
+      "step": 2888
+    },
+    {
+      "epoch": 0.8002770083102493,
+      "grad_norm": 0.7077227234840393,
+      "learning_rate": 4.7963784386713095e-06,
+      "loss": 0.6777,
+      "step": 2889
+    },
+    {
+      "epoch": 0.8005540166204986,
+      "grad_norm": 0.6867684125900269,
+      "learning_rate": 4.79623441189664e-06,
+      "loss": 0.6972,
+      "step": 2890
+    },
+    {
+      "epoch": 0.8008310249307479,
+      "grad_norm": 0.6424522399902344,
+      "learning_rate": 4.796090336367048e-06,
+      "loss": 0.6423,
+      "step": 2891
+    },
+    {
+      "epoch": 0.8011080332409972,
+      "grad_norm": 0.7177652716636658,
+      "learning_rate": 4.795946212085596e-06,
+      "loss": 0.6744,
+      "step": 2892
+    },
+    {
+      "epoch": 0.8013850415512466,
+      "grad_norm": 0.6668875813484192,
+      "learning_rate": 4.795802039055343e-06,
+      "loss": 0.6788,
+      "step": 2893
+    },
+    {
+      "epoch": 0.8016620498614958,
+      "grad_norm": 0.6542249321937561,
+      "learning_rate": 4.795657817279349e-06,
+      "loss": 0.7118,
+      "step": 2894
+    },
+    {
+      "epoch": 0.8019390581717452,
+      "grad_norm": 0.6446555256843567,
+      "learning_rate": 4.795513546760677e-06,
+      "loss": 0.6142,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8022160664819945,
+      "grad_norm": 0.6913520693778992,
+      "learning_rate": 4.7953692275023915e-06,
+      "loss": 0.6781,
+      "step": 2896
+    },
+    {
+      "epoch": 0.8024930747922437,
+      "grad_norm": 0.6808624863624573,
+      "learning_rate": 4.795224859507553e-06,
+      "loss": 0.7138,
+      "step": 2897
+    },
+    {
+      "epoch": 0.8027700831024931,
+      "grad_norm": 0.6847794651985168,
+      "learning_rate": 4.795080442779232e-06,
+      "loss": 0.7004,
+      "step": 2898
+    },
+    {
+      "epoch": 0.8030470914127423,
+      "grad_norm": 0.7110046148300171,
+      "learning_rate": 4.794935977320491e-06,
+      "loss": 0.6779,
+      "step": 2899
+    },
+    {
+      "epoch": 0.8033240997229917,
+      "grad_norm": 0.6769573092460632,
+      "learning_rate": 4.794791463134399e-06,
+      "loss": 0.6365,
+      "step": 2900
+    },
+    {
+      "epoch": 0.803601108033241,
+      "grad_norm": 0.6371225118637085,
+      "learning_rate": 4.794646900224023e-06,
+      "loss": 0.6444,
+      "step": 2901
+    },
+    {
+      "epoch": 0.8038781163434903,
+      "grad_norm": 0.7346711754798889,
+      "learning_rate": 4.794502288592435e-06,
+      "loss": 0.6966,
+      "step": 2902
+    },
+    {
+      "epoch": 0.8041551246537396,
+      "grad_norm": 0.6535019278526306,
+      "learning_rate": 4.794357628242703e-06,
+      "loss": 0.6423,
+      "step": 2903
+    },
+    {
+      "epoch": 0.804432132963989,
+      "grad_norm": 0.6877285242080688,
+      "learning_rate": 4.7942129191778995e-06,
+      "loss": 0.6663,
+      "step": 2904
+    },
+    {
+      "epoch": 0.8047091412742382,
+      "grad_norm": 0.6775194406509399,
+      "learning_rate": 4.794068161401097e-06,
+      "loss": 0.7229,
+      "step": 2905
+    },
+    {
+      "epoch": 0.8049861495844876,
+      "grad_norm": 0.685641884803772,
+      "learning_rate": 4.793923354915369e-06,
+      "loss": 0.6879,
+      "step": 2906
+    },
+    {
+      "epoch": 0.8052631578947368,
+      "grad_norm": 0.6785410642623901,
+      "learning_rate": 4.7937784997237894e-06,
+      "loss": 0.7212,
+      "step": 2907
+    },
+    {
+      "epoch": 0.8055401662049861,
+      "grad_norm": 0.6728705167770386,
+      "learning_rate": 4.793633595829435e-06,
+      "loss": 0.6853,
+      "step": 2908
+    },
+    {
+      "epoch": 0.8058171745152355,
+      "grad_norm": 0.6703183054924011,
+      "learning_rate": 4.793488643235383e-06,
+      "loss": 0.6484,
+      "step": 2909
+    },
+    {
+      "epoch": 0.8060941828254847,
+      "grad_norm": 0.6292587518692017,
+      "learning_rate": 4.793343641944709e-06,
+      "loss": 0.6437,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8063711911357341,
+      "grad_norm": 0.6794230937957764,
+      "learning_rate": 4.793198591960494e-06,
+      "loss": 0.6689,
+      "step": 2911
+    },
+    {
+      "epoch": 0.8066481994459834,
+      "grad_norm": 0.7074090838432312,
+      "learning_rate": 4.793053493285816e-06,
+      "loss": 0.691,
+      "step": 2912
+    },
+    {
+      "epoch": 0.8069252077562327,
+      "grad_norm": 0.6919972896575928,
+      "learning_rate": 4.792908345923757e-06,
+      "loss": 0.6969,
+      "step": 2913
+    },
+    {
+      "epoch": 0.807202216066482,
+      "grad_norm": 0.6462123990058899,
+      "learning_rate": 4.7927631498773975e-06,
+      "loss": 0.6998,
+      "step": 2914
+    },
+    {
+      "epoch": 0.8074792243767313,
+      "grad_norm": 0.7100611925125122,
+      "learning_rate": 4.792617905149822e-06,
+      "loss": 0.7316,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8077562326869806,
+      "grad_norm": 0.6892986297607422,
+      "learning_rate": 4.7924726117441135e-06,
+      "loss": 0.6959,
+      "step": 2916
+    },
+    {
+      "epoch": 0.80803324099723,
+      "grad_norm": 0.7016753554344177,
+      "learning_rate": 4.792327269663357e-06,
+      "loss": 0.764,
+      "step": 2917
+    },
+    {
+      "epoch": 0.8083102493074792,
+      "grad_norm": 0.6642240285873413,
+      "learning_rate": 4.7921818789106385e-06,
+      "loss": 0.7016,
+      "step": 2918
+    },
+    {
+      "epoch": 0.8085872576177285,
+      "grad_norm": 0.6750599145889282,
+      "learning_rate": 4.792036439489045e-06,
+      "loss": 0.6751,
+      "step": 2919
+    },
+    {
+      "epoch": 0.8088642659279779,
+      "grad_norm": 0.720486044883728,
+      "learning_rate": 4.791890951401664e-06,
+      "loss": 0.6709,
+      "step": 2920
+    },
+    {
+      "epoch": 0.8091412742382271,
+      "grad_norm": 0.6749086380004883,
+      "learning_rate": 4.791745414651588e-06,
+      "loss": 0.6387,
+      "step": 2921
+    },
+    {
+      "epoch": 0.8094182825484765,
+      "grad_norm": 0.6732552647590637,
+      "learning_rate": 4.791599829241902e-06,
+      "loss": 0.673,
+      "step": 2922
+    },
+    {
+      "epoch": 0.8096952908587257,
+      "grad_norm": 0.7399674654006958,
+      "learning_rate": 4.7914541951757e-06,
+      "loss": 0.6923,
+      "step": 2923
+    },
+    {
+      "epoch": 0.8099722991689751,
+      "grad_norm": 0.6551409959793091,
+      "learning_rate": 4.791308512456074e-06,
+      "loss": 0.651,
+      "step": 2924
+    },
+    {
+      "epoch": 0.8102493074792244,
+      "grad_norm": 0.69414883852005,
+      "learning_rate": 4.7911627810861165e-06,
+      "loss": 0.6788,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8105263157894737,
+      "grad_norm": 0.7157154083251953,
+      "learning_rate": 4.791017001068922e-06,
+      "loss": 0.6912,
+      "step": 2926
+    },
+    {
+      "epoch": 0.810803324099723,
+      "grad_norm": 0.6880971193313599,
+      "learning_rate": 4.7908711724075865e-06,
+      "loss": 0.6905,
+      "step": 2927
+    },
+    {
+      "epoch": 0.8110803324099723,
+      "grad_norm": 0.6524692177772522,
+      "learning_rate": 4.790725295105205e-06,
+      "loss": 0.6504,
+      "step": 2928
+    },
+    {
+      "epoch": 0.8113573407202216,
+      "grad_norm": 0.6358970403671265,
+      "learning_rate": 4.790579369164876e-06,
+      "loss": 0.6964,
+      "step": 2929
+    },
+    {
+      "epoch": 0.8116343490304709,
+      "grad_norm": 0.6947261691093445,
+      "learning_rate": 4.790433394589698e-06,
+      "loss": 0.6553,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8119113573407202,
+      "grad_norm": 0.681126594543457,
+      "learning_rate": 4.790287371382769e-06,
+      "loss": 0.7046,
+      "step": 2931
+    },
+    {
+      "epoch": 0.8121883656509695,
+      "grad_norm": 0.6445147395133972,
+      "learning_rate": 4.790141299547191e-06,
+      "loss": 0.6449,
+      "step": 2932
+    },
+    {
+      "epoch": 0.8124653739612189,
+      "grad_norm": 0.6821317076683044,
+      "learning_rate": 4.789995179086065e-06,
+      "loss": 0.7093,
+      "step": 2933
+    },
+    {
+      "epoch": 0.8127423822714681,
+      "grad_norm": 0.675289511680603,
+      "learning_rate": 4.789849010002493e-06,
+      "loss": 0.6994,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8130193905817175,
+      "grad_norm": 0.7069120407104492,
+      "learning_rate": 4.789702792299579e-06,
+      "loss": 0.6969,
+      "step": 2935
+    },
+    {
+      "epoch": 0.8132963988919668,
+      "grad_norm": 0.6931147575378418,
+      "learning_rate": 4.7895565259804275e-06,
+      "loss": 0.6342,
+      "step": 2936
+    },
+    {
+      "epoch": 0.813573407202216,
+      "grad_norm": 0.6551658511161804,
+      "learning_rate": 4.789410211048144e-06,
+      "loss": 0.6957,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8138504155124654,
+      "grad_norm": 0.6618546843528748,
+      "learning_rate": 4.789263847505835e-06,
+      "loss": 0.6414,
+      "step": 2938
+    },
+    {
+      "epoch": 0.8141274238227146,
+      "grad_norm": 0.6185067296028137,
+      "learning_rate": 4.789117435356608e-06,
+      "loss": 0.6516,
+      "step": 2939
+    },
+    {
+      "epoch": 0.814404432132964,
+      "grad_norm": 0.6497803330421448,
+      "learning_rate": 4.788970974603573e-06,
+      "loss": 0.6501,
+      "step": 2940
+    },
+    {
+      "epoch": 0.8146814404432133,
+      "grad_norm": 0.7077935934066772,
+      "learning_rate": 4.788824465249838e-06,
+      "loss": 0.6295,
+      "step": 2941
+    },
+    {
+      "epoch": 0.8149584487534626,
+      "grad_norm": 0.6624383926391602,
+      "learning_rate": 4.788677907298516e-06,
+      "loss": 0.6637,
+      "step": 2942
+    },
+    {
+      "epoch": 0.8152354570637119,
+      "grad_norm": 0.6660526394844055,
+      "learning_rate": 4.788531300752715e-06,
+      "loss": 0.7064,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8155124653739613,
+      "grad_norm": 0.6447203755378723,
+      "learning_rate": 4.788384645615552e-06,
+      "loss": 0.5935,
+      "step": 2944
+    },
+    {
+      "epoch": 0.8157894736842105,
+      "grad_norm": 0.6713807582855225,
+      "learning_rate": 4.788237941890139e-06,
+      "loss": 0.6769,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8160664819944599,
+      "grad_norm": 0.6947737336158752,
+      "learning_rate": 4.788091189579591e-06,
+      "loss": 0.6391,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8163434903047091,
+      "grad_norm": 0.686058521270752,
+      "learning_rate": 4.787944388687023e-06,
+      "loss": 0.7027,
+      "step": 2947
+    },
+    {
+      "epoch": 0.8166204986149584,
+      "grad_norm": 0.685586154460907,
+      "learning_rate": 4.7877975392155544e-06,
+      "loss": 0.7066,
+      "step": 2948
+    },
+    {
+      "epoch": 0.8168975069252078,
+      "grad_norm": 0.6798444390296936,
+      "learning_rate": 4.7876506411683e-06,
+      "loss": 0.7094,
+      "step": 2949
+    },
+    {
+      "epoch": 0.817174515235457,
+      "grad_norm": 0.67017662525177,
+      "learning_rate": 4.787503694548381e-06,
+      "loss": 0.6798,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8174515235457064,
+      "grad_norm": 0.6418706774711609,
+      "learning_rate": 4.787356699358917e-06,
+      "loss": 0.684,
+      "step": 2951
+    },
+    {
+      "epoch": 0.8177285318559557,
+      "grad_norm": 0.6584951877593994,
+      "learning_rate": 4.787209655603029e-06,
+      "loss": 0.6451,
+      "step": 2952
+    },
+    {
+      "epoch": 0.818005540166205,
+      "grad_norm": 0.6570129990577698,
+      "learning_rate": 4.78706256328384e-06,
+      "loss": 0.6856,
+      "step": 2953
+    },
+    {
+      "epoch": 0.8182825484764543,
+      "grad_norm": 0.67294842004776,
+      "learning_rate": 4.78691542240447e-06,
+      "loss": 0.6419,
+      "step": 2954
+    },
+    {
+      "epoch": 0.8185595567867036,
+      "grad_norm": 0.6495556831359863,
+      "learning_rate": 4.786768232968047e-06,
+      "loss": 0.6566,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8188365650969529,
+      "grad_norm": 0.6849834322929382,
+      "learning_rate": 4.786620994977695e-06,
+      "loss": 0.6939,
+      "step": 2956
+    },
+    {
+      "epoch": 0.8191135734072023,
+      "grad_norm": 0.6808969378471375,
+      "learning_rate": 4.78647370843654e-06,
+      "loss": 0.6968,
+      "step": 2957
+    },
+    {
+      "epoch": 0.8193905817174515,
+      "grad_norm": 0.672578752040863,
+      "learning_rate": 4.786326373347708e-06,
+      "loss": 0.6602,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8196675900277008,
+      "grad_norm": 0.6576619744300842,
+      "learning_rate": 4.786178989714328e-06,
+      "loss": 0.6875,
+      "step": 2959
+    },
+    {
+      "epoch": 0.8199445983379502,
+      "grad_norm": 0.7055615186691284,
+      "learning_rate": 4.786031557539532e-06,
+      "loss": 0.7054,
+      "step": 2960
+    },
+    {
+      "epoch": 0.8202216066481994,
+      "grad_norm": 0.6421630382537842,
+      "learning_rate": 4.785884076826446e-06,
+      "loss": 0.6487,
+      "step": 2961
+    },
+    {
+      "epoch": 0.8204986149584488,
+      "grad_norm": 0.6846792101860046,
+      "learning_rate": 4.7857365475782045e-06,
+      "loss": 0.6168,
+      "step": 2962
+    },
+    {
+      "epoch": 0.820775623268698,
+      "grad_norm": 0.6749417185783386,
+      "learning_rate": 4.7855889697979394e-06,
+      "loss": 0.6087,
+      "step": 2963
+    },
+    {
+      "epoch": 0.8210526315789474,
+      "grad_norm": 0.6619457006454468,
+      "learning_rate": 4.7854413434887834e-06,
+      "loss": 0.69,
+      "step": 2964
+    },
+    {
+      "epoch": 0.8213296398891967,
+      "grad_norm": 0.6408563852310181,
+      "learning_rate": 4.7852936686538705e-06,
+      "loss": 0.6095,
+      "step": 2965
+    },
+    {
+      "epoch": 0.821606648199446,
+      "grad_norm": 0.6566827297210693,
+      "learning_rate": 4.785145945296338e-06,
+      "loss": 0.6519,
+      "step": 2966
+    },
+    {
+      "epoch": 0.8218836565096953,
+      "grad_norm": 0.6920059323310852,
+      "learning_rate": 4.784998173419321e-06,
+      "loss": 0.666,
+      "step": 2967
+    },
+    {
+      "epoch": 0.8221606648199447,
+      "grad_norm": 0.6602988839149475,
+      "learning_rate": 4.784850353025958e-06,
+      "loss": 0.6775,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8224376731301939,
+      "grad_norm": 0.6310914754867554,
+      "learning_rate": 4.784702484119387e-06,
+      "loss": 0.6774,
+      "step": 2969
+    },
+    {
+      "epoch": 0.8227146814404432,
+      "grad_norm": 0.6583638787269592,
+      "learning_rate": 4.784554566702747e-06,
+      "loss": 0.6435,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8229916897506925,
+      "grad_norm": 0.6954817771911621,
+      "learning_rate": 4.78440660077918e-06,
+      "loss": 0.6828,
+      "step": 2971
+    },
+    {
+      "epoch": 0.8232686980609418,
+      "grad_norm": 0.675418496131897,
+      "learning_rate": 4.7842585863518266e-06,
+      "loss": 0.6384,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8235457063711912,
+      "grad_norm": 0.6548784375190735,
+      "learning_rate": 4.7841105234238305e-06,
+      "loss": 0.6704,
+      "step": 2973
+    },
+    {
+      "epoch": 0.8238227146814404,
+      "grad_norm": 0.6904942393302917,
+      "learning_rate": 4.7839624119983355e-06,
+      "loss": 0.6832,
+      "step": 2974
+    },
+    {
+      "epoch": 0.8240997229916898,
+      "grad_norm": 0.6817192435264587,
+      "learning_rate": 4.7838142520784855e-06,
+      "loss": 0.6846,
+      "step": 2975
+    },
+    {
+      "epoch": 0.824376731301939,
+      "grad_norm": 0.6640516519546509,
+      "learning_rate": 4.783666043667427e-06,
+      "loss": 0.6778,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8246537396121884,
+      "grad_norm": 0.6464129686355591,
+      "learning_rate": 4.7835177867683054e-06,
+      "loss": 0.6745,
+      "step": 2977
+    },
+    {
+      "epoch": 0.8249307479224377,
+      "grad_norm": 0.6703967452049255,
+      "learning_rate": 4.7833694813842704e-06,
+      "loss": 0.6806,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8252077562326869,
+      "grad_norm": 0.7072887420654297,
+      "learning_rate": 4.78322112751847e-06,
+      "loss": 0.6715,
+      "step": 2979
+    },
+    {
+      "epoch": 0.8254847645429363,
+      "grad_norm": 0.6501997709274292,
+      "learning_rate": 4.783072725174055e-06,
+      "loss": 0.6848,
+      "step": 2980
+    },
+    {
+      "epoch": 0.8257617728531856,
+      "grad_norm": 0.7048507928848267,
+      "learning_rate": 4.782924274354175e-06,
+      "loss": 0.6898,
+      "step": 2981
+    },
+    {
+      "epoch": 0.8260387811634349,
+      "grad_norm": 0.6500950455665588,
+      "learning_rate": 4.782775775061983e-06,
+      "loss": 0.6722,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8263157894736842,
+      "grad_norm": 0.6806106567382812,
+      "learning_rate": 4.782627227300633e-06,
+      "loss": 0.6942,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8265927977839335,
+      "grad_norm": 0.6816373467445374,
+      "learning_rate": 4.782478631073276e-06,
+      "loss": 0.7197,
+      "step": 2984
+    },
+    {
+      "epoch": 0.8268698060941828,
+      "grad_norm": 0.6448298096656799,
+      "learning_rate": 4.782329986383069e-06,
+      "loss": 0.676,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8271468144044322,
+      "grad_norm": 0.6851121187210083,
+      "learning_rate": 4.782181293233168e-06,
+      "loss": 0.6175,
+      "step": 2986
+    },
+    {
+      "epoch": 0.8274238227146814,
+      "grad_norm": 0.6796623468399048,
+      "learning_rate": 4.782032551626731e-06,
+      "loss": 0.6643,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8277008310249307,
+      "grad_norm": 0.7001053094863892,
+      "learning_rate": 4.781883761566915e-06,
+      "loss": 0.6902,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8279778393351801,
+      "grad_norm": 0.6801310181617737,
+      "learning_rate": 4.781734923056879e-06,
+      "loss": 0.6573,
+      "step": 2989
+    },
+    {
+      "epoch": 0.8282548476454293,
+      "grad_norm": 0.6727662086486816,
+      "learning_rate": 4.781586036099784e-06,
+      "loss": 0.6617,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8285318559556787,
+      "grad_norm": 0.6912431120872498,
+      "learning_rate": 4.78143710069879e-06,
+      "loss": 0.6612,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8288088642659279,
+      "grad_norm": 0.6520569324493408,
+      "learning_rate": 4.781288116857061e-06,
+      "loss": 0.6571,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8290858725761773,
+      "grad_norm": 0.7189939618110657,
+      "learning_rate": 4.781139084577759e-06,
+      "loss": 0.6676,
+      "step": 2993
+    },
+    {
+      "epoch": 0.8293628808864266,
+      "grad_norm": 0.6563591957092285,
+      "learning_rate": 4.78099000386405e-06,
+      "loss": 0.6675,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8296398891966759,
+      "grad_norm": 0.6994754076004028,
+      "learning_rate": 4.780840874719097e-06,
+      "loss": 0.6568,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8299168975069252,
+      "grad_norm": 0.6558755040168762,
+      "learning_rate": 4.780691697146068e-06,
+      "loss": 0.6082,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8301939058171746,
+      "grad_norm": 0.6601712703704834,
+      "learning_rate": 4.780542471148131e-06,
+      "loss": 0.668,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8304709141274238,
+      "grad_norm": 0.6975407004356384,
+      "learning_rate": 4.780393196728452e-06,
+      "loss": 0.6882,
+      "step": 2998
+    },
+    {
+      "epoch": 0.8307479224376731,
+      "grad_norm": 0.6977407932281494,
+      "learning_rate": 4.780243873890203e-06,
+      "loss": 0.6895,
+      "step": 2999
+    },
+    {
+      "epoch": 0.8310249307479224,
+      "grad_norm": 0.6513993740081787,
+      "learning_rate": 4.780094502636553e-06,
+      "loss": 0.6568,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8313019390581717,
+      "grad_norm": 0.7048574686050415,
+      "learning_rate": 4.779945082970673e-06,
+      "loss": 0.6626,
+      "step": 3001
+    },
+    {
+      "epoch": 0.8315789473684211,
+      "grad_norm": 0.6709084510803223,
+      "learning_rate": 4.779795614895738e-06,
+      "loss": 0.6928,
+      "step": 3002
+    },
+    {
+      "epoch": 0.8318559556786703,
+      "grad_norm": 0.7100402116775513,
+      "learning_rate": 4.779646098414919e-06,
+      "loss": 0.6443,
+      "step": 3003
+    },
+    {
+      "epoch": 0.8321329639889197,
+      "grad_norm": 0.6646271347999573,
+      "learning_rate": 4.779496533531393e-06,
+      "loss": 0.6267,
+      "step": 3004
+    },
+    {
+      "epoch": 0.832409972299169,
+      "grad_norm": 0.6824468970298767,
+      "learning_rate": 4.779346920248333e-06,
+      "loss": 0.6945,
+      "step": 3005
+    },
+    {
+      "epoch": 0.8326869806094183,
+      "grad_norm": 0.6574153900146484,
+      "learning_rate": 4.779197258568918e-06,
+      "loss": 0.6556,
+      "step": 3006
+    },
+    {
+      "epoch": 0.8329639889196676,
+      "grad_norm": 0.6881409883499146,
+      "learning_rate": 4.779047548496325e-06,
+      "loss": 0.6532,
+      "step": 3007
+    },
+    {
+      "epoch": 0.8332409972299168,
+      "grad_norm": 0.6868331432342529,
+      "learning_rate": 4.778897790033732e-06,
+      "loss": 0.6402,
+      "step": 3008
+    },
+    {
+      "epoch": 0.8335180055401662,
+      "grad_norm": 0.6438159942626953,
+      "learning_rate": 4.778747983184319e-06,
+      "loss": 0.6512,
+      "step": 3009
+    },
+    {
+      "epoch": 0.8337950138504155,
+      "grad_norm": 0.6847102642059326,
+      "learning_rate": 4.778598127951268e-06,
+      "loss": 0.6651,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8340720221606648,
+      "grad_norm": 0.6552219390869141,
+      "learning_rate": 4.778448224337759e-06,
+      "loss": 0.6574,
+      "step": 3011
+    },
+    {
+      "epoch": 0.8343490304709141,
+      "grad_norm": 0.6644777655601501,
+      "learning_rate": 4.778298272346975e-06,
+      "loss": 0.6665,
+      "step": 3012
+    },
+    {
+      "epoch": 0.8346260387811635,
+      "grad_norm": 0.6513178944587708,
+      "learning_rate": 4.778148271982102e-06,
+      "loss": 0.6535,
+      "step": 3013
+    },
+    {
+      "epoch": 0.8349030470914127,
+      "grad_norm": 0.6750456690788269,
+      "learning_rate": 4.777998223246323e-06,
+      "loss": 0.6848,
+      "step": 3014
+    },
+    {
+      "epoch": 0.8351800554016621,
+      "grad_norm": 0.7080809473991394,
+      "learning_rate": 4.777848126142824e-06,
+      "loss": 0.6663,
+      "step": 3015
+    },
+    {
+      "epoch": 0.8354570637119113,
+      "grad_norm": 0.725734293460846,
+      "learning_rate": 4.777697980674793e-06,
+      "loss": 0.7176,
+      "step": 3016
+    },
+    {
+      "epoch": 0.8357340720221607,
+      "grad_norm": 0.7126418352127075,
+      "learning_rate": 4.777547786845416e-06,
+      "loss": 0.7036,
+      "step": 3017
+    },
+    {
+      "epoch": 0.83601108033241,
+      "grad_norm": 0.6786972284317017,
+      "learning_rate": 4.777397544657884e-06,
+      "loss": 0.6533,
+      "step": 3018
+    },
+    {
+      "epoch": 0.8362880886426592,
+      "grad_norm": 0.6766000390052795,
+      "learning_rate": 4.7772472541153865e-06,
+      "loss": 0.6674,
+      "step": 3019
+    },
+    {
+      "epoch": 0.8365650969529086,
+      "grad_norm": 0.667149007320404,
+      "learning_rate": 4.777096915221114e-06,
+      "loss": 0.7067,
+      "step": 3020
+    },
+    {
+      "epoch": 0.8368421052631579,
+      "grad_norm": 0.7148330211639404,
+      "learning_rate": 4.776946527978259e-06,
+      "loss": 0.6662,
+      "step": 3021
+    },
+    {
+      "epoch": 0.8371191135734072,
+      "grad_norm": 0.6931552886962891,
+      "learning_rate": 4.776796092390015e-06,
+      "loss": 0.6841,
+      "step": 3022
+    },
+    {
+      "epoch": 0.8373961218836565,
+      "grad_norm": 0.7319605350494385,
+      "learning_rate": 4.776645608459575e-06,
+      "loss": 0.6989,
+      "step": 3023
+    },
+    {
+      "epoch": 0.8376731301939058,
+      "grad_norm": 0.6943248510360718,
+      "learning_rate": 4.7764950761901355e-06,
+      "loss": 0.6323,
+      "step": 3024
+    },
+    {
+      "epoch": 0.8379501385041551,
+      "grad_norm": 0.6594962477684021,
+      "learning_rate": 4.776344495584891e-06,
+      "loss": 0.6597,
+      "step": 3025
+    },
+    {
+      "epoch": 0.8382271468144045,
+      "grad_norm": 0.7413750290870667,
+      "learning_rate": 4.7761938666470405e-06,
+      "loss": 0.6685,
+      "step": 3026
+    },
+    {
+      "epoch": 0.8385041551246537,
+      "grad_norm": 0.6912171244621277,
+      "learning_rate": 4.776043189379781e-06,
+      "loss": 0.6683,
+      "step": 3027
+    },
+    {
+      "epoch": 0.838781163434903,
+      "grad_norm": 0.6637134552001953,
+      "learning_rate": 4.775892463786312e-06,
+      "loss": 0.721,
+      "step": 3028
+    },
+    {
+      "epoch": 0.8390581717451524,
+      "grad_norm": 0.6988323330879211,
+      "learning_rate": 4.775741689869835e-06,
+      "loss": 0.6792,
+      "step": 3029
+    },
+    {
+      "epoch": 0.8393351800554016,
+      "grad_norm": 0.6290362477302551,
+      "learning_rate": 4.77559086763355e-06,
+      "loss": 0.5723,
+      "step": 3030
+    },
+    {
+      "epoch": 0.839612188365651,
+      "grad_norm": 0.644509494304657,
+      "learning_rate": 4.775439997080659e-06,
+      "loss": 0.6326,
+      "step": 3031
+    },
+    {
+      "epoch": 0.8398891966759002,
+      "grad_norm": 0.6706430912017822,
+      "learning_rate": 4.775289078214366e-06,
+      "loss": 0.6629,
+      "step": 3032
+    },
+    {
+      "epoch": 0.8401662049861496,
+      "grad_norm": 0.7159662842750549,
+      "learning_rate": 4.775138111037876e-06,
+      "loss": 0.6711,
+      "step": 3033
+    },
+    {
+      "epoch": 0.8404432132963989,
+      "grad_norm": 0.669113039970398,
+      "learning_rate": 4.774987095554393e-06,
+      "loss": 0.6027,
+      "step": 3034
+    },
+    {
+      "epoch": 0.8407202216066482,
+      "grad_norm": 0.6876832246780396,
+      "learning_rate": 4.774836031767125e-06,
+      "loss": 0.6608,
+      "step": 3035
+    },
+    {
+      "epoch": 0.8409972299168975,
+      "grad_norm": 0.7057596445083618,
+      "learning_rate": 4.774684919679279e-06,
+      "loss": 0.7192,
+      "step": 3036
+    },
+    {
+      "epoch": 0.8412742382271469,
+      "grad_norm": 0.6566253900527954,
+      "learning_rate": 4.7745337592940635e-06,
+      "loss": 0.6393,
+      "step": 3037
+    },
+    {
+      "epoch": 0.8415512465373961,
+      "grad_norm": 0.6582949757575989,
+      "learning_rate": 4.774382550614687e-06,
+      "loss": 0.6249,
+      "step": 3038
+    },
+    {
+      "epoch": 0.8418282548476455,
+      "grad_norm": 0.6579154133796692,
+      "learning_rate": 4.77423129364436e-06,
+      "loss": 0.6777,
+      "step": 3039
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.6580916047096252,
+      "learning_rate": 4.7740799883862966e-06,
+      "loss": 0.6998,
+      "step": 3040
+    },
+    {
+      "epoch": 0.842382271468144,
+      "grad_norm": 0.6692470908164978,
+      "learning_rate": 4.773928634843706e-06,
+      "loss": 0.6524,
+      "step": 3041
+    },
+    {
+      "epoch": 0.8426592797783934,
+      "grad_norm": 0.6601621508598328,
+      "learning_rate": 4.773777233019805e-06,
+      "loss": 0.6935,
+      "step": 3042
+    },
+    {
+      "epoch": 0.8429362880886426,
+      "grad_norm": 0.6472424268722534,
+      "learning_rate": 4.773625782917806e-06,
+      "loss": 0.6437,
+      "step": 3043
+    },
+    {
+      "epoch": 0.843213296398892,
+      "grad_norm": 0.6353874206542969,
+      "learning_rate": 4.773474284540926e-06,
+      "loss": 0.7074,
+      "step": 3044
+    },
+    {
+      "epoch": 0.8434903047091413,
+      "grad_norm": 0.6795141100883484,
+      "learning_rate": 4.773322737892381e-06,
+      "loss": 0.6591,
+      "step": 3045
+    },
+    {
+      "epoch": 0.8437673130193906,
+      "grad_norm": 0.6520573496818542,
+      "learning_rate": 4.773171142975388e-06,
+      "loss": 0.6911,
+      "step": 3046
+    },
+    {
+      "epoch": 0.8440443213296399,
+      "grad_norm": 0.6691749095916748,
+      "learning_rate": 4.773019499793167e-06,
+      "loss": 0.6495,
+      "step": 3047
+    },
+    {
+      "epoch": 0.8443213296398892,
+      "grad_norm": 0.6517788171768188,
+      "learning_rate": 4.7728678083489375e-06,
+      "loss": 0.6762,
+      "step": 3048
+    },
+    {
+      "epoch": 0.8445983379501385,
+      "grad_norm": 0.6732432842254639,
+      "learning_rate": 4.7727160686459205e-06,
+      "loss": 0.6815,
+      "step": 3049
+    },
+    {
+      "epoch": 0.8448753462603878,
+      "grad_norm": 0.6788813471794128,
+      "learning_rate": 4.7725642806873375e-06,
+      "loss": 0.6832,
+      "step": 3050
+    },
+    {
+      "epoch": 0.8451523545706371,
+      "grad_norm": 0.6740108728408813,
+      "learning_rate": 4.772412444476411e-06,
+      "loss": 0.6829,
+      "step": 3051
+    },
+    {
+      "epoch": 0.8454293628808864,
+      "grad_norm": 0.7147025465965271,
+      "learning_rate": 4.772260560016365e-06,
+      "loss": 0.6787,
+      "step": 3052
+    },
+    {
+      "epoch": 0.8457063711911358,
+      "grad_norm": 0.6838698983192444,
+      "learning_rate": 4.7721086273104245e-06,
+      "loss": 0.6355,
+      "step": 3053
+    },
+    {
+      "epoch": 0.845983379501385,
+      "grad_norm": 0.7325090169906616,
+      "learning_rate": 4.771956646361816e-06,
+      "loss": 0.6596,
+      "step": 3054
+    },
+    {
+      "epoch": 0.8462603878116344,
+      "grad_norm": 0.6333363056182861,
+      "learning_rate": 4.771804617173765e-06,
+      "loss": 0.6532,
+      "step": 3055
+    },
+    {
+      "epoch": 0.8465373961218836,
+      "grad_norm": 0.6893544793128967,
+      "learning_rate": 4.771652539749502e-06,
+      "loss": 0.6605,
+      "step": 3056
+    },
+    {
+      "epoch": 0.846814404432133,
+      "grad_norm": 0.6349977850914001,
+      "learning_rate": 4.771500414092253e-06,
+      "loss": 0.6158,
+      "step": 3057
+    },
+    {
+      "epoch": 0.8470914127423823,
+      "grad_norm": 0.6788435578346252,
+      "learning_rate": 4.7713482402052505e-06,
+      "loss": 0.6811,
+      "step": 3058
+    },
+    {
+      "epoch": 0.8473684210526315,
+      "grad_norm": 0.7373189330101013,
+      "learning_rate": 4.771196018091724e-06,
+      "loss": 0.7169,
+      "step": 3059
+    },
+    {
+      "epoch": 0.8476454293628809,
+      "grad_norm": 0.6911075115203857,
+      "learning_rate": 4.7710437477549055e-06,
+      "loss": 0.6328,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8479224376731302,
+      "grad_norm": 0.7234703898429871,
+      "learning_rate": 4.770891429198029e-06,
+      "loss": 0.6492,
+      "step": 3061
+    },
+    {
+      "epoch": 0.8481994459833795,
+      "grad_norm": 0.7154641151428223,
+      "learning_rate": 4.770739062424328e-06,
+      "loss": 0.6541,
+      "step": 3062
+    },
+    {
+      "epoch": 0.8484764542936288,
+      "grad_norm": 0.7392740845680237,
+      "learning_rate": 4.7705866474370386e-06,
+      "loss": 0.6636,
+      "step": 3063
+    },
+    {
+      "epoch": 0.8487534626038781,
+      "grad_norm": 0.6674519181251526,
+      "learning_rate": 4.770434184239396e-06,
+      "loss": 0.6505,
+      "step": 3064
+    },
+    {
+      "epoch": 0.8490304709141274,
+      "grad_norm": 0.703151285648346,
+      "learning_rate": 4.770281672834638e-06,
+      "loss": 0.6472,
+      "step": 3065
+    },
+    {
+      "epoch": 0.8493074792243768,
+      "grad_norm": 0.6965990662574768,
+      "learning_rate": 4.770129113226001e-06,
+      "loss": 0.6578,
+      "step": 3066
+    },
+    {
+      "epoch": 0.849584487534626,
+      "grad_norm": 0.6909862756729126,
+      "learning_rate": 4.7699765054167265e-06,
+      "loss": 0.6242,
+      "step": 3067
+    },
+    {
+      "epoch": 0.8498614958448754,
+      "grad_norm": 0.6947312355041504,
+      "learning_rate": 4.769823849410054e-06,
+      "loss": 0.6758,
+      "step": 3068
+    },
+    {
+      "epoch": 0.8501385041551247,
+      "grad_norm": 0.6394481062889099,
+      "learning_rate": 4.769671145209225e-06,
+      "loss": 0.66,
+      "step": 3069
+    },
+    {
+      "epoch": 0.850415512465374,
+      "grad_norm": 0.6775333881378174,
+      "learning_rate": 4.7695183928174804e-06,
+      "loss": 0.7078,
+      "step": 3070
+    },
+    {
+      "epoch": 0.8506925207756233,
+      "grad_norm": 0.7309216260910034,
+      "learning_rate": 4.7693655922380655e-06,
+      "loss": 0.6954,
+      "step": 3071
+    },
+    {
+      "epoch": 0.8509695290858725,
+      "grad_norm": 0.7183328866958618,
+      "learning_rate": 4.7692127434742234e-06,
+      "loss": 0.6424,
+      "step": 3072
+    },
+    {
+      "epoch": 0.8512465373961219,
+      "grad_norm": 0.6703770160675049,
+      "learning_rate": 4.769059846529201e-06,
+      "loss": 0.6586,
+      "step": 3073
+    },
+    {
+      "epoch": 0.8515235457063712,
+      "grad_norm": 0.7268140912055969,
+      "learning_rate": 4.768906901406242e-06,
+      "loss": 0.6998,
+      "step": 3074
+    },
+    {
+      "epoch": 0.8518005540166205,
+      "grad_norm": 0.7222980856895447,
+      "learning_rate": 4.768753908108596e-06,
+      "loss": 0.7278,
+      "step": 3075
+    },
+    {
+      "epoch": 0.8520775623268698,
+      "grad_norm": 0.6823320984840393,
+      "learning_rate": 4.76860086663951e-06,
+      "loss": 0.638,
+      "step": 3076
+    },
+    {
+      "epoch": 0.8523545706371192,
+      "grad_norm": 0.6753017902374268,
+      "learning_rate": 4.768447777002234e-06,
+      "loss": 0.6778,
+      "step": 3077
+    },
+    {
+      "epoch": 0.8526315789473684,
+      "grad_norm": 0.6737832427024841,
+      "learning_rate": 4.76829463920002e-06,
+      "loss": 0.6892,
+      "step": 3078
+    },
+    {
+      "epoch": 0.8529085872576178,
+      "grad_norm": 0.6660850644111633,
+      "learning_rate": 4.768141453236118e-06,
+      "loss": 0.6461,
+      "step": 3079
+    },
+    {
+      "epoch": 0.853185595567867,
+      "grad_norm": 0.6556453704833984,
+      "learning_rate": 4.767988219113781e-06,
+      "loss": 0.6457,
+      "step": 3080
+    },
+    {
+      "epoch": 0.8534626038781163,
+      "grad_norm": 0.6840183734893799,
+      "learning_rate": 4.767834936836261e-06,
+      "loss": 0.6799,
+      "step": 3081
+    },
+    {
+      "epoch": 0.8537396121883657,
+      "grad_norm": 0.6308120489120483,
+      "learning_rate": 4.767681606406814e-06,
+      "loss": 0.617,
+      "step": 3082
+    },
+    {
+      "epoch": 0.8540166204986149,
+      "grad_norm": 0.676609218120575,
+      "learning_rate": 4.767528227828697e-06,
+      "loss": 0.6684,
+      "step": 3083
+    },
+    {
+      "epoch": 0.8542936288088643,
+      "grad_norm": 0.6895732283592224,
+      "learning_rate": 4.767374801105163e-06,
+      "loss": 0.68,
+      "step": 3084
+    },
+    {
+      "epoch": 0.8545706371191135,
+      "grad_norm": 0.6707661747932434,
+      "learning_rate": 4.767221326239474e-06,
+      "loss": 0.667,
+      "step": 3085
+    },
+    {
+      "epoch": 0.8548476454293629,
+      "grad_norm": 0.6718268990516663,
+      "learning_rate": 4.767067803234885e-06,
+      "loss": 0.6455,
+      "step": 3086
+    },
+    {
+      "epoch": 0.8551246537396122,
+      "grad_norm": 0.7019315361976624,
+      "learning_rate": 4.766914232094657e-06,
+      "loss": 0.6911,
+      "step": 3087
+    },
+    {
+      "epoch": 0.8554016620498615,
+      "grad_norm": 0.685129702091217,
+      "learning_rate": 4.766760612822051e-06,
+      "loss": 0.7099,
+      "step": 3088
+    },
+    {
+      "epoch": 0.8556786703601108,
+      "grad_norm": 0.6947982311248779,
+      "learning_rate": 4.766606945420329e-06,
+      "loss": 0.7263,
+      "step": 3089
+    },
+    {
+      "epoch": 0.8559556786703602,
+      "grad_norm": 0.6423403024673462,
+      "learning_rate": 4.766453229892753e-06,
+      "loss": 0.6686,
+      "step": 3090
+    },
+    {
+      "epoch": 0.8562326869806094,
+      "grad_norm": 0.6994284987449646,
+      "learning_rate": 4.766299466242587e-06,
+      "loss": 0.659,
+      "step": 3091
+    },
+    {
+      "epoch": 0.8565096952908587,
+      "grad_norm": 0.6681022047996521,
+      "learning_rate": 4.766145654473096e-06,
+      "loss": 0.6579,
+      "step": 3092
+    },
+    {
+      "epoch": 0.856786703601108,
+      "grad_norm": 0.6877515912055969,
+      "learning_rate": 4.7659917945875455e-06,
+      "loss": 0.6182,
+      "step": 3093
+    },
+    {
+      "epoch": 0.8570637119113573,
+      "grad_norm": 0.668996274471283,
+      "learning_rate": 4.765837886589202e-06,
+      "loss": 0.6618,
+      "step": 3094
+    },
+    {
+      "epoch": 0.8573407202216067,
+      "grad_norm": 0.6529152393341064,
+      "learning_rate": 4.765683930481335e-06,
+      "loss": 0.688,
+      "step": 3095
+    },
+    {
+      "epoch": 0.8576177285318559,
+      "grad_norm": 0.7194386124610901,
+      "learning_rate": 4.765529926267211e-06,
+      "loss": 0.645,
+      "step": 3096
+    },
+    {
+      "epoch": 0.8578947368421053,
+      "grad_norm": 0.6833193302154541,
+      "learning_rate": 4.765375873950102e-06,
+      "loss": 0.6941,
+      "step": 3097
+    },
+    {
+      "epoch": 0.8581717451523546,
+      "grad_norm": 0.6345197558403015,
+      "learning_rate": 4.765221773533277e-06,
+      "loss": 0.6294,
+      "step": 3098
+    },
+    {
+      "epoch": 0.8584487534626039,
+      "grad_norm": 0.7031524777412415,
+      "learning_rate": 4.76506762502001e-06,
+      "loss": 0.6644,
+      "step": 3099
+    },
+    {
+      "epoch": 0.8587257617728532,
+      "grad_norm": 0.646828293800354,
+      "learning_rate": 4.764913428413572e-06,
+      "loss": 0.6695,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8590027700831024,
+      "grad_norm": 0.6990915536880493,
+      "learning_rate": 4.764759183717239e-06,
+      "loss": 0.6898,
+      "step": 3101
+    },
+    {
+      "epoch": 0.8592797783933518,
+      "grad_norm": 0.7449160814285278,
+      "learning_rate": 4.764604890934285e-06,
+      "loss": 0.6352,
+      "step": 3102
+    },
+    {
+      "epoch": 0.8595567867036011,
+      "grad_norm": 0.6356722116470337,
+      "learning_rate": 4.764450550067986e-06,
+      "loss": 0.6182,
+      "step": 3103
+    },
+    {
+      "epoch": 0.8598337950138504,
+      "grad_norm": 0.6818626523017883,
+      "learning_rate": 4.764296161121618e-06,
+      "loss": 0.6556,
+      "step": 3104
+    },
+    {
+      "epoch": 0.8601108033240997,
+      "grad_norm": 0.7065897583961487,
+      "learning_rate": 4.764141724098461e-06,
+      "loss": 0.6528,
+      "step": 3105
+    },
+    {
+      "epoch": 0.8603878116343491,
+      "grad_norm": 0.7592453956604004,
+      "learning_rate": 4.763987239001793e-06,
+      "loss": 0.657,
+      "step": 3106
+    },
+    {
+      "epoch": 0.8606648199445983,
+      "grad_norm": 0.643108069896698,
+      "learning_rate": 4.7638327058348945e-06,
+      "loss": 0.682,
+      "step": 3107
+    },
+    {
+      "epoch": 0.8609418282548477,
+      "grad_norm": 0.6628758907318115,
+      "learning_rate": 4.763678124601046e-06,
+      "loss": 0.6405,
+      "step": 3108
+    },
+    {
+      "epoch": 0.8612188365650969,
+      "grad_norm": 0.7099124789237976,
+      "learning_rate": 4.7635234953035304e-06,
+      "loss": 0.6575,
+      "step": 3109
+    },
+    {
+      "epoch": 0.8614958448753463,
+      "grad_norm": 0.7039968371391296,
+      "learning_rate": 4.763368817945631e-06,
+      "loss": 0.7117,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8617728531855956,
+      "grad_norm": 0.6697438955307007,
+      "learning_rate": 4.76321409253063e-06,
+      "loss": 0.6456,
+      "step": 3111
+    },
+    {
+      "epoch": 0.8620498614958448,
+      "grad_norm": 0.6728848814964294,
+      "learning_rate": 4.763059319061816e-06,
+      "loss": 0.5569,
+      "step": 3112
+    },
+    {
+      "epoch": 0.8623268698060942,
+      "grad_norm": 0.6149548888206482,
+      "learning_rate": 4.762904497542472e-06,
+      "loss": 0.6469,
+      "step": 3113
+    },
+    {
+      "epoch": 0.8626038781163435,
+      "grad_norm": 0.6806933283805847,
+      "learning_rate": 4.762749627975888e-06,
+      "loss": 0.7133,
+      "step": 3114
+    },
+    {
+      "epoch": 0.8628808864265928,
+      "grad_norm": 0.6651535034179688,
+      "learning_rate": 4.76259471036535e-06,
+      "loss": 0.6436,
+      "step": 3115
+    },
+    {
+      "epoch": 0.8631578947368421,
+      "grad_norm": 0.6518577337265015,
+      "learning_rate": 4.762439744714149e-06,
+      "loss": 0.661,
+      "step": 3116
+    },
+    {
+      "epoch": 0.8634349030470914,
+      "grad_norm": 0.641085684299469,
+      "learning_rate": 4.762284731025574e-06,
+      "loss": 0.6133,
+      "step": 3117
+    },
+    {
+      "epoch": 0.8637119113573407,
+      "grad_norm": 0.6976808309555054,
+      "learning_rate": 4.762129669302917e-06,
+      "loss": 0.6968,
+      "step": 3118
+    },
+    {
+      "epoch": 0.8639889196675901,
+      "grad_norm": 0.6310849785804749,
+      "learning_rate": 4.761974559549471e-06,
+      "loss": 0.6388,
+      "step": 3119
+    },
+    {
+      "epoch": 0.8642659279778393,
+      "grad_norm": 0.6805459260940552,
+      "learning_rate": 4.7618194017685285e-06,
+      "loss": 0.713,
+      "step": 3120
+    },
+    {
+      "epoch": 0.8645429362880886,
+      "grad_norm": 0.6955195069313049,
+      "learning_rate": 4.761664195963383e-06,
+      "loss": 0.6308,
+      "step": 3121
+    },
+    {
+      "epoch": 0.864819944598338,
+      "grad_norm": 0.6788739562034607,
+      "learning_rate": 4.761508942137332e-06,
+      "loss": 0.6916,
+      "step": 3122
+    },
+    {
+      "epoch": 0.8650969529085872,
+      "grad_norm": 0.7025331854820251,
+      "learning_rate": 4.76135364029367e-06,
+      "loss": 0.679,
+      "step": 3123
+    },
+    {
+      "epoch": 0.8653739612188366,
+      "grad_norm": 0.6896618604660034,
+      "learning_rate": 4.761198290435696e-06,
+      "loss": 0.6763,
+      "step": 3124
+    },
+    {
+      "epoch": 0.8656509695290858,
+      "grad_norm": 0.6727938652038574,
+      "learning_rate": 4.761042892566707e-06,
+      "loss": 0.6373,
+      "step": 3125
+    },
+    {
+      "epoch": 0.8659279778393352,
+      "grad_norm": 0.6701960563659668,
+      "learning_rate": 4.760887446690005e-06,
+      "loss": 0.7016,
+      "step": 3126
+    },
+    {
+      "epoch": 0.8662049861495845,
+      "grad_norm": 0.6560091376304626,
+      "learning_rate": 4.7607319528088875e-06,
+      "loss": 0.6484,
+      "step": 3127
+    },
+    {
+      "epoch": 0.8664819944598338,
+      "grad_norm": 0.6654905080795288,
+      "learning_rate": 4.760576410926657e-06,
+      "loss": 0.6749,
+      "step": 3128
+    },
+    {
+      "epoch": 0.8667590027700831,
+      "grad_norm": 0.6756736040115356,
+      "learning_rate": 4.7604208210466175e-06,
+      "loss": 0.6182,
+      "step": 3129
+    },
+    {
+      "epoch": 0.8670360110803325,
+      "grad_norm": 0.6454859375953674,
+      "learning_rate": 4.760265183172072e-06,
+      "loss": 0.7207,
+      "step": 3130
+    },
+    {
+      "epoch": 0.8673130193905817,
+      "grad_norm": 0.6561083197593689,
+      "learning_rate": 4.760109497306323e-06,
+      "loss": 0.6288,
+      "step": 3131
+    },
+    {
+      "epoch": 0.867590027700831,
+      "grad_norm": 0.6179625391960144,
+      "learning_rate": 4.759953763452678e-06,
+      "loss": 0.6647,
+      "step": 3132
+    },
+    {
+      "epoch": 0.8678670360110803,
+      "grad_norm": 0.7050300240516663,
+      "learning_rate": 4.7597979816144435e-06,
+      "loss": 0.6844,
+      "step": 3133
+    },
+    {
+      "epoch": 0.8681440443213296,
+      "grad_norm": 0.6492171883583069,
+      "learning_rate": 4.759642151794928e-06,
+      "loss": 0.6574,
+      "step": 3134
+    },
+    {
+      "epoch": 0.868421052631579,
+      "grad_norm": 0.6972433924674988,
+      "learning_rate": 4.759486273997438e-06,
+      "loss": 0.6451,
+      "step": 3135
+    },
+    {
+      "epoch": 0.8686980609418282,
+      "grad_norm": 0.7063480615615845,
+      "learning_rate": 4.7593303482252835e-06,
+      "loss": 0.667,
+      "step": 3136
+    },
+    {
+      "epoch": 0.8689750692520776,
+      "grad_norm": 0.6627654433250427,
+      "learning_rate": 4.759174374481778e-06,
+      "loss": 0.6138,
+      "step": 3137
+    },
+    {
+      "epoch": 0.8692520775623269,
+      "grad_norm": 0.6340380311012268,
+      "learning_rate": 4.759018352770229e-06,
+      "loss": 0.6416,
+      "step": 3138
+    },
+    {
+      "epoch": 0.8695290858725762,
+      "grad_norm": 0.7147533297538757,
+      "learning_rate": 4.758862283093954e-06,
+      "loss": 0.6789,
+      "step": 3139
+    },
+    {
+      "epoch": 0.8698060941828255,
+      "grad_norm": 0.6682170629501343,
+      "learning_rate": 4.758706165456262e-06,
+      "loss": 0.661,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8700831024930747,
+      "grad_norm": 0.6603893637657166,
+      "learning_rate": 4.758549999860471e-06,
+      "loss": 0.6074,
+      "step": 3141
+    },
+    {
+      "epoch": 0.8703601108033241,
+      "grad_norm": 0.6531762480735779,
+      "learning_rate": 4.7583937863098965e-06,
+      "loss": 0.6717,
+      "step": 3142
+    },
+    {
+      "epoch": 0.8706371191135734,
+      "grad_norm": 0.7047556638717651,
+      "learning_rate": 4.758237524807853e-06,
+      "loss": 0.7041,
+      "step": 3143
+    },
+    {
+      "epoch": 0.8709141274238227,
+      "grad_norm": 0.6751700043678284,
+      "learning_rate": 4.758081215357661e-06,
+      "loss": 0.6753,
+      "step": 3144
+    },
+    {
+      "epoch": 0.871191135734072,
+      "grad_norm": 0.6937599778175354,
+      "learning_rate": 4.757924857962638e-06,
+      "loss": 0.697,
+      "step": 3145
+    },
+    {
+      "epoch": 0.8714681440443214,
+      "grad_norm": 0.6919699907302856,
+      "learning_rate": 4.757768452626104e-06,
+      "loss": 0.6793,
+      "step": 3146
+    },
+    {
+      "epoch": 0.8717451523545706,
+      "grad_norm": 0.625059962272644,
+      "learning_rate": 4.757611999351381e-06,
+      "loss": 0.6872,
+      "step": 3147
+    },
+    {
+      "epoch": 0.87202216066482,
+      "grad_norm": 0.660816490650177,
+      "learning_rate": 4.757455498141789e-06,
+      "loss": 0.6999,
+      "step": 3148
+    },
+    {
+      "epoch": 0.8722991689750692,
+      "grad_norm": 0.6542554497718811,
+      "learning_rate": 4.757298949000652e-06,
+      "loss": 0.6592,
+      "step": 3149
+    },
+    {
+      "epoch": 0.8725761772853186,
+      "grad_norm": 0.6849338412284851,
+      "learning_rate": 4.757142351931293e-06,
+      "loss": 0.6313,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8728531855955679,
+      "grad_norm": 0.685217559337616,
+      "learning_rate": 4.7569857069370395e-06,
+      "loss": 0.6299,
+      "step": 3151
+    },
+    {
+      "epoch": 0.8731301939058171,
+      "grad_norm": 0.6493719220161438,
+      "learning_rate": 4.756829014021215e-06,
+      "loss": 0.6399,
+      "step": 3152
+    },
+    {
+      "epoch": 0.8734072022160665,
+      "grad_norm": 0.6607489585876465,
+      "learning_rate": 4.756672273187146e-06,
+      "loss": 0.6659,
+      "step": 3153
+    },
+    {
+      "epoch": 0.8736842105263158,
+      "grad_norm": 0.6733990907669067,
+      "learning_rate": 4.756515484438163e-06,
+      "loss": 0.6533,
+      "step": 3154
+    },
+    {
+      "epoch": 0.8739612188365651,
+      "grad_norm": 0.6592486500740051,
+      "learning_rate": 4.7563586477775935e-06,
+      "loss": 0.6512,
+      "step": 3155
+    },
+    {
+      "epoch": 0.8742382271468144,
+      "grad_norm": 0.693362832069397,
+      "learning_rate": 4.756201763208768e-06,
+      "loss": 0.6779,
+      "step": 3156
+    },
+    {
+      "epoch": 0.8745152354570637,
+      "grad_norm": 0.6807980537414551,
+      "learning_rate": 4.756044830735017e-06,
+      "loss": 0.6771,
+      "step": 3157
+    },
+    {
+      "epoch": 0.874792243767313,
+      "grad_norm": 0.6866021156311035,
+      "learning_rate": 4.755887850359673e-06,
+      "loss": 0.6586,
+      "step": 3158
+    },
+    {
+      "epoch": 0.8750692520775624,
+      "grad_norm": 0.6628581881523132,
+      "learning_rate": 4.75573082208607e-06,
+      "loss": 0.6858,
+      "step": 3159
+    },
+    {
+      "epoch": 0.8753462603878116,
+      "grad_norm": 0.7030439972877502,
+      "learning_rate": 4.75557374591754e-06,
+      "loss": 0.6817,
+      "step": 3160
+    },
+    {
+      "epoch": 0.875623268698061,
+      "grad_norm": 0.6881664991378784,
+      "learning_rate": 4.7554166218574196e-06,
+      "loss": 0.6963,
+      "step": 3161
+    },
+    {
+      "epoch": 0.8759002770083103,
+      "grad_norm": 0.6781759262084961,
+      "learning_rate": 4.755259449909044e-06,
+      "loss": 0.6433,
+      "step": 3162
+    },
+    {
+      "epoch": 0.8761772853185595,
+      "grad_norm": 0.7240464687347412,
+      "learning_rate": 4.755102230075752e-06,
+      "loss": 0.6605,
+      "step": 3163
+    },
+    {
+      "epoch": 0.8764542936288089,
+      "grad_norm": 0.6497237086296082,
+      "learning_rate": 4.75494496236088e-06,
+      "loss": 0.6626,
+      "step": 3164
+    },
+    {
+      "epoch": 0.8767313019390581,
+      "grad_norm": 0.6879202723503113,
+      "learning_rate": 4.7547876467677685e-06,
+      "loss": 0.7042,
+      "step": 3165
+    },
+    {
+      "epoch": 0.8770083102493075,
+      "grad_norm": 0.6922905445098877,
+      "learning_rate": 4.754630283299757e-06,
+      "loss": 0.6756,
+      "step": 3166
+    },
+    {
+      "epoch": 0.8772853185595568,
+      "grad_norm": 0.6731038689613342,
+      "learning_rate": 4.754472871960186e-06,
+      "loss": 0.6272,
+      "step": 3167
+    },
+    {
+      "epoch": 0.8775623268698061,
+      "grad_norm": 0.6296544075012207,
+      "learning_rate": 4.7543154127524e-06,
+      "loss": 0.6471,
+      "step": 3168
+    },
+    {
+      "epoch": 0.8778393351800554,
+      "grad_norm": 0.7404965758323669,
+      "learning_rate": 4.754157905679741e-06,
+      "loss": 0.7242,
+      "step": 3169
+    },
+    {
+      "epoch": 0.8781163434903048,
+      "grad_norm": 0.6243695616722107,
+      "learning_rate": 4.754000350745551e-06,
+      "loss": 0.636,
+      "step": 3170
+    },
+    {
+      "epoch": 0.878393351800554,
+      "grad_norm": 0.6488210558891296,
+      "learning_rate": 4.753842747953179e-06,
+      "loss": 0.6139,
+      "step": 3171
+    },
+    {
+      "epoch": 0.8786703601108034,
+      "grad_norm": 0.6636707186698914,
+      "learning_rate": 4.75368509730597e-06,
+      "loss": 0.6445,
+      "step": 3172
+    },
+    {
+      "epoch": 0.8789473684210526,
+      "grad_norm": 0.6736623644828796,
+      "learning_rate": 4.7535273988072715e-06,
+      "loss": 0.6883,
+      "step": 3173
+    },
+    {
+      "epoch": 0.8792243767313019,
+      "grad_norm": 0.7248889803886414,
+      "learning_rate": 4.75336965246043e-06,
+      "loss": 0.6976,
+      "step": 3174
+    },
+    {
+      "epoch": 0.8795013850415513,
+      "grad_norm": 0.7072076201438904,
+      "learning_rate": 4.753211858268797e-06,
+      "loss": 0.6583,
+      "step": 3175
+    },
+    {
+      "epoch": 0.8797783933518005,
+      "grad_norm": 0.6631415486335754,
+      "learning_rate": 4.753054016235723e-06,
+      "loss": 0.6573,
+      "step": 3176
+    },
+    {
+      "epoch": 0.8800554016620499,
+      "grad_norm": 0.7018452286720276,
+      "learning_rate": 4.7528961263645575e-06,
+      "loss": 0.6408,
+      "step": 3177
+    },
+    {
+      "epoch": 0.8803324099722992,
+      "grad_norm": 0.6875770092010498,
+      "learning_rate": 4.752738188658654e-06,
+      "loss": 0.6701,
+      "step": 3178
+    },
+    {
+      "epoch": 0.8806094182825485,
+      "grad_norm": 0.6721932291984558,
+      "learning_rate": 4.752580203121367e-06,
+      "loss": 0.6711,
+      "step": 3179
+    },
+    {
+      "epoch": 0.8808864265927978,
+      "grad_norm": 0.7087852358818054,
+      "learning_rate": 4.752422169756048e-06,
+      "loss": 0.6526,
+      "step": 3180
+    },
+    {
+      "epoch": 0.881163434903047,
+      "grad_norm": 0.7229830026626587,
+      "learning_rate": 4.7522640885660556e-06,
+      "loss": 0.6688,
+      "step": 3181
+    },
+    {
+      "epoch": 0.8814404432132964,
+      "grad_norm": 0.6967926621437073,
+      "learning_rate": 4.752105959554745e-06,
+      "loss": 0.6511,
+      "step": 3182
+    },
+    {
+      "epoch": 0.8817174515235457,
+      "grad_norm": 0.7231464385986328,
+      "learning_rate": 4.751947782725472e-06,
+      "loss": 0.6702,
+      "step": 3183
+    },
+    {
+      "epoch": 0.881994459833795,
+      "grad_norm": 0.6495847105979919,
+      "learning_rate": 4.7517895580815985e-06,
+      "loss": 0.6626,
+      "step": 3184
+    },
+    {
+      "epoch": 0.8822714681440443,
+      "grad_norm": 0.6687087416648865,
+      "learning_rate": 4.751631285626482e-06,
+      "loss": 0.6823,
+      "step": 3185
+    },
+    {
+      "epoch": 0.8825484764542936,
+      "grad_norm": 0.6756120920181274,
+      "learning_rate": 4.751472965363482e-06,
+      "loss": 0.7317,
+      "step": 3186
+    },
+    {
+      "epoch": 0.8828254847645429,
+      "grad_norm": 0.6763913631439209,
+      "learning_rate": 4.751314597295963e-06,
+      "loss": 0.6534,
+      "step": 3187
+    },
+    {
+      "epoch": 0.8831024930747923,
+      "grad_norm": 0.6849490404129028,
+      "learning_rate": 4.751156181427285e-06,
+      "loss": 0.6579,
+      "step": 3188
+    },
+    {
+      "epoch": 0.8833795013850415,
+      "grad_norm": 0.6868329644203186,
+      "learning_rate": 4.750997717760813e-06,
+      "loss": 0.6889,
+      "step": 3189
+    },
+    {
+      "epoch": 0.8836565096952909,
+      "grad_norm": 0.6511489748954773,
+      "learning_rate": 4.75083920629991e-06,
+      "loss": 0.7108,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8839335180055402,
+      "grad_norm": 0.6739808917045593,
+      "learning_rate": 4.750680647047944e-06,
+      "loss": 0.6885,
+      "step": 3191
+    },
+    {
+      "epoch": 0.8842105263157894,
+      "grad_norm": 0.6702025532722473,
+      "learning_rate": 4.750522040008279e-06,
+      "loss": 0.6607,
+      "step": 3192
+    },
+    {
+      "epoch": 0.8844875346260388,
+      "grad_norm": 0.6719469428062439,
+      "learning_rate": 4.750363385184285e-06,
+      "loss": 0.6732,
+      "step": 3193
+    },
+    {
+      "epoch": 0.884764542936288,
+      "grad_norm": 0.7011950612068176,
+      "learning_rate": 4.750204682579328e-06,
+      "loss": 0.6364,
+      "step": 3194
+    },
+    {
+      "epoch": 0.8850415512465374,
+      "grad_norm": 0.6801823973655701,
+      "learning_rate": 4.750045932196781e-06,
+      "loss": 0.6367,
+      "step": 3195
+    },
+    {
+      "epoch": 0.8853185595567867,
+      "grad_norm": 0.718298614025116,
+      "learning_rate": 4.749887134040011e-06,
+      "loss": 0.7011,
+      "step": 3196
+    },
+    {
+      "epoch": 0.885595567867036,
+      "grad_norm": 0.7182875275611877,
+      "learning_rate": 4.7497282881123925e-06,
+      "loss": 0.6881,
+      "step": 3197
+    },
+    {
+      "epoch": 0.8858725761772853,
+      "grad_norm": 0.6788876056671143,
+      "learning_rate": 4.7495693944172975e-06,
+      "loss": 0.7216,
+      "step": 3198
+    },
+    {
+      "epoch": 0.8861495844875347,
+      "grad_norm": 0.6881414651870728,
+      "learning_rate": 4.749410452958099e-06,
+      "loss": 0.6589,
+      "step": 3199
+    },
+    {
+      "epoch": 0.8864265927977839,
+      "grad_norm": 0.639951229095459,
+      "learning_rate": 4.7492514637381725e-06,
+      "loss": 0.6664,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8867036011080333,
+      "grad_norm": 0.6903998851776123,
+      "learning_rate": 4.749092426760894e-06,
+      "loss": 0.6704,
+      "step": 3201
+    },
+    {
+      "epoch": 0.8869806094182825,
+      "grad_norm": 0.6713690161705017,
+      "learning_rate": 4.748933342029639e-06,
+      "loss": 0.6829,
+      "step": 3202
+    },
+    {
+      "epoch": 0.8872576177285318,
+      "grad_norm": 0.7007774114608765,
+      "learning_rate": 4.7487742095477865e-06,
+      "loss": 0.6929,
+      "step": 3203
+    },
+    {
+      "epoch": 0.8875346260387812,
+      "grad_norm": 0.6859935522079468,
+      "learning_rate": 4.748615029318714e-06,
+      "loss": 0.6402,
+      "step": 3204
+    },
+    {
+      "epoch": 0.8878116343490304,
+      "grad_norm": 0.7014520168304443,
+      "learning_rate": 4.7484558013458025e-06,
+      "loss": 0.6759,
+      "step": 3205
+    },
+    {
+      "epoch": 0.8880886426592798,
+      "grad_norm": 0.651506781578064,
+      "learning_rate": 4.748296525632432e-06,
+      "loss": 0.6491,
+      "step": 3206
+    },
+    {
+      "epoch": 0.8883656509695291,
+      "grad_norm": 0.6785458326339722,
+      "learning_rate": 4.748137202181985e-06,
+      "loss": 0.6821,
+      "step": 3207
+    },
+    {
+      "epoch": 0.8886426592797784,
+      "grad_norm": 0.6694476008415222,
+      "learning_rate": 4.747977830997845e-06,
+      "loss": 0.6845,
+      "step": 3208
+    },
+    {
+      "epoch": 0.8889196675900277,
+      "grad_norm": 0.7240982055664062,
+      "learning_rate": 4.747818412083395e-06,
+      "loss": 0.6531,
+      "step": 3209
+    },
+    {
+      "epoch": 0.889196675900277,
+      "grad_norm": 0.6883110404014587,
+      "learning_rate": 4.747658945442018e-06,
+      "loss": 0.6696,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8894736842105263,
+      "grad_norm": 0.6668145656585693,
+      "learning_rate": 4.747499431077103e-06,
+      "loss": 0.6814,
+      "step": 3211
+    },
+    {
+      "epoch": 0.8897506925207757,
+      "grad_norm": 0.6918444037437439,
+      "learning_rate": 4.7473398689920344e-06,
+      "loss": 0.686,
+      "step": 3212
+    },
+    {
+      "epoch": 0.8900277008310249,
+      "grad_norm": 0.6626413464546204,
+      "learning_rate": 4.747180259190203e-06,
+      "loss": 0.6781,
+      "step": 3213
+    },
+    {
+      "epoch": 0.8903047091412742,
+      "grad_norm": 0.6800508499145508,
+      "learning_rate": 4.747020601674996e-06,
+      "loss": 0.6627,
+      "step": 3214
+    },
+    {
+      "epoch": 0.8905817174515236,
+      "grad_norm": 0.638883113861084,
+      "learning_rate": 4.746860896449802e-06,
+      "loss": 0.6777,
+      "step": 3215
+    },
+    {
+      "epoch": 0.8908587257617728,
+      "grad_norm": 0.6268462538719177,
+      "learning_rate": 4.746701143518013e-06,
+      "loss": 0.567,
+      "step": 3216
+    },
+    {
+      "epoch": 0.8911357340720222,
+      "grad_norm": 0.6455804109573364,
+      "learning_rate": 4.746541342883023e-06,
+      "loss": 0.6246,
+      "step": 3217
+    },
+    {
+      "epoch": 0.8914127423822714,
+      "grad_norm": 0.667750358581543,
+      "learning_rate": 4.746381494548223e-06,
+      "loss": 0.6854,
+      "step": 3218
+    },
+    {
+      "epoch": 0.8916897506925208,
+      "grad_norm": 0.7158799171447754,
+      "learning_rate": 4.746221598517005e-06,
+      "loss": 0.6546,
+      "step": 3219
+    },
+    {
+      "epoch": 0.8919667590027701,
+      "grad_norm": 0.687025785446167,
+      "learning_rate": 4.746061654792769e-06,
+      "loss": 0.6289,
+      "step": 3220
+    },
+    {
+      "epoch": 0.8922437673130194,
+      "grad_norm": 0.7073338627815247,
+      "learning_rate": 4.745901663378907e-06,
+      "loss": 0.7161,
+      "step": 3221
+    },
+    {
+      "epoch": 0.8925207756232687,
+      "grad_norm": 0.6431207060813904,
+      "learning_rate": 4.745741624278817e-06,
+      "loss": 0.6723,
+      "step": 3222
+    },
+    {
+      "epoch": 0.892797783933518,
+      "grad_norm": 0.658841073513031,
+      "learning_rate": 4.745581537495898e-06,
+      "loss": 0.6587,
+      "step": 3223
+    },
+    {
+      "epoch": 0.8930747922437673,
+      "grad_norm": 0.6782752275466919,
+      "learning_rate": 4.745421403033548e-06,
+      "loss": 0.6692,
+      "step": 3224
+    },
+    {
+      "epoch": 0.8933518005540166,
+      "grad_norm": 0.6738010048866272,
+      "learning_rate": 4.745261220895168e-06,
+      "loss": 0.6891,
+      "step": 3225
+    },
+    {
+      "epoch": 0.8936288088642659,
+      "grad_norm": 0.6247903108596802,
+      "learning_rate": 4.745100991084157e-06,
+      "loss": 0.6206,
+      "step": 3226
+    },
+    {
+      "epoch": 0.8939058171745152,
+      "grad_norm": 0.6673998236656189,
+      "learning_rate": 4.744940713603921e-06,
+      "loss": 0.6054,
+      "step": 3227
+    },
+    {
+      "epoch": 0.8941828254847646,
+      "grad_norm": 0.7002072930335999,
+      "learning_rate": 4.744780388457859e-06,
+      "loss": 0.6975,
+      "step": 3228
+    },
+    {
+      "epoch": 0.8944598337950138,
+      "grad_norm": 0.6853911876678467,
+      "learning_rate": 4.7446200156493775e-06,
+      "loss": 0.6347,
+      "step": 3229
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.6441922187805176,
+      "learning_rate": 4.744459595181879e-06,
+      "loss": 0.7001,
+      "step": 3230
+    },
+    {
+      "epoch": 0.8950138504155125,
+      "grad_norm": 0.6561487317085266,
+      "learning_rate": 4.744299127058774e-06,
+      "loss": 0.6439,
+      "step": 3231
+    },
+    {
+      "epoch": 0.8952908587257618,
+      "grad_norm": 0.6780251860618591,
+      "learning_rate": 4.744138611283467e-06,
+      "loss": 0.7057,
+      "step": 3232
+    },
+    {
+      "epoch": 0.8955678670360111,
+      "grad_norm": 0.6251480579376221,
+      "learning_rate": 4.743978047859365e-06,
+      "loss": 0.6118,
+      "step": 3233
+    },
+    {
+      "epoch": 0.8958448753462603,
+      "grad_norm": 0.7261717915534973,
+      "learning_rate": 4.743817436789879e-06,
+      "loss": 0.7222,
+      "step": 3234
+    },
+    {
+      "epoch": 0.8961218836565097,
+      "grad_norm": 0.6909134387969971,
+      "learning_rate": 4.74365677807842e-06,
+      "loss": 0.6874,
+      "step": 3235
+    },
+    {
+      "epoch": 0.896398891966759,
+      "grad_norm": 0.621599018573761,
+      "learning_rate": 4.743496071728396e-06,
+      "loss": 0.632,
+      "step": 3236
+    },
+    {
+      "epoch": 0.8966759002770083,
+      "grad_norm": 0.6764978766441345,
+      "learning_rate": 4.743335317743223e-06,
+      "loss": 0.6776,
+      "step": 3237
+    },
+    {
+      "epoch": 0.8969529085872576,
+      "grad_norm": 0.6836737990379333,
+      "learning_rate": 4.743174516126311e-06,
+      "loss": 0.7023,
+      "step": 3238
+    },
+    {
+      "epoch": 0.897229916897507,
+      "grad_norm": 0.6584877967834473,
+      "learning_rate": 4.7430136668810764e-06,
+      "loss": 0.6377,
+      "step": 3239
+    },
+    {
+      "epoch": 0.8975069252077562,
+      "grad_norm": 0.6757174134254456,
+      "learning_rate": 4.742852770010933e-06,
+      "loss": 0.6419,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8977839335180056,
+      "grad_norm": 0.6961698532104492,
+      "learning_rate": 4.742691825519298e-06,
+      "loss": 0.676,
+      "step": 3241
+    },
+    {
+      "epoch": 0.8980609418282548,
+      "grad_norm": 0.6550590991973877,
+      "learning_rate": 4.742530833409587e-06,
+      "loss": 0.6471,
+      "step": 3242
+    },
+    {
+      "epoch": 0.8983379501385041,
+      "grad_norm": 0.7002878189086914,
+      "learning_rate": 4.742369793685222e-06,
+      "loss": 0.7044,
+      "step": 3243
+    },
+    {
+      "epoch": 0.8986149584487535,
+      "grad_norm": 0.7033624649047852,
+      "learning_rate": 4.742208706349619e-06,
+      "loss": 0.664,
+      "step": 3244
+    },
+    {
+      "epoch": 0.8988919667590027,
+      "grad_norm": 0.6921766996383667,
+      "learning_rate": 4.742047571406198e-06,
+      "loss": 0.6838,
+      "step": 3245
+    },
+    {
+      "epoch": 0.8991689750692521,
+      "grad_norm": 0.6530360579490662,
+      "learning_rate": 4.741886388858384e-06,
+      "loss": 0.6551,
+      "step": 3246
+    },
+    {
+      "epoch": 0.8994459833795014,
+      "grad_norm": 0.6938414573669434,
+      "learning_rate": 4.741725158709594e-06,
+      "loss": 0.6538,
+      "step": 3247
+    },
+    {
+      "epoch": 0.8997229916897507,
+      "grad_norm": 0.6908465623855591,
+      "learning_rate": 4.741563880963256e-06,
+      "loss": 0.684,
+      "step": 3248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6564576625823975,
+      "learning_rate": 4.741402555622791e-06,
+      "loss": 0.7037,
+      "step": 3249
+    },
+    {
+      "epoch": 0.9002770083102493,
+      "grad_norm": 0.7001242637634277,
+      "learning_rate": 4.741241182691627e-06,
+      "loss": 0.696,
+      "step": 3250
+    },
+    {
+      "epoch": 0.9005540166204986,
+      "grad_norm": 0.6578646898269653,
+      "learning_rate": 4.741079762173189e-06,
+      "loss": 0.6147,
+      "step": 3251
+    },
+    {
+      "epoch": 0.900831024930748,
+      "grad_norm": 0.5905128717422485,
+      "learning_rate": 4.7409182940709046e-06,
+      "loss": 0.6315,
+      "step": 3252
+    },
+    {
+      "epoch": 0.9011080332409972,
+      "grad_norm": 0.6931217908859253,
+      "learning_rate": 4.740756778388201e-06,
+      "loss": 0.6798,
+      "step": 3253
+    },
+    {
+      "epoch": 0.9013850415512465,
+      "grad_norm": 0.6865070462226868,
+      "learning_rate": 4.74059521512851e-06,
+      "loss": 0.6773,
+      "step": 3254
+    },
+    {
+      "epoch": 0.9016620498614959,
+      "grad_norm": 0.703082799911499,
+      "learning_rate": 4.74043360429526e-06,
+      "loss": 0.6764,
+      "step": 3255
+    },
+    {
+      "epoch": 0.9019390581717451,
+      "grad_norm": 0.7222578525543213,
+      "learning_rate": 4.7402719458918845e-06,
+      "loss": 0.7117,
+      "step": 3256
+    },
+    {
+      "epoch": 0.9022160664819945,
+      "grad_norm": 0.6672399044036865,
+      "learning_rate": 4.740110239921813e-06,
+      "loss": 0.6806,
+      "step": 3257
+    },
+    {
+      "epoch": 0.9024930747922437,
+      "grad_norm": 0.6685750484466553,
+      "learning_rate": 4.7399484863884815e-06,
+      "loss": 0.6392,
+      "step": 3258
+    },
+    {
+      "epoch": 0.9027700831024931,
+      "grad_norm": 0.6670814156532288,
+      "learning_rate": 4.739786685295323e-06,
+      "loss": 0.6684,
+      "step": 3259
+    },
+    {
+      "epoch": 0.9030470914127424,
+      "grad_norm": 0.6570850610733032,
+      "learning_rate": 4.739624836645773e-06,
+      "loss": 0.6547,
+      "step": 3260
+    },
+    {
+      "epoch": 0.9033240997229917,
+      "grad_norm": 0.7073267102241516,
+      "learning_rate": 4.739462940443269e-06,
+      "loss": 0.6942,
+      "step": 3261
+    },
+    {
+      "epoch": 0.903601108033241,
+      "grad_norm": 0.702384889125824,
+      "learning_rate": 4.739300996691247e-06,
+      "loss": 0.6594,
+      "step": 3262
+    },
+    {
+      "epoch": 0.9038781163434904,
+      "grad_norm": 0.6831039190292358,
+      "learning_rate": 4.739139005393148e-06,
+      "loss": 0.6795,
+      "step": 3263
+    },
+    {
+      "epoch": 0.9041551246537396,
+      "grad_norm": 0.7129132151603699,
+      "learning_rate": 4.738976966552407e-06,
+      "loss": 0.6783,
+      "step": 3264
+    },
+    {
+      "epoch": 0.9044321329639889,
+      "grad_norm": 0.706730842590332,
+      "learning_rate": 4.7388148801724695e-06,
+      "loss": 0.7052,
+      "step": 3265
+    },
+    {
+      "epoch": 0.9047091412742382,
+      "grad_norm": 0.6315455436706543,
+      "learning_rate": 4.738652746256774e-06,
+      "loss": 0.6489,
+      "step": 3266
+    },
+    {
+      "epoch": 0.9049861495844875,
+      "grad_norm": 0.6674370169639587,
+      "learning_rate": 4.7384905648087645e-06,
+      "loss": 0.6139,
+      "step": 3267
+    },
+    {
+      "epoch": 0.9052631578947369,
+      "grad_norm": 0.654758095741272,
+      "learning_rate": 4.738328335831883e-06,
+      "loss": 0.6819,
+      "step": 3268
+    },
+    {
+      "epoch": 0.9055401662049861,
+      "grad_norm": 0.6712855100631714,
+      "learning_rate": 4.738166059329575e-06,
+      "loss": 0.6338,
+      "step": 3269
+    },
+    {
+      "epoch": 0.9058171745152355,
+      "grad_norm": 0.6991258263587952,
+      "learning_rate": 4.738003735305287e-06,
+      "loss": 0.7033,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9060941828254848,
+      "grad_norm": 0.7066357731819153,
+      "learning_rate": 4.7378413637624635e-06,
+      "loss": 0.6532,
+      "step": 3271
+    },
+    {
+      "epoch": 0.9063711911357341,
+      "grad_norm": 0.6450583338737488,
+      "learning_rate": 4.737678944704552e-06,
+      "loss": 0.6506,
+      "step": 3272
+    },
+    {
+      "epoch": 0.9066481994459834,
+      "grad_norm": 0.6878437399864197,
+      "learning_rate": 4.737516478135003e-06,
+      "loss": 0.6555,
+      "step": 3273
+    },
+    {
+      "epoch": 0.9069252077562326,
+      "grad_norm": 0.686655580997467,
+      "learning_rate": 4.737353964057266e-06,
+      "loss": 0.6948,
+      "step": 3274
+    },
+    {
+      "epoch": 0.907202216066482,
+      "grad_norm": 0.7044336795806885,
+      "learning_rate": 4.73719140247479e-06,
+      "loss": 0.7176,
+      "step": 3275
+    },
+    {
+      "epoch": 0.9074792243767313,
+      "grad_norm": 0.6736946105957031,
+      "learning_rate": 4.737028793391027e-06,
+      "loss": 0.6026,
+      "step": 3276
+    },
+    {
+      "epoch": 0.9077562326869806,
+      "grad_norm": 0.6809951663017273,
+      "learning_rate": 4.736866136809431e-06,
+      "loss": 0.6892,
+      "step": 3277
+    },
+    {
+      "epoch": 0.9080332409972299,
+      "grad_norm": 0.6795469522476196,
+      "learning_rate": 4.736703432733454e-06,
+      "loss": 0.6348,
+      "step": 3278
+    },
+    {
+      "epoch": 0.9083102493074793,
+      "grad_norm": 0.674238383769989,
+      "learning_rate": 4.736540681166551e-06,
+      "loss": 0.6695,
+      "step": 3279
+    },
+    {
+      "epoch": 0.9085872576177285,
+      "grad_norm": 0.6498761177062988,
+      "learning_rate": 4.736377882112179e-06,
+      "loss": 0.6518,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9088642659279779,
+      "grad_norm": 0.6777954697608948,
+      "learning_rate": 4.736215035573792e-06,
+      "loss": 0.6654,
+      "step": 3281
+    },
+    {
+      "epoch": 0.9091412742382271,
+      "grad_norm": 0.6966802477836609,
+      "learning_rate": 4.736052141554849e-06,
+      "loss": 0.7087,
+      "step": 3282
+    },
+    {
+      "epoch": 0.9094182825484765,
+      "grad_norm": 0.6832250356674194,
+      "learning_rate": 4.73588920005881e-06,
+      "loss": 0.6806,
+      "step": 3283
+    },
+    {
+      "epoch": 0.9096952908587258,
+      "grad_norm": 0.6498856544494629,
+      "learning_rate": 4.735726211089132e-06,
+      "loss": 0.5972,
+      "step": 3284
+    },
+    {
+      "epoch": 0.909972299168975,
+      "grad_norm": 0.6577579975128174,
+      "learning_rate": 4.735563174649278e-06,
+      "loss": 0.6761,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9102493074792244,
+      "grad_norm": 0.718026876449585,
+      "learning_rate": 4.7354000907427095e-06,
+      "loss": 0.7497,
+      "step": 3286
+    },
+    {
+      "epoch": 0.9105263157894737,
+      "grad_norm": 0.6861085891723633,
+      "learning_rate": 4.7352369593728875e-06,
+      "loss": 0.7054,
+      "step": 3287
+    },
+    {
+      "epoch": 0.910803324099723,
+      "grad_norm": 0.663247287273407,
+      "learning_rate": 4.735073780543277e-06,
+      "loss": 0.695,
+      "step": 3288
+    },
+    {
+      "epoch": 0.9110803324099723,
+      "grad_norm": 0.6910631656646729,
+      "learning_rate": 4.734910554257342e-06,
+      "loss": 0.6695,
+      "step": 3289
+    },
+    {
+      "epoch": 0.9113573407202216,
+      "grad_norm": 0.6977118849754333,
+      "learning_rate": 4.734747280518549e-06,
+      "loss": 0.6954,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9116343490304709,
+      "grad_norm": 0.6666392683982849,
+      "learning_rate": 4.734583959330364e-06,
+      "loss": 0.6922,
+      "step": 3291
+    },
+    {
+      "epoch": 0.9119113573407203,
+      "grad_norm": 0.6943200826644897,
+      "learning_rate": 4.734420590696256e-06,
+      "loss": 0.6756,
+      "step": 3292
+    },
+    {
+      "epoch": 0.9121883656509695,
+      "grad_norm": 0.689194917678833,
+      "learning_rate": 4.734257174619692e-06,
+      "loss": 0.6736,
+      "step": 3293
+    },
+    {
+      "epoch": 0.9124653739612189,
+      "grad_norm": 0.6545650362968445,
+      "learning_rate": 4.7340937111041426e-06,
+      "loss": 0.6403,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9127423822714681,
+      "grad_norm": 0.6711616516113281,
+      "learning_rate": 4.733930200153078e-06,
+      "loss": 0.635,
+      "step": 3295
+    },
+    {
+      "epoch": 0.9130193905817174,
+      "grad_norm": 0.6374195218086243,
+      "learning_rate": 4.7337666417699715e-06,
+      "loss": 0.6438,
+      "step": 3296
+    },
+    {
+      "epoch": 0.9132963988919668,
+      "grad_norm": 0.6885934472084045,
+      "learning_rate": 4.733603035958293e-06,
+      "loss": 0.6421,
+      "step": 3297
+    },
+    {
+      "epoch": 0.913573407202216,
+      "grad_norm": 0.6831621527671814,
+      "learning_rate": 4.73343938272152e-06,
+      "loss": 0.6664,
+      "step": 3298
+    },
+    {
+      "epoch": 0.9138504155124654,
+      "grad_norm": 0.70931077003479,
+      "learning_rate": 4.733275682063124e-06,
+      "loss": 0.6999,
+      "step": 3299
+    },
+    {
+      "epoch": 0.9141274238227147,
+      "grad_norm": 0.6842132806777954,
+      "learning_rate": 4.733111933986583e-06,
+      "loss": 0.6761,
+      "step": 3300
+    },
+    {
+      "epoch": 0.914404432132964,
+      "grad_norm": 0.6851639151573181,
+      "learning_rate": 4.732948138495373e-06,
+      "loss": 0.6951,
+      "step": 3301
+    },
+    {
+      "epoch": 0.9146814404432133,
+      "grad_norm": 0.6987138986587524,
+      "learning_rate": 4.732784295592971e-06,
+      "loss": 0.6979,
+      "step": 3302
+    },
+    {
+      "epoch": 0.9149584487534625,
+      "grad_norm": 0.6682677865028381,
+      "learning_rate": 4.732620405282856e-06,
+      "loss": 0.6923,
+      "step": 3303
+    },
+    {
+      "epoch": 0.9152354570637119,
+      "grad_norm": 0.6620498895645142,
+      "learning_rate": 4.732456467568509e-06,
+      "loss": 0.6737,
+      "step": 3304
+    },
+    {
+      "epoch": 0.9155124653739612,
+      "grad_norm": 0.6930484175682068,
+      "learning_rate": 4.73229248245341e-06,
+      "loss": 0.6294,
+      "step": 3305
+    },
+    {
+      "epoch": 0.9157894736842105,
+      "grad_norm": 0.6822475790977478,
+      "learning_rate": 4.732128449941041e-06,
+      "loss": 0.6196,
+      "step": 3306
+    },
+    {
+      "epoch": 0.9160664819944598,
+      "grad_norm": 0.6686652898788452,
+      "learning_rate": 4.731964370034885e-06,
+      "loss": 0.635,
+      "step": 3307
+    },
+    {
+      "epoch": 0.9163434903047092,
+      "grad_norm": 0.6809656620025635,
+      "learning_rate": 4.7318002427384255e-06,
+      "loss": 0.6955,
+      "step": 3308
+    },
+    {
+      "epoch": 0.9166204986149584,
+      "grad_norm": 0.6582330465316772,
+      "learning_rate": 4.7316360680551475e-06,
+      "loss": 0.614,
+      "step": 3309
+    },
+    {
+      "epoch": 0.9168975069252078,
+      "grad_norm": 0.7251480221748352,
+      "learning_rate": 4.731471845988536e-06,
+      "loss": 0.6824,
+      "step": 3310
+    },
+    {
+      "epoch": 0.917174515235457,
+      "grad_norm": 0.645293116569519,
+      "learning_rate": 4.731307576542079e-06,
+      "loss": 0.6755,
+      "step": 3311
+    },
+    {
+      "epoch": 0.9174515235457064,
+      "grad_norm": 0.6753968596458435,
+      "learning_rate": 4.7311432597192655e-06,
+      "loss": 0.7234,
+      "step": 3312
+    },
+    {
+      "epoch": 0.9177285318559557,
+      "grad_norm": 0.6604711413383484,
+      "learning_rate": 4.730978895523581e-06,
+      "loss": 0.6388,
+      "step": 3313
+    },
+    {
+      "epoch": 0.918005540166205,
+      "grad_norm": 0.683600664138794,
+      "learning_rate": 4.7308144839585175e-06,
+      "loss": 0.681,
+      "step": 3314
+    },
+    {
+      "epoch": 0.9182825484764543,
+      "grad_norm": 0.6972975730895996,
+      "learning_rate": 4.7306500250275664e-06,
+      "loss": 0.6756,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9185595567867036,
+      "grad_norm": 0.671061635017395,
+      "learning_rate": 4.730485518734218e-06,
+      "loss": 0.6665,
+      "step": 3316
+    },
+    {
+      "epoch": 0.9188365650969529,
+      "grad_norm": 0.6876084208488464,
+      "learning_rate": 4.730320965081967e-06,
+      "loss": 0.7245,
+      "step": 3317
+    },
+    {
+      "epoch": 0.9191135734072022,
+      "grad_norm": 0.6627107858657837,
+      "learning_rate": 4.730156364074306e-06,
+      "loss": 0.6679,
+      "step": 3318
+    },
+    {
+      "epoch": 0.9193905817174515,
+      "grad_norm": 0.6769272685050964,
+      "learning_rate": 4.72999171571473e-06,
+      "loss": 0.6321,
+      "step": 3319
+    },
+    {
+      "epoch": 0.9196675900277008,
+      "grad_norm": 0.6696699261665344,
+      "learning_rate": 4.729827020006735e-06,
+      "loss": 0.6608,
+      "step": 3320
+    },
+    {
+      "epoch": 0.9199445983379502,
+      "grad_norm": 0.6594526767730713,
+      "learning_rate": 4.729662276953818e-06,
+      "loss": 0.6813,
+      "step": 3321
+    },
+    {
+      "epoch": 0.9202216066481994,
+      "grad_norm": 0.7049792408943176,
+      "learning_rate": 4.729497486559477e-06,
+      "loss": 0.6602,
+      "step": 3322
+    },
+    {
+      "epoch": 0.9204986149584488,
+      "grad_norm": 0.6542843580245972,
+      "learning_rate": 4.72933264882721e-06,
+      "loss": 0.6122,
+      "step": 3323
+    },
+    {
+      "epoch": 0.9207756232686981,
+      "grad_norm": 0.6764165759086609,
+      "learning_rate": 4.729167763760519e-06,
+      "loss": 0.7229,
+      "step": 3324
+    },
+    {
+      "epoch": 0.9210526315789473,
+      "grad_norm": 0.6534932255744934,
+      "learning_rate": 4.729002831362903e-06,
+      "loss": 0.6313,
+      "step": 3325
+    },
+    {
+      "epoch": 0.9213296398891967,
+      "grad_norm": 0.6753476858139038,
+      "learning_rate": 4.7288378516378655e-06,
+      "loss": 0.6742,
+      "step": 3326
+    },
+    {
+      "epoch": 0.9216066481994459,
+      "grad_norm": 0.6870114803314209,
+      "learning_rate": 4.728672824588908e-06,
+      "loss": 0.6941,
+      "step": 3327
+    },
+    {
+      "epoch": 0.9218836565096953,
+      "grad_norm": 0.6485161781311035,
+      "learning_rate": 4.728507750219535e-06,
+      "loss": 0.6246,
+      "step": 3328
+    },
+    {
+      "epoch": 0.9221606648199446,
+      "grad_norm": 0.6565285921096802,
+      "learning_rate": 4.7283426285332514e-06,
+      "loss": 0.6628,
+      "step": 3329
+    },
+    {
+      "epoch": 0.9224376731301939,
+      "grad_norm": 0.6443964242935181,
+      "learning_rate": 4.728177459533564e-06,
+      "loss": 0.6868,
+      "step": 3330
+    },
+    {
+      "epoch": 0.9227146814404432,
+      "grad_norm": 0.6890116333961487,
+      "learning_rate": 4.728012243223978e-06,
+      "loss": 0.7014,
+      "step": 3331
+    },
+    {
+      "epoch": 0.9229916897506926,
+      "grad_norm": 0.6694427132606506,
+      "learning_rate": 4.727846979608003e-06,
+      "loss": 0.6275,
+      "step": 3332
+    },
+    {
+      "epoch": 0.9232686980609418,
+      "grad_norm": 0.733548641204834,
+      "learning_rate": 4.7276816686891465e-06,
+      "loss": 0.6947,
+      "step": 3333
+    },
+    {
+      "epoch": 0.9235457063711912,
+      "grad_norm": 0.7278211116790771,
+      "learning_rate": 4.72751631047092e-06,
+      "loss": 0.6703,
+      "step": 3334
+    },
+    {
+      "epoch": 0.9238227146814404,
+      "grad_norm": 0.6326905488967896,
+      "learning_rate": 4.7273509049568334e-06,
+      "loss": 0.616,
+      "step": 3335
+    },
+    {
+      "epoch": 0.9240997229916897,
+      "grad_norm": 0.6611707210540771,
+      "learning_rate": 4.7271854521504e-06,
+      "loss": 0.6735,
+      "step": 3336
+    },
+    {
+      "epoch": 0.9243767313019391,
+      "grad_norm": 0.7007359266281128,
+      "learning_rate": 4.727019952055131e-06,
+      "loss": 0.6975,
+      "step": 3337
+    },
+    {
+      "epoch": 0.9246537396121883,
+      "grad_norm": 0.713338315486908,
+      "learning_rate": 4.726854404674541e-06,
+      "loss": 0.7236,
+      "step": 3338
+    },
+    {
+      "epoch": 0.9249307479224377,
+      "grad_norm": 0.6926329135894775,
+      "learning_rate": 4.7266888100121455e-06,
+      "loss": 0.6923,
+      "step": 3339
+    },
+    {
+      "epoch": 0.925207756232687,
+      "grad_norm": 0.6664144992828369,
+      "learning_rate": 4.7265231680714605e-06,
+      "loss": 0.691,
+      "step": 3340
+    },
+    {
+      "epoch": 0.9254847645429363,
+      "grad_norm": 0.6612799763679504,
+      "learning_rate": 4.726357478856002e-06,
+      "loss": 0.6582,
+      "step": 3341
+    },
+    {
+      "epoch": 0.9257617728531856,
+      "grad_norm": 0.6593889594078064,
+      "learning_rate": 4.726191742369289e-06,
+      "loss": 0.6536,
+      "step": 3342
+    },
+    {
+      "epoch": 0.9260387811634349,
+      "grad_norm": 0.6893621683120728,
+      "learning_rate": 4.7260259586148405e-06,
+      "loss": 0.6766,
+      "step": 3343
+    },
+    {
+      "epoch": 0.9263157894736842,
+      "grad_norm": 0.7140992879867554,
+      "learning_rate": 4.725860127596176e-06,
+      "loss": 0.6448,
+      "step": 3344
+    },
+    {
+      "epoch": 0.9265927977839336,
+      "grad_norm": 0.6934571862220764,
+      "learning_rate": 4.7256942493168176e-06,
+      "loss": 0.6908,
+      "step": 3345
+    },
+    {
+      "epoch": 0.9268698060941828,
+      "grad_norm": 0.644855260848999,
+      "learning_rate": 4.725528323780285e-06,
+      "loss": 0.6571,
+      "step": 3346
+    },
+    {
+      "epoch": 0.9271468144044321,
+      "grad_norm": 0.6686730980873108,
+      "learning_rate": 4.725362350990104e-06,
+      "loss": 0.6717,
+      "step": 3347
+    },
+    {
+      "epoch": 0.9274238227146815,
+      "grad_norm": 0.6989774703979492,
+      "learning_rate": 4.725196330949798e-06,
+      "loss": 0.6618,
+      "step": 3348
+    },
+    {
+      "epoch": 0.9277008310249307,
+      "grad_norm": 0.6894549131393433,
+      "learning_rate": 4.7250302636628895e-06,
+      "loss": 0.6222,
+      "step": 3349
+    },
+    {
+      "epoch": 0.9279778393351801,
+      "grad_norm": 0.7115414142608643,
+      "learning_rate": 4.724864149132907e-06,
+      "loss": 0.6424,
+      "step": 3350
+    },
+    {
+      "epoch": 0.9282548476454293,
+      "grad_norm": 0.6588413715362549,
+      "learning_rate": 4.7246979873633766e-06,
+      "loss": 0.6775,
+      "step": 3351
+    },
+    {
+      "epoch": 0.9285318559556787,
+      "grad_norm": 0.6895866990089417,
+      "learning_rate": 4.724531778357827e-06,
+      "loss": 0.6668,
+      "step": 3352
+    },
+    {
+      "epoch": 0.928808864265928,
+      "grad_norm": 0.7034494876861572,
+      "learning_rate": 4.724365522119787e-06,
+      "loss": 0.6918,
+      "step": 3353
+    },
+    {
+      "epoch": 0.9290858725761773,
+      "grad_norm": 0.685469925403595,
+      "learning_rate": 4.724199218652787e-06,
+      "loss": 0.6665,
+      "step": 3354
+    },
+    {
+      "epoch": 0.9293628808864266,
+      "grad_norm": 0.6953645348548889,
+      "learning_rate": 4.724032867960357e-06,
+      "loss": 0.62,
+      "step": 3355
+    },
+    {
+      "epoch": 0.929639889196676,
+      "grad_norm": 0.6760872006416321,
+      "learning_rate": 4.72386647004603e-06,
+      "loss": 0.6415,
+      "step": 3356
+    },
+    {
+      "epoch": 0.9299168975069252,
+      "grad_norm": 0.7333945035934448,
+      "learning_rate": 4.723700024913339e-06,
+      "loss": 0.6358,
+      "step": 3357
+    },
+    {
+      "epoch": 0.9301939058171745,
+      "grad_norm": 0.6890769600868225,
+      "learning_rate": 4.723533532565817e-06,
+      "loss": 0.6157,
+      "step": 3358
+    },
+    {
+      "epoch": 0.9304709141274238,
+      "grad_norm": 0.7347223162651062,
+      "learning_rate": 4.723366993007e-06,
+      "loss": 0.6789,
+      "step": 3359
+    },
+    {
+      "epoch": 0.9307479224376731,
+      "grad_norm": 0.6665700078010559,
+      "learning_rate": 4.7232004062404235e-06,
+      "loss": 0.6589,
+      "step": 3360
+    },
+    {
+      "epoch": 0.9310249307479225,
+      "grad_norm": 0.618806779384613,
+      "learning_rate": 4.723033772269626e-06,
+      "loss": 0.6269,
+      "step": 3361
+    },
+    {
+      "epoch": 0.9313019390581717,
+      "grad_norm": 0.6587586998939514,
+      "learning_rate": 4.722867091098144e-06,
+      "loss": 0.6907,
+      "step": 3362
+    },
+    {
+      "epoch": 0.9315789473684211,
+      "grad_norm": 0.6528121829032898,
+      "learning_rate": 4.722700362729516e-06,
+      "loss": 0.6062,
+      "step": 3363
+    },
+    {
+      "epoch": 0.9318559556786704,
+      "grad_norm": 0.6848546862602234,
+      "learning_rate": 4.722533587167285e-06,
+      "loss": 0.6603,
+      "step": 3364
+    },
+    {
+      "epoch": 0.9321329639889196,
+      "grad_norm": 0.6691559553146362,
+      "learning_rate": 4.722366764414989e-06,
+      "loss": 0.7152,
+      "step": 3365
+    },
+    {
+      "epoch": 0.932409972299169,
+      "grad_norm": 0.7005979418754578,
+      "learning_rate": 4.722199894476172e-06,
+      "loss": 0.6947,
+      "step": 3366
+    },
+    {
+      "epoch": 0.9326869806094182,
+      "grad_norm": 0.6714586615562439,
+      "learning_rate": 4.722032977354376e-06,
+      "loss": 0.6685,
+      "step": 3367
+    },
+    {
+      "epoch": 0.9329639889196676,
+      "grad_norm": 0.6560215353965759,
+      "learning_rate": 4.721866013053146e-06,
+      "loss": 0.6315,
+      "step": 3368
+    },
+    {
+      "epoch": 0.9332409972299169,
+      "grad_norm": 0.7032321095466614,
+      "learning_rate": 4.721699001576026e-06,
+      "loss": 0.6492,
+      "step": 3369
+    },
+    {
+      "epoch": 0.9335180055401662,
+      "grad_norm": 0.6863848567008972,
+      "learning_rate": 4.721531942926561e-06,
+      "loss": 0.6619,
+      "step": 3370
+    },
+    {
+      "epoch": 0.9337950138504155,
+      "grad_norm": 0.65123051404953,
+      "learning_rate": 4.721364837108302e-06,
+      "loss": 0.6821,
+      "step": 3371
+    },
+    {
+      "epoch": 0.9340720221606649,
+      "grad_norm": 0.6889357566833496,
+      "learning_rate": 4.7211976841247945e-06,
+      "loss": 0.6522,
+      "step": 3372
+    },
+    {
+      "epoch": 0.9343490304709141,
+      "grad_norm": 0.6711523532867432,
+      "learning_rate": 4.721030483979588e-06,
+      "loss": 0.6599,
+      "step": 3373
+    },
+    {
+      "epoch": 0.9346260387811635,
+      "grad_norm": 0.6738723516464233,
+      "learning_rate": 4.7208632366762305e-06,
+      "loss": 0.6697,
+      "step": 3374
+    },
+    {
+      "epoch": 0.9349030470914127,
+      "grad_norm": 0.6907211542129517,
+      "learning_rate": 4.7206959422182764e-06,
+      "loss": 0.6722,
+      "step": 3375
+    },
+    {
+      "epoch": 0.935180055401662,
+      "grad_norm": 0.6392459869384766,
+      "learning_rate": 4.720528600609276e-06,
+      "loss": 0.6543,
+      "step": 3376
+    },
+    {
+      "epoch": 0.9354570637119114,
+      "grad_norm": 0.7044897079467773,
+      "learning_rate": 4.720361211852784e-06,
+      "loss": 0.6878,
+      "step": 3377
+    },
+    {
+      "epoch": 0.9357340720221606,
+      "grad_norm": 0.7036895751953125,
+      "learning_rate": 4.720193775952352e-06,
+      "loss": 0.6862,
+      "step": 3378
+    },
+    {
+      "epoch": 0.93601108033241,
+      "grad_norm": 0.6759652495384216,
+      "learning_rate": 4.720026292911537e-06,
+      "loss": 0.6192,
+      "step": 3379
+    },
+    {
+      "epoch": 0.9362880886426593,
+      "grad_norm": 0.6304213404655457,
+      "learning_rate": 4.719858762733894e-06,
+      "loss": 0.6286,
+      "step": 3380
+    },
+    {
+      "epoch": 0.9365650969529086,
+      "grad_norm": 0.6903957724571228,
+      "learning_rate": 4.719691185422981e-06,
+      "loss": 0.7026,
+      "step": 3381
+    },
+    {
+      "epoch": 0.9368421052631579,
+      "grad_norm": 0.6814022660255432,
+      "learning_rate": 4.719523560982355e-06,
+      "loss": 0.6974,
+      "step": 3382
+    },
+    {
+      "epoch": 0.9371191135734072,
+      "grad_norm": 0.6745713949203491,
+      "learning_rate": 4.719355889415577e-06,
+      "loss": 0.6544,
+      "step": 3383
+    },
+    {
+      "epoch": 0.9373961218836565,
+      "grad_norm": 0.7096647620201111,
+      "learning_rate": 4.719188170726204e-06,
+      "loss": 0.7011,
+      "step": 3384
+    },
+    {
+      "epoch": 0.9376731301939059,
+      "grad_norm": 0.7082591652870178,
+      "learning_rate": 4.719020404917801e-06,
+      "loss": 0.596,
+      "step": 3385
+    },
+    {
+      "epoch": 0.9379501385041551,
+      "grad_norm": 0.6712037324905396,
+      "learning_rate": 4.718852591993927e-06,
+      "loss": 0.6941,
+      "step": 3386
+    },
+    {
+      "epoch": 0.9382271468144044,
+      "grad_norm": 0.6850805878639221,
+      "learning_rate": 4.718684731958146e-06,
+      "loss": 0.654,
+      "step": 3387
+    },
+    {
+      "epoch": 0.9385041551246538,
+      "grad_norm": 0.6575900316238403,
+      "learning_rate": 4.718516824814022e-06,
+      "loss": 0.7019,
+      "step": 3388
+    },
+    {
+      "epoch": 0.938781163434903,
+      "grad_norm": 0.6687057614326477,
+      "learning_rate": 4.7183488705651216e-06,
+      "loss": 0.6384,
+      "step": 3389
+    },
+    {
+      "epoch": 0.9390581717451524,
+      "grad_norm": 0.6599818468093872,
+      "learning_rate": 4.718180869215008e-06,
+      "loss": 0.6785,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9393351800554016,
+      "grad_norm": 0.6838821172714233,
+      "learning_rate": 4.718012820767251e-06,
+      "loss": 0.7252,
+      "step": 3391
+    },
+    {
+      "epoch": 0.939612188365651,
+      "grad_norm": 0.697323203086853,
+      "learning_rate": 4.717844725225418e-06,
+      "loss": 0.6596,
+      "step": 3392
+    },
+    {
+      "epoch": 0.9398891966759003,
+      "grad_norm": 0.6465617418289185,
+      "learning_rate": 4.717676582593076e-06,
+      "loss": 0.6527,
+      "step": 3393
+    },
+    {
+      "epoch": 0.9401662049861496,
+      "grad_norm": 0.6578206419944763,
+      "learning_rate": 4.7175083928737984e-06,
+      "loss": 0.6432,
+      "step": 3394
+    },
+    {
+      "epoch": 0.9404432132963989,
+      "grad_norm": 0.670072615146637,
+      "learning_rate": 4.717340156071154e-06,
+      "loss": 0.6704,
+      "step": 3395
+    },
+    {
+      "epoch": 0.9407202216066483,
+      "grad_norm": 0.6486279964447021,
+      "learning_rate": 4.717171872188716e-06,
+      "loss": 0.6744,
+      "step": 3396
+    },
+    {
+      "epoch": 0.9409972299168975,
+      "grad_norm": 0.681574285030365,
+      "learning_rate": 4.717003541230057e-06,
+      "loss": 0.6656,
+      "step": 3397
+    },
+    {
+      "epoch": 0.9412742382271468,
+      "grad_norm": 0.6697400808334351,
+      "learning_rate": 4.71683516319875e-06,
+      "loss": 0.6755,
+      "step": 3398
+    },
+    {
+      "epoch": 0.9415512465373961,
+      "grad_norm": 0.7256743907928467,
+      "learning_rate": 4.716666738098373e-06,
+      "loss": 0.69,
+      "step": 3399
+    },
+    {
+      "epoch": 0.9418282548476454,
+      "grad_norm": 0.670193612575531,
+      "learning_rate": 4.716498265932501e-06,
+      "loss": 0.6904,
+      "step": 3400
+    },
+    {
+      "epoch": 0.9421052631578948,
+      "grad_norm": 0.6912937760353088,
+      "learning_rate": 4.716329746704709e-06,
+      "loss": 0.7133,
+      "step": 3401
+    },
+    {
+      "epoch": 0.942382271468144,
+      "grad_norm": 0.6947979927062988,
+      "learning_rate": 4.716161180418577e-06,
+      "loss": 0.6339,
+      "step": 3402
+    },
+    {
+      "epoch": 0.9426592797783934,
+      "grad_norm": 0.7074637413024902,
+      "learning_rate": 4.715992567077683e-06,
+      "loss": 0.6328,
+      "step": 3403
+    },
+    {
+      "epoch": 0.9429362880886426,
+      "grad_norm": 0.6708641648292542,
+      "learning_rate": 4.7158239066856096e-06,
+      "loss": 0.645,
+      "step": 3404
+    },
+    {
+      "epoch": 0.943213296398892,
+      "grad_norm": 0.6959966421127319,
+      "learning_rate": 4.7156551992459346e-06,
+      "loss": 0.6736,
+      "step": 3405
+    },
+    {
+      "epoch": 0.9434903047091413,
+      "grad_norm": 0.682021975517273,
+      "learning_rate": 4.715486444762242e-06,
+      "loss": 0.6571,
+      "step": 3406
+    },
+    {
+      "epoch": 0.9437673130193905,
+      "grad_norm": 0.7637782096862793,
+      "learning_rate": 4.715317643238114e-06,
+      "loss": 0.6335,
+      "step": 3407
+    },
+    {
+      "epoch": 0.9440443213296399,
+      "grad_norm": 0.6714920401573181,
+      "learning_rate": 4.715148794677135e-06,
+      "loss": 0.7047,
+      "step": 3408
+    },
+    {
+      "epoch": 0.9443213296398892,
+      "grad_norm": 0.6588884592056274,
+      "learning_rate": 4.714979899082891e-06,
+      "loss": 0.6834,
+      "step": 3409
+    },
+    {
+      "epoch": 0.9445983379501385,
+      "grad_norm": 0.6203243732452393,
+      "learning_rate": 4.714810956458967e-06,
+      "loss": 0.5646,
+      "step": 3410
+    },
+    {
+      "epoch": 0.9448753462603878,
+      "grad_norm": 0.6979451179504395,
+      "learning_rate": 4.71464196680895e-06,
+      "loss": 0.6774,
+      "step": 3411
+    },
+    {
+      "epoch": 0.9451523545706371,
+      "grad_norm": 0.6701903343200684,
+      "learning_rate": 4.714472930136429e-06,
+      "loss": 0.6498,
+      "step": 3412
+    },
+    {
+      "epoch": 0.9454293628808864,
+      "grad_norm": 0.7109444737434387,
+      "learning_rate": 4.714303846444992e-06,
+      "loss": 0.6503,
+      "step": 3413
+    },
+    {
+      "epoch": 0.9457063711911358,
+      "grad_norm": 0.7005951404571533,
+      "learning_rate": 4.71413471573823e-06,
+      "loss": 0.6476,
+      "step": 3414
+    },
+    {
+      "epoch": 0.945983379501385,
+      "grad_norm": 0.6723399758338928,
+      "learning_rate": 4.7139655380197336e-06,
+      "loss": 0.6349,
+      "step": 3415
+    },
+    {
+      "epoch": 0.9462603878116344,
+      "grad_norm": 0.7047520875930786,
+      "learning_rate": 4.713796313293096e-06,
+      "loss": 0.7124,
+      "step": 3416
+    },
+    {
+      "epoch": 0.9465373961218837,
+      "grad_norm": 0.6956657767295837,
+      "learning_rate": 4.7136270415619075e-06,
+      "loss": 0.6001,
+      "step": 3417
+    },
+    {
+      "epoch": 0.9468144044321329,
+      "grad_norm": 0.71816486120224,
+      "learning_rate": 4.713457722829765e-06,
+      "loss": 0.6532,
+      "step": 3418
+    },
+    {
+      "epoch": 0.9470914127423823,
+      "grad_norm": 0.6962186098098755,
+      "learning_rate": 4.713288357100262e-06,
+      "loss": 0.6849,
+      "step": 3419
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.746861457824707,
+      "learning_rate": 4.713118944376996e-06,
+      "loss": 0.6546,
+      "step": 3420
+    },
+    {
+      "epoch": 0.9476454293628809,
+      "grad_norm": 0.6704689264297485,
+      "learning_rate": 4.712949484663561e-06,
+      "loss": 0.6357,
+      "step": 3421
+    },
+    {
+      "epoch": 0.9479224376731302,
+      "grad_norm": 0.6442332863807678,
+      "learning_rate": 4.712779977963559e-06,
+      "loss": 0.6806,
+      "step": 3422
+    },
+    {
+      "epoch": 0.9481994459833795,
+      "grad_norm": 0.6512526869773865,
+      "learning_rate": 4.7126104242805866e-06,
+      "loss": 0.6601,
+      "step": 3423
+    },
+    {
+      "epoch": 0.9484764542936288,
+      "grad_norm": 0.6544331908226013,
+      "learning_rate": 4.712440823618245e-06,
+      "loss": 0.6574,
+      "step": 3424
+    },
+    {
+      "epoch": 0.9487534626038782,
+      "grad_norm": 0.7471049427986145,
+      "learning_rate": 4.712271175980134e-06,
+      "loss": 0.6827,
+      "step": 3425
+    },
+    {
+      "epoch": 0.9490304709141274,
+      "grad_norm": 0.6754305362701416,
+      "learning_rate": 4.712101481369857e-06,
+      "loss": 0.7138,
+      "step": 3426
+    },
+    {
+      "epoch": 0.9493074792243767,
+      "grad_norm": 0.6700847148895264,
+      "learning_rate": 4.711931739791017e-06,
+      "loss": 0.6737,
+      "step": 3427
+    },
+    {
+      "epoch": 0.949584487534626,
+      "grad_norm": 0.6627695560455322,
+      "learning_rate": 4.711761951247216e-06,
+      "loss": 0.6215,
+      "step": 3428
+    },
+    {
+      "epoch": 0.9498614958448753,
+      "grad_norm": 0.6418578028678894,
+      "learning_rate": 4.711592115742061e-06,
+      "loss": 0.6435,
+      "step": 3429
+    },
+    {
+      "epoch": 0.9501385041551247,
+      "grad_norm": 0.666086733341217,
+      "learning_rate": 4.711422233279158e-06,
+      "loss": 0.666,
+      "step": 3430
+    },
+    {
+      "epoch": 0.9504155124653739,
+      "grad_norm": 0.6819273829460144,
+      "learning_rate": 4.711252303862114e-06,
+      "loss": 0.6654,
+      "step": 3431
+    },
+    {
+      "epoch": 0.9506925207756233,
+      "grad_norm": 0.69044429063797,
+      "learning_rate": 4.711082327494536e-06,
+      "loss": 0.6637,
+      "step": 3432
+    },
+    {
+      "epoch": 0.9509695290858726,
+      "grad_norm": 0.6500899791717529,
+      "learning_rate": 4.710912304180034e-06,
+      "loss": 0.6722,
+      "step": 3433
+    },
+    {
+      "epoch": 0.9512465373961219,
+      "grad_norm": 0.6798520684242249,
+      "learning_rate": 4.710742233922217e-06,
+      "loss": 0.6408,
+      "step": 3434
+    },
+    {
+      "epoch": 0.9515235457063712,
+      "grad_norm": 0.6969449520111084,
+      "learning_rate": 4.710572116724698e-06,
+      "loss": 0.6702,
+      "step": 3435
+    },
+    {
+      "epoch": 0.9518005540166204,
+      "grad_norm": 0.7223117351531982,
+      "learning_rate": 4.710401952591088e-06,
+      "loss": 0.6647,
+      "step": 3436
+    },
+    {
+      "epoch": 0.9520775623268698,
+      "grad_norm": 0.6953335404396057,
+      "learning_rate": 4.710231741524999e-06,
+      "loss": 0.6811,
+      "step": 3437
+    },
+    {
+      "epoch": 0.9523545706371191,
+      "grad_norm": 0.6859951019287109,
+      "learning_rate": 4.7100614835300454e-06,
+      "loss": 0.6611,
+      "step": 3438
+    },
+    {
+      "epoch": 0.9526315789473684,
+      "grad_norm": 0.6977407336235046,
+      "learning_rate": 4.709891178609843e-06,
+      "loss": 0.7132,
+      "step": 3439
+    },
+    {
+      "epoch": 0.9529085872576177,
+      "grad_norm": 0.7158917784690857,
+      "learning_rate": 4.709720826768007e-06,
+      "loss": 0.655,
+      "step": 3440
+    },
+    {
+      "epoch": 0.9531855955678671,
+      "grad_norm": 0.6699243187904358,
+      "learning_rate": 4.709550428008155e-06,
+      "loss": 0.6466,
+      "step": 3441
+    },
+    {
+      "epoch": 0.9534626038781163,
+      "grad_norm": 0.6597692966461182,
+      "learning_rate": 4.709379982333906e-06,
+      "loss": 0.6597,
+      "step": 3442
+    },
+    {
+      "epoch": 0.9537396121883657,
+      "grad_norm": 0.6464130282402039,
+      "learning_rate": 4.709209489748877e-06,
+      "loss": 0.6717,
+      "step": 3443
+    },
+    {
+      "epoch": 0.9540166204986149,
+      "grad_norm": 0.6933079957962036,
+      "learning_rate": 4.7090389502566884e-06,
+      "loss": 0.6533,
+      "step": 3444
+    },
+    {
+      "epoch": 0.9542936288088643,
+      "grad_norm": 0.6588965058326721,
+      "learning_rate": 4.708868363860962e-06,
+      "loss": 0.6764,
+      "step": 3445
+    },
+    {
+      "epoch": 0.9545706371191136,
+      "grad_norm": 0.6584217548370361,
+      "learning_rate": 4.7086977305653195e-06,
+      "loss": 0.6812,
+      "step": 3446
+    },
+    {
+      "epoch": 0.9548476454293628,
+      "grad_norm": 0.6467680335044861,
+      "learning_rate": 4.7085270503733835e-06,
+      "loss": 0.7133,
+      "step": 3447
+    },
+    {
+      "epoch": 0.9551246537396122,
+      "grad_norm": 0.6523286700248718,
+      "learning_rate": 4.708356323288779e-06,
+      "loss": 0.6373,
+      "step": 3448
+    },
+    {
+      "epoch": 0.9554016620498615,
+      "grad_norm": 0.6429436206817627,
+      "learning_rate": 4.708185549315129e-06,
+      "loss": 0.6679,
+      "step": 3449
+    },
+    {
+      "epoch": 0.9556786703601108,
+      "grad_norm": 0.6593552231788635,
+      "learning_rate": 4.708014728456062e-06,
+      "loss": 0.6264,
+      "step": 3450
+    },
+    {
+      "epoch": 0.9559556786703601,
+      "grad_norm": 0.692894697189331,
+      "learning_rate": 4.707843860715204e-06,
+      "loss": 0.6142,
+      "step": 3451
+    },
+    {
+      "epoch": 0.9562326869806094,
+      "grad_norm": 0.6877021193504333,
+      "learning_rate": 4.70767294609618e-06,
+      "loss": 0.5669,
+      "step": 3452
+    },
+    {
+      "epoch": 0.9565096952908587,
+      "grad_norm": 0.7020219564437866,
+      "learning_rate": 4.707501984602624e-06,
+      "loss": 0.6703,
+      "step": 3453
+    },
+    {
+      "epoch": 0.9567867036011081,
+      "grad_norm": 0.7109591960906982,
+      "learning_rate": 4.707330976238162e-06,
+      "loss": 0.6907,
+      "step": 3454
+    },
+    {
+      "epoch": 0.9570637119113573,
+      "grad_norm": 0.7097887396812439,
+      "learning_rate": 4.7071599210064275e-06,
+      "loss": 0.6712,
+      "step": 3455
+    },
+    {
+      "epoch": 0.9573407202216067,
+      "grad_norm": 0.7130573391914368,
+      "learning_rate": 4.7069888189110516e-06,
+      "loss": 0.6989,
+      "step": 3456
+    },
+    {
+      "epoch": 0.957617728531856,
+      "grad_norm": 0.6701627969741821,
+      "learning_rate": 4.706817669955667e-06,
+      "loss": 0.704,
+      "step": 3457
+    },
+    {
+      "epoch": 0.9578947368421052,
+      "grad_norm": 0.6597571969032288,
+      "learning_rate": 4.7066464741439064e-06,
+      "loss": 0.6759,
+      "step": 3458
+    },
+    {
+      "epoch": 0.9581717451523546,
+      "grad_norm": 0.6537261009216309,
+      "learning_rate": 4.706475231479407e-06,
+      "loss": 0.6853,
+      "step": 3459
+    },
+    {
+      "epoch": 0.9584487534626038,
+      "grad_norm": 0.6804844737052917,
+      "learning_rate": 4.706303941965804e-06,
+      "loss": 0.6645,
+      "step": 3460
+    },
+    {
+      "epoch": 0.9587257617728532,
+      "grad_norm": 0.6807675361633301,
+      "learning_rate": 4.706132605606733e-06,
+      "loss": 0.6846,
+      "step": 3461
+    },
+    {
+      "epoch": 0.9590027700831025,
+      "grad_norm": 0.7017925381660461,
+      "learning_rate": 4.7059612224058335e-06,
+      "loss": 0.6329,
+      "step": 3462
+    },
+    {
+      "epoch": 0.9592797783933518,
+      "grad_norm": 0.6846858859062195,
+      "learning_rate": 4.705789792366744e-06,
+      "loss": 0.6858,
+      "step": 3463
+    },
+    {
+      "epoch": 0.9595567867036011,
+      "grad_norm": 0.7076050639152527,
+      "learning_rate": 4.705618315493104e-06,
+      "loss": 0.653,
+      "step": 3464
+    },
+    {
+      "epoch": 0.9598337950138505,
+      "grad_norm": 0.6549046039581299,
+      "learning_rate": 4.7054467917885545e-06,
+      "loss": 0.6799,
+      "step": 3465
+    },
+    {
+      "epoch": 0.9601108033240997,
+      "grad_norm": 0.668217658996582,
+      "learning_rate": 4.705275221256738e-06,
+      "loss": 0.6367,
+      "step": 3466
+    },
+    {
+      "epoch": 0.960387811634349,
+      "grad_norm": 0.6847381591796875,
+      "learning_rate": 4.705103603901296e-06,
+      "loss": 0.6493,
+      "step": 3467
+    },
+    {
+      "epoch": 0.9606648199445983,
+      "grad_norm": 0.7230570316314697,
+      "learning_rate": 4.704931939725874e-06,
+      "loss": 0.6785,
+      "step": 3468
+    },
+    {
+      "epoch": 0.9609418282548476,
+      "grad_norm": 0.6642930507659912,
+      "learning_rate": 4.704760228734116e-06,
+      "loss": 0.6195,
+      "step": 3469
+    },
+    {
+      "epoch": 0.961218836565097,
+      "grad_norm": 0.6809485554695129,
+      "learning_rate": 4.704588470929668e-06,
+      "loss": 0.6773,
+      "step": 3470
+    },
+    {
+      "epoch": 0.9614958448753462,
+      "grad_norm": 0.6308563351631165,
+      "learning_rate": 4.704416666316177e-06,
+      "loss": 0.6313,
+      "step": 3471
+    },
+    {
+      "epoch": 0.9617728531855956,
+      "grad_norm": 0.6679580211639404,
+      "learning_rate": 4.704244814897291e-06,
+      "loss": 0.6274,
+      "step": 3472
+    },
+    {
+      "epoch": 0.9620498614958449,
+      "grad_norm": 0.7183241844177246,
+      "learning_rate": 4.704072916676658e-06,
+      "loss": 0.6432,
+      "step": 3473
+    },
+    {
+      "epoch": 0.9623268698060942,
+      "grad_norm": 0.6756312847137451,
+      "learning_rate": 4.703900971657929e-06,
+      "loss": 0.6477,
+      "step": 3474
+    },
+    {
+      "epoch": 0.9626038781163435,
+      "grad_norm": 0.7039386034011841,
+      "learning_rate": 4.703728979844753e-06,
+      "loss": 0.6681,
+      "step": 3475
+    },
+    {
+      "epoch": 0.9628808864265928,
+      "grad_norm": 0.6932883262634277,
+      "learning_rate": 4.703556941240784e-06,
+      "loss": 0.6858,
+      "step": 3476
+    },
+    {
+      "epoch": 0.9631578947368421,
+      "grad_norm": 0.9859476089477539,
+      "learning_rate": 4.7033848558496735e-06,
+      "loss": 0.6897,
+      "step": 3477
+    },
+    {
+      "epoch": 0.9634349030470915,
+      "grad_norm": 0.7097819447517395,
+      "learning_rate": 4.703212723675076e-06,
+      "loss": 0.6639,
+      "step": 3478
+    },
+    {
+      "epoch": 0.9637119113573407,
+      "grad_norm": 0.6593777537345886,
+      "learning_rate": 4.7030405447206465e-06,
+      "loss": 0.6823,
+      "step": 3479
+    },
+    {
+      "epoch": 0.96398891966759,
+      "grad_norm": 0.696010410785675,
+      "learning_rate": 4.702868318990038e-06,
+      "loss": 0.6707,
+      "step": 3480
+    },
+    {
+      "epoch": 0.9642659279778394,
+      "grad_norm": 0.681574821472168,
+      "learning_rate": 4.702696046486912e-06,
+      "loss": 0.6909,
+      "step": 3481
+    },
+    {
+      "epoch": 0.9645429362880886,
+      "grad_norm": 0.6492648720741272,
+      "learning_rate": 4.702523727214922e-06,
+      "loss": 0.6436,
+      "step": 3482
+    },
+    {
+      "epoch": 0.964819944598338,
+      "grad_norm": 0.6883756518363953,
+      "learning_rate": 4.7023513611777305e-06,
+      "loss": 0.6718,
+      "step": 3483
+    },
+    {
+      "epoch": 0.9650969529085872,
+      "grad_norm": 0.6744667887687683,
+      "learning_rate": 4.702178948378995e-06,
+      "loss": 0.6664,
+      "step": 3484
+    },
+    {
+      "epoch": 0.9653739612188366,
+      "grad_norm": 0.676249086856842,
+      "learning_rate": 4.702006488822376e-06,
+      "loss": 0.6607,
+      "step": 3485
+    },
+    {
+      "epoch": 0.9656509695290859,
+      "grad_norm": 0.6973958015441895,
+      "learning_rate": 4.701833982511537e-06,
+      "loss": 0.6672,
+      "step": 3486
+    },
+    {
+      "epoch": 0.9659279778393352,
+      "grad_norm": 0.6347668170928955,
+      "learning_rate": 4.701661429450139e-06,
+      "loss": 0.5975,
+      "step": 3487
+    },
+    {
+      "epoch": 0.9662049861495845,
+      "grad_norm": 0.6446684002876282,
+      "learning_rate": 4.701488829641845e-06,
+      "loss": 0.6622,
+      "step": 3488
+    },
+    {
+      "epoch": 0.9664819944598338,
+      "grad_norm": 0.7026192545890808,
+      "learning_rate": 4.7013161830903235e-06,
+      "loss": 0.6786,
+      "step": 3489
+    },
+    {
+      "epoch": 0.9667590027700831,
+      "grad_norm": 0.6588451862335205,
+      "learning_rate": 4.701143489799236e-06,
+      "loss": 0.6729,
+      "step": 3490
+    },
+    {
+      "epoch": 0.9670360110803324,
+      "grad_norm": 0.6893823146820068,
+      "learning_rate": 4.700970749772253e-06,
+      "loss": 0.6637,
+      "step": 3491
+    },
+    {
+      "epoch": 0.9673130193905817,
+      "grad_norm": 0.7043506503105164,
+      "learning_rate": 4.700797963013039e-06,
+      "loss": 0.6648,
+      "step": 3492
+    },
+    {
+      "epoch": 0.967590027700831,
+      "grad_norm": 0.6461666226387024,
+      "learning_rate": 4.700625129525263e-06,
+      "loss": 0.6728,
+      "step": 3493
+    },
+    {
+      "epoch": 0.9678670360110804,
+      "grad_norm": 0.6638358235359192,
+      "learning_rate": 4.7004522493125975e-06,
+      "loss": 0.6815,
+      "step": 3494
+    },
+    {
+      "epoch": 0.9681440443213296,
+      "grad_norm": 0.6599012017250061,
+      "learning_rate": 4.700279322378712e-06,
+      "loss": 0.6831,
+      "step": 3495
+    },
+    {
+      "epoch": 0.968421052631579,
+      "grad_norm": 0.7229272127151489,
+      "learning_rate": 4.700106348727276e-06,
+      "loss": 0.6823,
+      "step": 3496
+    },
+    {
+      "epoch": 0.9686980609418283,
+      "grad_norm": 0.6756724119186401,
+      "learning_rate": 4.699933328361964e-06,
+      "loss": 0.6549,
+      "step": 3497
+    },
+    {
+      "epoch": 0.9689750692520775,
+      "grad_norm": 0.6939228773117065,
+      "learning_rate": 4.699760261286449e-06,
+      "loss": 0.7094,
+      "step": 3498
+    },
+    {
+      "epoch": 0.9692520775623269,
+      "grad_norm": 0.6404763460159302,
+      "learning_rate": 4.699587147504407e-06,
+      "loss": 0.6724,
+      "step": 3499
+    },
+    {
+      "epoch": 0.9695290858725761,
+      "grad_norm": 0.7101907134056091,
+      "learning_rate": 4.6994139870195125e-06,
+      "loss": 0.712,
+      "step": 3500
+    },
+    {
+      "epoch": 0.9698060941828255,
+      "grad_norm": 0.6645216345787048,
+      "learning_rate": 4.699240779835443e-06,
+      "loss": 0.6843,
+      "step": 3501
+    },
+    {
+      "epoch": 0.9700831024930748,
+      "grad_norm": 0.6481513381004333,
+      "learning_rate": 4.699067525955874e-06,
+      "loss": 0.6724,
+      "step": 3502
+    },
+    {
+      "epoch": 0.9703601108033241,
+      "grad_norm": 0.7024312615394592,
+      "learning_rate": 4.698894225384487e-06,
+      "loss": 0.6837,
+      "step": 3503
+    },
+    {
+      "epoch": 0.9706371191135734,
+      "grad_norm": 0.6771408319473267,
+      "learning_rate": 4.6987208781249595e-06,
+      "loss": 0.6889,
+      "step": 3504
+    },
+    {
+      "epoch": 0.9709141274238227,
+      "grad_norm": 0.6590652465820312,
+      "learning_rate": 4.698547484180973e-06,
+      "loss": 0.6646,
+      "step": 3505
+    },
+    {
+      "epoch": 0.971191135734072,
+      "grad_norm": 0.7020836472511292,
+      "learning_rate": 4.698374043556209e-06,
+      "loss": 0.7133,
+      "step": 3506
+    },
+    {
+      "epoch": 0.9714681440443214,
+      "grad_norm": 0.6353387236595154,
+      "learning_rate": 4.69820055625435e-06,
+      "loss": 0.6207,
+      "step": 3507
+    },
+    {
+      "epoch": 0.9717451523545706,
+      "grad_norm": 0.6599223613739014,
+      "learning_rate": 4.698027022279079e-06,
+      "loss": 0.6412,
+      "step": 3508
+    },
+    {
+      "epoch": 0.97202216066482,
+      "grad_norm": 0.6925457715988159,
+      "learning_rate": 4.697853441634082e-06,
+      "loss": 0.6453,
+      "step": 3509
+    },
+    {
+      "epoch": 0.9722991689750693,
+      "grad_norm": 0.6730129718780518,
+      "learning_rate": 4.697679814323044e-06,
+      "loss": 0.6777,
+      "step": 3510
+    },
+    {
+      "epoch": 0.9725761772853185,
+      "grad_norm": 0.6823524832725525,
+      "learning_rate": 4.69750614034965e-06,
+      "loss": 0.7242,
+      "step": 3511
+    },
+    {
+      "epoch": 0.9728531855955679,
+      "grad_norm": 0.6779269576072693,
+      "learning_rate": 4.697332419717589e-06,
+      "loss": 0.6541,
+      "step": 3512
+    },
+    {
+      "epoch": 0.9731301939058171,
+      "grad_norm": 0.6031319499015808,
+      "learning_rate": 4.69715865243055e-06,
+      "loss": 0.6121,
+      "step": 3513
+    },
+    {
+      "epoch": 0.9734072022160665,
+      "grad_norm": 0.6461656093597412,
+      "learning_rate": 4.696984838492222e-06,
+      "loss": 0.7018,
+      "step": 3514
+    },
+    {
+      "epoch": 0.9736842105263158,
+      "grad_norm": 0.6950158476829529,
+      "learning_rate": 4.696810977906296e-06,
+      "loss": 0.6666,
+      "step": 3515
+    },
+    {
+      "epoch": 0.9739612188365651,
+      "grad_norm": 0.658356785774231,
+      "learning_rate": 4.696637070676462e-06,
+      "loss": 0.6518,
+      "step": 3516
+    },
+    {
+      "epoch": 0.9742382271468144,
+      "grad_norm": 0.7311317920684814,
+      "learning_rate": 4.696463116806414e-06,
+      "loss": 0.7112,
+      "step": 3517
+    },
+    {
+      "epoch": 0.9745152354570638,
+      "grad_norm": 0.6782693266868591,
+      "learning_rate": 4.696289116299844e-06,
+      "loss": 0.6726,
+      "step": 3518
+    },
+    {
+      "epoch": 0.974792243767313,
+      "grad_norm": 0.6792790293693542,
+      "learning_rate": 4.696115069160448e-06,
+      "loss": 0.648,
+      "step": 3519
+    },
+    {
+      "epoch": 0.9750692520775623,
+      "grad_norm": 0.7299467325210571,
+      "learning_rate": 4.695940975391919e-06,
+      "loss": 0.6764,
+      "step": 3520
+    },
+    {
+      "epoch": 0.9753462603878116,
+      "grad_norm": 0.6922098398208618,
+      "learning_rate": 4.695766834997957e-06,
+      "loss": 0.6705,
+      "step": 3521
+    },
+    {
+      "epoch": 0.9756232686980609,
+      "grad_norm": 0.6680386662483215,
+      "learning_rate": 4.695592647982258e-06,
+      "loss": 0.6282,
+      "step": 3522
+    },
+    {
+      "epoch": 0.9759002770083103,
+      "grad_norm": 0.6693275570869446,
+      "learning_rate": 4.695418414348519e-06,
+      "loss": 0.6997,
+      "step": 3523
+    },
+    {
+      "epoch": 0.9761772853185595,
+      "grad_norm": 0.6710549592971802,
+      "learning_rate": 4.695244134100442e-06,
+      "loss": 0.6322,
+      "step": 3524
+    },
+    {
+      "epoch": 0.9764542936288089,
+      "grad_norm": 0.6785879135131836,
+      "learning_rate": 4.695069807241724e-06,
+      "loss": 0.6538,
+      "step": 3525
+    },
+    {
+      "epoch": 0.9767313019390582,
+      "grad_norm": 0.6887544989585876,
+      "learning_rate": 4.694895433776071e-06,
+      "loss": 0.6663,
+      "step": 3526
+    },
+    {
+      "epoch": 0.9770083102493075,
+      "grad_norm": 0.7070327997207642,
+      "learning_rate": 4.694721013707181e-06,
+      "loss": 0.6693,
+      "step": 3527
+    },
+    {
+      "epoch": 0.9772853185595568,
+      "grad_norm": 0.6848759651184082,
+      "learning_rate": 4.6945465470387595e-06,
+      "loss": 0.6784,
+      "step": 3528
+    },
+    {
+      "epoch": 0.977562326869806,
+      "grad_norm": 0.6864799857139587,
+      "learning_rate": 4.69437203377451e-06,
+      "loss": 0.6472,
+      "step": 3529
+    },
+    {
+      "epoch": 0.9778393351800554,
+      "grad_norm": 0.65505051612854,
+      "learning_rate": 4.694197473918139e-06,
+      "loss": 0.6529,
+      "step": 3530
+    },
+    {
+      "epoch": 0.9781163434903047,
+      "grad_norm": 0.6809942126274109,
+      "learning_rate": 4.694022867473352e-06,
+      "loss": 0.669,
+      "step": 3531
+    },
+    {
+      "epoch": 0.978393351800554,
+      "grad_norm": 0.7042862176895142,
+      "learning_rate": 4.693848214443858e-06,
+      "loss": 0.6792,
+      "step": 3532
+    },
+    {
+      "epoch": 0.9786703601108033,
+      "grad_norm": 0.6662670373916626,
+      "learning_rate": 4.693673514833363e-06,
+      "loss": 0.6066,
+      "step": 3533
+    },
+    {
+      "epoch": 0.9789473684210527,
+      "grad_norm": 0.6451749205589294,
+      "learning_rate": 4.693498768645577e-06,
+      "loss": 0.6622,
+      "step": 3534
+    },
+    {
+      "epoch": 0.9792243767313019,
+      "grad_norm": 0.6562368273735046,
+      "learning_rate": 4.69332397588421e-06,
+      "loss": 0.6225,
+      "step": 3535
+    },
+    {
+      "epoch": 0.9795013850415513,
+      "grad_norm": 0.7036115527153015,
+      "learning_rate": 4.693149136552975e-06,
+      "loss": 0.6379,
+      "step": 3536
+    },
+    {
+      "epoch": 0.9797783933518005,
+      "grad_norm": 0.7075529098510742,
+      "learning_rate": 4.692974250655582e-06,
+      "loss": 0.6859,
+      "step": 3537
+    },
+    {
+      "epoch": 0.9800554016620499,
+      "grad_norm": 0.7317038774490356,
+      "learning_rate": 4.6927993181957456e-06,
+      "loss": 0.6833,
+      "step": 3538
+    },
+    {
+      "epoch": 0.9803324099722992,
+      "grad_norm": 0.6922315955162048,
+      "learning_rate": 4.692624339177181e-06,
+      "loss": 0.6971,
+      "step": 3539
+    },
+    {
+      "epoch": 0.9806094182825484,
+      "grad_norm": 0.6648818254470825,
+      "learning_rate": 4.692449313603601e-06,
+      "loss": 0.6506,
+      "step": 3540
+    },
+    {
+      "epoch": 0.9808864265927978,
+      "grad_norm": 0.6564773917198181,
+      "learning_rate": 4.6922742414787235e-06,
+      "loss": 0.6696,
+      "step": 3541
+    },
+    {
+      "epoch": 0.9811634349030471,
+      "grad_norm": 0.6901625990867615,
+      "learning_rate": 4.6920991228062664e-06,
+      "loss": 0.6548,
+      "step": 3542
+    },
+    {
+      "epoch": 0.9814404432132964,
+      "grad_norm": 0.6555216908454895,
+      "learning_rate": 4.691923957589946e-06,
+      "loss": 0.6035,
+      "step": 3543
+    },
+    {
+      "epoch": 0.9817174515235457,
+      "grad_norm": 0.6624067425727844,
+      "learning_rate": 4.6917487458334826e-06,
+      "loss": 0.6725,
+      "step": 3544
+    },
+    {
+      "epoch": 0.981994459833795,
+      "grad_norm": 0.691081702709198,
+      "learning_rate": 4.691573487540596e-06,
+      "loss": 0.6944,
+      "step": 3545
+    },
+    {
+      "epoch": 0.9822714681440443,
+      "grad_norm": 0.6511529684066772,
+      "learning_rate": 4.691398182715008e-06,
+      "loss": 0.661,
+      "step": 3546
+    },
+    {
+      "epoch": 0.9825484764542937,
+      "grad_norm": 0.6528334617614746,
+      "learning_rate": 4.69122283136044e-06,
+      "loss": 0.6942,
+      "step": 3547
+    },
+    {
+      "epoch": 0.9828254847645429,
+      "grad_norm": 0.6623634696006775,
+      "learning_rate": 4.691047433480616e-06,
+      "loss": 0.6701,
+      "step": 3548
+    },
+    {
+      "epoch": 0.9831024930747922,
+      "grad_norm": 0.6403034329414368,
+      "learning_rate": 4.69087198907926e-06,
+      "loss": 0.6517,
+      "step": 3549
+    },
+    {
+      "epoch": 0.9833795013850416,
+      "grad_norm": 0.6423508524894714,
+      "learning_rate": 4.690696498160096e-06,
+      "loss": 0.6115,
+      "step": 3550
+    },
+    {
+      "epoch": 0.9836565096952908,
+      "grad_norm": 0.6804136037826538,
+      "learning_rate": 4.69052096072685e-06,
+      "loss": 0.699,
+      "step": 3551
+    },
+    {
+      "epoch": 0.9839335180055402,
+      "grad_norm": 0.6722675561904907,
+      "learning_rate": 4.6903453767832514e-06,
+      "loss": 0.667,
+      "step": 3552
+    },
+    {
+      "epoch": 0.9842105263157894,
+      "grad_norm": 0.6349672675132751,
+      "learning_rate": 4.690169746333027e-06,
+      "loss": 0.6409,
+      "step": 3553
+    },
+    {
+      "epoch": 0.9844875346260388,
+      "grad_norm": 0.6569719910621643,
+      "learning_rate": 4.689994069379905e-06,
+      "loss": 0.5844,
+      "step": 3554
+    },
+    {
+      "epoch": 0.9847645429362881,
+      "grad_norm": 0.6513596773147583,
+      "learning_rate": 4.689818345927617e-06,
+      "loss": 0.6613,
+      "step": 3555
+    },
+    {
+      "epoch": 0.9850415512465374,
+      "grad_norm": 0.6963537931442261,
+      "learning_rate": 4.6896425759798925e-06,
+      "loss": 0.6691,
+      "step": 3556
+    },
+    {
+      "epoch": 0.9853185595567867,
+      "grad_norm": 0.6589019894599915,
+      "learning_rate": 4.6894667595404654e-06,
+      "loss": 0.6759,
+      "step": 3557
+    },
+    {
+      "epoch": 0.9855955678670361,
+      "grad_norm": 0.711513876914978,
+      "learning_rate": 4.6892908966130665e-06,
+      "loss": 0.6919,
+      "step": 3558
+    },
+    {
+      "epoch": 0.9858725761772853,
+      "grad_norm": 0.6605263352394104,
+      "learning_rate": 4.689114987201432e-06,
+      "loss": 0.6374,
+      "step": 3559
+    },
+    {
+      "epoch": 0.9861495844875346,
+      "grad_norm": 0.6809438467025757,
+      "learning_rate": 4.688939031309296e-06,
+      "loss": 0.7072,
+      "step": 3560
+    },
+    {
+      "epoch": 0.9864265927977839,
+      "grad_norm": 0.6980892419815063,
+      "learning_rate": 4.688763028940394e-06,
+      "loss": 0.6791,
+      "step": 3561
+    },
+    {
+      "epoch": 0.9867036011080332,
+      "grad_norm": 0.6453639268875122,
+      "learning_rate": 4.6885869800984634e-06,
+      "loss": 0.6198,
+      "step": 3562
+    },
+    {
+      "epoch": 0.9869806094182826,
+      "grad_norm": 0.7025441527366638,
+      "learning_rate": 4.6884108847872425e-06,
+      "loss": 0.6646,
+      "step": 3563
+    },
+    {
+      "epoch": 0.9872576177285318,
+      "grad_norm": 0.6834899187088013,
+      "learning_rate": 4.688234743010469e-06,
+      "loss": 0.6535,
+      "step": 3564
+    },
+    {
+      "epoch": 0.9875346260387812,
+      "grad_norm": 0.6900626420974731,
+      "learning_rate": 4.6880585547718845e-06,
+      "loss": 0.6207,
+      "step": 3565
+    },
+    {
+      "epoch": 0.9878116343490305,
+      "grad_norm": 0.683000385761261,
+      "learning_rate": 4.687882320075229e-06,
+      "loss": 0.6541,
+      "step": 3566
+    },
+    {
+      "epoch": 0.9880886426592798,
+      "grad_norm": 0.6564151048660278,
+      "learning_rate": 4.687706038924244e-06,
+      "loss": 0.636,
+      "step": 3567
+    },
+    {
+      "epoch": 0.9883656509695291,
+      "grad_norm": 0.6594918370246887,
+      "learning_rate": 4.687529711322675e-06,
+      "loss": 0.6138,
+      "step": 3568
+    },
+    {
+      "epoch": 0.9886426592797783,
+      "grad_norm": 0.6680994629859924,
+      "learning_rate": 4.687353337274262e-06,
+      "loss": 0.6581,
+      "step": 3569
+    },
+    {
+      "epoch": 0.9889196675900277,
+      "grad_norm": 0.6732953786849976,
+      "learning_rate": 4.687176916782753e-06,
+      "loss": 0.6795,
+      "step": 3570
+    },
+    {
+      "epoch": 0.989196675900277,
+      "grad_norm": 0.679394006729126,
+      "learning_rate": 4.687000449851892e-06,
+      "loss": 0.6376,
+      "step": 3571
+    },
+    {
+      "epoch": 0.9894736842105263,
+      "grad_norm": 0.7351693511009216,
+      "learning_rate": 4.686823936485426e-06,
+      "loss": 0.6874,
+      "step": 3572
+    },
+    {
+      "epoch": 0.9897506925207756,
+      "grad_norm": 0.7224422693252563,
+      "learning_rate": 4.686647376687104e-06,
+      "loss": 0.6522,
+      "step": 3573
+    },
+    {
+      "epoch": 0.990027700831025,
+      "grad_norm": 0.6476606726646423,
+      "learning_rate": 4.686470770460674e-06,
+      "loss": 0.6529,
+      "step": 3574
+    },
+    {
+      "epoch": 0.9903047091412742,
+      "grad_norm": 0.6556124091148376,
+      "learning_rate": 4.6862941178098855e-06,
+      "loss": 0.5976,
+      "step": 3575
+    },
+    {
+      "epoch": 0.9905817174515236,
+      "grad_norm": 0.6533079147338867,
+      "learning_rate": 4.686117418738489e-06,
+      "loss": 0.6716,
+      "step": 3576
+    },
+    {
+      "epoch": 0.9908587257617728,
+      "grad_norm": 0.6209880709648132,
+      "learning_rate": 4.685940673250239e-06,
+      "loss": 0.5748,
+      "step": 3577
+    },
+    {
+      "epoch": 0.9911357340720222,
+      "grad_norm": 0.6872321367263794,
+      "learning_rate": 4.685763881348885e-06,
+      "loss": 0.6536,
+      "step": 3578
+    },
+    {
+      "epoch": 0.9914127423822715,
+      "grad_norm": 0.7070121169090271,
+      "learning_rate": 4.685587043038182e-06,
+      "loss": 0.6236,
+      "step": 3579
+    },
+    {
+      "epoch": 0.9916897506925207,
+      "grad_norm": 0.6352293491363525,
+      "learning_rate": 4.685410158321885e-06,
+      "loss": 0.6185,
+      "step": 3580
+    },
+    {
+      "epoch": 0.9919667590027701,
+      "grad_norm": 0.6836139559745789,
+      "learning_rate": 4.685233227203749e-06,
+      "loss": 0.6413,
+      "step": 3581
+    },
+    {
+      "epoch": 0.9922437673130194,
+      "grad_norm": 0.6847363710403442,
+      "learning_rate": 4.685056249687531e-06,
+      "loss": 0.671,
+      "step": 3582
+    },
+    {
+      "epoch": 0.9925207756232687,
+      "grad_norm": 0.6621556282043457,
+      "learning_rate": 4.684879225776989e-06,
+      "loss": 0.6815,
+      "step": 3583
+    },
+    {
+      "epoch": 0.992797783933518,
+      "grad_norm": 0.6949865221977234,
+      "learning_rate": 4.684702155475882e-06,
+      "loss": 0.6364,
+      "step": 3584
+    },
+    {
+      "epoch": 0.9930747922437673,
+      "grad_norm": 0.6503170728683472,
+      "learning_rate": 4.684525038787968e-06,
+      "loss": 0.6872,
+      "step": 3585
+    },
+    {
+      "epoch": 0.9933518005540166,
+      "grad_norm": 0.6437188386917114,
+      "learning_rate": 4.6843478757170095e-06,
+      "loss": 0.6351,
+      "step": 3586
+    },
+    {
+      "epoch": 0.993628808864266,
+      "grad_norm": 0.6677717566490173,
+      "learning_rate": 4.684170666266767e-06,
+      "loss": 0.7332,
+      "step": 3587
+    },
+    {
+      "epoch": 0.9939058171745152,
+      "grad_norm": 0.6620640754699707,
+      "learning_rate": 4.683993410441004e-06,
+      "loss": 0.6612,
+      "step": 3588
+    },
+    {
+      "epoch": 0.9941828254847646,
+      "grad_norm": 0.6833710074424744,
+      "learning_rate": 4.683816108243484e-06,
+      "loss": 0.6662,
+      "step": 3589
+    },
+    {
+      "epoch": 0.9944598337950139,
+      "grad_norm": 0.7151123881340027,
+      "learning_rate": 4.68363875967797e-06,
+      "loss": 0.6797,
+      "step": 3590
+    },
+    {
+      "epoch": 0.9947368421052631,
+      "grad_norm": 0.7432909607887268,
+      "learning_rate": 4.683461364748229e-06,
+      "loss": 0.6629,
+      "step": 3591
+    },
+    {
+      "epoch": 0.9950138504155125,
+      "grad_norm": 0.6998159289360046,
+      "learning_rate": 4.683283923458028e-06,
+      "loss": 0.6761,
+      "step": 3592
+    },
+    {
+      "epoch": 0.9952908587257617,
+      "grad_norm": 0.670008659362793,
+      "learning_rate": 4.683106435811133e-06,
+      "loss": 0.6579,
+      "step": 3593
+    },
+    {
+      "epoch": 0.9955678670360111,
+      "grad_norm": 0.6878396272659302,
+      "learning_rate": 4.682928901811315e-06,
+      "loss": 0.7188,
+      "step": 3594
+    },
+    {
+      "epoch": 0.9958448753462604,
+      "grad_norm": 0.7093462347984314,
+      "learning_rate": 4.68275132146234e-06,
+      "loss": 0.6486,
+      "step": 3595
+    },
+    {
+      "epoch": 0.9961218836565097,
+      "grad_norm": 0.7437435388565063,
+      "learning_rate": 4.68257369476798e-06,
+      "loss": 0.6911,
+      "step": 3596
+    },
+    {
+      "epoch": 0.996398891966759,
+      "grad_norm": 0.6841958165168762,
+      "learning_rate": 4.682396021732008e-06,
+      "loss": 0.6565,
+      "step": 3597
+    },
+    {
+      "epoch": 0.9966759002770084,
+      "grad_norm": 0.6469076871871948,
+      "learning_rate": 4.6822183023581945e-06,
+      "loss": 0.713,
+      "step": 3598
+    },
+    {
+      "epoch": 0.9969529085872576,
+      "grad_norm": 0.6475204229354858,
+      "learning_rate": 4.6820405366503145e-06,
+      "loss": 0.6907,
+      "step": 3599
+    },
+    {
+      "epoch": 0.997229916897507,
+      "grad_norm": 0.7169883251190186,
+      "learning_rate": 4.6818627246121404e-06,
+      "loss": 0.6636,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9975069252077562,
+      "grad_norm": 0.6753641366958618,
+      "learning_rate": 4.68168486624745e-06,
+      "loss": 0.6268,
+      "step": 3601
+    },
+    {
+      "epoch": 0.9977839335180055,
+      "grad_norm": 0.6558805108070374,
+      "learning_rate": 4.681506961560018e-06,
+      "loss": 0.682,
+      "step": 3602
+    },
+    {
+      "epoch": 0.9980609418282549,
+      "grad_norm": 0.7053347229957581,
+      "learning_rate": 4.681329010553622e-06,
+      "loss": 0.7016,
+      "step": 3603
+    },
+    {
+      "epoch": 0.9983379501385041,
+      "grad_norm": 0.6714473366737366,
+      "learning_rate": 4.681151013232041e-06,
+      "loss": 0.6639,
+      "step": 3604
+    },
+    {
+      "epoch": 0.9986149584487535,
+      "grad_norm": 0.6419603824615479,
+      "learning_rate": 4.680972969599053e-06,
+      "loss": 0.6326,
+      "step": 3605
+    },
+    {
+      "epoch": 0.9988919667590028,
+      "grad_norm": 0.6689679622650146,
+      "learning_rate": 4.68079487965844e-06,
+      "loss": 0.651,
+      "step": 3606
+    },
+    {
+      "epoch": 0.9991689750692521,
+      "grad_norm": 0.6623345613479614,
+      "learning_rate": 4.680616743413982e-06,
+      "loss": 0.6543,
+      "step": 3607
+    },
+    {
+      "epoch": 0.9994459833795014,
+      "grad_norm": 0.6692014932632446,
+      "learning_rate": 4.680438560869463e-06,
+      "loss": 0.6411,
+      "step": 3608
+    },
+    {
+      "epoch": 0.9997229916897507,
+      "grad_norm": 0.7073262333869934,
+      "learning_rate": 4.680260332028664e-06,
+      "loss": 0.6296,
+      "step": 3609
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.698577880859375,
+      "learning_rate": 4.680082056895371e-06,
+      "loss": 0.6524,
+      "step": 3610
+    },
+    {
+      "epoch": 1.0002770083102492,
+      "grad_norm": 0.6731170415878296,
+      "learning_rate": 4.679903735473369e-06,
+      "loss": 0.6464,
+      "step": 3611
+    },
+    {
+      "epoch": 1.0005540166204987,
+      "grad_norm": 0.6520345211029053,
+      "learning_rate": 4.679725367766443e-06,
+      "loss": 0.6776,
+      "step": 3612
+    },
+    {
+      "epoch": 1.000831024930748,
+      "grad_norm": 0.6469970941543579,
+      "learning_rate": 4.679546953778381e-06,
+      "loss": 0.6336,
+      "step": 3613
+    },
+    {
+      "epoch": 1.0011080332409972,
+      "grad_norm": 0.627105176448822,
+      "learning_rate": 4.679368493512972e-06,
+      "loss": 0.6445,
+      "step": 3614
+    },
+    {
+      "epoch": 1.0013850415512466,
+      "grad_norm": 0.6837867498397827,
+      "learning_rate": 4.679189986974004e-06,
+      "loss": 0.6986,
+      "step": 3615
+    },
+    {
+      "epoch": 1.0016620498614959,
+      "grad_norm": 0.6424873471260071,
+      "learning_rate": 4.679011434165267e-06,
+      "loss": 0.5822,
+      "step": 3616
+    },
+    {
+      "epoch": 1.0019390581717451,
+      "grad_norm": 0.6437291502952576,
+      "learning_rate": 4.6788328350905535e-06,
+      "loss": 0.6721,
+      "step": 3617
+    },
+    {
+      "epoch": 1.0022160664819943,
+      "grad_norm": 0.6488319635391235,
+      "learning_rate": 4.678654189753654e-06,
+      "loss": 0.6695,
+      "step": 3618
+    },
+    {
+      "epoch": 1.0024930747922438,
+      "grad_norm": 0.7707616090774536,
+      "learning_rate": 4.678475498158363e-06,
+      "loss": 0.6881,
+      "step": 3619
+    },
+    {
+      "epoch": 1.002770083102493,
+      "grad_norm": 0.6751238703727722,
+      "learning_rate": 4.678296760308474e-06,
+      "loss": 0.5906,
+      "step": 3620
+    },
+    {
+      "epoch": 1.0030470914127423,
+      "grad_norm": 0.6540700793266296,
+      "learning_rate": 4.678117976207781e-06,
+      "loss": 0.5704,
+      "step": 3621
+    },
+    {
+      "epoch": 1.0033240997229917,
+      "grad_norm": 0.6810097694396973,
+      "learning_rate": 4.677939145860082e-06,
+      "loss": 0.6235,
+      "step": 3622
+    },
+    {
+      "epoch": 1.003601108033241,
+      "grad_norm": 0.6593797206878662,
+      "learning_rate": 4.677760269269173e-06,
+      "loss": 0.6184,
+      "step": 3623
+    },
+    {
+      "epoch": 1.0038781163434902,
+      "grad_norm": 0.6637284755706787,
+      "learning_rate": 4.677581346438852e-06,
+      "loss": 0.6539,
+      "step": 3624
+    },
+    {
+      "epoch": 1.0041551246537397,
+      "grad_norm": 0.6583884954452515,
+      "learning_rate": 4.677402377372918e-06,
+      "loss": 0.5725,
+      "step": 3625
+    },
+    {
+      "epoch": 1.004432132963989,
+      "grad_norm": 0.7044674754142761,
+      "learning_rate": 4.677223362075171e-06,
+      "loss": 0.7045,
+      "step": 3626
+    },
+    {
+      "epoch": 1.0047091412742382,
+      "grad_norm": 0.7017073631286621,
+      "learning_rate": 4.677044300549412e-06,
+      "loss": 0.6892,
+      "step": 3627
+    },
+    {
+      "epoch": 1.0049861495844876,
+      "grad_norm": 0.6834173798561096,
+      "learning_rate": 4.676865192799443e-06,
+      "loss": 0.5998,
+      "step": 3628
+    },
+    {
+      "epoch": 1.0052631578947369,
+      "grad_norm": 0.694014310836792,
+      "learning_rate": 4.6766860388290665e-06,
+      "loss": 0.6685,
+      "step": 3629
+    },
+    {
+      "epoch": 1.005540166204986,
+      "grad_norm": 0.7017285823822021,
+      "learning_rate": 4.676506838642088e-06,
+      "loss": 0.6742,
+      "step": 3630
+    },
+    {
+      "epoch": 1.0058171745152356,
+      "grad_norm": 0.6789602041244507,
+      "learning_rate": 4.6763275922423105e-06,
+      "loss": 0.732,
+      "step": 3631
+    },
+    {
+      "epoch": 1.0060941828254848,
+      "grad_norm": 0.724633514881134,
+      "learning_rate": 4.67614829963354e-06,
+      "loss": 0.6794,
+      "step": 3632
+    },
+    {
+      "epoch": 1.006371191135734,
+      "grad_norm": 0.6596531867980957,
+      "learning_rate": 4.675968960819584e-06,
+      "loss": 0.6641,
+      "step": 3633
+    },
+    {
+      "epoch": 1.0066481994459833,
+      "grad_norm": 0.7008945345878601,
+      "learning_rate": 4.67578957580425e-06,
+      "loss": 0.6609,
+      "step": 3634
+    },
+    {
+      "epoch": 1.0069252077562327,
+      "grad_norm": 0.662340521812439,
+      "learning_rate": 4.675610144591347e-06,
+      "loss": 0.6137,
+      "step": 3635
+    },
+    {
+      "epoch": 1.007202216066482,
+      "grad_norm": 0.703514575958252,
+      "learning_rate": 4.675430667184685e-06,
+      "loss": 0.61,
+      "step": 3636
+    },
+    {
+      "epoch": 1.0074792243767312,
+      "grad_norm": 0.675926148891449,
+      "learning_rate": 4.675251143588074e-06,
+      "loss": 0.6713,
+      "step": 3637
+    },
+    {
+      "epoch": 1.0077562326869807,
+      "grad_norm": 0.6686869263648987,
+      "learning_rate": 4.675071573805327e-06,
+      "loss": 0.6487,
+      "step": 3638
+    },
+    {
+      "epoch": 1.00803324099723,
+      "grad_norm": 0.7041005492210388,
+      "learning_rate": 4.674891957840255e-06,
+      "loss": 0.6618,
+      "step": 3639
+    },
+    {
+      "epoch": 1.0083102493074791,
+      "grad_norm": 0.7006294131278992,
+      "learning_rate": 4.674712295696674e-06,
+      "loss": 0.6877,
+      "step": 3640
+    },
+    {
+      "epoch": 1.0085872576177286,
+      "grad_norm": 0.6780726909637451,
+      "learning_rate": 4.674532587378396e-06,
+      "loss": 0.6457,
+      "step": 3641
+    },
+    {
+      "epoch": 1.0088642659279778,
+      "grad_norm": 0.6753490567207336,
+      "learning_rate": 4.674352832889239e-06,
+      "loss": 0.6326,
+      "step": 3642
+    },
+    {
+      "epoch": 1.009141274238227,
+      "grad_norm": 0.7531911730766296,
+      "learning_rate": 4.674173032233018e-06,
+      "loss": 0.6644,
+      "step": 3643
+    },
+    {
+      "epoch": 1.0094182825484765,
+      "grad_norm": 0.6730330586433411,
+      "learning_rate": 4.673993185413552e-06,
+      "loss": 0.6329,
+      "step": 3644
+    },
+    {
+      "epoch": 1.0096952908587258,
+      "grad_norm": 0.7184603214263916,
+      "learning_rate": 4.673813292434658e-06,
+      "loss": 0.6643,
+      "step": 3645
+    },
+    {
+      "epoch": 1.009972299168975,
+      "grad_norm": 0.7501634359359741,
+      "learning_rate": 4.673633353300157e-06,
+      "loss": 0.6286,
+      "step": 3646
+    },
+    {
+      "epoch": 1.0102493074792245,
+      "grad_norm": 0.7249453067779541,
+      "learning_rate": 4.67345336801387e-06,
+      "loss": 0.6791,
+      "step": 3647
+    },
+    {
+      "epoch": 1.0105263157894737,
+      "grad_norm": 0.6539937853813171,
+      "learning_rate": 4.673273336579616e-06,
+      "loss": 0.583,
+      "step": 3648
+    },
+    {
+      "epoch": 1.010803324099723,
+      "grad_norm": 0.6474573016166687,
+      "learning_rate": 4.673093259001219e-06,
+      "loss": 0.6132,
+      "step": 3649
+    },
+    {
+      "epoch": 1.0110803324099722,
+      "grad_norm": 0.684526264667511,
+      "learning_rate": 4.672913135282504e-06,
+      "loss": 0.5963,
+      "step": 3650
+    },
+    {
+      "epoch": 1.0113573407202217,
+      "grad_norm": 0.7160072922706604,
+      "learning_rate": 4.672732965427294e-06,
+      "loss": 0.6651,
+      "step": 3651
+    },
+    {
+      "epoch": 1.011634349030471,
+      "grad_norm": 0.6646113991737366,
+      "learning_rate": 4.672552749439413e-06,
+      "loss": 0.6264,
+      "step": 3652
+    },
+    {
+      "epoch": 1.0119113573407201,
+      "grad_norm": 0.6403873562812805,
+      "learning_rate": 4.672372487322691e-06,
+      "loss": 0.6132,
+      "step": 3653
+    },
+    {
+      "epoch": 1.0121883656509696,
+      "grad_norm": 0.6751793622970581,
+      "learning_rate": 4.672192179080952e-06,
+      "loss": 0.6898,
+      "step": 3654
+    },
+    {
+      "epoch": 1.0124653739612188,
+      "grad_norm": 0.7224442958831787,
+      "learning_rate": 4.6720118247180265e-06,
+      "loss": 0.6685,
+      "step": 3655
+    },
+    {
+      "epoch": 1.012742382271468,
+      "grad_norm": 0.7179122567176819,
+      "learning_rate": 4.671831424237743e-06,
+      "loss": 0.6731,
+      "step": 3656
+    },
+    {
+      "epoch": 1.0130193905817175,
+      "grad_norm": 0.6969308853149414,
+      "learning_rate": 4.671650977643932e-06,
+      "loss": 0.6385,
+      "step": 3657
+    },
+    {
+      "epoch": 1.0132963988919668,
+      "grad_norm": 0.7241250872612,
+      "learning_rate": 4.671470484940426e-06,
+      "loss": 0.6828,
+      "step": 3658
+    },
+    {
+      "epoch": 1.013573407202216,
+      "grad_norm": 0.6995605230331421,
+      "learning_rate": 4.671289946131055e-06,
+      "loss": 0.6591,
+      "step": 3659
+    },
+    {
+      "epoch": 1.0138504155124655,
+      "grad_norm": 0.6992942690849304,
+      "learning_rate": 4.671109361219655e-06,
+      "loss": 0.6478,
+      "step": 3660
+    },
+    {
+      "epoch": 1.0141274238227147,
+      "grad_norm": 0.6783244609832764,
+      "learning_rate": 4.670928730210057e-06,
+      "loss": 0.6515,
+      "step": 3661
+    },
+    {
+      "epoch": 1.014404432132964,
+      "grad_norm": 0.6964882016181946,
+      "learning_rate": 4.670748053106099e-06,
+      "loss": 0.6639,
+      "step": 3662
+    },
+    {
+      "epoch": 1.0146814404432134,
+      "grad_norm": 0.7556734681129456,
+      "learning_rate": 4.670567329911616e-06,
+      "loss": 0.6472,
+      "step": 3663
+    },
+    {
+      "epoch": 1.0149584487534626,
+      "grad_norm": 0.6755543351173401,
+      "learning_rate": 4.670386560630446e-06,
+      "loss": 0.6372,
+      "step": 3664
+    },
+    {
+      "epoch": 1.0152354570637119,
+      "grad_norm": 0.6663053631782532,
+      "learning_rate": 4.670205745266427e-06,
+      "loss": 0.6537,
+      "step": 3665
+    },
+    {
+      "epoch": 1.0155124653739611,
+      "grad_norm": 0.6746417880058289,
+      "learning_rate": 4.670024883823397e-06,
+      "loss": 0.6397,
+      "step": 3666
+    },
+    {
+      "epoch": 1.0157894736842106,
+      "grad_norm": 0.6543819904327393,
+      "learning_rate": 4.669843976305199e-06,
+      "loss": 0.6905,
+      "step": 3667
+    },
+    {
+      "epoch": 1.0160664819944598,
+      "grad_norm": 0.6693983674049377,
+      "learning_rate": 4.66966302271567e-06,
+      "loss": 0.6429,
+      "step": 3668
+    },
+    {
+      "epoch": 1.016343490304709,
+      "grad_norm": 0.6612565517425537,
+      "learning_rate": 4.669482023058656e-06,
+      "loss": 0.6052,
+      "step": 3669
+    },
+    {
+      "epoch": 1.0166204986149585,
+      "grad_norm": 0.6457585096359253,
+      "learning_rate": 4.6693009773379975e-06,
+      "loss": 0.6464,
+      "step": 3670
+    },
+    {
+      "epoch": 1.0168975069252078,
+      "grad_norm": 0.691620945930481,
+      "learning_rate": 4.669119885557539e-06,
+      "loss": 0.6357,
+      "step": 3671
+    },
+    {
+      "epoch": 1.017174515235457,
+      "grad_norm": 0.6583114862442017,
+      "learning_rate": 4.668938747721127e-06,
+      "loss": 0.6049,
+      "step": 3672
+    },
+    {
+      "epoch": 1.0174515235457064,
+      "grad_norm": 0.7024505734443665,
+      "learning_rate": 4.668757563832606e-06,
+      "loss": 0.6073,
+      "step": 3673
+    },
+    {
+      "epoch": 1.0177285318559557,
+      "grad_norm": 0.7239954471588135,
+      "learning_rate": 4.668576333895824e-06,
+      "loss": 0.672,
+      "step": 3674
+    },
+    {
+      "epoch": 1.018005540166205,
+      "grad_norm": 0.7404740452766418,
+      "learning_rate": 4.668395057914627e-06,
+      "loss": 0.671,
+      "step": 3675
+    },
+    {
+      "epoch": 1.0182825484764544,
+      "grad_norm": 0.692775547504425,
+      "learning_rate": 4.668213735892867e-06,
+      "loss": 0.6503,
+      "step": 3676
+    },
+    {
+      "epoch": 1.0185595567867036,
+      "grad_norm": 0.7027653455734253,
+      "learning_rate": 4.668032367834392e-06,
+      "loss": 0.6527,
+      "step": 3677
+    },
+    {
+      "epoch": 1.0188365650969529,
+      "grad_norm": 0.6956178545951843,
+      "learning_rate": 4.667850953743053e-06,
+      "loss": 0.6622,
+      "step": 3678
+    },
+    {
+      "epoch": 1.0191135734072023,
+      "grad_norm": 0.6842972040176392,
+      "learning_rate": 4.667669493622703e-06,
+      "loss": 0.6211,
+      "step": 3679
+    },
+    {
+      "epoch": 1.0193905817174516,
+      "grad_norm": 0.6833047866821289,
+      "learning_rate": 4.6674879874771925e-06,
+      "loss": 0.6739,
+      "step": 3680
+    },
+    {
+      "epoch": 1.0196675900277008,
+      "grad_norm": 0.7252957224845886,
+      "learning_rate": 4.667306435310379e-06,
+      "loss": 0.6258,
+      "step": 3681
+    },
+    {
+      "epoch": 1.01994459833795,
+      "grad_norm": 0.6874592900276184,
+      "learning_rate": 4.667124837126114e-06,
+      "loss": 0.6801,
+      "step": 3682
+    },
+    {
+      "epoch": 1.0202216066481995,
+      "grad_norm": 0.8430461883544922,
+      "learning_rate": 4.666943192928254e-06,
+      "loss": 0.6102,
+      "step": 3683
+    },
+    {
+      "epoch": 1.0204986149584487,
+      "grad_norm": 0.7195748090744019,
+      "learning_rate": 4.666761502720657e-06,
+      "loss": 0.6824,
+      "step": 3684
+    },
+    {
+      "epoch": 1.020775623268698,
+      "grad_norm": 0.6784130930900574,
+      "learning_rate": 4.666579766507179e-06,
+      "loss": 0.6756,
+      "step": 3685
+    },
+    {
+      "epoch": 1.0210526315789474,
+      "grad_norm": 0.7014387845993042,
+      "learning_rate": 4.66639798429168e-06,
+      "loss": 0.6086,
+      "step": 3686
+    },
+    {
+      "epoch": 1.0213296398891967,
+      "grad_norm": 0.6831591129302979,
+      "learning_rate": 4.66621615607802e-06,
+      "loss": 0.6833,
+      "step": 3687
+    },
+    {
+      "epoch": 1.021606648199446,
+      "grad_norm": 0.6831374168395996,
+      "learning_rate": 4.6660342818700596e-06,
+      "loss": 0.6541,
+      "step": 3688
+    },
+    {
+      "epoch": 1.0218836565096954,
+      "grad_norm": 0.6437398791313171,
+      "learning_rate": 4.665852361671659e-06,
+      "loss": 0.6125,
+      "step": 3689
+    },
+    {
+      "epoch": 1.0221606648199446,
+      "grad_norm": 0.670684278011322,
+      "learning_rate": 4.665670395486683e-06,
+      "loss": 0.6064,
+      "step": 3690
+    },
+    {
+      "epoch": 1.0224376731301938,
+      "grad_norm": 0.6811419725418091,
+      "learning_rate": 4.665488383318992e-06,
+      "loss": 0.6238,
+      "step": 3691
+    },
+    {
+      "epoch": 1.0227146814404433,
+      "grad_norm": 0.690031111240387,
+      "learning_rate": 4.665306325172453e-06,
+      "loss": 0.6418,
+      "step": 3692
+    },
+    {
+      "epoch": 1.0229916897506925,
+      "grad_norm": 0.6779752969741821,
+      "learning_rate": 4.665124221050931e-06,
+      "loss": 0.6248,
+      "step": 3693
+    },
+    {
+      "epoch": 1.0232686980609418,
+      "grad_norm": 0.726576030254364,
+      "learning_rate": 4.6649420709582945e-06,
+      "loss": 0.6345,
+      "step": 3694
+    },
+    {
+      "epoch": 1.0235457063711912,
+      "grad_norm": 0.7005026340484619,
+      "learning_rate": 4.664759874898408e-06,
+      "loss": 0.6493,
+      "step": 3695
+    },
+    {
+      "epoch": 1.0238227146814405,
+      "grad_norm": 0.7161099314689636,
+      "learning_rate": 4.664577632875141e-06,
+      "loss": 0.6785,
+      "step": 3696
+    },
+    {
+      "epoch": 1.0240997229916897,
+      "grad_norm": 0.6901417970657349,
+      "learning_rate": 4.664395344892363e-06,
+      "loss": 0.6378,
+      "step": 3697
+    },
+    {
+      "epoch": 1.024376731301939,
+      "grad_norm": 0.687687337398529,
+      "learning_rate": 4.664213010953945e-06,
+      "loss": 0.6133,
+      "step": 3698
+    },
+    {
+      "epoch": 1.0246537396121884,
+      "grad_norm": 0.7225800156593323,
+      "learning_rate": 4.664030631063758e-06,
+      "loss": 0.6473,
+      "step": 3699
+    },
+    {
+      "epoch": 1.0249307479224377,
+      "grad_norm": 0.6865379214286804,
+      "learning_rate": 4.663848205225674e-06,
+      "loss": 0.6705,
+      "step": 3700
+    },
+    {
+      "epoch": 1.025207756232687,
+      "grad_norm": 0.6914872527122498,
+      "learning_rate": 4.6636657334435675e-06,
+      "loss": 0.6339,
+      "step": 3701
+    },
+    {
+      "epoch": 1.0254847645429364,
+      "grad_norm": 0.687730073928833,
+      "learning_rate": 4.663483215721311e-06,
+      "loss": 0.6886,
+      "step": 3702
+    },
+    {
+      "epoch": 1.0257617728531856,
+      "grad_norm": 0.6916108727455139,
+      "learning_rate": 4.663300652062782e-06,
+      "loss": 0.6786,
+      "step": 3703
+    },
+    {
+      "epoch": 1.0260387811634348,
+      "grad_norm": 0.7067069411277771,
+      "learning_rate": 4.663118042471854e-06,
+      "loss": 0.657,
+      "step": 3704
+    },
+    {
+      "epoch": 1.0263157894736843,
+      "grad_norm": 0.7158690094947815,
+      "learning_rate": 4.662935386952408e-06,
+      "loss": 0.5958,
+      "step": 3705
+    },
+    {
+      "epoch": 1.0265927977839335,
+      "grad_norm": 0.72146075963974,
+      "learning_rate": 4.6627526855083196e-06,
+      "loss": 0.6631,
+      "step": 3706
+    },
+    {
+      "epoch": 1.0268698060941828,
+      "grad_norm": 0.685907244682312,
+      "learning_rate": 4.662569938143468e-06,
+      "loss": 0.6688,
+      "step": 3707
+    },
+    {
+      "epoch": 1.0271468144044322,
+      "grad_norm": 0.6628360748291016,
+      "learning_rate": 4.6623871448617345e-06,
+      "loss": 0.6427,
+      "step": 3708
+    },
+    {
+      "epoch": 1.0274238227146815,
+      "grad_norm": 0.7283874154090881,
+      "learning_rate": 4.662204305666999e-06,
+      "loss": 0.595,
+      "step": 3709
+    },
+    {
+      "epoch": 1.0277008310249307,
+      "grad_norm": 0.7178004384040833,
+      "learning_rate": 4.662021420563145e-06,
+      "loss": 0.6544,
+      "step": 3710
+    },
+    {
+      "epoch": 1.02797783933518,
+      "grad_norm": 0.6764352321624756,
+      "learning_rate": 4.6618384895540554e-06,
+      "loss": 0.5956,
+      "step": 3711
+    },
+    {
+      "epoch": 1.0282548476454294,
+      "grad_norm": 0.7078940868377686,
+      "learning_rate": 4.661655512643614e-06,
+      "loss": 0.6723,
+      "step": 3712
+    },
+    {
+      "epoch": 1.0285318559556786,
+      "grad_norm": 0.7133868336677551,
+      "learning_rate": 4.661472489835705e-06,
+      "loss": 0.6424,
+      "step": 3713
+    },
+    {
+      "epoch": 1.0288088642659279,
+      "grad_norm": 0.6933633685112,
+      "learning_rate": 4.661289421134215e-06,
+      "loss": 0.634,
+      "step": 3714
+    },
+    {
+      "epoch": 1.0290858725761773,
+      "grad_norm": 0.6836314797401428,
+      "learning_rate": 4.661106306543033e-06,
+      "loss": 0.6222,
+      "step": 3715
+    },
+    {
+      "epoch": 1.0293628808864266,
+      "grad_norm": 0.6919975876808167,
+      "learning_rate": 4.660923146066044e-06,
+      "loss": 0.6397,
+      "step": 3716
+    },
+    {
+      "epoch": 1.0296398891966758,
+      "grad_norm": 0.6854752898216248,
+      "learning_rate": 4.660739939707138e-06,
+      "loss": 0.6793,
+      "step": 3717
+    },
+    {
+      "epoch": 1.0299168975069253,
+      "grad_norm": 0.6768549680709839,
+      "learning_rate": 4.660556687470206e-06,
+      "loss": 0.6808,
+      "step": 3718
+    },
+    {
+      "epoch": 1.0301939058171745,
+      "grad_norm": 0.7218035459518433,
+      "learning_rate": 4.660373389359137e-06,
+      "loss": 0.6675,
+      "step": 3719
+    },
+    {
+      "epoch": 1.0304709141274238,
+      "grad_norm": 0.6770936250686646,
+      "learning_rate": 4.660190045377825e-06,
+      "loss": 0.6293,
+      "step": 3720
+    },
+    {
+      "epoch": 1.0307479224376732,
+      "grad_norm": 0.6360563635826111,
+      "learning_rate": 4.660006655530161e-06,
+      "loss": 0.5993,
+      "step": 3721
+    },
+    {
+      "epoch": 1.0310249307479225,
+      "grad_norm": 0.6777840852737427,
+      "learning_rate": 4.65982321982004e-06,
+      "loss": 0.6694,
+      "step": 3722
+    },
+    {
+      "epoch": 1.0313019390581717,
+      "grad_norm": 0.7174295783042908,
+      "learning_rate": 4.659639738251356e-06,
+      "loss": 0.6579,
+      "step": 3723
+    },
+    {
+      "epoch": 1.0315789473684212,
+      "grad_norm": 0.7186366319656372,
+      "learning_rate": 4.659456210828006e-06,
+      "loss": 0.689,
+      "step": 3724
+    },
+    {
+      "epoch": 1.0318559556786704,
+      "grad_norm": 0.6772783398628235,
+      "learning_rate": 4.659272637553885e-06,
+      "loss": 0.611,
+      "step": 3725
+    },
+    {
+      "epoch": 1.0321329639889196,
+      "grad_norm": 0.7406778335571289,
+      "learning_rate": 4.659089018432893e-06,
+      "loss": 0.6801,
+      "step": 3726
+    },
+    {
+      "epoch": 1.032409972299169,
+      "grad_norm": 0.7386046051979065,
+      "learning_rate": 4.6589053534689266e-06,
+      "loss": 0.6512,
+      "step": 3727
+    },
+    {
+      "epoch": 1.0326869806094183,
+      "grad_norm": 0.6807011365890503,
+      "learning_rate": 4.658721642665887e-06,
+      "loss": 0.5922,
+      "step": 3728
+    },
+    {
+      "epoch": 1.0329639889196676,
+      "grad_norm": 0.7346801161766052,
+      "learning_rate": 4.658537886027673e-06,
+      "loss": 0.6891,
+      "step": 3729
+    },
+    {
+      "epoch": 1.0332409972299168,
+      "grad_norm": 0.7221609354019165,
+      "learning_rate": 4.6583540835581885e-06,
+      "loss": 0.6662,
+      "step": 3730
+    },
+    {
+      "epoch": 1.0335180055401663,
+      "grad_norm": 0.6834601759910583,
+      "learning_rate": 4.6581702352613345e-06,
+      "loss": 0.6249,
+      "step": 3731
+    },
+    {
+      "epoch": 1.0337950138504155,
+      "grad_norm": 0.6952083110809326,
+      "learning_rate": 4.657986341141015e-06,
+      "loss": 0.6403,
+      "step": 3732
+    },
+    {
+      "epoch": 1.0340720221606647,
+      "grad_norm": 0.7092493772506714,
+      "learning_rate": 4.657802401201135e-06,
+      "loss": 0.6295,
+      "step": 3733
+    },
+    {
+      "epoch": 1.0343490304709142,
+      "grad_norm": 0.7145261764526367,
+      "learning_rate": 4.657618415445599e-06,
+      "loss": 0.6624,
+      "step": 3734
+    },
+    {
+      "epoch": 1.0346260387811634,
+      "grad_norm": 0.7718804478645325,
+      "learning_rate": 4.657434383878314e-06,
+      "loss": 0.6584,
+      "step": 3735
+    },
+    {
+      "epoch": 1.0349030470914127,
+      "grad_norm": 0.7761196494102478,
+      "learning_rate": 4.657250306503189e-06,
+      "loss": 0.6235,
+      "step": 3736
+    },
+    {
+      "epoch": 1.0351800554016621,
+      "grad_norm": 0.6836169362068176,
+      "learning_rate": 4.6570661833241295e-06,
+      "loss": 0.6169,
+      "step": 3737
+    },
+    {
+      "epoch": 1.0354570637119114,
+      "grad_norm": 0.7584412097930908,
+      "learning_rate": 4.656882014345046e-06,
+      "loss": 0.6676,
+      "step": 3738
+    },
+    {
+      "epoch": 1.0357340720221606,
+      "grad_norm": 0.733180820941925,
+      "learning_rate": 4.656697799569851e-06,
+      "loss": 0.6969,
+      "step": 3739
+    },
+    {
+      "epoch": 1.03601108033241,
+      "grad_norm": 0.7027671933174133,
+      "learning_rate": 4.656513539002452e-06,
+      "loss": 0.6319,
+      "step": 3740
+    },
+    {
+      "epoch": 1.0362880886426593,
+      "grad_norm": 0.7461979985237122,
+      "learning_rate": 4.656329232646765e-06,
+      "loss": 0.6715,
+      "step": 3741
+    },
+    {
+      "epoch": 1.0365650969529085,
+      "grad_norm": 0.7621151208877563,
+      "learning_rate": 4.656144880506701e-06,
+      "loss": 0.6923,
+      "step": 3742
+    },
+    {
+      "epoch": 1.0368421052631578,
+      "grad_norm": 0.7271478176116943,
+      "learning_rate": 4.655960482586175e-06,
+      "loss": 0.6726,
+      "step": 3743
+    },
+    {
+      "epoch": 1.0371191135734072,
+      "grad_norm": 0.6788163185119629,
+      "learning_rate": 4.655776038889103e-06,
+      "loss": 0.6465,
+      "step": 3744
+    },
+    {
+      "epoch": 1.0373961218836565,
+      "grad_norm": 0.6891070604324341,
+      "learning_rate": 4.6555915494193996e-06,
+      "loss": 0.6635,
+      "step": 3745
+    },
+    {
+      "epoch": 1.0376731301939057,
+      "grad_norm": 0.6785054802894592,
+      "learning_rate": 4.655407014180984e-06,
+      "loss": 0.6266,
+      "step": 3746
+    },
+    {
+      "epoch": 1.0379501385041552,
+      "grad_norm": 0.6859633922576904,
+      "learning_rate": 4.655222433177772e-06,
+      "loss": 0.683,
+      "step": 3747
+    },
+    {
+      "epoch": 1.0382271468144044,
+      "grad_norm": 0.6922705173492432,
+      "learning_rate": 4.655037806413684e-06,
+      "loss": 0.6311,
+      "step": 3748
+    },
+    {
+      "epoch": 1.0385041551246537,
+      "grad_norm": 0.6790265440940857,
+      "learning_rate": 4.654853133892639e-06,
+      "loss": 0.6471,
+      "step": 3749
+    },
+    {
+      "epoch": 1.0387811634349031,
+      "grad_norm": 0.705460786819458,
+      "learning_rate": 4.654668415618561e-06,
+      "loss": 0.6347,
+      "step": 3750
+    },
+    {
+      "epoch": 1.0390581717451524,
+      "grad_norm": 0.7426798343658447,
+      "learning_rate": 4.6544836515953695e-06,
+      "loss": 0.7269,
+      "step": 3751
+    },
+    {
+      "epoch": 1.0393351800554016,
+      "grad_norm": 0.7155759334564209,
+      "learning_rate": 4.654298841826988e-06,
+      "loss": 0.6246,
+      "step": 3752
+    },
+    {
+      "epoch": 1.039612188365651,
+      "grad_norm": 0.6881625652313232,
+      "learning_rate": 4.654113986317341e-06,
+      "loss": 0.6356,
+      "step": 3753
+    },
+    {
+      "epoch": 1.0398891966759003,
+      "grad_norm": 0.7087181210517883,
+      "learning_rate": 4.653929085070352e-06,
+      "loss": 0.6633,
+      "step": 3754
+    },
+    {
+      "epoch": 1.0401662049861495,
+      "grad_norm": 0.7184396386146545,
+      "learning_rate": 4.653744138089949e-06,
+      "loss": 0.6779,
+      "step": 3755
+    },
+    {
+      "epoch": 1.040443213296399,
+      "grad_norm": 0.7010110020637512,
+      "learning_rate": 4.653559145380058e-06,
+      "loss": 0.671,
+      "step": 3756
+    },
+    {
+      "epoch": 1.0407202216066482,
+      "grad_norm": 0.7426394820213318,
+      "learning_rate": 4.653374106944606e-06,
+      "loss": 0.6824,
+      "step": 3757
+    },
+    {
+      "epoch": 1.0409972299168975,
+      "grad_norm": 0.671944797039032,
+      "learning_rate": 4.653189022787523e-06,
+      "loss": 0.6487,
+      "step": 3758
+    },
+    {
+      "epoch": 1.0412742382271467,
+      "grad_norm": 0.6617016792297363,
+      "learning_rate": 4.6530038929127385e-06,
+      "loss": 0.6362,
+      "step": 3759
+    },
+    {
+      "epoch": 1.0415512465373962,
+      "grad_norm": 0.6958049535751343,
+      "learning_rate": 4.652818717324182e-06,
+      "loss": 0.7073,
+      "step": 3760
+    },
+    {
+      "epoch": 1.0418282548476454,
+      "grad_norm": 0.6825550198554993,
+      "learning_rate": 4.652633496025788e-06,
+      "loss": 0.6001,
+      "step": 3761
+    },
+    {
+      "epoch": 1.0421052631578946,
+      "grad_norm": 0.6797915101051331,
+      "learning_rate": 4.652448229021488e-06,
+      "loss": 0.6436,
+      "step": 3762
+    },
+    {
+      "epoch": 1.042382271468144,
+      "grad_norm": 0.6714237332344055,
+      "learning_rate": 4.652262916315213e-06,
+      "loss": 0.6443,
+      "step": 3763
+    },
+    {
+      "epoch": 1.0426592797783933,
+      "grad_norm": 0.6620033979415894,
+      "learning_rate": 4.652077557910902e-06,
+      "loss": 0.6283,
+      "step": 3764
+    },
+    {
+      "epoch": 1.0429362880886426,
+      "grad_norm": 0.6751744747161865,
+      "learning_rate": 4.6518921538124885e-06,
+      "loss": 0.6158,
+      "step": 3765
+    },
+    {
+      "epoch": 1.043213296398892,
+      "grad_norm": 0.723050594329834,
+      "learning_rate": 4.651706704023908e-06,
+      "loss": 0.6551,
+      "step": 3766
+    },
+    {
+      "epoch": 1.0434903047091413,
+      "grad_norm": 0.6662552952766418,
+      "learning_rate": 4.651521208549101e-06,
+      "loss": 0.6189,
+      "step": 3767
+    },
+    {
+      "epoch": 1.0437673130193905,
+      "grad_norm": 0.6784934401512146,
+      "learning_rate": 4.651335667392003e-06,
+      "loss": 0.6286,
+      "step": 3768
+    },
+    {
+      "epoch": 1.04404432132964,
+      "grad_norm": 0.6911255121231079,
+      "learning_rate": 4.6511500805565555e-06,
+      "loss": 0.6145,
+      "step": 3769
+    },
+    {
+      "epoch": 1.0443213296398892,
+      "grad_norm": 0.7118515372276306,
+      "learning_rate": 4.6509644480466984e-06,
+      "loss": 0.6615,
+      "step": 3770
+    },
+    {
+      "epoch": 1.0445983379501385,
+      "grad_norm": 0.6700476408004761,
+      "learning_rate": 4.650778769866372e-06,
+      "loss": 0.6113,
+      "step": 3771
+    },
+    {
+      "epoch": 1.044875346260388,
+      "grad_norm": 0.7546408176422119,
+      "learning_rate": 4.650593046019521e-06,
+      "loss": 0.6646,
+      "step": 3772
+    },
+    {
+      "epoch": 1.0451523545706372,
+      "grad_norm": 0.7420517206192017,
+      "learning_rate": 4.650407276510088e-06,
+      "loss": 0.7298,
+      "step": 3773
+    },
+    {
+      "epoch": 1.0454293628808864,
+      "grad_norm": 0.6817473769187927,
+      "learning_rate": 4.6502214613420164e-06,
+      "loss": 0.6105,
+      "step": 3774
+    },
+    {
+      "epoch": 1.0457063711911356,
+      "grad_norm": 0.7240308523178101,
+      "learning_rate": 4.6500356005192514e-06,
+      "loss": 0.6463,
+      "step": 3775
+    },
+    {
+      "epoch": 1.045983379501385,
+      "grad_norm": 0.6789447069168091,
+      "learning_rate": 4.649849694045741e-06,
+      "loss": 0.654,
+      "step": 3776
+    },
+    {
+      "epoch": 1.0462603878116343,
+      "grad_norm": 0.6913023591041565,
+      "learning_rate": 4.649663741925431e-06,
+      "loss": 0.6083,
+      "step": 3777
+    },
+    {
+      "epoch": 1.0465373961218836,
+      "grad_norm": 0.6867649555206299,
+      "learning_rate": 4.649477744162271e-06,
+      "loss": 0.6776,
+      "step": 3778
+    },
+    {
+      "epoch": 1.046814404432133,
+      "grad_norm": 0.6704279780387878,
+      "learning_rate": 4.6492917007602075e-06,
+      "loss": 0.6611,
+      "step": 3779
+    },
+    {
+      "epoch": 1.0470914127423823,
+      "grad_norm": 0.6176725029945374,
+      "learning_rate": 4.6491056117231936e-06,
+      "loss": 0.5878,
+      "step": 3780
+    },
+    {
+      "epoch": 1.0473684210526315,
+      "grad_norm": 0.6866704821586609,
+      "learning_rate": 4.648919477055179e-06,
+      "loss": 0.6346,
+      "step": 3781
+    },
+    {
+      "epoch": 1.047645429362881,
+      "grad_norm": 0.6737282872200012,
+      "learning_rate": 4.648733296760117e-06,
+      "loss": 0.6675,
+      "step": 3782
+    },
+    {
+      "epoch": 1.0479224376731302,
+      "grad_norm": 0.6623876690864563,
+      "learning_rate": 4.6485470708419595e-06,
+      "loss": 0.6386,
+      "step": 3783
+    },
+    {
+      "epoch": 1.0481994459833794,
+      "grad_norm": 0.6881468892097473,
+      "learning_rate": 4.648360799304661e-06,
+      "loss": 0.6733,
+      "step": 3784
+    },
+    {
+      "epoch": 1.048476454293629,
+      "grad_norm": 0.7181234955787659,
+      "learning_rate": 4.648174482152176e-06,
+      "loss": 0.6804,
+      "step": 3785
+    },
+    {
+      "epoch": 1.0487534626038781,
+      "grad_norm": 0.7076879739761353,
+      "learning_rate": 4.647988119388461e-06,
+      "loss": 0.6428,
+      "step": 3786
+    },
+    {
+      "epoch": 1.0490304709141274,
+      "grad_norm": 0.7163893580436707,
+      "learning_rate": 4.647801711017473e-06,
+      "loss": 0.659,
+      "step": 3787
+    },
+    {
+      "epoch": 1.0493074792243768,
+      "grad_norm": 0.6863222122192383,
+      "learning_rate": 4.64761525704317e-06,
+      "loss": 0.6169,
+      "step": 3788
+    },
+    {
+      "epoch": 1.049584487534626,
+      "grad_norm": 0.6541346907615662,
+      "learning_rate": 4.64742875746951e-06,
+      "loss": 0.6575,
+      "step": 3789
+    },
+    {
+      "epoch": 1.0498614958448753,
+      "grad_norm": 0.7049393057823181,
+      "learning_rate": 4.647242212300453e-06,
+      "loss": 0.6744,
+      "step": 3790
+    },
+    {
+      "epoch": 1.0501385041551246,
+      "grad_norm": 0.697417140007019,
+      "learning_rate": 4.647055621539962e-06,
+      "loss": 0.6321,
+      "step": 3791
+    },
+    {
+      "epoch": 1.050415512465374,
+      "grad_norm": 0.7294650077819824,
+      "learning_rate": 4.646868985191996e-06,
+      "loss": 0.7125,
+      "step": 3792
+    },
+    {
+      "epoch": 1.0506925207756233,
+      "grad_norm": 0.6390814185142517,
+      "learning_rate": 4.64668230326052e-06,
+      "loss": 0.5747,
+      "step": 3793
+    },
+    {
+      "epoch": 1.0509695290858725,
+      "grad_norm": 0.6700131893157959,
+      "learning_rate": 4.6464955757494955e-06,
+      "loss": 0.6229,
+      "step": 3794
+    },
+    {
+      "epoch": 1.051246537396122,
+      "grad_norm": 0.7338569760322571,
+      "learning_rate": 4.646308802662889e-06,
+      "loss": 0.6474,
+      "step": 3795
+    },
+    {
+      "epoch": 1.0515235457063712,
+      "grad_norm": 0.6908360123634338,
+      "learning_rate": 4.646121984004666e-06,
+      "loss": 0.6504,
+      "step": 3796
+    },
+    {
+      "epoch": 1.0518005540166204,
+      "grad_norm": 0.6733314394950867,
+      "learning_rate": 4.645935119778792e-06,
+      "loss": 0.665,
+      "step": 3797
+    },
+    {
+      "epoch": 1.0520775623268699,
+      "grad_norm": 0.7230425477027893,
+      "learning_rate": 4.6457482099892355e-06,
+      "loss": 0.6875,
+      "step": 3798
+    },
+    {
+      "epoch": 1.0523545706371191,
+      "grad_norm": 0.6543940305709839,
+      "learning_rate": 4.645561254639965e-06,
+      "loss": 0.648,
+      "step": 3799
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.718337893486023,
+      "learning_rate": 4.645374253734949e-06,
+      "loss": 0.6047,
+      "step": 3800
+    },
+    {
+      "epoch": 1.0529085872576178,
+      "grad_norm": 0.6767460703849792,
+      "learning_rate": 4.64518720727816e-06,
+      "loss": 0.6132,
+      "step": 3801
+    },
+    {
+      "epoch": 1.053185595567867,
+      "grad_norm": 0.6737857460975647,
+      "learning_rate": 4.645000115273568e-06,
+      "loss": 0.6429,
+      "step": 3802
+    },
+    {
+      "epoch": 1.0534626038781163,
+      "grad_norm": 0.7279229760169983,
+      "learning_rate": 4.644812977725146e-06,
+      "loss": 0.6202,
+      "step": 3803
+    },
+    {
+      "epoch": 1.0537396121883655,
+      "grad_norm": 0.6861441731452942,
+      "learning_rate": 4.644625794636867e-06,
+      "loss": 0.6381,
+      "step": 3804
+    },
+    {
+      "epoch": 1.054016620498615,
+      "grad_norm": 0.6784233450889587,
+      "learning_rate": 4.644438566012706e-06,
+      "loss": 0.6411,
+      "step": 3805
+    },
+    {
+      "epoch": 1.0542936288088642,
+      "grad_norm": 0.672813892364502,
+      "learning_rate": 4.644251291856637e-06,
+      "loss": 0.6455,
+      "step": 3806
+    },
+    {
+      "epoch": 1.0545706371191135,
+      "grad_norm": 0.6707661747932434,
+      "learning_rate": 4.6440639721726375e-06,
+      "loss": 0.648,
+      "step": 3807
+    },
+    {
+      "epoch": 1.054847645429363,
+      "grad_norm": 0.7086052894592285,
+      "learning_rate": 4.643876606964685e-06,
+      "loss": 0.6721,
+      "step": 3808
+    },
+    {
+      "epoch": 1.0551246537396122,
+      "grad_norm": 0.688983142375946,
+      "learning_rate": 4.643689196236757e-06,
+      "loss": 0.6487,
+      "step": 3809
+    },
+    {
+      "epoch": 1.0554016620498614,
+      "grad_norm": 0.6924951672554016,
+      "learning_rate": 4.643501739992834e-06,
+      "loss": 0.6605,
+      "step": 3810
+    },
+    {
+      "epoch": 1.0556786703601109,
+      "grad_norm": 0.6518945693969727,
+      "learning_rate": 4.643314238236893e-06,
+      "loss": 0.5965,
+      "step": 3811
+    },
+    {
+      "epoch": 1.05595567867036,
+      "grad_norm": 0.6460810899734497,
+      "learning_rate": 4.643126690972919e-06,
+      "loss": 0.6296,
+      "step": 3812
+    },
+    {
+      "epoch": 1.0562326869806093,
+      "grad_norm": 0.671865701675415,
+      "learning_rate": 4.642939098204892e-06,
+      "loss": 0.6657,
+      "step": 3813
+    },
+    {
+      "epoch": 1.0565096952908588,
+      "grad_norm": 0.6928796172142029,
+      "learning_rate": 4.642751459936796e-06,
+      "loss": 0.6628,
+      "step": 3814
+    },
+    {
+      "epoch": 1.056786703601108,
+      "grad_norm": 0.7006489634513855,
+      "learning_rate": 4.642563776172614e-06,
+      "loss": 0.6696,
+      "step": 3815
+    },
+    {
+      "epoch": 1.0570637119113573,
+      "grad_norm": 0.7267358303070068,
+      "learning_rate": 4.642376046916331e-06,
+      "loss": 0.6708,
+      "step": 3816
+    },
+    {
+      "epoch": 1.0573407202216067,
+      "grad_norm": 0.6783285737037659,
+      "learning_rate": 4.642188272171933e-06,
+      "loss": 0.6437,
+      "step": 3817
+    },
+    {
+      "epoch": 1.057617728531856,
+      "grad_norm": 0.6970617771148682,
+      "learning_rate": 4.642000451943409e-06,
+      "loss": 0.6242,
+      "step": 3818
+    },
+    {
+      "epoch": 1.0578947368421052,
+      "grad_norm": 0.7154437899589539,
+      "learning_rate": 4.641812586234745e-06,
+      "loss": 0.6422,
+      "step": 3819
+    },
+    {
+      "epoch": 1.0581717451523547,
+      "grad_norm": 0.6935372948646545,
+      "learning_rate": 4.64162467504993e-06,
+      "loss": 0.6757,
+      "step": 3820
+    },
+    {
+      "epoch": 1.058448753462604,
+      "grad_norm": 0.7333564758300781,
+      "learning_rate": 4.641436718392954e-06,
+      "loss": 0.632,
+      "step": 3821
+    },
+    {
+      "epoch": 1.0587257617728532,
+      "grad_norm": 0.7167134881019592,
+      "learning_rate": 4.641248716267808e-06,
+      "loss": 0.6473,
+      "step": 3822
+    },
+    {
+      "epoch": 1.0590027700831024,
+      "grad_norm": 0.7148312330245972,
+      "learning_rate": 4.641060668678483e-06,
+      "loss": 0.6946,
+      "step": 3823
+    },
+    {
+      "epoch": 1.0592797783933519,
+      "grad_norm": 0.7001869678497314,
+      "learning_rate": 4.640872575628973e-06,
+      "loss": 0.6365,
+      "step": 3824
+    },
+    {
+      "epoch": 1.059556786703601,
+      "grad_norm": 0.680051326751709,
+      "learning_rate": 4.640684437123271e-06,
+      "loss": 0.6184,
+      "step": 3825
+    },
+    {
+      "epoch": 1.0598337950138503,
+      "grad_norm": 0.715980589389801,
+      "learning_rate": 4.640496253165371e-06,
+      "loss": 0.6697,
+      "step": 3826
+    },
+    {
+      "epoch": 1.0601108033240998,
+      "grad_norm": 0.70005863904953,
+      "learning_rate": 4.6403080237592704e-06,
+      "loss": 0.6385,
+      "step": 3827
+    },
+    {
+      "epoch": 1.060387811634349,
+      "grad_norm": 0.67072993516922,
+      "learning_rate": 4.640119748908964e-06,
+      "loss": 0.58,
+      "step": 3828
+    },
+    {
+      "epoch": 1.0606648199445983,
+      "grad_norm": 0.6752764582633972,
+      "learning_rate": 4.6399314286184496e-06,
+      "loss": 0.634,
+      "step": 3829
+    },
+    {
+      "epoch": 1.0609418282548477,
+      "grad_norm": 0.676100492477417,
+      "learning_rate": 4.639743062891727e-06,
+      "loss": 0.6329,
+      "step": 3830
+    },
+    {
+      "epoch": 1.061218836565097,
+      "grad_norm": 0.6580525636672974,
+      "learning_rate": 4.639554651732794e-06,
+      "loss": 0.5743,
+      "step": 3831
+    },
+    {
+      "epoch": 1.0614958448753462,
+      "grad_norm": 0.7047877311706543,
+      "learning_rate": 4.639366195145653e-06,
+      "loss": 0.6801,
+      "step": 3832
+    },
+    {
+      "epoch": 1.0617728531855957,
+      "grad_norm": 0.6486926078796387,
+      "learning_rate": 4.639177693134304e-06,
+      "loss": 0.6117,
+      "step": 3833
+    },
+    {
+      "epoch": 1.062049861495845,
+      "grad_norm": 0.704925000667572,
+      "learning_rate": 4.638989145702749e-06,
+      "loss": 0.6711,
+      "step": 3834
+    },
+    {
+      "epoch": 1.0623268698060941,
+      "grad_norm": 0.6557621359825134,
+      "learning_rate": 4.638800552854993e-06,
+      "loss": 0.5999,
+      "step": 3835
+    },
+    {
+      "epoch": 1.0626038781163434,
+      "grad_norm": 0.6929731369018555,
+      "learning_rate": 4.638611914595038e-06,
+      "loss": 0.6597,
+      "step": 3836
+    },
+    {
+      "epoch": 1.0628808864265928,
+      "grad_norm": 0.7148190140724182,
+      "learning_rate": 4.638423230926891e-06,
+      "loss": 0.6758,
+      "step": 3837
+    },
+    {
+      "epoch": 1.063157894736842,
+      "grad_norm": 0.6981825828552246,
+      "learning_rate": 4.638234501854559e-06,
+      "loss": 0.6246,
+      "step": 3838
+    },
+    {
+      "epoch": 1.0634349030470913,
+      "grad_norm": 0.6970887780189514,
+      "learning_rate": 4.638045727382048e-06,
+      "loss": 0.6225,
+      "step": 3839
+    },
+    {
+      "epoch": 1.0637119113573408,
+      "grad_norm": 0.6663398146629333,
+      "learning_rate": 4.637856907513366e-06,
+      "loss": 0.6148,
+      "step": 3840
+    },
+    {
+      "epoch": 1.06398891966759,
+      "grad_norm": 0.7192431092262268,
+      "learning_rate": 4.637668042252522e-06,
+      "loss": 0.6335,
+      "step": 3841
+    },
+    {
+      "epoch": 1.0642659279778393,
+      "grad_norm": 0.6780756115913391,
+      "learning_rate": 4.637479131603526e-06,
+      "loss": 0.6896,
+      "step": 3842
+    },
+    {
+      "epoch": 1.0645429362880887,
+      "grad_norm": 0.7636496424674988,
+      "learning_rate": 4.63729017557039e-06,
+      "loss": 0.6415,
+      "step": 3843
+    },
+    {
+      "epoch": 1.064819944598338,
+      "grad_norm": 0.6857926845550537,
+      "learning_rate": 4.637101174157126e-06,
+      "loss": 0.5899,
+      "step": 3844
+    },
+    {
+      "epoch": 1.0650969529085872,
+      "grad_norm": 0.6750476360321045,
+      "learning_rate": 4.636912127367746e-06,
+      "loss": 0.6186,
+      "step": 3845
+    },
+    {
+      "epoch": 1.0653739612188367,
+      "grad_norm": 0.6938686966896057,
+      "learning_rate": 4.636723035206265e-06,
+      "loss": 0.6909,
+      "step": 3846
+    },
+    {
+      "epoch": 1.065650969529086,
+      "grad_norm": 0.7091078162193298,
+      "learning_rate": 4.636533897676696e-06,
+      "loss": 0.6578,
+      "step": 3847
+    },
+    {
+      "epoch": 1.0659279778393351,
+      "grad_norm": 0.6912356615066528,
+      "learning_rate": 4.636344714783058e-06,
+      "loss": 0.6169,
+      "step": 3848
+    },
+    {
+      "epoch": 1.0662049861495846,
+      "grad_norm": 0.6690759658813477,
+      "learning_rate": 4.636155486529365e-06,
+      "loss": 0.6258,
+      "step": 3849
+    },
+    {
+      "epoch": 1.0664819944598338,
+      "grad_norm": 0.704440712928772,
+      "learning_rate": 4.6359662129196355e-06,
+      "loss": 0.6696,
+      "step": 3850
+    },
+    {
+      "epoch": 1.066759002770083,
+      "grad_norm": 0.6986096501350403,
+      "learning_rate": 4.63577689395789e-06,
+      "loss": 0.6342,
+      "step": 3851
+    },
+    {
+      "epoch": 1.0670360110803325,
+      "grad_norm": 0.7509374618530273,
+      "learning_rate": 4.635587529648146e-06,
+      "loss": 0.6382,
+      "step": 3852
+    },
+    {
+      "epoch": 1.0673130193905818,
+      "grad_norm": 0.6818401217460632,
+      "learning_rate": 4.635398119994425e-06,
+      "loss": 0.6343,
+      "step": 3853
+    },
+    {
+      "epoch": 1.067590027700831,
+      "grad_norm": 0.6807084083557129,
+      "learning_rate": 4.6352086650007484e-06,
+      "loss": 0.6749,
+      "step": 3854
+    },
+    {
+      "epoch": 1.0678670360110802,
+      "grad_norm": 0.6697701811790466,
+      "learning_rate": 4.635019164671141e-06,
+      "loss": 0.6586,
+      "step": 3855
+    },
+    {
+      "epoch": 1.0681440443213297,
+      "grad_norm": 0.6940995454788208,
+      "learning_rate": 4.634829619009622e-06,
+      "loss": 0.6672,
+      "step": 3856
+    },
+    {
+      "epoch": 1.068421052631579,
+      "grad_norm": 0.7035492658615112,
+      "learning_rate": 4.63464002802022e-06,
+      "loss": 0.6256,
+      "step": 3857
+    },
+    {
+      "epoch": 1.0686980609418282,
+      "grad_norm": 0.7253122329711914,
+      "learning_rate": 4.634450391706959e-06,
+      "loss": 0.6276,
+      "step": 3858
+    },
+    {
+      "epoch": 1.0689750692520776,
+      "grad_norm": 0.7146328091621399,
+      "learning_rate": 4.6342607100738645e-06,
+      "loss": 0.6419,
+      "step": 3859
+    },
+    {
+      "epoch": 1.0692520775623269,
+      "grad_norm": 0.7397799491882324,
+      "learning_rate": 4.634070983124965e-06,
+      "loss": 0.6172,
+      "step": 3860
+    },
+    {
+      "epoch": 1.0695290858725761,
+      "grad_norm": 0.6944563388824463,
+      "learning_rate": 4.633881210864289e-06,
+      "loss": 0.6411,
+      "step": 3861
+    },
+    {
+      "epoch": 1.0698060941828256,
+      "grad_norm": 0.7174142599105835,
+      "learning_rate": 4.633691393295865e-06,
+      "loss": 0.6253,
+      "step": 3862
+    },
+    {
+      "epoch": 1.0700831024930748,
+      "grad_norm": 0.6977531313896179,
+      "learning_rate": 4.633501530423724e-06,
+      "loss": 0.6395,
+      "step": 3863
+    },
+    {
+      "epoch": 1.070360110803324,
+      "grad_norm": 0.6565867066383362,
+      "learning_rate": 4.6333116222518985e-06,
+      "loss": 0.6364,
+      "step": 3864
+    },
+    {
+      "epoch": 1.0706371191135735,
+      "grad_norm": 0.724623441696167,
+      "learning_rate": 4.633121668784417e-06,
+      "loss": 0.6946,
+      "step": 3865
+    },
+    {
+      "epoch": 1.0709141274238227,
+      "grad_norm": 0.6574693918228149,
+      "learning_rate": 4.632931670025317e-06,
+      "loss": 0.5963,
+      "step": 3866
+    },
+    {
+      "epoch": 1.071191135734072,
+      "grad_norm": 0.6685285568237305,
+      "learning_rate": 4.63274162597863e-06,
+      "loss": 0.6138,
+      "step": 3867
+    },
+    {
+      "epoch": 1.0714681440443212,
+      "grad_norm": 0.679956316947937,
+      "learning_rate": 4.632551536648392e-06,
+      "loss": 0.6449,
+      "step": 3868
+    },
+    {
+      "epoch": 1.0717451523545707,
+      "grad_norm": 0.7049286961555481,
+      "learning_rate": 4.6323614020386385e-06,
+      "loss": 0.6302,
+      "step": 3869
+    },
+    {
+      "epoch": 1.07202216066482,
+      "grad_norm": 0.6563496589660645,
+      "learning_rate": 4.632171222153407e-06,
+      "loss": 0.657,
+      "step": 3870
+    },
+    {
+      "epoch": 1.0722991689750692,
+      "grad_norm": 0.6608796119689941,
+      "learning_rate": 4.631980996996737e-06,
+      "loss": 0.6301,
+      "step": 3871
+    },
+    {
+      "epoch": 1.0725761772853186,
+      "grad_norm": 0.7565630674362183,
+      "learning_rate": 4.6317907265726645e-06,
+      "loss": 0.6184,
+      "step": 3872
+    },
+    {
+      "epoch": 1.0728531855955679,
+      "grad_norm": 0.6970774531364441,
+      "learning_rate": 4.631600410885231e-06,
+      "loss": 0.6511,
+      "step": 3873
+    },
+    {
+      "epoch": 1.073130193905817,
+      "grad_norm": 0.6813990473747253,
+      "learning_rate": 4.6314100499384775e-06,
+      "loss": 0.6551,
+      "step": 3874
+    },
+    {
+      "epoch": 1.0734072022160666,
+      "grad_norm": 0.655053436756134,
+      "learning_rate": 4.631219643736445e-06,
+      "loss": 0.6385,
+      "step": 3875
+    },
+    {
+      "epoch": 1.0736842105263158,
+      "grad_norm": 0.6602583527565002,
+      "learning_rate": 4.631029192283178e-06,
+      "loss": 0.6619,
+      "step": 3876
+    },
+    {
+      "epoch": 1.073961218836565,
+      "grad_norm": 0.6874777674674988,
+      "learning_rate": 4.630838695582718e-06,
+      "loss": 0.6414,
+      "step": 3877
+    },
+    {
+      "epoch": 1.0742382271468145,
+      "grad_norm": 0.6865057349205017,
+      "learning_rate": 4.6306481536391125e-06,
+      "loss": 0.6874,
+      "step": 3878
+    },
+    {
+      "epoch": 1.0745152354570637,
+      "grad_norm": 0.6830350160598755,
+      "learning_rate": 4.630457566456405e-06,
+      "loss": 0.6246,
+      "step": 3879
+    },
+    {
+      "epoch": 1.074792243767313,
+      "grad_norm": 0.6801131367683411,
+      "learning_rate": 4.630266934038642e-06,
+      "loss": 0.6192,
+      "step": 3880
+    },
+    {
+      "epoch": 1.0750692520775624,
+      "grad_norm": 0.7242031693458557,
+      "learning_rate": 4.630076256389872e-06,
+      "loss": 0.6639,
+      "step": 3881
+    },
+    {
+      "epoch": 1.0753462603878117,
+      "grad_norm": 0.6798548698425293,
+      "learning_rate": 4.629885533514143e-06,
+      "loss": 0.6434,
+      "step": 3882
+    },
+    {
+      "epoch": 1.075623268698061,
+      "grad_norm": 0.65250164270401,
+      "learning_rate": 4.629694765415506e-06,
+      "loss": 0.663,
+      "step": 3883
+    },
+    {
+      "epoch": 1.0759002770083101,
+      "grad_norm": 0.7037781476974487,
+      "learning_rate": 4.629503952098011e-06,
+      "loss": 0.6613,
+      "step": 3884
+    },
+    {
+      "epoch": 1.0761772853185596,
+      "grad_norm": 0.7319658994674683,
+      "learning_rate": 4.6293130935657086e-06,
+      "loss": 0.6283,
+      "step": 3885
+    },
+    {
+      "epoch": 1.0764542936288088,
+      "grad_norm": 0.7035661935806274,
+      "learning_rate": 4.6291221898226505e-06,
+      "loss": 0.6061,
+      "step": 3886
+    },
+    {
+      "epoch": 1.076731301939058,
+      "grad_norm": 0.6649536490440369,
+      "learning_rate": 4.628931240872892e-06,
+      "loss": 0.6054,
+      "step": 3887
+    },
+    {
+      "epoch": 1.0770083102493075,
+      "grad_norm": 0.6910920143127441,
+      "learning_rate": 4.628740246720486e-06,
+      "loss": 0.638,
+      "step": 3888
+    },
+    {
+      "epoch": 1.0772853185595568,
+      "grad_norm": 0.6866145133972168,
+      "learning_rate": 4.628549207369489e-06,
+      "loss": 0.6503,
+      "step": 3889
+    },
+    {
+      "epoch": 1.077562326869806,
+      "grad_norm": 0.6962095499038696,
+      "learning_rate": 4.628358122823956e-06,
+      "loss": 0.6534,
+      "step": 3890
+    },
+    {
+      "epoch": 1.0778393351800555,
+      "grad_norm": 0.6696187853813171,
+      "learning_rate": 4.628166993087945e-06,
+      "loss": 0.664,
+      "step": 3891
+    },
+    {
+      "epoch": 1.0781163434903047,
+      "grad_norm": 0.6706064343452454,
+      "learning_rate": 4.627975818165514e-06,
+      "loss": 0.6615,
+      "step": 3892
+    },
+    {
+      "epoch": 1.078393351800554,
+      "grad_norm": 0.7948892712593079,
+      "learning_rate": 4.627784598060723e-06,
+      "loss": 0.6638,
+      "step": 3893
+    },
+    {
+      "epoch": 1.0786703601108034,
+      "grad_norm": 0.7281169891357422,
+      "learning_rate": 4.62759333277763e-06,
+      "loss": 0.6554,
+      "step": 3894
+    },
+    {
+      "epoch": 1.0789473684210527,
+      "grad_norm": 0.673255980014801,
+      "learning_rate": 4.627402022320298e-06,
+      "loss": 0.6496,
+      "step": 3895
+    },
+    {
+      "epoch": 1.079224376731302,
+      "grad_norm": 0.7022432088851929,
+      "learning_rate": 4.627210666692787e-06,
+      "loss": 0.6301,
+      "step": 3896
+    },
+    {
+      "epoch": 1.0795013850415511,
+      "grad_norm": 0.6839655637741089,
+      "learning_rate": 4.627019265899163e-06,
+      "loss": 0.6362,
+      "step": 3897
+    },
+    {
+      "epoch": 1.0797783933518006,
+      "grad_norm": 0.7075784802436829,
+      "learning_rate": 4.626827819943488e-06,
+      "loss": 0.6091,
+      "step": 3898
+    },
+    {
+      "epoch": 1.0800554016620498,
+      "grad_norm": 0.7186004519462585,
+      "learning_rate": 4.626636328829826e-06,
+      "loss": 0.6602,
+      "step": 3899
+    },
+    {
+      "epoch": 1.080332409972299,
+      "grad_norm": 0.7156257033348083,
+      "learning_rate": 4.6264447925622445e-06,
+      "loss": 0.6283,
+      "step": 3900
+    },
+    {
+      "epoch": 1.0806094182825485,
+      "grad_norm": 0.6951586008071899,
+      "learning_rate": 4.626253211144809e-06,
+      "loss": 0.6733,
+      "step": 3901
+    },
+    {
+      "epoch": 1.0808864265927978,
+      "grad_norm": 0.7055180072784424,
+      "learning_rate": 4.626061584581589e-06,
+      "loss": 0.6538,
+      "step": 3902
+    },
+    {
+      "epoch": 1.081163434903047,
+      "grad_norm": 0.7053461074829102,
+      "learning_rate": 4.625869912876651e-06,
+      "loss": 0.6503,
+      "step": 3903
+    },
+    {
+      "epoch": 1.0814404432132965,
+      "grad_norm": 0.6794017553329468,
+      "learning_rate": 4.6256781960340665e-06,
+      "loss": 0.6354,
+      "step": 3904
+    },
+    {
+      "epoch": 1.0817174515235457,
+      "grad_norm": 0.7033610939979553,
+      "learning_rate": 4.625486434057906e-06,
+      "loss": 0.7112,
+      "step": 3905
+    },
+    {
+      "epoch": 1.081994459833795,
+      "grad_norm": 0.7239475846290588,
+      "learning_rate": 4.6252946269522406e-06,
+      "loss": 0.6583,
+      "step": 3906
+    },
+    {
+      "epoch": 1.0822714681440444,
+      "grad_norm": 0.6499510407447815,
+      "learning_rate": 4.625102774721142e-06,
+      "loss": 0.6023,
+      "step": 3907
+    },
+    {
+      "epoch": 1.0825484764542936,
+      "grad_norm": 0.6923518776893616,
+      "learning_rate": 4.6249108773686846e-06,
+      "loss": 0.6733,
+      "step": 3908
+    },
+    {
+      "epoch": 1.0828254847645429,
+      "grad_norm": 0.6752975583076477,
+      "learning_rate": 4.624718934898944e-06,
+      "loss": 0.6746,
+      "step": 3909
+    },
+    {
+      "epoch": 1.0831024930747923,
+      "grad_norm": 0.6957419514656067,
+      "learning_rate": 4.624526947315994e-06,
+      "loss": 0.6517,
+      "step": 3910
+    },
+    {
+      "epoch": 1.0833795013850416,
+      "grad_norm": 0.7053093910217285,
+      "learning_rate": 4.62433491462391e-06,
+      "loss": 0.5989,
+      "step": 3911
+    },
+    {
+      "epoch": 1.0836565096952908,
+      "grad_norm": 0.7093052268028259,
+      "learning_rate": 4.624142836826773e-06,
+      "loss": 0.6169,
+      "step": 3912
+    },
+    {
+      "epoch": 1.0839335180055403,
+      "grad_norm": 0.6362714767456055,
+      "learning_rate": 4.623950713928658e-06,
+      "loss": 0.5909,
+      "step": 3913
+    },
+    {
+      "epoch": 1.0842105263157895,
+      "grad_norm": 0.6835925579071045,
+      "learning_rate": 4.623758545933645e-06,
+      "loss": 0.6006,
+      "step": 3914
+    },
+    {
+      "epoch": 1.0844875346260388,
+      "grad_norm": 0.7387256622314453,
+      "learning_rate": 4.623566332845815e-06,
+      "loss": 0.6884,
+      "step": 3915
+    },
+    {
+      "epoch": 1.084764542936288,
+      "grad_norm": 0.6731743216514587,
+      "learning_rate": 4.6233740746692495e-06,
+      "loss": 0.6236,
+      "step": 3916
+    },
+    {
+      "epoch": 1.0850415512465375,
+      "grad_norm": 0.667972981929779,
+      "learning_rate": 4.623181771408029e-06,
+      "loss": 0.6199,
+      "step": 3917
+    },
+    {
+      "epoch": 1.0853185595567867,
+      "grad_norm": 0.6656007766723633,
+      "learning_rate": 4.622989423066238e-06,
+      "loss": 0.6551,
+      "step": 3918
+    },
+    {
+      "epoch": 1.085595567867036,
+      "grad_norm": 0.6904618740081787,
+      "learning_rate": 4.622797029647958e-06,
+      "loss": 0.6831,
+      "step": 3919
+    },
+    {
+      "epoch": 1.0858725761772854,
+      "grad_norm": 0.6999261975288391,
+      "learning_rate": 4.622604591157279e-06,
+      "loss": 0.6216,
+      "step": 3920
+    },
+    {
+      "epoch": 1.0861495844875346,
+      "grad_norm": 0.7652238011360168,
+      "learning_rate": 4.622412107598282e-06,
+      "loss": 0.6708,
+      "step": 3921
+    },
+    {
+      "epoch": 1.0864265927977839,
+      "grad_norm": 0.6936379671096802,
+      "learning_rate": 4.622219578975057e-06,
+      "loss": 0.6148,
+      "step": 3922
+    },
+    {
+      "epoch": 1.0867036011080333,
+      "grad_norm": 0.7080146074295044,
+      "learning_rate": 4.622027005291691e-06,
+      "loss": 0.6505,
+      "step": 3923
+    },
+    {
+      "epoch": 1.0869806094182826,
+      "grad_norm": 0.7341291308403015,
+      "learning_rate": 4.621834386552271e-06,
+      "loss": 0.6059,
+      "step": 3924
+    },
+    {
+      "epoch": 1.0872576177285318,
+      "grad_norm": 0.7292314767837524,
+      "learning_rate": 4.62164172276089e-06,
+      "loss": 0.6471,
+      "step": 3925
+    },
+    {
+      "epoch": 1.0875346260387813,
+      "grad_norm": 0.6754866242408752,
+      "learning_rate": 4.621449013921636e-06,
+      "loss": 0.6214,
+      "step": 3926
+    },
+    {
+      "epoch": 1.0878116343490305,
+      "grad_norm": 0.6615110039710999,
+      "learning_rate": 4.621256260038602e-06,
+      "loss": 0.6625,
+      "step": 3927
+    },
+    {
+      "epoch": 1.0880886426592797,
+      "grad_norm": 0.6877681016921997,
+      "learning_rate": 4.621063461115882e-06,
+      "loss": 0.6483,
+      "step": 3928
+    },
+    {
+      "epoch": 1.088365650969529,
+      "grad_norm": 0.687233030796051,
+      "learning_rate": 4.6208706171575665e-06,
+      "loss": 0.6702,
+      "step": 3929
+    },
+    {
+      "epoch": 1.0886426592797784,
+      "grad_norm": 0.6554688215255737,
+      "learning_rate": 4.6206777281677525e-06,
+      "loss": 0.6056,
+      "step": 3930
+    },
+    {
+      "epoch": 1.0889196675900277,
+      "grad_norm": 0.6812751889228821,
+      "learning_rate": 4.620484794150534e-06,
+      "loss": 0.6578,
+      "step": 3931
+    },
+    {
+      "epoch": 1.089196675900277,
+      "grad_norm": 0.7219202518463135,
+      "learning_rate": 4.6202918151100085e-06,
+      "loss": 0.6641,
+      "step": 3932
+    },
+    {
+      "epoch": 1.0894736842105264,
+      "grad_norm": 0.68025141954422,
+      "learning_rate": 4.6200987910502735e-06,
+      "loss": 0.6567,
+      "step": 3933
+    },
+    {
+      "epoch": 1.0897506925207756,
+      "grad_norm": 0.7100472450256348,
+      "learning_rate": 4.6199057219754265e-06,
+      "loss": 0.6972,
+      "step": 3934
+    },
+    {
+      "epoch": 1.0900277008310248,
+      "grad_norm": 0.7465969920158386,
+      "learning_rate": 4.619712607889568e-06,
+      "loss": 0.623,
+      "step": 3935
+    },
+    {
+      "epoch": 1.0903047091412743,
+      "grad_norm": 0.7210541367530823,
+      "learning_rate": 4.619519448796797e-06,
+      "loss": 0.6612,
+      "step": 3936
+    },
+    {
+      "epoch": 1.0905817174515235,
+      "grad_norm": 0.6653692722320557,
+      "learning_rate": 4.619326244701216e-06,
+      "loss": 0.5711,
+      "step": 3937
+    },
+    {
+      "epoch": 1.0908587257617728,
+      "grad_norm": 0.7188334465026855,
+      "learning_rate": 4.619132995606926e-06,
+      "loss": 0.6296,
+      "step": 3938
+    },
+    {
+      "epoch": 1.0911357340720222,
+      "grad_norm": 0.7091854214668274,
+      "learning_rate": 4.618939701518032e-06,
+      "loss": 0.6647,
+      "step": 3939
+    },
+    {
+      "epoch": 1.0914127423822715,
+      "grad_norm": 0.7106387615203857,
+      "learning_rate": 4.618746362438636e-06,
+      "loss": 0.6711,
+      "step": 3940
+    },
+    {
+      "epoch": 1.0916897506925207,
+      "grad_norm": 0.6800446510314941,
+      "learning_rate": 4.618552978372844e-06,
+      "loss": 0.6541,
+      "step": 3941
+    },
+    {
+      "epoch": 1.0919667590027702,
+      "grad_norm": 0.6591231822967529,
+      "learning_rate": 4.618359549324762e-06,
+      "loss": 0.6561,
+      "step": 3942
+    },
+    {
+      "epoch": 1.0922437673130194,
+      "grad_norm": 0.6881027221679688,
+      "learning_rate": 4.618166075298498e-06,
+      "loss": 0.6558,
+      "step": 3943
+    },
+    {
+      "epoch": 1.0925207756232687,
+      "grad_norm": 0.7059905529022217,
+      "learning_rate": 4.617972556298158e-06,
+      "loss": 0.6297,
+      "step": 3944
+    },
+    {
+      "epoch": 1.0927977839335181,
+      "grad_norm": 0.695412278175354,
+      "learning_rate": 4.617778992327852e-06,
+      "loss": 0.6747,
+      "step": 3945
+    },
+    {
+      "epoch": 1.0930747922437674,
+      "grad_norm": 0.6626036763191223,
+      "learning_rate": 4.61758538339169e-06,
+      "loss": 0.6412,
+      "step": 3946
+    },
+    {
+      "epoch": 1.0933518005540166,
+      "grad_norm": 0.7440068125724792,
+      "learning_rate": 4.617391729493783e-06,
+      "loss": 0.6632,
+      "step": 3947
+    },
+    {
+      "epoch": 1.0936288088642658,
+      "grad_norm": 0.6759104132652283,
+      "learning_rate": 4.617198030638242e-06,
+      "loss": 0.5918,
+      "step": 3948
+    },
+    {
+      "epoch": 1.0939058171745153,
+      "grad_norm": 0.6452819108963013,
+      "learning_rate": 4.61700428682918e-06,
+      "loss": 0.615,
+      "step": 3949
+    },
+    {
+      "epoch": 1.0941828254847645,
+      "grad_norm": 0.6776657700538635,
+      "learning_rate": 4.6168104980707105e-06,
+      "loss": 0.6719,
+      "step": 3950
+    },
+    {
+      "epoch": 1.0944598337950138,
+      "grad_norm": 0.7503976821899414,
+      "learning_rate": 4.616616664366949e-06,
+      "loss": 0.6442,
+      "step": 3951
+    },
+    {
+      "epoch": 1.0947368421052632,
+      "grad_norm": 0.7288950085639954,
+      "learning_rate": 4.61642278572201e-06,
+      "loss": 0.6398,
+      "step": 3952
+    },
+    {
+      "epoch": 1.0950138504155125,
+      "grad_norm": 0.6706618070602417,
+      "learning_rate": 4.616228862140011e-06,
+      "loss": 0.6616,
+      "step": 3953
+    },
+    {
+      "epoch": 1.0952908587257617,
+      "grad_norm": 0.6500645875930786,
+      "learning_rate": 4.616034893625069e-06,
+      "loss": 0.6003,
+      "step": 3954
+    },
+    {
+      "epoch": 1.0955678670360112,
+      "grad_norm": 0.7065473794937134,
+      "learning_rate": 4.6158408801813015e-06,
+      "loss": 0.663,
+      "step": 3955
+    },
+    {
+      "epoch": 1.0958448753462604,
+      "grad_norm": 0.6544113755226135,
+      "learning_rate": 4.6156468218128296e-06,
+      "loss": 0.6369,
+      "step": 3956
+    },
+    {
+      "epoch": 1.0961218836565096,
+      "grad_norm": 0.6566190719604492,
+      "learning_rate": 4.615452718523773e-06,
+      "loss": 0.6221,
+      "step": 3957
+    },
+    {
+      "epoch": 1.096398891966759,
+      "grad_norm": 0.6906711459159851,
+      "learning_rate": 4.615258570318253e-06,
+      "loss": 0.6615,
+      "step": 3958
+    },
+    {
+      "epoch": 1.0966759002770083,
+      "grad_norm": 0.7138991951942444,
+      "learning_rate": 4.615064377200391e-06,
+      "loss": 0.6421,
+      "step": 3959
+    },
+    {
+      "epoch": 1.0969529085872576,
+      "grad_norm": 0.6587837934494019,
+      "learning_rate": 4.614870139174311e-06,
+      "loss": 0.6443,
+      "step": 3960
+    },
+    {
+      "epoch": 1.0972299168975068,
+      "grad_norm": 0.6848393082618713,
+      "learning_rate": 4.614675856244138e-06,
+      "loss": 0.5864,
+      "step": 3961
+    },
+    {
+      "epoch": 1.0975069252077563,
+      "grad_norm": 0.752826988697052,
+      "learning_rate": 4.614481528413995e-06,
+      "loss": 0.6359,
+      "step": 3962
+    },
+    {
+      "epoch": 1.0977839335180055,
+      "grad_norm": 0.7032961249351501,
+      "learning_rate": 4.614287155688011e-06,
+      "loss": 0.6627,
+      "step": 3963
+    },
+    {
+      "epoch": 1.0980609418282548,
+      "grad_norm": 0.6835402846336365,
+      "learning_rate": 4.614092738070311e-06,
+      "loss": 0.6927,
+      "step": 3964
+    },
+    {
+      "epoch": 1.0983379501385042,
+      "grad_norm": 0.6948668956756592,
+      "learning_rate": 4.613898275565022e-06,
+      "loss": 0.6491,
+      "step": 3965
+    },
+    {
+      "epoch": 1.0986149584487535,
+      "grad_norm": 0.6744138598442078,
+      "learning_rate": 4.613703768176276e-06,
+      "loss": 0.6413,
+      "step": 3966
+    },
+    {
+      "epoch": 1.0988919667590027,
+      "grad_norm": 0.7132924199104309,
+      "learning_rate": 4.613509215908201e-06,
+      "loss": 0.6307,
+      "step": 3967
+    },
+    {
+      "epoch": 1.0991689750692522,
+      "grad_norm": 0.6486527919769287,
+      "learning_rate": 4.613314618764927e-06,
+      "loss": 0.6537,
+      "step": 3968
+    },
+    {
+      "epoch": 1.0994459833795014,
+      "grad_norm": 0.693931519985199,
+      "learning_rate": 4.613119976750588e-06,
+      "loss": 0.6721,
+      "step": 3969
+    },
+    {
+      "epoch": 1.0997229916897506,
+      "grad_norm": 0.6713244318962097,
+      "learning_rate": 4.612925289869314e-06,
+      "loss": 0.6658,
+      "step": 3970
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6835699081420898,
+      "learning_rate": 4.6127305581252414e-06,
+      "loss": 0.6678,
+      "step": 3971
+    },
+    {
+      "epoch": 1.1002770083102493,
+      "grad_norm": 0.7053037285804749,
+      "learning_rate": 4.612535781522504e-06,
+      "loss": 0.6403,
+      "step": 3972
+    },
+    {
+      "epoch": 1.1005540166204986,
+      "grad_norm": 0.6976191997528076,
+      "learning_rate": 4.612340960065236e-06,
+      "loss": 0.6336,
+      "step": 3973
+    },
+    {
+      "epoch": 1.100831024930748,
+      "grad_norm": 0.716823935508728,
+      "learning_rate": 4.612146093757576e-06,
+      "loss": 0.6577,
+      "step": 3974
+    },
+    {
+      "epoch": 1.1011080332409973,
+      "grad_norm": 0.6851913928985596,
+      "learning_rate": 4.61195118260366e-06,
+      "loss": 0.6297,
+      "step": 3975
+    },
+    {
+      "epoch": 1.1013850415512465,
+      "grad_norm": 0.7345995306968689,
+      "learning_rate": 4.611756226607627e-06,
+      "loss": 0.6526,
+      "step": 3976
+    },
+    {
+      "epoch": 1.101662049861496,
+      "grad_norm": 0.696278989315033,
+      "learning_rate": 4.611561225773617e-06,
+      "loss": 0.6526,
+      "step": 3977
+    },
+    {
+      "epoch": 1.1019390581717452,
+      "grad_norm": 0.6616877317428589,
+      "learning_rate": 4.61136618010577e-06,
+      "loss": 0.6268,
+      "step": 3978
+    },
+    {
+      "epoch": 1.1022160664819944,
+      "grad_norm": 0.688875138759613,
+      "learning_rate": 4.611171089608226e-06,
+      "loss": 0.6434,
+      "step": 3979
+    },
+    {
+      "epoch": 1.1024930747922437,
+      "grad_norm": 0.7036260366439819,
+      "learning_rate": 4.610975954285129e-06,
+      "loss": 0.6641,
+      "step": 3980
+    },
+    {
+      "epoch": 1.1027700831024931,
+      "grad_norm": 0.6837557554244995,
+      "learning_rate": 4.610780774140623e-06,
+      "loss": 0.6485,
+      "step": 3981
+    },
+    {
+      "epoch": 1.1030470914127424,
+      "grad_norm": 0.7002822756767273,
+      "learning_rate": 4.61058554917885e-06,
+      "loss": 0.6616,
+      "step": 3982
+    },
+    {
+      "epoch": 1.1033240997229916,
+      "grad_norm": 0.6870222687721252,
+      "learning_rate": 4.610390279403955e-06,
+      "loss": 0.6418,
+      "step": 3983
+    },
+    {
+      "epoch": 1.103601108033241,
+      "grad_norm": 0.727690577507019,
+      "learning_rate": 4.610194964820086e-06,
+      "loss": 0.6673,
+      "step": 3984
+    },
+    {
+      "epoch": 1.1038781163434903,
+      "grad_norm": 0.7704971432685852,
+      "learning_rate": 4.609999605431389e-06,
+      "loss": 0.6677,
+      "step": 3985
+    },
+    {
+      "epoch": 1.1041551246537396,
+      "grad_norm": 0.7142701745033264,
+      "learning_rate": 4.6098042012420115e-06,
+      "loss": 0.6743,
+      "step": 3986
+    },
+    {
+      "epoch": 1.104432132963989,
+      "grad_norm": 0.6958649754524231,
+      "learning_rate": 4.609608752256103e-06,
+      "loss": 0.6094,
+      "step": 3987
+    },
+    {
+      "epoch": 1.1047091412742382,
+      "grad_norm": 0.6734898686408997,
+      "learning_rate": 4.609413258477814e-06,
+      "loss": 0.667,
+      "step": 3988
+    },
+    {
+      "epoch": 1.1049861495844875,
+      "grad_norm": 0.6905788779258728,
+      "learning_rate": 4.609217719911293e-06,
+      "loss": 0.6301,
+      "step": 3989
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.7093279957771301,
+      "learning_rate": 4.609022136560695e-06,
+      "loss": 0.6646,
+      "step": 3990
+    },
+    {
+      "epoch": 1.1055401662049862,
+      "grad_norm": 0.6776278018951416,
+      "learning_rate": 4.608826508430171e-06,
+      "loss": 0.5878,
+      "step": 3991
+    },
+    {
+      "epoch": 1.1058171745152354,
+      "grad_norm": 0.6801807284355164,
+      "learning_rate": 4.608630835523875e-06,
+      "loss": 0.6163,
+      "step": 3992
+    },
+    {
+      "epoch": 1.1060941828254847,
+      "grad_norm": 0.7167036533355713,
+      "learning_rate": 4.608435117845961e-06,
+      "loss": 0.6394,
+      "step": 3993
+    },
+    {
+      "epoch": 1.1063711911357341,
+      "grad_norm": 0.6812177896499634,
+      "learning_rate": 4.6082393554005855e-06,
+      "loss": 0.6594,
+      "step": 3994
+    },
+    {
+      "epoch": 1.1066481994459834,
+      "grad_norm": 0.7752406001091003,
+      "learning_rate": 4.608043548191904e-06,
+      "loss": 0.6783,
+      "step": 3995
+    },
+    {
+      "epoch": 1.1069252077562326,
+      "grad_norm": 0.6911202669143677,
+      "learning_rate": 4.607847696224076e-06,
+      "loss": 0.6464,
+      "step": 3996
+    },
+    {
+      "epoch": 1.107202216066482,
+      "grad_norm": 0.6800339818000793,
+      "learning_rate": 4.607651799501257e-06,
+      "loss": 0.6368,
+      "step": 3997
+    },
+    {
+      "epoch": 1.1074792243767313,
+      "grad_norm": 0.7221301794052124,
+      "learning_rate": 4.607455858027608e-06,
+      "loss": 0.6962,
+      "step": 3998
+    },
+    {
+      "epoch": 1.1077562326869805,
+      "grad_norm": 0.7182201743125916,
+      "learning_rate": 4.60725987180729e-06,
+      "loss": 0.6438,
+      "step": 3999
+    },
+    {
+      "epoch": 1.10803324099723,
+      "grad_norm": 0.6805998086929321,
+      "learning_rate": 4.6070638408444635e-06,
+      "loss": 0.5945,
+      "step": 4000
+    },
+    {
+      "epoch": 1.1083102493074792,
+      "grad_norm": 0.7249844074249268,
+      "learning_rate": 4.6068677651432915e-06,
+      "loss": 0.6334,
+      "step": 4001
+    },
+    {
+      "epoch": 1.1085872576177285,
+      "grad_norm": 0.6792013049125671,
+      "learning_rate": 4.606671644707935e-06,
+      "loss": 0.6646,
+      "step": 4002
+    },
+    {
+      "epoch": 1.108864265927978,
+      "grad_norm": 0.7077414989471436,
+      "learning_rate": 4.6064754795425606e-06,
+      "loss": 0.6588,
+      "step": 4003
+    },
+    {
+      "epoch": 1.1091412742382272,
+      "grad_norm": 0.6874167919158936,
+      "learning_rate": 4.606279269651332e-06,
+      "loss": 0.649,
+      "step": 4004
+    },
+    {
+      "epoch": 1.1094182825484764,
+      "grad_norm": 0.6752064228057861,
+      "learning_rate": 4.606083015038416e-06,
+      "loss": 0.6216,
+      "step": 4005
+    },
+    {
+      "epoch": 1.1096952908587259,
+      "grad_norm": 0.7003196477890015,
+      "learning_rate": 4.605886715707979e-06,
+      "loss": 0.6617,
+      "step": 4006
+    },
+    {
+      "epoch": 1.109972299168975,
+      "grad_norm": 0.727846622467041,
+      "learning_rate": 4.605690371664189e-06,
+      "loss": 0.6484,
+      "step": 4007
+    },
+    {
+      "epoch": 1.1102493074792243,
+      "grad_norm": 0.7528225779533386,
+      "learning_rate": 4.605493982911216e-06,
+      "loss": 0.7293,
+      "step": 4008
+    },
+    {
+      "epoch": 1.1105263157894736,
+      "grad_norm": 0.6964515447616577,
+      "learning_rate": 4.605297549453228e-06,
+      "loss": 0.6379,
+      "step": 4009
+    },
+    {
+      "epoch": 1.110803324099723,
+      "grad_norm": 0.7198163866996765,
+      "learning_rate": 4.605101071294397e-06,
+      "loss": 0.6376,
+      "step": 4010
+    },
+    {
+      "epoch": 1.1110803324099723,
+      "grad_norm": 0.6878595352172852,
+      "learning_rate": 4.604904548438894e-06,
+      "loss": 0.6473,
+      "step": 4011
+    },
+    {
+      "epoch": 1.1113573407202215,
+      "grad_norm": 0.6337695717811584,
+      "learning_rate": 4.604707980890892e-06,
+      "loss": 0.5738,
+      "step": 4012
+    },
+    {
+      "epoch": 1.111634349030471,
+      "grad_norm": 0.6804854869842529,
+      "learning_rate": 4.604511368654565e-06,
+      "loss": 0.605,
+      "step": 4013
+    },
+    {
+      "epoch": 1.1119113573407202,
+      "grad_norm": 0.6958677768707275,
+      "learning_rate": 4.604314711734088e-06,
+      "loss": 0.6795,
+      "step": 4014
+    },
+    {
+      "epoch": 1.1121883656509695,
+      "grad_norm": 0.7208279371261597,
+      "learning_rate": 4.604118010133634e-06,
+      "loss": 0.6608,
+      "step": 4015
+    },
+    {
+      "epoch": 1.112465373961219,
+      "grad_norm": 0.752122163772583,
+      "learning_rate": 4.6039212638573835e-06,
+      "loss": 0.6512,
+      "step": 4016
+    },
+    {
+      "epoch": 1.1127423822714682,
+      "grad_norm": 0.7201291918754578,
+      "learning_rate": 4.6037244729095104e-06,
+      "loss": 0.6476,
+      "step": 4017
+    },
+    {
+      "epoch": 1.1130193905817174,
+      "grad_norm": 0.7278525829315186,
+      "learning_rate": 4.603527637294195e-06,
+      "loss": 0.5995,
+      "step": 4018
+    },
+    {
+      "epoch": 1.1132963988919669,
+      "grad_norm": 0.7007999420166016,
+      "learning_rate": 4.603330757015616e-06,
+      "loss": 0.6716,
+      "step": 4019
+    },
+    {
+      "epoch": 1.113573407202216,
+      "grad_norm": 0.6690526008605957,
+      "learning_rate": 4.603133832077953e-06,
+      "loss": 0.6476,
+      "step": 4020
+    },
+    {
+      "epoch": 1.1138504155124653,
+      "grad_norm": 0.7289823889732361,
+      "learning_rate": 4.6029368624853885e-06,
+      "loss": 0.6441,
+      "step": 4021
+    },
+    {
+      "epoch": 1.1141274238227146,
+      "grad_norm": 0.674523651599884,
+      "learning_rate": 4.602739848242105e-06,
+      "loss": 0.626,
+      "step": 4022
+    },
+    {
+      "epoch": 1.114404432132964,
+      "grad_norm": 0.7001068592071533,
+      "learning_rate": 4.602542789352283e-06,
+      "loss": 0.6816,
+      "step": 4023
+    },
+    {
+      "epoch": 1.1146814404432133,
+      "grad_norm": 0.7051699161529541,
+      "learning_rate": 4.602345685820109e-06,
+      "loss": 0.6761,
+      "step": 4024
+    },
+    {
+      "epoch": 1.1149584487534625,
+      "grad_norm": 0.699550449848175,
+      "learning_rate": 4.602148537649768e-06,
+      "loss": 0.6442,
+      "step": 4025
+    },
+    {
+      "epoch": 1.115235457063712,
+      "grad_norm": 0.7217336893081665,
+      "learning_rate": 4.601951344845445e-06,
+      "loss": 0.6612,
+      "step": 4026
+    },
+    {
+      "epoch": 1.1155124653739612,
+      "grad_norm": 0.7049715518951416,
+      "learning_rate": 4.601754107411326e-06,
+      "loss": 0.5907,
+      "step": 4027
+    },
+    {
+      "epoch": 1.1157894736842104,
+      "grad_norm": 0.7184882164001465,
+      "learning_rate": 4.6015568253516005e-06,
+      "loss": 0.6234,
+      "step": 4028
+    },
+    {
+      "epoch": 1.11606648199446,
+      "grad_norm": 0.6838163733482361,
+      "learning_rate": 4.601359498670457e-06,
+      "loss": 0.6032,
+      "step": 4029
+    },
+    {
+      "epoch": 1.1163434903047091,
+      "grad_norm": 0.6776624321937561,
+      "learning_rate": 4.601162127372085e-06,
+      "loss": 0.664,
+      "step": 4030
+    },
+    {
+      "epoch": 1.1166204986149584,
+      "grad_norm": 0.6719553470611572,
+      "learning_rate": 4.600964711460675e-06,
+      "loss": 0.6842,
+      "step": 4031
+    },
+    {
+      "epoch": 1.1168975069252078,
+      "grad_norm": 0.7092738747596741,
+      "learning_rate": 4.60076725094042e-06,
+      "loss": 0.6525,
+      "step": 4032
+    },
+    {
+      "epoch": 1.117174515235457,
+      "grad_norm": 0.6958156824111938,
+      "learning_rate": 4.600569745815511e-06,
+      "loss": 0.6388,
+      "step": 4033
+    },
+    {
+      "epoch": 1.1174515235457063,
+      "grad_norm": 0.7308716773986816,
+      "learning_rate": 4.600372196090142e-06,
+      "loss": 0.659,
+      "step": 4034
+    },
+    {
+      "epoch": 1.1177285318559558,
+      "grad_norm": 0.6616610884666443,
+      "learning_rate": 4.600174601768506e-06,
+      "loss": 0.6024,
+      "step": 4035
+    },
+    {
+      "epoch": 1.118005540166205,
+      "grad_norm": 0.6373092532157898,
+      "learning_rate": 4.599976962854802e-06,
+      "loss": 0.6173,
+      "step": 4036
+    },
+    {
+      "epoch": 1.1182825484764543,
+      "grad_norm": 0.6987490653991699,
+      "learning_rate": 4.599779279353224e-06,
+      "loss": 0.6275,
+      "step": 4037
+    },
+    {
+      "epoch": 1.1185595567867037,
+      "grad_norm": 0.6965712904930115,
+      "learning_rate": 4.599581551267969e-06,
+      "loss": 0.6338,
+      "step": 4038
+    },
+    {
+      "epoch": 1.118836565096953,
+      "grad_norm": 0.6864861249923706,
+      "learning_rate": 4.5993837786032355e-06,
+      "loss": 0.6241,
+      "step": 4039
+    },
+    {
+      "epoch": 1.1191135734072022,
+      "grad_norm": 0.717839241027832,
+      "learning_rate": 4.599185961363224e-06,
+      "loss": 0.6982,
+      "step": 4040
+    },
+    {
+      "epoch": 1.1193905817174514,
+      "grad_norm": 0.7153016924858093,
+      "learning_rate": 4.598988099552133e-06,
+      "loss": 0.687,
+      "step": 4041
+    },
+    {
+      "epoch": 1.1196675900277009,
+      "grad_norm": 0.7333785891532898,
+      "learning_rate": 4.5987901931741655e-06,
+      "loss": 0.6703,
+      "step": 4042
+    },
+    {
+      "epoch": 1.1199445983379501,
+      "grad_norm": 0.6576941609382629,
+      "learning_rate": 4.598592242233522e-06,
+      "loss": 0.6394,
+      "step": 4043
+    },
+    {
+      "epoch": 1.1202216066481994,
+      "grad_norm": 0.6860183477401733,
+      "learning_rate": 4.598394246734406e-06,
+      "loss": 0.6788,
+      "step": 4044
+    },
+    {
+      "epoch": 1.1204986149584488,
+      "grad_norm": 0.6848655343055725,
+      "learning_rate": 4.598196206681021e-06,
+      "loss": 0.6213,
+      "step": 4045
+    },
+    {
+      "epoch": 1.120775623268698,
+      "grad_norm": 0.6786518692970276,
+      "learning_rate": 4.597998122077573e-06,
+      "loss": 0.6325,
+      "step": 4046
+    },
+    {
+      "epoch": 1.1210526315789473,
+      "grad_norm": 0.7070546746253967,
+      "learning_rate": 4.597799992928266e-06,
+      "loss": 0.67,
+      "step": 4047
+    },
+    {
+      "epoch": 1.1213296398891968,
+      "grad_norm": 0.6792629361152649,
+      "learning_rate": 4.597601819237309e-06,
+      "loss": 0.6638,
+      "step": 4048
+    },
+    {
+      "epoch": 1.121606648199446,
+      "grad_norm": 0.6677889823913574,
+      "learning_rate": 4.597403601008908e-06,
+      "loss": 0.6391,
+      "step": 4049
+    },
+    {
+      "epoch": 1.1218836565096952,
+      "grad_norm": 0.7167168259620667,
+      "learning_rate": 4.597205338247274e-06,
+      "loss": 0.6547,
+      "step": 4050
+    },
+    {
+      "epoch": 1.1221606648199447,
+      "grad_norm": 0.6899300217628479,
+      "learning_rate": 4.597007030956612e-06,
+      "loss": 0.6093,
+      "step": 4051
+    },
+    {
+      "epoch": 1.122437673130194,
+      "grad_norm": 0.6876311898231506,
+      "learning_rate": 4.596808679141138e-06,
+      "loss": 0.6596,
+      "step": 4052
+    },
+    {
+      "epoch": 1.1227146814404432,
+      "grad_norm": 0.6713271737098694,
+      "learning_rate": 4.596610282805061e-06,
+      "loss": 0.6324,
+      "step": 4053
+    },
+    {
+      "epoch": 1.1229916897506924,
+      "grad_norm": 0.7074865698814392,
+      "learning_rate": 4.5964118419525925e-06,
+      "loss": 0.605,
+      "step": 4054
+    },
+    {
+      "epoch": 1.1232686980609419,
+      "grad_norm": 0.708007276058197,
+      "learning_rate": 4.596213356587948e-06,
+      "loss": 0.7101,
+      "step": 4055
+    },
+    {
+      "epoch": 1.123545706371191,
+      "grad_norm": 0.6596668362617493,
+      "learning_rate": 4.59601482671534e-06,
+      "loss": 0.6427,
+      "step": 4056
+    },
+    {
+      "epoch": 1.1238227146814403,
+      "grad_norm": 0.6609944701194763,
+      "learning_rate": 4.5958162523389845e-06,
+      "loss": 0.6166,
+      "step": 4057
+    },
+    {
+      "epoch": 1.1240997229916898,
+      "grad_norm": 0.6856138110160828,
+      "learning_rate": 4.595617633463099e-06,
+      "loss": 0.6993,
+      "step": 4058
+    },
+    {
+      "epoch": 1.124376731301939,
+      "grad_norm": 0.6793034672737122,
+      "learning_rate": 4.595418970091898e-06,
+      "loss": 0.6075,
+      "step": 4059
+    },
+    {
+      "epoch": 1.1246537396121883,
+      "grad_norm": 0.6801674962043762,
+      "learning_rate": 4.5952202622296015e-06,
+      "loss": 0.6518,
+      "step": 4060
+    },
+    {
+      "epoch": 1.1249307479224377,
+      "grad_norm": 0.7081454992294312,
+      "learning_rate": 4.595021509880429e-06,
+      "loss": 0.6733,
+      "step": 4061
+    },
+    {
+      "epoch": 1.125207756232687,
+      "grad_norm": 0.7073140740394592,
+      "learning_rate": 4.594822713048599e-06,
+      "loss": 0.6269,
+      "step": 4062
+    },
+    {
+      "epoch": 1.1254847645429362,
+      "grad_norm": 0.7026985287666321,
+      "learning_rate": 4.594623871738334e-06,
+      "loss": 0.6626,
+      "step": 4063
+    },
+    {
+      "epoch": 1.1257617728531857,
+      "grad_norm": 0.6450572609901428,
+      "learning_rate": 4.594424985953855e-06,
+      "loss": 0.6018,
+      "step": 4064
+    },
+    {
+      "epoch": 1.126038781163435,
+      "grad_norm": 0.6867948770523071,
+      "learning_rate": 4.594226055699384e-06,
+      "loss": 0.6411,
+      "step": 4065
+    },
+    {
+      "epoch": 1.1263157894736842,
+      "grad_norm": 0.6572390198707581,
+      "learning_rate": 4.594027080979147e-06,
+      "loss": 0.6537,
+      "step": 4066
+    },
+    {
+      "epoch": 1.1265927977839336,
+      "grad_norm": 0.6882357597351074,
+      "learning_rate": 4.593828061797366e-06,
+      "loss": 0.6781,
+      "step": 4067
+    },
+    {
+      "epoch": 1.1268698060941829,
+      "grad_norm": 0.7123607397079468,
+      "learning_rate": 4.59362899815827e-06,
+      "loss": 0.6529,
+      "step": 4068
+    },
+    {
+      "epoch": 1.127146814404432,
+      "grad_norm": 0.7132445573806763,
+      "learning_rate": 4.593429890066083e-06,
+      "loss": 0.6516,
+      "step": 4069
+    },
+    {
+      "epoch": 1.1274238227146816,
+      "grad_norm": 0.6925243735313416,
+      "learning_rate": 4.593230737525032e-06,
+      "loss": 0.6489,
+      "step": 4070
+    },
+    {
+      "epoch": 1.1277008310249308,
+      "grad_norm": 0.7030454874038696,
+      "learning_rate": 4.593031540539349e-06,
+      "loss": 0.6918,
+      "step": 4071
+    },
+    {
+      "epoch": 1.12797783933518,
+      "grad_norm": 0.7076012492179871,
+      "learning_rate": 4.59283229911326e-06,
+      "loss": 0.6412,
+      "step": 4072
+    },
+    {
+      "epoch": 1.1282548476454293,
+      "grad_norm": 0.6813455820083618,
+      "learning_rate": 4.592633013250997e-06,
+      "loss": 0.5779,
+      "step": 4073
+    },
+    {
+      "epoch": 1.1285318559556787,
+      "grad_norm": 0.7080217003822327,
+      "learning_rate": 4.592433682956792e-06,
+      "loss": 0.6446,
+      "step": 4074
+    },
+    {
+      "epoch": 1.128808864265928,
+      "grad_norm": 0.699512243270874,
+      "learning_rate": 4.592234308234876e-06,
+      "loss": 0.6654,
+      "step": 4075
+    },
+    {
+      "epoch": 1.1290858725761772,
+      "grad_norm": 0.7339920997619629,
+      "learning_rate": 4.592034889089482e-06,
+      "loss": 0.6161,
+      "step": 4076
+    },
+    {
+      "epoch": 1.1293628808864267,
+      "grad_norm": 0.6406992077827454,
+      "learning_rate": 4.591835425524845e-06,
+      "loss": 0.5995,
+      "step": 4077
+    },
+    {
+      "epoch": 1.129639889196676,
+      "grad_norm": 0.6801450252532959,
+      "learning_rate": 4.591635917545201e-06,
+      "loss": 0.6729,
+      "step": 4078
+    },
+    {
+      "epoch": 1.1299168975069251,
+      "grad_norm": 0.7015729546546936,
+      "learning_rate": 4.5914363651547846e-06,
+      "loss": 0.6838,
+      "step": 4079
+    },
+    {
+      "epoch": 1.1301939058171746,
+      "grad_norm": 0.6717569828033447,
+      "learning_rate": 4.591236768357833e-06,
+      "loss": 0.5861,
+      "step": 4080
+    },
+    {
+      "epoch": 1.1304709141274238,
+      "grad_norm": 0.7090352177619934,
+      "learning_rate": 4.591037127158584e-06,
+      "loss": 0.644,
+      "step": 4081
+    },
+    {
+      "epoch": 1.130747922437673,
+      "grad_norm": 0.662742555141449,
+      "learning_rate": 4.590837441561277e-06,
+      "loss": 0.6224,
+      "step": 4082
+    },
+    {
+      "epoch": 1.1310249307479223,
+      "grad_norm": 0.7132098078727722,
+      "learning_rate": 4.590637711570152e-06,
+      "loss": 0.6407,
+      "step": 4083
+    },
+    {
+      "epoch": 1.1313019390581718,
+      "grad_norm": 0.7207354307174683,
+      "learning_rate": 4.590437937189449e-06,
+      "loss": 0.6605,
+      "step": 4084
+    },
+    {
+      "epoch": 1.131578947368421,
+      "grad_norm": 0.6709021329879761,
+      "learning_rate": 4.5902381184234116e-06,
+      "loss": 0.6759,
+      "step": 4085
+    },
+    {
+      "epoch": 1.1318559556786703,
+      "grad_norm": 0.6779177188873291,
+      "learning_rate": 4.590038255276279e-06,
+      "loss": 0.6342,
+      "step": 4086
+    },
+    {
+      "epoch": 1.1321329639889197,
+      "grad_norm": 0.6814398169517517,
+      "learning_rate": 4.589838347752298e-06,
+      "loss": 0.6431,
+      "step": 4087
+    },
+    {
+      "epoch": 1.132409972299169,
+      "grad_norm": 0.6873770952224731,
+      "learning_rate": 4.589638395855712e-06,
+      "loss": 0.6253,
+      "step": 4088
+    },
+    {
+      "epoch": 1.1326869806094182,
+      "grad_norm": 0.6760203242301941,
+      "learning_rate": 4.589438399590766e-06,
+      "loss": 0.6744,
+      "step": 4089
+    },
+    {
+      "epoch": 1.1329639889196677,
+      "grad_norm": 0.7229450345039368,
+      "learning_rate": 4.589238358961707e-06,
+      "loss": 0.6845,
+      "step": 4090
+    },
+    {
+      "epoch": 1.133240997229917,
+      "grad_norm": 0.7091704607009888,
+      "learning_rate": 4.589038273972783e-06,
+      "loss": 0.5978,
+      "step": 4091
+    },
+    {
+      "epoch": 1.1335180055401661,
+      "grad_norm": 0.6956828236579895,
+      "learning_rate": 4.588838144628241e-06,
+      "loss": 0.646,
+      "step": 4092
+    },
+    {
+      "epoch": 1.1337950138504156,
+      "grad_norm": 0.6587649583816528,
+      "learning_rate": 4.588637970932332e-06,
+      "loss": 0.6249,
+      "step": 4093
+    },
+    {
+      "epoch": 1.1340720221606648,
+      "grad_norm": 0.690669596195221,
+      "learning_rate": 4.588437752889305e-06,
+      "loss": 0.6517,
+      "step": 4094
+    },
+    {
+      "epoch": 1.134349030470914,
+      "grad_norm": 0.6745144724845886,
+      "learning_rate": 4.588237490503409e-06,
+      "loss": 0.6255,
+      "step": 4095
+    },
+    {
+      "epoch": 1.1346260387811635,
+      "grad_norm": 0.6996001601219177,
+      "learning_rate": 4.5880371837789e-06,
+      "loss": 0.6747,
+      "step": 4096
+    },
+    {
+      "epoch": 1.1349030470914128,
+      "grad_norm": 0.7193899154663086,
+      "learning_rate": 4.587836832720029e-06,
+      "loss": 0.6316,
+      "step": 4097
+    },
+    {
+      "epoch": 1.135180055401662,
+      "grad_norm": 0.716244637966156,
+      "learning_rate": 4.587636437331051e-06,
+      "loss": 0.6515,
+      "step": 4098
+    },
+    {
+      "epoch": 1.1354570637119115,
+      "grad_norm": 0.6850689053535461,
+      "learning_rate": 4.587435997616219e-06,
+      "loss": 0.6456,
+      "step": 4099
+    },
+    {
+      "epoch": 1.1357340720221607,
+      "grad_norm": 0.6978815197944641,
+      "learning_rate": 4.587235513579791e-06,
+      "loss": 0.6909,
+      "step": 4100
+    },
+    {
+      "epoch": 1.13601108033241,
+      "grad_norm": 0.692414402961731,
+      "learning_rate": 4.587034985226022e-06,
+      "loss": 0.6706,
+      "step": 4101
+    },
+    {
+      "epoch": 1.1362880886426594,
+      "grad_norm": 0.7044286131858826,
+      "learning_rate": 4.586834412559171e-06,
+      "loss": 0.651,
+      "step": 4102
+    },
+    {
+      "epoch": 1.1365650969529086,
+      "grad_norm": 0.7050232887268066,
+      "learning_rate": 4.586633795583496e-06,
+      "loss": 0.6158,
+      "step": 4103
+    },
+    {
+      "epoch": 1.1368421052631579,
+      "grad_norm": 0.7236217856407166,
+      "learning_rate": 4.586433134303257e-06,
+      "loss": 0.6257,
+      "step": 4104
+    },
+    {
+      "epoch": 1.1371191135734071,
+      "grad_norm": 0.7170664668083191,
+      "learning_rate": 4.5862324287227145e-06,
+      "loss": 0.6654,
+      "step": 4105
+    },
+    {
+      "epoch": 1.1373961218836566,
+      "grad_norm": 0.6874409317970276,
+      "learning_rate": 4.58603167884613e-06,
+      "loss": 0.5662,
+      "step": 4106
+    },
+    {
+      "epoch": 1.1376731301939058,
+      "grad_norm": 0.6953471302986145,
+      "learning_rate": 4.585830884677765e-06,
+      "loss": 0.6646,
+      "step": 4107
+    },
+    {
+      "epoch": 1.137950138504155,
+      "grad_norm": 0.7238764762878418,
+      "learning_rate": 4.585630046221884e-06,
+      "loss": 0.6542,
+      "step": 4108
+    },
+    {
+      "epoch": 1.1382271468144045,
+      "grad_norm": 0.6525726914405823,
+      "learning_rate": 4.585429163482752e-06,
+      "loss": 0.6734,
+      "step": 4109
+    },
+    {
+      "epoch": 1.1385041551246537,
+      "grad_norm": 0.682162880897522,
+      "learning_rate": 4.585228236464632e-06,
+      "loss": 0.614,
+      "step": 4110
+    },
+    {
+      "epoch": 1.138781163434903,
+      "grad_norm": 0.6903233528137207,
+      "learning_rate": 4.585027265171793e-06,
+      "loss": 0.6457,
+      "step": 4111
+    },
+    {
+      "epoch": 1.1390581717451524,
+      "grad_norm": 0.713100016117096,
+      "learning_rate": 4.5848262496084995e-06,
+      "loss": 0.6154,
+      "step": 4112
+    },
+    {
+      "epoch": 1.1393351800554017,
+      "grad_norm": 0.6778297424316406,
+      "learning_rate": 4.58462518977902e-06,
+      "loss": 0.5965,
+      "step": 4113
+    },
+    {
+      "epoch": 1.139612188365651,
+      "grad_norm": 0.6337829232215881,
+      "learning_rate": 4.584424085687625e-06,
+      "loss": 0.5248,
+      "step": 4114
+    },
+    {
+      "epoch": 1.1398891966759002,
+      "grad_norm": 0.693355143070221,
+      "learning_rate": 4.584222937338584e-06,
+      "loss": 0.6749,
+      "step": 4115
+    },
+    {
+      "epoch": 1.1401662049861496,
+      "grad_norm": 0.7208781838417053,
+      "learning_rate": 4.584021744736167e-06,
+      "loss": 0.6706,
+      "step": 4116
+    },
+    {
+      "epoch": 1.1404432132963989,
+      "grad_norm": 0.7035034894943237,
+      "learning_rate": 4.583820507884646e-06,
+      "loss": 0.5913,
+      "step": 4117
+    },
+    {
+      "epoch": 1.140720221606648,
+      "grad_norm": 0.7289274334907532,
+      "learning_rate": 4.583619226788294e-06,
+      "loss": 0.6392,
+      "step": 4118
+    },
+    {
+      "epoch": 1.1409972299168976,
+      "grad_norm": 0.7026898264884949,
+      "learning_rate": 4.583417901451386e-06,
+      "loss": 0.6338,
+      "step": 4119
+    },
+    {
+      "epoch": 1.1412742382271468,
+      "grad_norm": 0.6910617351531982,
+      "learning_rate": 4.583216531878195e-06,
+      "loss": 0.6512,
+      "step": 4120
+    },
+    {
+      "epoch": 1.141551246537396,
+      "grad_norm": 0.6606723666191101,
+      "learning_rate": 4.5830151180729964e-06,
+      "loss": 0.6401,
+      "step": 4121
+    },
+    {
+      "epoch": 1.1418282548476455,
+      "grad_norm": 0.7039617896080017,
+      "learning_rate": 4.582813660040068e-06,
+      "loss": 0.6573,
+      "step": 4122
+    },
+    {
+      "epoch": 1.1421052631578947,
+      "grad_norm": 0.7446121573448181,
+      "learning_rate": 4.582612157783687e-06,
+      "loss": 0.6593,
+      "step": 4123
+    },
+    {
+      "epoch": 1.142382271468144,
+      "grad_norm": 0.6769001483917236,
+      "learning_rate": 4.58241061130813e-06,
+      "loss": 0.6238,
+      "step": 4124
+    },
+    {
+      "epoch": 1.1426592797783934,
+      "grad_norm": 0.7253504395484924,
+      "learning_rate": 4.582209020617679e-06,
+      "loss": 0.6597,
+      "step": 4125
+    },
+    {
+      "epoch": 1.1429362880886427,
+      "grad_norm": 0.7634405493736267,
+      "learning_rate": 4.582007385716614e-06,
+      "loss": 0.6808,
+      "step": 4126
+    },
+    {
+      "epoch": 1.143213296398892,
+      "grad_norm": 0.7294143438339233,
+      "learning_rate": 4.581805706609213e-06,
+      "loss": 0.614,
+      "step": 4127
+    },
+    {
+      "epoch": 1.1434903047091414,
+      "grad_norm": 0.700397253036499,
+      "learning_rate": 4.5816039832997625e-06,
+      "loss": 0.662,
+      "step": 4128
+    },
+    {
+      "epoch": 1.1437673130193906,
+      "grad_norm": 0.747981071472168,
+      "learning_rate": 4.581402215792543e-06,
+      "loss": 0.6671,
+      "step": 4129
+    },
+    {
+      "epoch": 1.1440443213296398,
+      "grad_norm": 0.6845051646232605,
+      "learning_rate": 4.581200404091839e-06,
+      "loss": 0.6468,
+      "step": 4130
+    },
+    {
+      "epoch": 1.1443213296398893,
+      "grad_norm": 0.6928357481956482,
+      "learning_rate": 4.580998548201936e-06,
+      "loss": 0.674,
+      "step": 4131
+    },
+    {
+      "epoch": 1.1445983379501385,
+      "grad_norm": 0.6433817744255066,
+      "learning_rate": 4.5807966481271184e-06,
+      "loss": 0.6253,
+      "step": 4132
+    },
+    {
+      "epoch": 1.1448753462603878,
+      "grad_norm": 0.6889892816543579,
+      "learning_rate": 4.5805947038716754e-06,
+      "loss": 0.6596,
+      "step": 4133
+    },
+    {
+      "epoch": 1.1451523545706372,
+      "grad_norm": 0.6575981974601746,
+      "learning_rate": 4.580392715439893e-06,
+      "loss": 0.633,
+      "step": 4134
+    },
+    {
+      "epoch": 1.1454293628808865,
+      "grad_norm": 0.6686013340950012,
+      "learning_rate": 4.580190682836061e-06,
+      "loss": 0.6355,
+      "step": 4135
+    },
+    {
+      "epoch": 1.1457063711911357,
+      "grad_norm": 0.6737796068191528,
+      "learning_rate": 4.579988606064468e-06,
+      "loss": 0.6032,
+      "step": 4136
+    },
+    {
+      "epoch": 1.145983379501385,
+      "grad_norm": 0.7231971621513367,
+      "learning_rate": 4.579786485129405e-06,
+      "loss": 0.6805,
+      "step": 4137
+    },
+    {
+      "epoch": 1.1462603878116344,
+      "grad_norm": 0.6811356544494629,
+      "learning_rate": 4.579584320035164e-06,
+      "loss": 0.6046,
+      "step": 4138
+    },
+    {
+      "epoch": 1.1465373961218837,
+      "grad_norm": 0.7172678112983704,
+      "learning_rate": 4.579382110786038e-06,
+      "loss": 0.6553,
+      "step": 4139
+    },
+    {
+      "epoch": 1.146814404432133,
+      "grad_norm": 0.6764629483222961,
+      "learning_rate": 4.579179857386319e-06,
+      "loss": 0.6216,
+      "step": 4140
+    },
+    {
+      "epoch": 1.1470914127423824,
+      "grad_norm": 0.6475736498832703,
+      "learning_rate": 4.578977559840302e-06,
+      "loss": 0.6237,
+      "step": 4141
+    },
+    {
+      "epoch": 1.1473684210526316,
+      "grad_norm": 0.7342414855957031,
+      "learning_rate": 4.578775218152282e-06,
+      "loss": 0.6655,
+      "step": 4142
+    },
+    {
+      "epoch": 1.1476454293628808,
+      "grad_norm": 0.7185686826705933,
+      "learning_rate": 4.578572832326555e-06,
+      "loss": 0.6126,
+      "step": 4143
+    },
+    {
+      "epoch": 1.14792243767313,
+      "grad_norm": 0.7499321699142456,
+      "learning_rate": 4.57837040236742e-06,
+      "loss": 0.6875,
+      "step": 4144
+    },
+    {
+      "epoch": 1.1481994459833795,
+      "grad_norm": 0.683867335319519,
+      "learning_rate": 4.578167928279174e-06,
+      "loss": 0.6397,
+      "step": 4145
+    },
+    {
+      "epoch": 1.1484764542936288,
+      "grad_norm": 0.6831228137016296,
+      "learning_rate": 4.577965410066114e-06,
+      "loss": 0.662,
+      "step": 4146
+    },
+    {
+      "epoch": 1.148753462603878,
+      "grad_norm": 0.686525821685791,
+      "learning_rate": 4.577762847732543e-06,
+      "loss": 0.6128,
+      "step": 4147
+    },
+    {
+      "epoch": 1.1490304709141275,
+      "grad_norm": 0.6958313584327698,
+      "learning_rate": 4.57756024128276e-06,
+      "loss": 0.6187,
+      "step": 4148
+    },
+    {
+      "epoch": 1.1493074792243767,
+      "grad_norm": 0.6785929799079895,
+      "learning_rate": 4.577357590721069e-06,
+      "loss": 0.6062,
+      "step": 4149
+    },
+    {
+      "epoch": 1.149584487534626,
+      "grad_norm": 0.7083978056907654,
+      "learning_rate": 4.57715489605177e-06,
+      "loss": 0.6558,
+      "step": 4150
+    },
+    {
+      "epoch": 1.1498614958448754,
+      "grad_norm": 0.6996467113494873,
+      "learning_rate": 4.576952157279169e-06,
+      "loss": 0.6411,
+      "step": 4151
+    },
+    {
+      "epoch": 1.1501385041551246,
+      "grad_norm": 0.7247840166091919,
+      "learning_rate": 4.576749374407569e-06,
+      "loss": 0.6825,
+      "step": 4152
+    },
+    {
+      "epoch": 1.1504155124653739,
+      "grad_norm": 0.712798535823822,
+      "learning_rate": 4.576546547441276e-06,
+      "loss": 0.6215,
+      "step": 4153
+    },
+    {
+      "epoch": 1.1506925207756233,
+      "grad_norm": 0.7341089248657227,
+      "learning_rate": 4.576343676384598e-06,
+      "loss": 0.6737,
+      "step": 4154
+    },
+    {
+      "epoch": 1.1509695290858726,
+      "grad_norm": 0.6770998239517212,
+      "learning_rate": 4.576140761241841e-06,
+      "loss": 0.5728,
+      "step": 4155
+    },
+    {
+      "epoch": 1.1512465373961218,
+      "grad_norm": 0.7066571712493896,
+      "learning_rate": 4.575937802017313e-06,
+      "loss": 0.6122,
+      "step": 4156
+    },
+    {
+      "epoch": 1.1515235457063713,
+      "grad_norm": 0.7351990938186646,
+      "learning_rate": 4.575734798715325e-06,
+      "loss": 0.6717,
+      "step": 4157
+    },
+    {
+      "epoch": 1.1518005540166205,
+      "grad_norm": 0.6787825226783752,
+      "learning_rate": 4.5755317513401855e-06,
+      "loss": 0.6366,
+      "step": 4158
+    },
+    {
+      "epoch": 1.1520775623268698,
+      "grad_norm": 0.6576148271560669,
+      "learning_rate": 4.575328659896207e-06,
+      "loss": 0.6444,
+      "step": 4159
+    },
+    {
+      "epoch": 1.1523545706371192,
+      "grad_norm": 0.6681926846504211,
+      "learning_rate": 4.575125524387702e-06,
+      "loss": 0.6042,
+      "step": 4160
+    },
+    {
+      "epoch": 1.1526315789473685,
+      "grad_norm": 0.67631596326828,
+      "learning_rate": 4.574922344818982e-06,
+      "loss": 0.6265,
+      "step": 4161
+    },
+    {
+      "epoch": 1.1529085872576177,
+      "grad_norm": 0.6682254076004028,
+      "learning_rate": 4.574719121194362e-06,
+      "loss": 0.6007,
+      "step": 4162
+    },
+    {
+      "epoch": 1.1531855955678671,
+      "grad_norm": 0.6621627807617188,
+      "learning_rate": 4.574515853518157e-06,
+      "loss": 0.6509,
+      "step": 4163
+    },
+    {
+      "epoch": 1.1534626038781164,
+      "grad_norm": 0.7396254539489746,
+      "learning_rate": 4.574312541794683e-06,
+      "loss": 0.6482,
+      "step": 4164
+    },
+    {
+      "epoch": 1.1537396121883656,
+      "grad_norm": 0.7040379643440247,
+      "learning_rate": 4.574109186028255e-06,
+      "loss": 0.6774,
+      "step": 4165
+    },
+    {
+      "epoch": 1.1540166204986149,
+      "grad_norm": 0.7313653230667114,
+      "learning_rate": 4.573905786223194e-06,
+      "loss": 0.6583,
+      "step": 4166
+    },
+    {
+      "epoch": 1.1542936288088643,
+      "grad_norm": 0.6954412460327148,
+      "learning_rate": 4.573702342383816e-06,
+      "loss": 0.6682,
+      "step": 4167
+    },
+    {
+      "epoch": 1.1545706371191136,
+      "grad_norm": 0.7022963166236877,
+      "learning_rate": 4.573498854514442e-06,
+      "loss": 0.6149,
+      "step": 4168
+    },
+    {
+      "epoch": 1.1548476454293628,
+      "grad_norm": 0.6929730176925659,
+      "learning_rate": 4.5732953226193925e-06,
+      "loss": 0.5669,
+      "step": 4169
+    },
+    {
+      "epoch": 1.1551246537396123,
+      "grad_norm": 0.688026487827301,
+      "learning_rate": 4.573091746702988e-06,
+      "loss": 0.6446,
+      "step": 4170
+    },
+    {
+      "epoch": 1.1554016620498615,
+      "grad_norm": 0.7175707221031189,
+      "learning_rate": 4.572888126769552e-06,
+      "loss": 0.6663,
+      "step": 4171
+    },
+    {
+      "epoch": 1.1556786703601107,
+      "grad_norm": 0.6707695126533508,
+      "learning_rate": 4.5726844628234075e-06,
+      "loss": 0.6774,
+      "step": 4172
+    },
+    {
+      "epoch": 1.1559556786703602,
+      "grad_norm": 0.6873745322227478,
+      "learning_rate": 4.572480754868879e-06,
+      "loss": 0.6272,
+      "step": 4173
+    },
+    {
+      "epoch": 1.1562326869806094,
+      "grad_norm": 0.6898880004882812,
+      "learning_rate": 4.572277002910291e-06,
+      "loss": 0.5811,
+      "step": 4174
+    },
+    {
+      "epoch": 1.1565096952908587,
+      "grad_norm": 0.7220594882965088,
+      "learning_rate": 4.57207320695197e-06,
+      "loss": 0.6973,
+      "step": 4175
+    },
+    {
+      "epoch": 1.156786703601108,
+      "grad_norm": 0.6856008172035217,
+      "learning_rate": 4.571869366998244e-06,
+      "loss": 0.6165,
+      "step": 4176
+    },
+    {
+      "epoch": 1.1570637119113574,
+      "grad_norm": 0.6618354916572571,
+      "learning_rate": 4.57166548305344e-06,
+      "loss": 0.5833,
+      "step": 4177
+    },
+    {
+      "epoch": 1.1573407202216066,
+      "grad_norm": 0.7045774459838867,
+      "learning_rate": 4.571461555121888e-06,
+      "loss": 0.6329,
+      "step": 4178
+    },
+    {
+      "epoch": 1.1576177285318558,
+      "grad_norm": 0.7029802799224854,
+      "learning_rate": 4.571257583207917e-06,
+      "loss": 0.637,
+      "step": 4179
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.6747250556945801,
+      "learning_rate": 4.571053567315858e-06,
+      "loss": 0.6293,
+      "step": 4180
+    },
+    {
+      "epoch": 1.1581717451523545,
+      "grad_norm": 0.6726357340812683,
+      "learning_rate": 4.570849507450042e-06,
+      "loss": 0.6595,
+      "step": 4181
+    },
+    {
+      "epoch": 1.1584487534626038,
+      "grad_norm": 0.7162138223648071,
+      "learning_rate": 4.570645403614803e-06,
+      "loss": 0.677,
+      "step": 4182
+    },
+    {
+      "epoch": 1.1587257617728532,
+      "grad_norm": 0.6681135892868042,
+      "learning_rate": 4.5704412558144744e-06,
+      "loss": 0.6358,
+      "step": 4183
+    },
+    {
+      "epoch": 1.1590027700831025,
+      "grad_norm": 0.7122220993041992,
+      "learning_rate": 4.570237064053391e-06,
+      "loss": 0.6678,
+      "step": 4184
+    },
+    {
+      "epoch": 1.1592797783933517,
+      "grad_norm": 0.6930856704711914,
+      "learning_rate": 4.570032828335887e-06,
+      "loss": 0.6539,
+      "step": 4185
+    },
+    {
+      "epoch": 1.1595567867036012,
+      "grad_norm": 0.7084271311759949,
+      "learning_rate": 4.569828548666301e-06,
+      "loss": 0.6265,
+      "step": 4186
+    },
+    {
+      "epoch": 1.1598337950138504,
+      "grad_norm": 0.6743333339691162,
+      "learning_rate": 4.569624225048968e-06,
+      "loss": 0.5993,
+      "step": 4187
+    },
+    {
+      "epoch": 1.1601108033240997,
+      "grad_norm": 0.7010241746902466,
+      "learning_rate": 4.569419857488228e-06,
+      "loss": 0.6733,
+      "step": 4188
+    },
+    {
+      "epoch": 1.1603878116343491,
+      "grad_norm": 0.656864583492279,
+      "learning_rate": 4.56921544598842e-06,
+      "loss": 0.5831,
+      "step": 4189
+    },
+    {
+      "epoch": 1.1606648199445984,
+      "grad_norm": 0.6931899189949036,
+      "learning_rate": 4.569010990553883e-06,
+      "loss": 0.6362,
+      "step": 4190
+    },
+    {
+      "epoch": 1.1609418282548476,
+      "grad_norm": 0.7144986987113953,
+      "learning_rate": 4.5688064911889605e-06,
+      "loss": 0.635,
+      "step": 4191
+    },
+    {
+      "epoch": 1.161218836565097,
+      "grad_norm": 0.6765926480293274,
+      "learning_rate": 4.5686019478979915e-06,
+      "loss": 0.6082,
+      "step": 4192
+    },
+    {
+      "epoch": 1.1614958448753463,
+      "grad_norm": 0.7012190222740173,
+      "learning_rate": 4.5683973606853216e-06,
+      "loss": 0.6636,
+      "step": 4193
+    },
+    {
+      "epoch": 1.1617728531855955,
+      "grad_norm": 0.6823146939277649,
+      "learning_rate": 4.568192729555293e-06,
+      "loss": 0.6071,
+      "step": 4194
+    },
+    {
+      "epoch": 1.162049861495845,
+      "grad_norm": 0.6781933903694153,
+      "learning_rate": 4.567988054512251e-06,
+      "loss": 0.6544,
+      "step": 4195
+    },
+    {
+      "epoch": 1.1623268698060942,
+      "grad_norm": 0.6953796744346619,
+      "learning_rate": 4.567783335560542e-06,
+      "loss": 0.6036,
+      "step": 4196
+    },
+    {
+      "epoch": 1.1626038781163435,
+      "grad_norm": 0.7140665650367737,
+      "learning_rate": 4.567578572704512e-06,
+      "loss": 0.6562,
+      "step": 4197
+    },
+    {
+      "epoch": 1.1628808864265927,
+      "grad_norm": 0.7328235507011414,
+      "learning_rate": 4.567373765948509e-06,
+      "loss": 0.6365,
+      "step": 4198
+    },
+    {
+      "epoch": 1.1631578947368422,
+      "grad_norm": 0.7307348847389221,
+      "learning_rate": 4.567168915296881e-06,
+      "loss": 0.6155,
+      "step": 4199
+    },
+    {
+      "epoch": 1.1634349030470914,
+      "grad_norm": 0.6891008019447327,
+      "learning_rate": 4.566964020753979e-06,
+      "loss": 0.6396,
+      "step": 4200
+    },
+    {
+      "epoch": 1.1637119113573406,
+      "grad_norm": 0.7145599126815796,
+      "learning_rate": 4.5667590823241515e-06,
+      "loss": 0.6082,
+      "step": 4201
+    },
+    {
+      "epoch": 1.16398891966759,
+      "grad_norm": 0.6793602705001831,
+      "learning_rate": 4.5665541000117504e-06,
+      "loss": 0.6393,
+      "step": 4202
+    },
+    {
+      "epoch": 1.1642659279778393,
+      "grad_norm": 0.6986454129219055,
+      "learning_rate": 4.566349073821129e-06,
+      "loss": 0.6861,
+      "step": 4203
+    },
+    {
+      "epoch": 1.1645429362880886,
+      "grad_norm": 0.6601034998893738,
+      "learning_rate": 4.566144003756641e-06,
+      "loss": 0.6556,
+      "step": 4204
+    },
+    {
+      "epoch": 1.164819944598338,
+      "grad_norm": 0.7221328020095825,
+      "learning_rate": 4.565938889822637e-06,
+      "loss": 0.6427,
+      "step": 4205
+    },
+    {
+      "epoch": 1.1650969529085873,
+      "grad_norm": 0.6929381489753723,
+      "learning_rate": 4.5657337320234765e-06,
+      "loss": 0.6507,
+      "step": 4206
+    },
+    {
+      "epoch": 1.1653739612188365,
+      "grad_norm": 0.716958224773407,
+      "learning_rate": 4.565528530363513e-06,
+      "loss": 0.6626,
+      "step": 4207
+    },
+    {
+      "epoch": 1.1656509695290858,
+      "grad_norm": 0.6676911115646362,
+      "learning_rate": 4.5653232848471044e-06,
+      "loss": 0.6325,
+      "step": 4208
+    },
+    {
+      "epoch": 1.1659279778393352,
+      "grad_norm": 0.6989991068840027,
+      "learning_rate": 4.565117995478607e-06,
+      "loss": 0.6409,
+      "step": 4209
+    },
+    {
+      "epoch": 1.1662049861495845,
+      "grad_norm": 0.6747390627861023,
+      "learning_rate": 4.564912662262382e-06,
+      "loss": 0.6289,
+      "step": 4210
+    },
+    {
+      "epoch": 1.1664819944598337,
+      "grad_norm": 0.7510681748390198,
+      "learning_rate": 4.564707285202787e-06,
+      "loss": 0.6718,
+      "step": 4211
+    },
+    {
+      "epoch": 1.1667590027700832,
+      "grad_norm": 0.6956333518028259,
+      "learning_rate": 4.564501864304184e-06,
+      "loss": 0.6466,
+      "step": 4212
+    },
+    {
+      "epoch": 1.1670360110803324,
+      "grad_norm": 0.6801543831825256,
+      "learning_rate": 4.564296399570935e-06,
+      "loss": 0.6505,
+      "step": 4213
+    },
+    {
+      "epoch": 1.1673130193905816,
+      "grad_norm": 0.6749382019042969,
+      "learning_rate": 4.564090891007401e-06,
+      "loss": 0.6512,
+      "step": 4214
+    },
+    {
+      "epoch": 1.167590027700831,
+      "grad_norm": 0.6772417426109314,
+      "learning_rate": 4.563885338617946e-06,
+      "loss": 0.6516,
+      "step": 4215
+    },
+    {
+      "epoch": 1.1678670360110803,
+      "grad_norm": 0.709536612033844,
+      "learning_rate": 4.563679742406935e-06,
+      "loss": 0.6672,
+      "step": 4216
+    },
+    {
+      "epoch": 1.1681440443213296,
+      "grad_norm": 0.6753586530685425,
+      "learning_rate": 4.563474102378733e-06,
+      "loss": 0.5978,
+      "step": 4217
+    },
+    {
+      "epoch": 1.168421052631579,
+      "grad_norm": 0.6711388230323792,
+      "learning_rate": 4.563268418537707e-06,
+      "loss": 0.6517,
+      "step": 4218
+    },
+    {
+      "epoch": 1.1686980609418283,
+      "grad_norm": 0.710477352142334,
+      "learning_rate": 4.563062690888223e-06,
+      "loss": 0.5786,
+      "step": 4219
+    },
+    {
+      "epoch": 1.1689750692520775,
+      "grad_norm": 0.7528419494628906,
+      "learning_rate": 4.562856919434649e-06,
+      "loss": 0.6847,
+      "step": 4220
+    },
+    {
+      "epoch": 1.169252077562327,
+      "grad_norm": 0.6778903603553772,
+      "learning_rate": 4.562651104181355e-06,
+      "loss": 0.6534,
+      "step": 4221
+    },
+    {
+      "epoch": 1.1695290858725762,
+      "grad_norm": 0.7068110704421997,
+      "learning_rate": 4.562445245132711e-06,
+      "loss": 0.6424,
+      "step": 4222
+    },
+    {
+      "epoch": 1.1698060941828254,
+      "grad_norm": 0.7029364705085754,
+      "learning_rate": 4.562239342293087e-06,
+      "loss": 0.604,
+      "step": 4223
+    },
+    {
+      "epoch": 1.170083102493075,
+      "grad_norm": 0.6927694082260132,
+      "learning_rate": 4.562033395666856e-06,
+      "loss": 0.631,
+      "step": 4224
+    },
+    {
+      "epoch": 1.1703601108033241,
+      "grad_norm": 0.6557900905609131,
+      "learning_rate": 4.56182740525839e-06,
+      "loss": 0.5642,
+      "step": 4225
+    },
+    {
+      "epoch": 1.1706371191135734,
+      "grad_norm": 0.7124279141426086,
+      "learning_rate": 4.561621371072062e-06,
+      "loss": 0.616,
+      "step": 4226
+    },
+    {
+      "epoch": 1.1709141274238228,
+      "grad_norm": 0.687511146068573,
+      "learning_rate": 4.561415293112248e-06,
+      "loss": 0.6791,
+      "step": 4227
+    },
+    {
+      "epoch": 1.171191135734072,
+      "grad_norm": 0.735288679599762,
+      "learning_rate": 4.561209171383323e-06,
+      "loss": 0.6629,
+      "step": 4228
+    },
+    {
+      "epoch": 1.1714681440443213,
+      "grad_norm": 0.7048251032829285,
+      "learning_rate": 4.561003005889664e-06,
+      "loss": 0.6281,
+      "step": 4229
+    },
+    {
+      "epoch": 1.1717451523545706,
+      "grad_norm": 0.7248666882514954,
+      "learning_rate": 4.5607967966356485e-06,
+      "loss": 0.6081,
+      "step": 4230
+    },
+    {
+      "epoch": 1.17202216066482,
+      "grad_norm": 0.7453956604003906,
+      "learning_rate": 4.5605905436256525e-06,
+      "loss": 0.6674,
+      "step": 4231
+    },
+    {
+      "epoch": 1.1722991689750693,
+      "grad_norm": 0.7330853343009949,
+      "learning_rate": 4.560384246864058e-06,
+      "loss": 0.6476,
+      "step": 4232
+    },
+    {
+      "epoch": 1.1725761772853185,
+      "grad_norm": 0.6860072016716003,
+      "learning_rate": 4.560177906355244e-06,
+      "loss": 0.6248,
+      "step": 4233
+    },
+    {
+      "epoch": 1.172853185595568,
+      "grad_norm": 0.7117403149604797,
+      "learning_rate": 4.559971522103593e-06,
+      "loss": 0.6479,
+      "step": 4234
+    },
+    {
+      "epoch": 1.1731301939058172,
+      "grad_norm": 0.7106106877326965,
+      "learning_rate": 4.559765094113485e-06,
+      "loss": 0.6356,
+      "step": 4235
+    },
+    {
+      "epoch": 1.1734072022160664,
+      "grad_norm": 0.7224343419075012,
+      "learning_rate": 4.559558622389304e-06,
+      "loss": 0.6826,
+      "step": 4236
+    },
+    {
+      "epoch": 1.1736842105263159,
+      "grad_norm": 0.6818200945854187,
+      "learning_rate": 4.559352106935435e-06,
+      "loss": 0.665,
+      "step": 4237
+    },
+    {
+      "epoch": 1.1739612188365651,
+      "grad_norm": 0.6833382844924927,
+      "learning_rate": 4.559145547756261e-06,
+      "loss": 0.626,
+      "step": 4238
+    },
+    {
+      "epoch": 1.1742382271468144,
+      "grad_norm": 0.6887573599815369,
+      "learning_rate": 4.5589389448561685e-06,
+      "loss": 0.5914,
+      "step": 4239
+    },
+    {
+      "epoch": 1.1745152354570636,
+      "grad_norm": 0.6644566059112549,
+      "learning_rate": 4.558732298239544e-06,
+      "loss": 0.6298,
+      "step": 4240
+    },
+    {
+      "epoch": 1.174792243767313,
+      "grad_norm": 0.6625338792800903,
+      "learning_rate": 4.5585256079107754e-06,
+      "loss": 0.6454,
+      "step": 4241
+    },
+    {
+      "epoch": 1.1750692520775623,
+      "grad_norm": 0.693747878074646,
+      "learning_rate": 4.558318873874251e-06,
+      "loss": 0.636,
+      "step": 4242
+    },
+    {
+      "epoch": 1.1753462603878115,
+      "grad_norm": 0.664527952671051,
+      "learning_rate": 4.558112096134361e-06,
+      "loss": 0.6218,
+      "step": 4243
+    },
+    {
+      "epoch": 1.175623268698061,
+      "grad_norm": 0.7000016570091248,
+      "learning_rate": 4.557905274695496e-06,
+      "loss": 0.5644,
+      "step": 4244
+    },
+    {
+      "epoch": 1.1759002770083102,
+      "grad_norm": 0.679958701133728,
+      "learning_rate": 4.557698409562045e-06,
+      "loss": 0.6628,
+      "step": 4245
+    },
+    {
+      "epoch": 1.1761772853185595,
+      "grad_norm": 0.7532631158828735,
+      "learning_rate": 4.557491500738404e-06,
+      "loss": 0.6723,
+      "step": 4246
+    },
+    {
+      "epoch": 1.176454293628809,
+      "grad_norm": 0.717473566532135,
+      "learning_rate": 4.557284548228962e-06,
+      "loss": 0.6573,
+      "step": 4247
+    },
+    {
+      "epoch": 1.1767313019390582,
+      "grad_norm": 0.6726430058479309,
+      "learning_rate": 4.557077552038117e-06,
+      "loss": 0.641,
+      "step": 4248
+    },
+    {
+      "epoch": 1.1770083102493074,
+      "grad_norm": 0.6901071667671204,
+      "learning_rate": 4.556870512170261e-06,
+      "loss": 0.6325,
+      "step": 4249
+    },
+    {
+      "epoch": 1.1772853185595569,
+      "grad_norm": 0.743581235408783,
+      "learning_rate": 4.5566634286297925e-06,
+      "loss": 0.6499,
+      "step": 4250
+    },
+    {
+      "epoch": 1.177562326869806,
+      "grad_norm": 0.6675125360488892,
+      "learning_rate": 4.556456301421107e-06,
+      "loss": 0.6098,
+      "step": 4251
+    },
+    {
+      "epoch": 1.1778393351800553,
+      "grad_norm": 0.7380349636077881,
+      "learning_rate": 4.556249130548603e-06,
+      "loss": 0.6552,
+      "step": 4252
+    },
+    {
+      "epoch": 1.1781163434903048,
+      "grad_norm": 0.6815469264984131,
+      "learning_rate": 4.556041916016678e-06,
+      "loss": 0.6356,
+      "step": 4253
+    },
+    {
+      "epoch": 1.178393351800554,
+      "grad_norm": 0.7289326190948486,
+      "learning_rate": 4.555834657829733e-06,
+      "loss": 0.6627,
+      "step": 4254
+    },
+    {
+      "epoch": 1.1786703601108033,
+      "grad_norm": 0.6645580530166626,
+      "learning_rate": 4.555627355992168e-06,
+      "loss": 0.6166,
+      "step": 4255
+    },
+    {
+      "epoch": 1.1789473684210527,
+      "grad_norm": 0.69098961353302,
+      "learning_rate": 4.555420010508386e-06,
+      "loss": 0.6364,
+      "step": 4256
+    },
+    {
+      "epoch": 1.179224376731302,
+      "grad_norm": 0.701954185962677,
+      "learning_rate": 4.555212621382786e-06,
+      "loss": 0.6556,
+      "step": 4257
+    },
+    {
+      "epoch": 1.1795013850415512,
+      "grad_norm": 0.7003589868545532,
+      "learning_rate": 4.555005188619776e-06,
+      "loss": 0.6613,
+      "step": 4258
+    },
+    {
+      "epoch": 1.1797783933518005,
+      "grad_norm": 0.6646780967712402,
+      "learning_rate": 4.554797712223756e-06,
+      "loss": 0.637,
+      "step": 4259
+    },
+    {
+      "epoch": 1.18005540166205,
+      "grad_norm": 0.6946302652359009,
+      "learning_rate": 4.554590192199134e-06,
+      "loss": 0.6475,
+      "step": 4260
+    },
+    {
+      "epoch": 1.1803324099722992,
+      "grad_norm": 0.6732128858566284,
+      "learning_rate": 4.554382628550315e-06,
+      "loss": 0.6355,
+      "step": 4261
+    },
+    {
+      "epoch": 1.1806094182825484,
+      "grad_norm": 0.7041682004928589,
+      "learning_rate": 4.5541750212817065e-06,
+      "loss": 0.6679,
+      "step": 4262
+    },
+    {
+      "epoch": 1.1808864265927979,
+      "grad_norm": 0.6972526907920837,
+      "learning_rate": 4.553967370397717e-06,
+      "loss": 0.6495,
+      "step": 4263
+    },
+    {
+      "epoch": 1.181163434903047,
+      "grad_norm": 0.6706414222717285,
+      "learning_rate": 4.553759675902754e-06,
+      "loss": 0.648,
+      "step": 4264
+    },
+    {
+      "epoch": 1.1814404432132963,
+      "grad_norm": 0.6960473656654358,
+      "learning_rate": 4.553551937801229e-06,
+      "loss": 0.597,
+      "step": 4265
+    },
+    {
+      "epoch": 1.1817174515235458,
+      "grad_norm": 0.7602198123931885,
+      "learning_rate": 4.5533441560975525e-06,
+      "loss": 0.6275,
+      "step": 4266
+    },
+    {
+      "epoch": 1.181994459833795,
+      "grad_norm": 0.7271232008934021,
+      "learning_rate": 4.5531363307961355e-06,
+      "loss": 0.6614,
+      "step": 4267
+    },
+    {
+      "epoch": 1.1822714681440443,
+      "grad_norm": 0.6794989109039307,
+      "learning_rate": 4.552928461901391e-06,
+      "loss": 0.6522,
+      "step": 4268
+    },
+    {
+      "epoch": 1.1825484764542935,
+      "grad_norm": 0.6836474537849426,
+      "learning_rate": 4.552720549417733e-06,
+      "loss": 0.6167,
+      "step": 4269
+    },
+    {
+      "epoch": 1.182825484764543,
+      "grad_norm": 0.7154577970504761,
+      "learning_rate": 4.552512593349575e-06,
+      "loss": 0.6265,
+      "step": 4270
+    },
+    {
+      "epoch": 1.1831024930747922,
+      "grad_norm": 0.7174842953681946,
+      "learning_rate": 4.552304593701334e-06,
+      "loss": 0.6631,
+      "step": 4271
+    },
+    {
+      "epoch": 1.1833795013850414,
+      "grad_norm": 0.702748715877533,
+      "learning_rate": 4.5520965504774254e-06,
+      "loss": 0.6121,
+      "step": 4272
+    },
+    {
+      "epoch": 1.183656509695291,
+      "grad_norm": 0.7215221524238586,
+      "learning_rate": 4.551888463682266e-06,
+      "loss": 0.6173,
+      "step": 4273
+    },
+    {
+      "epoch": 1.1839335180055401,
+      "grad_norm": 0.6870599985122681,
+      "learning_rate": 4.551680333320274e-06,
+      "loss": 0.632,
+      "step": 4274
+    },
+    {
+      "epoch": 1.1842105263157894,
+      "grad_norm": 0.7133249640464783,
+      "learning_rate": 4.55147215939587e-06,
+      "loss": 0.6538,
+      "step": 4275
+    },
+    {
+      "epoch": 1.1844875346260388,
+      "grad_norm": 0.7360088229179382,
+      "learning_rate": 4.551263941913473e-06,
+      "loss": 0.6628,
+      "step": 4276
+    },
+    {
+      "epoch": 1.184764542936288,
+      "grad_norm": 0.6721140742301941,
+      "learning_rate": 4.551055680877504e-06,
+      "loss": 0.6347,
+      "step": 4277
+    },
+    {
+      "epoch": 1.1850415512465373,
+      "grad_norm": 0.6926547884941101,
+      "learning_rate": 4.550847376292385e-06,
+      "loss": 0.6544,
+      "step": 4278
+    },
+    {
+      "epoch": 1.1853185595567868,
+      "grad_norm": 0.7329055070877075,
+      "learning_rate": 4.550639028162539e-06,
+      "loss": 0.669,
+      "step": 4279
+    },
+    {
+      "epoch": 1.185595567867036,
+      "grad_norm": 0.7039414644241333,
+      "learning_rate": 4.55043063649239e-06,
+      "loss": 0.6161,
+      "step": 4280
+    },
+    {
+      "epoch": 1.1858725761772853,
+      "grad_norm": 0.7472238540649414,
+      "learning_rate": 4.550222201286362e-06,
+      "loss": 0.657,
+      "step": 4281
+    },
+    {
+      "epoch": 1.1861495844875347,
+      "grad_norm": 0.7102679014205933,
+      "learning_rate": 4.550013722548882e-06,
+      "loss": 0.6748,
+      "step": 4282
+    },
+    {
+      "epoch": 1.186426592797784,
+      "grad_norm": 0.6597498655319214,
+      "learning_rate": 4.549805200284374e-06,
+      "loss": 0.6175,
+      "step": 4283
+    },
+    {
+      "epoch": 1.1867036011080332,
+      "grad_norm": 0.7128549814224243,
+      "learning_rate": 4.5495966344972675e-06,
+      "loss": 0.6658,
+      "step": 4284
+    },
+    {
+      "epoch": 1.1869806094182827,
+      "grad_norm": 0.6920933723449707,
+      "learning_rate": 4.54938802519199e-06,
+      "loss": 0.6846,
+      "step": 4285
+    },
+    {
+      "epoch": 1.1872576177285319,
+      "grad_norm": 0.7122675776481628,
+      "learning_rate": 4.549179372372971e-06,
+      "loss": 0.598,
+      "step": 4286
+    },
+    {
+      "epoch": 1.1875346260387811,
+      "grad_norm": 0.6924179792404175,
+      "learning_rate": 4.548970676044642e-06,
+      "loss": 0.6782,
+      "step": 4287
+    },
+    {
+      "epoch": 1.1878116343490306,
+      "grad_norm": 0.6880548596382141,
+      "learning_rate": 4.548761936211433e-06,
+      "loss": 0.6455,
+      "step": 4288
+    },
+    {
+      "epoch": 1.1880886426592798,
+      "grad_norm": 0.6973972320556641,
+      "learning_rate": 4.5485531528777754e-06,
+      "loss": 0.6654,
+      "step": 4289
+    },
+    {
+      "epoch": 1.188365650969529,
+      "grad_norm": 0.7235435843467712,
+      "learning_rate": 4.548344326048103e-06,
+      "loss": 0.6529,
+      "step": 4290
+    },
+    {
+      "epoch": 1.1886426592797783,
+      "grad_norm": 0.7232111096382141,
+      "learning_rate": 4.54813545572685e-06,
+      "loss": 0.6503,
+      "step": 4291
+    },
+    {
+      "epoch": 1.1889196675900278,
+      "grad_norm": 0.7104105353355408,
+      "learning_rate": 4.547926541918451e-06,
+      "loss": 0.617,
+      "step": 4292
+    },
+    {
+      "epoch": 1.189196675900277,
+      "grad_norm": 0.7474246025085449,
+      "learning_rate": 4.547717584627341e-06,
+      "loss": 0.6935,
+      "step": 4293
+    },
+    {
+      "epoch": 1.1894736842105262,
+      "grad_norm": 0.666694164276123,
+      "learning_rate": 4.547508583857958e-06,
+      "loss": 0.6286,
+      "step": 4294
+    },
+    {
+      "epoch": 1.1897506925207757,
+      "grad_norm": 0.7198405861854553,
+      "learning_rate": 4.547299539614739e-06,
+      "loss": 0.6494,
+      "step": 4295
+    },
+    {
+      "epoch": 1.190027700831025,
+      "grad_norm": 0.6890691518783569,
+      "learning_rate": 4.547090451902122e-06,
+      "loss": 0.6632,
+      "step": 4296
+    },
+    {
+      "epoch": 1.1903047091412742,
+      "grad_norm": 0.7404274940490723,
+      "learning_rate": 4.546881320724547e-06,
+      "loss": 0.6546,
+      "step": 4297
+    },
+    {
+      "epoch": 1.1905817174515236,
+      "grad_norm": 0.7668154835700989,
+      "learning_rate": 4.5466721460864545e-06,
+      "loss": 0.6071,
+      "step": 4298
+    },
+    {
+      "epoch": 1.1908587257617729,
+      "grad_norm": 0.6996212601661682,
+      "learning_rate": 4.546462927992287e-06,
+      "loss": 0.5975,
+      "step": 4299
+    },
+    {
+      "epoch": 1.1911357340720221,
+      "grad_norm": 0.7173477411270142,
+      "learning_rate": 4.546253666446484e-06,
+      "loss": 0.6498,
+      "step": 4300
+    },
+    {
+      "epoch": 1.1914127423822714,
+      "grad_norm": 0.6888937950134277,
+      "learning_rate": 4.54604436145349e-06,
+      "loss": 0.6264,
+      "step": 4301
+    },
+    {
+      "epoch": 1.1916897506925208,
+      "grad_norm": 0.7148381471633911,
+      "learning_rate": 4.54583501301775e-06,
+      "loss": 0.6041,
+      "step": 4302
+    },
+    {
+      "epoch": 1.19196675900277,
+      "grad_norm": 0.714459240436554,
+      "learning_rate": 4.545625621143707e-06,
+      "loss": 0.6545,
+      "step": 4303
+    },
+    {
+      "epoch": 1.1922437673130193,
+      "grad_norm": 0.6980823874473572,
+      "learning_rate": 4.54541618583581e-06,
+      "loss": 0.6409,
+      "step": 4304
+    },
+    {
+      "epoch": 1.1925207756232687,
+      "grad_norm": 0.70297771692276,
+      "learning_rate": 4.545206707098503e-06,
+      "loss": 0.696,
+      "step": 4305
+    },
+    {
+      "epoch": 1.192797783933518,
+      "grad_norm": 0.6873321533203125,
+      "learning_rate": 4.5449971849362345e-06,
+      "loss": 0.6614,
+      "step": 4306
+    },
+    {
+      "epoch": 1.1930747922437672,
+      "grad_norm": 0.6853544116020203,
+      "learning_rate": 4.5447876193534535e-06,
+      "loss": 0.6193,
+      "step": 4307
+    },
+    {
+      "epoch": 1.1933518005540167,
+      "grad_norm": 0.7040695548057556,
+      "learning_rate": 4.54457801035461e-06,
+      "loss": 0.65,
+      "step": 4308
+    },
+    {
+      "epoch": 1.193628808864266,
+      "grad_norm": 0.7038838863372803,
+      "learning_rate": 4.544368357944153e-06,
+      "loss": 0.5984,
+      "step": 4309
+    },
+    {
+      "epoch": 1.1939058171745152,
+      "grad_norm": 0.7128795385360718,
+      "learning_rate": 4.5441586621265355e-06,
+      "loss": 0.6457,
+      "step": 4310
+    },
+    {
+      "epoch": 1.1941828254847646,
+      "grad_norm": 0.6804617643356323,
+      "learning_rate": 4.543948922906209e-06,
+      "loss": 0.6331,
+      "step": 4311
+    },
+    {
+      "epoch": 1.1944598337950139,
+      "grad_norm": 0.7417746186256409,
+      "learning_rate": 4.543739140287628e-06,
+      "loss": 0.6492,
+      "step": 4312
+    },
+    {
+      "epoch": 1.194736842105263,
+      "grad_norm": 0.6482412815093994,
+      "learning_rate": 4.543529314275246e-06,
+      "loss": 0.6603,
+      "step": 4313
+    },
+    {
+      "epoch": 1.1950138504155126,
+      "grad_norm": 0.718251645565033,
+      "learning_rate": 4.543319444873517e-06,
+      "loss": 0.6177,
+      "step": 4314
+    },
+    {
+      "epoch": 1.1952908587257618,
+      "grad_norm": 0.701597273349762,
+      "learning_rate": 4.5431095320869e-06,
+      "loss": 0.646,
+      "step": 4315
+    },
+    {
+      "epoch": 1.195567867036011,
+      "grad_norm": 0.6754698753356934,
+      "learning_rate": 4.542899575919848e-06,
+      "loss": 0.6194,
+      "step": 4316
+    },
+    {
+      "epoch": 1.1958448753462605,
+      "grad_norm": 0.7565904855728149,
+      "learning_rate": 4.542689576376822e-06,
+      "loss": 0.6346,
+      "step": 4317
+    },
+    {
+      "epoch": 1.1961218836565097,
+      "grad_norm": 0.6827279329299927,
+      "learning_rate": 4.5424795334622796e-06,
+      "loss": 0.5896,
+      "step": 4318
+    },
+    {
+      "epoch": 1.196398891966759,
+      "grad_norm": 0.6626576781272888,
+      "learning_rate": 4.542269447180682e-06,
+      "loss": 0.5791,
+      "step": 4319
+    },
+    {
+      "epoch": 1.1966759002770084,
+      "grad_norm": 0.6990970969200134,
+      "learning_rate": 4.542059317536487e-06,
+      "loss": 0.6393,
+      "step": 4320
+    },
+    {
+      "epoch": 1.1969529085872577,
+      "grad_norm": 0.6745477318763733,
+      "learning_rate": 4.541849144534158e-06,
+      "loss": 0.63,
+      "step": 4321
+    },
+    {
+      "epoch": 1.197229916897507,
+      "grad_norm": 0.6842837929725647,
+      "learning_rate": 4.541638928178158e-06,
+      "loss": 0.6406,
+      "step": 4322
+    },
+    {
+      "epoch": 1.1975069252077561,
+      "grad_norm": 0.7515599727630615,
+      "learning_rate": 4.54142866847295e-06,
+      "loss": 0.7141,
+      "step": 4323
+    },
+    {
+      "epoch": 1.1977839335180056,
+      "grad_norm": 0.6852219104766846,
+      "learning_rate": 4.541218365422997e-06,
+      "loss": 0.6726,
+      "step": 4324
+    },
+    {
+      "epoch": 1.1980609418282548,
+      "grad_norm": 0.692903995513916,
+      "learning_rate": 4.541008019032766e-06,
+      "loss": 0.6796,
+      "step": 4325
+    },
+    {
+      "epoch": 1.198337950138504,
+      "grad_norm": 0.6800821423530579,
+      "learning_rate": 4.540797629306722e-06,
+      "loss": 0.6205,
+      "step": 4326
+    },
+    {
+      "epoch": 1.1986149584487535,
+      "grad_norm": 0.6892538070678711,
+      "learning_rate": 4.540587196249334e-06,
+      "loss": 0.6659,
+      "step": 4327
+    },
+    {
+      "epoch": 1.1988919667590028,
+      "grad_norm": 0.6764261722564697,
+      "learning_rate": 4.5403767198650685e-06,
+      "loss": 0.6432,
+      "step": 4328
+    },
+    {
+      "epoch": 1.199168975069252,
+      "grad_norm": 0.6850671768188477,
+      "learning_rate": 4.540166200158394e-06,
+      "loss": 0.6389,
+      "step": 4329
+    },
+    {
+      "epoch": 1.1994459833795015,
+      "grad_norm": 0.7733122706413269,
+      "learning_rate": 4.539955637133781e-06,
+      "loss": 0.6897,
+      "step": 4330
+    },
+    {
+      "epoch": 1.1997229916897507,
+      "grad_norm": 0.7287300825119019,
+      "learning_rate": 4.539745030795701e-06,
+      "loss": 0.6121,
+      "step": 4331
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7426024079322815,
+      "learning_rate": 4.539534381148625e-06,
+      "loss": 0.6242,
+      "step": 4332
+    },
+    {
+      "epoch": 1.2002770083102492,
+      "grad_norm": 0.6916001439094543,
+      "learning_rate": 4.539323688197025e-06,
+      "loss": 0.6173,
+      "step": 4333
+    },
+    {
+      "epoch": 1.2005540166204987,
+      "grad_norm": 0.6753976941108704,
+      "learning_rate": 4.5391129519453755e-06,
+      "loss": 0.5832,
+      "step": 4334
+    },
+    {
+      "epoch": 1.200831024930748,
+      "grad_norm": 0.7031921148300171,
+      "learning_rate": 4.538902172398151e-06,
+      "loss": 0.7186,
+      "step": 4335
+    },
+    {
+      "epoch": 1.2011080332409971,
+      "grad_norm": 0.6825315952301025,
+      "learning_rate": 4.538691349559826e-06,
+      "loss": 0.6368,
+      "step": 4336
+    },
+    {
+      "epoch": 1.2013850415512466,
+      "grad_norm": 0.7218365669250488,
+      "learning_rate": 4.538480483434878e-06,
+      "loss": 0.6509,
+      "step": 4337
+    },
+    {
+      "epoch": 1.2016620498614958,
+      "grad_norm": 0.6999350786209106,
+      "learning_rate": 4.538269574027784e-06,
+      "loss": 0.6409,
+      "step": 4338
+    },
+    {
+      "epoch": 1.201939058171745,
+      "grad_norm": 0.6903960108757019,
+      "learning_rate": 4.538058621343021e-06,
+      "loss": 0.5946,
+      "step": 4339
+    },
+    {
+      "epoch": 1.2022160664819945,
+      "grad_norm": 0.7130528688430786,
+      "learning_rate": 4.537847625385069e-06,
+      "loss": 0.6628,
+      "step": 4340
+    },
+    {
+      "epoch": 1.2024930747922438,
+      "grad_norm": 0.6918487548828125,
+      "learning_rate": 4.537636586158408e-06,
+      "loss": 0.6841,
+      "step": 4341
+    },
+    {
+      "epoch": 1.202770083102493,
+      "grad_norm": 0.7330001592636108,
+      "learning_rate": 4.5374255036675185e-06,
+      "loss": 0.6573,
+      "step": 4342
+    },
+    {
+      "epoch": 1.2030470914127425,
+      "grad_norm": 0.6802187561988831,
+      "learning_rate": 4.537214377916882e-06,
+      "loss": 0.6342,
+      "step": 4343
+    },
+    {
+      "epoch": 1.2033240997229917,
+      "grad_norm": 0.6788448691368103,
+      "learning_rate": 4.537003208910983e-06,
+      "loss": 0.6528,
+      "step": 4344
+    },
+    {
+      "epoch": 1.203601108033241,
+      "grad_norm": 0.6833314299583435,
+      "learning_rate": 4.536791996654303e-06,
+      "loss": 0.6437,
+      "step": 4345
+    },
+    {
+      "epoch": 1.2038781163434904,
+      "grad_norm": 0.6793355345726013,
+      "learning_rate": 4.536580741151328e-06,
+      "loss": 0.6384,
+      "step": 4346
+    },
+    {
+      "epoch": 1.2041551246537396,
+      "grad_norm": 0.7549615502357483,
+      "learning_rate": 4.536369442406543e-06,
+      "loss": 0.6657,
+      "step": 4347
+    },
+    {
+      "epoch": 1.2044321329639889,
+      "grad_norm": 0.6814740896224976,
+      "learning_rate": 4.536158100424434e-06,
+      "loss": 0.6212,
+      "step": 4348
+    },
+    {
+      "epoch": 1.2047091412742383,
+      "grad_norm": 0.7178139686584473,
+      "learning_rate": 4.535946715209488e-06,
+      "loss": 0.6641,
+      "step": 4349
+    },
+    {
+      "epoch": 1.2049861495844876,
+      "grad_norm": 0.7358893752098083,
+      "learning_rate": 4.535735286766196e-06,
+      "loss": 0.6558,
+      "step": 4350
+    },
+    {
+      "epoch": 1.2052631578947368,
+      "grad_norm": 0.6513990759849548,
+      "learning_rate": 4.535523815099044e-06,
+      "loss": 0.6421,
+      "step": 4351
+    },
+    {
+      "epoch": 1.2055401662049863,
+      "grad_norm": 0.7175055146217346,
+      "learning_rate": 4.535312300212523e-06,
+      "loss": 0.6676,
+      "step": 4352
+    },
+    {
+      "epoch": 1.2058171745152355,
+      "grad_norm": 0.6629988551139832,
+      "learning_rate": 4.535100742111125e-06,
+      "loss": 0.5907,
+      "step": 4353
+    },
+    {
+      "epoch": 1.2060941828254848,
+      "grad_norm": 0.7017353177070618,
+      "learning_rate": 4.534889140799341e-06,
+      "loss": 0.6615,
+      "step": 4354
+    },
+    {
+      "epoch": 1.206371191135734,
+      "grad_norm": 0.7436637878417969,
+      "learning_rate": 4.5346774962816635e-06,
+      "loss": 0.681,
+      "step": 4355
+    },
+    {
+      "epoch": 1.2066481994459834,
+      "grad_norm": 0.6942512392997742,
+      "learning_rate": 4.534465808562587e-06,
+      "loss": 0.6619,
+      "step": 4356
+    },
+    {
+      "epoch": 1.2069252077562327,
+      "grad_norm": 0.6875748038291931,
+      "learning_rate": 4.534254077646606e-06,
+      "loss": 0.664,
+      "step": 4357
+    },
+    {
+      "epoch": 1.207202216066482,
+      "grad_norm": 0.6879202127456665,
+      "learning_rate": 4.5340423035382165e-06,
+      "loss": 0.6087,
+      "step": 4358
+    },
+    {
+      "epoch": 1.2074792243767314,
+      "grad_norm": 0.6686105132102966,
+      "learning_rate": 4.533830486241914e-06,
+      "loss": 0.6218,
+      "step": 4359
+    },
+    {
+      "epoch": 1.2077562326869806,
+      "grad_norm": 0.6839017868041992,
+      "learning_rate": 4.533618625762196e-06,
+      "loss": 0.6579,
+      "step": 4360
+    },
+    {
+      "epoch": 1.2080332409972299,
+      "grad_norm": 0.6563047170639038,
+      "learning_rate": 4.533406722103562e-06,
+      "loss": 0.6073,
+      "step": 4361
+    },
+    {
+      "epoch": 1.208310249307479,
+      "grad_norm": 0.7116596102714539,
+      "learning_rate": 4.53319477527051e-06,
+      "loss": 0.6353,
+      "step": 4362
+    },
+    {
+      "epoch": 1.2085872576177286,
+      "grad_norm": 0.6606056690216064,
+      "learning_rate": 4.5329827852675415e-06,
+      "loss": 0.5481,
+      "step": 4363
+    },
+    {
+      "epoch": 1.2088642659279778,
+      "grad_norm": 0.7383421659469604,
+      "learning_rate": 4.532770752099156e-06,
+      "loss": 0.6742,
+      "step": 4364
+    },
+    {
+      "epoch": 1.209141274238227,
+      "grad_norm": 0.6783084273338318,
+      "learning_rate": 4.532558675769857e-06,
+      "loss": 0.6384,
+      "step": 4365
+    },
+    {
+      "epoch": 1.2094182825484765,
+      "grad_norm": 0.6640183329582214,
+      "learning_rate": 4.532346556284146e-06,
+      "loss": 0.6,
+      "step": 4366
+    },
+    {
+      "epoch": 1.2096952908587257,
+      "grad_norm": 0.6977359652519226,
+      "learning_rate": 4.5321343936465275e-06,
+      "loss": 0.5709,
+      "step": 4367
+    },
+    {
+      "epoch": 1.209972299168975,
+      "grad_norm": 0.6893478631973267,
+      "learning_rate": 4.531922187861507e-06,
+      "loss": 0.6805,
+      "step": 4368
+    },
+    {
+      "epoch": 1.2102493074792244,
+      "grad_norm": 0.7190584540367126,
+      "learning_rate": 4.531709938933589e-06,
+      "loss": 0.6757,
+      "step": 4369
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.7101868391036987,
+      "learning_rate": 4.531497646867281e-06,
+      "loss": 0.647,
+      "step": 4370
+    },
+    {
+      "epoch": 1.210803324099723,
+      "grad_norm": 0.7230154871940613,
+      "learning_rate": 4.53128531166709e-06,
+      "loss": 0.6387,
+      "step": 4371
+    },
+    {
+      "epoch": 1.2110803324099724,
+      "grad_norm": 0.6733981966972351,
+      "learning_rate": 4.5310729333375245e-06,
+      "loss": 0.6259,
+      "step": 4372
+    },
+    {
+      "epoch": 1.2113573407202216,
+      "grad_norm": 0.7233732342720032,
+      "learning_rate": 4.530860511883093e-06,
+      "loss": 0.6457,
+      "step": 4373
+    },
+    {
+      "epoch": 1.2116343490304708,
+      "grad_norm": 0.6873233318328857,
+      "learning_rate": 4.530648047308308e-06,
+      "loss": 0.6845,
+      "step": 4374
+    },
+    {
+      "epoch": 1.2119113573407203,
+      "grad_norm": 0.6871516108512878,
+      "learning_rate": 4.530435539617678e-06,
+      "loss": 0.6357,
+      "step": 4375
+    },
+    {
+      "epoch": 1.2121883656509695,
+      "grad_norm": 0.6907857656478882,
+      "learning_rate": 4.530222988815717e-06,
+      "loss": 0.6181,
+      "step": 4376
+    },
+    {
+      "epoch": 1.2124653739612188,
+      "grad_norm": 0.680697500705719,
+      "learning_rate": 4.530010394906938e-06,
+      "loss": 0.6281,
+      "step": 4377
+    },
+    {
+      "epoch": 1.2127423822714682,
+      "grad_norm": 0.6937142014503479,
+      "learning_rate": 4.529797757895852e-06,
+      "loss": 0.6251,
+      "step": 4378
+    },
+    {
+      "epoch": 1.2130193905817175,
+      "grad_norm": 0.6891060471534729,
+      "learning_rate": 4.529585077786977e-06,
+      "loss": 0.5744,
+      "step": 4379
+    },
+    {
+      "epoch": 1.2132963988919667,
+      "grad_norm": 0.6695826053619385,
+      "learning_rate": 4.529372354584829e-06,
+      "loss": 0.6819,
+      "step": 4380
+    },
+    {
+      "epoch": 1.2135734072022162,
+      "grad_norm": 0.7129724621772766,
+      "learning_rate": 4.529159588293922e-06,
+      "loss": 0.6574,
+      "step": 4381
+    },
+    {
+      "epoch": 1.2138504155124654,
+      "grad_norm": 0.7002400159835815,
+      "learning_rate": 4.528946778918777e-06,
+      "loss": 0.646,
+      "step": 4382
+    },
+    {
+      "epoch": 1.2141274238227147,
+      "grad_norm": 0.7137943506240845,
+      "learning_rate": 4.528733926463909e-06,
+      "loss": 0.6311,
+      "step": 4383
+    },
+    {
+      "epoch": 1.214404432132964,
+      "grad_norm": 0.6910654306411743,
+      "learning_rate": 4.528521030933839e-06,
+      "loss": 0.6364,
+      "step": 4384
+    },
+    {
+      "epoch": 1.2146814404432134,
+      "grad_norm": 0.7344447374343872,
+      "learning_rate": 4.528308092333087e-06,
+      "loss": 0.6668,
+      "step": 4385
+    },
+    {
+      "epoch": 1.2149584487534626,
+      "grad_norm": 0.7259987592697144,
+      "learning_rate": 4.5280951106661755e-06,
+      "loss": 0.6296,
+      "step": 4386
+    },
+    {
+      "epoch": 1.2152354570637118,
+      "grad_norm": 0.7200382351875305,
+      "learning_rate": 4.527882085937625e-06,
+      "loss": 0.6375,
+      "step": 4387
+    },
+    {
+      "epoch": 1.2155124653739613,
+      "grad_norm": 0.7581033110618591,
+      "learning_rate": 4.527669018151959e-06,
+      "loss": 0.6453,
+      "step": 4388
+    },
+    {
+      "epoch": 1.2157894736842105,
+      "grad_norm": 0.6925063133239746,
+      "learning_rate": 4.527455907313702e-06,
+      "loss": 0.6299,
+      "step": 4389
+    },
+    {
+      "epoch": 1.2160664819944598,
+      "grad_norm": 0.7005012631416321,
+      "learning_rate": 4.527242753427378e-06,
+      "loss": 0.6334,
+      "step": 4390
+    },
+    {
+      "epoch": 1.2163434903047092,
+      "grad_norm": 0.6756923198699951,
+      "learning_rate": 4.527029556497513e-06,
+      "loss": 0.6363,
+      "step": 4391
+    },
+    {
+      "epoch": 1.2166204986149585,
+      "grad_norm": 0.7093691825866699,
+      "learning_rate": 4.526816316528635e-06,
+      "loss": 0.6279,
+      "step": 4392
+    },
+    {
+      "epoch": 1.2168975069252077,
+      "grad_norm": 0.7289251089096069,
+      "learning_rate": 4.526603033525271e-06,
+      "loss": 0.6371,
+      "step": 4393
+    },
+    {
+      "epoch": 1.217174515235457,
+      "grad_norm": 0.6732731461524963,
+      "learning_rate": 4.526389707491948e-06,
+      "loss": 0.6124,
+      "step": 4394
+    },
+    {
+      "epoch": 1.2174515235457064,
+      "grad_norm": 0.6666122674942017,
+      "learning_rate": 4.5261763384331975e-06,
+      "loss": 0.6345,
+      "step": 4395
+    },
+    {
+      "epoch": 1.2177285318559556,
+      "grad_norm": 0.6899173855781555,
+      "learning_rate": 4.525962926353549e-06,
+      "loss": 0.6911,
+      "step": 4396
+    },
+    {
+      "epoch": 1.2180055401662049,
+      "grad_norm": 0.673346996307373,
+      "learning_rate": 4.525749471257534e-06,
+      "loss": 0.6024,
+      "step": 4397
+    },
+    {
+      "epoch": 1.2182825484764543,
+      "grad_norm": 0.695513904094696,
+      "learning_rate": 4.5255359731496835e-06,
+      "loss": 0.6529,
+      "step": 4398
+    },
+    {
+      "epoch": 1.2185595567867036,
+      "grad_norm": 0.6898539066314697,
+      "learning_rate": 4.5253224320345324e-06,
+      "loss": 0.673,
+      "step": 4399
+    },
+    {
+      "epoch": 1.2188365650969528,
+      "grad_norm": 0.6731051206588745,
+      "learning_rate": 4.525108847916614e-06,
+      "loss": 0.6152,
+      "step": 4400
+    },
+    {
+      "epoch": 1.2191135734072023,
+      "grad_norm": 0.6579740047454834,
+      "learning_rate": 4.524895220800463e-06,
+      "loss": 0.6188,
+      "step": 4401
+    },
+    {
+      "epoch": 1.2193905817174515,
+      "grad_norm": 0.6864678263664246,
+      "learning_rate": 4.524681550690615e-06,
+      "loss": 0.6384,
+      "step": 4402
+    },
+    {
+      "epoch": 1.2196675900277008,
+      "grad_norm": 0.7058321237564087,
+      "learning_rate": 4.524467837591608e-06,
+      "loss": 0.6311,
+      "step": 4403
+    },
+    {
+      "epoch": 1.2199445983379502,
+      "grad_norm": 0.7127742171287537,
+      "learning_rate": 4.524254081507978e-06,
+      "loss": 0.646,
+      "step": 4404
+    },
+    {
+      "epoch": 1.2202216066481995,
+      "grad_norm": 0.6770905256271362,
+      "learning_rate": 4.524040282444265e-06,
+      "loss": 0.6787,
+      "step": 4405
+    },
+    {
+      "epoch": 1.2204986149584487,
+      "grad_norm": 0.6978400349617004,
+      "learning_rate": 4.523826440405009e-06,
+      "loss": 0.6089,
+      "step": 4406
+    },
+    {
+      "epoch": 1.2207756232686982,
+      "grad_norm": 0.6935478448867798,
+      "learning_rate": 4.523612555394748e-06,
+      "loss": 0.6656,
+      "step": 4407
+    },
+    {
+      "epoch": 1.2210526315789474,
+      "grad_norm": 0.6988148093223572,
+      "learning_rate": 4.523398627418025e-06,
+      "loss": 0.6174,
+      "step": 4408
+    },
+    {
+      "epoch": 1.2213296398891966,
+      "grad_norm": 0.67249596118927,
+      "learning_rate": 4.523184656479382e-06,
+      "loss": 0.5974,
+      "step": 4409
+    },
+    {
+      "epoch": 1.221606648199446,
+      "grad_norm": 0.6931259632110596,
+      "learning_rate": 4.522970642583362e-06,
+      "loss": 0.6077,
+      "step": 4410
+    },
+    {
+      "epoch": 1.2218836565096953,
+      "grad_norm": 0.6827836632728577,
+      "learning_rate": 4.522756585734509e-06,
+      "loss": 0.6427,
+      "step": 4411
+    },
+    {
+      "epoch": 1.2221606648199446,
+      "grad_norm": 0.7271913886070251,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.6013,
+      "step": 4412
+    },
+    {
+      "epoch": 1.222437673130194,
+      "grad_norm": 0.7043451070785522,
+      "learning_rate": 4.522328343196486e-06,
+      "loss": 0.6546,
+      "step": 4413
+    },
+    {
+      "epoch": 1.2227146814404433,
+      "grad_norm": 0.7179906964302063,
+      "learning_rate": 4.522114157516409e-06,
+      "loss": 0.6769,
+      "step": 4414
+    },
+    {
+      "epoch": 1.2229916897506925,
+      "grad_norm": 0.7093173265457153,
+      "learning_rate": 4.521899928901683e-06,
+      "loss": 0.6705,
+      "step": 4415
+    },
+    {
+      "epoch": 1.2232686980609417,
+      "grad_norm": 0.6971815228462219,
+      "learning_rate": 4.5216856573568594e-06,
+      "loss": 0.6267,
+      "step": 4416
+    },
+    {
+      "epoch": 1.2235457063711912,
+      "grad_norm": 0.7456490993499756,
+      "learning_rate": 4.521471342886486e-06,
+      "loss": 0.6268,
+      "step": 4417
+    },
+    {
+      "epoch": 1.2238227146814404,
+      "grad_norm": 0.7302700877189636,
+      "learning_rate": 4.521256985495114e-06,
+      "loss": 0.6061,
+      "step": 4418
+    },
+    {
+      "epoch": 1.2240997229916897,
+      "grad_norm": 0.6957641243934631,
+      "learning_rate": 4.5210425851872944e-06,
+      "loss": 0.6209,
+      "step": 4419
+    },
+    {
+      "epoch": 1.2243767313019391,
+      "grad_norm": 0.7002754807472229,
+      "learning_rate": 4.520828141967579e-06,
+      "loss": 0.6072,
+      "step": 4420
+    },
+    {
+      "epoch": 1.2246537396121884,
+      "grad_norm": 0.7033765316009521,
+      "learning_rate": 4.5206136558405214e-06,
+      "loss": 0.6292,
+      "step": 4421
+    },
+    {
+      "epoch": 1.2249307479224376,
+      "grad_norm": 0.7024313807487488,
+      "learning_rate": 4.520399126810676e-06,
+      "loss": 0.6282,
+      "step": 4422
+    },
+    {
+      "epoch": 1.225207756232687,
+      "grad_norm": 0.6639774441719055,
+      "learning_rate": 4.520184554882598e-06,
+      "loss": 0.6386,
+      "step": 4423
+    },
+    {
+      "epoch": 1.2254847645429363,
+      "grad_norm": 0.6910221576690674,
+      "learning_rate": 4.519969940060843e-06,
+      "loss": 0.6593,
+      "step": 4424
+    },
+    {
+      "epoch": 1.2257617728531855,
+      "grad_norm": 0.6906338930130005,
+      "learning_rate": 4.519755282349967e-06,
+      "loss": 0.6654,
+      "step": 4425
+    },
+    {
+      "epoch": 1.2260387811634348,
+      "grad_norm": 0.6946108937263489,
+      "learning_rate": 4.519540581754529e-06,
+      "loss": 0.6468,
+      "step": 4426
+    },
+    {
+      "epoch": 1.2263157894736842,
+      "grad_norm": 0.683337926864624,
+      "learning_rate": 4.519325838279087e-06,
+      "loss": 0.6455,
+      "step": 4427
+    },
+    {
+      "epoch": 1.2265927977839335,
+      "grad_norm": 0.7096542119979858,
+      "learning_rate": 4.519111051928201e-06,
+      "loss": 0.5897,
+      "step": 4428
+    },
+    {
+      "epoch": 1.2268698060941827,
+      "grad_norm": 0.7188792824745178,
+      "learning_rate": 4.518896222706431e-06,
+      "loss": 0.6503,
+      "step": 4429
+    },
+    {
+      "epoch": 1.2271468144044322,
+      "grad_norm": 0.6780487895011902,
+      "learning_rate": 4.518681350618338e-06,
+      "loss": 0.6146,
+      "step": 4430
+    },
+    {
+      "epoch": 1.2274238227146814,
+      "grad_norm": 0.7028935551643372,
+      "learning_rate": 4.518466435668485e-06,
+      "loss": 0.7035,
+      "step": 4431
+    },
+    {
+      "epoch": 1.2277008310249307,
+      "grad_norm": 0.673099935054779,
+      "learning_rate": 4.518251477861435e-06,
+      "loss": 0.6031,
+      "step": 4432
+    },
+    {
+      "epoch": 1.2279778393351801,
+      "grad_norm": 0.7044011354446411,
+      "learning_rate": 4.518036477201752e-06,
+      "loss": 0.6526,
+      "step": 4433
+    },
+    {
+      "epoch": 1.2282548476454294,
+      "grad_norm": 0.7253800630569458,
+      "learning_rate": 4.5178214336940015e-06,
+      "loss": 0.6041,
+      "step": 4434
+    },
+    {
+      "epoch": 1.2285318559556786,
+      "grad_norm": 0.7392945885658264,
+      "learning_rate": 4.5176063473427485e-06,
+      "loss": 0.607,
+      "step": 4435
+    },
+    {
+      "epoch": 1.228808864265928,
+      "grad_norm": 0.6777207255363464,
+      "learning_rate": 4.51739121815256e-06,
+      "loss": 0.6576,
+      "step": 4436
+    },
+    {
+      "epoch": 1.2290858725761773,
+      "grad_norm": 0.7337944507598877,
+      "learning_rate": 4.517176046128005e-06,
+      "loss": 0.6573,
+      "step": 4437
+    },
+    {
+      "epoch": 1.2293628808864265,
+      "grad_norm": 0.6642884612083435,
+      "learning_rate": 4.516960831273651e-06,
+      "loss": 0.5576,
+      "step": 4438
+    },
+    {
+      "epoch": 1.229639889196676,
+      "grad_norm": 0.7272876501083374,
+      "learning_rate": 4.516745573594068e-06,
+      "loss": 0.6473,
+      "step": 4439
+    },
+    {
+      "epoch": 1.2299168975069252,
+      "grad_norm": 0.6578417420387268,
+      "learning_rate": 4.516530273093826e-06,
+      "loss": 0.648,
+      "step": 4440
+    },
+    {
+      "epoch": 1.2301939058171745,
+      "grad_norm": 0.6466295123100281,
+      "learning_rate": 4.516314929777496e-06,
+      "loss": 0.6004,
+      "step": 4441
+    },
+    {
+      "epoch": 1.230470914127424,
+      "grad_norm": 0.6575599908828735,
+      "learning_rate": 4.516099543649651e-06,
+      "loss": 0.5946,
+      "step": 4442
+    },
+    {
+      "epoch": 1.2307479224376732,
+      "grad_norm": 0.7217696905136108,
+      "learning_rate": 4.515884114714865e-06,
+      "loss": 0.6447,
+      "step": 4443
+    },
+    {
+      "epoch": 1.2310249307479224,
+      "grad_norm": 0.7224985957145691,
+      "learning_rate": 4.515668642977711e-06,
+      "loss": 0.6679,
+      "step": 4444
+    },
+    {
+      "epoch": 1.2313019390581719,
+      "grad_norm": 0.7255591750144958,
+      "learning_rate": 4.515453128442764e-06,
+      "loss": 0.6278,
+      "step": 4445
+    },
+    {
+      "epoch": 1.231578947368421,
+      "grad_norm": 0.6969859004020691,
+      "learning_rate": 4.5152375711146e-06,
+      "loss": 0.6217,
+      "step": 4446
+    },
+    {
+      "epoch": 1.2318559556786703,
+      "grad_norm": 0.731724739074707,
+      "learning_rate": 4.515021970997797e-06,
+      "loss": 0.6845,
+      "step": 4447
+    },
+    {
+      "epoch": 1.2321329639889196,
+      "grad_norm": 0.7150566577911377,
+      "learning_rate": 4.5148063280969315e-06,
+      "loss": 0.6456,
+      "step": 4448
+    },
+    {
+      "epoch": 1.232409972299169,
+      "grad_norm": 0.6822801232337952,
+      "learning_rate": 4.5145906424165815e-06,
+      "loss": 0.6243,
+      "step": 4449
+    },
+    {
+      "epoch": 1.2326869806094183,
+      "grad_norm": 0.7677393555641174,
+      "learning_rate": 4.514374913961328e-06,
+      "loss": 0.6645,
+      "step": 4450
+    },
+    {
+      "epoch": 1.2329639889196675,
+      "grad_norm": 0.71131432056427,
+      "learning_rate": 4.514159142735751e-06,
+      "loss": 0.6702,
+      "step": 4451
+    },
+    {
+      "epoch": 1.233240997229917,
+      "grad_norm": 0.7009198069572449,
+      "learning_rate": 4.5139433287444316e-06,
+      "loss": 0.6511,
+      "step": 4452
+    },
+    {
+      "epoch": 1.2335180055401662,
+      "grad_norm": 0.6575670838356018,
+      "learning_rate": 4.5137274719919526e-06,
+      "loss": 0.6374,
+      "step": 4453
+    },
+    {
+      "epoch": 1.2337950138504155,
+      "grad_norm": 0.721043586730957,
+      "learning_rate": 4.513511572482896e-06,
+      "loss": 0.609,
+      "step": 4454
+    },
+    {
+      "epoch": 1.234072022160665,
+      "grad_norm": 0.7106706500053406,
+      "learning_rate": 4.513295630221847e-06,
+      "loss": 0.6202,
+      "step": 4455
+    },
+    {
+      "epoch": 1.2343490304709142,
+      "grad_norm": 0.6895164251327515,
+      "learning_rate": 4.513079645213391e-06,
+      "loss": 0.6248,
+      "step": 4456
+    },
+    {
+      "epoch": 1.2346260387811634,
+      "grad_norm": 0.7164800763130188,
+      "learning_rate": 4.512863617462113e-06,
+      "loss": 0.6702,
+      "step": 4457
+    },
+    {
+      "epoch": 1.2349030470914126,
+      "grad_norm": 0.6920846700668335,
+      "learning_rate": 4.512647546972601e-06,
+      "loss": 0.6542,
+      "step": 4458
+    },
+    {
+      "epoch": 1.235180055401662,
+      "grad_norm": 0.6987187266349792,
+      "learning_rate": 4.5124314337494405e-06,
+      "loss": 0.6786,
+      "step": 4459
+    },
+    {
+      "epoch": 1.2354570637119113,
+      "grad_norm": 0.7604243159294128,
+      "learning_rate": 4.5122152777972215e-06,
+      "loss": 0.654,
+      "step": 4460
+    },
+    {
+      "epoch": 1.2357340720221606,
+      "grad_norm": 0.6927931308746338,
+      "learning_rate": 4.5119990791205335e-06,
+      "loss": 0.6272,
+      "step": 4461
+    },
+    {
+      "epoch": 1.23601108033241,
+      "grad_norm": 0.7042927742004395,
+      "learning_rate": 4.511782837723968e-06,
+      "loss": 0.6764,
+      "step": 4462
+    },
+    {
+      "epoch": 1.2362880886426593,
+      "grad_norm": 0.7222568988800049,
+      "learning_rate": 4.5115665536121145e-06,
+      "loss": 0.6519,
+      "step": 4463
+    },
+    {
+      "epoch": 1.2365650969529085,
+      "grad_norm": 0.7020226120948792,
+      "learning_rate": 4.511350226789566e-06,
+      "loss": 0.6217,
+      "step": 4464
+    },
+    {
+      "epoch": 1.236842105263158,
+      "grad_norm": 0.7381118535995483,
+      "learning_rate": 4.511133857260917e-06,
+      "loss": 0.6656,
+      "step": 4465
+    },
+    {
+      "epoch": 1.2371191135734072,
+      "grad_norm": 0.7156130075454712,
+      "learning_rate": 4.510917445030759e-06,
+      "loss": 0.5877,
+      "step": 4466
+    },
+    {
+      "epoch": 1.2373961218836564,
+      "grad_norm": 0.682116687297821,
+      "learning_rate": 4.510700990103689e-06,
+      "loss": 0.5929,
+      "step": 4467
+    },
+    {
+      "epoch": 1.237673130193906,
+      "grad_norm": 0.7004262208938599,
+      "learning_rate": 4.510484492484302e-06,
+      "loss": 0.6637,
+      "step": 4468
+    },
+    {
+      "epoch": 1.2379501385041551,
+      "grad_norm": 0.7237050533294678,
+      "learning_rate": 4.510267952177195e-06,
+      "loss": 0.6004,
+      "step": 4469
+    },
+    {
+      "epoch": 1.2382271468144044,
+      "grad_norm": 0.7384737730026245,
+      "learning_rate": 4.510051369186965e-06,
+      "loss": 0.6523,
+      "step": 4470
+    },
+    {
+      "epoch": 1.2385041551246538,
+      "grad_norm": 0.6756117343902588,
+      "learning_rate": 4.509834743518212e-06,
+      "loss": 0.6028,
+      "step": 4471
+    },
+    {
+      "epoch": 1.238781163434903,
+      "grad_norm": 0.656095564365387,
+      "learning_rate": 4.5096180751755345e-06,
+      "loss": 0.5923,
+      "step": 4472
+    },
+    {
+      "epoch": 1.2390581717451523,
+      "grad_norm": 0.7115722894668579,
+      "learning_rate": 4.509401364163534e-06,
+      "loss": 0.6651,
+      "step": 4473
+    },
+    {
+      "epoch": 1.2393351800554018,
+      "grad_norm": 0.694269597530365,
+      "learning_rate": 4.509184610486811e-06,
+      "loss": 0.6496,
+      "step": 4474
+    },
+    {
+      "epoch": 1.239612188365651,
+      "grad_norm": 0.8465580940246582,
+      "learning_rate": 4.508967814149967e-06,
+      "loss": 0.6187,
+      "step": 4475
+    },
+    {
+      "epoch": 1.2398891966759003,
+      "grad_norm": 0.7109991312026978,
+      "learning_rate": 4.508750975157606e-06,
+      "loss": 0.6393,
+      "step": 4476
+    },
+    {
+      "epoch": 1.2401662049861495,
+      "grad_norm": 0.6904207468032837,
+      "learning_rate": 4.508534093514333e-06,
+      "loss": 0.6075,
+      "step": 4477
+    },
+    {
+      "epoch": 1.240443213296399,
+      "grad_norm": 0.7130043506622314,
+      "learning_rate": 4.508317169224752e-06,
+      "loss": 0.6706,
+      "step": 4478
+    },
+    {
+      "epoch": 1.2407202216066482,
+      "grad_norm": 0.6982183456420898,
+      "learning_rate": 4.508100202293468e-06,
+      "loss": 0.6405,
+      "step": 4479
+    },
+    {
+      "epoch": 1.2409972299168974,
+      "grad_norm": 0.697863757610321,
+      "learning_rate": 4.507883192725089e-06,
+      "loss": 0.6528,
+      "step": 4480
+    },
+    {
+      "epoch": 1.2412742382271469,
+      "grad_norm": 0.7692074179649353,
+      "learning_rate": 4.5076661405242215e-06,
+      "loss": 0.6911,
+      "step": 4481
+    },
+    {
+      "epoch": 1.2415512465373961,
+      "grad_norm": 0.6971391439437866,
+      "learning_rate": 4.507449045695475e-06,
+      "loss": 0.6221,
+      "step": 4482
+    },
+    {
+      "epoch": 1.2418282548476454,
+      "grad_norm": 0.7123902440071106,
+      "learning_rate": 4.507231908243459e-06,
+      "loss": 0.6344,
+      "step": 4483
+    },
+    {
+      "epoch": 1.2421052631578948,
+      "grad_norm": 0.6307297348976135,
+      "learning_rate": 4.507014728172784e-06,
+      "loss": 0.5585,
+      "step": 4484
+    },
+    {
+      "epoch": 1.242382271468144,
+      "grad_norm": 0.6647037863731384,
+      "learning_rate": 4.506797505488061e-06,
+      "loss": 0.611,
+      "step": 4485
+    },
+    {
+      "epoch": 1.2426592797783933,
+      "grad_norm": 0.7097634673118591,
+      "learning_rate": 4.506580240193902e-06,
+      "loss": 0.6426,
+      "step": 4486
+    },
+    {
+      "epoch": 1.2429362880886425,
+      "grad_norm": 0.7197310924530029,
+      "learning_rate": 4.50636293229492e-06,
+      "loss": 0.6631,
+      "step": 4487
+    },
+    {
+      "epoch": 1.243213296398892,
+      "grad_norm": 0.733557403087616,
+      "learning_rate": 4.50614558179573e-06,
+      "loss": 0.7106,
+      "step": 4488
+    },
+    {
+      "epoch": 1.2434903047091412,
+      "grad_norm": 0.6763700246810913,
+      "learning_rate": 4.505928188700946e-06,
+      "loss": 0.641,
+      "step": 4489
+    },
+    {
+      "epoch": 1.2437673130193905,
+      "grad_norm": 0.6792541146278381,
+      "learning_rate": 4.505710753015183e-06,
+      "loss": 0.6299,
+      "step": 4490
+    },
+    {
+      "epoch": 1.24404432132964,
+      "grad_norm": 0.6903266310691833,
+      "learning_rate": 4.505493274743059e-06,
+      "loss": 0.6145,
+      "step": 4491
+    },
+    {
+      "epoch": 1.2443213296398892,
+      "grad_norm": 0.683650553226471,
+      "learning_rate": 4.505275753889193e-06,
+      "loss": 0.6278,
+      "step": 4492
+    },
+    {
+      "epoch": 1.2445983379501384,
+      "grad_norm": 0.6768408417701721,
+      "learning_rate": 4.5050581904581995e-06,
+      "loss": 0.6286,
+      "step": 4493
+    },
+    {
+      "epoch": 1.2448753462603879,
+      "grad_norm": 0.7079578042030334,
+      "learning_rate": 4.5048405844547e-06,
+      "loss": 0.677,
+      "step": 4494
+    },
+    {
+      "epoch": 1.245152354570637,
+      "grad_norm": 0.7316173315048218,
+      "learning_rate": 4.504622935883317e-06,
+      "loss": 0.6535,
+      "step": 4495
+    },
+    {
+      "epoch": 1.2454293628808863,
+      "grad_norm": 0.710331380367279,
+      "learning_rate": 4.504405244748669e-06,
+      "loss": 0.6437,
+      "step": 4496
+    },
+    {
+      "epoch": 1.2457063711911358,
+      "grad_norm": 0.7169858813285828,
+      "learning_rate": 4.504187511055379e-06,
+      "loss": 0.6394,
+      "step": 4497
+    },
+    {
+      "epoch": 1.245983379501385,
+      "grad_norm": 0.7219961285591125,
+      "learning_rate": 4.50396973480807e-06,
+      "loss": 0.6137,
+      "step": 4498
+    },
+    {
+      "epoch": 1.2462603878116343,
+      "grad_norm": 0.6625384092330933,
+      "learning_rate": 4.503751916011366e-06,
+      "loss": 0.6166,
+      "step": 4499
+    },
+    {
+      "epoch": 1.2465373961218837,
+      "grad_norm": 0.6969285011291504,
+      "learning_rate": 4.5035340546698915e-06,
+      "loss": 0.6553,
+      "step": 4500
+    },
+    {
+      "epoch": 1.246814404432133,
+      "grad_norm": 0.7164677381515503,
+      "learning_rate": 4.5033161507882735e-06,
+      "loss": 0.6162,
+      "step": 4501
+    },
+    {
+      "epoch": 1.2470914127423822,
+      "grad_norm": 0.6780024766921997,
+      "learning_rate": 4.503098204371137e-06,
+      "loss": 0.6477,
+      "step": 4502
+    },
+    {
+      "epoch": 1.2473684210526317,
+      "grad_norm": 0.6793859601020813,
+      "learning_rate": 4.50288021542311e-06,
+      "loss": 0.639,
+      "step": 4503
+    },
+    {
+      "epoch": 1.247645429362881,
+      "grad_norm": 0.6692623496055603,
+      "learning_rate": 4.5026621839488216e-06,
+      "loss": 0.6533,
+      "step": 4504
+    },
+    {
+      "epoch": 1.2479224376731302,
+      "grad_norm": 0.6580060720443726,
+      "learning_rate": 4.502444109952901e-06,
+      "loss": 0.635,
+      "step": 4505
+    },
+    {
+      "epoch": 1.2481994459833796,
+      "grad_norm": 0.6871946454048157,
+      "learning_rate": 4.502225993439978e-06,
+      "loss": 0.6506,
+      "step": 4506
+    },
+    {
+      "epoch": 1.2484764542936289,
+      "grad_norm": 0.6828311085700989,
+      "learning_rate": 4.502007834414684e-06,
+      "loss": 0.6284,
+      "step": 4507
+    },
+    {
+      "epoch": 1.248753462603878,
+      "grad_norm": 0.6815675497055054,
+      "learning_rate": 4.5017896328816515e-06,
+      "loss": 0.6438,
+      "step": 4508
+    },
+    {
+      "epoch": 1.2490304709141273,
+      "grad_norm": 0.6794790029525757,
+      "learning_rate": 4.501571388845513e-06,
+      "loss": 0.6353,
+      "step": 4509
+    },
+    {
+      "epoch": 1.2493074792243768,
+      "grad_norm": 0.6996496319770813,
+      "learning_rate": 4.501353102310901e-06,
+      "loss": 0.5925,
+      "step": 4510
+    },
+    {
+      "epoch": 1.249584487534626,
+      "grad_norm": 0.7551298141479492,
+      "learning_rate": 4.501134773282453e-06,
+      "loss": 0.6096,
+      "step": 4511
+    },
+    {
+      "epoch": 1.2498614958448753,
+      "grad_norm": 0.6573173403739929,
+      "learning_rate": 4.500916401764803e-06,
+      "loss": 0.6343,
+      "step": 4512
+    },
+    {
+      "epoch": 1.2501385041551247,
+      "grad_norm": 0.7132793068885803,
+      "learning_rate": 4.500697987762589e-06,
+      "loss": 0.6373,
+      "step": 4513
+    },
+    {
+      "epoch": 1.250415512465374,
+      "grad_norm": 0.6720200181007385,
+      "learning_rate": 4.500479531280447e-06,
+      "loss": 0.628,
+      "step": 4514
+    },
+    {
+      "epoch": 1.2506925207756232,
+      "grad_norm": 0.6969121098518372,
+      "learning_rate": 4.500261032323015e-06,
+      "loss": 0.655,
+      "step": 4515
+    },
+    {
+      "epoch": 1.2509695290858724,
+      "grad_norm": 0.6909615993499756,
+      "learning_rate": 4.500042490894935e-06,
+      "loss": 0.6342,
+      "step": 4516
+    },
+    {
+      "epoch": 1.251246537396122,
+      "grad_norm": 0.7056781053543091,
+      "learning_rate": 4.499823907000844e-06,
+      "loss": 0.6741,
+      "step": 4517
+    },
+    {
+      "epoch": 1.2515235457063711,
+      "grad_norm": 0.6950377821922302,
+      "learning_rate": 4.499605280645385e-06,
+      "loss": 0.6234,
+      "step": 4518
+    },
+    {
+      "epoch": 1.2518005540166204,
+      "grad_norm": 0.7027127146720886,
+      "learning_rate": 4.499386611833199e-06,
+      "loss": 0.6298,
+      "step": 4519
+    },
+    {
+      "epoch": 1.2520775623268698,
+      "grad_norm": 0.7132726311683655,
+      "learning_rate": 4.499167900568929e-06,
+      "loss": 0.6417,
+      "step": 4520
+    },
+    {
+      "epoch": 1.252354570637119,
+      "grad_norm": 0.7127083539962769,
+      "learning_rate": 4.49894914685722e-06,
+      "loss": 0.6289,
+      "step": 4521
+    },
+    {
+      "epoch": 1.2526315789473683,
+      "grad_norm": 0.6999812126159668,
+      "learning_rate": 4.4987303507027155e-06,
+      "loss": 0.6629,
+      "step": 4522
+    },
+    {
+      "epoch": 1.2529085872576178,
+      "grad_norm": 0.7140102982521057,
+      "learning_rate": 4.4985115121100605e-06,
+      "loss": 0.6367,
+      "step": 4523
+    },
+    {
+      "epoch": 1.253185595567867,
+      "grad_norm": 0.6769750714302063,
+      "learning_rate": 4.4982926310839035e-06,
+      "loss": 0.6438,
+      "step": 4524
+    },
+    {
+      "epoch": 1.2534626038781163,
+      "grad_norm": 0.6684278249740601,
+      "learning_rate": 4.498073707628891e-06,
+      "loss": 0.6069,
+      "step": 4525
+    },
+    {
+      "epoch": 1.2537396121883657,
+      "grad_norm": 0.6692405939102173,
+      "learning_rate": 4.497854741749671e-06,
+      "loss": 0.615,
+      "step": 4526
+    },
+    {
+      "epoch": 1.254016620498615,
+      "grad_norm": 0.7372734546661377,
+      "learning_rate": 4.497635733450893e-06,
+      "loss": 0.5969,
+      "step": 4527
+    },
+    {
+      "epoch": 1.2542936288088642,
+      "grad_norm": 0.7145940661430359,
+      "learning_rate": 4.497416682737207e-06,
+      "loss": 0.6445,
+      "step": 4528
+    },
+    {
+      "epoch": 1.2545706371191137,
+      "grad_norm": 0.7145653963088989,
+      "learning_rate": 4.497197589613263e-06,
+      "loss": 0.6326,
+      "step": 4529
+    },
+    {
+      "epoch": 1.254847645429363,
+      "grad_norm": 0.6717661619186401,
+      "learning_rate": 4.496978454083716e-06,
+      "loss": 0.7085,
+      "step": 4530
+    },
+    {
+      "epoch": 1.2551246537396121,
+      "grad_norm": 0.6757134199142456,
+      "learning_rate": 4.496759276153215e-06,
+      "loss": 0.6329,
+      "step": 4531
+    },
+    {
+      "epoch": 1.2554016620498616,
+      "grad_norm": 0.7108446955680847,
+      "learning_rate": 4.4965400558264155e-06,
+      "loss": 0.6525,
+      "step": 4532
+    },
+    {
+      "epoch": 1.2556786703601108,
+      "grad_norm": 0.7189573049545288,
+      "learning_rate": 4.496320793107972e-06,
+      "loss": 0.6302,
+      "step": 4533
+    },
+    {
+      "epoch": 1.25595567867036,
+      "grad_norm": 0.7077795267105103,
+      "learning_rate": 4.496101488002541e-06,
+      "loss": 0.6495,
+      "step": 4534
+    },
+    {
+      "epoch": 1.2562326869806095,
+      "grad_norm": 0.722058892250061,
+      "learning_rate": 4.495882140514777e-06,
+      "loss": 0.6663,
+      "step": 4535
+    },
+    {
+      "epoch": 1.2565096952908588,
+      "grad_norm": 0.7322894334793091,
+      "learning_rate": 4.495662750649339e-06,
+      "loss": 0.6126,
+      "step": 4536
+    },
+    {
+      "epoch": 1.256786703601108,
+      "grad_norm": 0.7008985877037048,
+      "learning_rate": 4.495443318410883e-06,
+      "loss": 0.6237,
+      "step": 4537
+    },
+    {
+      "epoch": 1.2570637119113575,
+      "grad_norm": 0.7410510778427124,
+      "learning_rate": 4.4952238438040705e-06,
+      "loss": 0.6669,
+      "step": 4538
+    },
+    {
+      "epoch": 1.2573407202216067,
+      "grad_norm": 0.7088368535041809,
+      "learning_rate": 4.495004326833561e-06,
+      "loss": 0.5999,
+      "step": 4539
+    },
+    {
+      "epoch": 1.257617728531856,
+      "grad_norm": 0.712836742401123,
+      "learning_rate": 4.4947847675040135e-06,
+      "loss": 0.6145,
+      "step": 4540
+    },
+    {
+      "epoch": 1.2578947368421054,
+      "grad_norm": 0.6875491142272949,
+      "learning_rate": 4.494565165820093e-06,
+      "loss": 0.6499,
+      "step": 4541
+    },
+    {
+      "epoch": 1.2581717451523546,
+      "grad_norm": 0.7181865572929382,
+      "learning_rate": 4.49434552178646e-06,
+      "loss": 0.6741,
+      "step": 4542
+    },
+    {
+      "epoch": 1.2584487534626039,
+      "grad_norm": 0.639896810054779,
+      "learning_rate": 4.4941258354077774e-06,
+      "loss": 0.5928,
+      "step": 4543
+    },
+    {
+      "epoch": 1.2587257617728531,
+      "grad_norm": 0.7280617952346802,
+      "learning_rate": 4.493906106688712e-06,
+      "loss": 0.6431,
+      "step": 4544
+    },
+    {
+      "epoch": 1.2590027700831026,
+      "grad_norm": 0.6944310665130615,
+      "learning_rate": 4.493686335633928e-06,
+      "loss": 0.614,
+      "step": 4545
+    },
+    {
+      "epoch": 1.2592797783933518,
+      "grad_norm": 0.6686841249465942,
+      "learning_rate": 4.4934665222480914e-06,
+      "loss": 0.5946,
+      "step": 4546
+    },
+    {
+      "epoch": 1.259556786703601,
+      "grad_norm": 0.711683452129364,
+      "learning_rate": 4.49324666653587e-06,
+      "loss": 0.6002,
+      "step": 4547
+    },
+    {
+      "epoch": 1.2598337950138503,
+      "grad_norm": 0.7249209880828857,
+      "learning_rate": 4.493026768501931e-06,
+      "loss": 0.6424,
+      "step": 4548
+    },
+    {
+      "epoch": 1.2601108033240997,
+      "grad_norm": 0.6927335858345032,
+      "learning_rate": 4.492806828150945e-06,
+      "loss": 0.6605,
+      "step": 4549
+    },
+    {
+      "epoch": 1.260387811634349,
+      "grad_norm": 0.689573347568512,
+      "learning_rate": 4.4925868454875806e-06,
+      "loss": 0.6113,
+      "step": 4550
+    },
+    {
+      "epoch": 1.2606648199445982,
+      "grad_norm": 0.7189673781394958,
+      "learning_rate": 4.492366820516509e-06,
+      "loss": 0.6747,
+      "step": 4551
+    },
+    {
+      "epoch": 1.2609418282548477,
+      "grad_norm": 0.7115404009819031,
+      "learning_rate": 4.492146753242402e-06,
+      "loss": 0.6049,
+      "step": 4552
+    },
+    {
+      "epoch": 1.261218836565097,
+      "grad_norm": 0.6188634037971497,
+      "learning_rate": 4.491926643669931e-06,
+      "loss": 0.5518,
+      "step": 4553
+    },
+    {
+      "epoch": 1.2614958448753462,
+      "grad_norm": 0.7529398798942566,
+      "learning_rate": 4.491706491803772e-06,
+      "loss": 0.7053,
+      "step": 4554
+    },
+    {
+      "epoch": 1.2617728531855956,
+      "grad_norm": 0.6784997582435608,
+      "learning_rate": 4.4914862976485975e-06,
+      "loss": 0.6631,
+      "step": 4555
+    },
+    {
+      "epoch": 1.2620498614958449,
+      "grad_norm": 0.6835280656814575,
+      "learning_rate": 4.491266061209083e-06,
+      "loss": 0.6227,
+      "step": 4556
+    },
+    {
+      "epoch": 1.262326869806094,
+      "grad_norm": 0.7191388010978699,
+      "learning_rate": 4.4910457824899045e-06,
+      "loss": 0.6378,
+      "step": 4557
+    },
+    {
+      "epoch": 1.2626038781163436,
+      "grad_norm": 0.6956838965415955,
+      "learning_rate": 4.49082546149574e-06,
+      "loss": 0.6674,
+      "step": 4558
+    },
+    {
+      "epoch": 1.2628808864265928,
+      "grad_norm": 0.7452521324157715,
+      "learning_rate": 4.490605098231267e-06,
+      "loss": 0.6648,
+      "step": 4559
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.7133474946022034,
+      "learning_rate": 4.490384692701164e-06,
+      "loss": 0.5917,
+      "step": 4560
+    },
+    {
+      "epoch": 1.2634349030470915,
+      "grad_norm": 1.4470252990722656,
+      "learning_rate": 4.490164244910112e-06,
+      "loss": 0.6172,
+      "step": 4561
+    },
+    {
+      "epoch": 1.2637119113573407,
+      "grad_norm": 0.6451750993728638,
+      "learning_rate": 4.489943754862789e-06,
+      "loss": 0.5528,
+      "step": 4562
+    },
+    {
+      "epoch": 1.26398891966759,
+      "grad_norm": 0.7071599960327148,
+      "learning_rate": 4.489723222563879e-06,
+      "loss": 0.6244,
+      "step": 4563
+    },
+    {
+      "epoch": 1.2642659279778394,
+      "grad_norm": 0.716210663318634,
+      "learning_rate": 4.489502648018064e-06,
+      "loss": 0.6003,
+      "step": 4564
+    },
+    {
+      "epoch": 1.2645429362880887,
+      "grad_norm": 0.7129285335540771,
+      "learning_rate": 4.489282031230028e-06,
+      "loss": 0.6613,
+      "step": 4565
+    },
+    {
+      "epoch": 1.264819944598338,
+      "grad_norm": 0.7196843028068542,
+      "learning_rate": 4.4890613722044526e-06,
+      "loss": 0.626,
+      "step": 4566
+    },
+    {
+      "epoch": 1.2650969529085874,
+      "grad_norm": 0.6882171034812927,
+      "learning_rate": 4.488840670946026e-06,
+      "loss": 0.6542,
+      "step": 4567
+    },
+    {
+      "epoch": 1.2653739612188366,
+      "grad_norm": 0.7489578127861023,
+      "learning_rate": 4.488619927459433e-06,
+      "loss": 0.6923,
+      "step": 4568
+    },
+    {
+      "epoch": 1.2656509695290858,
+      "grad_norm": 0.6784943342208862,
+      "learning_rate": 4.48839914174936e-06,
+      "loss": 0.6528,
+      "step": 4569
+    },
+    {
+      "epoch": 1.2659279778393353,
+      "grad_norm": 0.7308769822120667,
+      "learning_rate": 4.488178313820496e-06,
+      "loss": 0.6789,
+      "step": 4570
+    },
+    {
+      "epoch": 1.2662049861495845,
+      "grad_norm": 0.7080146074295044,
+      "learning_rate": 4.487957443677529e-06,
+      "loss": 0.5962,
+      "step": 4571
+    },
+    {
+      "epoch": 1.2664819944598338,
+      "grad_norm": 0.7015159726142883,
+      "learning_rate": 4.487736531325149e-06,
+      "loss": 0.6787,
+      "step": 4572
+    },
+    {
+      "epoch": 1.266759002770083,
+      "grad_norm": 0.768822431564331,
+      "learning_rate": 4.487515576768046e-06,
+      "loss": 0.6674,
+      "step": 4573
+    },
+    {
+      "epoch": 1.2670360110803325,
+      "grad_norm": 0.6876410841941833,
+      "learning_rate": 4.487294580010911e-06,
+      "loss": 0.6382,
+      "step": 4574
+    },
+    {
+      "epoch": 1.2673130193905817,
+      "grad_norm": 0.6838921308517456,
+      "learning_rate": 4.487073541058438e-06,
+      "loss": 0.6155,
+      "step": 4575
+    },
+    {
+      "epoch": 1.267590027700831,
+      "grad_norm": 0.6943739652633667,
+      "learning_rate": 4.486852459915319e-06,
+      "loss": 0.6397,
+      "step": 4576
+    },
+    {
+      "epoch": 1.2678670360110802,
+      "grad_norm": 0.6962082386016846,
+      "learning_rate": 4.486631336586248e-06,
+      "loss": 0.6101,
+      "step": 4577
+    },
+    {
+      "epoch": 1.2681440443213297,
+      "grad_norm": 0.6831097602844238,
+      "learning_rate": 4.486410171075921e-06,
+      "loss": 0.6382,
+      "step": 4578
+    },
+    {
+      "epoch": 1.268421052631579,
+      "grad_norm": 0.6977545022964478,
+      "learning_rate": 4.486188963389034e-06,
+      "loss": 0.625,
+      "step": 4579
+    },
+    {
+      "epoch": 1.2686980609418281,
+      "grad_norm": 0.7188796997070312,
+      "learning_rate": 4.485967713530282e-06,
+      "loss": 0.6152,
+      "step": 4580
+    },
+    {
+      "epoch": 1.2689750692520776,
+      "grad_norm": 0.7160911560058594,
+      "learning_rate": 4.485746421504365e-06,
+      "loss": 0.6692,
+      "step": 4581
+    },
+    {
+      "epoch": 1.2692520775623268,
+      "grad_norm": 0.7020512223243713,
+      "learning_rate": 4.485525087315979e-06,
+      "loss": 0.6369,
+      "step": 4582
+    },
+    {
+      "epoch": 1.269529085872576,
+      "grad_norm": 0.7281891107559204,
+      "learning_rate": 4.485303710969825e-06,
+      "loss": 0.6518,
+      "step": 4583
+    },
+    {
+      "epoch": 1.2698060941828255,
+      "grad_norm": 0.6871817111968994,
+      "learning_rate": 4.485082292470604e-06,
+      "loss": 0.6156,
+      "step": 4584
+    },
+    {
+      "epoch": 1.2700831024930748,
+      "grad_norm": 0.7031907439231873,
+      "learning_rate": 4.484860831823017e-06,
+      "loss": 0.6422,
+      "step": 4585
+    },
+    {
+      "epoch": 1.270360110803324,
+      "grad_norm": 0.7040826678276062,
+      "learning_rate": 4.484639329031765e-06,
+      "loss": 0.6438,
+      "step": 4586
+    },
+    {
+      "epoch": 1.2706371191135735,
+      "grad_norm": 0.691525399684906,
+      "learning_rate": 4.4844177841015516e-06,
+      "loss": 0.6267,
+      "step": 4587
+    },
+    {
+      "epoch": 1.2709141274238227,
+      "grad_norm": 0.683026134967804,
+      "learning_rate": 4.484196197037082e-06,
+      "loss": 0.6341,
+      "step": 4588
+    },
+    {
+      "epoch": 1.271191135734072,
+      "grad_norm": 0.696869969367981,
+      "learning_rate": 4.483974567843059e-06,
+      "loss": 0.6229,
+      "step": 4589
+    },
+    {
+      "epoch": 1.2714681440443214,
+      "grad_norm": 0.765889048576355,
+      "learning_rate": 4.48375289652419e-06,
+      "loss": 0.6088,
+      "step": 4590
+    },
+    {
+      "epoch": 1.2717451523545706,
+      "grad_norm": 0.6921378374099731,
+      "learning_rate": 4.483531183085181e-06,
+      "loss": 0.5697,
+      "step": 4591
+    },
+    {
+      "epoch": 1.2720221606648199,
+      "grad_norm": 0.7679482698440552,
+      "learning_rate": 4.483309427530739e-06,
+      "loss": 0.6527,
+      "step": 4592
+    },
+    {
+      "epoch": 1.2722991689750693,
+      "grad_norm": 0.6497597098350525,
+      "learning_rate": 4.483087629865573e-06,
+      "loss": 0.526,
+      "step": 4593
+    },
+    {
+      "epoch": 1.2725761772853186,
+      "grad_norm": 0.7268680334091187,
+      "learning_rate": 4.482865790094394e-06,
+      "loss": 0.687,
+      "step": 4594
+    },
+    {
+      "epoch": 1.2728531855955678,
+      "grad_norm": 0.6968439221382141,
+      "learning_rate": 4.48264390822191e-06,
+      "loss": 0.6423,
+      "step": 4595
+    },
+    {
+      "epoch": 1.2731301939058173,
+      "grad_norm": 0.6854841709136963,
+      "learning_rate": 4.4824219842528315e-06,
+      "loss": 0.6677,
+      "step": 4596
+    },
+    {
+      "epoch": 1.2734072022160665,
+      "grad_norm": 0.6899726390838623,
+      "learning_rate": 4.482200018191872e-06,
+      "loss": 0.6266,
+      "step": 4597
+    },
+    {
+      "epoch": 1.2736842105263158,
+      "grad_norm": 0.7343031764030457,
+      "learning_rate": 4.481978010043745e-06,
+      "loss": 0.6459,
+      "step": 4598
+    },
+    {
+      "epoch": 1.2739612188365652,
+      "grad_norm": 0.6642977595329285,
+      "learning_rate": 4.4817559598131635e-06,
+      "loss": 0.6194,
+      "step": 4599
+    },
+    {
+      "epoch": 1.2742382271468145,
+      "grad_norm": 0.7220334410667419,
+      "learning_rate": 4.481533867504841e-06,
+      "loss": 0.6689,
+      "step": 4600
+    },
+    {
+      "epoch": 1.2745152354570637,
+      "grad_norm": 0.6452382206916809,
+      "learning_rate": 4.481311733123495e-06,
+      "loss": 0.6143,
+      "step": 4601
+    },
+    {
+      "epoch": 1.2747922437673131,
+      "grad_norm": 0.6728619337081909,
+      "learning_rate": 4.481089556673841e-06,
+      "loss": 0.6198,
+      "step": 4602
+    },
+    {
+      "epoch": 1.2750692520775624,
+      "grad_norm": 0.6677701473236084,
+      "learning_rate": 4.480867338160596e-06,
+      "loss": 0.5959,
+      "step": 4603
+    },
+    {
+      "epoch": 1.2753462603878116,
+      "grad_norm": 0.7178294062614441,
+      "learning_rate": 4.480645077588479e-06,
+      "loss": 0.6414,
+      "step": 4604
+    },
+    {
+      "epoch": 1.2756232686980609,
+      "grad_norm": 0.7325197458267212,
+      "learning_rate": 4.48042277496221e-06,
+      "loss": 0.6589,
+      "step": 4605
+    },
+    {
+      "epoch": 1.2759002770083103,
+      "grad_norm": 0.7167461514472961,
+      "learning_rate": 4.480200430286508e-06,
+      "loss": 0.6728,
+      "step": 4606
+    },
+    {
+      "epoch": 1.2761772853185596,
+      "grad_norm": 0.6576178073883057,
+      "learning_rate": 4.479978043566094e-06,
+      "loss": 0.655,
+      "step": 4607
+    },
+    {
+      "epoch": 1.2764542936288088,
+      "grad_norm": 0.6865749359130859,
+      "learning_rate": 4.4797556148056884e-06,
+      "loss": 0.6594,
+      "step": 4608
+    },
+    {
+      "epoch": 1.276731301939058,
+      "grad_norm": 0.7056397795677185,
+      "learning_rate": 4.479533144010016e-06,
+      "loss": 0.6249,
+      "step": 4609
+    },
+    {
+      "epoch": 1.2770083102493075,
+      "grad_norm": 0.6812729239463806,
+      "learning_rate": 4.4793106311838e-06,
+      "loss": 0.6174,
+      "step": 4610
+    },
+    {
+      "epoch": 1.2772853185595567,
+      "grad_norm": 0.6877213716506958,
+      "learning_rate": 4.4790880763317646e-06,
+      "loss": 0.6007,
+      "step": 4611
+    },
+    {
+      "epoch": 1.277562326869806,
+      "grad_norm": 0.7136980295181274,
+      "learning_rate": 4.478865479458635e-06,
+      "loss": 0.6883,
+      "step": 4612
+    },
+    {
+      "epoch": 1.2778393351800554,
+      "grad_norm": 0.7126525640487671,
+      "learning_rate": 4.4786428405691385e-06,
+      "loss": 0.6294,
+      "step": 4613
+    },
+    {
+      "epoch": 1.2781163434903047,
+      "grad_norm": 0.6921951770782471,
+      "learning_rate": 4.478420159668001e-06,
+      "loss": 0.6899,
+      "step": 4614
+    },
+    {
+      "epoch": 1.278393351800554,
+      "grad_norm": 0.7217085361480713,
+      "learning_rate": 4.478197436759951e-06,
+      "loss": 0.6774,
+      "step": 4615
+    },
+    {
+      "epoch": 1.2786703601108034,
+      "grad_norm": 0.6517031788825989,
+      "learning_rate": 4.477974671849718e-06,
+      "loss": 0.575,
+      "step": 4616
+    },
+    {
+      "epoch": 1.2789473684210526,
+      "grad_norm": 0.680118203163147,
+      "learning_rate": 4.477751864942032e-06,
+      "loss": 0.6299,
+      "step": 4617
+    },
+    {
+      "epoch": 1.2792243767313018,
+      "grad_norm": 0.7211265563964844,
+      "learning_rate": 4.477529016041623e-06,
+      "loss": 0.6159,
+      "step": 4618
+    },
+    {
+      "epoch": 1.2795013850415513,
+      "grad_norm": 0.7222548127174377,
+      "learning_rate": 4.477306125153223e-06,
+      "loss": 0.6827,
+      "step": 4619
+    },
+    {
+      "epoch": 1.2797783933518005,
+      "grad_norm": 0.7041400074958801,
+      "learning_rate": 4.477083192281564e-06,
+      "loss": 0.5943,
+      "step": 4620
+    },
+    {
+      "epoch": 1.2800554016620498,
+      "grad_norm": 0.6987934708595276,
+      "learning_rate": 4.476860217431381e-06,
+      "loss": 0.6259,
+      "step": 4621
+    },
+    {
+      "epoch": 1.2803324099722992,
+      "grad_norm": 0.7352740168571472,
+      "learning_rate": 4.4766372006074055e-06,
+      "loss": 0.6532,
+      "step": 4622
+    },
+    {
+      "epoch": 1.2806094182825485,
+      "grad_norm": 0.6798954010009766,
+      "learning_rate": 4.476414141814375e-06,
+      "loss": 0.6638,
+      "step": 4623
+    },
+    {
+      "epoch": 1.2808864265927977,
+      "grad_norm": 0.6747403740882874,
+      "learning_rate": 4.476191041057026e-06,
+      "loss": 0.6807,
+      "step": 4624
+    },
+    {
+      "epoch": 1.2811634349030472,
+      "grad_norm": 0.685394823551178,
+      "learning_rate": 4.475967898340094e-06,
+      "loss": 0.6137,
+      "step": 4625
+    },
+    {
+      "epoch": 1.2814404432132964,
+      "grad_norm": 0.7123876810073853,
+      "learning_rate": 4.4757447136683175e-06,
+      "loss": 0.5759,
+      "step": 4626
+    },
+    {
+      "epoch": 1.2817174515235457,
+      "grad_norm": 0.7212755084037781,
+      "learning_rate": 4.475521487046435e-06,
+      "loss": 0.6571,
+      "step": 4627
+    },
+    {
+      "epoch": 1.2819944598337951,
+      "grad_norm": 0.7282657027244568,
+      "learning_rate": 4.475298218479186e-06,
+      "loss": 0.6568,
+      "step": 4628
+    },
+    {
+      "epoch": 1.2822714681440444,
+      "grad_norm": 0.7013609409332275,
+      "learning_rate": 4.4750749079713124e-06,
+      "loss": 0.6175,
+      "step": 4629
+    },
+    {
+      "epoch": 1.2825484764542936,
+      "grad_norm": 0.6803429126739502,
+      "learning_rate": 4.474851555527554e-06,
+      "loss": 0.6389,
+      "step": 4630
+    },
+    {
+      "epoch": 1.282825484764543,
+      "grad_norm": 0.6631330251693726,
+      "learning_rate": 4.474628161152654e-06,
+      "loss": 0.6118,
+      "step": 4631
+    },
+    {
+      "epoch": 1.2831024930747923,
+      "grad_norm": 0.7086408734321594,
+      "learning_rate": 4.474404724851356e-06,
+      "loss": 0.612,
+      "step": 4632
+    },
+    {
+      "epoch": 1.2833795013850415,
+      "grad_norm": 0.7163555026054382,
+      "learning_rate": 4.474181246628404e-06,
+      "loss": 0.6794,
+      "step": 4633
+    },
+    {
+      "epoch": 1.283656509695291,
+      "grad_norm": 0.7133662104606628,
+      "learning_rate": 4.4739577264885415e-06,
+      "loss": 0.6819,
+      "step": 4634
+    },
+    {
+      "epoch": 1.2839335180055402,
+      "grad_norm": 0.7242811918258667,
+      "learning_rate": 4.473734164436516e-06,
+      "loss": 0.6825,
+      "step": 4635
+    },
+    {
+      "epoch": 1.2842105263157895,
+      "grad_norm": 0.6506803631782532,
+      "learning_rate": 4.473510560477073e-06,
+      "loss": 0.5928,
+      "step": 4636
+    },
+    {
+      "epoch": 1.2844875346260387,
+      "grad_norm": 0.6938612461090088,
+      "learning_rate": 4.473286914614962e-06,
+      "loss": 0.6575,
+      "step": 4637
+    },
+    {
+      "epoch": 1.2847645429362882,
+      "grad_norm": 0.679517924785614,
+      "learning_rate": 4.4730632268549305e-06,
+      "loss": 0.6791,
+      "step": 4638
+    },
+    {
+      "epoch": 1.2850415512465374,
+      "grad_norm": 0.6626391410827637,
+      "learning_rate": 4.472839497201728e-06,
+      "loss": 0.5879,
+      "step": 4639
+    },
+    {
+      "epoch": 1.2853185595567866,
+      "grad_norm": 0.677143394947052,
+      "learning_rate": 4.472615725660104e-06,
+      "loss": 0.6404,
+      "step": 4640
+    },
+    {
+      "epoch": 1.2855955678670359,
+      "grad_norm": 0.6869742274284363,
+      "learning_rate": 4.472391912234812e-06,
+      "loss": 0.5803,
+      "step": 4641
+    },
+    {
+      "epoch": 1.2858725761772853,
+      "grad_norm": 0.7295581102371216,
+      "learning_rate": 4.472168056930602e-06,
+      "loss": 0.6676,
+      "step": 4642
+    },
+    {
+      "epoch": 1.2861495844875346,
+      "grad_norm": 0.6919876337051392,
+      "learning_rate": 4.471944159752228e-06,
+      "loss": 0.6327,
+      "step": 4643
+    },
+    {
+      "epoch": 1.2864265927977838,
+      "grad_norm": 0.6825493574142456,
+      "learning_rate": 4.471720220704443e-06,
+      "loss": 0.6457,
+      "step": 4644
+    },
+    {
+      "epoch": 1.2867036011080333,
+      "grad_norm": 0.6969654560089111,
+      "learning_rate": 4.471496239792004e-06,
+      "loss": 0.6363,
+      "step": 4645
+    },
+    {
+      "epoch": 1.2869806094182825,
+      "grad_norm": 0.7185841798782349,
+      "learning_rate": 4.471272217019664e-06,
+      "loss": 0.6138,
+      "step": 4646
+    },
+    {
+      "epoch": 1.2872576177285318,
+      "grad_norm": 0.6602373719215393,
+      "learning_rate": 4.4710481523921816e-06,
+      "loss": 0.6046,
+      "step": 4647
+    },
+    {
+      "epoch": 1.2875346260387812,
+      "grad_norm": 0.7047215104103088,
+      "learning_rate": 4.470824045914312e-06,
+      "loss": 0.6184,
+      "step": 4648
+    },
+    {
+      "epoch": 1.2878116343490305,
+      "grad_norm": 0.7656547427177429,
+      "learning_rate": 4.470599897590816e-06,
+      "loss": 0.6555,
+      "step": 4649
+    },
+    {
+      "epoch": 1.2880886426592797,
+      "grad_norm": 0.7046623826026917,
+      "learning_rate": 4.470375707426452e-06,
+      "loss": 0.6694,
+      "step": 4650
+    },
+    {
+      "epoch": 1.2883656509695292,
+      "grad_norm": 0.6936684846878052,
+      "learning_rate": 4.470151475425979e-06,
+      "loss": 0.6186,
+      "step": 4651
+    },
+    {
+      "epoch": 1.2886426592797784,
+      "grad_norm": 0.7161130905151367,
+      "learning_rate": 4.46992720159416e-06,
+      "loss": 0.6827,
+      "step": 4652
+    },
+    {
+      "epoch": 1.2889196675900276,
+      "grad_norm": 0.670766294002533,
+      "learning_rate": 4.469702885935755e-06,
+      "loss": 0.5889,
+      "step": 4653
+    },
+    {
+      "epoch": 1.289196675900277,
+      "grad_norm": 0.6470814347267151,
+      "learning_rate": 4.469478528455529e-06,
+      "loss": 0.6004,
+      "step": 4654
+    },
+    {
+      "epoch": 1.2894736842105263,
+      "grad_norm": 0.7229974865913391,
+      "learning_rate": 4.469254129158242e-06,
+      "loss": 0.6121,
+      "step": 4655
+    },
+    {
+      "epoch": 1.2897506925207756,
+      "grad_norm": 0.6630874276161194,
+      "learning_rate": 4.469029688048663e-06,
+      "loss": 0.6844,
+      "step": 4656
+    },
+    {
+      "epoch": 1.290027700831025,
+      "grad_norm": 0.6928765773773193,
+      "learning_rate": 4.468805205131555e-06,
+      "loss": 0.6489,
+      "step": 4657
+    },
+    {
+      "epoch": 1.2903047091412743,
+      "grad_norm": 0.7021420001983643,
+      "learning_rate": 4.4685806804116835e-06,
+      "loss": 0.624,
+      "step": 4658
+    },
+    {
+      "epoch": 1.2905817174515235,
+      "grad_norm": 0.7199434041976929,
+      "learning_rate": 4.468356113893818e-06,
+      "loss": 0.6424,
+      "step": 4659
+    },
+    {
+      "epoch": 1.290858725761773,
+      "grad_norm": 0.678230881690979,
+      "learning_rate": 4.4681315055827255e-06,
+      "loss": 0.5948,
+      "step": 4660
+    },
+    {
+      "epoch": 1.2911357340720222,
+      "grad_norm": 0.6820054650306702,
+      "learning_rate": 4.467906855483175e-06,
+      "loss": 0.648,
+      "step": 4661
+    },
+    {
+      "epoch": 1.2914127423822714,
+      "grad_norm": 0.7020586729049683,
+      "learning_rate": 4.467682163599937e-06,
+      "loss": 0.648,
+      "step": 4662
+    },
+    {
+      "epoch": 1.291689750692521,
+      "grad_norm": 0.7439424991607666,
+      "learning_rate": 4.467457429937781e-06,
+      "loss": 0.6649,
+      "step": 4663
+    },
+    {
+      "epoch": 1.2919667590027701,
+      "grad_norm": 0.7275158166885376,
+      "learning_rate": 4.46723265450148e-06,
+      "loss": 0.6516,
+      "step": 4664
+    },
+    {
+      "epoch": 1.2922437673130194,
+      "grad_norm": 0.6814894676208496,
+      "learning_rate": 4.467007837295805e-06,
+      "loss": 0.6106,
+      "step": 4665
+    },
+    {
+      "epoch": 1.2925207756232688,
+      "grad_norm": 0.721579372882843,
+      "learning_rate": 4.466782978325531e-06,
+      "loss": 0.6157,
+      "step": 4666
+    },
+    {
+      "epoch": 1.292797783933518,
+      "grad_norm": 0.7090845108032227,
+      "learning_rate": 4.466558077595432e-06,
+      "loss": 0.685,
+      "step": 4667
+    },
+    {
+      "epoch": 1.2930747922437673,
+      "grad_norm": 0.661973237991333,
+      "learning_rate": 4.466333135110283e-06,
+      "loss": 0.6404,
+      "step": 4668
+    },
+    {
+      "epoch": 1.2933518005540166,
+      "grad_norm": 0.7091084718704224,
+      "learning_rate": 4.466108150874859e-06,
+      "loss": 0.6804,
+      "step": 4669
+    },
+    {
+      "epoch": 1.293628808864266,
+      "grad_norm": 0.6906969547271729,
+      "learning_rate": 4.465883124893939e-06,
+      "loss": 0.6756,
+      "step": 4670
+    },
+    {
+      "epoch": 1.2939058171745152,
+      "grad_norm": 0.6625245809555054,
+      "learning_rate": 4.465658057172301e-06,
+      "loss": 0.6563,
+      "step": 4671
+    },
+    {
+      "epoch": 1.2941828254847645,
+      "grad_norm": 0.6626020669937134,
+      "learning_rate": 4.46543294771472e-06,
+      "loss": 0.6524,
+      "step": 4672
+    },
+    {
+      "epoch": 1.2944598337950137,
+      "grad_norm": 0.7468387484550476,
+      "learning_rate": 4.465207796525981e-06,
+      "loss": 0.6784,
+      "step": 4673
+    },
+    {
+      "epoch": 1.2947368421052632,
+      "grad_norm": 0.7022821307182312,
+      "learning_rate": 4.46498260361086e-06,
+      "loss": 0.6629,
+      "step": 4674
+    },
+    {
+      "epoch": 1.2950138504155124,
+      "grad_norm": 0.7802258729934692,
+      "learning_rate": 4.464757368974142e-06,
+      "loss": 0.6862,
+      "step": 4675
+    },
+    {
+      "epoch": 1.2952908587257617,
+      "grad_norm": 0.7289448976516724,
+      "learning_rate": 4.464532092620607e-06,
+      "loss": 0.6343,
+      "step": 4676
+    },
+    {
+      "epoch": 1.2955678670360111,
+      "grad_norm": 0.68182373046875,
+      "learning_rate": 4.464306774555038e-06,
+      "loss": 0.6641,
+      "step": 4677
+    },
+    {
+      "epoch": 1.2958448753462604,
+      "grad_norm": 0.6709579229354858,
+      "learning_rate": 4.46408141478222e-06,
+      "loss": 0.6494,
+      "step": 4678
+    },
+    {
+      "epoch": 1.2961218836565096,
+      "grad_norm": 0.724901556968689,
+      "learning_rate": 4.463856013306939e-06,
+      "loss": 0.6506,
+      "step": 4679
+    },
+    {
+      "epoch": 1.296398891966759,
+      "grad_norm": 0.7291810512542725,
+      "learning_rate": 4.463630570133978e-06,
+      "loss": 0.6156,
+      "step": 4680
+    },
+    {
+      "epoch": 1.2966759002770083,
+      "grad_norm": 0.7094226479530334,
+      "learning_rate": 4.463405085268127e-06,
+      "loss": 0.6355,
+      "step": 4681
+    },
+    {
+      "epoch": 1.2969529085872575,
+      "grad_norm": 0.7184739112854004,
+      "learning_rate": 4.463179558714171e-06,
+      "loss": 0.6385,
+      "step": 4682
+    },
+    {
+      "epoch": 1.297229916897507,
+      "grad_norm": 0.6918147802352905,
+      "learning_rate": 4.4629539904769e-06,
+      "loss": 0.6432,
+      "step": 4683
+    },
+    {
+      "epoch": 1.2975069252077562,
+      "grad_norm": 0.6969414949417114,
+      "learning_rate": 4.462728380561103e-06,
+      "loss": 0.621,
+      "step": 4684
+    },
+    {
+      "epoch": 1.2977839335180055,
+      "grad_norm": 0.7299537062644958,
+      "learning_rate": 4.46250272897157e-06,
+      "loss": 0.7035,
+      "step": 4685
+    },
+    {
+      "epoch": 1.298060941828255,
+      "grad_norm": 0.7747761607170105,
+      "learning_rate": 4.462277035713093e-06,
+      "loss": 0.587,
+      "step": 4686
+    },
+    {
+      "epoch": 1.2983379501385042,
+      "grad_norm": 0.7294924855232239,
+      "learning_rate": 4.462051300790463e-06,
+      "loss": 0.6202,
+      "step": 4687
+    },
+    {
+      "epoch": 1.2986149584487534,
+      "grad_norm": 0.7223609685897827,
+      "learning_rate": 4.461825524208473e-06,
+      "loss": 0.6515,
+      "step": 4688
+    },
+    {
+      "epoch": 1.2988919667590029,
+      "grad_norm": 0.6905951499938965,
+      "learning_rate": 4.461599705971918e-06,
+      "loss": 0.6466,
+      "step": 4689
+    },
+    {
+      "epoch": 1.299168975069252,
+      "grad_norm": 0.7002838253974915,
+      "learning_rate": 4.4613738460855905e-06,
+      "loss": 0.6579,
+      "step": 4690
+    },
+    {
+      "epoch": 1.2994459833795013,
+      "grad_norm": 0.6855634450912476,
+      "learning_rate": 4.461147944554288e-06,
+      "loss": 0.6083,
+      "step": 4691
+    },
+    {
+      "epoch": 1.2997229916897508,
+      "grad_norm": 0.6717777848243713,
+      "learning_rate": 4.460922001382807e-06,
+      "loss": 0.6412,
+      "step": 4692
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.685459554195404,
+      "learning_rate": 4.460696016575944e-06,
+      "loss": 0.6333,
+      "step": 4693
+    },
+    {
+      "epoch": 1.3002770083102493,
+      "grad_norm": 0.6794373393058777,
+      "learning_rate": 4.460469990138497e-06,
+      "loss": 0.6755,
+      "step": 4694
+    },
+    {
+      "epoch": 1.3005540166204987,
+      "grad_norm": 0.6875943541526794,
+      "learning_rate": 4.4602439220752665e-06,
+      "loss": 0.6295,
+      "step": 4695
+    },
+    {
+      "epoch": 1.300831024930748,
+      "grad_norm": 0.6423923373222351,
+      "learning_rate": 4.460017812391051e-06,
+      "loss": 0.6523,
+      "step": 4696
+    },
+    {
+      "epoch": 1.3011080332409972,
+      "grad_norm": 0.6820371150970459,
+      "learning_rate": 4.459791661090652e-06,
+      "loss": 0.5964,
+      "step": 4697
+    },
+    {
+      "epoch": 1.3013850415512465,
+      "grad_norm": 0.6558367013931274,
+      "learning_rate": 4.4595654681788715e-06,
+      "loss": 0.6141,
+      "step": 4698
+    },
+    {
+      "epoch": 1.301662049861496,
+      "grad_norm": 0.7163952589035034,
+      "learning_rate": 4.459339233660512e-06,
+      "loss": 0.6544,
+      "step": 4699
+    },
+    {
+      "epoch": 1.3019390581717452,
+      "grad_norm": 0.7021366357803345,
+      "learning_rate": 4.459112957540377e-06,
+      "loss": 0.6427,
+      "step": 4700
+    },
+    {
+      "epoch": 1.3022160664819944,
+      "grad_norm": 0.7018139362335205,
+      "learning_rate": 4.45888663982327e-06,
+      "loss": 0.6041,
+      "step": 4701
+    },
+    {
+      "epoch": 1.3024930747922436,
+      "grad_norm": 0.6925555467605591,
+      "learning_rate": 4.458660280513998e-06,
+      "loss": 0.6071,
+      "step": 4702
+    },
+    {
+      "epoch": 1.302770083102493,
+      "grad_norm": 0.7855971455574036,
+      "learning_rate": 4.458433879617367e-06,
+      "loss": 0.6324,
+      "step": 4703
+    },
+    {
+      "epoch": 1.3030470914127423,
+      "grad_norm": 0.6972799301147461,
+      "learning_rate": 4.458207437138182e-06,
+      "loss": 0.6331,
+      "step": 4704
+    },
+    {
+      "epoch": 1.3033240997229916,
+      "grad_norm": 0.6723865866661072,
+      "learning_rate": 4.4579809530812535e-06,
+      "loss": 0.6135,
+      "step": 4705
+    },
+    {
+      "epoch": 1.303601108033241,
+      "grad_norm": 0.7214810848236084,
+      "learning_rate": 4.457754427451389e-06,
+      "loss": 0.6299,
+      "step": 4706
+    },
+    {
+      "epoch": 1.3038781163434903,
+      "grad_norm": 0.7141785621643066,
+      "learning_rate": 4.4575278602533975e-06,
+      "loss": 0.6764,
+      "step": 4707
+    },
+    {
+      "epoch": 1.3041551246537395,
+      "grad_norm": 0.6679520010948181,
+      "learning_rate": 4.457301251492092e-06,
+      "loss": 0.636,
+      "step": 4708
+    },
+    {
+      "epoch": 1.304432132963989,
+      "grad_norm": 0.695136308670044,
+      "learning_rate": 4.457074601172282e-06,
+      "loss": 0.6668,
+      "step": 4709
+    },
+    {
+      "epoch": 1.3047091412742382,
+      "grad_norm": 0.7452983260154724,
+      "learning_rate": 4.45684790929878e-06,
+      "loss": 0.63,
+      "step": 4710
+    },
+    {
+      "epoch": 1.3049861495844874,
+      "grad_norm": 0.673240602016449,
+      "learning_rate": 4.4566211758764e-06,
+      "loss": 0.6409,
+      "step": 4711
+    },
+    {
+      "epoch": 1.305263157894737,
+      "grad_norm": 0.6913855671882629,
+      "learning_rate": 4.456394400909956e-06,
+      "loss": 0.6333,
+      "step": 4712
+    },
+    {
+      "epoch": 1.3055401662049861,
+      "grad_norm": 0.6479515433311462,
+      "learning_rate": 4.456167584404262e-06,
+      "loss": 0.5998,
+      "step": 4713
+    },
+    {
+      "epoch": 1.3058171745152354,
+      "grad_norm": 0.7309516072273254,
+      "learning_rate": 4.455940726364135e-06,
+      "loss": 0.6845,
+      "step": 4714
+    },
+    {
+      "epoch": 1.3060941828254848,
+      "grad_norm": 0.6839296817779541,
+      "learning_rate": 4.455713826794391e-06,
+      "loss": 0.6567,
+      "step": 4715
+    },
+    {
+      "epoch": 1.306371191135734,
+      "grad_norm": 0.6930821537971497,
+      "learning_rate": 4.455486885699849e-06,
+      "loss": 0.5875,
+      "step": 4716
+    },
+    {
+      "epoch": 1.3066481994459833,
+      "grad_norm": 0.7218079566955566,
+      "learning_rate": 4.455259903085326e-06,
+      "loss": 0.6062,
+      "step": 4717
+    },
+    {
+      "epoch": 1.3069252077562328,
+      "grad_norm": 0.6791582107543945,
+      "learning_rate": 4.455032878955643e-06,
+      "loss": 0.6485,
+      "step": 4718
+    },
+    {
+      "epoch": 1.307202216066482,
+      "grad_norm": 0.690004825592041,
+      "learning_rate": 4.454805813315619e-06,
+      "loss": 0.6284,
+      "step": 4719
+    },
+    {
+      "epoch": 1.3074792243767313,
+      "grad_norm": 0.7091121673583984,
+      "learning_rate": 4.454578706170075e-06,
+      "loss": 0.5778,
+      "step": 4720
+    },
+    {
+      "epoch": 1.3077562326869807,
+      "grad_norm": 0.6884284615516663,
+      "learning_rate": 4.4543515575238336e-06,
+      "loss": 0.6224,
+      "step": 4721
+    },
+    {
+      "epoch": 1.30803324099723,
+      "grad_norm": 0.7247439622879028,
+      "learning_rate": 4.454124367381718e-06,
+      "loss": 0.6518,
+      "step": 4722
+    },
+    {
+      "epoch": 1.3083102493074792,
+      "grad_norm": 0.7225871682167053,
+      "learning_rate": 4.453897135748552e-06,
+      "loss": 0.6391,
+      "step": 4723
+    },
+    {
+      "epoch": 1.3085872576177286,
+      "grad_norm": 0.7072814702987671,
+      "learning_rate": 4.45366986262916e-06,
+      "loss": 0.6758,
+      "step": 4724
+    },
+    {
+      "epoch": 1.3088642659279779,
+      "grad_norm": 0.7352576851844788,
+      "learning_rate": 4.453442548028367e-06,
+      "loss": 0.6547,
+      "step": 4725
+    },
+    {
+      "epoch": 1.3091412742382271,
+      "grad_norm": 0.7073747515678406,
+      "learning_rate": 4.453215191951002e-06,
+      "loss": 0.594,
+      "step": 4726
+    },
+    {
+      "epoch": 1.3094182825484766,
+      "grad_norm": 0.6708406805992126,
+      "learning_rate": 4.452987794401889e-06,
+      "loss": 0.6303,
+      "step": 4727
+    },
+    {
+      "epoch": 1.3096952908587258,
+      "grad_norm": 0.7301291823387146,
+      "learning_rate": 4.452760355385858e-06,
+      "loss": 0.6228,
+      "step": 4728
+    },
+    {
+      "epoch": 1.309972299168975,
+      "grad_norm": 0.7150014638900757,
+      "learning_rate": 4.452532874907738e-06,
+      "loss": 0.6325,
+      "step": 4729
+    },
+    {
+      "epoch": 1.3102493074792243,
+      "grad_norm": 0.6688306331634521,
+      "learning_rate": 4.452305352972359e-06,
+      "loss": 0.6629,
+      "step": 4730
+    },
+    {
+      "epoch": 1.3105263157894738,
+      "grad_norm": 0.6686022877693176,
+      "learning_rate": 4.452077789584551e-06,
+      "loss": 0.6015,
+      "step": 4731
+    },
+    {
+      "epoch": 1.310803324099723,
+      "grad_norm": 0.728378176689148,
+      "learning_rate": 4.451850184749147e-06,
+      "loss": 0.6728,
+      "step": 4732
+    },
+    {
+      "epoch": 1.3110803324099722,
+      "grad_norm": 0.6999015212059021,
+      "learning_rate": 4.451622538470979e-06,
+      "loss": 0.6272,
+      "step": 4733
+    },
+    {
+      "epoch": 1.3113573407202215,
+      "grad_norm": 0.7011626362800598,
+      "learning_rate": 4.451394850754881e-06,
+      "loss": 0.6667,
+      "step": 4734
+    },
+    {
+      "epoch": 1.311634349030471,
+      "grad_norm": 0.7036128044128418,
+      "learning_rate": 4.4511671216056865e-06,
+      "loss": 0.6376,
+      "step": 4735
+    },
+    {
+      "epoch": 1.3119113573407202,
+      "grad_norm": 0.6627241373062134,
+      "learning_rate": 4.450939351028231e-06,
+      "loss": 0.6393,
+      "step": 4736
+    },
+    {
+      "epoch": 1.3121883656509694,
+      "grad_norm": 0.692579448223114,
+      "learning_rate": 4.450711539027352e-06,
+      "loss": 0.6152,
+      "step": 4737
+    },
+    {
+      "epoch": 1.3124653739612189,
+      "grad_norm": 0.6422987580299377,
+      "learning_rate": 4.450483685607884e-06,
+      "loss": 0.6209,
+      "step": 4738
+    },
+    {
+      "epoch": 1.3127423822714681,
+      "grad_norm": 0.6677369475364685,
+      "learning_rate": 4.450255790774667e-06,
+      "loss": 0.6189,
+      "step": 4739
+    },
+    {
+      "epoch": 1.3130193905817173,
+      "grad_norm": 0.696083664894104,
+      "learning_rate": 4.450027854532539e-06,
+      "loss": 0.6347,
+      "step": 4740
+    },
+    {
+      "epoch": 1.3132963988919668,
+      "grad_norm": 0.702476441860199,
+      "learning_rate": 4.449799876886341e-06,
+      "loss": 0.6398,
+      "step": 4741
+    },
+    {
+      "epoch": 1.313573407202216,
+      "grad_norm": 0.7123687267303467,
+      "learning_rate": 4.449571857840911e-06,
+      "loss": 0.6288,
+      "step": 4742
+    },
+    {
+      "epoch": 1.3138504155124653,
+      "grad_norm": 0.7451086044311523,
+      "learning_rate": 4.449343797401092e-06,
+      "loss": 0.656,
+      "step": 4743
+    },
+    {
+      "epoch": 1.3141274238227147,
+      "grad_norm": 0.7208288311958313,
+      "learning_rate": 4.449115695571727e-06,
+      "loss": 0.6844,
+      "step": 4744
+    },
+    {
+      "epoch": 1.314404432132964,
+      "grad_norm": 0.6745233535766602,
+      "learning_rate": 4.448887552357658e-06,
+      "loss": 0.6571,
+      "step": 4745
+    },
+    {
+      "epoch": 1.3146814404432132,
+      "grad_norm": 0.7078210115432739,
+      "learning_rate": 4.448659367763729e-06,
+      "loss": 0.6656,
+      "step": 4746
+    },
+    {
+      "epoch": 1.3149584487534627,
+      "grad_norm": 0.6784594058990479,
+      "learning_rate": 4.448431141794784e-06,
+      "loss": 0.6252,
+      "step": 4747
+    },
+    {
+      "epoch": 1.315235457063712,
+      "grad_norm": 0.7477366328239441,
+      "learning_rate": 4.4482028744556725e-06,
+      "loss": 0.647,
+      "step": 4748
+    },
+    {
+      "epoch": 1.3155124653739612,
+      "grad_norm": 0.7212636470794678,
+      "learning_rate": 4.447974565751237e-06,
+      "loss": 0.6541,
+      "step": 4749
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.7099882960319519,
+      "learning_rate": 4.447746215686328e-06,
+      "loss": 0.625,
+      "step": 4750
+    },
+    {
+      "epoch": 1.3160664819944599,
+      "grad_norm": 0.6715225577354431,
+      "learning_rate": 4.447517824265793e-06,
+      "loss": 0.6326,
+      "step": 4751
+    },
+    {
+      "epoch": 1.316343490304709,
+      "grad_norm": 0.7152224779129028,
+      "learning_rate": 4.44728939149448e-06,
+      "loss": 0.6436,
+      "step": 4752
+    },
+    {
+      "epoch": 1.3166204986149586,
+      "grad_norm": 0.6906647086143494,
+      "learning_rate": 4.44706091737724e-06,
+      "loss": 0.6627,
+      "step": 4753
+    },
+    {
+      "epoch": 1.3168975069252078,
+      "grad_norm": 0.7126035094261169,
+      "learning_rate": 4.446832401918927e-06,
+      "loss": 0.6472,
+      "step": 4754
+    },
+    {
+      "epoch": 1.317174515235457,
+      "grad_norm": 0.6664754748344421,
+      "learning_rate": 4.446603845124388e-06,
+      "loss": 0.6275,
+      "step": 4755
+    },
+    {
+      "epoch": 1.3174515235457065,
+      "grad_norm": 0.654424786567688,
+      "learning_rate": 4.4463752469984796e-06,
+      "loss": 0.6258,
+      "step": 4756
+    },
+    {
+      "epoch": 1.3177285318559557,
+      "grad_norm": 0.7223296165466309,
+      "learning_rate": 4.446146607546053e-06,
+      "loss": 0.6377,
+      "step": 4757
+    },
+    {
+      "epoch": 1.318005540166205,
+      "grad_norm": 0.7383649349212646,
+      "learning_rate": 4.445917926771966e-06,
+      "loss": 0.6745,
+      "step": 4758
+    },
+    {
+      "epoch": 1.3182825484764544,
+      "grad_norm": 0.6735358834266663,
+      "learning_rate": 4.445689204681071e-06,
+      "loss": 0.6315,
+      "step": 4759
+    },
+    {
+      "epoch": 1.3185595567867037,
+      "grad_norm": 0.6980936527252197,
+      "learning_rate": 4.445460441278225e-06,
+      "loss": 0.6214,
+      "step": 4760
+    },
+    {
+      "epoch": 1.318836565096953,
+      "grad_norm": 0.7276691794395447,
+      "learning_rate": 4.445231636568286e-06,
+      "loss": 0.671,
+      "step": 4761
+    },
+    {
+      "epoch": 1.3191135734072021,
+      "grad_norm": 0.7002887725830078,
+      "learning_rate": 4.445002790556111e-06,
+      "loss": 0.6264,
+      "step": 4762
+    },
+    {
+      "epoch": 1.3193905817174516,
+      "grad_norm": 0.7233803868293762,
+      "learning_rate": 4.444773903246561e-06,
+      "loss": 0.6311,
+      "step": 4763
+    },
+    {
+      "epoch": 1.3196675900277008,
+      "grad_norm": 0.7100148797035217,
+      "learning_rate": 4.444544974644493e-06,
+      "loss": 0.6274,
+      "step": 4764
+    },
+    {
+      "epoch": 1.31994459833795,
+      "grad_norm": 0.6861152648925781,
+      "learning_rate": 4.4443160047547716e-06,
+      "loss": 0.6194,
+      "step": 4765
+    },
+    {
+      "epoch": 1.3202216066481993,
+      "grad_norm": 0.7061241865158081,
+      "learning_rate": 4.444086993582255e-06,
+      "loss": 0.6851,
+      "step": 4766
+    },
+    {
+      "epoch": 1.3204986149584488,
+      "grad_norm": 0.6619032621383667,
+      "learning_rate": 4.443857941131807e-06,
+      "loss": 0.6126,
+      "step": 4767
+    },
+    {
+      "epoch": 1.320775623268698,
+      "grad_norm": 0.7326112985610962,
+      "learning_rate": 4.4436288474082905e-06,
+      "loss": 0.6274,
+      "step": 4768
+    },
+    {
+      "epoch": 1.3210526315789473,
+      "grad_norm": 0.6847018599510193,
+      "learning_rate": 4.44339971241657e-06,
+      "loss": 0.6524,
+      "step": 4769
+    },
+    {
+      "epoch": 1.3213296398891967,
+      "grad_norm": 0.6778106689453125,
+      "learning_rate": 4.443170536161512e-06,
+      "loss": 0.6166,
+      "step": 4770
+    },
+    {
+      "epoch": 1.321606648199446,
+      "grad_norm": 0.6772767901420593,
+      "learning_rate": 4.44294131864798e-06,
+      "loss": 0.6339,
+      "step": 4771
+    },
+    {
+      "epoch": 1.3218836565096952,
+      "grad_norm": 0.6837650537490845,
+      "learning_rate": 4.442712059880842e-06,
+      "loss": 0.6382,
+      "step": 4772
+    },
+    {
+      "epoch": 1.3221606648199447,
+      "grad_norm": 0.6830680966377258,
+      "learning_rate": 4.442482759864966e-06,
+      "loss": 0.6558,
+      "step": 4773
+    },
+    {
+      "epoch": 1.322437673130194,
+      "grad_norm": 0.7215142250061035,
+      "learning_rate": 4.442253418605221e-06,
+      "loss": 0.634,
+      "step": 4774
+    },
+    {
+      "epoch": 1.3227146814404431,
+      "grad_norm": 0.6660498380661011,
+      "learning_rate": 4.442024036106476e-06,
+      "loss": 0.6656,
+      "step": 4775
+    },
+    {
+      "epoch": 1.3229916897506926,
+      "grad_norm": 0.6940247416496277,
+      "learning_rate": 4.441794612373601e-06,
+      "loss": 0.6765,
+      "step": 4776
+    },
+    {
+      "epoch": 1.3232686980609418,
+      "grad_norm": 0.7326346635818481,
+      "learning_rate": 4.4415651474114664e-06,
+      "loss": 0.6547,
+      "step": 4777
+    },
+    {
+      "epoch": 1.323545706371191,
+      "grad_norm": 0.6846688389778137,
+      "learning_rate": 4.441335641224947e-06,
+      "loss": 0.6461,
+      "step": 4778
+    },
+    {
+      "epoch": 1.3238227146814405,
+      "grad_norm": 0.696536660194397,
+      "learning_rate": 4.441106093818914e-06,
+      "loss": 0.6428,
+      "step": 4779
+    },
+    {
+      "epoch": 1.3240997229916898,
+      "grad_norm": 0.7244516015052795,
+      "learning_rate": 4.440876505198242e-06,
+      "loss": 0.6311,
+      "step": 4780
+    },
+    {
+      "epoch": 1.324376731301939,
+      "grad_norm": 0.7271976470947266,
+      "learning_rate": 4.440646875367805e-06,
+      "loss": 0.6445,
+      "step": 4781
+    },
+    {
+      "epoch": 1.3246537396121885,
+      "grad_norm": 0.6603023409843445,
+      "learning_rate": 4.440417204332477e-06,
+      "loss": 0.5799,
+      "step": 4782
+    },
+    {
+      "epoch": 1.3249307479224377,
+      "grad_norm": 0.6802365183830261,
+      "learning_rate": 4.440187492097139e-06,
+      "loss": 0.6408,
+      "step": 4783
+    },
+    {
+      "epoch": 1.325207756232687,
+      "grad_norm": 0.685371458530426,
+      "learning_rate": 4.4399577386666645e-06,
+      "loss": 0.6284,
+      "step": 4784
+    },
+    {
+      "epoch": 1.3254847645429364,
+      "grad_norm": 0.752849280834198,
+      "learning_rate": 4.439727944045934e-06,
+      "loss": 0.6522,
+      "step": 4785
+    },
+    {
+      "epoch": 1.3257617728531856,
+      "grad_norm": 0.7088627815246582,
+      "learning_rate": 4.4394981082398254e-06,
+      "loss": 0.6221,
+      "step": 4786
+    },
+    {
+      "epoch": 1.3260387811634349,
+      "grad_norm": 0.6534282565116882,
+      "learning_rate": 4.439268231253219e-06,
+      "loss": 0.6515,
+      "step": 4787
+    },
+    {
+      "epoch": 1.3263157894736843,
+      "grad_norm": 0.7163136005401611,
+      "learning_rate": 4.439038313090996e-06,
+      "loss": 0.6562,
+      "step": 4788
+    },
+    {
+      "epoch": 1.3265927977839336,
+      "grad_norm": 0.7224626541137695,
+      "learning_rate": 4.438808353758038e-06,
+      "loss": 0.6291,
+      "step": 4789
+    },
+    {
+      "epoch": 1.3268698060941828,
+      "grad_norm": 0.7121604084968567,
+      "learning_rate": 4.438578353259227e-06,
+      "loss": 0.6915,
+      "step": 4790
+    },
+    {
+      "epoch": 1.327146814404432,
+      "grad_norm": 0.7014841437339783,
+      "learning_rate": 4.438348311599446e-06,
+      "loss": 0.6353,
+      "step": 4791
+    },
+    {
+      "epoch": 1.3274238227146815,
+      "grad_norm": 0.7054080367088318,
+      "learning_rate": 4.438118228783582e-06,
+      "loss": 0.6322,
+      "step": 4792
+    },
+    {
+      "epoch": 1.3277008310249307,
+      "grad_norm": 0.7404279708862305,
+      "learning_rate": 4.437888104816518e-06,
+      "loss": 0.6305,
+      "step": 4793
+    },
+    {
+      "epoch": 1.32797783933518,
+      "grad_norm": 0.7308083772659302,
+      "learning_rate": 4.4376579397031406e-06,
+      "loss": 0.6363,
+      "step": 4794
+    },
+    {
+      "epoch": 1.3282548476454292,
+      "grad_norm": 0.702226996421814,
+      "learning_rate": 4.437427733448337e-06,
+      "loss": 0.5752,
+      "step": 4795
+    },
+    {
+      "epoch": 1.3285318559556787,
+      "grad_norm": 0.6674501895904541,
+      "learning_rate": 4.437197486056994e-06,
+      "loss": 0.6565,
+      "step": 4796
+    },
+    {
+      "epoch": 1.328808864265928,
+      "grad_norm": 0.6679547429084778,
+      "learning_rate": 4.436967197534003e-06,
+      "loss": 0.6591,
+      "step": 4797
+    },
+    {
+      "epoch": 1.3290858725761772,
+      "grad_norm": 0.6631932854652405,
+      "learning_rate": 4.436736867884252e-06,
+      "loss": 0.6463,
+      "step": 4798
+    },
+    {
+      "epoch": 1.3293628808864266,
+      "grad_norm": 0.7053595185279846,
+      "learning_rate": 4.436506497112629e-06,
+      "loss": 0.6494,
+      "step": 4799
+    },
+    {
+      "epoch": 1.3296398891966759,
+      "grad_norm": 0.7127822637557983,
+      "learning_rate": 4.43627608522403e-06,
+      "loss": 0.6781,
+      "step": 4800
+    },
+    {
+      "epoch": 1.329916897506925,
+      "grad_norm": 0.667141318321228,
+      "learning_rate": 4.436045632223344e-06,
+      "loss": 0.617,
+      "step": 4801
+    },
+    {
+      "epoch": 1.3301939058171746,
+      "grad_norm": 0.6789475083351135,
+      "learning_rate": 4.4358151381154654e-06,
+      "loss": 0.6538,
+      "step": 4802
+    },
+    {
+      "epoch": 1.3304709141274238,
+      "grad_norm": 0.6923264861106873,
+      "learning_rate": 4.435584602905289e-06,
+      "loss": 0.6294,
+      "step": 4803
+    },
+    {
+      "epoch": 1.330747922437673,
+      "grad_norm": 0.7080187797546387,
+      "learning_rate": 4.435354026597707e-06,
+      "loss": 0.625,
+      "step": 4804
+    },
+    {
+      "epoch": 1.3310249307479225,
+      "grad_norm": 0.7222117185592651,
+      "learning_rate": 4.435123409197617e-06,
+      "loss": 0.6589,
+      "step": 4805
+    },
+    {
+      "epoch": 1.3313019390581717,
+      "grad_norm": 0.6662753820419312,
+      "learning_rate": 4.434892750709915e-06,
+      "loss": 0.5916,
+      "step": 4806
+    },
+    {
+      "epoch": 1.331578947368421,
+      "grad_norm": 0.730897068977356,
+      "learning_rate": 4.434662051139499e-06,
+      "loss": 0.6537,
+      "step": 4807
+    },
+    {
+      "epoch": 1.3318559556786704,
+      "grad_norm": 0.719174861907959,
+      "learning_rate": 4.434431310491267e-06,
+      "loss": 0.6498,
+      "step": 4808
+    },
+    {
+      "epoch": 1.3321329639889197,
+      "grad_norm": 0.6620305776596069,
+      "learning_rate": 4.434200528770118e-06,
+      "loss": 0.6247,
+      "step": 4809
+    },
+    {
+      "epoch": 1.332409972299169,
+      "grad_norm": 0.6843666434288025,
+      "learning_rate": 4.433969705980953e-06,
+      "loss": 0.6083,
+      "step": 4810
+    },
+    {
+      "epoch": 1.3326869806094184,
+      "grad_norm": 0.7134045362472534,
+      "learning_rate": 4.4337388421286724e-06,
+      "loss": 0.6956,
+      "step": 4811
+    },
+    {
+      "epoch": 1.3329639889196676,
+      "grad_norm": 0.6772432327270508,
+      "learning_rate": 4.433507937218177e-06,
+      "loss": 0.652,
+      "step": 4812
+    },
+    {
+      "epoch": 1.3332409972299168,
+      "grad_norm": 0.6777689456939697,
+      "learning_rate": 4.433276991254371e-06,
+      "loss": 0.6256,
+      "step": 4813
+    },
+    {
+      "epoch": 1.3335180055401663,
+      "grad_norm": 0.7466452717781067,
+      "learning_rate": 4.4330460042421584e-06,
+      "loss": 0.6231,
+      "step": 4814
+    },
+    {
+      "epoch": 1.3337950138504155,
+      "grad_norm": 0.6792182326316833,
+      "learning_rate": 4.4328149761864415e-06,
+      "loss": 0.6025,
+      "step": 4815
+    },
+    {
+      "epoch": 1.3340720221606648,
+      "grad_norm": 0.6836180090904236,
+      "learning_rate": 4.432583907092127e-06,
+      "loss": 0.6066,
+      "step": 4816
+    },
+    {
+      "epoch": 1.3343490304709142,
+      "grad_norm": 0.7040029168128967,
+      "learning_rate": 4.432352796964121e-06,
+      "loss": 0.662,
+      "step": 4817
+    },
+    {
+      "epoch": 1.3346260387811635,
+      "grad_norm": 0.6517881751060486,
+      "learning_rate": 4.432121645807331e-06,
+      "loss": 0.5577,
+      "step": 4818
+    },
+    {
+      "epoch": 1.3349030470914127,
+      "grad_norm": 0.6596935987472534,
+      "learning_rate": 4.431890453626664e-06,
+      "loss": 0.6183,
+      "step": 4819
+    },
+    {
+      "epoch": 1.3351800554016622,
+      "grad_norm": 0.6841338276863098,
+      "learning_rate": 4.4316592204270284e-06,
+      "loss": 0.6077,
+      "step": 4820
+    },
+    {
+      "epoch": 1.3354570637119114,
+      "grad_norm": 0.6922977566719055,
+      "learning_rate": 4.4314279462133345e-06,
+      "loss": 0.6343,
+      "step": 4821
+    },
+    {
+      "epoch": 1.3357340720221607,
+      "grad_norm": 0.7033997178077698,
+      "learning_rate": 4.431196630990494e-06,
+      "loss": 0.6277,
+      "step": 4822
+    },
+    {
+      "epoch": 1.33601108033241,
+      "grad_norm": 0.7048894762992859,
+      "learning_rate": 4.430965274763417e-06,
+      "loss": 0.6105,
+      "step": 4823
+    },
+    {
+      "epoch": 1.3362880886426594,
+      "grad_norm": 0.6983682513237,
+      "learning_rate": 4.430733877537016e-06,
+      "loss": 0.6152,
+      "step": 4824
+    },
+    {
+      "epoch": 1.3365650969529086,
+      "grad_norm": 0.7080015540122986,
+      "learning_rate": 4.430502439316204e-06,
+      "loss": 0.6336,
+      "step": 4825
+    },
+    {
+      "epoch": 1.3368421052631578,
+      "grad_norm": 0.6375042796134949,
+      "learning_rate": 4.4302709601058955e-06,
+      "loss": 0.6112,
+      "step": 4826
+    },
+    {
+      "epoch": 1.337119113573407,
+      "grad_norm": 0.6347265243530273,
+      "learning_rate": 4.4300394399110055e-06,
+      "loss": 0.6033,
+      "step": 4827
+    },
+    {
+      "epoch": 1.3373961218836565,
+      "grad_norm": 0.7151878476142883,
+      "learning_rate": 4.429807878736448e-06,
+      "loss": 0.6392,
+      "step": 4828
+    },
+    {
+      "epoch": 1.3376731301939058,
+      "grad_norm": 0.6534563899040222,
+      "learning_rate": 4.429576276587143e-06,
+      "loss": 0.645,
+      "step": 4829
+    },
+    {
+      "epoch": 1.337950138504155,
+      "grad_norm": 0.7110674977302551,
+      "learning_rate": 4.429344633468005e-06,
+      "loss": 0.6271,
+      "step": 4830
+    },
+    {
+      "epoch": 1.3382271468144045,
+      "grad_norm": 0.6874481439590454,
+      "learning_rate": 4.429112949383953e-06,
+      "loss": 0.6397,
+      "step": 4831
+    },
+    {
+      "epoch": 1.3385041551246537,
+      "grad_norm": 0.6635070443153381,
+      "learning_rate": 4.428881224339907e-06,
+      "loss": 0.6297,
+      "step": 4832
+    },
+    {
+      "epoch": 1.338781163434903,
+      "grad_norm": 0.6539145112037659,
+      "learning_rate": 4.4286494583407875e-06,
+      "loss": 0.585,
+      "step": 4833
+    },
+    {
+      "epoch": 1.3390581717451524,
+      "grad_norm": 0.7016680836677551,
+      "learning_rate": 4.428417651391514e-06,
+      "loss": 0.575,
+      "step": 4834
+    },
+    {
+      "epoch": 1.3393351800554016,
+      "grad_norm": 0.7028341889381409,
+      "learning_rate": 4.428185803497009e-06,
+      "loss": 0.5788,
+      "step": 4835
+    },
+    {
+      "epoch": 1.3396121883656509,
+      "grad_norm": 0.6498736143112183,
+      "learning_rate": 4.427953914662196e-06,
+      "loss": 0.6205,
+      "step": 4836
+    },
+    {
+      "epoch": 1.3398891966759003,
+      "grad_norm": 0.6495695114135742,
+      "learning_rate": 4.427721984891998e-06,
+      "loss": 0.5968,
+      "step": 4837
+    },
+    {
+      "epoch": 1.3401662049861496,
+      "grad_norm": 0.6960228681564331,
+      "learning_rate": 4.427490014191339e-06,
+      "loss": 0.6364,
+      "step": 4838
+    },
+    {
+      "epoch": 1.3404432132963988,
+      "grad_norm": 0.681462287902832,
+      "learning_rate": 4.4272580025651445e-06,
+      "loss": 0.6549,
+      "step": 4839
+    },
+    {
+      "epoch": 1.3407202216066483,
+      "grad_norm": 0.6907786130905151,
+      "learning_rate": 4.427025950018341e-06,
+      "loss": 0.6461,
+      "step": 4840
+    },
+    {
+      "epoch": 1.3409972299168975,
+      "grad_norm": 0.7233545780181885,
+      "learning_rate": 4.4267938565558555e-06,
+      "loss": 0.6735,
+      "step": 4841
+    },
+    {
+      "epoch": 1.3412742382271468,
+      "grad_norm": 0.6877539157867432,
+      "learning_rate": 4.426561722182616e-06,
+      "loss": 0.6173,
+      "step": 4842
+    },
+    {
+      "epoch": 1.3415512465373962,
+      "grad_norm": 0.7026717662811279,
+      "learning_rate": 4.426329546903552e-06,
+      "loss": 0.6466,
+      "step": 4843
+    },
+    {
+      "epoch": 1.3418282548476455,
+      "grad_norm": 0.7012684941291809,
+      "learning_rate": 4.426097330723591e-06,
+      "loss": 0.6601,
+      "step": 4844
+    },
+    {
+      "epoch": 1.3421052631578947,
+      "grad_norm": 0.6691379547119141,
+      "learning_rate": 4.425865073647666e-06,
+      "loss": 0.6278,
+      "step": 4845
+    },
+    {
+      "epoch": 1.3423822714681442,
+      "grad_norm": 0.7390052676200867,
+      "learning_rate": 4.4256327756807075e-06,
+      "loss": 0.664,
+      "step": 4846
+    },
+    {
+      "epoch": 1.3426592797783934,
+      "grad_norm": 0.6644761562347412,
+      "learning_rate": 4.425400436827647e-06,
+      "loss": 0.6076,
+      "step": 4847
+    },
+    {
+      "epoch": 1.3429362880886426,
+      "grad_norm": 0.705655038356781,
+      "learning_rate": 4.425168057093418e-06,
+      "loss": 0.6452,
+      "step": 4848
+    },
+    {
+      "epoch": 1.343213296398892,
+      "grad_norm": 0.7211757898330688,
+      "learning_rate": 4.424935636482955e-06,
+      "loss": 0.6577,
+      "step": 4849
+    },
+    {
+      "epoch": 1.3434903047091413,
+      "grad_norm": 0.668464720249176,
+      "learning_rate": 4.424703175001194e-06,
+      "loss": 0.5973,
+      "step": 4850
+    },
+    {
+      "epoch": 1.3437673130193906,
+      "grad_norm": 0.6970594525337219,
+      "learning_rate": 4.4244706726530676e-06,
+      "loss": 0.6887,
+      "step": 4851
+    },
+    {
+      "epoch": 1.34404432132964,
+      "grad_norm": 0.6972764134407043,
+      "learning_rate": 4.424238129443515e-06,
+      "loss": 0.6646,
+      "step": 4852
+    },
+    {
+      "epoch": 1.3443213296398893,
+      "grad_norm": 0.6718432307243347,
+      "learning_rate": 4.424005545377473e-06,
+      "loss": 0.5996,
+      "step": 4853
+    },
+    {
+      "epoch": 1.3445983379501385,
+      "grad_norm": 0.6845982074737549,
+      "learning_rate": 4.42377292045988e-06,
+      "loss": 0.6474,
+      "step": 4854
+    },
+    {
+      "epoch": 1.3448753462603877,
+      "grad_norm": 0.6861254572868347,
+      "learning_rate": 4.4235402546956756e-06,
+      "loss": 0.6814,
+      "step": 4855
+    },
+    {
+      "epoch": 1.3451523545706372,
+      "grad_norm": 0.6887468099594116,
+      "learning_rate": 4.423307548089798e-06,
+      "loss": 0.6538,
+      "step": 4856
+    },
+    {
+      "epoch": 1.3454293628808864,
+      "grad_norm": 0.6625999212265015,
+      "learning_rate": 4.423074800647191e-06,
+      "loss": 0.6455,
+      "step": 4857
+    },
+    {
+      "epoch": 1.3457063711911357,
+      "grad_norm": 0.7102819681167603,
+      "learning_rate": 4.422842012372794e-06,
+      "loss": 0.6903,
+      "step": 4858
+    },
+    {
+      "epoch": 1.345983379501385,
+      "grad_norm": 0.687042236328125,
+      "learning_rate": 4.422609183271551e-06,
+      "loss": 0.6013,
+      "step": 4859
+    },
+    {
+      "epoch": 1.3462603878116344,
+      "grad_norm": 0.7196336984634399,
+      "learning_rate": 4.422376313348405e-06,
+      "loss": 0.6243,
+      "step": 4860
+    },
+    {
+      "epoch": 1.3465373961218836,
+      "grad_norm": 0.6799708008766174,
+      "learning_rate": 4.422143402608302e-06,
+      "loss": 0.6289,
+      "step": 4861
+    },
+    {
+      "epoch": 1.3468144044321329,
+      "grad_norm": 0.7131449580192566,
+      "learning_rate": 4.421910451056184e-06,
+      "loss": 0.6736,
+      "step": 4862
+    },
+    {
+      "epoch": 1.3470914127423823,
+      "grad_norm": 0.6889610290527344,
+      "learning_rate": 4.421677458697e-06,
+      "loss": 0.6175,
+      "step": 4863
+    },
+    {
+      "epoch": 1.3473684210526315,
+      "grad_norm": 0.6625426411628723,
+      "learning_rate": 4.421444425535696e-06,
+      "loss": 0.597,
+      "step": 4864
+    },
+    {
+      "epoch": 1.3476454293628808,
+      "grad_norm": 0.7058026790618896,
+      "learning_rate": 4.42121135157722e-06,
+      "loss": 0.6549,
+      "step": 4865
+    },
+    {
+      "epoch": 1.3479224376731302,
+      "grad_norm": 0.6828805208206177,
+      "learning_rate": 4.420978236826521e-06,
+      "loss": 0.6195,
+      "step": 4866
+    },
+    {
+      "epoch": 1.3481994459833795,
+      "grad_norm": 0.7049974203109741,
+      "learning_rate": 4.420745081288549e-06,
+      "loss": 0.6275,
+      "step": 4867
+    },
+    {
+      "epoch": 1.3484764542936287,
+      "grad_norm": 0.7258075475692749,
+      "learning_rate": 4.4205118849682524e-06,
+      "loss": 0.6864,
+      "step": 4868
+    },
+    {
+      "epoch": 1.3487534626038782,
+      "grad_norm": 0.721996545791626,
+      "learning_rate": 4.420278647870585e-06,
+      "loss": 0.6157,
+      "step": 4869
+    },
+    {
+      "epoch": 1.3490304709141274,
+      "grad_norm": 0.7096882462501526,
+      "learning_rate": 4.420045370000497e-06,
+      "loss": 0.6405,
+      "step": 4870
+    },
+    {
+      "epoch": 1.3493074792243767,
+      "grad_norm": 0.71756511926651,
+      "learning_rate": 4.419812051362944e-06,
+      "loss": 0.5924,
+      "step": 4871
+    },
+    {
+      "epoch": 1.3495844875346261,
+      "grad_norm": 0.7162507176399231,
+      "learning_rate": 4.419578691962878e-06,
+      "loss": 0.6705,
+      "step": 4872
+    },
+    {
+      "epoch": 1.3498614958448754,
+      "grad_norm": 0.7038041949272156,
+      "learning_rate": 4.419345291805254e-06,
+      "loss": 0.6778,
+      "step": 4873
+    },
+    {
+      "epoch": 1.3501385041551246,
+      "grad_norm": 0.6972736716270447,
+      "learning_rate": 4.4191118508950286e-06,
+      "loss": 0.6447,
+      "step": 4874
+    },
+    {
+      "epoch": 1.350415512465374,
+      "grad_norm": 0.7007998824119568,
+      "learning_rate": 4.418878369237157e-06,
+      "loss": 0.5849,
+      "step": 4875
+    },
+    {
+      "epoch": 1.3506925207756233,
+      "grad_norm": 0.6676672697067261,
+      "learning_rate": 4.418644846836597e-06,
+      "loss": 0.6379,
+      "step": 4876
+    },
+    {
+      "epoch": 1.3509695290858725,
+      "grad_norm": 0.6992767453193665,
+      "learning_rate": 4.418411283698309e-06,
+      "loss": 0.7062,
+      "step": 4877
+    },
+    {
+      "epoch": 1.351246537396122,
+      "grad_norm": 0.722588062286377,
+      "learning_rate": 4.418177679827248e-06,
+      "loss": 0.6823,
+      "step": 4878
+    },
+    {
+      "epoch": 1.3515235457063712,
+      "grad_norm": 0.676636815071106,
+      "learning_rate": 4.417944035228377e-06,
+      "loss": 0.6406,
+      "step": 4879
+    },
+    {
+      "epoch": 1.3518005540166205,
+      "grad_norm": 0.7693878412246704,
+      "learning_rate": 4.4177103499066575e-06,
+      "loss": 0.6849,
+      "step": 4880
+    },
+    {
+      "epoch": 1.35207756232687,
+      "grad_norm": 0.8314934968948364,
+      "learning_rate": 4.41747662386705e-06,
+      "loss": 0.6413,
+      "step": 4881
+    },
+    {
+      "epoch": 1.3523545706371192,
+      "grad_norm": 0.6808057427406311,
+      "learning_rate": 4.417242857114515e-06,
+      "loss": 0.6657,
+      "step": 4882
+    },
+    {
+      "epoch": 1.3526315789473684,
+      "grad_norm": 0.6932368278503418,
+      "learning_rate": 4.41700904965402e-06,
+      "loss": 0.6405,
+      "step": 4883
+    },
+    {
+      "epoch": 1.3529085872576179,
+      "grad_norm": 0.683689296245575,
+      "learning_rate": 4.416775201490527e-06,
+      "loss": 0.5958,
+      "step": 4884
+    },
+    {
+      "epoch": 1.353185595567867,
+      "grad_norm": 0.7213323712348938,
+      "learning_rate": 4.4165413126290015e-06,
+      "loss": 0.6122,
+      "step": 4885
+    },
+    {
+      "epoch": 1.3534626038781163,
+      "grad_norm": 0.6757032871246338,
+      "learning_rate": 4.416307383074408e-06,
+      "loss": 0.5911,
+      "step": 4886
+    },
+    {
+      "epoch": 1.3537396121883656,
+      "grad_norm": 0.6949636340141296,
+      "learning_rate": 4.416073412831717e-06,
+      "loss": 0.6261,
+      "step": 4887
+    },
+    {
+      "epoch": 1.3540166204986148,
+      "grad_norm": 0.7025136947631836,
+      "learning_rate": 4.415839401905894e-06,
+      "loss": 0.6555,
+      "step": 4888
+    },
+    {
+      "epoch": 1.3542936288088643,
+      "grad_norm": 0.6896710991859436,
+      "learning_rate": 4.415605350301908e-06,
+      "loss": 0.6389,
+      "step": 4889
+    },
+    {
+      "epoch": 1.3545706371191135,
+      "grad_norm": 0.6909484267234802,
+      "learning_rate": 4.415371258024727e-06,
+      "loss": 0.6401,
+      "step": 4890
+    },
+    {
+      "epoch": 1.3548476454293628,
+      "grad_norm": 0.7090306282043457,
+      "learning_rate": 4.415137125079325e-06,
+      "loss": 0.6582,
+      "step": 4891
+    },
+    {
+      "epoch": 1.3551246537396122,
+      "grad_norm": 0.685055136680603,
+      "learning_rate": 4.41490295147067e-06,
+      "loss": 0.5861,
+      "step": 4892
+    },
+    {
+      "epoch": 1.3554016620498615,
+      "grad_norm": 0.7134220600128174,
+      "learning_rate": 4.414668737203736e-06,
+      "loss": 0.6362,
+      "step": 4893
+    },
+    {
+      "epoch": 1.3556786703601107,
+      "grad_norm": 0.6816451549530029,
+      "learning_rate": 4.414434482283494e-06,
+      "loss": 0.6156,
+      "step": 4894
+    },
+    {
+      "epoch": 1.3559556786703602,
+      "grad_norm": 0.7040759921073914,
+      "learning_rate": 4.41420018671492e-06,
+      "loss": 0.6265,
+      "step": 4895
+    },
+    {
+      "epoch": 1.3562326869806094,
+      "grad_norm": 0.722031831741333,
+      "learning_rate": 4.413965850502987e-06,
+      "loss": 0.6878,
+      "step": 4896
+    },
+    {
+      "epoch": 1.3565096952908586,
+      "grad_norm": 0.6951577663421631,
+      "learning_rate": 4.413731473652672e-06,
+      "loss": 0.5759,
+      "step": 4897
+    },
+    {
+      "epoch": 1.356786703601108,
+      "grad_norm": 0.6912435293197632,
+      "learning_rate": 4.41349705616895e-06,
+      "loss": 0.6396,
+      "step": 4898
+    },
+    {
+      "epoch": 1.3570637119113573,
+      "grad_norm": 0.7001351714134216,
+      "learning_rate": 4.413262598056799e-06,
+      "loss": 0.6035,
+      "step": 4899
+    },
+    {
+      "epoch": 1.3573407202216066,
+      "grad_norm": 0.7252200245857239,
+      "learning_rate": 4.413028099321198e-06,
+      "loss": 0.6015,
+      "step": 4900
+    },
+    {
+      "epoch": 1.357617728531856,
+      "grad_norm": 0.7011927962303162,
+      "learning_rate": 4.412793559967124e-06,
+      "loss": 0.6121,
+      "step": 4901
+    },
+    {
+      "epoch": 1.3578947368421053,
+      "grad_norm": 0.7054765224456787,
+      "learning_rate": 4.4125589799995585e-06,
+      "loss": 0.6563,
+      "step": 4902
+    },
+    {
+      "epoch": 1.3581717451523545,
+      "grad_norm": 0.6914619207382202,
+      "learning_rate": 4.412324359423482e-06,
+      "loss": 0.6285,
+      "step": 4903
+    },
+    {
+      "epoch": 1.358448753462604,
+      "grad_norm": 0.7494106888771057,
+      "learning_rate": 4.4120896982438745e-06,
+      "loss": 0.6738,
+      "step": 4904
+    },
+    {
+      "epoch": 1.3587257617728532,
+      "grad_norm": 0.6712677478790283,
+      "learning_rate": 4.411854996465721e-06,
+      "loss": 0.6406,
+      "step": 4905
+    },
+    {
+      "epoch": 1.3590027700831024,
+      "grad_norm": 0.7037259936332703,
+      "learning_rate": 4.411620254094003e-06,
+      "loss": 0.6761,
+      "step": 4906
+    },
+    {
+      "epoch": 1.359279778393352,
+      "grad_norm": 0.7152909636497498,
+      "learning_rate": 4.411385471133706e-06,
+      "loss": 0.6637,
+      "step": 4907
+    },
+    {
+      "epoch": 1.3595567867036011,
+      "grad_norm": 0.6827277541160583,
+      "learning_rate": 4.411150647589813e-06,
+      "loss": 0.6316,
+      "step": 4908
+    },
+    {
+      "epoch": 1.3598337950138504,
+      "grad_norm": 0.7183017134666443,
+      "learning_rate": 4.410915783467312e-06,
+      "loss": 0.6172,
+      "step": 4909
+    },
+    {
+      "epoch": 1.3601108033240998,
+      "grad_norm": 0.6997835636138916,
+      "learning_rate": 4.41068087877119e-06,
+      "loss": 0.6005,
+      "step": 4910
+    },
+    {
+      "epoch": 1.360387811634349,
+      "grad_norm": 0.6682953834533691,
+      "learning_rate": 4.410445933506432e-06,
+      "loss": 0.6269,
+      "step": 4911
+    },
+    {
+      "epoch": 1.3606648199445983,
+      "grad_norm": 0.7462157011032104,
+      "learning_rate": 4.4102109476780284e-06,
+      "loss": 0.6336,
+      "step": 4912
+    },
+    {
+      "epoch": 1.3609418282548478,
+      "grad_norm": 0.6756572723388672,
+      "learning_rate": 4.409975921290969e-06,
+      "loss": 0.6441,
+      "step": 4913
+    },
+    {
+      "epoch": 1.361218836565097,
+      "grad_norm": 0.7158071398735046,
+      "learning_rate": 4.409740854350243e-06,
+      "loss": 0.6703,
+      "step": 4914
+    },
+    {
+      "epoch": 1.3614958448753463,
+      "grad_norm": 0.7397752404212952,
+      "learning_rate": 4.409505746860841e-06,
+      "loss": 0.6339,
+      "step": 4915
+    },
+    {
+      "epoch": 1.3617728531855955,
+      "grad_norm": 0.7081032991409302,
+      "learning_rate": 4.409270598827756e-06,
+      "loss": 0.674,
+      "step": 4916
+    },
+    {
+      "epoch": 1.362049861495845,
+      "grad_norm": 0.7220388054847717,
+      "learning_rate": 4.409035410255981e-06,
+      "loss": 0.647,
+      "step": 4917
+    },
+    {
+      "epoch": 1.3623268698060942,
+      "grad_norm": 0.7010590434074402,
+      "learning_rate": 4.408800181150509e-06,
+      "loss": 0.6112,
+      "step": 4918
+    },
+    {
+      "epoch": 1.3626038781163434,
+      "grad_norm": 0.7032333016395569,
+      "learning_rate": 4.408564911516333e-06,
+      "loss": 0.6207,
+      "step": 4919
+    },
+    {
+      "epoch": 1.3628808864265927,
+      "grad_norm": 0.7119109630584717,
+      "learning_rate": 4.408329601358452e-06,
+      "loss": 0.6234,
+      "step": 4920
+    },
+    {
+      "epoch": 1.3631578947368421,
+      "grad_norm": 0.7243717908859253,
+      "learning_rate": 4.408094250681859e-06,
+      "loss": 0.649,
+      "step": 4921
+    },
+    {
+      "epoch": 1.3634349030470914,
+      "grad_norm": 0.6528921127319336,
+      "learning_rate": 4.407858859491552e-06,
+      "loss": 0.5989,
+      "step": 4922
+    },
+    {
+      "epoch": 1.3637119113573406,
+      "grad_norm": 0.693771243095398,
+      "learning_rate": 4.407623427792531e-06,
+      "loss": 0.6507,
+      "step": 4923
+    },
+    {
+      "epoch": 1.36398891966759,
+      "grad_norm": 0.6874661445617676,
+      "learning_rate": 4.407387955589792e-06,
+      "loss": 0.6474,
+      "step": 4924
+    },
+    {
+      "epoch": 1.3642659279778393,
+      "grad_norm": 0.7141904830932617,
+      "learning_rate": 4.407152442888336e-06,
+      "loss": 0.6593,
+      "step": 4925
+    },
+    {
+      "epoch": 1.3645429362880885,
+      "grad_norm": 0.7314815521240234,
+      "learning_rate": 4.406916889693163e-06,
+      "loss": 0.6326,
+      "step": 4926
+    },
+    {
+      "epoch": 1.364819944598338,
+      "grad_norm": 0.6970093846321106,
+      "learning_rate": 4.4066812960092754e-06,
+      "loss": 0.5605,
+      "step": 4927
+    },
+    {
+      "epoch": 1.3650969529085872,
+      "grad_norm": 0.7311210632324219,
+      "learning_rate": 4.406445661841674e-06,
+      "loss": 0.6857,
+      "step": 4928
+    },
+    {
+      "epoch": 1.3653739612188365,
+      "grad_norm": 0.6865612864494324,
+      "learning_rate": 4.406209987195363e-06,
+      "loss": 0.6316,
+      "step": 4929
+    },
+    {
+      "epoch": 1.365650969529086,
+      "grad_norm": 0.7145235538482666,
+      "learning_rate": 4.405974272075347e-06,
+      "loss": 0.6357,
+      "step": 4930
+    },
+    {
+      "epoch": 1.3659279778393352,
+      "grad_norm": 0.6490334272384644,
+      "learning_rate": 4.405738516486628e-06,
+      "loss": 0.625,
+      "step": 4931
+    },
+    {
+      "epoch": 1.3662049861495844,
+      "grad_norm": 0.6927633881568909,
+      "learning_rate": 4.405502720434215e-06,
+      "loss": 0.6267,
+      "step": 4932
+    },
+    {
+      "epoch": 1.3664819944598339,
+      "grad_norm": 0.6824126839637756,
+      "learning_rate": 4.405266883923112e-06,
+      "loss": 0.6357,
+      "step": 4933
+    },
+    {
+      "epoch": 1.366759002770083,
+      "grad_norm": 0.7101442813873291,
+      "learning_rate": 4.405031006958328e-06,
+      "loss": 0.6518,
+      "step": 4934
+    },
+    {
+      "epoch": 1.3670360110803323,
+      "grad_norm": 0.7349445223808289,
+      "learning_rate": 4.40479508954487e-06,
+      "loss": 0.6223,
+      "step": 4935
+    },
+    {
+      "epoch": 1.3673130193905818,
+      "grad_norm": 0.6581106781959534,
+      "learning_rate": 4.404559131687749e-06,
+      "loss": 0.6091,
+      "step": 4936
+    },
+    {
+      "epoch": 1.367590027700831,
+      "grad_norm": 0.692273736000061,
+      "learning_rate": 4.404323133391974e-06,
+      "loss": 0.6242,
+      "step": 4937
+    },
+    {
+      "epoch": 1.3678670360110803,
+      "grad_norm": 0.6735270619392395,
+      "learning_rate": 4.404087094662556e-06,
+      "loss": 0.5825,
+      "step": 4938
+    },
+    {
+      "epoch": 1.3681440443213297,
+      "grad_norm": 0.6877938508987427,
+      "learning_rate": 4.403851015504506e-06,
+      "loss": 0.6088,
+      "step": 4939
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.690207302570343,
+      "learning_rate": 4.4036148959228365e-06,
+      "loss": 0.6462,
+      "step": 4940
+    },
+    {
+      "epoch": 1.3686980609418282,
+      "grad_norm": 0.6809142827987671,
+      "learning_rate": 4.4033787359225626e-06,
+      "loss": 0.649,
+      "step": 4941
+    },
+    {
+      "epoch": 1.3689750692520777,
+      "grad_norm": 0.7366628646850586,
+      "learning_rate": 4.403142535508697e-06,
+      "loss": 0.6731,
+      "step": 4942
+    },
+    {
+      "epoch": 1.369252077562327,
+      "grad_norm": 0.7015873789787292,
+      "learning_rate": 4.402906294686254e-06,
+      "loss": 0.6381,
+      "step": 4943
+    },
+    {
+      "epoch": 1.3695290858725762,
+      "grad_norm": 0.6807177662849426,
+      "learning_rate": 4.402670013460252e-06,
+      "loss": 0.6652,
+      "step": 4944
+    },
+    {
+      "epoch": 1.3698060941828256,
+      "grad_norm": 0.6987523436546326,
+      "learning_rate": 4.402433691835707e-06,
+      "loss": 0.614,
+      "step": 4945
+    },
+    {
+      "epoch": 1.3700831024930749,
+      "grad_norm": 0.7272536754608154,
+      "learning_rate": 4.402197329817636e-06,
+      "loss": 0.5836,
+      "step": 4946
+    },
+    {
+      "epoch": 1.370360110803324,
+      "grad_norm": 0.7454519867897034,
+      "learning_rate": 4.401960927411057e-06,
+      "loss": 0.6685,
+      "step": 4947
+    },
+    {
+      "epoch": 1.3706371191135733,
+      "grad_norm": 0.6987824440002441,
+      "learning_rate": 4.401724484620993e-06,
+      "loss": 0.6276,
+      "step": 4948
+    },
+    {
+      "epoch": 1.3709141274238228,
+      "grad_norm": 0.7243875861167908,
+      "learning_rate": 4.40148800145246e-06,
+      "loss": 0.6288,
+      "step": 4949
+    },
+    {
+      "epoch": 1.371191135734072,
+      "grad_norm": 0.6841127872467041,
+      "learning_rate": 4.4012514779104805e-06,
+      "loss": 0.6113,
+      "step": 4950
+    },
+    {
+      "epoch": 1.3714681440443213,
+      "grad_norm": 0.663722574710846,
+      "learning_rate": 4.401014914000078e-06,
+      "loss": 0.6448,
+      "step": 4951
+    },
+    {
+      "epoch": 1.3717451523545705,
+      "grad_norm": 0.7879506349563599,
+      "learning_rate": 4.400778309726273e-06,
+      "loss": 0.6207,
+      "step": 4952
+    },
+    {
+      "epoch": 1.37202216066482,
+      "grad_norm": 0.7135480046272278,
+      "learning_rate": 4.400541665094091e-06,
+      "loss": 0.6194,
+      "step": 4953
+    },
+    {
+      "epoch": 1.3722991689750692,
+      "grad_norm": 0.6983644366264343,
+      "learning_rate": 4.4003049801085565e-06,
+      "loss": 0.6689,
+      "step": 4954
+    },
+    {
+      "epoch": 1.3725761772853184,
+      "grad_norm": 0.6594923734664917,
+      "learning_rate": 4.4000682547746945e-06,
+      "loss": 0.6202,
+      "step": 4955
+    },
+    {
+      "epoch": 1.372853185595568,
+      "grad_norm": 0.6694162487983704,
+      "learning_rate": 4.39983148909753e-06,
+      "loss": 0.6257,
+      "step": 4956
+    },
+    {
+      "epoch": 1.3731301939058171,
+      "grad_norm": 0.7033407688140869,
+      "learning_rate": 4.399594683082092e-06,
+      "loss": 0.626,
+      "step": 4957
+    },
+    {
+      "epoch": 1.3734072022160664,
+      "grad_norm": 0.6754797101020813,
+      "learning_rate": 4.399357836733408e-06,
+      "loss": 0.6152,
+      "step": 4958
+    },
+    {
+      "epoch": 1.3736842105263158,
+      "grad_norm": 0.6850981712341309,
+      "learning_rate": 4.399120950056507e-06,
+      "loss": 0.6074,
+      "step": 4959
+    },
+    {
+      "epoch": 1.373961218836565,
+      "grad_norm": 0.6460973024368286,
+      "learning_rate": 4.398884023056418e-06,
+      "loss": 0.6314,
+      "step": 4960
+    },
+    {
+      "epoch": 1.3742382271468143,
+      "grad_norm": 0.7158933281898499,
+      "learning_rate": 4.398647055738172e-06,
+      "loss": 0.6515,
+      "step": 4961
+    },
+    {
+      "epoch": 1.3745152354570638,
+      "grad_norm": 0.7266000509262085,
+      "learning_rate": 4.3984100481068e-06,
+      "loss": 0.639,
+      "step": 4962
+    },
+    {
+      "epoch": 1.374792243767313,
+      "grad_norm": 0.7677624225616455,
+      "learning_rate": 4.398173000167335e-06,
+      "loss": 0.6462,
+      "step": 4963
+    },
+    {
+      "epoch": 1.3750692520775623,
+      "grad_norm": 0.6976761817932129,
+      "learning_rate": 4.39793591192481e-06,
+      "loss": 0.6105,
+      "step": 4964
+    },
+    {
+      "epoch": 1.3753462603878117,
+      "grad_norm": 0.6852526664733887,
+      "learning_rate": 4.397698783384259e-06,
+      "loss": 0.6208,
+      "step": 4965
+    },
+    {
+      "epoch": 1.375623268698061,
+      "grad_norm": 0.7142199277877808,
+      "learning_rate": 4.397461614550716e-06,
+      "loss": 0.6435,
+      "step": 4966
+    },
+    {
+      "epoch": 1.3759002770083102,
+      "grad_norm": 0.6593324542045593,
+      "learning_rate": 4.397224405429218e-06,
+      "loss": 0.6297,
+      "step": 4967
+    },
+    {
+      "epoch": 1.3761772853185597,
+      "grad_norm": 0.6773996353149414,
+      "learning_rate": 4.3969871560248e-06,
+      "loss": 0.6538,
+      "step": 4968
+    },
+    {
+      "epoch": 1.376454293628809,
+      "grad_norm": 0.7089106440544128,
+      "learning_rate": 4.396749866342501e-06,
+      "loss": 0.6292,
+      "step": 4969
+    },
+    {
+      "epoch": 1.3767313019390581,
+      "grad_norm": 0.6807532906532288,
+      "learning_rate": 4.396512536387358e-06,
+      "loss": 0.632,
+      "step": 4970
+    },
+    {
+      "epoch": 1.3770083102493076,
+      "grad_norm": 0.6472201347351074,
+      "learning_rate": 4.396275166164411e-06,
+      "loss": 0.627,
+      "step": 4971
+    },
+    {
+      "epoch": 1.3772853185595568,
+      "grad_norm": 0.675743043422699,
+      "learning_rate": 4.3960377556787e-06,
+      "loss": 0.6061,
+      "step": 4972
+    },
+    {
+      "epoch": 1.377562326869806,
+      "grad_norm": 0.6863962411880493,
+      "learning_rate": 4.395800304935265e-06,
+      "loss": 0.6055,
+      "step": 4973
+    },
+    {
+      "epoch": 1.3778393351800555,
+      "grad_norm": 0.6849420666694641,
+      "learning_rate": 4.395562813939148e-06,
+      "loss": 0.6504,
+      "step": 4974
+    },
+    {
+      "epoch": 1.3781163434903048,
+      "grad_norm": 0.6862570643424988,
+      "learning_rate": 4.395325282695393e-06,
+      "loss": 0.5924,
+      "step": 4975
+    },
+    {
+      "epoch": 1.378393351800554,
+      "grad_norm": 0.7083377838134766,
+      "learning_rate": 4.395087711209041e-06,
+      "loss": 0.61,
+      "step": 4976
+    },
+    {
+      "epoch": 1.3786703601108035,
+      "grad_norm": 0.6750888228416443,
+      "learning_rate": 4.394850099485138e-06,
+      "loss": 0.6368,
+      "step": 4977
+    },
+    {
+      "epoch": 1.3789473684210527,
+      "grad_norm": 0.6690243482589722,
+      "learning_rate": 4.3946124475287275e-06,
+      "loss": 0.62,
+      "step": 4978
+    },
+    {
+      "epoch": 1.379224376731302,
+      "grad_norm": 0.6947623491287231,
+      "learning_rate": 4.3943747553448566e-06,
+      "loss": 0.6352,
+      "step": 4979
+    },
+    {
+      "epoch": 1.3795013850415512,
+      "grad_norm": 0.7098008394241333,
+      "learning_rate": 4.3941370229385725e-06,
+      "loss": 0.6206,
+      "step": 4980
+    },
+    {
+      "epoch": 1.3797783933518006,
+      "grad_norm": 0.7828469276428223,
+      "learning_rate": 4.393899250314923e-06,
+      "loss": 0.6318,
+      "step": 4981
+    },
+    {
+      "epoch": 1.3800554016620499,
+      "grad_norm": 0.6869782209396362,
+      "learning_rate": 4.3936614374789545e-06,
+      "loss": 0.651,
+      "step": 4982
+    },
+    {
+      "epoch": 1.3803324099722991,
+      "grad_norm": 0.6738064885139465,
+      "learning_rate": 4.393423584435718e-06,
+      "loss": 0.6511,
+      "step": 4983
+    },
+    {
+      "epoch": 1.3806094182825484,
+      "grad_norm": 0.7082775831222534,
+      "learning_rate": 4.3931856911902635e-06,
+      "loss": 0.6642,
+      "step": 4984
+    },
+    {
+      "epoch": 1.3808864265927978,
+      "grad_norm": 0.7249225378036499,
+      "learning_rate": 4.392947757747643e-06,
+      "loss": 0.618,
+      "step": 4985
+    },
+    {
+      "epoch": 1.381163434903047,
+      "grad_norm": 0.6885894536972046,
+      "learning_rate": 4.392709784112907e-06,
+      "loss": 0.6463,
+      "step": 4986
+    },
+    {
+      "epoch": 1.3814404432132963,
+      "grad_norm": 0.7157624959945679,
+      "learning_rate": 4.392471770291109e-06,
+      "loss": 0.6746,
+      "step": 4987
+    },
+    {
+      "epoch": 1.3817174515235457,
+      "grad_norm": 0.7281468510627747,
+      "learning_rate": 4.392233716287302e-06,
+      "loss": 0.6293,
+      "step": 4988
+    },
+    {
+      "epoch": 1.381994459833795,
+      "grad_norm": 0.7012090086936951,
+      "learning_rate": 4.3919956221065415e-06,
+      "loss": 0.653,
+      "step": 4989
+    },
+    {
+      "epoch": 1.3822714681440442,
+      "grad_norm": 0.6941935420036316,
+      "learning_rate": 4.391757487753882e-06,
+      "loss": 0.6138,
+      "step": 4990
+    },
+    {
+      "epoch": 1.3825484764542937,
+      "grad_norm": 0.703411340713501,
+      "learning_rate": 4.39151931323438e-06,
+      "loss": 0.6536,
+      "step": 4991
+    },
+    {
+      "epoch": 1.382825484764543,
+      "grad_norm": 0.6732279658317566,
+      "learning_rate": 4.391281098553093e-06,
+      "loss": 0.6215,
+      "step": 4992
+    },
+    {
+      "epoch": 1.3831024930747922,
+      "grad_norm": 0.7355413436889648,
+      "learning_rate": 4.3910428437150776e-06,
+      "loss": 0.6319,
+      "step": 4993
+    },
+    {
+      "epoch": 1.3833795013850416,
+      "grad_norm": 0.6964555978775024,
+      "learning_rate": 4.390804548725394e-06,
+      "loss": 0.6176,
+      "step": 4994
+    },
+    {
+      "epoch": 1.3836565096952909,
+      "grad_norm": 0.7024789452552795,
+      "learning_rate": 4.390566213589101e-06,
+      "loss": 0.6563,
+      "step": 4995
+    },
+    {
+      "epoch": 1.38393351800554,
+      "grad_norm": 0.6982277035713196,
+      "learning_rate": 4.3903278383112606e-06,
+      "loss": 0.6427,
+      "step": 4996
+    },
+    {
+      "epoch": 1.3842105263157896,
+      "grad_norm": 0.7301534414291382,
+      "learning_rate": 4.390089422896931e-06,
+      "loss": 0.6351,
+      "step": 4997
+    },
+    {
+      "epoch": 1.3844875346260388,
+      "grad_norm": 0.7176867127418518,
+      "learning_rate": 4.389850967351177e-06,
+      "loss": 0.6316,
+      "step": 4998
+    },
+    {
+      "epoch": 1.384764542936288,
+      "grad_norm": 0.729347825050354,
+      "learning_rate": 4.3896124716790615e-06,
+      "loss": 0.6598,
+      "step": 4999
+    },
+    {
+      "epoch": 1.3850415512465375,
+      "grad_norm": 0.6763535737991333,
+      "learning_rate": 4.3893739358856465e-06,
+      "loss": 0.658,
+      "step": 5000
+    },
+    {
+      "epoch": 1.3853185595567867,
+      "grad_norm": 0.6909424066543579,
+      "learning_rate": 4.389135359975998e-06,
+      "loss": 0.5952,
+      "step": 5001
+    },
+    {
+      "epoch": 1.385595567867036,
+      "grad_norm": 0.6764463186264038,
+      "learning_rate": 4.388896743955182e-06,
+      "loss": 0.6299,
+      "step": 5002
+    },
+    {
+      "epoch": 1.3858725761772854,
+      "grad_norm": 0.7030847668647766,
+      "learning_rate": 4.388658087828264e-06,
+      "loss": 0.6338,
+      "step": 5003
+    },
+    {
+      "epoch": 1.3861495844875347,
+      "grad_norm": 0.670483410358429,
+      "learning_rate": 4.388419391600312e-06,
+      "loss": 0.6809,
+      "step": 5004
+    },
+    {
+      "epoch": 1.386426592797784,
+      "grad_norm": 0.6741538643836975,
+      "learning_rate": 4.388180655276394e-06,
+      "loss": 0.5736,
+      "step": 5005
+    },
+    {
+      "epoch": 1.3867036011080334,
+      "grad_norm": 0.7062891125679016,
+      "learning_rate": 4.387941878861578e-06,
+      "loss": 0.6475,
+      "step": 5006
+    },
+    {
+      "epoch": 1.3869806094182826,
+      "grad_norm": 0.7449402213096619,
+      "learning_rate": 4.387703062360935e-06,
+      "loss": 0.6448,
+      "step": 5007
+    },
+    {
+      "epoch": 1.3872576177285318,
+      "grad_norm": 0.714369535446167,
+      "learning_rate": 4.387464205779536e-06,
+      "loss": 0.6513,
+      "step": 5008
+    },
+    {
+      "epoch": 1.387534626038781,
+      "grad_norm": 0.6852971911430359,
+      "learning_rate": 4.387225309122451e-06,
+      "loss": 0.61,
+      "step": 5009
+    },
+    {
+      "epoch": 1.3878116343490305,
+      "grad_norm": 0.7403523325920105,
+      "learning_rate": 4.3869863723947536e-06,
+      "loss": 0.6809,
+      "step": 5010
+    },
+    {
+      "epoch": 1.3880886426592798,
+      "grad_norm": 0.6950234770774841,
+      "learning_rate": 4.386747395601517e-06,
+      "loss": 0.6178,
+      "step": 5011
+    },
+    {
+      "epoch": 1.388365650969529,
+      "grad_norm": 0.7244662642478943,
+      "learning_rate": 4.386508378747815e-06,
+      "loss": 0.687,
+      "step": 5012
+    },
+    {
+      "epoch": 1.3886426592797783,
+      "grad_norm": 0.7085997462272644,
+      "learning_rate": 4.386269321838722e-06,
+      "loss": 0.5976,
+      "step": 5013
+    },
+    {
+      "epoch": 1.3889196675900277,
+      "grad_norm": 0.7164833545684814,
+      "learning_rate": 4.386030224879314e-06,
+      "loss": 0.6344,
+      "step": 5014
+    },
+    {
+      "epoch": 1.389196675900277,
+      "grad_norm": 0.6968429088592529,
+      "learning_rate": 4.385791087874668e-06,
+      "loss": 0.6052,
+      "step": 5015
+    },
+    {
+      "epoch": 1.3894736842105262,
+      "grad_norm": 0.6692972779273987,
+      "learning_rate": 4.385551910829861e-06,
+      "loss": 0.5957,
+      "step": 5016
+    },
+    {
+      "epoch": 1.3897506925207757,
+      "grad_norm": 0.6999967694282532,
+      "learning_rate": 4.3853126937499735e-06,
+      "loss": 0.6415,
+      "step": 5017
+    },
+    {
+      "epoch": 1.390027700831025,
+      "grad_norm": 0.6935191750526428,
+      "learning_rate": 4.385073436640081e-06,
+      "loss": 0.6374,
+      "step": 5018
+    },
+    {
+      "epoch": 1.3903047091412741,
+      "grad_norm": 0.7233996391296387,
+      "learning_rate": 4.384834139505267e-06,
+      "loss": 0.6473,
+      "step": 5019
+    },
+    {
+      "epoch": 1.3905817174515236,
+      "grad_norm": 0.7004993557929993,
+      "learning_rate": 4.3845948023506106e-06,
+      "loss": 0.6411,
+      "step": 5020
+    },
+    {
+      "epoch": 1.3908587257617728,
+      "grad_norm": 0.7594379186630249,
+      "learning_rate": 4.384355425181193e-06,
+      "loss": 0.6059,
+      "step": 5021
+    },
+    {
+      "epoch": 1.391135734072022,
+      "grad_norm": 0.7419117093086243,
+      "learning_rate": 4.384116008002099e-06,
+      "loss": 0.671,
+      "step": 5022
+    },
+    {
+      "epoch": 1.3914127423822715,
+      "grad_norm": 0.6777958869934082,
+      "learning_rate": 4.3838765508184095e-06,
+      "loss": 0.6296,
+      "step": 5023
+    },
+    {
+      "epoch": 1.3916897506925208,
+      "grad_norm": 0.6925370693206787,
+      "learning_rate": 4.38363705363521e-06,
+      "loss": 0.6727,
+      "step": 5024
+    },
+    {
+      "epoch": 1.39196675900277,
+      "grad_norm": 0.6831212639808655,
+      "learning_rate": 4.383397516457586e-06,
+      "loss": 0.648,
+      "step": 5025
+    },
+    {
+      "epoch": 1.3922437673130195,
+      "grad_norm": 0.6936487555503845,
+      "learning_rate": 4.383157939290624e-06,
+      "loss": 0.6504,
+      "step": 5026
+    },
+    {
+      "epoch": 1.3925207756232687,
+      "grad_norm": 0.6868559718132019,
+      "learning_rate": 4.382918322139407e-06,
+      "loss": 0.5858,
+      "step": 5027
+    },
+    {
+      "epoch": 1.392797783933518,
+      "grad_norm": 0.7001911401748657,
+      "learning_rate": 4.382678665009028e-06,
+      "loss": 0.6252,
+      "step": 5028
+    },
+    {
+      "epoch": 1.3930747922437674,
+      "grad_norm": 0.6877859830856323,
+      "learning_rate": 4.382438967904572e-06,
+      "loss": 0.6169,
+      "step": 5029
+    },
+    {
+      "epoch": 1.3933518005540166,
+      "grad_norm": 0.7069692015647888,
+      "learning_rate": 4.38219923083113e-06,
+      "loss": 0.6361,
+      "step": 5030
+    },
+    {
+      "epoch": 1.3936288088642659,
+      "grad_norm": 0.7657400369644165,
+      "learning_rate": 4.381959453793792e-06,
+      "loss": 0.6686,
+      "step": 5031
+    },
+    {
+      "epoch": 1.3939058171745153,
+      "grad_norm": 0.7232145667076111,
+      "learning_rate": 4.381719636797648e-06,
+      "loss": 0.6654,
+      "step": 5032
+    },
+    {
+      "epoch": 1.3941828254847646,
+      "grad_norm": 0.676956057548523,
+      "learning_rate": 4.381479779847791e-06,
+      "loss": 0.6518,
+      "step": 5033
+    },
+    {
+      "epoch": 1.3944598337950138,
+      "grad_norm": 0.6840213537216187,
+      "learning_rate": 4.381239882949314e-06,
+      "loss": 0.596,
+      "step": 5034
+    },
+    {
+      "epoch": 1.3947368421052633,
+      "grad_norm": 0.6917917728424072,
+      "learning_rate": 4.380999946107309e-06,
+      "loss": 0.6598,
+      "step": 5035
+    },
+    {
+      "epoch": 1.3950138504155125,
+      "grad_norm": 0.6727468967437744,
+      "learning_rate": 4.380759969326873e-06,
+      "loss": 0.666,
+      "step": 5036
+    },
+    {
+      "epoch": 1.3952908587257618,
+      "grad_norm": 0.6889804601669312,
+      "learning_rate": 4.3805199526131e-06,
+      "loss": 0.687,
+      "step": 5037
+    },
+    {
+      "epoch": 1.3955678670360112,
+      "grad_norm": 0.6997852325439453,
+      "learning_rate": 4.380279895971085e-06,
+      "loss": 0.6766,
+      "step": 5038
+    },
+    {
+      "epoch": 1.3958448753462604,
+      "grad_norm": 0.760193407535553,
+      "learning_rate": 4.380039799405926e-06,
+      "loss": 0.6834,
+      "step": 5039
+    },
+    {
+      "epoch": 1.3961218836565097,
+      "grad_norm": 0.7114586234092712,
+      "learning_rate": 4.379799662922721e-06,
+      "loss": 0.6307,
+      "step": 5040
+    },
+    {
+      "epoch": 1.396398891966759,
+      "grad_norm": 0.6752461194992065,
+      "learning_rate": 4.379559486526569e-06,
+      "loss": 0.6107,
+      "step": 5041
+    },
+    {
+      "epoch": 1.3966759002770084,
+      "grad_norm": 0.7001312375068665,
+      "learning_rate": 4.379319270222569e-06,
+      "loss": 0.6305,
+      "step": 5042
+    },
+    {
+      "epoch": 1.3969529085872576,
+      "grad_norm": 0.7060052156448364,
+      "learning_rate": 4.379079014015822e-06,
+      "loss": 0.626,
+      "step": 5043
+    },
+    {
+      "epoch": 1.3972299168975069,
+      "grad_norm": 0.684003472328186,
+      "learning_rate": 4.378838717911429e-06,
+      "loss": 0.6246,
+      "step": 5044
+    },
+    {
+      "epoch": 1.397506925207756,
+      "grad_norm": 0.6880972385406494,
+      "learning_rate": 4.378598381914491e-06,
+      "loss": 0.628,
+      "step": 5045
+    },
+    {
+      "epoch": 1.3977839335180056,
+      "grad_norm": 0.7145376801490784,
+      "learning_rate": 4.3783580060301135e-06,
+      "loss": 0.6342,
+      "step": 5046
+    },
+    {
+      "epoch": 1.3980609418282548,
+      "grad_norm": 0.7057557106018066,
+      "learning_rate": 4.378117590263398e-06,
+      "loss": 0.6696,
+      "step": 5047
+    },
+    {
+      "epoch": 1.398337950138504,
+      "grad_norm": 0.6906470656394958,
+      "learning_rate": 4.37787713461945e-06,
+      "loss": 0.5806,
+      "step": 5048
+    },
+    {
+      "epoch": 1.3986149584487535,
+      "grad_norm": 0.669457733631134,
+      "learning_rate": 4.377636639103375e-06,
+      "loss": 0.5847,
+      "step": 5049
+    },
+    {
+      "epoch": 1.3988919667590027,
+      "grad_norm": 0.6445726752281189,
+      "learning_rate": 4.3773961037202784e-06,
+      "loss": 0.5876,
+      "step": 5050
+    },
+    {
+      "epoch": 1.399168975069252,
+      "grad_norm": 0.7106205224990845,
+      "learning_rate": 4.377155528475268e-06,
+      "loss": 0.6559,
+      "step": 5051
+    },
+    {
+      "epoch": 1.3994459833795014,
+      "grad_norm": 0.6805086135864258,
+      "learning_rate": 4.376914913373453e-06,
+      "loss": 0.6066,
+      "step": 5052
+    },
+    {
+      "epoch": 1.3997229916897507,
+      "grad_norm": 0.8467181921005249,
+      "learning_rate": 4.3766742584199415e-06,
+      "loss": 0.623,
+      "step": 5053
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.687349796295166,
+      "learning_rate": 4.376433563619843e-06,
+      "loss": 0.6407,
+      "step": 5054
+    },
+    {
+      "epoch": 1.4002770083102494,
+      "grad_norm": 0.7355000376701355,
+      "learning_rate": 4.376192828978266e-06,
+      "loss": 0.6078,
+      "step": 5055
+    },
+    {
+      "epoch": 1.4005540166204986,
+      "grad_norm": 0.6767886877059937,
+      "learning_rate": 4.375952054500326e-06,
+      "loss": 0.6645,
+      "step": 5056
+    },
+    {
+      "epoch": 1.4008310249307478,
+      "grad_norm": 0.7190877199172974,
+      "learning_rate": 4.3757112401911326e-06,
+      "loss": 0.6333,
+      "step": 5057
+    },
+    {
+      "epoch": 1.4011080332409973,
+      "grad_norm": 0.702785849571228,
+      "learning_rate": 4.3754703860558e-06,
+      "loss": 0.6458,
+      "step": 5058
+    },
+    {
+      "epoch": 1.4013850415512465,
+      "grad_norm": 0.6814035773277283,
+      "learning_rate": 4.375229492099441e-06,
+      "loss": 0.6532,
+      "step": 5059
+    },
+    {
+      "epoch": 1.4016620498614958,
+      "grad_norm": 0.6846272349357605,
+      "learning_rate": 4.374988558327171e-06,
+      "loss": 0.6325,
+      "step": 5060
+    },
+    {
+      "epoch": 1.4019390581717452,
+      "grad_norm": 0.6913299560546875,
+      "learning_rate": 4.374747584744106e-06,
+      "loss": 0.613,
+      "step": 5061
+    },
+    {
+      "epoch": 1.4022160664819945,
+      "grad_norm": 0.6918284893035889,
+      "learning_rate": 4.374506571355362e-06,
+      "loss": 0.6599,
+      "step": 5062
+    },
+    {
+      "epoch": 1.4024930747922437,
+      "grad_norm": 0.7418447136878967,
+      "learning_rate": 4.374265518166057e-06,
+      "loss": 0.6344,
+      "step": 5063
+    },
+    {
+      "epoch": 1.4027700831024932,
+      "grad_norm": 0.6935515403747559,
+      "learning_rate": 4.374024425181308e-06,
+      "loss": 0.6421,
+      "step": 5064
+    },
+    {
+      "epoch": 1.4030470914127424,
+      "grad_norm": 0.6994535326957703,
+      "learning_rate": 4.3737832924062365e-06,
+      "loss": 0.655,
+      "step": 5065
+    },
+    {
+      "epoch": 1.4033240997229917,
+      "grad_norm": 0.6631689667701721,
+      "learning_rate": 4.373542119845959e-06,
+      "loss": 0.5511,
+      "step": 5066
+    },
+    {
+      "epoch": 1.4036011080332411,
+      "grad_norm": 0.6856090426445007,
+      "learning_rate": 4.3733009075055975e-06,
+      "loss": 0.6661,
+      "step": 5067
+    },
+    {
+      "epoch": 1.4038781163434904,
+      "grad_norm": 0.693882405757904,
+      "learning_rate": 4.373059655390274e-06,
+      "loss": 0.625,
+      "step": 5068
+    },
+    {
+      "epoch": 1.4041551246537396,
+      "grad_norm": 0.6695311069488525,
+      "learning_rate": 4.372818363505112e-06,
+      "loss": 0.5803,
+      "step": 5069
+    },
+    {
+      "epoch": 1.404432132963989,
+      "grad_norm": 0.6360461115837097,
+      "learning_rate": 4.372577031855232e-06,
+      "loss": 0.6051,
+      "step": 5070
+    },
+    {
+      "epoch": 1.4047091412742383,
+      "grad_norm": 0.6905994415283203,
+      "learning_rate": 4.37233566044576e-06,
+      "loss": 0.6168,
+      "step": 5071
+    },
+    {
+      "epoch": 1.4049861495844875,
+      "grad_norm": 0.7065186500549316,
+      "learning_rate": 4.37209424928182e-06,
+      "loss": 0.6804,
+      "step": 5072
+    },
+    {
+      "epoch": 1.4052631578947368,
+      "grad_norm": 0.7095279097557068,
+      "learning_rate": 4.37185279836854e-06,
+      "loss": 0.6599,
+      "step": 5073
+    },
+    {
+      "epoch": 1.4055401662049862,
+      "grad_norm": 0.6844799518585205,
+      "learning_rate": 4.371611307711044e-06,
+      "loss": 0.642,
+      "step": 5074
+    },
+    {
+      "epoch": 1.4058171745152355,
+      "grad_norm": 0.6771243810653687,
+      "learning_rate": 4.37136977731446e-06,
+      "loss": 0.6468,
+      "step": 5075
+    },
+    {
+      "epoch": 1.4060941828254847,
+      "grad_norm": 0.6787841320037842,
+      "learning_rate": 4.371128207183917e-06,
+      "loss": 0.6116,
+      "step": 5076
+    },
+    {
+      "epoch": 1.406371191135734,
+      "grad_norm": 0.6891928315162659,
+      "learning_rate": 4.3708865973245436e-06,
+      "loss": 0.6625,
+      "step": 5077
+    },
+    {
+      "epoch": 1.4066481994459834,
+      "grad_norm": 0.6616782546043396,
+      "learning_rate": 4.3706449477414704e-06,
+      "loss": 0.5704,
+      "step": 5078
+    },
+    {
+      "epoch": 1.4069252077562326,
+      "grad_norm": 0.686374306678772,
+      "learning_rate": 4.370403258439827e-06,
+      "loss": 0.6485,
+      "step": 5079
+    },
+    {
+      "epoch": 1.4072022160664819,
+      "grad_norm": 0.7003771662712097,
+      "learning_rate": 4.370161529424747e-06,
+      "loss": 0.6582,
+      "step": 5080
+    },
+    {
+      "epoch": 1.4074792243767313,
+      "grad_norm": 0.6979756355285645,
+      "learning_rate": 4.369919760701361e-06,
+      "loss": 0.5902,
+      "step": 5081
+    },
+    {
+      "epoch": 1.4077562326869806,
+      "grad_norm": 0.7365483641624451,
+      "learning_rate": 4.369677952274803e-06,
+      "loss": 0.637,
+      "step": 5082
+    },
+    {
+      "epoch": 1.4080332409972298,
+      "grad_norm": 0.650162935256958,
+      "learning_rate": 4.369436104150208e-06,
+      "loss": 0.6277,
+      "step": 5083
+    },
+    {
+      "epoch": 1.4083102493074793,
+      "grad_norm": 0.7429336905479431,
+      "learning_rate": 4.369194216332712e-06,
+      "loss": 0.6687,
+      "step": 5084
+    },
+    {
+      "epoch": 1.4085872576177285,
+      "grad_norm": 0.7430762052536011,
+      "learning_rate": 4.368952288827447e-06,
+      "loss": 0.6399,
+      "step": 5085
+    },
+    {
+      "epoch": 1.4088642659279778,
+      "grad_norm": 0.680882453918457,
+      "learning_rate": 4.368710321639553e-06,
+      "loss": 0.6187,
+      "step": 5086
+    },
+    {
+      "epoch": 1.4091412742382272,
+      "grad_norm": 0.714248538017273,
+      "learning_rate": 4.368468314774167e-06,
+      "loss": 0.6387,
+      "step": 5087
+    },
+    {
+      "epoch": 1.4094182825484765,
+      "grad_norm": 0.6917892694473267,
+      "learning_rate": 4.368226268236427e-06,
+      "loss": 0.6168,
+      "step": 5088
+    },
+    {
+      "epoch": 1.4096952908587257,
+      "grad_norm": 0.6685717105865479,
+      "learning_rate": 4.367984182031472e-06,
+      "loss": 0.6311,
+      "step": 5089
+    },
+    {
+      "epoch": 1.4099722991689752,
+      "grad_norm": 0.7154122591018677,
+      "learning_rate": 4.367742056164443e-06,
+      "loss": 0.6407,
+      "step": 5090
+    },
+    {
+      "epoch": 1.4102493074792244,
+      "grad_norm": 0.73028165102005,
+      "learning_rate": 4.36749989064048e-06,
+      "loss": 0.6499,
+      "step": 5091
+    },
+    {
+      "epoch": 1.4105263157894736,
+      "grad_norm": 0.7013839483261108,
+      "learning_rate": 4.367257685464726e-06,
+      "loss": 0.658,
+      "step": 5092
+    },
+    {
+      "epoch": 1.410803324099723,
+      "grad_norm": 0.6611378788948059,
+      "learning_rate": 4.367015440642321e-06,
+      "loss": 0.6839,
+      "step": 5093
+    },
+    {
+      "epoch": 1.4110803324099723,
+      "grad_norm": 0.6703953146934509,
+      "learning_rate": 4.366773156178413e-06,
+      "loss": 0.5726,
+      "step": 5094
+    },
+    {
+      "epoch": 1.4113573407202216,
+      "grad_norm": 0.7304432392120361,
+      "learning_rate": 4.366530832078142e-06,
+      "loss": 0.6775,
+      "step": 5095
+    },
+    {
+      "epoch": 1.411634349030471,
+      "grad_norm": 0.6808119416236877,
+      "learning_rate": 4.366288468346655e-06,
+      "loss": 0.6244,
+      "step": 5096
+    },
+    {
+      "epoch": 1.4119113573407203,
+      "grad_norm": 0.7225993871688843,
+      "learning_rate": 4.366046064989099e-06,
+      "loss": 0.6391,
+      "step": 5097
+    },
+    {
+      "epoch": 1.4121883656509695,
+      "grad_norm": 0.7567205429077148,
+      "learning_rate": 4.365803622010618e-06,
+      "loss": 0.602,
+      "step": 5098
+    },
+    {
+      "epoch": 1.412465373961219,
+      "grad_norm": 0.7190272212028503,
+      "learning_rate": 4.365561139416362e-06,
+      "loss": 0.6829,
+      "step": 5099
+    },
+    {
+      "epoch": 1.4127423822714682,
+      "grad_norm": 0.7132216691970825,
+      "learning_rate": 4.365318617211479e-06,
+      "loss": 0.6793,
+      "step": 5100
+    },
+    {
+      "epoch": 1.4130193905817174,
+      "grad_norm": 0.7090128660202026,
+      "learning_rate": 4.365076055401118e-06,
+      "loss": 0.6117,
+      "step": 5101
+    },
+    {
+      "epoch": 1.4132963988919667,
+      "grad_norm": 0.658957302570343,
+      "learning_rate": 4.36483345399043e-06,
+      "loss": 0.6093,
+      "step": 5102
+    },
+    {
+      "epoch": 1.4135734072022161,
+      "grad_norm": 0.7175829410552979,
+      "learning_rate": 4.364590812984565e-06,
+      "loss": 0.6222,
+      "step": 5103
+    },
+    {
+      "epoch": 1.4138504155124654,
+      "grad_norm": 0.7089779376983643,
+      "learning_rate": 4.364348132388676e-06,
+      "loss": 0.6569,
+      "step": 5104
+    },
+    {
+      "epoch": 1.4141274238227146,
+      "grad_norm": 0.6942974925041199,
+      "learning_rate": 4.364105412207914e-06,
+      "loss": 0.6322,
+      "step": 5105
+    },
+    {
+      "epoch": 1.4144044321329639,
+      "grad_norm": 0.6663142442703247,
+      "learning_rate": 4.3638626524474345e-06,
+      "loss": 0.6599,
+      "step": 5106
+    },
+    {
+      "epoch": 1.4146814404432133,
+      "grad_norm": 0.6893264651298523,
+      "learning_rate": 4.363619853112391e-06,
+      "loss": 0.5942,
+      "step": 5107
+    },
+    {
+      "epoch": 1.4149584487534625,
+      "grad_norm": 0.7206471562385559,
+      "learning_rate": 4.363377014207939e-06,
+      "loss": 0.6543,
+      "step": 5108
+    },
+    {
+      "epoch": 1.4152354570637118,
+      "grad_norm": 0.7506195306777954,
+      "learning_rate": 4.363134135739234e-06,
+      "loss": 0.6474,
+      "step": 5109
+    },
+    {
+      "epoch": 1.4155124653739612,
+      "grad_norm": 0.6773307919502258,
+      "learning_rate": 4.362891217711434e-06,
+      "loss": 0.586,
+      "step": 5110
+    },
+    {
+      "epoch": 1.4157894736842105,
+      "grad_norm": 0.6862797141075134,
+      "learning_rate": 4.362648260129696e-06,
+      "loss": 0.633,
+      "step": 5111
+    },
+    {
+      "epoch": 1.4160664819944597,
+      "grad_norm": 0.6824764609336853,
+      "learning_rate": 4.3624052629991785e-06,
+      "loss": 0.6649,
+      "step": 5112
+    },
+    {
+      "epoch": 1.4163434903047092,
+      "grad_norm": 0.6997144222259521,
+      "learning_rate": 4.362162226325042e-06,
+      "loss": 0.6156,
+      "step": 5113
+    },
+    {
+      "epoch": 1.4166204986149584,
+      "grad_norm": 0.7159362435340881,
+      "learning_rate": 4.361919150112444e-06,
+      "loss": 0.6039,
+      "step": 5114
+    },
+    {
+      "epoch": 1.4168975069252077,
+      "grad_norm": 0.6508421897888184,
+      "learning_rate": 4.3616760343665495e-06,
+      "loss": 0.5546,
+      "step": 5115
+    },
+    {
+      "epoch": 1.4171745152354571,
+      "grad_norm": 0.7000342011451721,
+      "learning_rate": 4.361432879092518e-06,
+      "loss": 0.6295,
+      "step": 5116
+    },
+    {
+      "epoch": 1.4174515235457064,
+      "grad_norm": 0.6941149830818176,
+      "learning_rate": 4.361189684295513e-06,
+      "loss": 0.6089,
+      "step": 5117
+    },
+    {
+      "epoch": 1.4177285318559556,
+      "grad_norm": 0.7070619463920593,
+      "learning_rate": 4.360946449980698e-06,
+      "loss": 0.6032,
+      "step": 5118
+    },
+    {
+      "epoch": 1.418005540166205,
+      "grad_norm": 0.6684051752090454,
+      "learning_rate": 4.360703176153237e-06,
+      "loss": 0.6237,
+      "step": 5119
+    },
+    {
+      "epoch": 1.4182825484764543,
+      "grad_norm": 0.6901852488517761,
+      "learning_rate": 4.360459862818296e-06,
+      "loss": 0.6612,
+      "step": 5120
+    },
+    {
+      "epoch": 1.4185595567867035,
+      "grad_norm": 0.7107248306274414,
+      "learning_rate": 4.360216509981042e-06,
+      "loss": 0.6356,
+      "step": 5121
+    },
+    {
+      "epoch": 1.418836565096953,
+      "grad_norm": 0.7306237816810608,
+      "learning_rate": 4.35997311764664e-06,
+      "loss": 0.6512,
+      "step": 5122
+    },
+    {
+      "epoch": 1.4191135734072022,
+      "grad_norm": 0.7205201387405396,
+      "learning_rate": 4.359729685820259e-06,
+      "loss": 0.6506,
+      "step": 5123
+    },
+    {
+      "epoch": 1.4193905817174515,
+      "grad_norm": 0.7062286138534546,
+      "learning_rate": 4.359486214507068e-06,
+      "loss": 0.6119,
+      "step": 5124
+    },
+    {
+      "epoch": 1.419667590027701,
+      "grad_norm": 0.6981781125068665,
+      "learning_rate": 4.359242703712235e-06,
+      "loss": 0.5616,
+      "step": 5125
+    },
+    {
+      "epoch": 1.4199445983379502,
+      "grad_norm": 0.7032713890075684,
+      "learning_rate": 4.358999153440932e-06,
+      "loss": 0.5854,
+      "step": 5126
+    },
+    {
+      "epoch": 1.4202216066481994,
+      "grad_norm": 0.6792232990264893,
+      "learning_rate": 4.358755563698331e-06,
+      "loss": 0.6368,
+      "step": 5127
+    },
+    {
+      "epoch": 1.4204986149584489,
+      "grad_norm": 0.6883552670478821,
+      "learning_rate": 4.358511934489601e-06,
+      "loss": 0.5862,
+      "step": 5128
+    },
+    {
+      "epoch": 1.420775623268698,
+      "grad_norm": 0.7066919803619385,
+      "learning_rate": 4.3582682658199175e-06,
+      "loss": 0.6181,
+      "step": 5129
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.7226448059082031,
+      "learning_rate": 4.358024557694452e-06,
+      "loss": 0.6098,
+      "step": 5130
+    },
+    {
+      "epoch": 1.4213296398891968,
+      "grad_norm": 0.7035112380981445,
+      "learning_rate": 4.357780810118381e-06,
+      "loss": 0.6122,
+      "step": 5131
+    },
+    {
+      "epoch": 1.421606648199446,
+      "grad_norm": 0.6860722899436951,
+      "learning_rate": 4.35753702309688e-06,
+      "loss": 0.6283,
+      "step": 5132
+    },
+    {
+      "epoch": 1.4218836565096953,
+      "grad_norm": 0.6797297596931458,
+      "learning_rate": 4.357293196635122e-06,
+      "loss": 0.5826,
+      "step": 5133
+    },
+    {
+      "epoch": 1.4221606648199445,
+      "grad_norm": 0.6257482767105103,
+      "learning_rate": 4.357049330738288e-06,
+      "loss": 0.6165,
+      "step": 5134
+    },
+    {
+      "epoch": 1.422437673130194,
+      "grad_norm": 0.6722486615180969,
+      "learning_rate": 4.356805425411555e-06,
+      "loss": 0.6323,
+      "step": 5135
+    },
+    {
+      "epoch": 1.4227146814404432,
+      "grad_norm": 0.6958386301994324,
+      "learning_rate": 4.3565614806600995e-06,
+      "loss": 0.6151,
+      "step": 5136
+    },
+    {
+      "epoch": 1.4229916897506925,
+      "grad_norm": 0.7121996879577637,
+      "learning_rate": 4.356317496489104e-06,
+      "loss": 0.6806,
+      "step": 5137
+    },
+    {
+      "epoch": 1.4232686980609417,
+      "grad_norm": 0.7728790640830994,
+      "learning_rate": 4.356073472903747e-06,
+      "loss": 0.6119,
+      "step": 5138
+    },
+    {
+      "epoch": 1.4235457063711912,
+      "grad_norm": 0.6925826072692871,
+      "learning_rate": 4.3558294099092095e-06,
+      "loss": 0.6252,
+      "step": 5139
+    },
+    {
+      "epoch": 1.4238227146814404,
+      "grad_norm": 0.717306911945343,
+      "learning_rate": 4.355585307510675e-06,
+      "loss": 0.6625,
+      "step": 5140
+    },
+    {
+      "epoch": 1.4240997229916896,
+      "grad_norm": 0.6816144585609436,
+      "learning_rate": 4.355341165713326e-06,
+      "loss": 0.6055,
+      "step": 5141
+    },
+    {
+      "epoch": 1.424376731301939,
+      "grad_norm": 0.6522184610366821,
+      "learning_rate": 4.355096984522346e-06,
+      "loss": 0.5748,
+      "step": 5142
+    },
+    {
+      "epoch": 1.4246537396121883,
+      "grad_norm": 0.7107172012329102,
+      "learning_rate": 4.354852763942919e-06,
+      "loss": 0.6665,
+      "step": 5143
+    },
+    {
+      "epoch": 1.4249307479224376,
+      "grad_norm": 0.7126984596252441,
+      "learning_rate": 4.354608503980232e-06,
+      "loss": 0.6294,
+      "step": 5144
+    },
+    {
+      "epoch": 1.425207756232687,
+      "grad_norm": 0.7015909552574158,
+      "learning_rate": 4.3543642046394696e-06,
+      "loss": 0.6172,
+      "step": 5145
+    },
+    {
+      "epoch": 1.4254847645429363,
+      "grad_norm": 0.6696215271949768,
+      "learning_rate": 4.35411986592582e-06,
+      "loss": 0.635,
+      "step": 5146
+    },
+    {
+      "epoch": 1.4257617728531855,
+      "grad_norm": 0.6961736679077148,
+      "learning_rate": 4.35387548784447e-06,
+      "loss": 0.6564,
+      "step": 5147
+    },
+    {
+      "epoch": 1.426038781163435,
+      "grad_norm": 0.6732077598571777,
+      "learning_rate": 4.3536310704006106e-06,
+      "loss": 0.5947,
+      "step": 5148
+    },
+    {
+      "epoch": 1.4263157894736842,
+      "grad_norm": 0.6792954802513123,
+      "learning_rate": 4.3533866135994284e-06,
+      "loss": 0.594,
+      "step": 5149
+    },
+    {
+      "epoch": 1.4265927977839334,
+      "grad_norm": 0.747936487197876,
+      "learning_rate": 4.353142117446115e-06,
+      "loss": 0.6103,
+      "step": 5150
+    },
+    {
+      "epoch": 1.426869806094183,
+      "grad_norm": 0.6711264252662659,
+      "learning_rate": 4.352897581945864e-06,
+      "loss": 0.6538,
+      "step": 5151
+    },
+    {
+      "epoch": 1.4271468144044321,
+      "grad_norm": 0.721182107925415,
+      "learning_rate": 4.352653007103864e-06,
+      "loss": 0.6649,
+      "step": 5152
+    },
+    {
+      "epoch": 1.4274238227146814,
+      "grad_norm": 0.6888142824172974,
+      "learning_rate": 4.352408392925311e-06,
+      "loss": 0.6392,
+      "step": 5153
+    },
+    {
+      "epoch": 1.4277008310249308,
+      "grad_norm": 0.7058656215667725,
+      "learning_rate": 4.352163739415396e-06,
+      "loss": 0.6148,
+      "step": 5154
+    },
+    {
+      "epoch": 1.42797783933518,
+      "grad_norm": 0.7143036127090454,
+      "learning_rate": 4.351919046579315e-06,
+      "loss": 0.6343,
+      "step": 5155
+    },
+    {
+      "epoch": 1.4282548476454293,
+      "grad_norm": 0.7155359983444214,
+      "learning_rate": 4.351674314422264e-06,
+      "loss": 0.6475,
+      "step": 5156
+    },
+    {
+      "epoch": 1.4285318559556788,
+      "grad_norm": 0.6871824860572815,
+      "learning_rate": 4.351429542949438e-06,
+      "loss": 0.6427,
+      "step": 5157
+    },
+    {
+      "epoch": 1.428808864265928,
+      "grad_norm": 0.6846596598625183,
+      "learning_rate": 4.351184732166035e-06,
+      "loss": 0.5837,
+      "step": 5158
+    },
+    {
+      "epoch": 1.4290858725761773,
+      "grad_norm": 0.7276217937469482,
+      "learning_rate": 4.350939882077252e-06,
+      "loss": 0.5906,
+      "step": 5159
+    },
+    {
+      "epoch": 1.4293628808864267,
+      "grad_norm": 0.7454106211662292,
+      "learning_rate": 4.350694992688289e-06,
+      "loss": 0.6098,
+      "step": 5160
+    },
+    {
+      "epoch": 1.429639889196676,
+      "grad_norm": 0.6848341226577759,
+      "learning_rate": 4.3504500640043455e-06,
+      "loss": 0.6442,
+      "step": 5161
+    },
+    {
+      "epoch": 1.4299168975069252,
+      "grad_norm": 0.6798859238624573,
+      "learning_rate": 4.350205096030622e-06,
+      "loss": 0.619,
+      "step": 5162
+    },
+    {
+      "epoch": 1.4301939058171746,
+      "grad_norm": 0.6953790187835693,
+      "learning_rate": 4.349960088772318e-06,
+      "loss": 0.5988,
+      "step": 5163
+    },
+    {
+      "epoch": 1.4304709141274239,
+      "grad_norm": 0.7027685642242432,
+      "learning_rate": 4.349715042234638e-06,
+      "loss": 0.6127,
+      "step": 5164
+    },
+    {
+      "epoch": 1.4307479224376731,
+      "grad_norm": 0.7262356877326965,
+      "learning_rate": 4.349469956422784e-06,
+      "loss": 0.678,
+      "step": 5165
+    },
+    {
+      "epoch": 1.4310249307479224,
+      "grad_norm": 0.7216992974281311,
+      "learning_rate": 4.34922483134196e-06,
+      "loss": 0.6285,
+      "step": 5166
+    },
+    {
+      "epoch": 1.4313019390581718,
+      "grad_norm": 0.6989957690238953,
+      "learning_rate": 4.34897966699737e-06,
+      "loss": 0.6434,
+      "step": 5167
+    },
+    {
+      "epoch": 1.431578947368421,
+      "grad_norm": 0.726123571395874,
+      "learning_rate": 4.34873446339422e-06,
+      "loss": 0.6514,
+      "step": 5168
+    },
+    {
+      "epoch": 1.4318559556786703,
+      "grad_norm": 0.7209747433662415,
+      "learning_rate": 4.348489220537716e-06,
+      "loss": 0.635,
+      "step": 5169
+    },
+    {
+      "epoch": 1.4321329639889195,
+      "grad_norm": 0.6753904223442078,
+      "learning_rate": 4.348243938433065e-06,
+      "loss": 0.6142,
+      "step": 5170
+    },
+    {
+      "epoch": 1.432409972299169,
+      "grad_norm": 0.740227222442627,
+      "learning_rate": 4.347998617085477e-06,
+      "loss": 0.631,
+      "step": 5171
+    },
+    {
+      "epoch": 1.4326869806094182,
+      "grad_norm": 0.6714673638343811,
+      "learning_rate": 4.347753256500158e-06,
+      "loss": 0.602,
+      "step": 5172
+    },
+    {
+      "epoch": 1.4329639889196675,
+      "grad_norm": 0.6836201548576355,
+      "learning_rate": 4.347507856682318e-06,
+      "loss": 0.63,
+      "step": 5173
+    },
+    {
+      "epoch": 1.433240997229917,
+      "grad_norm": 0.7172759175300598,
+      "learning_rate": 4.347262417637169e-06,
+      "loss": 0.6962,
+      "step": 5174
+    },
+    {
+      "epoch": 1.4335180055401662,
+      "grad_norm": 0.6770889163017273,
+      "learning_rate": 4.347016939369921e-06,
+      "loss": 0.6794,
+      "step": 5175
+    },
+    {
+      "epoch": 1.4337950138504154,
+      "grad_norm": 0.738587498664856,
+      "learning_rate": 4.346771421885787e-06,
+      "loss": 0.6369,
+      "step": 5176
+    },
+    {
+      "epoch": 1.4340720221606649,
+      "grad_norm": 0.7032896876335144,
+      "learning_rate": 4.346525865189979e-06,
+      "loss": 0.6236,
+      "step": 5177
+    },
+    {
+      "epoch": 1.434349030470914,
+      "grad_norm": 0.7206043004989624,
+      "learning_rate": 4.346280269287712e-06,
+      "loss": 0.6393,
+      "step": 5178
+    },
+    {
+      "epoch": 1.4346260387811633,
+      "grad_norm": 0.7211987376213074,
+      "learning_rate": 4.3460346341842e-06,
+      "loss": 0.6348,
+      "step": 5179
+    },
+    {
+      "epoch": 1.4349030470914128,
+      "grad_norm": 0.6673139333724976,
+      "learning_rate": 4.345788959884658e-06,
+      "loss": 0.6243,
+      "step": 5180
+    },
+    {
+      "epoch": 1.435180055401662,
+      "grad_norm": 0.7322635054588318,
+      "learning_rate": 4.345543246394303e-06,
+      "loss": 0.6455,
+      "step": 5181
+    },
+    {
+      "epoch": 1.4354570637119113,
+      "grad_norm": 0.6816807985305786,
+      "learning_rate": 4.345297493718352e-06,
+      "loss": 0.6102,
+      "step": 5182
+    },
+    {
+      "epoch": 1.4357340720221607,
+      "grad_norm": 0.6793221831321716,
+      "learning_rate": 4.345051701862023e-06,
+      "loss": 0.6377,
+      "step": 5183
+    },
+    {
+      "epoch": 1.43601108033241,
+      "grad_norm": 0.6505659222602844,
+      "learning_rate": 4.344805870830534e-06,
+      "loss": 0.5451,
+      "step": 5184
+    },
+    {
+      "epoch": 1.4362880886426592,
+      "grad_norm": 0.7179732322692871,
+      "learning_rate": 4.344560000629105e-06,
+      "loss": 0.6237,
+      "step": 5185
+    },
+    {
+      "epoch": 1.4365650969529087,
+      "grad_norm": 0.7613140940666199,
+      "learning_rate": 4.344314091262958e-06,
+      "loss": 0.6479,
+      "step": 5186
+    },
+    {
+      "epoch": 1.436842105263158,
+      "grad_norm": 0.7074609398841858,
+      "learning_rate": 4.344068142737312e-06,
+      "loss": 0.6614,
+      "step": 5187
+    },
+    {
+      "epoch": 1.4371191135734072,
+      "grad_norm": 0.7174100279808044,
+      "learning_rate": 4.34382215505739e-06,
+      "loss": 0.622,
+      "step": 5188
+    },
+    {
+      "epoch": 1.4373961218836566,
+      "grad_norm": 0.6789400577545166,
+      "learning_rate": 4.343576128228416e-06,
+      "loss": 0.6135,
+      "step": 5189
+    },
+    {
+      "epoch": 1.4376731301939059,
+      "grad_norm": 0.6755404472351074,
+      "learning_rate": 4.343330062255612e-06,
+      "loss": 0.5856,
+      "step": 5190
+    },
+    {
+      "epoch": 1.437950138504155,
+      "grad_norm": 0.6998751163482666,
+      "learning_rate": 4.343083957144204e-06,
+      "loss": 0.6333,
+      "step": 5191
+    },
+    {
+      "epoch": 1.4382271468144046,
+      "grad_norm": 0.70278000831604,
+      "learning_rate": 4.3428378128994165e-06,
+      "loss": 0.5697,
+      "step": 5192
+    },
+    {
+      "epoch": 1.4385041551246538,
+      "grad_norm": 0.6691498160362244,
+      "learning_rate": 4.342591629526476e-06,
+      "loss": 0.6252,
+      "step": 5193
+    },
+    {
+      "epoch": 1.438781163434903,
+      "grad_norm": 0.6869835257530212,
+      "learning_rate": 4.34234540703061e-06,
+      "loss": 0.6272,
+      "step": 5194
+    },
+    {
+      "epoch": 1.4390581717451525,
+      "grad_norm": 0.663788914680481,
+      "learning_rate": 4.342099145417046e-06,
+      "loss": 0.6791,
+      "step": 5195
+    },
+    {
+      "epoch": 1.4393351800554017,
+      "grad_norm": 0.6984208822250366,
+      "learning_rate": 4.341852844691012e-06,
+      "loss": 0.6582,
+      "step": 5196
+    },
+    {
+      "epoch": 1.439612188365651,
+      "grad_norm": 0.6600984930992126,
+      "learning_rate": 4.34160650485774e-06,
+      "loss": 0.6389,
+      "step": 5197
+    },
+    {
+      "epoch": 1.4398891966759002,
+      "grad_norm": 0.7514785528182983,
+      "learning_rate": 4.341360125922458e-06,
+      "loss": 0.6174,
+      "step": 5198
+    },
+    {
+      "epoch": 1.4401662049861497,
+      "grad_norm": 0.7093493342399597,
+      "learning_rate": 4.3411137078904e-06,
+      "loss": 0.617,
+      "step": 5199
+    },
+    {
+      "epoch": 1.440443213296399,
+      "grad_norm": 0.639533281326294,
+      "learning_rate": 4.340867250766794e-06,
+      "loss": 0.5836,
+      "step": 5200
+    },
+    {
+      "epoch": 1.4407202216066481,
+      "grad_norm": 0.6882728934288025,
+      "learning_rate": 4.340620754556876e-06,
+      "loss": 0.6223,
+      "step": 5201
+    },
+    {
+      "epoch": 1.4409972299168974,
+      "grad_norm": 0.7091029286384583,
+      "learning_rate": 4.340374219265879e-06,
+      "loss": 0.6513,
+      "step": 5202
+    },
+    {
+      "epoch": 1.4412742382271468,
+      "grad_norm": 0.6923207640647888,
+      "learning_rate": 4.340127644899038e-06,
+      "loss": 0.6318,
+      "step": 5203
+    },
+    {
+      "epoch": 1.441551246537396,
+      "grad_norm": 0.7474663257598877,
+      "learning_rate": 4.339881031461588e-06,
+      "loss": 0.6651,
+      "step": 5204
+    },
+    {
+      "epoch": 1.4418282548476453,
+      "grad_norm": 0.7120457291603088,
+      "learning_rate": 4.339634378958765e-06,
+      "loss": 0.6202,
+      "step": 5205
+    },
+    {
+      "epoch": 1.4421052631578948,
+      "grad_norm": 0.7063796520233154,
+      "learning_rate": 4.339387687395805e-06,
+      "loss": 0.6105,
+      "step": 5206
+    },
+    {
+      "epoch": 1.442382271468144,
+      "grad_norm": 0.6743805408477783,
+      "learning_rate": 4.3391409567779485e-06,
+      "loss": 0.5991,
+      "step": 5207
+    },
+    {
+      "epoch": 1.4426592797783933,
+      "grad_norm": 0.6794716715812683,
+      "learning_rate": 4.338894187110432e-06,
+      "loss": 0.5837,
+      "step": 5208
+    },
+    {
+      "epoch": 1.4429362880886427,
+      "grad_norm": 0.709031879901886,
+      "learning_rate": 4.338647378398496e-06,
+      "loss": 0.5933,
+      "step": 5209
+    },
+    {
+      "epoch": 1.443213296398892,
+      "grad_norm": 0.6801929473876953,
+      "learning_rate": 4.338400530647382e-06,
+      "loss": 0.6406,
+      "step": 5210
+    },
+    {
+      "epoch": 1.4434903047091412,
+      "grad_norm": 0.7303329706192017,
+      "learning_rate": 4.3381536438623286e-06,
+      "loss": 0.6533,
+      "step": 5211
+    },
+    {
+      "epoch": 1.4437673130193907,
+      "grad_norm": 0.7102288603782654,
+      "learning_rate": 4.3379067180485795e-06,
+      "loss": 0.6582,
+      "step": 5212
+    },
+    {
+      "epoch": 1.44404432132964,
+      "grad_norm": 0.7001471519470215,
+      "learning_rate": 4.337659753211377e-06,
+      "loss": 0.6493,
+      "step": 5213
+    },
+    {
+      "epoch": 1.4443213296398891,
+      "grad_norm": 0.7027942538261414,
+      "learning_rate": 4.337412749355966e-06,
+      "loss": 0.6005,
+      "step": 5214
+    },
+    {
+      "epoch": 1.4445983379501386,
+      "grad_norm": 0.6876331567764282,
+      "learning_rate": 4.337165706487589e-06,
+      "loss": 0.669,
+      "step": 5215
+    },
+    {
+      "epoch": 1.4448753462603878,
+      "grad_norm": 0.707457959651947,
+      "learning_rate": 4.336918624611494e-06,
+      "loss": 0.6003,
+      "step": 5216
+    },
+    {
+      "epoch": 1.445152354570637,
+      "grad_norm": 0.7237023115158081,
+      "learning_rate": 4.336671503732924e-06,
+      "loss": 0.6595,
+      "step": 5217
+    },
+    {
+      "epoch": 1.4454293628808865,
+      "grad_norm": 0.7044596672058105,
+      "learning_rate": 4.336424343857128e-06,
+      "loss": 0.6118,
+      "step": 5218
+    },
+    {
+      "epoch": 1.4457063711911358,
+      "grad_norm": 0.719316303730011,
+      "learning_rate": 4.336177144989354e-06,
+      "loss": 0.6243,
+      "step": 5219
+    },
+    {
+      "epoch": 1.445983379501385,
+      "grad_norm": 0.6969180107116699,
+      "learning_rate": 4.335929907134849e-06,
+      "loss": 0.656,
+      "step": 5220
+    },
+    {
+      "epoch": 1.4462603878116345,
+      "grad_norm": 0.7368977665901184,
+      "learning_rate": 4.335682630298865e-06,
+      "loss": 0.6949,
+      "step": 5221
+    },
+    {
+      "epoch": 1.4465373961218837,
+      "grad_norm": 0.68007892370224,
+      "learning_rate": 4.33543531448665e-06,
+      "loss": 0.6299,
+      "step": 5222
+    },
+    {
+      "epoch": 1.446814404432133,
+      "grad_norm": 0.6848158836364746,
+      "learning_rate": 4.335187959703457e-06,
+      "loss": 0.6451,
+      "step": 5223
+    },
+    {
+      "epoch": 1.4470914127423824,
+      "grad_norm": 0.7511315941810608,
+      "learning_rate": 4.334940565954536e-06,
+      "loss": 0.651,
+      "step": 5224
+    },
+    {
+      "epoch": 1.4473684210526316,
+      "grad_norm": 0.6904433965682983,
+      "learning_rate": 4.334693133245141e-06,
+      "loss": 0.6535,
+      "step": 5225
+    },
+    {
+      "epoch": 1.4476454293628809,
+      "grad_norm": 0.7157757878303528,
+      "learning_rate": 4.334445661580527e-06,
+      "loss": 0.6092,
+      "step": 5226
+    },
+    {
+      "epoch": 1.4479224376731301,
+      "grad_norm": 0.7234309315681458,
+      "learning_rate": 4.334198150965946e-06,
+      "loss": 0.6465,
+      "step": 5227
+    },
+    {
+      "epoch": 1.4481994459833796,
+      "grad_norm": 0.7146196961402893,
+      "learning_rate": 4.333950601406653e-06,
+      "loss": 0.6399,
+      "step": 5228
+    },
+    {
+      "epoch": 1.4484764542936288,
+      "grad_norm": 0.7037093043327332,
+      "learning_rate": 4.333703012907907e-06,
+      "loss": 0.6356,
+      "step": 5229
+    },
+    {
+      "epoch": 1.448753462603878,
+      "grad_norm": 0.6986050605773926,
+      "learning_rate": 4.333455385474962e-06,
+      "loss": 0.6332,
+      "step": 5230
+    },
+    {
+      "epoch": 1.4490304709141273,
+      "grad_norm": 0.6712711453437805,
+      "learning_rate": 4.333207719113078e-06,
+      "loss": 0.6631,
+      "step": 5231
+    },
+    {
+      "epoch": 1.4493074792243767,
+      "grad_norm": 0.7003492116928101,
+      "learning_rate": 4.332960013827512e-06,
+      "loss": 0.6268,
+      "step": 5232
+    },
+    {
+      "epoch": 1.449584487534626,
+      "grad_norm": 0.705556333065033,
+      "learning_rate": 4.332712269623525e-06,
+      "loss": 0.6109,
+      "step": 5233
+    },
+    {
+      "epoch": 1.4498614958448752,
+      "grad_norm": 0.69657301902771,
+      "learning_rate": 4.332464486506375e-06,
+      "loss": 0.5833,
+      "step": 5234
+    },
+    {
+      "epoch": 1.4501385041551247,
+      "grad_norm": 0.682807445526123,
+      "learning_rate": 4.332216664481325e-06,
+      "loss": 0.5722,
+      "step": 5235
+    },
+    {
+      "epoch": 1.450415512465374,
+      "grad_norm": 0.6681585907936096,
+      "learning_rate": 4.331968803553636e-06,
+      "loss": 0.6012,
+      "step": 5236
+    },
+    {
+      "epoch": 1.4506925207756232,
+      "grad_norm": 0.6505764722824097,
+      "learning_rate": 4.331720903728571e-06,
+      "loss": 0.6116,
+      "step": 5237
+    },
+    {
+      "epoch": 1.4509695290858726,
+      "grad_norm": 0.7191979885101318,
+      "learning_rate": 4.331472965011394e-06,
+      "loss": 0.6577,
+      "step": 5238
+    },
+    {
+      "epoch": 1.4512465373961219,
+      "grad_norm": 0.6958252191543579,
+      "learning_rate": 4.331224987407369e-06,
+      "loss": 0.6515,
+      "step": 5239
+    },
+    {
+      "epoch": 1.451523545706371,
+      "grad_norm": 0.6717762351036072,
+      "learning_rate": 4.3309769709217605e-06,
+      "loss": 0.5901,
+      "step": 5240
+    },
+    {
+      "epoch": 1.4518005540166206,
+      "grad_norm": 0.716303825378418,
+      "learning_rate": 4.330728915559834e-06,
+      "loss": 0.652,
+      "step": 5241
+    },
+    {
+      "epoch": 1.4520775623268698,
+      "grad_norm": 0.6768020391464233,
+      "learning_rate": 4.33048082132686e-06,
+      "loss": 0.5722,
+      "step": 5242
+    },
+    {
+      "epoch": 1.452354570637119,
+      "grad_norm": 0.7079585790634155,
+      "learning_rate": 4.330232688228102e-06,
+      "loss": 0.6415,
+      "step": 5243
+    },
+    {
+      "epoch": 1.4526315789473685,
+      "grad_norm": 0.7371026277542114,
+      "learning_rate": 4.329984516268831e-06,
+      "loss": 0.6219,
+      "step": 5244
+    },
+    {
+      "epoch": 1.4529085872576177,
+      "grad_norm": 0.6973332166671753,
+      "learning_rate": 4.329736305454314e-06,
+      "loss": 0.6243,
+      "step": 5245
+    },
+    {
+      "epoch": 1.453185595567867,
+      "grad_norm": 0.6986327767372131,
+      "learning_rate": 4.329488055789824e-06,
+      "loss": 0.6318,
+      "step": 5246
+    },
+    {
+      "epoch": 1.4534626038781164,
+      "grad_norm": 0.6920615434646606,
+      "learning_rate": 4.3292397672806296e-06,
+      "loss": 0.6202,
+      "step": 5247
+    },
+    {
+      "epoch": 1.4537396121883657,
+      "grad_norm": 0.7052721977233887,
+      "learning_rate": 4.328991439932003e-06,
+      "loss": 0.6409,
+      "step": 5248
+    },
+    {
+      "epoch": 1.454016620498615,
+      "grad_norm": 0.6558910608291626,
+      "learning_rate": 4.328743073749219e-06,
+      "loss": 0.6174,
+      "step": 5249
+    },
+    {
+      "epoch": 1.4542936288088644,
+      "grad_norm": 0.696160078048706,
+      "learning_rate": 4.328494668737549e-06,
+      "loss": 0.6321,
+      "step": 5250
+    },
+    {
+      "epoch": 1.4545706371191136,
+      "grad_norm": 0.7242604494094849,
+      "learning_rate": 4.328246224902267e-06,
+      "loss": 0.6464,
+      "step": 5251
+    },
+    {
+      "epoch": 1.4548476454293628,
+      "grad_norm": 0.6753219366073608,
+      "learning_rate": 4.3279977422486495e-06,
+      "loss": 0.6298,
+      "step": 5252
+    },
+    {
+      "epoch": 1.4551246537396123,
+      "grad_norm": 0.7261792421340942,
+      "learning_rate": 4.327749220781972e-06,
+      "loss": 0.6281,
+      "step": 5253
+    },
+    {
+      "epoch": 1.4554016620498615,
+      "grad_norm": 0.7096898555755615,
+      "learning_rate": 4.327500660507511e-06,
+      "loss": 0.6053,
+      "step": 5254
+    },
+    {
+      "epoch": 1.4556786703601108,
+      "grad_norm": 0.6910092234611511,
+      "learning_rate": 4.327252061430543e-06,
+      "loss": 0.6212,
+      "step": 5255
+    },
+    {
+      "epoch": 1.4559556786703602,
+      "grad_norm": 0.6957384943962097,
+      "learning_rate": 4.327003423556349e-06,
+      "loss": 0.6263,
+      "step": 5256
+    },
+    {
+      "epoch": 1.4562326869806095,
+      "grad_norm": 0.6883211731910706,
+      "learning_rate": 4.326754746890205e-06,
+      "loss": 0.644,
+      "step": 5257
+    },
+    {
+      "epoch": 1.4565096952908587,
+      "grad_norm": 0.7107007503509521,
+      "learning_rate": 4.326506031437393e-06,
+      "loss": 0.6443,
+      "step": 5258
+    },
+    {
+      "epoch": 1.456786703601108,
+      "grad_norm": 0.708169162273407,
+      "learning_rate": 4.326257277203194e-06,
+      "loss": 0.6279,
+      "step": 5259
+    },
+    {
+      "epoch": 1.4570637119113574,
+      "grad_norm": 0.7273499369621277,
+      "learning_rate": 4.326008484192889e-06,
+      "loss": 0.684,
+      "step": 5260
+    },
+    {
+      "epoch": 1.4573407202216067,
+      "grad_norm": 0.7274107336997986,
+      "learning_rate": 4.325759652411762e-06,
+      "loss": 0.5453,
+      "step": 5261
+    },
+    {
+      "epoch": 1.457617728531856,
+      "grad_norm": 0.6847637891769409,
+      "learning_rate": 4.325510781865094e-06,
+      "loss": 0.6211,
+      "step": 5262
+    },
+    {
+      "epoch": 1.4578947368421051,
+      "grad_norm": 0.6615282893180847,
+      "learning_rate": 4.32526187255817e-06,
+      "loss": 0.5822,
+      "step": 5263
+    },
+    {
+      "epoch": 1.4581717451523546,
+      "grad_norm": 0.6708592176437378,
+      "learning_rate": 4.325012924496276e-06,
+      "loss": 0.637,
+      "step": 5264
+    },
+    {
+      "epoch": 1.4584487534626038,
+      "grad_norm": 0.7095727324485779,
+      "learning_rate": 4.324763937684696e-06,
+      "loss": 0.6257,
+      "step": 5265
+    },
+    {
+      "epoch": 1.458725761772853,
+      "grad_norm": 0.6905611753463745,
+      "learning_rate": 4.324514912128719e-06,
+      "loss": 0.6167,
+      "step": 5266
+    },
+    {
+      "epoch": 1.4590027700831025,
+      "grad_norm": 0.7230263352394104,
+      "learning_rate": 4.32426584783363e-06,
+      "loss": 0.5622,
+      "step": 5267
+    },
+    {
+      "epoch": 1.4592797783933518,
+      "grad_norm": 0.6828804612159729,
+      "learning_rate": 4.324016744804719e-06,
+      "loss": 0.6402,
+      "step": 5268
+    },
+    {
+      "epoch": 1.459556786703601,
+      "grad_norm": 0.7129217982292175,
+      "learning_rate": 4.323767603047274e-06,
+      "loss": 0.6777,
+      "step": 5269
+    },
+    {
+      "epoch": 1.4598337950138505,
+      "grad_norm": 0.6748174428939819,
+      "learning_rate": 4.323518422566586e-06,
+      "loss": 0.609,
+      "step": 5270
+    },
+    {
+      "epoch": 1.4601108033240997,
+      "grad_norm": 0.6736699342727661,
+      "learning_rate": 4.323269203367944e-06,
+      "loss": 0.6251,
+      "step": 5271
+    },
+    {
+      "epoch": 1.460387811634349,
+      "grad_norm": 0.706195592880249,
+      "learning_rate": 4.323019945456642e-06,
+      "loss": 0.5806,
+      "step": 5272
+    },
+    {
+      "epoch": 1.4606648199445984,
+      "grad_norm": 0.7386101484298706,
+      "learning_rate": 4.32277064883797e-06,
+      "loss": 0.6047,
+      "step": 5273
+    },
+    {
+      "epoch": 1.4609418282548476,
+      "grad_norm": 0.7328194975852966,
+      "learning_rate": 4.322521313517223e-06,
+      "loss": 0.6254,
+      "step": 5274
+    },
+    {
+      "epoch": 1.4612188365650969,
+      "grad_norm": 0.6765896081924438,
+      "learning_rate": 4.3222719394996944e-06,
+      "loss": 0.6273,
+      "step": 5275
+    },
+    {
+      "epoch": 1.4614958448753463,
+      "grad_norm": 0.6815961003303528,
+      "learning_rate": 4.322022526790679e-06,
+      "loss": 0.6604,
+      "step": 5276
+    },
+    {
+      "epoch": 1.4617728531855956,
+      "grad_norm": 0.7059053778648376,
+      "learning_rate": 4.321773075395473e-06,
+      "loss": 0.6445,
+      "step": 5277
+    },
+    {
+      "epoch": 1.4620498614958448,
+      "grad_norm": 0.6844702363014221,
+      "learning_rate": 4.321523585319371e-06,
+      "loss": 0.6199,
+      "step": 5278
+    },
+    {
+      "epoch": 1.4623268698060943,
+      "grad_norm": 0.6714056134223938,
+      "learning_rate": 4.321274056567672e-06,
+      "loss": 0.5777,
+      "step": 5279
+    },
+    {
+      "epoch": 1.4626038781163435,
+      "grad_norm": 0.7048715353012085,
+      "learning_rate": 4.321024489145674e-06,
+      "loss": 0.658,
+      "step": 5280
+    },
+    {
+      "epoch": 1.4628808864265928,
+      "grad_norm": 0.6695634126663208,
+      "learning_rate": 4.320774883058675e-06,
+      "loss": 0.6579,
+      "step": 5281
+    },
+    {
+      "epoch": 1.4631578947368422,
+      "grad_norm": 0.735461413860321,
+      "learning_rate": 4.320525238311976e-06,
+      "loss": 0.6551,
+      "step": 5282
+    },
+    {
+      "epoch": 1.4634349030470915,
+      "grad_norm": 0.6574238538742065,
+      "learning_rate": 4.320275554910876e-06,
+      "loss": 0.5934,
+      "step": 5283
+    },
+    {
+      "epoch": 1.4637119113573407,
+      "grad_norm": 0.6849135160446167,
+      "learning_rate": 4.320025832860679e-06,
+      "loss": 0.5843,
+      "step": 5284
+    },
+    {
+      "epoch": 1.4639889196675901,
+      "grad_norm": 0.7034268975257874,
+      "learning_rate": 4.319776072166686e-06,
+      "loss": 0.6099,
+      "step": 5285
+    },
+    {
+      "epoch": 1.4642659279778394,
+      "grad_norm": 0.7210450768470764,
+      "learning_rate": 4.319526272834198e-06,
+      "loss": 0.6202,
+      "step": 5286
+    },
+    {
+      "epoch": 1.4645429362880886,
+      "grad_norm": 0.7004812359809875,
+      "learning_rate": 4.319276434868522e-06,
+      "loss": 0.6692,
+      "step": 5287
+    },
+    {
+      "epoch": 1.464819944598338,
+      "grad_norm": 0.7177503705024719,
+      "learning_rate": 4.319026558274961e-06,
+      "loss": 0.6493,
+      "step": 5288
+    },
+    {
+      "epoch": 1.4650969529085873,
+      "grad_norm": 0.6847087740898132,
+      "learning_rate": 4.3187766430588215e-06,
+      "loss": 0.6499,
+      "step": 5289
+    },
+    {
+      "epoch": 1.4653739612188366,
+      "grad_norm": 0.7065343856811523,
+      "learning_rate": 4.318526689225408e-06,
+      "loss": 0.6308,
+      "step": 5290
+    },
+    {
+      "epoch": 1.4656509695290858,
+      "grad_norm": 0.687520444393158,
+      "learning_rate": 4.31827669678003e-06,
+      "loss": 0.66,
+      "step": 5291
+    },
+    {
+      "epoch": 1.4659279778393353,
+      "grad_norm": 0.6558403968811035,
+      "learning_rate": 4.318026665727993e-06,
+      "loss": 0.6164,
+      "step": 5292
+    },
+    {
+      "epoch": 1.4662049861495845,
+      "grad_norm": 0.7158898711204529,
+      "learning_rate": 4.317776596074609e-06,
+      "loss": 0.6015,
+      "step": 5293
+    },
+    {
+      "epoch": 1.4664819944598337,
+      "grad_norm": 0.6923995614051819,
+      "learning_rate": 4.317526487825185e-06,
+      "loss": 0.617,
+      "step": 5294
+    },
+    {
+      "epoch": 1.466759002770083,
+      "grad_norm": 0.6958532333374023,
+      "learning_rate": 4.317276340985032e-06,
+      "loss": 0.6142,
+      "step": 5295
+    },
+    {
+      "epoch": 1.4670360110803324,
+      "grad_norm": 0.6740797758102417,
+      "learning_rate": 4.3170261555594615e-06,
+      "loss": 0.6077,
+      "step": 5296
+    },
+    {
+      "epoch": 1.4673130193905817,
+      "grad_norm": 0.7106777429580688,
+      "learning_rate": 4.316775931553785e-06,
+      "loss": 0.6472,
+      "step": 5297
+    },
+    {
+      "epoch": 1.467590027700831,
+      "grad_norm": 0.6848693490028381,
+      "learning_rate": 4.316525668973317e-06,
+      "loss": 0.6531,
+      "step": 5298
+    },
+    {
+      "epoch": 1.4678670360110804,
+      "grad_norm": 0.6626380681991577,
+      "learning_rate": 4.316275367823369e-06,
+      "loss": 0.6212,
+      "step": 5299
+    },
+    {
+      "epoch": 1.4681440443213296,
+      "grad_norm": 0.7039237022399902,
+      "learning_rate": 4.3160250281092575e-06,
+      "loss": 0.653,
+      "step": 5300
+    },
+    {
+      "epoch": 1.4684210526315788,
+      "grad_norm": 0.6798372268676758,
+      "learning_rate": 4.315774649836297e-06,
+      "loss": 0.5864,
+      "step": 5301
+    },
+    {
+      "epoch": 1.4686980609418283,
+      "grad_norm": 0.6488889455795288,
+      "learning_rate": 4.315524233009803e-06,
+      "loss": 0.6323,
+      "step": 5302
+    },
+    {
+      "epoch": 1.4689750692520775,
+      "grad_norm": 0.6437256932258606,
+      "learning_rate": 4.3152737776350935e-06,
+      "loss": 0.5617,
+      "step": 5303
+    },
+    {
+      "epoch": 1.4692520775623268,
+      "grad_norm": 0.6839866638183594,
+      "learning_rate": 4.315023283717487e-06,
+      "loss": 0.6437,
+      "step": 5304
+    },
+    {
+      "epoch": 1.4695290858725762,
+      "grad_norm": 0.6503082513809204,
+      "learning_rate": 4.3147727512623e-06,
+      "loss": 0.6379,
+      "step": 5305
+    },
+    {
+      "epoch": 1.4698060941828255,
+      "grad_norm": 0.7124677300453186,
+      "learning_rate": 4.314522180274854e-06,
+      "loss": 0.6483,
+      "step": 5306
+    },
+    {
+      "epoch": 1.4700831024930747,
+      "grad_norm": 0.6634410619735718,
+      "learning_rate": 4.314271570760467e-06,
+      "loss": 0.6248,
+      "step": 5307
+    },
+    {
+      "epoch": 1.4703601108033242,
+      "grad_norm": 0.671969473361969,
+      "learning_rate": 4.314020922724462e-06,
+      "loss": 0.6142,
+      "step": 5308
+    },
+    {
+      "epoch": 1.4706371191135734,
+      "grad_norm": 0.7019748091697693,
+      "learning_rate": 4.313770236172161e-06,
+      "loss": 0.6396,
+      "step": 5309
+    },
+    {
+      "epoch": 1.4709141274238227,
+      "grad_norm": 0.7120486497879028,
+      "learning_rate": 4.313519511108886e-06,
+      "loss": 0.6307,
+      "step": 5310
+    },
+    {
+      "epoch": 1.4711911357340721,
+      "grad_norm": 0.7347435355186462,
+      "learning_rate": 4.313268747539959e-06,
+      "loss": 0.6312,
+      "step": 5311
+    },
+    {
+      "epoch": 1.4714681440443214,
+      "grad_norm": 0.6481282114982605,
+      "learning_rate": 4.3130179454707075e-06,
+      "loss": 0.6131,
+      "step": 5312
+    },
+    {
+      "epoch": 1.4717451523545706,
+      "grad_norm": 0.6802014708518982,
+      "learning_rate": 4.3127671049064546e-06,
+      "loss": 0.6703,
+      "step": 5313
+    },
+    {
+      "epoch": 1.47202216066482,
+      "grad_norm": 0.6929247975349426,
+      "learning_rate": 4.3125162258525265e-06,
+      "loss": 0.6416,
+      "step": 5314
+    },
+    {
+      "epoch": 1.4722991689750693,
+      "grad_norm": 0.7569963932037354,
+      "learning_rate": 4.312265308314251e-06,
+      "loss": 0.6125,
+      "step": 5315
+    },
+    {
+      "epoch": 1.4725761772853185,
+      "grad_norm": 0.6945342421531677,
+      "learning_rate": 4.312014352296954e-06,
+      "loss": 0.6376,
+      "step": 5316
+    },
+    {
+      "epoch": 1.472853185595568,
+      "grad_norm": 0.6991958618164062,
+      "learning_rate": 4.3117633578059645e-06,
+      "loss": 0.5975,
+      "step": 5317
+    },
+    {
+      "epoch": 1.4731301939058172,
+      "grad_norm": 0.709173321723938,
+      "learning_rate": 4.311512324846614e-06,
+      "loss": 0.648,
+      "step": 5318
+    },
+    {
+      "epoch": 1.4734072022160665,
+      "grad_norm": 0.7404986619949341,
+      "learning_rate": 4.311261253424229e-06,
+      "loss": 0.6158,
+      "step": 5319
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.7493718266487122,
+      "learning_rate": 4.311010143544143e-06,
+      "loss": 0.6611,
+      "step": 5320
+    },
+    {
+      "epoch": 1.4739612188365652,
+      "grad_norm": 0.7251834273338318,
+      "learning_rate": 4.310758995211687e-06,
+      "loss": 0.545,
+      "step": 5321
+    },
+    {
+      "epoch": 1.4742382271468144,
+      "grad_norm": 0.7508950233459473,
+      "learning_rate": 4.310507808432193e-06,
+      "loss": 0.645,
+      "step": 5322
+    },
+    {
+      "epoch": 1.4745152354570636,
+      "grad_norm": 0.7097291350364685,
+      "learning_rate": 4.310256583210995e-06,
+      "loss": 0.6233,
+      "step": 5323
+    },
+    {
+      "epoch": 1.4747922437673129,
+      "grad_norm": 0.6722292900085449,
+      "learning_rate": 4.3100053195534265e-06,
+      "loss": 0.6236,
+      "step": 5324
+    },
+    {
+      "epoch": 1.4750692520775623,
+      "grad_norm": 0.6865547895431519,
+      "learning_rate": 4.309754017464823e-06,
+      "loss": 0.6743,
+      "step": 5325
+    },
+    {
+      "epoch": 1.4753462603878116,
+      "grad_norm": 0.6935882568359375,
+      "learning_rate": 4.30950267695052e-06,
+      "loss": 0.634,
+      "step": 5326
+    },
+    {
+      "epoch": 1.4756232686980608,
+      "grad_norm": 0.7458108067512512,
+      "learning_rate": 4.3092512980158545e-06,
+      "loss": 0.6486,
+      "step": 5327
+    },
+    {
+      "epoch": 1.4759002770083103,
+      "grad_norm": 0.693493664264679,
+      "learning_rate": 4.308999880666163e-06,
+      "loss": 0.6331,
+      "step": 5328
+    },
+    {
+      "epoch": 1.4761772853185595,
+      "grad_norm": 0.6644585728645325,
+      "learning_rate": 4.308748424906785e-06,
+      "loss": 0.6216,
+      "step": 5329
+    },
+    {
+      "epoch": 1.4764542936288088,
+      "grad_norm": 0.7142424583435059,
+      "learning_rate": 4.308496930743058e-06,
+      "loss": 0.6425,
+      "step": 5330
+    },
+    {
+      "epoch": 1.4767313019390582,
+      "grad_norm": 0.7267554402351379,
+      "learning_rate": 4.308245398180323e-06,
+      "loss": 0.6313,
+      "step": 5331
+    },
+    {
+      "epoch": 1.4770083102493075,
+      "grad_norm": 0.666397213935852,
+      "learning_rate": 4.307993827223922e-06,
+      "loss": 0.6631,
+      "step": 5332
+    },
+    {
+      "epoch": 1.4772853185595567,
+      "grad_norm": 0.7025532722473145,
+      "learning_rate": 4.307742217879193e-06,
+      "loss": 0.6206,
+      "step": 5333
+    },
+    {
+      "epoch": 1.4775623268698062,
+      "grad_norm": 0.7297946810722351,
+      "learning_rate": 4.307490570151481e-06,
+      "loss": 0.5774,
+      "step": 5334
+    },
+    {
+      "epoch": 1.4778393351800554,
+      "grad_norm": 0.672098696231842,
+      "learning_rate": 4.307238884046129e-06,
+      "loss": 0.612,
+      "step": 5335
+    },
+    {
+      "epoch": 1.4781163434903046,
+      "grad_norm": 0.6860193610191345,
+      "learning_rate": 4.3069871595684795e-06,
+      "loss": 0.6446,
+      "step": 5336
+    },
+    {
+      "epoch": 1.478393351800554,
+      "grad_norm": 0.6797826886177063,
+      "learning_rate": 4.306735396723878e-06,
+      "loss": 0.6536,
+      "step": 5337
+    },
+    {
+      "epoch": 1.4786703601108033,
+      "grad_norm": 0.6978946924209595,
+      "learning_rate": 4.30648359551767e-06,
+      "loss": 0.6221,
+      "step": 5338
+    },
+    {
+      "epoch": 1.4789473684210526,
+      "grad_norm": 0.694804847240448,
+      "learning_rate": 4.306231755955202e-06,
+      "loss": 0.6012,
+      "step": 5339
+    },
+    {
+      "epoch": 1.479224376731302,
+      "grad_norm": 0.7225304841995239,
+      "learning_rate": 4.305979878041823e-06,
+      "loss": 0.6955,
+      "step": 5340
+    },
+    {
+      "epoch": 1.4795013850415513,
+      "grad_norm": 0.7552931904792786,
+      "learning_rate": 4.305727961782877e-06,
+      "loss": 0.5893,
+      "step": 5341
+    },
+    {
+      "epoch": 1.4797783933518005,
+      "grad_norm": 0.7309013605117798,
+      "learning_rate": 4.305476007183715e-06,
+      "loss": 0.6621,
+      "step": 5342
+    },
+    {
+      "epoch": 1.48005540166205,
+      "grad_norm": 0.6812788844108582,
+      "learning_rate": 4.305224014249688e-06,
+      "loss": 0.6633,
+      "step": 5343
+    },
+    {
+      "epoch": 1.4803324099722992,
+      "grad_norm": 0.7306602597236633,
+      "learning_rate": 4.304971982986145e-06,
+      "loss": 0.6242,
+      "step": 5344
+    },
+    {
+      "epoch": 1.4806094182825484,
+      "grad_norm": 0.6713859438896179,
+      "learning_rate": 4.304719913398438e-06,
+      "loss": 0.5877,
+      "step": 5345
+    },
+    {
+      "epoch": 1.480886426592798,
+      "grad_norm": 0.6959987878799438,
+      "learning_rate": 4.304467805491916e-06,
+      "loss": 0.648,
+      "step": 5346
+    },
+    {
+      "epoch": 1.4811634349030471,
+      "grad_norm": 0.6816545128822327,
+      "learning_rate": 4.304215659271936e-06,
+      "loss": 0.6096,
+      "step": 5347
+    },
+    {
+      "epoch": 1.4814404432132964,
+      "grad_norm": 0.7139344811439514,
+      "learning_rate": 4.303963474743851e-06,
+      "loss": 0.6545,
+      "step": 5348
+    },
+    {
+      "epoch": 1.4817174515235458,
+      "grad_norm": 0.7413231730461121,
+      "learning_rate": 4.303711251913013e-06,
+      "loss": 0.6533,
+      "step": 5349
+    },
+    {
+      "epoch": 1.481994459833795,
+      "grad_norm": 0.6867762804031372,
+      "learning_rate": 4.30345899078478e-06,
+      "loss": 0.6578,
+      "step": 5350
+    },
+    {
+      "epoch": 1.4822714681440443,
+      "grad_norm": 0.7088814377784729,
+      "learning_rate": 4.303206691364507e-06,
+      "loss": 0.6596,
+      "step": 5351
+    },
+    {
+      "epoch": 1.4825484764542936,
+      "grad_norm": 0.7046895027160645,
+      "learning_rate": 4.30295435365755e-06,
+      "loss": 0.6589,
+      "step": 5352
+    },
+    {
+      "epoch": 1.482825484764543,
+      "grad_norm": 0.6934412121772766,
+      "learning_rate": 4.302701977669269e-06,
+      "loss": 0.5886,
+      "step": 5353
+    },
+    {
+      "epoch": 1.4831024930747922,
+      "grad_norm": 0.7157440185546875,
+      "learning_rate": 4.302449563405021e-06,
+      "loss": 0.6188,
+      "step": 5354
+    },
+    {
+      "epoch": 1.4833795013850415,
+      "grad_norm": 0.6679389476776123,
+      "learning_rate": 4.302197110870165e-06,
+      "loss": 0.6092,
+      "step": 5355
+    },
+    {
+      "epoch": 1.4836565096952907,
+      "grad_norm": 0.6910407543182373,
+      "learning_rate": 4.301944620070063e-06,
+      "loss": 0.5916,
+      "step": 5356
+    },
+    {
+      "epoch": 1.4839335180055402,
+      "grad_norm": 0.7141585946083069,
+      "learning_rate": 4.301692091010075e-06,
+      "loss": 0.6066,
+      "step": 5357
+    },
+    {
+      "epoch": 1.4842105263157894,
+      "grad_norm": 0.7129756808280945,
+      "learning_rate": 4.3014395236955635e-06,
+      "loss": 0.6816,
+      "step": 5358
+    },
+    {
+      "epoch": 1.4844875346260387,
+      "grad_norm": 0.7230321168899536,
+      "learning_rate": 4.301186918131889e-06,
+      "loss": 0.5535,
+      "step": 5359
+    },
+    {
+      "epoch": 1.4847645429362881,
+      "grad_norm": 0.6658442616462708,
+      "learning_rate": 4.300934274324417e-06,
+      "loss": 0.6308,
+      "step": 5360
+    },
+    {
+      "epoch": 1.4850415512465374,
+      "grad_norm": 0.7296638488769531,
+      "learning_rate": 4.3006815922785115e-06,
+      "loss": 0.6358,
+      "step": 5361
+    },
+    {
+      "epoch": 1.4853185595567866,
+      "grad_norm": 0.7094091176986694,
+      "learning_rate": 4.300428871999538e-06,
+      "loss": 0.6209,
+      "step": 5362
+    },
+    {
+      "epoch": 1.485595567867036,
+      "grad_norm": 0.669418215751648,
+      "learning_rate": 4.300176113492861e-06,
+      "loss": 0.6032,
+      "step": 5363
+    },
+    {
+      "epoch": 1.4858725761772853,
+      "grad_norm": 0.7072867751121521,
+      "learning_rate": 4.299923316763848e-06,
+      "loss": 0.654,
+      "step": 5364
+    },
+    {
+      "epoch": 1.4861495844875345,
+      "grad_norm": 0.6897730827331543,
+      "learning_rate": 4.299670481817867e-06,
+      "loss": 0.6145,
+      "step": 5365
+    },
+    {
+      "epoch": 1.486426592797784,
+      "grad_norm": 0.6724587678909302,
+      "learning_rate": 4.299417608660285e-06,
+      "loss": 0.6323,
+      "step": 5366
+    },
+    {
+      "epoch": 1.4867036011080332,
+      "grad_norm": 0.6695003509521484,
+      "learning_rate": 4.299164697296473e-06,
+      "loss": 0.6383,
+      "step": 5367
+    },
+    {
+      "epoch": 1.4869806094182825,
+      "grad_norm": 0.6462554931640625,
+      "learning_rate": 4.298911747731798e-06,
+      "loss": 0.5728,
+      "step": 5368
+    },
+    {
+      "epoch": 1.487257617728532,
+      "grad_norm": 0.676162838935852,
+      "learning_rate": 4.298658759971635e-06,
+      "loss": 0.6123,
+      "step": 5369
+    },
+    {
+      "epoch": 1.4875346260387812,
+      "grad_norm": 0.7211152911186218,
+      "learning_rate": 4.298405734021352e-06,
+      "loss": 0.6803,
+      "step": 5370
+    },
+    {
+      "epoch": 1.4878116343490304,
+      "grad_norm": 0.7380316853523254,
+      "learning_rate": 4.298152669886323e-06,
+      "loss": 0.6448,
+      "step": 5371
+    },
+    {
+      "epoch": 1.4880886426592799,
+      "grad_norm": 0.7026437520980835,
+      "learning_rate": 4.297899567571921e-06,
+      "loss": 0.6442,
+      "step": 5372
+    },
+    {
+      "epoch": 1.488365650969529,
+      "grad_norm": 0.6982065439224243,
+      "learning_rate": 4.297646427083519e-06,
+      "loss": 0.6463,
+      "step": 5373
+    },
+    {
+      "epoch": 1.4886426592797783,
+      "grad_norm": 0.6661611199378967,
+      "learning_rate": 4.297393248426494e-06,
+      "loss": 0.6259,
+      "step": 5374
+    },
+    {
+      "epoch": 1.4889196675900278,
+      "grad_norm": 0.7075362801551819,
+      "learning_rate": 4.297140031606218e-06,
+      "loss": 0.6677,
+      "step": 5375
+    },
+    {
+      "epoch": 1.489196675900277,
+      "grad_norm": 0.7266721129417419,
+      "learning_rate": 4.296886776628072e-06,
+      "loss": 0.6782,
+      "step": 5376
+    },
+    {
+      "epoch": 1.4894736842105263,
+      "grad_norm": 0.7158677577972412,
+      "learning_rate": 4.2966334834974295e-06,
+      "loss": 0.6519,
+      "step": 5377
+    },
+    {
+      "epoch": 1.4897506925207757,
+      "grad_norm": 0.6620814204216003,
+      "learning_rate": 4.296380152219671e-06,
+      "loss": 0.6416,
+      "step": 5378
+    },
+    {
+      "epoch": 1.490027700831025,
+      "grad_norm": 0.74250727891922,
+      "learning_rate": 4.2961267828001744e-06,
+      "loss": 0.6586,
+      "step": 5379
+    },
+    {
+      "epoch": 1.4903047091412742,
+      "grad_norm": 0.6609101295471191,
+      "learning_rate": 4.295873375244319e-06,
+      "loss": 0.6372,
+      "step": 5380
+    },
+    {
+      "epoch": 1.4905817174515237,
+      "grad_norm": 0.7089360952377319,
+      "learning_rate": 4.295619929557486e-06,
+      "loss": 0.622,
+      "step": 5381
+    },
+    {
+      "epoch": 1.490858725761773,
+      "grad_norm": 0.6710687875747681,
+      "learning_rate": 4.295366445745056e-06,
+      "loss": 0.6346,
+      "step": 5382
+    },
+    {
+      "epoch": 1.4911357340720222,
+      "grad_norm": 0.7197329998016357,
+      "learning_rate": 4.295112923812412e-06,
+      "loss": 0.6322,
+      "step": 5383
+    },
+    {
+      "epoch": 1.4914127423822714,
+      "grad_norm": 0.7263287901878357,
+      "learning_rate": 4.294859363764936e-06,
+      "loss": 0.6461,
+      "step": 5384
+    },
+    {
+      "epoch": 1.4916897506925209,
+      "grad_norm": 0.6800792217254639,
+      "learning_rate": 4.294605765608013e-06,
+      "loss": 0.5554,
+      "step": 5385
+    },
+    {
+      "epoch": 1.49196675900277,
+      "grad_norm": 0.6973571181297302,
+      "learning_rate": 4.294352129347026e-06,
+      "loss": 0.6044,
+      "step": 5386
+    },
+    {
+      "epoch": 1.4922437673130193,
+      "grad_norm": 0.6802106499671936,
+      "learning_rate": 4.29409845498736e-06,
+      "loss": 0.6012,
+      "step": 5387
+    },
+    {
+      "epoch": 1.4925207756232686,
+      "grad_norm": 0.7250824570655823,
+      "learning_rate": 4.2938447425344034e-06,
+      "loss": 0.664,
+      "step": 5388
+    },
+    {
+      "epoch": 1.492797783933518,
+      "grad_norm": 0.6981179118156433,
+      "learning_rate": 4.293590991993542e-06,
+      "loss": 0.6245,
+      "step": 5389
+    },
+    {
+      "epoch": 1.4930747922437673,
+      "grad_norm": 0.7301192879676819,
+      "learning_rate": 4.293337203370162e-06,
+      "loss": 0.686,
+      "step": 5390
+    },
+    {
+      "epoch": 1.4933518005540165,
+      "grad_norm": 0.716961681842804,
+      "learning_rate": 4.293083376669655e-06,
+      "loss": 0.6187,
+      "step": 5391
+    },
+    {
+      "epoch": 1.493628808864266,
+      "grad_norm": 0.7076943516731262,
+      "learning_rate": 4.292829511897409e-06,
+      "loss": 0.6584,
+      "step": 5392
+    },
+    {
+      "epoch": 1.4939058171745152,
+      "grad_norm": 0.676984965801239,
+      "learning_rate": 4.292575609058814e-06,
+      "loss": 0.6488,
+      "step": 5393
+    },
+    {
+      "epoch": 1.4941828254847644,
+      "grad_norm": 0.6705701947212219,
+      "learning_rate": 4.292321668159261e-06,
+      "loss": 0.5845,
+      "step": 5394
+    },
+    {
+      "epoch": 1.494459833795014,
+      "grad_norm": 0.7139601111412048,
+      "learning_rate": 4.292067689204142e-06,
+      "loss": 0.6151,
+      "step": 5395
+    },
+    {
+      "epoch": 1.4947368421052631,
+      "grad_norm": 0.6521778702735901,
+      "learning_rate": 4.2918136721988475e-06,
+      "loss": 0.5775,
+      "step": 5396
+    },
+    {
+      "epoch": 1.4950138504155124,
+      "grad_norm": 0.7110154032707214,
+      "learning_rate": 4.291559617148775e-06,
+      "loss": 0.6636,
+      "step": 5397
+    },
+    {
+      "epoch": 1.4952908587257618,
+      "grad_norm": 0.672015368938446,
+      "learning_rate": 4.291305524059317e-06,
+      "loss": 0.6439,
+      "step": 5398
+    },
+    {
+      "epoch": 1.495567867036011,
+      "grad_norm": 0.6617342233657837,
+      "learning_rate": 4.2910513929358665e-06,
+      "loss": 0.6578,
+      "step": 5399
+    },
+    {
+      "epoch": 1.4958448753462603,
+      "grad_norm": 0.6488606929779053,
+      "learning_rate": 4.2907972237838225e-06,
+      "loss": 0.6104,
+      "step": 5400
+    },
+    {
+      "epoch": 1.4961218836565098,
+      "grad_norm": 0.7130759358406067,
+      "learning_rate": 4.29054301660858e-06,
+      "loss": 0.6166,
+      "step": 5401
+    },
+    {
+      "epoch": 1.496398891966759,
+      "grad_norm": 0.7276590466499329,
+      "learning_rate": 4.290288771415536e-06,
+      "loss": 0.6528,
+      "step": 5402
+    },
+    {
+      "epoch": 1.4966759002770083,
+      "grad_norm": 0.7055269479751587,
+      "learning_rate": 4.29003448821009e-06,
+      "loss": 0.6317,
+      "step": 5403
+    },
+    {
+      "epoch": 1.4969529085872577,
+      "grad_norm": 0.6901901364326477,
+      "learning_rate": 4.28978016699764e-06,
+      "loss": 0.642,
+      "step": 5404
+    },
+    {
+      "epoch": 1.497229916897507,
+      "grad_norm": 0.7070296406745911,
+      "learning_rate": 4.289525807783588e-06,
+      "loss": 0.6675,
+      "step": 5405
+    },
+    {
+      "epoch": 1.4975069252077562,
+      "grad_norm": 0.6661621332168579,
+      "learning_rate": 4.289271410573332e-06,
+      "loss": 0.6203,
+      "step": 5406
+    },
+    {
+      "epoch": 1.4977839335180057,
+      "grad_norm": 0.6991596221923828,
+      "learning_rate": 4.2890169753722746e-06,
+      "loss": 0.5962,
+      "step": 5407
+    },
+    {
+      "epoch": 1.4980609418282549,
+      "grad_norm": 0.7044360041618347,
+      "learning_rate": 4.288762502185818e-06,
+      "loss": 0.6288,
+      "step": 5408
+    },
+    {
+      "epoch": 1.4983379501385041,
+      "grad_norm": 0.6955025792121887,
+      "learning_rate": 4.288507991019366e-06,
+      "loss": 0.6694,
+      "step": 5409
+    },
+    {
+      "epoch": 1.4986149584487536,
+      "grad_norm": 0.7140067219734192,
+      "learning_rate": 4.2882534418783216e-06,
+      "loss": 0.6491,
+      "step": 5410
+    },
+    {
+      "epoch": 1.4988919667590028,
+      "grad_norm": 0.7064623236656189,
+      "learning_rate": 4.287998854768091e-06,
+      "loss": 0.6259,
+      "step": 5411
+    },
+    {
+      "epoch": 1.499168975069252,
+      "grad_norm": 0.6820512413978577,
+      "learning_rate": 4.287744229694077e-06,
+      "loss": 0.6292,
+      "step": 5412
+    },
+    {
+      "epoch": 1.4994459833795015,
+      "grad_norm": 0.7026885747909546,
+      "learning_rate": 4.287489566661689e-06,
+      "loss": 0.6626,
+      "step": 5413
+    },
+    {
+      "epoch": 1.4997229916897508,
+      "grad_norm": 0.6979086995124817,
+      "learning_rate": 4.287234865676332e-06,
+      "loss": 0.6331,
+      "step": 5414
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6782578825950623,
+      "learning_rate": 4.2869801267434145e-06,
+      "loss": 0.6163,
+      "step": 5415
+    },
+    {
+      "epoch": 1.5002770083102495,
+      "grad_norm": 0.7158074378967285,
+      "learning_rate": 4.2867253498683455e-06,
+      "loss": 0.6391,
+      "step": 5416
+    },
+    {
+      "epoch": 1.5005540166204985,
+      "grad_norm": 0.703403115272522,
+      "learning_rate": 4.286470535056536e-06,
+      "loss": 0.6348,
+      "step": 5417
+    },
+    {
+      "epoch": 1.500831024930748,
+      "grad_norm": 0.6722911596298218,
+      "learning_rate": 4.286215682313393e-06,
+      "loss": 0.6328,
+      "step": 5418
+    },
+    {
+      "epoch": 1.5011080332409972,
+      "grad_norm": 0.7120633125305176,
+      "learning_rate": 4.28596079164433e-06,
+      "loss": 0.6293,
+      "step": 5419
+    },
+    {
+      "epoch": 1.5013850415512464,
+      "grad_norm": 0.692536473274231,
+      "learning_rate": 4.285705863054759e-06,
+      "loss": 0.6305,
+      "step": 5420
+    },
+    {
+      "epoch": 1.5016620498614959,
+      "grad_norm": 0.685097873210907,
+      "learning_rate": 4.285450896550093e-06,
+      "loss": 0.6588,
+      "step": 5421
+    },
+    {
+      "epoch": 1.5019390581717451,
+      "grad_norm": 0.6913133263587952,
+      "learning_rate": 4.285195892135744e-06,
+      "loss": 0.6222,
+      "step": 5422
+    },
+    {
+      "epoch": 1.5022160664819943,
+      "grad_norm": 0.7574515342712402,
+      "learning_rate": 4.284940849817127e-06,
+      "loss": 0.6058,
+      "step": 5423
+    },
+    {
+      "epoch": 1.5024930747922438,
+      "grad_norm": 0.7188212871551514,
+      "learning_rate": 4.284685769599658e-06,
+      "loss": 0.6449,
+      "step": 5424
+    },
+    {
+      "epoch": 1.502770083102493,
+      "grad_norm": 0.7023669481277466,
+      "learning_rate": 4.284430651488753e-06,
+      "loss": 0.6061,
+      "step": 5425
+    },
+    {
+      "epoch": 1.5030470914127423,
+      "grad_norm": 0.6833453178405762,
+      "learning_rate": 4.2841754954898274e-06,
+      "loss": 0.6079,
+      "step": 5426
+    },
+    {
+      "epoch": 1.5033240997229917,
+      "grad_norm": 0.7508666515350342,
+      "learning_rate": 4.2839203016083e-06,
+      "loss": 0.6261,
+      "step": 5427
+    },
+    {
+      "epoch": 1.503601108033241,
+      "grad_norm": 0.6877349615097046,
+      "learning_rate": 4.283665069849589e-06,
+      "loss": 0.6221,
+      "step": 5428
+    },
+    {
+      "epoch": 1.5038781163434902,
+      "grad_norm": 0.6670334339141846,
+      "learning_rate": 4.283409800219114e-06,
+      "loss": 0.5992,
+      "step": 5429
+    },
+    {
+      "epoch": 1.5041551246537397,
+      "grad_norm": 0.7320896983146667,
+      "learning_rate": 4.283154492722294e-06,
+      "loss": 0.6989,
+      "step": 5430
+    },
+    {
+      "epoch": 1.504432132963989,
+      "grad_norm": 0.6807078719139099,
+      "learning_rate": 4.282899147364551e-06,
+      "loss": 0.6386,
+      "step": 5431
+    },
+    {
+      "epoch": 1.5047091412742382,
+      "grad_norm": 0.6722118854522705,
+      "learning_rate": 4.282643764151306e-06,
+      "loss": 0.5941,
+      "step": 5432
+    },
+    {
+      "epoch": 1.5049861495844876,
+      "grad_norm": 0.6991921663284302,
+      "learning_rate": 4.282388343087981e-06,
+      "loss": 0.6243,
+      "step": 5433
+    },
+    {
+      "epoch": 1.5052631578947369,
+      "grad_norm": 0.7761150002479553,
+      "learning_rate": 4.28213288418e-06,
+      "loss": 0.6775,
+      "step": 5434
+    },
+    {
+      "epoch": 1.505540166204986,
+      "grad_norm": 0.7118150591850281,
+      "learning_rate": 4.281877387432787e-06,
+      "loss": 0.6609,
+      "step": 5435
+    },
+    {
+      "epoch": 1.5058171745152356,
+      "grad_norm": 0.6934765577316284,
+      "learning_rate": 4.281621852851766e-06,
+      "loss": 0.6241,
+      "step": 5436
+    },
+    {
+      "epoch": 1.5060941828254848,
+      "grad_norm": 0.7331833839416504,
+      "learning_rate": 4.2813662804423645e-06,
+      "loss": 0.6768,
+      "step": 5437
+    },
+    {
+      "epoch": 1.506371191135734,
+      "grad_norm": 0.7009062767028809,
+      "learning_rate": 4.281110670210007e-06,
+      "loss": 0.6504,
+      "step": 5438
+    },
+    {
+      "epoch": 1.5066481994459835,
+      "grad_norm": 0.668209433555603,
+      "learning_rate": 4.280855022160121e-06,
+      "loss": 0.5986,
+      "step": 5439
+    },
+    {
+      "epoch": 1.5069252077562327,
+      "grad_norm": 0.6664574146270752,
+      "learning_rate": 4.2805993362981355e-06,
+      "loss": 0.5815,
+      "step": 5440
+    },
+    {
+      "epoch": 1.507202216066482,
+      "grad_norm": 0.6569491028785706,
+      "learning_rate": 4.28034361262948e-06,
+      "loss": 0.6364,
+      "step": 5441
+    },
+    {
+      "epoch": 1.5074792243767314,
+      "grad_norm": 0.6996865272521973,
+      "learning_rate": 4.280087851159582e-06,
+      "loss": 0.6371,
+      "step": 5442
+    },
+    {
+      "epoch": 1.5077562326869804,
+      "grad_norm": 0.7376276254653931,
+      "learning_rate": 4.279832051893873e-06,
+      "loss": 0.6692,
+      "step": 5443
+    },
+    {
+      "epoch": 1.50803324099723,
+      "grad_norm": 0.7286374568939209,
+      "learning_rate": 4.279576214837784e-06,
+      "loss": 0.6352,
+      "step": 5444
+    },
+    {
+      "epoch": 1.5083102493074794,
+      "grad_norm": 0.6960650682449341,
+      "learning_rate": 4.279320339996749e-06,
+      "loss": 0.5829,
+      "step": 5445
+    },
+    {
+      "epoch": 1.5085872576177284,
+      "grad_norm": 0.6842361688613892,
+      "learning_rate": 4.279064427376199e-06,
+      "loss": 0.6139,
+      "step": 5446
+    },
+    {
+      "epoch": 1.5088642659279778,
+      "grad_norm": 0.6859242916107178,
+      "learning_rate": 4.278808476981567e-06,
+      "loss": 0.6111,
+      "step": 5447
+    },
+    {
+      "epoch": 1.5091412742382273,
+      "grad_norm": 0.7144955992698669,
+      "learning_rate": 4.2785524888182885e-06,
+      "loss": 0.6538,
+      "step": 5448
+    },
+    {
+      "epoch": 1.5094182825484763,
+      "grad_norm": 0.7113151550292969,
+      "learning_rate": 4.278296462891798e-06,
+      "loss": 0.6355,
+      "step": 5449
+    },
+    {
+      "epoch": 1.5096952908587258,
+      "grad_norm": 0.7035449147224426,
+      "learning_rate": 4.278040399207533e-06,
+      "loss": 0.6406,
+      "step": 5450
+    },
+    {
+      "epoch": 1.509972299168975,
+      "grad_norm": 0.6973490118980408,
+      "learning_rate": 4.2777842977709305e-06,
+      "loss": 0.6132,
+      "step": 5451
+    },
+    {
+      "epoch": 1.5102493074792243,
+      "grad_norm": 0.7446228861808777,
+      "learning_rate": 4.2775281585874275e-06,
+      "loss": 0.6287,
+      "step": 5452
+    },
+    {
+      "epoch": 1.5105263157894737,
+      "grad_norm": 0.7096726894378662,
+      "learning_rate": 4.2772719816624615e-06,
+      "loss": 0.6358,
+      "step": 5453
+    },
+    {
+      "epoch": 1.510803324099723,
+      "grad_norm": 0.6858077049255371,
+      "learning_rate": 4.277015767001473e-06,
+      "loss": 0.6146,
+      "step": 5454
+    },
+    {
+      "epoch": 1.5110803324099722,
+      "grad_norm": 0.6863676905632019,
+      "learning_rate": 4.276759514609902e-06,
+      "loss": 0.6236,
+      "step": 5455
+    },
+    {
+      "epoch": 1.5113573407202217,
+      "grad_norm": 0.6914564967155457,
+      "learning_rate": 4.276503224493188e-06,
+      "loss": 0.6244,
+      "step": 5456
+    },
+    {
+      "epoch": 1.511634349030471,
+      "grad_norm": 0.6845644116401672,
+      "learning_rate": 4.276246896656775e-06,
+      "loss": 0.642,
+      "step": 5457
+    },
+    {
+      "epoch": 1.5119113573407201,
+      "grad_norm": 0.6796553134918213,
+      "learning_rate": 4.275990531106104e-06,
+      "loss": 0.648,
+      "step": 5458
+    },
+    {
+      "epoch": 1.5121883656509696,
+      "grad_norm": 0.6412548422813416,
+      "learning_rate": 4.275734127846619e-06,
+      "loss": 0.5683,
+      "step": 5459
+    },
+    {
+      "epoch": 1.5124653739612188,
+      "grad_norm": 0.725028932094574,
+      "learning_rate": 4.275477686883763e-06,
+      "loss": 0.6299,
+      "step": 5460
+    },
+    {
+      "epoch": 1.512742382271468,
+      "grad_norm": 0.7068825960159302,
+      "learning_rate": 4.275221208222982e-06,
+      "loss": 0.6453,
+      "step": 5461
+    },
+    {
+      "epoch": 1.5130193905817175,
+      "grad_norm": 0.7068927884101868,
+      "learning_rate": 4.274964691869722e-06,
+      "loss": 0.6541,
+      "step": 5462
+    },
+    {
+      "epoch": 1.5132963988919668,
+      "grad_norm": 0.7255887985229492,
+      "learning_rate": 4.2747081378294275e-06,
+      "loss": 0.6562,
+      "step": 5463
+    },
+    {
+      "epoch": 1.513573407202216,
+      "grad_norm": 0.7126734852790833,
+      "learning_rate": 4.274451546107548e-06,
+      "loss": 0.6605,
+      "step": 5464
+    },
+    {
+      "epoch": 1.5138504155124655,
+      "grad_norm": 0.6980812549591064,
+      "learning_rate": 4.274194916709531e-06,
+      "loss": 0.623,
+      "step": 5465
+    },
+    {
+      "epoch": 1.5141274238227147,
+      "grad_norm": 0.6746025681495667,
+      "learning_rate": 4.2739382496408245e-06,
+      "loss": 0.5671,
+      "step": 5466
+    },
+    {
+      "epoch": 1.514404432132964,
+      "grad_norm": 0.7552827000617981,
+      "learning_rate": 4.2736815449068794e-06,
+      "loss": 0.6663,
+      "step": 5467
+    },
+    {
+      "epoch": 1.5146814404432134,
+      "grad_norm": 0.704200267791748,
+      "learning_rate": 4.273424802513145e-06,
+      "loss": 0.6326,
+      "step": 5468
+    },
+    {
+      "epoch": 1.5149584487534626,
+      "grad_norm": 0.7429570555686951,
+      "learning_rate": 4.273168022465073e-06,
+      "loss": 0.6494,
+      "step": 5469
+    },
+    {
+      "epoch": 1.5152354570637119,
+      "grad_norm": 0.7143659591674805,
+      "learning_rate": 4.272911204768116e-06,
+      "loss": 0.6555,
+      "step": 5470
+    },
+    {
+      "epoch": 1.5155124653739613,
+      "grad_norm": 0.6867467164993286,
+      "learning_rate": 4.272654349427726e-06,
+      "loss": 0.6375,
+      "step": 5471
+    },
+    {
+      "epoch": 1.5157894736842106,
+      "grad_norm": 0.7136903405189514,
+      "learning_rate": 4.272397456449359e-06,
+      "loss": 0.6331,
+      "step": 5472
+    },
+    {
+      "epoch": 1.5160664819944598,
+      "grad_norm": 0.7227846384048462,
+      "learning_rate": 4.2721405258384666e-06,
+      "loss": 0.6297,
+      "step": 5473
+    },
+    {
+      "epoch": 1.5163434903047093,
+      "grad_norm": 0.717075526714325,
+      "learning_rate": 4.271883557600506e-06,
+      "loss": 0.6408,
+      "step": 5474
+    },
+    {
+      "epoch": 1.5166204986149583,
+      "grad_norm": 0.6881066560745239,
+      "learning_rate": 4.271626551740932e-06,
+      "loss": 0.6404,
+      "step": 5475
+    },
+    {
+      "epoch": 1.5168975069252078,
+      "grad_norm": 0.6838085651397705,
+      "learning_rate": 4.271369508265202e-06,
+      "loss": 0.6611,
+      "step": 5476
+    },
+    {
+      "epoch": 1.5171745152354572,
+      "grad_norm": 0.6700277924537659,
+      "learning_rate": 4.271112427178774e-06,
+      "loss": 0.6043,
+      "step": 5477
+    },
+    {
+      "epoch": 1.5174515235457062,
+      "grad_norm": 0.6892854571342468,
+      "learning_rate": 4.270855308487106e-06,
+      "loss": 0.6272,
+      "step": 5478
+    },
+    {
+      "epoch": 1.5177285318559557,
+      "grad_norm": 0.6893463730812073,
+      "learning_rate": 4.270598152195657e-06,
+      "loss": 0.6337,
+      "step": 5479
+    },
+    {
+      "epoch": 1.5180055401662051,
+      "grad_norm": 0.696556568145752,
+      "learning_rate": 4.270340958309889e-06,
+      "loss": 0.6119,
+      "step": 5480
+    },
+    {
+      "epoch": 1.5182825484764542,
+      "grad_norm": 0.6970028877258301,
+      "learning_rate": 4.2700837268352605e-06,
+      "loss": 0.6339,
+      "step": 5481
+    },
+    {
+      "epoch": 1.5185595567867036,
+      "grad_norm": 0.6837161779403687,
+      "learning_rate": 4.269826457777234e-06,
+      "loss": 0.6501,
+      "step": 5482
+    },
+    {
+      "epoch": 1.5188365650969529,
+      "grad_norm": 0.7038562297821045,
+      "learning_rate": 4.2695691511412726e-06,
+      "loss": 0.6385,
+      "step": 5483
+    },
+    {
+      "epoch": 1.519113573407202,
+      "grad_norm": 0.7245030403137207,
+      "learning_rate": 4.26931180693284e-06,
+      "loss": 0.6326,
+      "step": 5484
+    },
+    {
+      "epoch": 1.5193905817174516,
+      "grad_norm": 0.7005324959754944,
+      "learning_rate": 4.269054425157398e-06,
+      "loss": 0.6136,
+      "step": 5485
+    },
+    {
+      "epoch": 1.5196675900277008,
+      "grad_norm": 0.7006979584693909,
+      "learning_rate": 4.268797005820413e-06,
+      "loss": 0.5971,
+      "step": 5486
+    },
+    {
+      "epoch": 1.51994459833795,
+      "grad_norm": 0.7217789888381958,
+      "learning_rate": 4.268539548927351e-06,
+      "loss": 0.666,
+      "step": 5487
+    },
+    {
+      "epoch": 1.5202216066481995,
+      "grad_norm": 0.6616018414497375,
+      "learning_rate": 4.2682820544836785e-06,
+      "loss": 0.5997,
+      "step": 5488
+    },
+    {
+      "epoch": 1.5204986149584487,
+      "grad_norm": 0.6509618759155273,
+      "learning_rate": 4.2680245224948615e-06,
+      "loss": 0.6043,
+      "step": 5489
+    },
+    {
+      "epoch": 1.520775623268698,
+      "grad_norm": 0.7281323075294495,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.6311,
+      "step": 5490
+    },
+    {
+      "epoch": 1.5210526315789474,
+      "grad_norm": 0.6721056699752808,
+      "learning_rate": 4.2675093459036695e-06,
+      "loss": 0.6389,
+      "step": 5491
+    },
+    {
+      "epoch": 1.5213296398891967,
+      "grad_norm": 0.7493563890457153,
+      "learning_rate": 4.267251701312234e-06,
+      "loss": 0.5979,
+      "step": 5492
+    },
+    {
+      "epoch": 1.521606648199446,
+      "grad_norm": 0.6652003526687622,
+      "learning_rate": 4.266994019197531e-06,
+      "loss": 0.6315,
+      "step": 5493
+    },
+    {
+      "epoch": 1.5218836565096954,
+      "grad_norm": 0.713382363319397,
+      "learning_rate": 4.266736299565032e-06,
+      "loss": 0.6358,
+      "step": 5494
+    },
+    {
+      "epoch": 1.5221606648199446,
+      "grad_norm": 0.6911031007766724,
+      "learning_rate": 4.2664785424202105e-06,
+      "loss": 0.682,
+      "step": 5495
+    },
+    {
+      "epoch": 1.5224376731301938,
+      "grad_norm": 0.6979799270629883,
+      "learning_rate": 4.266220747768538e-06,
+      "loss": 0.6399,
+      "step": 5496
+    },
+    {
+      "epoch": 1.5227146814404433,
+      "grad_norm": 0.6786274313926697,
+      "learning_rate": 4.2659629156154894e-06,
+      "loss": 0.5986,
+      "step": 5497
+    },
+    {
+      "epoch": 1.5229916897506925,
+      "grad_norm": 0.6834037899971008,
+      "learning_rate": 4.2657050459665375e-06,
+      "loss": 0.6497,
+      "step": 5498
+    },
+    {
+      "epoch": 1.5232686980609418,
+      "grad_norm": 0.6988540887832642,
+      "learning_rate": 4.265447138827158e-06,
+      "loss": 0.6382,
+      "step": 5499
+    },
+    {
+      "epoch": 1.5235457063711912,
+      "grad_norm": 0.6756704449653625,
+      "learning_rate": 4.265189194202828e-06,
+      "loss": 0.62,
+      "step": 5500
+    },
+    {
+      "epoch": 1.5238227146814405,
+      "grad_norm": 0.6644415855407715,
+      "learning_rate": 4.264931212099023e-06,
+      "loss": 0.5966,
+      "step": 5501
+    },
+    {
+      "epoch": 1.5240997229916897,
+      "grad_norm": 0.704603910446167,
+      "learning_rate": 4.264673192521221e-06,
+      "loss": 0.6487,
+      "step": 5502
+    },
+    {
+      "epoch": 1.5243767313019392,
+      "grad_norm": 0.6660934090614319,
+      "learning_rate": 4.264415135474902e-06,
+      "loss": 0.6117,
+      "step": 5503
+    },
+    {
+      "epoch": 1.5246537396121884,
+      "grad_norm": 0.658829391002655,
+      "learning_rate": 4.264157040965543e-06,
+      "loss": 0.6567,
+      "step": 5504
+    },
+    {
+      "epoch": 1.5249307479224377,
+      "grad_norm": 0.738782525062561,
+      "learning_rate": 4.2638989089986245e-06,
+      "loss": 0.6261,
+      "step": 5505
+    },
+    {
+      "epoch": 1.5252077562326871,
+      "grad_norm": 0.7513110041618347,
+      "learning_rate": 4.263640739579629e-06,
+      "loss": 0.6354,
+      "step": 5506
+    },
+    {
+      "epoch": 1.5254847645429361,
+      "grad_norm": 0.6777371168136597,
+      "learning_rate": 4.2633825327140355e-06,
+      "loss": 0.6373,
+      "step": 5507
+    },
+    {
+      "epoch": 1.5257617728531856,
+      "grad_norm": 0.6883520483970642,
+      "learning_rate": 4.263124288407328e-06,
+      "loss": 0.628,
+      "step": 5508
+    },
+    {
+      "epoch": 1.526038781163435,
+      "grad_norm": 0.7177495956420898,
+      "learning_rate": 4.262866006664989e-06,
+      "loss": 0.6231,
+      "step": 5509
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.7112988233566284,
+      "learning_rate": 4.262607687492503e-06,
+      "loss": 0.6436,
+      "step": 5510
+    },
+    {
+      "epoch": 1.5265927977839335,
+      "grad_norm": 0.6998488306999207,
+      "learning_rate": 4.262349330895354e-06,
+      "loss": 0.5978,
+      "step": 5511
+    },
+    {
+      "epoch": 1.526869806094183,
+      "grad_norm": 0.7161896824836731,
+      "learning_rate": 4.26209093687903e-06,
+      "loss": 0.6391,
+      "step": 5512
+    },
+    {
+      "epoch": 1.527146814404432,
+      "grad_norm": 0.6859282851219177,
+      "learning_rate": 4.261832505449014e-06,
+      "loss": 0.6053,
+      "step": 5513
+    },
+    {
+      "epoch": 1.5274238227146815,
+      "grad_norm": 0.6807940006256104,
+      "learning_rate": 4.261574036610795e-06,
+      "loss": 0.6584,
+      "step": 5514
+    },
+    {
+      "epoch": 1.5277008310249307,
+      "grad_norm": 0.710959255695343,
+      "learning_rate": 4.261315530369861e-06,
+      "loss": 0.6322,
+      "step": 5515
+    },
+    {
+      "epoch": 1.52797783933518,
+      "grad_norm": 0.7400763034820557,
+      "learning_rate": 4.2610569867316994e-06,
+      "loss": 0.6589,
+      "step": 5516
+    },
+    {
+      "epoch": 1.5282548476454294,
+      "grad_norm": 0.7326130867004395,
+      "learning_rate": 4.260798405701802e-06,
+      "loss": 0.6851,
+      "step": 5517
+    },
+    {
+      "epoch": 1.5285318559556786,
+      "grad_norm": 0.6840008497238159,
+      "learning_rate": 4.2605397872856565e-06,
+      "loss": 0.6138,
+      "step": 5518
+    },
+    {
+      "epoch": 1.5288088642659279,
+      "grad_norm": 0.6635662317276001,
+      "learning_rate": 4.260281131488757e-06,
+      "loss": 0.6023,
+      "step": 5519
+    },
+    {
+      "epoch": 1.5290858725761773,
+      "grad_norm": 0.706579864025116,
+      "learning_rate": 4.260022438316593e-06,
+      "loss": 0.6307,
+      "step": 5520
+    },
+    {
+      "epoch": 1.5293628808864266,
+      "grad_norm": 0.7006165981292725,
+      "learning_rate": 4.259763707774658e-06,
+      "loss": 0.6226,
+      "step": 5521
+    },
+    {
+      "epoch": 1.5296398891966758,
+      "grad_norm": 0.6810389757156372,
+      "learning_rate": 4.259504939868445e-06,
+      "loss": 0.6825,
+      "step": 5522
+    },
+    {
+      "epoch": 1.5299168975069253,
+      "grad_norm": 0.6524600982666016,
+      "learning_rate": 4.25924613460345e-06,
+      "loss": 0.598,
+      "step": 5523
+    },
+    {
+      "epoch": 1.5301939058171745,
+      "grad_norm": 0.714477002620697,
+      "learning_rate": 4.258987291985166e-06,
+      "loss": 0.6285,
+      "step": 5524
+    },
+    {
+      "epoch": 1.5304709141274238,
+      "grad_norm": 0.6839551329612732,
+      "learning_rate": 4.258728412019089e-06,
+      "loss": 0.5905,
+      "step": 5525
+    },
+    {
+      "epoch": 1.5307479224376732,
+      "grad_norm": 0.695607602596283,
+      "learning_rate": 4.258469494710719e-06,
+      "loss": 0.5707,
+      "step": 5526
+    },
+    {
+      "epoch": 1.5310249307479225,
+      "grad_norm": 0.7326964139938354,
+      "learning_rate": 4.258210540065549e-06,
+      "loss": 0.6337,
+      "step": 5527
+    },
+    {
+      "epoch": 1.5313019390581717,
+      "grad_norm": 0.7270690202713013,
+      "learning_rate": 4.2579515480890805e-06,
+      "loss": 0.6443,
+      "step": 5528
+    },
+    {
+      "epoch": 1.5315789473684212,
+      "grad_norm": 0.7199828028678894,
+      "learning_rate": 4.25769251878681e-06,
+      "loss": 0.6347,
+      "step": 5529
+    },
+    {
+      "epoch": 1.5318559556786704,
+      "grad_norm": 0.7613186240196228,
+      "learning_rate": 4.2574334521642404e-06,
+      "loss": 0.6838,
+      "step": 5530
+    },
+    {
+      "epoch": 1.5321329639889196,
+      "grad_norm": 0.7245945930480957,
+      "learning_rate": 4.257174348226869e-06,
+      "loss": 0.644,
+      "step": 5531
+    },
+    {
+      "epoch": 1.532409972299169,
+      "grad_norm": 0.7069911360740662,
+      "learning_rate": 4.256915206980199e-06,
+      "loss": 0.6928,
+      "step": 5532
+    },
+    {
+      "epoch": 1.5326869806094183,
+      "grad_norm": 0.7057062387466431,
+      "learning_rate": 4.256656028429733e-06,
+      "loss": 0.628,
+      "step": 5533
+    },
+    {
+      "epoch": 1.5329639889196676,
+      "grad_norm": 0.7141075134277344,
+      "learning_rate": 4.2563968125809734e-06,
+      "loss": 0.6596,
+      "step": 5534
+    },
+    {
+      "epoch": 1.533240997229917,
+      "grad_norm": 0.6988198757171631,
+      "learning_rate": 4.256137559439424e-06,
+      "loss": 0.6343,
+      "step": 5535
+    },
+    {
+      "epoch": 1.533518005540166,
+      "grad_norm": 0.6692717671394348,
+      "learning_rate": 4.25587826901059e-06,
+      "loss": 0.6166,
+      "step": 5536
+    },
+    {
+      "epoch": 1.5337950138504155,
+      "grad_norm": 0.6966196894645691,
+      "learning_rate": 4.255618941299976e-06,
+      "loss": 0.607,
+      "step": 5537
+    },
+    {
+      "epoch": 1.534072022160665,
+      "grad_norm": 0.6670767664909363,
+      "learning_rate": 4.255359576313089e-06,
+      "loss": 0.5969,
+      "step": 5538
+    },
+    {
+      "epoch": 1.534349030470914,
+      "grad_norm": 0.6590955257415771,
+      "learning_rate": 4.255100174055434e-06,
+      "loss": 0.6277,
+      "step": 5539
+    },
+    {
+      "epoch": 1.5346260387811634,
+      "grad_norm": 0.6713624596595764,
+      "learning_rate": 4.254840734532521e-06,
+      "loss": 0.6466,
+      "step": 5540
+    },
+    {
+      "epoch": 1.534903047091413,
+      "grad_norm": 0.8212193250656128,
+      "learning_rate": 4.254581257749859e-06,
+      "loss": 0.6556,
+      "step": 5541
+    },
+    {
+      "epoch": 1.535180055401662,
+      "grad_norm": 0.7253426313400269,
+      "learning_rate": 4.254321743712955e-06,
+      "loss": 0.6175,
+      "step": 5542
+    },
+    {
+      "epoch": 1.5354570637119114,
+      "grad_norm": 0.6754900813102722,
+      "learning_rate": 4.254062192427321e-06,
+      "loss": 0.6497,
+      "step": 5543
+    },
+    {
+      "epoch": 1.5357340720221606,
+      "grad_norm": 0.6963987350463867,
+      "learning_rate": 4.2538026038984665e-06,
+      "loss": 0.6163,
+      "step": 5544
+    },
+    {
+      "epoch": 1.5360110803324099,
+      "grad_norm": 0.7260979413986206,
+      "learning_rate": 4.253542978131904e-06,
+      "loss": 0.6287,
+      "step": 5545
+    },
+    {
+      "epoch": 1.5362880886426593,
+      "grad_norm": 0.7109147906303406,
+      "learning_rate": 4.253283315133146e-06,
+      "loss": 0.658,
+      "step": 5546
+    },
+    {
+      "epoch": 1.5365650969529085,
+      "grad_norm": 0.7043456435203552,
+      "learning_rate": 4.253023614907707e-06,
+      "loss": 0.6204,
+      "step": 5547
+    },
+    {
+      "epoch": 1.5368421052631578,
+      "grad_norm": 0.6898895502090454,
+      "learning_rate": 4.252763877461099e-06,
+      "loss": 0.6093,
+      "step": 5548
+    },
+    {
+      "epoch": 1.5371191135734072,
+      "grad_norm": 0.7038882374763489,
+      "learning_rate": 4.252504102798838e-06,
+      "loss": 0.6297,
+      "step": 5549
+    },
+    {
+      "epoch": 1.5373961218836565,
+      "grad_norm": 0.7062937021255493,
+      "learning_rate": 4.25224429092644e-06,
+      "loss": 0.6274,
+      "step": 5550
+    },
+    {
+      "epoch": 1.5376731301939057,
+      "grad_norm": 0.7120965719223022,
+      "learning_rate": 4.25198444184942e-06,
+      "loss": 0.6055,
+      "step": 5551
+    },
+    {
+      "epoch": 1.5379501385041552,
+      "grad_norm": 0.7174949645996094,
+      "learning_rate": 4.251724555573297e-06,
+      "loss": 0.6823,
+      "step": 5552
+    },
+    {
+      "epoch": 1.5382271468144044,
+      "grad_norm": 0.7251701354980469,
+      "learning_rate": 4.251464632103587e-06,
+      "loss": 0.6277,
+      "step": 5553
+    },
+    {
+      "epoch": 1.5385041551246537,
+      "grad_norm": 0.7451249361038208,
+      "learning_rate": 4.251204671445811e-06,
+      "loss": 0.6663,
+      "step": 5554
+    },
+    {
+      "epoch": 1.5387811634349031,
+      "grad_norm": 0.72591632604599,
+      "learning_rate": 4.250944673605488e-06,
+      "loss": 0.6495,
+      "step": 5555
+    },
+    {
+      "epoch": 1.5390581717451524,
+      "grad_norm": 0.6595091223716736,
+      "learning_rate": 4.2506846385881375e-06,
+      "loss": 0.6072,
+      "step": 5556
+    },
+    {
+      "epoch": 1.5393351800554016,
+      "grad_norm": 0.705683171749115,
+      "learning_rate": 4.250424566399282e-06,
+      "loss": 0.6637,
+      "step": 5557
+    },
+    {
+      "epoch": 1.539612188365651,
+      "grad_norm": 0.6721247434616089,
+      "learning_rate": 4.250164457044443e-06,
+      "loss": 0.6715,
+      "step": 5558
+    },
+    {
+      "epoch": 1.5398891966759003,
+      "grad_norm": 0.6472792029380798,
+      "learning_rate": 4.249904310529143e-06,
+      "loss": 0.6056,
+      "step": 5559
+    },
+    {
+      "epoch": 1.5401662049861495,
+      "grad_norm": 0.7078953385353088,
+      "learning_rate": 4.249644126858905e-06,
+      "loss": 0.6375,
+      "step": 5560
+    },
+    {
+      "epoch": 1.540443213296399,
+      "grad_norm": 0.6639420986175537,
+      "learning_rate": 4.249383906039254e-06,
+      "loss": 0.6428,
+      "step": 5561
+    },
+    {
+      "epoch": 1.5407202216066482,
+      "grad_norm": 0.7171564698219299,
+      "learning_rate": 4.249123648075716e-06,
+      "loss": 0.5995,
+      "step": 5562
+    },
+    {
+      "epoch": 1.5409972299168975,
+      "grad_norm": 0.7072045803070068,
+      "learning_rate": 4.248863352973817e-06,
+      "loss": 0.634,
+      "step": 5563
+    },
+    {
+      "epoch": 1.541274238227147,
+      "grad_norm": 0.6976851224899292,
+      "learning_rate": 4.248603020739081e-06,
+      "loss": 0.6055,
+      "step": 5564
+    },
+    {
+      "epoch": 1.5415512465373962,
+      "grad_norm": 0.6449586153030396,
+      "learning_rate": 4.248342651377038e-06,
+      "loss": 0.6068,
+      "step": 5565
+    },
+    {
+      "epoch": 1.5418282548476454,
+      "grad_norm": 0.7084159851074219,
+      "learning_rate": 4.248082244893216e-06,
+      "loss": 0.6633,
+      "step": 5566
+    },
+    {
+      "epoch": 1.5421052631578949,
+      "grad_norm": 0.6994836926460266,
+      "learning_rate": 4.247821801293144e-06,
+      "loss": 0.613,
+      "step": 5567
+    },
+    {
+      "epoch": 1.5423822714681439,
+      "grad_norm": 0.7218542695045471,
+      "learning_rate": 4.247561320582352e-06,
+      "loss": 0.6203,
+      "step": 5568
+    },
+    {
+      "epoch": 1.5426592797783933,
+      "grad_norm": 0.6681088805198669,
+      "learning_rate": 4.24730080276637e-06,
+      "loss": 0.6178,
+      "step": 5569
+    },
+    {
+      "epoch": 1.5429362880886428,
+      "grad_norm": 0.6878266334533691,
+      "learning_rate": 4.247040247850729e-06,
+      "loss": 0.615,
+      "step": 5570
+    },
+    {
+      "epoch": 1.5432132963988918,
+      "grad_norm": 0.6583474278450012,
+      "learning_rate": 4.246779655840963e-06,
+      "loss": 0.5656,
+      "step": 5571
+    },
+    {
+      "epoch": 1.5434903047091413,
+      "grad_norm": 0.6586458086967468,
+      "learning_rate": 4.246519026742604e-06,
+      "loss": 0.5824,
+      "step": 5572
+    },
+    {
+      "epoch": 1.5437673130193907,
+      "grad_norm": 0.7186761498451233,
+      "learning_rate": 4.2462583605611865e-06,
+      "loss": 0.6897,
+      "step": 5573
+    },
+    {
+      "epoch": 1.5440443213296398,
+      "grad_norm": 0.6883975267410278,
+      "learning_rate": 4.245997657302243e-06,
+      "loss": 0.5985,
+      "step": 5574
+    },
+    {
+      "epoch": 1.5443213296398892,
+      "grad_norm": 0.7059137225151062,
+      "learning_rate": 4.245736916971312e-06,
+      "loss": 0.6583,
+      "step": 5575
+    },
+    {
+      "epoch": 1.5445983379501385,
+      "grad_norm": 0.7427789568901062,
+      "learning_rate": 4.245476139573928e-06,
+      "loss": 0.6418,
+      "step": 5576
+    },
+    {
+      "epoch": 1.5448753462603877,
+      "grad_norm": 0.708476185798645,
+      "learning_rate": 4.245215325115627e-06,
+      "loss": 0.5888,
+      "step": 5577
+    },
+    {
+      "epoch": 1.5451523545706372,
+      "grad_norm": 0.6729523539543152,
+      "learning_rate": 4.2449544736019486e-06,
+      "loss": 0.649,
+      "step": 5578
+    },
+    {
+      "epoch": 1.5454293628808864,
+      "grad_norm": 0.6773388385772705,
+      "learning_rate": 4.24469358503843e-06,
+      "loss": 0.6112,
+      "step": 5579
+    },
+    {
+      "epoch": 1.5457063711911356,
+      "grad_norm": 0.7275780439376831,
+      "learning_rate": 4.244432659430612e-06,
+      "loss": 0.6699,
+      "step": 5580
+    },
+    {
+      "epoch": 1.545983379501385,
+      "grad_norm": 0.7121980786323547,
+      "learning_rate": 4.244171696784033e-06,
+      "loss": 0.6851,
+      "step": 5581
+    },
+    {
+      "epoch": 1.5462603878116343,
+      "grad_norm": 0.6924628019332886,
+      "learning_rate": 4.243910697104235e-06,
+      "loss": 0.6616,
+      "step": 5582
+    },
+    {
+      "epoch": 1.5465373961218836,
+      "grad_norm": 0.6828587651252747,
+      "learning_rate": 4.24364966039676e-06,
+      "loss": 0.6252,
+      "step": 5583
+    },
+    {
+      "epoch": 1.546814404432133,
+      "grad_norm": 0.688407301902771,
+      "learning_rate": 4.24338858666715e-06,
+      "loss": 0.6092,
+      "step": 5584
+    },
+    {
+      "epoch": 1.5470914127423823,
+      "grad_norm": 0.722294807434082,
+      "learning_rate": 4.2431274759209475e-06,
+      "loss": 0.599,
+      "step": 5585
+    },
+    {
+      "epoch": 1.5473684210526315,
+      "grad_norm": 0.6793596744537354,
+      "learning_rate": 4.242866328163697e-06,
+      "loss": 0.5869,
+      "step": 5586
+    },
+    {
+      "epoch": 1.547645429362881,
+      "grad_norm": 0.6670454740524292,
+      "learning_rate": 4.242605143400944e-06,
+      "loss": 0.6207,
+      "step": 5587
+    },
+    {
+      "epoch": 1.5479224376731302,
+      "grad_norm": 0.7123885154724121,
+      "learning_rate": 4.242343921638235e-06,
+      "loss": 0.6038,
+      "step": 5588
+    },
+    {
+      "epoch": 1.5481994459833794,
+      "grad_norm": 0.7021961212158203,
+      "learning_rate": 4.242082662881113e-06,
+      "loss": 0.6067,
+      "step": 5589
+    },
+    {
+      "epoch": 1.548476454293629,
+      "grad_norm": 0.6837357878684998,
+      "learning_rate": 4.241821367135128e-06,
+      "loss": 0.6044,
+      "step": 5590
+    },
+    {
+      "epoch": 1.5487534626038781,
+      "grad_norm": 0.6798022389411926,
+      "learning_rate": 4.241560034405827e-06,
+      "loss": 0.6291,
+      "step": 5591
+    },
+    {
+      "epoch": 1.5490304709141274,
+      "grad_norm": 0.7640137076377869,
+      "learning_rate": 4.24129866469876e-06,
+      "loss": 0.607,
+      "step": 5592
+    },
+    {
+      "epoch": 1.5493074792243768,
+      "grad_norm": 0.6815189123153687,
+      "learning_rate": 4.241037258019474e-06,
+      "loss": 0.6259,
+      "step": 5593
+    },
+    {
+      "epoch": 1.549584487534626,
+      "grad_norm": 0.6690675616264343,
+      "learning_rate": 4.240775814373522e-06,
+      "loss": 0.6311,
+      "step": 5594
+    },
+    {
+      "epoch": 1.5498614958448753,
+      "grad_norm": 0.6691811680793762,
+      "learning_rate": 4.2405143337664535e-06,
+      "loss": 0.5653,
+      "step": 5595
+    },
+    {
+      "epoch": 1.5501385041551248,
+      "grad_norm": 0.6947237849235535,
+      "learning_rate": 4.240252816203821e-06,
+      "loss": 0.6133,
+      "step": 5596
+    },
+    {
+      "epoch": 1.550415512465374,
+      "grad_norm": 0.7483521699905396,
+      "learning_rate": 4.239991261691178e-06,
+      "loss": 0.5988,
+      "step": 5597
+    },
+    {
+      "epoch": 1.5506925207756233,
+      "grad_norm": 0.6979661583900452,
+      "learning_rate": 4.2397296702340755e-06,
+      "loss": 0.5872,
+      "step": 5598
+    },
+    {
+      "epoch": 1.5509695290858727,
+      "grad_norm": 0.7207865118980408,
+      "learning_rate": 4.239468041838071e-06,
+      "loss": 0.5924,
+      "step": 5599
+    },
+    {
+      "epoch": 1.5512465373961217,
+      "grad_norm": 0.7086422443389893,
+      "learning_rate": 4.239206376508716e-06,
+      "loss": 0.6513,
+      "step": 5600
+    },
+    {
+      "epoch": 1.5515235457063712,
+      "grad_norm": 0.688262403011322,
+      "learning_rate": 4.238944674251569e-06,
+      "loss": 0.6381,
+      "step": 5601
+    },
+    {
+      "epoch": 1.5518005540166206,
+      "grad_norm": 0.6768400073051453,
+      "learning_rate": 4.238682935072186e-06,
+      "loss": 0.6527,
+      "step": 5602
+    },
+    {
+      "epoch": 1.5520775623268697,
+      "grad_norm": 0.7429431676864624,
+      "learning_rate": 4.238421158976124e-06,
+      "loss": 0.6481,
+      "step": 5603
+    },
+    {
+      "epoch": 1.5523545706371191,
+      "grad_norm": 0.6837465167045593,
+      "learning_rate": 4.238159345968942e-06,
+      "loss": 0.6411,
+      "step": 5604
+    },
+    {
+      "epoch": 1.5526315789473686,
+      "grad_norm": 0.6943199634552002,
+      "learning_rate": 4.237897496056198e-06,
+      "loss": 0.6303,
+      "step": 5605
+    },
+    {
+      "epoch": 1.5529085872576176,
+      "grad_norm": 0.7053850889205933,
+      "learning_rate": 4.2376356092434514e-06,
+      "loss": 0.6542,
+      "step": 5606
+    },
+    {
+      "epoch": 1.553185595567867,
+      "grad_norm": 0.6722522377967834,
+      "learning_rate": 4.237373685536264e-06,
+      "loss": 0.6155,
+      "step": 5607
+    },
+    {
+      "epoch": 1.5534626038781163,
+      "grad_norm": 0.6928883790969849,
+      "learning_rate": 4.237111724940196e-06,
+      "loss": 0.6355,
+      "step": 5608
+    },
+    {
+      "epoch": 1.5537396121883655,
+      "grad_norm": 0.7059097290039062,
+      "learning_rate": 4.23684972746081e-06,
+      "loss": 0.6008,
+      "step": 5609
+    },
+    {
+      "epoch": 1.554016620498615,
+      "grad_norm": 0.6975715160369873,
+      "learning_rate": 4.23658769310367e-06,
+      "loss": 0.616,
+      "step": 5610
+    },
+    {
+      "epoch": 1.5542936288088642,
+      "grad_norm": 0.6692479252815247,
+      "learning_rate": 4.236325621874337e-06,
+      "loss": 0.626,
+      "step": 5611
+    },
+    {
+      "epoch": 1.5545706371191135,
+      "grad_norm": 0.6283546686172485,
+      "learning_rate": 4.236063513778378e-06,
+      "loss": 0.6119,
+      "step": 5612
+    },
+    {
+      "epoch": 1.554847645429363,
+      "grad_norm": 0.678083062171936,
+      "learning_rate": 4.235801368821356e-06,
+      "loss": 0.6154,
+      "step": 5613
+    },
+    {
+      "epoch": 1.5551246537396122,
+      "grad_norm": 0.7457267045974731,
+      "learning_rate": 4.23553918700884e-06,
+      "loss": 0.6322,
+      "step": 5614
+    },
+    {
+      "epoch": 1.5554016620498614,
+      "grad_norm": 0.721976637840271,
+      "learning_rate": 4.2352769683463935e-06,
+      "loss": 0.6533,
+      "step": 5615
+    },
+    {
+      "epoch": 1.5556786703601109,
+      "grad_norm": 0.655988872051239,
+      "learning_rate": 4.235014712839586e-06,
+      "loss": 0.6062,
+      "step": 5616
+    },
+    {
+      "epoch": 1.55595567867036,
+      "grad_norm": 0.6747307777404785,
+      "learning_rate": 4.234752420493985e-06,
+      "loss": 0.599,
+      "step": 5617
+    },
+    {
+      "epoch": 1.5562326869806093,
+      "grad_norm": 0.7203627824783325,
+      "learning_rate": 4.234490091315161e-06,
+      "loss": 0.6121,
+      "step": 5618
+    },
+    {
+      "epoch": 1.5565096952908588,
+      "grad_norm": 0.7498337030410767,
+      "learning_rate": 4.234227725308682e-06,
+      "loss": 0.6466,
+      "step": 5619
+    },
+    {
+      "epoch": 1.556786703601108,
+      "grad_norm": 0.6924552321434021,
+      "learning_rate": 4.23396532248012e-06,
+      "loss": 0.6343,
+      "step": 5620
+    },
+    {
+      "epoch": 1.5570637119113573,
+      "grad_norm": 0.7387464046478271,
+      "learning_rate": 4.2337028828350465e-06,
+      "loss": 0.6427,
+      "step": 5621
+    },
+    {
+      "epoch": 1.5573407202216067,
+      "grad_norm": 0.6742679476737976,
+      "learning_rate": 4.233440406379032e-06,
+      "loss": 0.6186,
+      "step": 5622
+    },
+    {
+      "epoch": 1.557617728531856,
+      "grad_norm": 0.6853709816932678,
+      "learning_rate": 4.233177893117653e-06,
+      "loss": 0.6418,
+      "step": 5623
+    },
+    {
+      "epoch": 1.5578947368421052,
+      "grad_norm": 0.709339439868927,
+      "learning_rate": 4.23291534305648e-06,
+      "loss": 0.6201,
+      "step": 5624
+    },
+    {
+      "epoch": 1.5581717451523547,
+      "grad_norm": 0.7180233001708984,
+      "learning_rate": 4.232652756201089e-06,
+      "loss": 0.6691,
+      "step": 5625
+    },
+    {
+      "epoch": 1.558448753462604,
+      "grad_norm": 0.7282020449638367,
+      "learning_rate": 4.232390132557056e-06,
+      "loss": 0.6479,
+      "step": 5626
+    },
+    {
+      "epoch": 1.5587257617728532,
+      "grad_norm": 0.7290409207344055,
+      "learning_rate": 4.232127472129955e-06,
+      "loss": 0.6528,
+      "step": 5627
+    },
+    {
+      "epoch": 1.5590027700831026,
+      "grad_norm": 0.7034227252006531,
+      "learning_rate": 4.231864774925366e-06,
+      "loss": 0.6252,
+      "step": 5628
+    },
+    {
+      "epoch": 1.5592797783933516,
+      "grad_norm": 0.7223827838897705,
+      "learning_rate": 4.231602040948864e-06,
+      "loss": 0.6817,
+      "step": 5629
+    },
+    {
+      "epoch": 1.559556786703601,
+      "grad_norm": 0.6938064694404602,
+      "learning_rate": 4.23133927020603e-06,
+      "loss": 0.6437,
+      "step": 5630
+    },
+    {
+      "epoch": 1.5598337950138506,
+      "grad_norm": 0.7022026777267456,
+      "learning_rate": 4.231076462702441e-06,
+      "loss": 0.6048,
+      "step": 5631
+    },
+    {
+      "epoch": 1.5601108033240996,
+      "grad_norm": 0.6670244932174683,
+      "learning_rate": 4.230813618443678e-06,
+      "loss": 0.6502,
+      "step": 5632
+    },
+    {
+      "epoch": 1.560387811634349,
+      "grad_norm": 0.6831268668174744,
+      "learning_rate": 4.230550737435321e-06,
+      "loss": 0.6364,
+      "step": 5633
+    },
+    {
+      "epoch": 1.5606648199445985,
+      "grad_norm": 0.6734561324119568,
+      "learning_rate": 4.230287819682953e-06,
+      "loss": 0.6567,
+      "step": 5634
+    },
+    {
+      "epoch": 1.5609418282548475,
+      "grad_norm": 0.69823157787323,
+      "learning_rate": 4.230024865192156e-06,
+      "loss": 0.6737,
+      "step": 5635
+    },
+    {
+      "epoch": 1.561218836565097,
+      "grad_norm": 0.6804174184799194,
+      "learning_rate": 4.229761873968514e-06,
+      "loss": 0.6264,
+      "step": 5636
+    },
+    {
+      "epoch": 1.5614958448753462,
+      "grad_norm": 0.6687811613082886,
+      "learning_rate": 4.22949884601761e-06,
+      "loss": 0.6173,
+      "step": 5637
+    },
+    {
+      "epoch": 1.5617728531855954,
+      "grad_norm": 0.6959174871444702,
+      "learning_rate": 4.229235781345028e-06,
+      "loss": 0.6483,
+      "step": 5638
+    },
+    {
+      "epoch": 1.562049861495845,
+      "grad_norm": 0.7006558179855347,
+      "learning_rate": 4.228972679956355e-06,
+      "loss": 0.6643,
+      "step": 5639
+    },
+    {
+      "epoch": 1.5623268698060941,
+      "grad_norm": 0.6768726706504822,
+      "learning_rate": 4.228709541857176e-06,
+      "loss": 0.6393,
+      "step": 5640
+    },
+    {
+      "epoch": 1.5626038781163434,
+      "grad_norm": 0.6525865793228149,
+      "learning_rate": 4.2284463670530795e-06,
+      "loss": 0.6102,
+      "step": 5641
+    },
+    {
+      "epoch": 1.5628808864265928,
+      "grad_norm": 0.6904488801956177,
+      "learning_rate": 4.228183155549652e-06,
+      "loss": 0.6141,
+      "step": 5642
+    },
+    {
+      "epoch": 1.563157894736842,
+      "grad_norm": 0.6868029832839966,
+      "learning_rate": 4.2279199073524836e-06,
+      "loss": 0.6719,
+      "step": 5643
+    },
+    {
+      "epoch": 1.5634349030470913,
+      "grad_norm": 0.6973175406455994,
+      "learning_rate": 4.227656622467162e-06,
+      "loss": 0.6318,
+      "step": 5644
+    },
+    {
+      "epoch": 1.5637119113573408,
+      "grad_norm": 0.7145452499389648,
+      "learning_rate": 4.227393300899279e-06,
+      "loss": 0.6533,
+      "step": 5645
+    },
+    {
+      "epoch": 1.56398891966759,
+      "grad_norm": 0.6662142276763916,
+      "learning_rate": 4.2271299426544246e-06,
+      "loss": 0.6255,
+      "step": 5646
+    },
+    {
+      "epoch": 1.5642659279778393,
+      "grad_norm": 0.6924449801445007,
+      "learning_rate": 4.226866547738191e-06,
+      "loss": 0.6635,
+      "step": 5647
+    },
+    {
+      "epoch": 1.5645429362880887,
+      "grad_norm": 0.7409833073616028,
+      "learning_rate": 4.22660311615617e-06,
+      "loss": 0.6129,
+      "step": 5648
+    },
+    {
+      "epoch": 1.564819944598338,
+      "grad_norm": 0.7077249884605408,
+      "learning_rate": 4.226339647913957e-06,
+      "loss": 0.6303,
+      "step": 5649
+    },
+    {
+      "epoch": 1.5650969529085872,
+      "grad_norm": 0.7363086938858032,
+      "learning_rate": 4.2260761430171436e-06,
+      "loss": 0.6602,
+      "step": 5650
+    },
+    {
+      "epoch": 1.5653739612188367,
+      "grad_norm": 0.6939164996147156,
+      "learning_rate": 4.225812601471326e-06,
+      "loss": 0.6097,
+      "step": 5651
+    },
+    {
+      "epoch": 1.565650969529086,
+      "grad_norm": 0.6998023390769958,
+      "learning_rate": 4.2255490232821e-06,
+      "loss": 0.6487,
+      "step": 5652
+    },
+    {
+      "epoch": 1.5659279778393351,
+      "grad_norm": 0.6921630501747131,
+      "learning_rate": 4.225285408455062e-06,
+      "loss": 0.6318,
+      "step": 5653
+    },
+    {
+      "epoch": 1.5662049861495846,
+      "grad_norm": 0.6838632822036743,
+      "learning_rate": 4.225021756995808e-06,
+      "loss": 0.604,
+      "step": 5654
+    },
+    {
+      "epoch": 1.5664819944598338,
+      "grad_norm": 0.687044620513916,
+      "learning_rate": 4.2247580689099376e-06,
+      "loss": 0.6232,
+      "step": 5655
+    },
+    {
+      "epoch": 1.566759002770083,
+      "grad_norm": 0.7102434635162354,
+      "learning_rate": 4.224494344203048e-06,
+      "loss": 0.6399,
+      "step": 5656
+    },
+    {
+      "epoch": 1.5670360110803325,
+      "grad_norm": 0.7067602276802063,
+      "learning_rate": 4.224230582880741e-06,
+      "loss": 0.6927,
+      "step": 5657
+    },
+    {
+      "epoch": 1.5673130193905818,
+      "grad_norm": 0.7272063493728638,
+      "learning_rate": 4.223966784948614e-06,
+      "loss": 0.6373,
+      "step": 5658
+    },
+    {
+      "epoch": 1.567590027700831,
+      "grad_norm": 0.6936042904853821,
+      "learning_rate": 4.22370295041227e-06,
+      "loss": 0.645,
+      "step": 5659
+    },
+    {
+      "epoch": 1.5678670360110805,
+      "grad_norm": 0.6833428740501404,
+      "learning_rate": 4.22343907927731e-06,
+      "loss": 0.5977,
+      "step": 5660
+    },
+    {
+      "epoch": 1.5681440443213295,
+      "grad_norm": 0.6883816123008728,
+      "learning_rate": 4.2231751715493385e-06,
+      "loss": 0.645,
+      "step": 5661
+    },
+    {
+      "epoch": 1.568421052631579,
+      "grad_norm": 0.6898177862167358,
+      "learning_rate": 4.2229112272339564e-06,
+      "loss": 0.5879,
+      "step": 5662
+    },
+    {
+      "epoch": 1.5686980609418284,
+      "grad_norm": 0.7070897817611694,
+      "learning_rate": 4.22264724633677e-06,
+      "loss": 0.6639,
+      "step": 5663
+    },
+    {
+      "epoch": 1.5689750692520774,
+      "grad_norm": 0.7003639936447144,
+      "learning_rate": 4.222383228863383e-06,
+      "loss": 0.6656,
+      "step": 5664
+    },
+    {
+      "epoch": 1.5692520775623269,
+      "grad_norm": 0.7034210562705994,
+      "learning_rate": 4.222119174819402e-06,
+      "loss": 0.621,
+      "step": 5665
+    },
+    {
+      "epoch": 1.5695290858725763,
+      "grad_norm": 0.6807825565338135,
+      "learning_rate": 4.221855084210433e-06,
+      "loss": 0.6402,
+      "step": 5666
+    },
+    {
+      "epoch": 1.5698060941828254,
+      "grad_norm": 0.7109666466712952,
+      "learning_rate": 4.221590957042083e-06,
+      "loss": 0.651,
+      "step": 5667
+    },
+    {
+      "epoch": 1.5700831024930748,
+      "grad_norm": 0.6760213375091553,
+      "learning_rate": 4.221326793319961e-06,
+      "loss": 0.5901,
+      "step": 5668
+    },
+    {
+      "epoch": 1.570360110803324,
+      "grad_norm": 0.6880183815956116,
+      "learning_rate": 4.2210625930496745e-06,
+      "loss": 0.6271,
+      "step": 5669
+    },
+    {
+      "epoch": 1.5706371191135733,
+      "grad_norm": 0.6721237301826477,
+      "learning_rate": 4.2207983562368345e-06,
+      "loss": 0.6651,
+      "step": 5670
+    },
+    {
+      "epoch": 1.5709141274238227,
+      "grad_norm": 0.6518550515174866,
+      "learning_rate": 4.220534082887051e-06,
+      "loss": 0.6267,
+      "step": 5671
+    },
+    {
+      "epoch": 1.571191135734072,
+      "grad_norm": 0.7003772854804993,
+      "learning_rate": 4.220269773005936e-06,
+      "loss": 0.6366,
+      "step": 5672
+    },
+    {
+      "epoch": 1.5714681440443212,
+      "grad_norm": 0.6701993346214294,
+      "learning_rate": 4.220005426599099e-06,
+      "loss": 0.6311,
+      "step": 5673
+    },
+    {
+      "epoch": 1.5717451523545707,
+      "grad_norm": 0.7315934896469116,
+      "learning_rate": 4.219741043672156e-06,
+      "loss": 0.6511,
+      "step": 5674
+    },
+    {
+      "epoch": 1.57202216066482,
+      "grad_norm": 0.6972675323486328,
+      "learning_rate": 4.219476624230717e-06,
+      "loss": 0.6386,
+      "step": 5675
+    },
+    {
+      "epoch": 1.5722991689750692,
+      "grad_norm": 0.6991989612579346,
+      "learning_rate": 4.2192121682803995e-06,
+      "loss": 0.6406,
+      "step": 5676
+    },
+    {
+      "epoch": 1.5725761772853186,
+      "grad_norm": 0.7138469219207764,
+      "learning_rate": 4.2189476758268174e-06,
+      "loss": 0.6512,
+      "step": 5677
+    },
+    {
+      "epoch": 1.5728531855955679,
+      "grad_norm": 0.7503437995910645,
+      "learning_rate": 4.218683146875586e-06,
+      "loss": 0.5921,
+      "step": 5678
+    },
+    {
+      "epoch": 1.573130193905817,
+      "grad_norm": 0.6732651591300964,
+      "learning_rate": 4.218418581432323e-06,
+      "loss": 0.5849,
+      "step": 5679
+    },
+    {
+      "epoch": 1.5734072022160666,
+      "grad_norm": 0.710514485836029,
+      "learning_rate": 4.2181539795026435e-06,
+      "loss": 0.6108,
+      "step": 5680
+    },
+    {
+      "epoch": 1.5736842105263158,
+      "grad_norm": 0.6641442179679871,
+      "learning_rate": 4.217889341092169e-06,
+      "loss": 0.6162,
+      "step": 5681
+    },
+    {
+      "epoch": 1.573961218836565,
+      "grad_norm": 0.6942542791366577,
+      "learning_rate": 4.217624666206516e-06,
+      "loss": 0.6349,
+      "step": 5682
+    },
+    {
+      "epoch": 1.5742382271468145,
+      "grad_norm": 0.7054006457328796,
+      "learning_rate": 4.217359954851305e-06,
+      "loss": 0.6625,
+      "step": 5683
+    },
+    {
+      "epoch": 1.5745152354570637,
+      "grad_norm": 0.7004750967025757,
+      "learning_rate": 4.217095207032157e-06,
+      "loss": 0.6717,
+      "step": 5684
+    },
+    {
+      "epoch": 1.574792243767313,
+      "grad_norm": 0.7107009887695312,
+      "learning_rate": 4.216830422754693e-06,
+      "loss": 0.6459,
+      "step": 5685
+    },
+    {
+      "epoch": 1.5750692520775624,
+      "grad_norm": 0.6618795394897461,
+      "learning_rate": 4.216565602024533e-06,
+      "loss": 0.6437,
+      "step": 5686
+    },
+    {
+      "epoch": 1.5753462603878117,
+      "grad_norm": 0.6664261817932129,
+      "learning_rate": 4.216300744847304e-06,
+      "loss": 0.6536,
+      "step": 5687
+    },
+    {
+      "epoch": 1.575623268698061,
+      "grad_norm": 0.70806884765625,
+      "learning_rate": 4.2160358512286266e-06,
+      "loss": 0.6301,
+      "step": 5688
+    },
+    {
+      "epoch": 1.5759002770083104,
+      "grad_norm": 0.697137176990509,
+      "learning_rate": 4.215770921174125e-06,
+      "loss": 0.6338,
+      "step": 5689
+    },
+    {
+      "epoch": 1.5761772853185596,
+      "grad_norm": 0.6988281607627869,
+      "learning_rate": 4.2155059546894254e-06,
+      "loss": 0.6173,
+      "step": 5690
+    },
+    {
+      "epoch": 1.5764542936288088,
+      "grad_norm": 0.7043812274932861,
+      "learning_rate": 4.215240951780154e-06,
+      "loss": 0.61,
+      "step": 5691
+    },
+    {
+      "epoch": 1.5767313019390583,
+      "grad_norm": 0.7179585695266724,
+      "learning_rate": 4.214975912451936e-06,
+      "loss": 0.6334,
+      "step": 5692
+    },
+    {
+      "epoch": 1.5770083102493073,
+      "grad_norm": 0.672431230545044,
+      "learning_rate": 4.214710836710401e-06,
+      "loss": 0.6086,
+      "step": 5693
+    },
+    {
+      "epoch": 1.5772853185595568,
+      "grad_norm": 0.7191444039344788,
+      "learning_rate": 4.214445724561175e-06,
+      "loss": 0.6191,
+      "step": 5694
+    },
+    {
+      "epoch": 1.5775623268698062,
+      "grad_norm": 0.7192513346672058,
+      "learning_rate": 4.214180576009888e-06,
+      "loss": 0.5964,
+      "step": 5695
+    },
+    {
+      "epoch": 1.5778393351800553,
+      "grad_norm": 0.6964846849441528,
+      "learning_rate": 4.213915391062171e-06,
+      "loss": 0.6572,
+      "step": 5696
+    },
+    {
+      "epoch": 1.5781163434903047,
+      "grad_norm": 0.7242207527160645,
+      "learning_rate": 4.213650169723652e-06,
+      "loss": 0.6375,
+      "step": 5697
+    },
+    {
+      "epoch": 1.5783933518005542,
+      "grad_norm": 0.6550120115280151,
+      "learning_rate": 4.213384911999965e-06,
+      "loss": 0.5441,
+      "step": 5698
+    },
+    {
+      "epoch": 1.5786703601108032,
+      "grad_norm": 0.72943115234375,
+      "learning_rate": 4.21311961789674e-06,
+      "loss": 0.637,
+      "step": 5699
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.6830201745033264,
+      "learning_rate": 4.212854287419611e-06,
+      "loss": 0.5859,
+      "step": 5700
+    },
+    {
+      "epoch": 1.579224376731302,
+      "grad_norm": 0.7375404238700867,
+      "learning_rate": 4.212588920574211e-06,
+      "loss": 0.6329,
+      "step": 5701
+    },
+    {
+      "epoch": 1.5795013850415511,
+      "grad_norm": 0.6917905807495117,
+      "learning_rate": 4.212323517366175e-06,
+      "loss": 0.6007,
+      "step": 5702
+    },
+    {
+      "epoch": 1.5797783933518006,
+      "grad_norm": 0.6942349076271057,
+      "learning_rate": 4.2120580778011375e-06,
+      "loss": 0.6181,
+      "step": 5703
+    },
+    {
+      "epoch": 1.5800554016620498,
+      "grad_norm": 0.6996932029724121,
+      "learning_rate": 4.211792601884736e-06,
+      "loss": 0.6177,
+      "step": 5704
+    },
+    {
+      "epoch": 1.580332409972299,
+      "grad_norm": 0.7083521485328674,
+      "learning_rate": 4.2115270896226045e-06,
+      "loss": 0.6672,
+      "step": 5705
+    },
+    {
+      "epoch": 1.5806094182825485,
+      "grad_norm": 0.7094027996063232,
+      "learning_rate": 4.211261541020383e-06,
+      "loss": 0.626,
+      "step": 5706
+    },
+    {
+      "epoch": 1.5808864265927978,
+      "grad_norm": 0.6889899373054504,
+      "learning_rate": 4.210995956083709e-06,
+      "loss": 0.6546,
+      "step": 5707
+    },
+    {
+      "epoch": 1.581163434903047,
+      "grad_norm": 0.7199320793151855,
+      "learning_rate": 4.2107303348182214e-06,
+      "loss": 0.5984,
+      "step": 5708
+    },
+    {
+      "epoch": 1.5814404432132965,
+      "grad_norm": 0.6789028644561768,
+      "learning_rate": 4.210464677229561e-06,
+      "loss": 0.6422,
+      "step": 5709
+    },
+    {
+      "epoch": 1.5817174515235457,
+      "grad_norm": 0.6862844824790955,
+      "learning_rate": 4.210198983323366e-06,
+      "loss": 0.568,
+      "step": 5710
+    },
+    {
+      "epoch": 1.581994459833795,
+      "grad_norm": 0.6758617162704468,
+      "learning_rate": 4.20993325310528e-06,
+      "loss": 0.6054,
+      "step": 5711
+    },
+    {
+      "epoch": 1.5822714681440444,
+      "grad_norm": 0.7183728814125061,
+      "learning_rate": 4.209667486580943e-06,
+      "loss": 0.6674,
+      "step": 5712
+    },
+    {
+      "epoch": 1.5825484764542936,
+      "grad_norm": 0.6828016638755798,
+      "learning_rate": 4.209401683756e-06,
+      "loss": 0.5502,
+      "step": 5713
+    },
+    {
+      "epoch": 1.5828254847645429,
+      "grad_norm": 0.6988704204559326,
+      "learning_rate": 4.209135844636095e-06,
+      "loss": 0.6313,
+      "step": 5714
+    },
+    {
+      "epoch": 1.5831024930747923,
+      "grad_norm": 0.7439329028129578,
+      "learning_rate": 4.2088699692268704e-06,
+      "loss": 0.6421,
+      "step": 5715
+    },
+    {
+      "epoch": 1.5833795013850416,
+      "grad_norm": 0.6711263060569763,
+      "learning_rate": 4.2086040575339726e-06,
+      "loss": 0.6273,
+      "step": 5716
+    },
+    {
+      "epoch": 1.5836565096952908,
+      "grad_norm": 0.7117868065834045,
+      "learning_rate": 4.2083381095630474e-06,
+      "loss": 0.6029,
+      "step": 5717
+    },
+    {
+      "epoch": 1.5839335180055403,
+      "grad_norm": 0.6777087450027466,
+      "learning_rate": 4.208072125319741e-06,
+      "loss": 0.5973,
+      "step": 5718
+    },
+    {
+      "epoch": 1.5842105263157895,
+      "grad_norm": 0.6707855463027954,
+      "learning_rate": 4.2078061048097015e-06,
+      "loss": 0.6251,
+      "step": 5719
+    },
+    {
+      "epoch": 1.5844875346260388,
+      "grad_norm": 0.6649163961410522,
+      "learning_rate": 4.207540048038578e-06,
+      "loss": 0.5889,
+      "step": 5720
+    },
+    {
+      "epoch": 1.5847645429362882,
+      "grad_norm": 0.6663191318511963,
+      "learning_rate": 4.207273955012018e-06,
+      "loss": 0.6105,
+      "step": 5721
+    },
+    {
+      "epoch": 1.5850415512465375,
+      "grad_norm": 0.6953136324882507,
+      "learning_rate": 4.2070078257356714e-06,
+      "loss": 0.6114,
+      "step": 5722
+    },
+    {
+      "epoch": 1.5853185595567867,
+      "grad_norm": 0.6712597012519836,
+      "learning_rate": 4.206741660215191e-06,
+      "loss": 0.6083,
+      "step": 5723
+    },
+    {
+      "epoch": 1.5855955678670361,
+      "grad_norm": 0.6431275606155396,
+      "learning_rate": 4.2064754584562255e-06,
+      "loss": 0.5897,
+      "step": 5724
+    },
+    {
+      "epoch": 1.5858725761772852,
+      "grad_norm": 0.7072885632514954,
+      "learning_rate": 4.206209220464429e-06,
+      "loss": 0.6651,
+      "step": 5725
+    },
+    {
+      "epoch": 1.5861495844875346,
+      "grad_norm": 0.6720908284187317,
+      "learning_rate": 4.205942946245453e-06,
+      "loss": 0.6164,
+      "step": 5726
+    },
+    {
+      "epoch": 1.586426592797784,
+      "grad_norm": 0.6930977702140808,
+      "learning_rate": 4.205676635804952e-06,
+      "loss": 0.6553,
+      "step": 5727
+    },
+    {
+      "epoch": 1.586703601108033,
+      "grad_norm": 0.691413938999176,
+      "learning_rate": 4.20541028914858e-06,
+      "loss": 0.6186,
+      "step": 5728
+    },
+    {
+      "epoch": 1.5869806094182826,
+      "grad_norm": 0.7159149646759033,
+      "learning_rate": 4.205143906281992e-06,
+      "loss": 0.6417,
+      "step": 5729
+    },
+    {
+      "epoch": 1.587257617728532,
+      "grad_norm": 0.6715986132621765,
+      "learning_rate": 4.204877487210845e-06,
+      "loss": 0.6001,
+      "step": 5730
+    },
+    {
+      "epoch": 1.587534626038781,
+      "grad_norm": 0.6652202010154724,
+      "learning_rate": 4.204611031940795e-06,
+      "loss": 0.6552,
+      "step": 5731
+    },
+    {
+      "epoch": 1.5878116343490305,
+      "grad_norm": 0.7530796527862549,
+      "learning_rate": 4.204344540477499e-06,
+      "loss": 0.6387,
+      "step": 5732
+    },
+    {
+      "epoch": 1.5880886426592797,
+      "grad_norm": 0.6929827928543091,
+      "learning_rate": 4.2040780128266165e-06,
+      "loss": 0.6227,
+      "step": 5733
+    },
+    {
+      "epoch": 1.588365650969529,
+      "grad_norm": 0.6594801545143127,
+      "learning_rate": 4.203811448993806e-06,
+      "loss": 0.5887,
+      "step": 5734
+    },
+    {
+      "epoch": 1.5886426592797784,
+      "grad_norm": 0.6690226197242737,
+      "learning_rate": 4.203544848984729e-06,
+      "loss": 0.5977,
+      "step": 5735
+    },
+    {
+      "epoch": 1.5889196675900277,
+      "grad_norm": 0.7046278119087219,
+      "learning_rate": 4.203278212805042e-06,
+      "loss": 0.6422,
+      "step": 5736
+    },
+    {
+      "epoch": 1.589196675900277,
+      "grad_norm": 0.6667335033416748,
+      "learning_rate": 4.20301154046041e-06,
+      "loss": 0.6092,
+      "step": 5737
+    },
+    {
+      "epoch": 1.5894736842105264,
+      "grad_norm": 0.66410893201828,
+      "learning_rate": 4.202744831956494e-06,
+      "loss": 0.6406,
+      "step": 5738
+    },
+    {
+      "epoch": 1.5897506925207756,
+      "grad_norm": 0.6676257252693176,
+      "learning_rate": 4.202478087298957e-06,
+      "loss": 0.6684,
+      "step": 5739
+    },
+    {
+      "epoch": 1.5900277008310248,
+      "grad_norm": 0.6923943758010864,
+      "learning_rate": 4.202211306493462e-06,
+      "loss": 0.5827,
+      "step": 5740
+    },
+    {
+      "epoch": 1.5903047091412743,
+      "grad_norm": 0.6694300770759583,
+      "learning_rate": 4.201944489545674e-06,
+      "loss": 0.6266,
+      "step": 5741
+    },
+    {
+      "epoch": 1.5905817174515235,
+      "grad_norm": 0.6623051762580872,
+      "learning_rate": 4.201677636461259e-06,
+      "loss": 0.6549,
+      "step": 5742
+    },
+    {
+      "epoch": 1.5908587257617728,
+      "grad_norm": 0.6522954106330872,
+      "learning_rate": 4.201410747245881e-06,
+      "loss": 0.6173,
+      "step": 5743
+    },
+    {
+      "epoch": 1.5911357340720222,
+      "grad_norm": 0.7510678768157959,
+      "learning_rate": 4.201143821905209e-06,
+      "loss": 0.6355,
+      "step": 5744
+    },
+    {
+      "epoch": 1.5914127423822715,
+      "grad_norm": 0.6667211651802063,
+      "learning_rate": 4.200876860444909e-06,
+      "loss": 0.6082,
+      "step": 5745
+    },
+    {
+      "epoch": 1.5916897506925207,
+      "grad_norm": 0.6723722219467163,
+      "learning_rate": 4.2006098628706496e-06,
+      "loss": 0.6256,
+      "step": 5746
+    },
+    {
+      "epoch": 1.5919667590027702,
+      "grad_norm": 0.6964446306228638,
+      "learning_rate": 4.2003428291881e-06,
+      "loss": 0.6059,
+      "step": 5747
+    },
+    {
+      "epoch": 1.5922437673130194,
+      "grad_norm": 0.73036128282547,
+      "learning_rate": 4.20007575940293e-06,
+      "loss": 0.6385,
+      "step": 5748
+    },
+    {
+      "epoch": 1.5925207756232687,
+      "grad_norm": 0.6907002329826355,
+      "learning_rate": 4.19980865352081e-06,
+      "loss": 0.6662,
+      "step": 5749
+    },
+    {
+      "epoch": 1.5927977839335181,
+      "grad_norm": 0.7342780232429504,
+      "learning_rate": 4.199541511547412e-06,
+      "loss": 0.6752,
+      "step": 5750
+    },
+    {
+      "epoch": 1.5930747922437674,
+      "grad_norm": 0.7659139633178711,
+      "learning_rate": 4.1992743334884065e-06,
+      "loss": 0.6259,
+      "step": 5751
+    },
+    {
+      "epoch": 1.5933518005540166,
+      "grad_norm": 0.6953239440917969,
+      "learning_rate": 4.199007119349468e-06,
+      "loss": 0.672,
+      "step": 5752
+    },
+    {
+      "epoch": 1.593628808864266,
+      "grad_norm": 0.6655431985855103,
+      "learning_rate": 4.19873986913627e-06,
+      "loss": 0.6207,
+      "step": 5753
+    },
+    {
+      "epoch": 1.593905817174515,
+      "grad_norm": 0.7081992030143738,
+      "learning_rate": 4.1984725828544855e-06,
+      "loss": 0.6568,
+      "step": 5754
+    },
+    {
+      "epoch": 1.5941828254847645,
+      "grad_norm": 0.6927662491798401,
+      "learning_rate": 4.198205260509791e-06,
+      "loss": 0.6323,
+      "step": 5755
+    },
+    {
+      "epoch": 1.594459833795014,
+      "grad_norm": 0.698785662651062,
+      "learning_rate": 4.197937902107863e-06,
+      "loss": 0.6653,
+      "step": 5756
+    },
+    {
+      "epoch": 1.594736842105263,
+      "grad_norm": 0.7155250906944275,
+      "learning_rate": 4.197670507654376e-06,
+      "loss": 0.6522,
+      "step": 5757
+    },
+    {
+      "epoch": 1.5950138504155125,
+      "grad_norm": 0.6735395789146423,
+      "learning_rate": 4.197403077155009e-06,
+      "loss": 0.5891,
+      "step": 5758
+    },
+    {
+      "epoch": 1.595290858725762,
+      "grad_norm": 0.6795505285263062,
+      "learning_rate": 4.19713561061544e-06,
+      "loss": 0.6047,
+      "step": 5759
+    },
+    {
+      "epoch": 1.595567867036011,
+      "grad_norm": 0.6992025971412659,
+      "learning_rate": 4.196868108041348e-06,
+      "loss": 0.6649,
+      "step": 5760
+    },
+    {
+      "epoch": 1.5958448753462604,
+      "grad_norm": 0.7018173336982727,
+      "learning_rate": 4.196600569438413e-06,
+      "loss": 0.6429,
+      "step": 5761
+    },
+    {
+      "epoch": 1.5961218836565096,
+      "grad_norm": 0.7325546145439148,
+      "learning_rate": 4.196332994812316e-06,
+      "loss": 0.6058,
+      "step": 5762
+    },
+    {
+      "epoch": 1.5963988919667589,
+      "grad_norm": 0.7186279892921448,
+      "learning_rate": 4.196065384168736e-06,
+      "loss": 0.6224,
+      "step": 5763
+    },
+    {
+      "epoch": 1.5966759002770083,
+      "grad_norm": 0.7321295142173767,
+      "learning_rate": 4.195797737513357e-06,
+      "loss": 0.6758,
+      "step": 5764
+    },
+    {
+      "epoch": 1.5969529085872576,
+      "grad_norm": 0.7160171866416931,
+      "learning_rate": 4.195530054851863e-06,
+      "loss": 0.6367,
+      "step": 5765
+    },
+    {
+      "epoch": 1.5972299168975068,
+      "grad_norm": 0.6524419784545898,
+      "learning_rate": 4.1952623361899345e-06,
+      "loss": 0.5977,
+      "step": 5766
+    },
+    {
+      "epoch": 1.5975069252077563,
+      "grad_norm": 0.6996549367904663,
+      "learning_rate": 4.1949945815332574e-06,
+      "loss": 0.627,
+      "step": 5767
+    },
+    {
+      "epoch": 1.5977839335180055,
+      "grad_norm": 0.6982312798500061,
+      "learning_rate": 4.194726790887517e-06,
+      "loss": 0.6677,
+      "step": 5768
+    },
+    {
+      "epoch": 1.5980609418282548,
+      "grad_norm": 0.6984411478042603,
+      "learning_rate": 4.1944589642583985e-06,
+      "loss": 0.5714,
+      "step": 5769
+    },
+    {
+      "epoch": 1.5983379501385042,
+      "grad_norm": 0.662164568901062,
+      "learning_rate": 4.1941911016515905e-06,
+      "loss": 0.5776,
+      "step": 5770
+    },
+    {
+      "epoch": 1.5986149584487535,
+      "grad_norm": 0.7050111889839172,
+      "learning_rate": 4.193923203072778e-06,
+      "loss": 0.6099,
+      "step": 5771
+    },
+    {
+      "epoch": 1.5988919667590027,
+      "grad_norm": 0.69181227684021,
+      "learning_rate": 4.19365526852765e-06,
+      "loss": 0.6415,
+      "step": 5772
+    },
+    {
+      "epoch": 1.5991689750692522,
+      "grad_norm": 0.6893934011459351,
+      "learning_rate": 4.193387298021895e-06,
+      "loss": 0.6694,
+      "step": 5773
+    },
+    {
+      "epoch": 1.5994459833795014,
+      "grad_norm": 0.7136704325675964,
+      "learning_rate": 4.193119291561204e-06,
+      "loss": 0.6497,
+      "step": 5774
+    },
+    {
+      "epoch": 1.5997229916897506,
+      "grad_norm": 0.717369556427002,
+      "learning_rate": 4.192851249151266e-06,
+      "loss": 0.6128,
+      "step": 5775
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7070300579071045,
+      "learning_rate": 4.192583170797775e-06,
+      "loss": 0.6448,
+      "step": 5776
+    },
+    {
+      "epoch": 1.6002770083102493,
+      "grad_norm": 0.7221513986587524,
+      "learning_rate": 4.192315056506418e-06,
+      "loss": 0.6756,
+      "step": 5777
+    },
+    {
+      "epoch": 1.6005540166204986,
+      "grad_norm": 0.7016468048095703,
+      "learning_rate": 4.1920469062828924e-06,
+      "loss": 0.6226,
+      "step": 5778
+    },
+    {
+      "epoch": 1.600831024930748,
+      "grad_norm": 0.7056813836097717,
+      "learning_rate": 4.1917787201328894e-06,
+      "loss": 0.6313,
+      "step": 5779
+    },
+    {
+      "epoch": 1.6011080332409973,
+      "grad_norm": 0.7116861939430237,
+      "learning_rate": 4.191510498062104e-06,
+      "loss": 0.6539,
+      "step": 5780
+    },
+    {
+      "epoch": 1.6013850415512465,
+      "grad_norm": 0.6958196759223938,
+      "learning_rate": 4.191242240076231e-06,
+      "loss": 0.6612,
+      "step": 5781
+    },
+    {
+      "epoch": 1.601662049861496,
+      "grad_norm": 0.707463800907135,
+      "learning_rate": 4.190973946180966e-06,
+      "loss": 0.6052,
+      "step": 5782
+    },
+    {
+      "epoch": 1.6019390581717452,
+      "grad_norm": 0.6749431490898132,
+      "learning_rate": 4.190705616382006e-06,
+      "loss": 0.6278,
+      "step": 5783
+    },
+    {
+      "epoch": 1.6022160664819944,
+      "grad_norm": 0.727830171585083,
+      "learning_rate": 4.190437250685049e-06,
+      "loss": 0.6063,
+      "step": 5784
+    },
+    {
+      "epoch": 1.602493074792244,
+      "grad_norm": 0.7113375663757324,
+      "learning_rate": 4.190168849095791e-06,
+      "loss": 0.645,
+      "step": 5785
+    },
+    {
+      "epoch": 1.602770083102493,
+      "grad_norm": 0.6866186261177063,
+      "learning_rate": 4.1899004116199335e-06,
+      "loss": 0.6072,
+      "step": 5786
+    },
+    {
+      "epoch": 1.6030470914127424,
+      "grad_norm": 0.7033379673957825,
+      "learning_rate": 4.189631938263173e-06,
+      "loss": 0.6638,
+      "step": 5787
+    },
+    {
+      "epoch": 1.6033240997229918,
+      "grad_norm": 0.6700825691223145,
+      "learning_rate": 4.1893634290312135e-06,
+      "loss": 0.6362,
+      "step": 5788
+    },
+    {
+      "epoch": 1.6036011080332409,
+      "grad_norm": 0.7247818112373352,
+      "learning_rate": 4.189094883929753e-06,
+      "loss": 0.6437,
+      "step": 5789
+    },
+    {
+      "epoch": 1.6038781163434903,
+      "grad_norm": 0.6923117637634277,
+      "learning_rate": 4.188826302964495e-06,
+      "loss": 0.6261,
+      "step": 5790
+    },
+    {
+      "epoch": 1.6041551246537398,
+      "grad_norm": 0.6868463754653931,
+      "learning_rate": 4.188557686141141e-06,
+      "loss": 0.6412,
+      "step": 5791
+    },
+    {
+      "epoch": 1.6044321329639888,
+      "grad_norm": 0.6793109774589539,
+      "learning_rate": 4.188289033465396e-06,
+      "loss": 0.6606,
+      "step": 5792
+    },
+    {
+      "epoch": 1.6047091412742382,
+      "grad_norm": 0.7150652408599854,
+      "learning_rate": 4.188020344942964e-06,
+      "loss": 0.6581,
+      "step": 5793
+    },
+    {
+      "epoch": 1.6049861495844875,
+      "grad_norm": 0.6954755783081055,
+      "learning_rate": 4.187751620579549e-06,
+      "loss": 0.6559,
+      "step": 5794
+    },
+    {
+      "epoch": 1.6052631578947367,
+      "grad_norm": 0.691980242729187,
+      "learning_rate": 4.187482860380857e-06,
+      "loss": 0.6021,
+      "step": 5795
+    },
+    {
+      "epoch": 1.6055401662049862,
+      "grad_norm": 0.670291006565094,
+      "learning_rate": 4.187214064352594e-06,
+      "loss": 0.6163,
+      "step": 5796
+    },
+    {
+      "epoch": 1.6058171745152354,
+      "grad_norm": 0.6734989881515503,
+      "learning_rate": 4.186945232500468e-06,
+      "loss": 0.6721,
+      "step": 5797
+    },
+    {
+      "epoch": 1.6060941828254847,
+      "grad_norm": 0.6668962240219116,
+      "learning_rate": 4.186676364830187e-06,
+      "loss": 0.5777,
+      "step": 5798
+    },
+    {
+      "epoch": 1.6063711911357341,
+      "grad_norm": 0.722041666507721,
+      "learning_rate": 4.186407461347459e-06,
+      "loss": 0.6112,
+      "step": 5799
+    },
+    {
+      "epoch": 1.6066481994459834,
+      "grad_norm": 0.73446124792099,
+      "learning_rate": 4.186138522057994e-06,
+      "loss": 0.6116,
+      "step": 5800
+    },
+    {
+      "epoch": 1.6069252077562326,
+      "grad_norm": 0.6750873923301697,
+      "learning_rate": 4.185869546967502e-06,
+      "loss": 0.6499,
+      "step": 5801
+    },
+    {
+      "epoch": 1.607202216066482,
+      "grad_norm": 0.6893951892852783,
+      "learning_rate": 4.185600536081694e-06,
+      "loss": 0.6315,
+      "step": 5802
+    },
+    {
+      "epoch": 1.6074792243767313,
+      "grad_norm": 0.7207143902778625,
+      "learning_rate": 4.185331489406283e-06,
+      "loss": 0.6556,
+      "step": 5803
+    },
+    {
+      "epoch": 1.6077562326869805,
+      "grad_norm": 0.6945677399635315,
+      "learning_rate": 4.1850624069469805e-06,
+      "loss": 0.5956,
+      "step": 5804
+    },
+    {
+      "epoch": 1.60803324099723,
+      "grad_norm": 0.6582666039466858,
+      "learning_rate": 4.1847932887095e-06,
+      "loss": 0.6429,
+      "step": 5805
+    },
+    {
+      "epoch": 1.6083102493074792,
+      "grad_norm": 0.6848458647727966,
+      "learning_rate": 4.184524134699556e-06,
+      "loss": 0.6427,
+      "step": 5806
+    },
+    {
+      "epoch": 1.6085872576177285,
+      "grad_norm": 0.6576165556907654,
+      "learning_rate": 4.184254944922862e-06,
+      "loss": 0.6079,
+      "step": 5807
+    },
+    {
+      "epoch": 1.608864265927978,
+      "grad_norm": 0.7028154134750366,
+      "learning_rate": 4.1839857193851345e-06,
+      "loss": 0.6434,
+      "step": 5808
+    },
+    {
+      "epoch": 1.6091412742382272,
+      "grad_norm": 0.684888482093811,
+      "learning_rate": 4.183716458092089e-06,
+      "loss": 0.6032,
+      "step": 5809
+    },
+    {
+      "epoch": 1.6094182825484764,
+      "grad_norm": 0.7266806364059448,
+      "learning_rate": 4.183447161049445e-06,
+      "loss": 0.6548,
+      "step": 5810
+    },
+    {
+      "epoch": 1.6096952908587259,
+      "grad_norm": 0.6765641570091248,
+      "learning_rate": 4.183177828262916e-06,
+      "loss": 0.6468,
+      "step": 5811
+    },
+    {
+      "epoch": 1.609972299168975,
+      "grad_norm": 0.6718610525131226,
+      "learning_rate": 4.182908459738226e-06,
+      "loss": 0.5708,
+      "step": 5812
+    },
+    {
+      "epoch": 1.6102493074792243,
+      "grad_norm": 0.714972734451294,
+      "learning_rate": 4.1826390554810905e-06,
+      "loss": 0.6184,
+      "step": 5813
+    },
+    {
+      "epoch": 1.6105263157894738,
+      "grad_norm": 0.6993780136108398,
+      "learning_rate": 4.182369615497232e-06,
+      "loss": 0.665,
+      "step": 5814
+    },
+    {
+      "epoch": 1.610803324099723,
+      "grad_norm": 0.6753741502761841,
+      "learning_rate": 4.182100139792369e-06,
+      "loss": 0.6266,
+      "step": 5815
+    },
+    {
+      "epoch": 1.6110803324099723,
+      "grad_norm": 0.6745509505271912,
+      "learning_rate": 4.1818306283722245e-06,
+      "loss": 0.6409,
+      "step": 5816
+    },
+    {
+      "epoch": 1.6113573407202217,
+      "grad_norm": 0.705803394317627,
+      "learning_rate": 4.181561081242521e-06,
+      "loss": 0.6208,
+      "step": 5817
+    },
+    {
+      "epoch": 1.6116343490304708,
+      "grad_norm": 0.7307119369506836,
+      "learning_rate": 4.181291498408983e-06,
+      "loss": 0.6313,
+      "step": 5818
+    },
+    {
+      "epoch": 1.6119113573407202,
+      "grad_norm": 0.6954278945922852,
+      "learning_rate": 4.181021879877331e-06,
+      "loss": 0.6508,
+      "step": 5819
+    },
+    {
+      "epoch": 1.6121883656509697,
+      "grad_norm": 0.7063873410224915,
+      "learning_rate": 4.1807522256532925e-06,
+      "loss": 0.6352,
+      "step": 5820
+    },
+    {
+      "epoch": 1.6124653739612187,
+      "grad_norm": 0.7044040560722351,
+      "learning_rate": 4.180482535742591e-06,
+      "loss": 0.603,
+      "step": 5821
+    },
+    {
+      "epoch": 1.6127423822714682,
+      "grad_norm": 0.6697425246238708,
+      "learning_rate": 4.1802128101509556e-06,
+      "loss": 0.5898,
+      "step": 5822
+    },
+    {
+      "epoch": 1.6130193905817176,
+      "grad_norm": 0.6932482719421387,
+      "learning_rate": 4.17994304888411e-06,
+      "loss": 0.6418,
+      "step": 5823
+    },
+    {
+      "epoch": 1.6132963988919666,
+      "grad_norm": 0.702626645565033,
+      "learning_rate": 4.179673251947784e-06,
+      "loss": 0.6289,
+      "step": 5824
+    },
+    {
+      "epoch": 1.613573407202216,
+      "grad_norm": 0.6617067456245422,
+      "learning_rate": 4.179403419347705e-06,
+      "loss": 0.621,
+      "step": 5825
+    },
+    {
+      "epoch": 1.6138504155124653,
+      "grad_norm": 0.6601985096931458,
+      "learning_rate": 4.179133551089604e-06,
+      "loss": 0.6575,
+      "step": 5826
+    },
+    {
+      "epoch": 1.6141274238227146,
+      "grad_norm": 0.6812746524810791,
+      "learning_rate": 4.178863647179208e-06,
+      "loss": 0.6089,
+      "step": 5827
+    },
+    {
+      "epoch": 1.614404432132964,
+      "grad_norm": 0.7015700936317444,
+      "learning_rate": 4.17859370762225e-06,
+      "loss": 0.5839,
+      "step": 5828
+    },
+    {
+      "epoch": 1.6146814404432133,
+      "grad_norm": 0.6997047066688538,
+      "learning_rate": 4.1783237324244605e-06,
+      "loss": 0.6244,
+      "step": 5829
+    },
+    {
+      "epoch": 1.6149584487534625,
+      "grad_norm": 0.6965237259864807,
+      "learning_rate": 4.178053721591574e-06,
+      "loss": 0.5986,
+      "step": 5830
+    },
+    {
+      "epoch": 1.615235457063712,
+      "grad_norm": 0.69244784116745,
+      "learning_rate": 4.17778367512932e-06,
+      "loss": 0.5947,
+      "step": 5831
+    },
+    {
+      "epoch": 1.6155124653739612,
+      "grad_norm": 0.7107717990875244,
+      "learning_rate": 4.177513593043434e-06,
+      "loss": 0.6405,
+      "step": 5832
+    },
+    {
+      "epoch": 1.6157894736842104,
+      "grad_norm": 0.6921395659446716,
+      "learning_rate": 4.177243475339651e-06,
+      "loss": 0.6364,
+      "step": 5833
+    },
+    {
+      "epoch": 1.61606648199446,
+      "grad_norm": 0.7329249382019043,
+      "learning_rate": 4.1769733220237055e-06,
+      "loss": 0.6618,
+      "step": 5834
+    },
+    {
+      "epoch": 1.6163434903047091,
+      "grad_norm": 0.6867843270301819,
+      "learning_rate": 4.176703133101334e-06,
+      "loss": 0.5642,
+      "step": 5835
+    },
+    {
+      "epoch": 1.6166204986149584,
+      "grad_norm": 0.6759707927703857,
+      "learning_rate": 4.176432908578274e-06,
+      "loss": 0.6479,
+      "step": 5836
+    },
+    {
+      "epoch": 1.6168975069252078,
+      "grad_norm": 0.7146587371826172,
+      "learning_rate": 4.176162648460261e-06,
+      "loss": 0.6392,
+      "step": 5837
+    },
+    {
+      "epoch": 1.617174515235457,
+      "grad_norm": 0.7169117331504822,
+      "learning_rate": 4.175892352753035e-06,
+      "loss": 0.6448,
+      "step": 5838
+    },
+    {
+      "epoch": 1.6174515235457063,
+      "grad_norm": 0.700145423412323,
+      "learning_rate": 4.175622021462335e-06,
+      "loss": 0.6157,
+      "step": 5839
+    },
+    {
+      "epoch": 1.6177285318559558,
+      "grad_norm": 0.7134767174720764,
+      "learning_rate": 4.1753516545939e-06,
+      "loss": 0.6314,
+      "step": 5840
+    },
+    {
+      "epoch": 1.618005540166205,
+      "grad_norm": 0.6458624005317688,
+      "learning_rate": 4.17508125215347e-06,
+      "loss": 0.5763,
+      "step": 5841
+    },
+    {
+      "epoch": 1.6182825484764543,
+      "grad_norm": 0.6700456142425537,
+      "learning_rate": 4.174810814146789e-06,
+      "loss": 0.6501,
+      "step": 5842
+    },
+    {
+      "epoch": 1.6185595567867037,
+      "grad_norm": 0.6677443981170654,
+      "learning_rate": 4.174540340579596e-06,
+      "loss": 0.6479,
+      "step": 5843
+    },
+    {
+      "epoch": 1.618836565096953,
+      "grad_norm": 0.6755135655403137,
+      "learning_rate": 4.174269831457636e-06,
+      "loss": 0.5886,
+      "step": 5844
+    },
+    {
+      "epoch": 1.6191135734072022,
+      "grad_norm": 0.7005934715270996,
+      "learning_rate": 4.173999286786652e-06,
+      "loss": 0.6009,
+      "step": 5845
+    },
+    {
+      "epoch": 1.6193905817174516,
+      "grad_norm": 0.682978630065918,
+      "learning_rate": 4.173728706572388e-06,
+      "loss": 0.6118,
+      "step": 5846
+    },
+    {
+      "epoch": 1.6196675900277007,
+      "grad_norm": 0.6874737739562988,
+      "learning_rate": 4.173458090820589e-06,
+      "loss": 0.6296,
+      "step": 5847
+    },
+    {
+      "epoch": 1.6199445983379501,
+      "grad_norm": 0.6969865560531616,
+      "learning_rate": 4.173187439537001e-06,
+      "loss": 0.6379,
+      "step": 5848
+    },
+    {
+      "epoch": 1.6202216066481996,
+      "grad_norm": 0.7983630895614624,
+      "learning_rate": 4.172916752727372e-06,
+      "loss": 0.6125,
+      "step": 5849
+    },
+    {
+      "epoch": 1.6204986149584486,
+      "grad_norm": 0.6705129742622375,
+      "learning_rate": 4.172646030397447e-06,
+      "loss": 0.5946,
+      "step": 5850
+    },
+    {
+      "epoch": 1.620775623268698,
+      "grad_norm": 0.6784124970436096,
+      "learning_rate": 4.172375272552976e-06,
+      "loss": 0.604,
+      "step": 5851
+    },
+    {
+      "epoch": 1.6210526315789475,
+      "grad_norm": 0.7067432403564453,
+      "learning_rate": 4.172104479199707e-06,
+      "loss": 0.667,
+      "step": 5852
+    },
+    {
+      "epoch": 1.6213296398891965,
+      "grad_norm": 0.7362405061721802,
+      "learning_rate": 4.17183365034339e-06,
+      "loss": 0.6497,
+      "step": 5853
+    },
+    {
+      "epoch": 1.621606648199446,
+      "grad_norm": 0.6884350180625916,
+      "learning_rate": 4.171562785989774e-06,
+      "loss": 0.6158,
+      "step": 5854
+    },
+    {
+      "epoch": 1.6218836565096952,
+      "grad_norm": 0.699510931968689,
+      "learning_rate": 4.171291886144614e-06,
+      "loss": 0.598,
+      "step": 5855
+    },
+    {
+      "epoch": 1.6221606648199445,
+      "grad_norm": 0.6367965936660767,
+      "learning_rate": 4.171020950813657e-06,
+      "loss": 0.6063,
+      "step": 5856
+    },
+    {
+      "epoch": 1.622437673130194,
+      "grad_norm": 0.6910621523857117,
+      "learning_rate": 4.170749980002658e-06,
+      "loss": 0.6346,
+      "step": 5857
+    },
+    {
+      "epoch": 1.6227146814404432,
+      "grad_norm": 0.702388346195221,
+      "learning_rate": 4.170478973717371e-06,
+      "loss": 0.6157,
+      "step": 5858
+    },
+    {
+      "epoch": 1.6229916897506924,
+      "grad_norm": 0.7175259590148926,
+      "learning_rate": 4.1702079319635495e-06,
+      "loss": 0.6365,
+      "step": 5859
+    },
+    {
+      "epoch": 1.6232686980609419,
+      "grad_norm": 0.6877427101135254,
+      "learning_rate": 4.169936854746947e-06,
+      "loss": 0.6248,
+      "step": 5860
+    },
+    {
+      "epoch": 1.623545706371191,
+      "grad_norm": 0.7271732091903687,
+      "learning_rate": 4.169665742073322e-06,
+      "loss": 0.6386,
+      "step": 5861
+    },
+    {
+      "epoch": 1.6238227146814403,
+      "grad_norm": 0.6602852940559387,
+      "learning_rate": 4.169394593948429e-06,
+      "loss": 0.6184,
+      "step": 5862
+    },
+    {
+      "epoch": 1.6240997229916898,
+      "grad_norm": 0.7164726853370667,
+      "learning_rate": 4.169123410378026e-06,
+      "loss": 0.6213,
+      "step": 5863
+    },
+    {
+      "epoch": 1.624376731301939,
+      "grad_norm": 0.6855493187904358,
+      "learning_rate": 4.1688521913678706e-06,
+      "loss": 0.6311,
+      "step": 5864
+    },
+    {
+      "epoch": 1.6246537396121883,
+      "grad_norm": 0.661792516708374,
+      "learning_rate": 4.1685809369237204e-06,
+      "loss": 0.6065,
+      "step": 5865
+    },
+    {
+      "epoch": 1.6249307479224377,
+      "grad_norm": 0.6726290583610535,
+      "learning_rate": 4.1683096470513365e-06,
+      "loss": 0.6425,
+      "step": 5866
+    },
+    {
+      "epoch": 1.625207756232687,
+      "grad_norm": 0.7385438084602356,
+      "learning_rate": 4.168038321756478e-06,
+      "loss": 0.6404,
+      "step": 5867
+    },
+    {
+      "epoch": 1.6254847645429362,
+      "grad_norm": 0.6656984090805054,
+      "learning_rate": 4.167766961044906e-06,
+      "loss": 0.6575,
+      "step": 5868
+    },
+    {
+      "epoch": 1.6257617728531857,
+      "grad_norm": 0.6919022798538208,
+      "learning_rate": 4.167495564922383e-06,
+      "loss": 0.6485,
+      "step": 5869
+    },
+    {
+      "epoch": 1.626038781163435,
+      "grad_norm": 0.7050084471702576,
+      "learning_rate": 4.1672241333946715e-06,
+      "loss": 0.6616,
+      "step": 5870
+    },
+    {
+      "epoch": 1.6263157894736842,
+      "grad_norm": 0.7358623147010803,
+      "learning_rate": 4.166952666467533e-06,
+      "loss": 0.6555,
+      "step": 5871
+    },
+    {
+      "epoch": 1.6265927977839336,
+      "grad_norm": 0.656574010848999,
+      "learning_rate": 4.166681164146733e-06,
+      "loss": 0.633,
+      "step": 5872
+    },
+    {
+      "epoch": 1.6268698060941829,
+      "grad_norm": 0.6728115677833557,
+      "learning_rate": 4.166409626438035e-06,
+      "loss": 0.5646,
+      "step": 5873
+    },
+    {
+      "epoch": 1.627146814404432,
+      "grad_norm": 0.6639528870582581,
+      "learning_rate": 4.166138053347205e-06,
+      "loss": 0.5575,
+      "step": 5874
+    },
+    {
+      "epoch": 1.6274238227146816,
+      "grad_norm": 0.6908798217773438,
+      "learning_rate": 4.1658664448800105e-06,
+      "loss": 0.5618,
+      "step": 5875
+    },
+    {
+      "epoch": 1.6277008310249308,
+      "grad_norm": 0.7277550101280212,
+      "learning_rate": 4.165594801042217e-06,
+      "loss": 0.5629,
+      "step": 5876
+    },
+    {
+      "epoch": 1.62797783933518,
+      "grad_norm": 0.7118584513664246,
+      "learning_rate": 4.165323121839591e-06,
+      "loss": 0.6638,
+      "step": 5877
+    },
+    {
+      "epoch": 1.6282548476454295,
+      "grad_norm": 0.6751552820205688,
+      "learning_rate": 4.165051407277903e-06,
+      "loss": 0.5756,
+      "step": 5878
+    },
+    {
+      "epoch": 1.6285318559556785,
+      "grad_norm": 0.6624975204467773,
+      "learning_rate": 4.164779657362922e-06,
+      "loss": 0.6035,
+      "step": 5879
+    },
+    {
+      "epoch": 1.628808864265928,
+      "grad_norm": 0.682131826877594,
+      "learning_rate": 4.164507872100417e-06,
+      "loss": 0.5847,
+      "step": 5880
+    },
+    {
+      "epoch": 1.6290858725761774,
+      "grad_norm": 0.6885446310043335,
+      "learning_rate": 4.16423605149616e-06,
+      "loss": 0.6109,
+      "step": 5881
+    },
+    {
+      "epoch": 1.6293628808864264,
+      "grad_norm": 0.6840854287147522,
+      "learning_rate": 4.1639641955559205e-06,
+      "loss": 0.6098,
+      "step": 5882
+    },
+    {
+      "epoch": 1.629639889196676,
+      "grad_norm": 0.6697250604629517,
+      "learning_rate": 4.163692304285473e-06,
+      "loss": 0.6197,
+      "step": 5883
+    },
+    {
+      "epoch": 1.6299168975069254,
+      "grad_norm": 0.6881322264671326,
+      "learning_rate": 4.163420377690588e-06,
+      "loss": 0.6151,
+      "step": 5884
+    },
+    {
+      "epoch": 1.6301939058171744,
+      "grad_norm": 0.69287109375,
+      "learning_rate": 4.163148415777042e-06,
+      "loss": 0.6573,
+      "step": 5885
+    },
+    {
+      "epoch": 1.6304709141274238,
+      "grad_norm": 0.6807845234870911,
+      "learning_rate": 4.162876418550606e-06,
+      "loss": 0.6245,
+      "step": 5886
+    },
+    {
+      "epoch": 1.630747922437673,
+      "grad_norm": 0.6973237991333008,
+      "learning_rate": 4.162604386017059e-06,
+      "loss": 0.6596,
+      "step": 5887
+    },
+    {
+      "epoch": 1.6310249307479223,
+      "grad_norm": 0.7001870274543762,
+      "learning_rate": 4.1623323181821736e-06,
+      "loss": 0.6374,
+      "step": 5888
+    },
+    {
+      "epoch": 1.6313019390581718,
+      "grad_norm": 0.6852225065231323,
+      "learning_rate": 4.162060215051729e-06,
+      "loss": 0.6621,
+      "step": 5889
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.7498638033866882,
+      "learning_rate": 4.161788076631501e-06,
+      "loss": 0.6141,
+      "step": 5890
+    },
+    {
+      "epoch": 1.6318559556786703,
+      "grad_norm": 0.7050655484199524,
+      "learning_rate": 4.1615159029272675e-06,
+      "loss": 0.6357,
+      "step": 5891
+    },
+    {
+      "epoch": 1.6321329639889197,
+      "grad_norm": 0.7564597725868225,
+      "learning_rate": 4.161243693944809e-06,
+      "loss": 0.6705,
+      "step": 5892
+    },
+    {
+      "epoch": 1.632409972299169,
+      "grad_norm": 0.6575698852539062,
+      "learning_rate": 4.1609714496899055e-06,
+      "loss": 0.6403,
+      "step": 5893
+    },
+    {
+      "epoch": 1.6326869806094182,
+      "grad_norm": 0.6732252836227417,
+      "learning_rate": 4.160699170168337e-06,
+      "loss": 0.6116,
+      "step": 5894
+    },
+    {
+      "epoch": 1.6329639889196677,
+      "grad_norm": 0.698630690574646,
+      "learning_rate": 4.160426855385882e-06,
+      "loss": 0.6599,
+      "step": 5895
+    },
+    {
+      "epoch": 1.633240997229917,
+      "grad_norm": 0.6858016848564148,
+      "learning_rate": 4.160154505348326e-06,
+      "loss": 0.641,
+      "step": 5896
+    },
+    {
+      "epoch": 1.6335180055401661,
+      "grad_norm": 0.6913908123970032,
+      "learning_rate": 4.159882120061449e-06,
+      "loss": 0.6289,
+      "step": 5897
+    },
+    {
+      "epoch": 1.6337950138504156,
+      "grad_norm": 0.7428088188171387,
+      "learning_rate": 4.159609699531038e-06,
+      "loss": 0.652,
+      "step": 5898
+    },
+    {
+      "epoch": 1.6340720221606648,
+      "grad_norm": 0.7286577820777893,
+      "learning_rate": 4.159337243762872e-06,
+      "loss": 0.6971,
+      "step": 5899
+    },
+    {
+      "epoch": 1.634349030470914,
+      "grad_norm": 0.6914637684822083,
+      "learning_rate": 4.15906475276274e-06,
+      "loss": 0.6216,
+      "step": 5900
+    },
+    {
+      "epoch": 1.6346260387811635,
+      "grad_norm": 0.6664144992828369,
+      "learning_rate": 4.158792226536428e-06,
+      "loss": 0.6421,
+      "step": 5901
+    },
+    {
+      "epoch": 1.6349030470914128,
+      "grad_norm": 0.707007646560669,
+      "learning_rate": 4.158519665089719e-06,
+      "loss": 0.6282,
+      "step": 5902
+    },
+    {
+      "epoch": 1.635180055401662,
+      "grad_norm": 0.6818870902061462,
+      "learning_rate": 4.158247068428403e-06,
+      "loss": 0.6298,
+      "step": 5903
+    },
+    {
+      "epoch": 1.6354570637119115,
+      "grad_norm": 0.7262736558914185,
+      "learning_rate": 4.157974436558267e-06,
+      "loss": 0.671,
+      "step": 5904
+    },
+    {
+      "epoch": 1.6357340720221607,
+      "grad_norm": 0.6701874136924744,
+      "learning_rate": 4.1577017694851e-06,
+      "loss": 0.5893,
+      "step": 5905
+    },
+    {
+      "epoch": 1.63601108033241,
+      "grad_norm": 0.7156980037689209,
+      "learning_rate": 4.1574290672146905e-06,
+      "loss": 0.6744,
+      "step": 5906
+    },
+    {
+      "epoch": 1.6362880886426594,
+      "grad_norm": 0.703990638256073,
+      "learning_rate": 4.157156329752829e-06,
+      "loss": 0.6432,
+      "step": 5907
+    },
+    {
+      "epoch": 1.6365650969529086,
+      "grad_norm": 0.7014253735542297,
+      "learning_rate": 4.156883557105308e-06,
+      "loss": 0.6514,
+      "step": 5908
+    },
+    {
+      "epoch": 1.6368421052631579,
+      "grad_norm": 0.6927618384361267,
+      "learning_rate": 4.1566107492779175e-06,
+      "loss": 0.6527,
+      "step": 5909
+    },
+    {
+      "epoch": 1.6371191135734073,
+      "grad_norm": 0.7299452424049377,
+      "learning_rate": 4.156337906276449e-06,
+      "loss": 0.6105,
+      "step": 5910
+    },
+    {
+      "epoch": 1.6373961218836564,
+      "grad_norm": 0.7313836812973022,
+      "learning_rate": 4.156065028106699e-06,
+      "loss": 0.6661,
+      "step": 5911
+    },
+    {
+      "epoch": 1.6376731301939058,
+      "grad_norm": 0.7255539298057556,
+      "learning_rate": 4.1557921147744575e-06,
+      "loss": 0.585,
+      "step": 5912
+    },
+    {
+      "epoch": 1.6379501385041553,
+      "grad_norm": 0.7102248668670654,
+      "learning_rate": 4.155519166285522e-06,
+      "loss": 0.5964,
+      "step": 5913
+    },
+    {
+      "epoch": 1.6382271468144043,
+      "grad_norm": 0.7126936316490173,
+      "learning_rate": 4.155246182645686e-06,
+      "loss": 0.6335,
+      "step": 5914
+    },
+    {
+      "epoch": 1.6385041551246537,
+      "grad_norm": 0.6877632737159729,
+      "learning_rate": 4.154973163860747e-06,
+      "loss": 0.6224,
+      "step": 5915
+    },
+    {
+      "epoch": 1.6387811634349032,
+      "grad_norm": 0.7150420546531677,
+      "learning_rate": 4.154700109936501e-06,
+      "loss": 0.6443,
+      "step": 5916
+    },
+    {
+      "epoch": 1.6390581717451522,
+      "grad_norm": 0.7128241658210754,
+      "learning_rate": 4.1544270208787465e-06,
+      "loss": 0.6616,
+      "step": 5917
+    },
+    {
+      "epoch": 1.6393351800554017,
+      "grad_norm": 0.7850664258003235,
+      "learning_rate": 4.154153896693282e-06,
+      "loss": 0.6313,
+      "step": 5918
+    },
+    {
+      "epoch": 1.639612188365651,
+      "grad_norm": 0.7086406350135803,
+      "learning_rate": 4.153880737385905e-06,
+      "loss": 0.6439,
+      "step": 5919
+    },
+    {
+      "epoch": 1.6398891966759002,
+      "grad_norm": 0.6542023420333862,
+      "learning_rate": 4.153607542962417e-06,
+      "loss": 0.5952,
+      "step": 5920
+    },
+    {
+      "epoch": 1.6401662049861496,
+      "grad_norm": 0.7138886451721191,
+      "learning_rate": 4.153334313428616e-06,
+      "loss": 0.6411,
+      "step": 5921
+    },
+    {
+      "epoch": 1.6404432132963989,
+      "grad_norm": 0.6809943318367004,
+      "learning_rate": 4.153061048790307e-06,
+      "loss": 0.6428,
+      "step": 5922
+    },
+    {
+      "epoch": 1.640720221606648,
+      "grad_norm": 0.6998105645179749,
+      "learning_rate": 4.152787749053292e-06,
+      "loss": 0.634,
+      "step": 5923
+    },
+    {
+      "epoch": 1.6409972299168976,
+      "grad_norm": 0.6871700882911682,
+      "learning_rate": 4.152514414223371e-06,
+      "loss": 0.5989,
+      "step": 5924
+    },
+    {
+      "epoch": 1.6412742382271468,
+      "grad_norm": 0.6837252974510193,
+      "learning_rate": 4.152241044306347e-06,
+      "loss": 0.5922,
+      "step": 5925
+    },
+    {
+      "epoch": 1.641551246537396,
+      "grad_norm": 0.7211995720863342,
+      "learning_rate": 4.151967639308029e-06,
+      "loss": 0.634,
+      "step": 5926
+    },
+    {
+      "epoch": 1.6418282548476455,
+      "grad_norm": 0.6932203769683838,
+      "learning_rate": 4.151694199234218e-06,
+      "loss": 0.6271,
+      "step": 5927
+    },
+    {
+      "epoch": 1.6421052631578947,
+      "grad_norm": 0.6809141635894775,
+      "learning_rate": 4.151420724090722e-06,
+      "loss": 0.6408,
+      "step": 5928
+    },
+    {
+      "epoch": 1.642382271468144,
+      "grad_norm": 0.6621232032775879,
+      "learning_rate": 4.151147213883346e-06,
+      "loss": 0.6117,
+      "step": 5929
+    },
+    {
+      "epoch": 1.6426592797783934,
+      "grad_norm": 0.6841561794281006,
+      "learning_rate": 4.150873668617899e-06,
+      "loss": 0.6742,
+      "step": 5930
+    },
+    {
+      "epoch": 1.6429362880886427,
+      "grad_norm": 0.713799774646759,
+      "learning_rate": 4.150600088300188e-06,
+      "loss": 0.6447,
+      "step": 5931
+    },
+    {
+      "epoch": 1.643213296398892,
+      "grad_norm": 0.7068320512771606,
+      "learning_rate": 4.150326472936021e-06,
+      "loss": 0.6563,
+      "step": 5932
+    },
+    {
+      "epoch": 1.6434903047091414,
+      "grad_norm": 0.736694872379303,
+      "learning_rate": 4.150052822531209e-06,
+      "loss": 0.6337,
+      "step": 5933
+    },
+    {
+      "epoch": 1.6437673130193906,
+      "grad_norm": 0.6896104216575623,
+      "learning_rate": 4.149779137091562e-06,
+      "loss": 0.6386,
+      "step": 5934
+    },
+    {
+      "epoch": 1.6440443213296398,
+      "grad_norm": 0.709987998008728,
+      "learning_rate": 4.14950541662289e-06,
+      "loss": 0.6438,
+      "step": 5935
+    },
+    {
+      "epoch": 1.6443213296398893,
+      "grad_norm": 0.7147083282470703,
+      "learning_rate": 4.149231661131007e-06,
+      "loss": 0.6226,
+      "step": 5936
+    },
+    {
+      "epoch": 1.6445983379501385,
+      "grad_norm": 0.7098004221916199,
+      "learning_rate": 4.148957870621723e-06,
+      "loss": 0.6436,
+      "step": 5937
+    },
+    {
+      "epoch": 1.6448753462603878,
+      "grad_norm": 0.677734375,
+      "learning_rate": 4.148684045100853e-06,
+      "loss": 0.5906,
+      "step": 5938
+    },
+    {
+      "epoch": 1.6451523545706372,
+      "grad_norm": 0.740846574306488,
+      "learning_rate": 4.148410184574211e-06,
+      "loss": 0.6357,
+      "step": 5939
+    },
+    {
+      "epoch": 1.6454293628808865,
+      "grad_norm": 0.6789989471435547,
+      "learning_rate": 4.148136289047609e-06,
+      "loss": 0.6532,
+      "step": 5940
+    },
+    {
+      "epoch": 1.6457063711911357,
+      "grad_norm": 0.6899898052215576,
+      "learning_rate": 4.147862358526867e-06,
+      "loss": 0.6571,
+      "step": 5941
+    },
+    {
+      "epoch": 1.6459833795013852,
+      "grad_norm": 0.6977363228797913,
+      "learning_rate": 4.147588393017798e-06,
+      "loss": 0.6629,
+      "step": 5942
+    },
+    {
+      "epoch": 1.6462603878116342,
+      "grad_norm": 0.7238714098930359,
+      "learning_rate": 4.147314392526221e-06,
+      "loss": 0.6504,
+      "step": 5943
+    },
+    {
+      "epoch": 1.6465373961218837,
+      "grad_norm": 0.6461185216903687,
+      "learning_rate": 4.147040357057951e-06,
+      "loss": 0.5962,
+      "step": 5944
+    },
+    {
+      "epoch": 1.6468144044321331,
+      "grad_norm": 0.7050908803939819,
+      "learning_rate": 4.146766286618808e-06,
+      "loss": 0.6689,
+      "step": 5945
+    },
+    {
+      "epoch": 1.6470914127423821,
+      "grad_norm": 0.6823935508728027,
+      "learning_rate": 4.146492181214613e-06,
+      "loss": 0.6609,
+      "step": 5946
+    },
+    {
+      "epoch": 1.6473684210526316,
+      "grad_norm": 0.7076802253723145,
+      "learning_rate": 4.146218040851185e-06,
+      "loss": 0.6364,
+      "step": 5947
+    },
+    {
+      "epoch": 1.6476454293628808,
+      "grad_norm": 0.6914806365966797,
+      "learning_rate": 4.145943865534341e-06,
+      "loss": 0.6272,
+      "step": 5948
+    },
+    {
+      "epoch": 1.64792243767313,
+      "grad_norm": 0.6949312090873718,
+      "learning_rate": 4.145669655269908e-06,
+      "loss": 0.6245,
+      "step": 5949
+    },
+    {
+      "epoch": 1.6481994459833795,
+      "grad_norm": 0.6770291924476624,
+      "learning_rate": 4.145395410063704e-06,
+      "loss": 0.6596,
+      "step": 5950
+    },
+    {
+      "epoch": 1.6484764542936288,
+      "grad_norm": 0.7212213277816772,
+      "learning_rate": 4.145121129921554e-06,
+      "loss": 0.6292,
+      "step": 5951
+    },
+    {
+      "epoch": 1.648753462603878,
+      "grad_norm": 0.6756653785705566,
+      "learning_rate": 4.144846814849282e-06,
+      "loss": 0.6754,
+      "step": 5952
+    },
+    {
+      "epoch": 1.6490304709141275,
+      "grad_norm": 0.7480176091194153,
+      "learning_rate": 4.144572464852711e-06,
+      "loss": 0.689,
+      "step": 5953
+    },
+    {
+      "epoch": 1.6493074792243767,
+      "grad_norm": 0.6765139698982239,
+      "learning_rate": 4.144298079937668e-06,
+      "loss": 0.5313,
+      "step": 5954
+    },
+    {
+      "epoch": 1.649584487534626,
+      "grad_norm": 0.7239553928375244,
+      "learning_rate": 4.144023660109976e-06,
+      "loss": 0.6449,
+      "step": 5955
+    },
+    {
+      "epoch": 1.6498614958448754,
+      "grad_norm": 0.6811375617980957,
+      "learning_rate": 4.143749205375464e-06,
+      "loss": 0.5836,
+      "step": 5956
+    },
+    {
+      "epoch": 1.6501385041551246,
+      "grad_norm": 0.7435874938964844,
+      "learning_rate": 4.143474715739958e-06,
+      "loss": 0.615,
+      "step": 5957
+    },
+    {
+      "epoch": 1.6504155124653739,
+      "grad_norm": 0.7210352420806885,
+      "learning_rate": 4.143200191209287e-06,
+      "loss": 0.6824,
+      "step": 5958
+    },
+    {
+      "epoch": 1.6506925207756233,
+      "grad_norm": 0.675553023815155,
+      "learning_rate": 4.142925631789281e-06,
+      "loss": 0.6526,
+      "step": 5959
+    },
+    {
+      "epoch": 1.6509695290858726,
+      "grad_norm": 0.6919329762458801,
+      "learning_rate": 4.142651037485766e-06,
+      "loss": 0.6208,
+      "step": 5960
+    },
+    {
+      "epoch": 1.6512465373961218,
+      "grad_norm": 0.6395980715751648,
+      "learning_rate": 4.142376408304577e-06,
+      "loss": 0.5776,
+      "step": 5961
+    },
+    {
+      "epoch": 1.6515235457063713,
+      "grad_norm": 0.6764992475509644,
+      "learning_rate": 4.142101744251541e-06,
+      "loss": 0.5655,
+      "step": 5962
+    },
+    {
+      "epoch": 1.6518005540166205,
+      "grad_norm": 0.6732890605926514,
+      "learning_rate": 4.141827045332493e-06,
+      "loss": 0.6311,
+      "step": 5963
+    },
+    {
+      "epoch": 1.6520775623268698,
+      "grad_norm": 0.6796002984046936,
+      "learning_rate": 4.141552311553262e-06,
+      "loss": 0.5606,
+      "step": 5964
+    },
+    {
+      "epoch": 1.6523545706371192,
+      "grad_norm": 0.7205073833465576,
+      "learning_rate": 4.141277542919685e-06,
+      "loss": 0.6066,
+      "step": 5965
+    },
+    {
+      "epoch": 1.6526315789473685,
+      "grad_norm": 0.6774910092353821,
+      "learning_rate": 4.141002739437594e-06,
+      "loss": 0.6294,
+      "step": 5966
+    },
+    {
+      "epoch": 1.6529085872576177,
+      "grad_norm": 0.6876561641693115,
+      "learning_rate": 4.140727901112824e-06,
+      "loss": 0.6169,
+      "step": 5967
+    },
+    {
+      "epoch": 1.6531855955678671,
+      "grad_norm": 0.7015146017074585,
+      "learning_rate": 4.140453027951211e-06,
+      "loss": 0.6245,
+      "step": 5968
+    },
+    {
+      "epoch": 1.6534626038781164,
+      "grad_norm": 0.6706700325012207,
+      "learning_rate": 4.140178119958591e-06,
+      "loss": 0.623,
+      "step": 5969
+    },
+    {
+      "epoch": 1.6537396121883656,
+      "grad_norm": 0.6837378740310669,
+      "learning_rate": 4.1399031771408005e-06,
+      "loss": 0.6311,
+      "step": 5970
+    },
+    {
+      "epoch": 1.654016620498615,
+      "grad_norm": 0.7064226865768433,
+      "learning_rate": 4.139628199503678e-06,
+      "loss": 0.6196,
+      "step": 5971
+    },
+    {
+      "epoch": 1.654293628808864,
+      "grad_norm": 0.7537009119987488,
+      "learning_rate": 4.139353187053061e-06,
+      "loss": 0.6401,
+      "step": 5972
+    },
+    {
+      "epoch": 1.6545706371191136,
+      "grad_norm": 0.694089949131012,
+      "learning_rate": 4.13907813979479e-06,
+      "loss": 0.6339,
+      "step": 5973
+    },
+    {
+      "epoch": 1.654847645429363,
+      "grad_norm": 0.7288523316383362,
+      "learning_rate": 4.138803057734705e-06,
+      "loss": 0.5988,
+      "step": 5974
+    },
+    {
+      "epoch": 1.655124653739612,
+      "grad_norm": 0.7179703116416931,
+      "learning_rate": 4.138527940878645e-06,
+      "loss": 0.6612,
+      "step": 5975
+    },
+    {
+      "epoch": 1.6554016620498615,
+      "grad_norm": 0.6668789982795715,
+      "learning_rate": 4.138252789232453e-06,
+      "loss": 0.6397,
+      "step": 5976
+    },
+    {
+      "epoch": 1.655678670360111,
+      "grad_norm": 0.6644570231437683,
+      "learning_rate": 4.137977602801971e-06,
+      "loss": 0.6482,
+      "step": 5977
+    },
+    {
+      "epoch": 1.65595567867036,
+      "grad_norm": 0.6780247092247009,
+      "learning_rate": 4.13770238159304e-06,
+      "loss": 0.6181,
+      "step": 5978
+    },
+    {
+      "epoch": 1.6562326869806094,
+      "grad_norm": 0.7090055346488953,
+      "learning_rate": 4.1374271256115066e-06,
+      "loss": 0.607,
+      "step": 5979
+    },
+    {
+      "epoch": 1.6565096952908587,
+      "grad_norm": 0.7301034927368164,
+      "learning_rate": 4.137151834863213e-06,
+      "loss": 0.6075,
+      "step": 5980
+    },
+    {
+      "epoch": 1.656786703601108,
+      "grad_norm": 0.6959425806999207,
+      "learning_rate": 4.136876509354005e-06,
+      "loss": 0.6439,
+      "step": 5981
+    },
+    {
+      "epoch": 1.6570637119113574,
+      "grad_norm": 0.709457516670227,
+      "learning_rate": 4.13660114908973e-06,
+      "loss": 0.6018,
+      "step": 5982
+    },
+    {
+      "epoch": 1.6573407202216066,
+      "grad_norm": 0.7351607084274292,
+      "learning_rate": 4.136325754076232e-06,
+      "loss": 0.6567,
+      "step": 5983
+    },
+    {
+      "epoch": 1.6576177285318558,
+      "grad_norm": 0.7283446192741394,
+      "learning_rate": 4.136050324319359e-06,
+      "loss": 0.6451,
+      "step": 5984
+    },
+    {
+      "epoch": 1.6578947368421053,
+      "grad_norm": 0.6645892858505249,
+      "learning_rate": 4.135774859824962e-06,
+      "loss": 0.622,
+      "step": 5985
+    },
+    {
+      "epoch": 1.6581717451523545,
+      "grad_norm": 0.7328609228134155,
+      "learning_rate": 4.135499360598884e-06,
+      "loss": 0.6407,
+      "step": 5986
+    },
+    {
+      "epoch": 1.6584487534626038,
+      "grad_norm": 0.7020002603530884,
+      "learning_rate": 4.13522382664698e-06,
+      "loss": 0.6337,
+      "step": 5987
+    },
+    {
+      "epoch": 1.6587257617728532,
+      "grad_norm": 0.6328570246696472,
+      "learning_rate": 4.1349482579750984e-06,
+      "loss": 0.5864,
+      "step": 5988
+    },
+    {
+      "epoch": 1.6590027700831025,
+      "grad_norm": 0.6584644317626953,
+      "learning_rate": 4.13467265458909e-06,
+      "loss": 0.6348,
+      "step": 5989
+    },
+    {
+      "epoch": 1.6592797783933517,
+      "grad_norm": 0.7188389897346497,
+      "learning_rate": 4.134397016494807e-06,
+      "loss": 0.6192,
+      "step": 5990
+    },
+    {
+      "epoch": 1.6595567867036012,
+      "grad_norm": 0.6980937719345093,
+      "learning_rate": 4.1341213436981e-06,
+      "loss": 0.6191,
+      "step": 5991
+    },
+    {
+      "epoch": 1.6598337950138504,
+      "grad_norm": 0.6944648027420044,
+      "learning_rate": 4.133845636204825e-06,
+      "loss": 0.608,
+      "step": 5992
+    },
+    {
+      "epoch": 1.6601108033240997,
+      "grad_norm": 0.7136276364326477,
+      "learning_rate": 4.1335698940208344e-06,
+      "loss": 0.6054,
+      "step": 5993
+    },
+    {
+      "epoch": 1.6603878116343491,
+      "grad_norm": 0.6488214731216431,
+      "learning_rate": 4.133294117151983e-06,
+      "loss": 0.5935,
+      "step": 5994
+    },
+    {
+      "epoch": 1.6606648199445984,
+      "grad_norm": 0.6956791877746582,
+      "learning_rate": 4.1330183056041275e-06,
+      "loss": 0.6339,
+      "step": 5995
+    },
+    {
+      "epoch": 1.6609418282548476,
+      "grad_norm": 0.7471428513526917,
+      "learning_rate": 4.132742459383122e-06,
+      "loss": 0.5852,
+      "step": 5996
+    },
+    {
+      "epoch": 1.661218836565097,
+      "grad_norm": 0.71979159116745,
+      "learning_rate": 4.132466578494825e-06,
+      "loss": 0.6563,
+      "step": 5997
+    },
+    {
+      "epoch": 1.6614958448753463,
+      "grad_norm": 0.6717140674591064,
+      "learning_rate": 4.132190662945093e-06,
+      "loss": 0.6134,
+      "step": 5998
+    },
+    {
+      "epoch": 1.6617728531855955,
+      "grad_norm": 0.6809756755828857,
+      "learning_rate": 4.131914712739786e-06,
+      "loss": 0.6311,
+      "step": 5999
+    },
+    {
+      "epoch": 1.662049861495845,
+      "grad_norm": 0.6761506795883179,
+      "learning_rate": 4.131638727884762e-06,
+      "loss": 0.6154,
+      "step": 6000
+    },
+    {
+      "epoch": 1.6623268698060942,
+      "grad_norm": 0.6920652389526367,
+      "learning_rate": 4.131362708385881e-06,
+      "loss": 0.6319,
+      "step": 6001
+    },
+    {
+      "epoch": 1.6626038781163435,
+      "grad_norm": 0.7034053206443787,
+      "learning_rate": 4.131086654249003e-06,
+      "loss": 0.5533,
+      "step": 6002
+    },
+    {
+      "epoch": 1.662880886426593,
+      "grad_norm": 0.7007937431335449,
+      "learning_rate": 4.130810565479991e-06,
+      "loss": 0.5813,
+      "step": 6003
+    },
+    {
+      "epoch": 1.663157894736842,
+      "grad_norm": 0.6734853386878967,
+      "learning_rate": 4.130534442084706e-06,
+      "loss": 0.5991,
+      "step": 6004
+    },
+    {
+      "epoch": 1.6634349030470914,
+      "grad_norm": 0.7057620286941528,
+      "learning_rate": 4.130258284069009e-06,
+      "loss": 0.5601,
+      "step": 6005
+    },
+    {
+      "epoch": 1.6637119113573409,
+      "grad_norm": 0.6708484292030334,
+      "learning_rate": 4.129982091438768e-06,
+      "loss": 0.613,
+      "step": 6006
+    },
+    {
+      "epoch": 1.6639889196675899,
+      "grad_norm": 0.682408332824707,
+      "learning_rate": 4.129705864199844e-06,
+      "loss": 0.5747,
+      "step": 6007
+    },
+    {
+      "epoch": 1.6642659279778393,
+      "grad_norm": 0.7112895846366882,
+      "learning_rate": 4.129429602358101e-06,
+      "loss": 0.6028,
+      "step": 6008
+    },
+    {
+      "epoch": 1.6645429362880888,
+      "grad_norm": 0.721603512763977,
+      "learning_rate": 4.129153305919408e-06,
+      "loss": 0.6482,
+      "step": 6009
+    },
+    {
+      "epoch": 1.6648199445983378,
+      "grad_norm": 0.6883713603019714,
+      "learning_rate": 4.12887697488963e-06,
+      "loss": 0.6276,
+      "step": 6010
+    },
+    {
+      "epoch": 1.6650969529085873,
+      "grad_norm": 0.6822021007537842,
+      "learning_rate": 4.128600609274633e-06,
+      "loss": 0.6277,
+      "step": 6011
+    },
+    {
+      "epoch": 1.6653739612188365,
+      "grad_norm": 0.7424676418304443,
+      "learning_rate": 4.128324209080288e-06,
+      "loss": 0.6671,
+      "step": 6012
+    },
+    {
+      "epoch": 1.6656509695290858,
+      "grad_norm": 0.6871791481971741,
+      "learning_rate": 4.12804777431246e-06,
+      "loss": 0.6383,
+      "step": 6013
+    },
+    {
+      "epoch": 1.6659279778393352,
+      "grad_norm": 0.7980043292045593,
+      "learning_rate": 4.127771304977021e-06,
+      "loss": 0.6583,
+      "step": 6014
+    },
+    {
+      "epoch": 1.6662049861495845,
+      "grad_norm": 0.6801428198814392,
+      "learning_rate": 4.12749480107984e-06,
+      "loss": 0.6019,
+      "step": 6015
+    },
+    {
+      "epoch": 1.6664819944598337,
+      "grad_norm": 0.6864743828773499,
+      "learning_rate": 4.127218262626788e-06,
+      "loss": 0.6221,
+      "step": 6016
+    },
+    {
+      "epoch": 1.6667590027700832,
+      "grad_norm": 0.7071241140365601,
+      "learning_rate": 4.126941689623737e-06,
+      "loss": 0.6464,
+      "step": 6017
+    },
+    {
+      "epoch": 1.6670360110803324,
+      "grad_norm": 0.6872143149375916,
+      "learning_rate": 4.126665082076559e-06,
+      "loss": 0.5996,
+      "step": 6018
+    },
+    {
+      "epoch": 1.6673130193905816,
+      "grad_norm": 0.6813023686408997,
+      "learning_rate": 4.126388439991128e-06,
+      "loss": 0.6098,
+      "step": 6019
+    },
+    {
+      "epoch": 1.667590027700831,
+      "grad_norm": 0.7234155535697937,
+      "learning_rate": 4.126111763373316e-06,
+      "loss": 0.5886,
+      "step": 6020
+    },
+    {
+      "epoch": 1.6678670360110803,
+      "grad_norm": 0.7097522020339966,
+      "learning_rate": 4.125835052228999e-06,
+      "loss": 0.6022,
+      "step": 6021
+    },
+    {
+      "epoch": 1.6681440443213296,
+      "grad_norm": 0.7106451392173767,
+      "learning_rate": 4.125558306564053e-06,
+      "loss": 0.6189,
+      "step": 6022
+    },
+    {
+      "epoch": 1.668421052631579,
+      "grad_norm": 0.6844403743743896,
+      "learning_rate": 4.125281526384352e-06,
+      "loss": 0.6278,
+      "step": 6023
+    },
+    {
+      "epoch": 1.6686980609418283,
+      "grad_norm": 0.6791717410087585,
+      "learning_rate": 4.125004711695773e-06,
+      "loss": 0.6461,
+      "step": 6024
+    },
+    {
+      "epoch": 1.6689750692520775,
+      "grad_norm": 0.6500537991523743,
+      "learning_rate": 4.124727862504196e-06,
+      "loss": 0.5991,
+      "step": 6025
+    },
+    {
+      "epoch": 1.669252077562327,
+      "grad_norm": 0.7625722289085388,
+      "learning_rate": 4.124450978815495e-06,
+      "loss": 0.6168,
+      "step": 6026
+    },
+    {
+      "epoch": 1.6695290858725762,
+      "grad_norm": 0.6701847910881042,
+      "learning_rate": 4.124174060635553e-06,
+      "loss": 0.6208,
+      "step": 6027
+    },
+    {
+      "epoch": 1.6698060941828254,
+      "grad_norm": 0.7106515765190125,
+      "learning_rate": 4.123897107970248e-06,
+      "loss": 0.6424,
+      "step": 6028
+    },
+    {
+      "epoch": 1.670083102493075,
+      "grad_norm": 0.6984321475028992,
+      "learning_rate": 4.123620120825459e-06,
+      "loss": 0.6732,
+      "step": 6029
+    },
+    {
+      "epoch": 1.6703601108033241,
+      "grad_norm": 0.7095187306404114,
+      "learning_rate": 4.12334309920707e-06,
+      "loss": 0.6538,
+      "step": 6030
+    },
+    {
+      "epoch": 1.6706371191135734,
+      "grad_norm": 0.7166954874992371,
+      "learning_rate": 4.1230660431209615e-06,
+      "loss": 0.6093,
+      "step": 6031
+    },
+    {
+      "epoch": 1.6709141274238228,
+      "grad_norm": 0.7276129722595215,
+      "learning_rate": 4.1227889525730156e-06,
+      "loss": 0.6483,
+      "step": 6032
+    },
+    {
+      "epoch": 1.671191135734072,
+      "grad_norm": 0.6965504884719849,
+      "learning_rate": 4.122511827569115e-06,
+      "loss": 0.5949,
+      "step": 6033
+    },
+    {
+      "epoch": 1.6714681440443213,
+      "grad_norm": 0.6962923407554626,
+      "learning_rate": 4.122234668115147e-06,
+      "loss": 0.6356,
+      "step": 6034
+    },
+    {
+      "epoch": 1.6717451523545708,
+      "grad_norm": 0.7430986762046814,
+      "learning_rate": 4.121957474216992e-06,
+      "loss": 0.6417,
+      "step": 6035
+    },
+    {
+      "epoch": 1.6720221606648198,
+      "grad_norm": 0.7319906949996948,
+      "learning_rate": 4.121680245880539e-06,
+      "loss": 0.6445,
+      "step": 6036
+    },
+    {
+      "epoch": 1.6722991689750693,
+      "grad_norm": 0.7240617275238037,
+      "learning_rate": 4.121402983111673e-06,
+      "loss": 0.6573,
+      "step": 6037
+    },
+    {
+      "epoch": 1.6725761772853187,
+      "grad_norm": 0.7075072526931763,
+      "learning_rate": 4.121125685916281e-06,
+      "loss": 0.6018,
+      "step": 6038
+    },
+    {
+      "epoch": 1.6728531855955677,
+      "grad_norm": 0.6763313412666321,
+      "learning_rate": 4.1208483543002514e-06,
+      "loss": 0.6343,
+      "step": 6039
+    },
+    {
+      "epoch": 1.6731301939058172,
+      "grad_norm": 0.7229200601577759,
+      "learning_rate": 4.120570988269472e-06,
+      "loss": 0.5932,
+      "step": 6040
+    },
+    {
+      "epoch": 1.6734072022160666,
+      "grad_norm": 0.7736415266990662,
+      "learning_rate": 4.120293587829831e-06,
+      "loss": 0.6361,
+      "step": 6041
+    },
+    {
+      "epoch": 1.6736842105263157,
+      "grad_norm": 0.6882472038269043,
+      "learning_rate": 4.12001615298722e-06,
+      "loss": 0.6095,
+      "step": 6042
+    },
+    {
+      "epoch": 1.6739612188365651,
+      "grad_norm": 0.7033710479736328,
+      "learning_rate": 4.119738683747528e-06,
+      "loss": 0.6688,
+      "step": 6043
+    },
+    {
+      "epoch": 1.6742382271468144,
+      "grad_norm": 0.6977055072784424,
+      "learning_rate": 4.119461180116649e-06,
+      "loss": 0.6318,
+      "step": 6044
+    },
+    {
+      "epoch": 1.6745152354570636,
+      "grad_norm": 0.7398744225502014,
+      "learning_rate": 4.1191836421004726e-06,
+      "loss": 0.6096,
+      "step": 6045
+    },
+    {
+      "epoch": 1.674792243767313,
+      "grad_norm": 0.6894487738609314,
+      "learning_rate": 4.118906069704892e-06,
+      "loss": 0.6415,
+      "step": 6046
+    },
+    {
+      "epoch": 1.6750692520775623,
+      "grad_norm": 0.6845730543136597,
+      "learning_rate": 4.118628462935803e-06,
+      "loss": 0.6172,
+      "step": 6047
+    },
+    {
+      "epoch": 1.6753462603878115,
+      "grad_norm": 0.6749402284622192,
+      "learning_rate": 4.1183508217990964e-06,
+      "loss": 0.6417,
+      "step": 6048
+    },
+    {
+      "epoch": 1.675623268698061,
+      "grad_norm": 0.6568241715431213,
+      "learning_rate": 4.118073146300671e-06,
+      "loss": 0.6279,
+      "step": 6049
+    },
+    {
+      "epoch": 1.6759002770083102,
+      "grad_norm": 0.7095878720283508,
+      "learning_rate": 4.117795436446418e-06,
+      "loss": 0.6294,
+      "step": 6050
+    },
+    {
+      "epoch": 1.6761772853185595,
+      "grad_norm": 0.7073789238929749,
+      "learning_rate": 4.117517692242239e-06,
+      "loss": 0.637,
+      "step": 6051
+    },
+    {
+      "epoch": 1.676454293628809,
+      "grad_norm": 0.6689110398292542,
+      "learning_rate": 4.117239913694027e-06,
+      "loss": 0.6159,
+      "step": 6052
+    },
+    {
+      "epoch": 1.6767313019390582,
+      "grad_norm": 0.7140373587608337,
+      "learning_rate": 4.116962100807682e-06,
+      "loss": 0.594,
+      "step": 6053
+    },
+    {
+      "epoch": 1.6770083102493074,
+      "grad_norm": 0.7038289904594421,
+      "learning_rate": 4.116684253589102e-06,
+      "loss": 0.5982,
+      "step": 6054
+    },
+    {
+      "epoch": 1.6772853185595569,
+      "grad_norm": 0.6871324777603149,
+      "learning_rate": 4.116406372044188e-06,
+      "loss": 0.6365,
+      "step": 6055
+    },
+    {
+      "epoch": 1.677562326869806,
+      "grad_norm": 0.6705648899078369,
+      "learning_rate": 4.1161284561788385e-06,
+      "loss": 0.6244,
+      "step": 6056
+    },
+    {
+      "epoch": 1.6778393351800553,
+      "grad_norm": 0.6981586813926697,
+      "learning_rate": 4.115850505998954e-06,
+      "loss": 0.6711,
+      "step": 6057
+    },
+    {
+      "epoch": 1.6781163434903048,
+      "grad_norm": 0.6764851212501526,
+      "learning_rate": 4.115572521510438e-06,
+      "loss": 0.6196,
+      "step": 6058
+    },
+    {
+      "epoch": 1.678393351800554,
+      "grad_norm": 0.6880233883857727,
+      "learning_rate": 4.11529450271919e-06,
+      "loss": 0.6334,
+      "step": 6059
+    },
+    {
+      "epoch": 1.6786703601108033,
+      "grad_norm": 0.6908855438232422,
+      "learning_rate": 4.115016449631116e-06,
+      "loss": 0.6166,
+      "step": 6060
+    },
+    {
+      "epoch": 1.6789473684210527,
+      "grad_norm": 0.6814182996749878,
+      "learning_rate": 4.1147383622521174e-06,
+      "loss": 0.594,
+      "step": 6061
+    },
+    {
+      "epoch": 1.679224376731302,
+      "grad_norm": 0.6829125285148621,
+      "learning_rate": 4.114460240588101e-06,
+      "loss": 0.5736,
+      "step": 6062
+    },
+    {
+      "epoch": 1.6795013850415512,
+      "grad_norm": 0.7400720119476318,
+      "learning_rate": 4.11418208464497e-06,
+      "loss": 0.6244,
+      "step": 6063
+    },
+    {
+      "epoch": 1.6797783933518007,
+      "grad_norm": 0.7241595983505249,
+      "learning_rate": 4.113903894428632e-06,
+      "loss": 0.6691,
+      "step": 6064
+    },
+    {
+      "epoch": 1.6800554016620497,
+      "grad_norm": 0.7304342985153198,
+      "learning_rate": 4.113625669944992e-06,
+      "loss": 0.6692,
+      "step": 6065
+    },
+    {
+      "epoch": 1.6803324099722992,
+      "grad_norm": 0.7147141098976135,
+      "learning_rate": 4.1133474111999584e-06,
+      "loss": 0.6329,
+      "step": 6066
+    },
+    {
+      "epoch": 1.6806094182825486,
+      "grad_norm": 0.7014601826667786,
+      "learning_rate": 4.1130691181994395e-06,
+      "loss": 0.6164,
+      "step": 6067
+    },
+    {
+      "epoch": 1.6808864265927976,
+      "grad_norm": 0.7008762359619141,
+      "learning_rate": 4.112790790949344e-06,
+      "loss": 0.6395,
+      "step": 6068
+    },
+    {
+      "epoch": 1.681163434903047,
+      "grad_norm": 0.726631760597229,
+      "learning_rate": 4.112512429455581e-06,
+      "loss": 0.6034,
+      "step": 6069
+    },
+    {
+      "epoch": 1.6814404432132966,
+      "grad_norm": 0.6704750061035156,
+      "learning_rate": 4.112234033724061e-06,
+      "loss": 0.6053,
+      "step": 6070
+    },
+    {
+      "epoch": 1.6817174515235456,
+      "grad_norm": 0.6997814178466797,
+      "learning_rate": 4.111955603760696e-06,
+      "loss": 0.6357,
+      "step": 6071
+    },
+    {
+      "epoch": 1.681994459833795,
+      "grad_norm": 0.682888388633728,
+      "learning_rate": 4.111677139571396e-06,
+      "loss": 0.5929,
+      "step": 6072
+    },
+    {
+      "epoch": 1.6822714681440443,
+      "grad_norm": 0.6795300245285034,
+      "learning_rate": 4.111398641162076e-06,
+      "loss": 0.6298,
+      "step": 6073
+    },
+    {
+      "epoch": 1.6825484764542935,
+      "grad_norm": 0.6853460669517517,
+      "learning_rate": 4.111120108538647e-06,
+      "loss": 0.627,
+      "step": 6074
+    },
+    {
+      "epoch": 1.682825484764543,
+      "grad_norm": 0.6448818445205688,
+      "learning_rate": 4.110841541707024e-06,
+      "loss": 0.5968,
+      "step": 6075
+    },
+    {
+      "epoch": 1.6831024930747922,
+      "grad_norm": 0.7064438462257385,
+      "learning_rate": 4.110562940673121e-06,
+      "loss": 0.6412,
+      "step": 6076
+    },
+    {
+      "epoch": 1.6833795013850414,
+      "grad_norm": 0.6554502248764038,
+      "learning_rate": 4.110284305442855e-06,
+      "loss": 0.5882,
+      "step": 6077
+    },
+    {
+      "epoch": 1.683656509695291,
+      "grad_norm": 0.7326308488845825,
+      "learning_rate": 4.110005636022138e-06,
+      "loss": 0.6333,
+      "step": 6078
+    },
+    {
+      "epoch": 1.6839335180055401,
+      "grad_norm": 0.6282971501350403,
+      "learning_rate": 4.109726932416893e-06,
+      "loss": 0.5892,
+      "step": 6079
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.6800413727760315,
+      "learning_rate": 4.109448194633033e-06,
+      "loss": 0.557,
+      "step": 6080
+    },
+    {
+      "epoch": 1.6844875346260388,
+      "grad_norm": 0.6659765243530273,
+      "learning_rate": 4.1091694226764784e-06,
+      "loss": 0.5852,
+      "step": 6081
+    },
+    {
+      "epoch": 1.684764542936288,
+      "grad_norm": 0.6778327226638794,
+      "learning_rate": 4.108890616553147e-06,
+      "loss": 0.6176,
+      "step": 6082
+    },
+    {
+      "epoch": 1.6850415512465373,
+      "grad_norm": 0.707545280456543,
+      "learning_rate": 4.1086117762689595e-06,
+      "loss": 0.6196,
+      "step": 6083
+    },
+    {
+      "epoch": 1.6853185595567868,
+      "grad_norm": 0.7330470085144043,
+      "learning_rate": 4.1083329018298356e-06,
+      "loss": 0.6483,
+      "step": 6084
+    },
+    {
+      "epoch": 1.685595567867036,
+      "grad_norm": 0.7315000891685486,
+      "learning_rate": 4.1080539932416974e-06,
+      "loss": 0.6321,
+      "step": 6085
+    },
+    {
+      "epoch": 1.6858725761772853,
+      "grad_norm": 0.6998733878135681,
+      "learning_rate": 4.107775050510467e-06,
+      "loss": 0.6541,
+      "step": 6086
+    },
+    {
+      "epoch": 1.6861495844875347,
+      "grad_norm": 0.6499519348144531,
+      "learning_rate": 4.107496073642064e-06,
+      "loss": 0.5905,
+      "step": 6087
+    },
+    {
+      "epoch": 1.686426592797784,
+      "grad_norm": 0.7056503295898438,
+      "learning_rate": 4.1072170626424166e-06,
+      "loss": 0.6332,
+      "step": 6088
+    },
+    {
+      "epoch": 1.6867036011080332,
+      "grad_norm": 0.6869191527366638,
+      "learning_rate": 4.106938017517446e-06,
+      "loss": 0.6453,
+      "step": 6089
+    },
+    {
+      "epoch": 1.6869806094182827,
+      "grad_norm": 0.7189652919769287,
+      "learning_rate": 4.106658938273078e-06,
+      "loss": 0.5499,
+      "step": 6090
+    },
+    {
+      "epoch": 1.6872576177285319,
+      "grad_norm": 0.6633101105690002,
+      "learning_rate": 4.106379824915237e-06,
+      "loss": 0.6392,
+      "step": 6091
+    },
+    {
+      "epoch": 1.6875346260387811,
+      "grad_norm": 0.7263966798782349,
+      "learning_rate": 4.106100677449851e-06,
+      "loss": 0.6561,
+      "step": 6092
+    },
+    {
+      "epoch": 1.6878116343490306,
+      "grad_norm": 0.7036128640174866,
+      "learning_rate": 4.105821495882845e-06,
+      "loss": 0.6357,
+      "step": 6093
+    },
+    {
+      "epoch": 1.6880886426592798,
+      "grad_norm": 0.6799331307411194,
+      "learning_rate": 4.105542280220148e-06,
+      "loss": 0.6067,
+      "step": 6094
+    },
+    {
+      "epoch": 1.688365650969529,
+      "grad_norm": 0.6819090843200684,
+      "learning_rate": 4.105263030467689e-06,
+      "loss": 0.6249,
+      "step": 6095
+    },
+    {
+      "epoch": 1.6886426592797785,
+      "grad_norm": 0.6934465169906616,
+      "learning_rate": 4.104983746631396e-06,
+      "loss": 0.6438,
+      "step": 6096
+    },
+    {
+      "epoch": 1.6889196675900275,
+      "grad_norm": 0.6890839338302612,
+      "learning_rate": 4.1047044287172e-06,
+      "loss": 0.6263,
+      "step": 6097
+    },
+    {
+      "epoch": 1.689196675900277,
+      "grad_norm": 0.7147068977355957,
+      "learning_rate": 4.10442507673103e-06,
+      "loss": 0.6331,
+      "step": 6098
+    },
+    {
+      "epoch": 1.6894736842105265,
+      "grad_norm": 0.6808082461357117,
+      "learning_rate": 4.104145690678819e-06,
+      "loss": 0.6405,
+      "step": 6099
+    },
+    {
+      "epoch": 1.6897506925207755,
+      "grad_norm": 0.6937206387519836,
+      "learning_rate": 4.103866270566498e-06,
+      "loss": 0.6308,
+      "step": 6100
+    },
+    {
+      "epoch": 1.690027700831025,
+      "grad_norm": 0.715852677822113,
+      "learning_rate": 4.103586816400001e-06,
+      "loss": 0.614,
+      "step": 6101
+    },
+    {
+      "epoch": 1.6903047091412744,
+      "grad_norm": 0.7389687299728394,
+      "learning_rate": 4.10330732818526e-06,
+      "loss": 0.6356,
+      "step": 6102
+    },
+    {
+      "epoch": 1.6905817174515234,
+      "grad_norm": 0.7088740468025208,
+      "learning_rate": 4.10302780592821e-06,
+      "loss": 0.6506,
+      "step": 6103
+    },
+    {
+      "epoch": 1.6908587257617729,
+      "grad_norm": 0.6934943199157715,
+      "learning_rate": 4.102748249634787e-06,
+      "loss": 0.6239,
+      "step": 6104
+    },
+    {
+      "epoch": 1.6911357340720221,
+      "grad_norm": 0.6720358729362488,
+      "learning_rate": 4.102468659310925e-06,
+      "loss": 0.6261,
+      "step": 6105
+    },
+    {
+      "epoch": 1.6914127423822714,
+      "grad_norm": 0.7021436095237732,
+      "learning_rate": 4.102189034962561e-06,
+      "loss": 0.6418,
+      "step": 6106
+    },
+    {
+      "epoch": 1.6916897506925208,
+      "grad_norm": 0.6990009546279907,
+      "learning_rate": 4.101909376595632e-06,
+      "loss": 0.6781,
+      "step": 6107
+    },
+    {
+      "epoch": 1.69196675900277,
+      "grad_norm": 0.6986948251724243,
+      "learning_rate": 4.101629684216076e-06,
+      "loss": 0.6628,
+      "step": 6108
+    },
+    {
+      "epoch": 1.6922437673130193,
+      "grad_norm": 0.6557834148406982,
+      "learning_rate": 4.101349957829833e-06,
+      "loss": 0.6234,
+      "step": 6109
+    },
+    {
+      "epoch": 1.6925207756232687,
+      "grad_norm": 0.7131497263908386,
+      "learning_rate": 4.1010701974428405e-06,
+      "loss": 0.6557,
+      "step": 6110
+    },
+    {
+      "epoch": 1.692797783933518,
+      "grad_norm": 0.7220354676246643,
+      "learning_rate": 4.100790403061038e-06,
+      "loss": 0.6288,
+      "step": 6111
+    },
+    {
+      "epoch": 1.6930747922437672,
+      "grad_norm": 0.7219642996788025,
+      "learning_rate": 4.100510574690368e-06,
+      "loss": 0.6509,
+      "step": 6112
+    },
+    {
+      "epoch": 1.6933518005540167,
+      "grad_norm": 0.770431399345398,
+      "learning_rate": 4.1002307123367716e-06,
+      "loss": 0.6359,
+      "step": 6113
+    },
+    {
+      "epoch": 1.693628808864266,
+      "grad_norm": 0.682773232460022,
+      "learning_rate": 4.099950816006189e-06,
+      "loss": 0.6319,
+      "step": 6114
+    },
+    {
+      "epoch": 1.6939058171745152,
+      "grad_norm": 0.6756408214569092,
+      "learning_rate": 4.099670885704566e-06,
+      "loss": 0.6226,
+      "step": 6115
+    },
+    {
+      "epoch": 1.6941828254847646,
+      "grad_norm": 0.7353326082229614,
+      "learning_rate": 4.099390921437845e-06,
+      "loss": 0.6501,
+      "step": 6116
+    },
+    {
+      "epoch": 1.6944598337950139,
+      "grad_norm": 0.7328805327415466,
+      "learning_rate": 4.099110923211971e-06,
+      "loss": 0.6325,
+      "step": 6117
+    },
+    {
+      "epoch": 1.694736842105263,
+      "grad_norm": 0.6881809234619141,
+      "learning_rate": 4.098830891032888e-06,
+      "loss": 0.6609,
+      "step": 6118
+    },
+    {
+      "epoch": 1.6950138504155126,
+      "grad_norm": 0.8122984766960144,
+      "learning_rate": 4.098550824906542e-06,
+      "loss": 0.6201,
+      "step": 6119
+    },
+    {
+      "epoch": 1.6952908587257618,
+      "grad_norm": 0.6945459842681885,
+      "learning_rate": 4.098270724838879e-06,
+      "loss": 0.5877,
+      "step": 6120
+    },
+    {
+      "epoch": 1.695567867036011,
+      "grad_norm": 0.6746237874031067,
+      "learning_rate": 4.097990590835847e-06,
+      "loss": 0.6191,
+      "step": 6121
+    },
+    {
+      "epoch": 1.6958448753462605,
+      "grad_norm": 0.7451200485229492,
+      "learning_rate": 4.097710422903395e-06,
+      "loss": 0.6232,
+      "step": 6122
+    },
+    {
+      "epoch": 1.6961218836565097,
+      "grad_norm": 0.7273604273796082,
+      "learning_rate": 4.0974302210474705e-06,
+      "loss": 0.6481,
+      "step": 6123
+    },
+    {
+      "epoch": 1.696398891966759,
+      "grad_norm": 0.6815860271453857,
+      "learning_rate": 4.0971499852740224e-06,
+      "loss": 0.5715,
+      "step": 6124
+    },
+    {
+      "epoch": 1.6966759002770084,
+      "grad_norm": 0.6721625328063965,
+      "learning_rate": 4.096869715589002e-06,
+      "loss": 0.6237,
+      "step": 6125
+    },
+    {
+      "epoch": 1.6969529085872577,
+      "grad_norm": 0.7191078066825867,
+      "learning_rate": 4.096589411998359e-06,
+      "loss": 0.6062,
+      "step": 6126
+    },
+    {
+      "epoch": 1.697229916897507,
+      "grad_norm": 0.7018681168556213,
+      "learning_rate": 4.096309074508046e-06,
+      "loss": 0.6266,
+      "step": 6127
+    },
+    {
+      "epoch": 1.6975069252077564,
+      "grad_norm": 0.6845425963401794,
+      "learning_rate": 4.096028703124014e-06,
+      "loss": 0.6761,
+      "step": 6128
+    },
+    {
+      "epoch": 1.6977839335180054,
+      "grad_norm": 0.7229567170143127,
+      "learning_rate": 4.095748297852218e-06,
+      "loss": 0.5982,
+      "step": 6129
+    },
+    {
+      "epoch": 1.6980609418282548,
+      "grad_norm": 0.6780923008918762,
+      "learning_rate": 4.09546785869861e-06,
+      "loss": 0.6243,
+      "step": 6130
+    },
+    {
+      "epoch": 1.6983379501385043,
+      "grad_norm": 0.7234211564064026,
+      "learning_rate": 4.095187385669145e-06,
+      "loss": 0.6738,
+      "step": 6131
+    },
+    {
+      "epoch": 1.6986149584487533,
+      "grad_norm": 0.6908296346664429,
+      "learning_rate": 4.094906878769779e-06,
+      "loss": 0.6236,
+      "step": 6132
+    },
+    {
+      "epoch": 1.6988919667590028,
+      "grad_norm": 0.6617946624755859,
+      "learning_rate": 4.094626338006466e-06,
+      "loss": 0.6388,
+      "step": 6133
+    },
+    {
+      "epoch": 1.6991689750692522,
+      "grad_norm": 0.7265128493309021,
+      "learning_rate": 4.094345763385163e-06,
+      "loss": 0.6631,
+      "step": 6134
+    },
+    {
+      "epoch": 1.6994459833795013,
+      "grad_norm": 0.6898639798164368,
+      "learning_rate": 4.094065154911829e-06,
+      "loss": 0.6443,
+      "step": 6135
+    },
+    {
+      "epoch": 1.6997229916897507,
+      "grad_norm": 0.6893454194068909,
+      "learning_rate": 4.09378451259242e-06,
+      "loss": 0.6131,
+      "step": 6136
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.71174556016922,
+      "learning_rate": 4.093503836432897e-06,
+      "loss": 0.5904,
+      "step": 6137
+    },
+    {
+      "epoch": 1.7002770083102492,
+      "grad_norm": 0.7047994136810303,
+      "learning_rate": 4.093223126439217e-06,
+      "loss": 0.5866,
+      "step": 6138
+    },
+    {
+      "epoch": 1.7005540166204987,
+      "grad_norm": 0.7247219681739807,
+      "learning_rate": 4.092942382617342e-06,
+      "loss": 0.6405,
+      "step": 6139
+    },
+    {
+      "epoch": 1.700831024930748,
+      "grad_norm": 0.7214411497116089,
+      "learning_rate": 4.092661604973232e-06,
+      "loss": 0.6256,
+      "step": 6140
+    },
+    {
+      "epoch": 1.7011080332409971,
+      "grad_norm": 0.6483976244926453,
+      "learning_rate": 4.09238079351285e-06,
+      "loss": 0.6162,
+      "step": 6141
+    },
+    {
+      "epoch": 1.7013850415512466,
+      "grad_norm": 0.6869223713874817,
+      "learning_rate": 4.092099948242156e-06,
+      "loss": 0.6235,
+      "step": 6142
+    },
+    {
+      "epoch": 1.7016620498614958,
+      "grad_norm": 0.7167564034461975,
+      "learning_rate": 4.0918190691671145e-06,
+      "loss": 0.6512,
+      "step": 6143
+    },
+    {
+      "epoch": 1.701939058171745,
+      "grad_norm": 0.6702075600624084,
+      "learning_rate": 4.09153815629369e-06,
+      "loss": 0.6293,
+      "step": 6144
+    },
+    {
+      "epoch": 1.7022160664819945,
+      "grad_norm": 0.6951959729194641,
+      "learning_rate": 4.091257209627845e-06,
+      "loss": 0.5995,
+      "step": 6145
+    },
+    {
+      "epoch": 1.7024930747922438,
+      "grad_norm": 0.7316385507583618,
+      "learning_rate": 4.090976229175546e-06,
+      "loss": 0.6548,
+      "step": 6146
+    },
+    {
+      "epoch": 1.702770083102493,
+      "grad_norm": 0.7311107516288757,
+      "learning_rate": 4.090695214942759e-06,
+      "loss": 0.6865,
+      "step": 6147
+    },
+    {
+      "epoch": 1.7030470914127425,
+      "grad_norm": 0.663872241973877,
+      "learning_rate": 4.09041416693545e-06,
+      "loss": 0.6137,
+      "step": 6148
+    },
+    {
+      "epoch": 1.7033240997229917,
+      "grad_norm": 0.6457836627960205,
+      "learning_rate": 4.0901330851595874e-06,
+      "loss": 0.6149,
+      "step": 6149
+    },
+    {
+      "epoch": 1.703601108033241,
+      "grad_norm": 0.6875283122062683,
+      "learning_rate": 4.089851969621138e-06,
+      "loss": 0.6416,
+      "step": 6150
+    },
+    {
+      "epoch": 1.7038781163434904,
+      "grad_norm": 0.6898400783538818,
+      "learning_rate": 4.089570820326072e-06,
+      "loss": 0.5711,
+      "step": 6151
+    },
+    {
+      "epoch": 1.7041551246537396,
+      "grad_norm": 0.7291202545166016,
+      "learning_rate": 4.089289637280358e-06,
+      "loss": 0.6284,
+      "step": 6152
+    },
+    {
+      "epoch": 1.7044321329639889,
+      "grad_norm": 0.7050850987434387,
+      "learning_rate": 4.089008420489966e-06,
+      "loss": 0.6316,
+      "step": 6153
+    },
+    {
+      "epoch": 1.7047091412742383,
+      "grad_norm": 0.6834461092948914,
+      "learning_rate": 4.0887271699608675e-06,
+      "loss": 0.6267,
+      "step": 6154
+    },
+    {
+      "epoch": 1.7049861495844876,
+      "grad_norm": 0.7263997197151184,
+      "learning_rate": 4.088445885699034e-06,
+      "loss": 0.648,
+      "step": 6155
+    },
+    {
+      "epoch": 1.7052631578947368,
+      "grad_norm": 0.7006638646125793,
+      "learning_rate": 4.088164567710438e-06,
+      "loss": 0.6418,
+      "step": 6156
+    },
+    {
+      "epoch": 1.7055401662049863,
+      "grad_norm": 0.6719088554382324,
+      "learning_rate": 4.087883216001052e-06,
+      "loss": 0.6188,
+      "step": 6157
+    },
+    {
+      "epoch": 1.7058171745152355,
+      "grad_norm": 0.6723303198814392,
+      "learning_rate": 4.08760183057685e-06,
+      "loss": 0.5958,
+      "step": 6158
+    },
+    {
+      "epoch": 1.7060941828254848,
+      "grad_norm": 0.7296972274780273,
+      "learning_rate": 4.087320411443808e-06,
+      "loss": 0.6476,
+      "step": 6159
+    },
+    {
+      "epoch": 1.7063711911357342,
+      "grad_norm": 0.7308874130249023,
+      "learning_rate": 4.0870389586079e-06,
+      "loss": 0.6427,
+      "step": 6160
+    },
+    {
+      "epoch": 1.7066481994459832,
+      "grad_norm": 0.6853946447372437,
+      "learning_rate": 4.086757472075102e-06,
+      "loss": 0.6455,
+      "step": 6161
+    },
+    {
+      "epoch": 1.7069252077562327,
+      "grad_norm": 0.6351714730262756,
+      "learning_rate": 4.08647595185139e-06,
+      "loss": 0.6193,
+      "step": 6162
+    },
+    {
+      "epoch": 1.7072022160664821,
+      "grad_norm": 0.7018414735794067,
+      "learning_rate": 4.0861943979427424e-06,
+      "loss": 0.6611,
+      "step": 6163
+    },
+    {
+      "epoch": 1.7074792243767312,
+      "grad_norm": 0.742825448513031,
+      "learning_rate": 4.085912810355137e-06,
+      "loss": 0.6531,
+      "step": 6164
+    },
+    {
+      "epoch": 1.7077562326869806,
+      "grad_norm": 0.685265302658081,
+      "learning_rate": 4.085631189094553e-06,
+      "loss": 0.6116,
+      "step": 6165
+    },
+    {
+      "epoch": 1.7080332409972299,
+      "grad_norm": 0.7166582345962524,
+      "learning_rate": 4.08534953416697e-06,
+      "loss": 0.6211,
+      "step": 6166
+    },
+    {
+      "epoch": 1.708310249307479,
+      "grad_norm": 0.7106016874313354,
+      "learning_rate": 4.085067845578368e-06,
+      "loss": 0.6188,
+      "step": 6167
+    },
+    {
+      "epoch": 1.7085872576177286,
+      "grad_norm": 0.6873186230659485,
+      "learning_rate": 4.084786123334726e-06,
+      "loss": 0.5869,
+      "step": 6168
+    },
+    {
+      "epoch": 1.7088642659279778,
+      "grad_norm": 0.670590877532959,
+      "learning_rate": 4.084504367442028e-06,
+      "loss": 0.6009,
+      "step": 6169
+    },
+    {
+      "epoch": 1.709141274238227,
+      "grad_norm": 0.668232262134552,
+      "learning_rate": 4.084222577906257e-06,
+      "loss": 0.6048,
+      "step": 6170
+    },
+    {
+      "epoch": 1.7094182825484765,
+      "grad_norm": 0.6711857914924622,
+      "learning_rate": 4.083940754733394e-06,
+      "loss": 0.6295,
+      "step": 6171
+    },
+    {
+      "epoch": 1.7096952908587257,
+      "grad_norm": 0.7261093258857727,
+      "learning_rate": 4.083658897929425e-06,
+      "loss": 0.6218,
+      "step": 6172
+    },
+    {
+      "epoch": 1.709972299168975,
+      "grad_norm": 0.7620695233345032,
+      "learning_rate": 4.083377007500333e-06,
+      "loss": 0.6085,
+      "step": 6173
+    },
+    {
+      "epoch": 1.7102493074792244,
+      "grad_norm": 0.7104936242103577,
+      "learning_rate": 4.083095083452103e-06,
+      "loss": 0.6665,
+      "step": 6174
+    },
+    {
+      "epoch": 1.7105263157894737,
+      "grad_norm": 0.6915191411972046,
+      "learning_rate": 4.082813125790722e-06,
+      "loss": 0.628,
+      "step": 6175
+    },
+    {
+      "epoch": 1.710803324099723,
+      "grad_norm": 0.6785016059875488,
+      "learning_rate": 4.082531134522176e-06,
+      "loss": 0.6056,
+      "step": 6176
+    },
+    {
+      "epoch": 1.7110803324099724,
+      "grad_norm": 0.7069738507270813,
+      "learning_rate": 4.082249109652453e-06,
+      "loss": 0.6552,
+      "step": 6177
+    },
+    {
+      "epoch": 1.7113573407202216,
+      "grad_norm": 0.6891471743583679,
+      "learning_rate": 4.081967051187541e-06,
+      "loss": 0.6242,
+      "step": 6178
+    },
+    {
+      "epoch": 1.7116343490304708,
+      "grad_norm": 0.7005449533462524,
+      "learning_rate": 4.0816849591334284e-06,
+      "loss": 0.6649,
+      "step": 6179
+    },
+    {
+      "epoch": 1.7119113573407203,
+      "grad_norm": 0.717957615852356,
+      "learning_rate": 4.081402833496105e-06,
+      "loss": 0.6159,
+      "step": 6180
+    },
+    {
+      "epoch": 1.7121883656509695,
+      "grad_norm": 0.6962898969650269,
+      "learning_rate": 4.08112067428156e-06,
+      "loss": 0.6841,
+      "step": 6181
+    },
+    {
+      "epoch": 1.7124653739612188,
+      "grad_norm": 0.677981972694397,
+      "learning_rate": 4.080838481495786e-06,
+      "loss": 0.595,
+      "step": 6182
+    },
+    {
+      "epoch": 1.7127423822714682,
+      "grad_norm": 0.6928604245185852,
+      "learning_rate": 4.080556255144775e-06,
+      "loss": 0.6615,
+      "step": 6183
+    },
+    {
+      "epoch": 1.7130193905817175,
+      "grad_norm": 0.7054057121276855,
+      "learning_rate": 4.0802739952345175e-06,
+      "loss": 0.5946,
+      "step": 6184
+    },
+    {
+      "epoch": 1.7132963988919667,
+      "grad_norm": 0.7249715328216553,
+      "learning_rate": 4.079991701771007e-06,
+      "loss": 0.6842,
+      "step": 6185
+    },
+    {
+      "epoch": 1.7135734072022162,
+      "grad_norm": 0.6929246187210083,
+      "learning_rate": 4.079709374760239e-06,
+      "loss": 0.6191,
+      "step": 6186
+    },
+    {
+      "epoch": 1.7138504155124654,
+      "grad_norm": 0.7572833299636841,
+      "learning_rate": 4.079427014208206e-06,
+      "loss": 0.6109,
+      "step": 6187
+    },
+    {
+      "epoch": 1.7141274238227147,
+      "grad_norm": 0.6429564356803894,
+      "learning_rate": 4.079144620120904e-06,
+      "loss": 0.5983,
+      "step": 6188
+    },
+    {
+      "epoch": 1.7144044321329641,
+      "grad_norm": 0.6844918727874756,
+      "learning_rate": 4.0788621925043294e-06,
+      "loss": 0.6029,
+      "step": 6189
+    },
+    {
+      "epoch": 1.7146814404432131,
+      "grad_norm": 0.6932576298713684,
+      "learning_rate": 4.078579731364478e-06,
+      "loss": 0.6118,
+      "step": 6190
+    },
+    {
+      "epoch": 1.7149584487534626,
+      "grad_norm": 0.691020131111145,
+      "learning_rate": 4.078297236707348e-06,
+      "loss": 0.6609,
+      "step": 6191
+    },
+    {
+      "epoch": 1.715235457063712,
+      "grad_norm": 0.7045336365699768,
+      "learning_rate": 4.078014708538938e-06,
+      "loss": 0.6435,
+      "step": 6192
+    },
+    {
+      "epoch": 1.715512465373961,
+      "grad_norm": 0.6789292097091675,
+      "learning_rate": 4.0777321468652444e-06,
+      "loss": 0.6268,
+      "step": 6193
+    },
+    {
+      "epoch": 1.7157894736842105,
+      "grad_norm": 0.6860518455505371,
+      "learning_rate": 4.077449551692268e-06,
+      "loss": 0.6457,
+      "step": 6194
+    },
+    {
+      "epoch": 1.71606648199446,
+      "grad_norm": 0.6922401189804077,
+      "learning_rate": 4.077166923026012e-06,
+      "loss": 0.6315,
+      "step": 6195
+    },
+    {
+      "epoch": 1.716343490304709,
+      "grad_norm": 0.7087437510490417,
+      "learning_rate": 4.076884260872472e-06,
+      "loss": 0.6008,
+      "step": 6196
+    },
+    {
+      "epoch": 1.7166204986149585,
+      "grad_norm": 0.6820953488349915,
+      "learning_rate": 4.076601565237653e-06,
+      "loss": 0.6221,
+      "step": 6197
+    },
+    {
+      "epoch": 1.7168975069252077,
+      "grad_norm": 0.7251583337783813,
+      "learning_rate": 4.076318836127556e-06,
+      "loss": 0.6201,
+      "step": 6198
+    },
+    {
+      "epoch": 1.717174515235457,
+      "grad_norm": 0.7204704880714417,
+      "learning_rate": 4.076036073548185e-06,
+      "loss": 0.6396,
+      "step": 6199
+    },
+    {
+      "epoch": 1.7174515235457064,
+      "grad_norm": 0.6726484894752502,
+      "learning_rate": 4.075753277505544e-06,
+      "loss": 0.6495,
+      "step": 6200
+    },
+    {
+      "epoch": 1.7177285318559556,
+      "grad_norm": 0.7026734352111816,
+      "learning_rate": 4.075470448005636e-06,
+      "loss": 0.6056,
+      "step": 6201
+    },
+    {
+      "epoch": 1.7180055401662049,
+      "grad_norm": 0.704633891582489,
+      "learning_rate": 4.075187585054468e-06,
+      "loss": 0.5745,
+      "step": 6202
+    },
+    {
+      "epoch": 1.7182825484764543,
+      "grad_norm": 0.7028330564498901,
+      "learning_rate": 4.0749046886580444e-06,
+      "loss": 0.6352,
+      "step": 6203
+    },
+    {
+      "epoch": 1.7185595567867036,
+      "grad_norm": 0.6898366808891296,
+      "learning_rate": 4.074621758822373e-06,
+      "loss": 0.6102,
+      "step": 6204
+    },
+    {
+      "epoch": 1.7188365650969528,
+      "grad_norm": 0.727149248123169,
+      "learning_rate": 4.074338795553459e-06,
+      "loss": 0.626,
+      "step": 6205
+    },
+    {
+      "epoch": 1.7191135734072023,
+      "grad_norm": 0.6894062161445618,
+      "learning_rate": 4.074055798857314e-06,
+      "loss": 0.6292,
+      "step": 6206
+    },
+    {
+      "epoch": 1.7193905817174515,
+      "grad_norm": 0.9477318525314331,
+      "learning_rate": 4.073772768739943e-06,
+      "loss": 0.6284,
+      "step": 6207
+    },
+    {
+      "epoch": 1.7196675900277008,
+      "grad_norm": 0.7123240232467651,
+      "learning_rate": 4.073489705207358e-06,
+      "loss": 0.6025,
+      "step": 6208
+    },
+    {
+      "epoch": 1.7199445983379502,
+      "grad_norm": 0.7271552085876465,
+      "learning_rate": 4.073206608265567e-06,
+      "loss": 0.6025,
+      "step": 6209
+    },
+    {
+      "epoch": 1.7202216066481995,
+      "grad_norm": 0.7003310918807983,
+      "learning_rate": 4.072923477920584e-06,
+      "loss": 0.6437,
+      "step": 6210
+    },
+    {
+      "epoch": 1.7204986149584487,
+      "grad_norm": 0.6652795672416687,
+      "learning_rate": 4.072640314178418e-06,
+      "loss": 0.5818,
+      "step": 6211
+    },
+    {
+      "epoch": 1.7207756232686982,
+      "grad_norm": 0.7066246867179871,
+      "learning_rate": 4.072357117045082e-06,
+      "loss": 0.65,
+      "step": 6212
+    },
+    {
+      "epoch": 1.7210526315789474,
+      "grad_norm": 0.7616447806358337,
+      "learning_rate": 4.072073886526589e-06,
+      "loss": 0.6075,
+      "step": 6213
+    },
+    {
+      "epoch": 1.7213296398891966,
+      "grad_norm": 0.7162469029426575,
+      "learning_rate": 4.071790622628953e-06,
+      "loss": 0.6344,
+      "step": 6214
+    },
+    {
+      "epoch": 1.721606648199446,
+      "grad_norm": 0.6894190311431885,
+      "learning_rate": 4.071507325358188e-06,
+      "loss": 0.5557,
+      "step": 6215
+    },
+    {
+      "epoch": 1.7218836565096953,
+      "grad_norm": 0.7193700075149536,
+      "learning_rate": 4.071223994720309e-06,
+      "loss": 0.6059,
+      "step": 6216
+    },
+    {
+      "epoch": 1.7221606648199446,
+      "grad_norm": 0.7540478706359863,
+      "learning_rate": 4.070940630721333e-06,
+      "loss": 0.6264,
+      "step": 6217
+    },
+    {
+      "epoch": 1.722437673130194,
+      "grad_norm": 0.6955430507659912,
+      "learning_rate": 4.070657233367274e-06,
+      "loss": 0.6087,
+      "step": 6218
+    },
+    {
+      "epoch": 1.7227146814404433,
+      "grad_norm": 0.6968384385108948,
+      "learning_rate": 4.070373802664153e-06,
+      "loss": 0.6091,
+      "step": 6219
+    },
+    {
+      "epoch": 1.7229916897506925,
+      "grad_norm": 0.7160845398902893,
+      "learning_rate": 4.070090338617985e-06,
+      "loss": 0.6538,
+      "step": 6220
+    },
+    {
+      "epoch": 1.723268698060942,
+      "grad_norm": 0.7012823820114136,
+      "learning_rate": 4.069806841234789e-06,
+      "loss": 0.6606,
+      "step": 6221
+    },
+    {
+      "epoch": 1.723545706371191,
+      "grad_norm": 0.6941177845001221,
+      "learning_rate": 4.069523310520585e-06,
+      "loss": 0.6361,
+      "step": 6222
+    },
+    {
+      "epoch": 1.7238227146814404,
+      "grad_norm": 0.6887818574905396,
+      "learning_rate": 4.069239746481394e-06,
+      "loss": 0.6145,
+      "step": 6223
+    },
+    {
+      "epoch": 1.72409972299169,
+      "grad_norm": 0.7334598302841187,
+      "learning_rate": 4.068956149123235e-06,
+      "loss": 0.6769,
+      "step": 6224
+    },
+    {
+      "epoch": 1.724376731301939,
+      "grad_norm": 0.697456419467926,
+      "learning_rate": 4.06867251845213e-06,
+      "loss": 0.5673,
+      "step": 6225
+    },
+    {
+      "epoch": 1.7246537396121884,
+      "grad_norm": 0.6590297222137451,
+      "learning_rate": 4.0683888544741024e-06,
+      "loss": 0.5588,
+      "step": 6226
+    },
+    {
+      "epoch": 1.7249307479224378,
+      "grad_norm": 0.7389965653419495,
+      "learning_rate": 4.068105157195173e-06,
+      "loss": 0.6293,
+      "step": 6227
+    },
+    {
+      "epoch": 1.7252077562326869,
+      "grad_norm": 0.6632516980171204,
+      "learning_rate": 4.067821426621368e-06,
+      "loss": 0.6334,
+      "step": 6228
+    },
+    {
+      "epoch": 1.7254847645429363,
+      "grad_norm": 0.7266767024993896,
+      "learning_rate": 4.0675376627587095e-06,
+      "loss": 0.6003,
+      "step": 6229
+    },
+    {
+      "epoch": 1.7257617728531855,
+      "grad_norm": 0.6739516854286194,
+      "learning_rate": 4.067253865613223e-06,
+      "loss": 0.6102,
+      "step": 6230
+    },
+    {
+      "epoch": 1.7260387811634348,
+      "grad_norm": 0.6888447403907776,
+      "learning_rate": 4.066970035190935e-06,
+      "loss": 0.627,
+      "step": 6231
+    },
+    {
+      "epoch": 1.7263157894736842,
+      "grad_norm": 0.7709148526191711,
+      "learning_rate": 4.066686171497872e-06,
+      "loss": 0.6326,
+      "step": 6232
+    },
+    {
+      "epoch": 1.7265927977839335,
+      "grad_norm": 0.6897366046905518,
+      "learning_rate": 4.06640227454006e-06,
+      "loss": 0.6276,
+      "step": 6233
+    },
+    {
+      "epoch": 1.7268698060941827,
+      "grad_norm": 0.6722098588943481,
+      "learning_rate": 4.066118344323528e-06,
+      "loss": 0.6011,
+      "step": 6234
+    },
+    {
+      "epoch": 1.7271468144044322,
+      "grad_norm": 0.7252819538116455,
+      "learning_rate": 4.065834380854305e-06,
+      "loss": 0.6208,
+      "step": 6235
+    },
+    {
+      "epoch": 1.7274238227146814,
+      "grad_norm": 0.7158248424530029,
+      "learning_rate": 4.065550384138417e-06,
+      "loss": 0.6293,
+      "step": 6236
+    },
+    {
+      "epoch": 1.7277008310249307,
+      "grad_norm": 0.7303574681282043,
+      "learning_rate": 4.065266354181898e-06,
+      "loss": 0.6289,
+      "step": 6237
+    },
+    {
+      "epoch": 1.7279778393351801,
+      "grad_norm": 0.6661691069602966,
+      "learning_rate": 4.064982290990777e-06,
+      "loss": 0.6345,
+      "step": 6238
+    },
+    {
+      "epoch": 1.7282548476454294,
+      "grad_norm": 0.7096925377845764,
+      "learning_rate": 4.064698194571085e-06,
+      "loss": 0.6219,
+      "step": 6239
+    },
+    {
+      "epoch": 1.7285318559556786,
+      "grad_norm": 0.7123353481292725,
+      "learning_rate": 4.064414064928854e-06,
+      "loss": 0.6726,
+      "step": 6240
+    },
+    {
+      "epoch": 1.728808864265928,
+      "grad_norm": 0.6776770353317261,
+      "learning_rate": 4.064129902070118e-06,
+      "loss": 0.6325,
+      "step": 6241
+    },
+    {
+      "epoch": 1.7290858725761773,
+      "grad_norm": 0.727475643157959,
+      "learning_rate": 4.063845706000911e-06,
+      "loss": 0.6176,
+      "step": 6242
+    },
+    {
+      "epoch": 1.7293628808864265,
+      "grad_norm": 0.6917515397071838,
+      "learning_rate": 4.063561476727265e-06,
+      "loss": 0.623,
+      "step": 6243
+    },
+    {
+      "epoch": 1.729639889196676,
+      "grad_norm": 0.6911443471908569,
+      "learning_rate": 4.0632772142552155e-06,
+      "loss": 0.6079,
+      "step": 6244
+    },
+    {
+      "epoch": 1.7299168975069252,
+      "grad_norm": 0.6569199562072754,
+      "learning_rate": 4.062992918590799e-06,
+      "loss": 0.6144,
+      "step": 6245
+    },
+    {
+      "epoch": 1.7301939058171745,
+      "grad_norm": 0.6976788640022278,
+      "learning_rate": 4.062708589740051e-06,
+      "loss": 0.6127,
+      "step": 6246
+    },
+    {
+      "epoch": 1.730470914127424,
+      "grad_norm": 0.6994149684906006,
+      "learning_rate": 4.06242422770901e-06,
+      "loss": 0.6107,
+      "step": 6247
+    },
+    {
+      "epoch": 1.7307479224376732,
+      "grad_norm": 0.7247564792633057,
+      "learning_rate": 4.062139832503711e-06,
+      "loss": 0.653,
+      "step": 6248
+    },
+    {
+      "epoch": 1.7310249307479224,
+      "grad_norm": 0.6729344129562378,
+      "learning_rate": 4.061855404130195e-06,
+      "loss": 0.6358,
+      "step": 6249
+    },
+    {
+      "epoch": 1.7313019390581719,
+      "grad_norm": 0.7238703370094299,
+      "learning_rate": 4.0615709425945e-06,
+      "loss": 0.6623,
+      "step": 6250
+    },
+    {
+      "epoch": 1.731578947368421,
+      "grad_norm": 0.6958584785461426,
+      "learning_rate": 4.061286447902667e-06,
+      "loss": 0.6162,
+      "step": 6251
+    },
+    {
+      "epoch": 1.7318559556786703,
+      "grad_norm": 0.7200596332550049,
+      "learning_rate": 4.061001920060735e-06,
+      "loss": 0.6552,
+      "step": 6252
+    },
+    {
+      "epoch": 1.7321329639889198,
+      "grad_norm": 0.6930348873138428,
+      "learning_rate": 4.0607173590747455e-06,
+      "loss": 0.5934,
+      "step": 6253
+    },
+    {
+      "epoch": 1.7324099722991688,
+      "grad_norm": 0.6601454615592957,
+      "learning_rate": 4.060432764950741e-06,
+      "loss": 0.6083,
+      "step": 6254
+    },
+    {
+      "epoch": 1.7326869806094183,
+      "grad_norm": 0.6737957000732422,
+      "learning_rate": 4.0601481376947634e-06,
+      "loss": 0.633,
+      "step": 6255
+    },
+    {
+      "epoch": 1.7329639889196677,
+      "grad_norm": 0.6880027651786804,
+      "learning_rate": 4.059863477312858e-06,
+      "loss": 0.607,
+      "step": 6256
+    },
+    {
+      "epoch": 1.7332409972299168,
+      "grad_norm": 0.6753184199333191,
+      "learning_rate": 4.059578783811067e-06,
+      "loss": 0.6336,
+      "step": 6257
+    },
+    {
+      "epoch": 1.7335180055401662,
+      "grad_norm": 0.6858214735984802,
+      "learning_rate": 4.059294057195435e-06,
+      "loss": 0.6702,
+      "step": 6258
+    },
+    {
+      "epoch": 1.7337950138504157,
+      "grad_norm": 0.6795182228088379,
+      "learning_rate": 4.059009297472009e-06,
+      "loss": 0.6329,
+      "step": 6259
+    },
+    {
+      "epoch": 1.7340720221606647,
+      "grad_norm": 0.6773170232772827,
+      "learning_rate": 4.058724504646834e-06,
+      "loss": 0.6339,
+      "step": 6260
+    },
+    {
+      "epoch": 1.7343490304709142,
+      "grad_norm": 0.6879174709320068,
+      "learning_rate": 4.058439678725957e-06,
+      "loss": 0.6261,
+      "step": 6261
+    },
+    {
+      "epoch": 1.7346260387811634,
+      "grad_norm": 0.7096613645553589,
+      "learning_rate": 4.058154819715426e-06,
+      "loss": 0.647,
+      "step": 6262
+    },
+    {
+      "epoch": 1.7349030470914126,
+      "grad_norm": 0.695104718208313,
+      "learning_rate": 4.05786992762129e-06,
+      "loss": 0.6142,
+      "step": 6263
+    },
+    {
+      "epoch": 1.735180055401662,
+      "grad_norm": 0.6822800636291504,
+      "learning_rate": 4.057585002449596e-06,
+      "loss": 0.6559,
+      "step": 6264
+    },
+    {
+      "epoch": 1.7354570637119113,
+      "grad_norm": 0.7518125176429749,
+      "learning_rate": 4.057300044206395e-06,
+      "loss": 0.636,
+      "step": 6265
+    },
+    {
+      "epoch": 1.7357340720221606,
+      "grad_norm": 0.6837979555130005,
+      "learning_rate": 4.057015052897737e-06,
+      "loss": 0.6183,
+      "step": 6266
+    },
+    {
+      "epoch": 1.73601108033241,
+      "grad_norm": 0.7142424583435059,
+      "learning_rate": 4.056730028529673e-06,
+      "loss": 0.6499,
+      "step": 6267
+    },
+    {
+      "epoch": 1.7362880886426593,
+      "grad_norm": 0.7081639170646667,
+      "learning_rate": 4.0564449711082555e-06,
+      "loss": 0.6267,
+      "step": 6268
+    },
+    {
+      "epoch": 1.7365650969529085,
+      "grad_norm": 0.6069878935813904,
+      "learning_rate": 4.0561598806395365e-06,
+      "loss": 0.5292,
+      "step": 6269
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.6596729755401611,
+      "learning_rate": 4.055874757129569e-06,
+      "loss": 0.591,
+      "step": 6270
+    },
+    {
+      "epoch": 1.7371191135734072,
+      "grad_norm": 0.7232509851455688,
+      "learning_rate": 4.055589600584407e-06,
+      "loss": 0.6362,
+      "step": 6271
+    },
+    {
+      "epoch": 1.7373961218836564,
+      "grad_norm": 0.6596758961677551,
+      "learning_rate": 4.055304411010106e-06,
+      "loss": 0.627,
+      "step": 6272
+    },
+    {
+      "epoch": 1.737673130193906,
+      "grad_norm": 0.6614773273468018,
+      "learning_rate": 4.05501918841272e-06,
+      "loss": 0.5894,
+      "step": 6273
+    },
+    {
+      "epoch": 1.7379501385041551,
+      "grad_norm": 0.7271131873130798,
+      "learning_rate": 4.054733932798306e-06,
+      "loss": 0.6117,
+      "step": 6274
+    },
+    {
+      "epoch": 1.7382271468144044,
+      "grad_norm": 0.6962509155273438,
+      "learning_rate": 4.0544486441729204e-06,
+      "loss": 0.6136,
+      "step": 6275
+    },
+    {
+      "epoch": 1.7385041551246538,
+      "grad_norm": 0.7124279737472534,
+      "learning_rate": 4.054163322542619e-06,
+      "loss": 0.6045,
+      "step": 6276
+    },
+    {
+      "epoch": 1.738781163434903,
+      "grad_norm": 0.7210063338279724,
+      "learning_rate": 4.053877967913463e-06,
+      "loss": 0.6177,
+      "step": 6277
+    },
+    {
+      "epoch": 1.7390581717451523,
+      "grad_norm": 0.7171337008476257,
+      "learning_rate": 4.053592580291509e-06,
+      "loss": 0.6144,
+      "step": 6278
+    },
+    {
+      "epoch": 1.7393351800554018,
+      "grad_norm": 0.6800563931465149,
+      "learning_rate": 4.053307159682817e-06,
+      "loss": 0.6012,
+      "step": 6279
+    },
+    {
+      "epoch": 1.739612188365651,
+      "grad_norm": 0.725743293762207,
+      "learning_rate": 4.053021706093447e-06,
+      "loss": 0.6271,
+      "step": 6280
+    },
+    {
+      "epoch": 1.7398891966759003,
+      "grad_norm": 0.7025867700576782,
+      "learning_rate": 4.05273621952946e-06,
+      "loss": 0.5914,
+      "step": 6281
+    },
+    {
+      "epoch": 1.7401662049861497,
+      "grad_norm": 0.6952527761459351,
+      "learning_rate": 4.0524506999969185e-06,
+      "loss": 0.6406,
+      "step": 6282
+    },
+    {
+      "epoch": 1.7404432132963987,
+      "grad_norm": 0.7095229029655457,
+      "learning_rate": 4.052165147501884e-06,
+      "loss": 0.6267,
+      "step": 6283
+    },
+    {
+      "epoch": 1.7407202216066482,
+      "grad_norm": 0.6685070991516113,
+      "learning_rate": 4.051879562050419e-06,
+      "loss": 0.6507,
+      "step": 6284
+    },
+    {
+      "epoch": 1.7409972299168976,
+      "grad_norm": 0.7472874522209167,
+      "learning_rate": 4.051593943648588e-06,
+      "loss": 0.6187,
+      "step": 6285
+    },
+    {
+      "epoch": 1.7412742382271467,
+      "grad_norm": 0.7431471943855286,
+      "learning_rate": 4.051308292302456e-06,
+      "loss": 0.6263,
+      "step": 6286
+    },
+    {
+      "epoch": 1.7415512465373961,
+      "grad_norm": 0.6761013269424438,
+      "learning_rate": 4.051022608018087e-06,
+      "loss": 0.6467,
+      "step": 6287
+    },
+    {
+      "epoch": 1.7418282548476456,
+      "grad_norm": 0.6962248086929321,
+      "learning_rate": 4.050736890801547e-06,
+      "loss": 0.6051,
+      "step": 6288
+    },
+    {
+      "epoch": 1.7421052631578946,
+      "grad_norm": 0.7127392292022705,
+      "learning_rate": 4.0504511406589025e-06,
+      "loss": 0.6463,
+      "step": 6289
+    },
+    {
+      "epoch": 1.742382271468144,
+      "grad_norm": 0.656398594379425,
+      "learning_rate": 4.050165357596222e-06,
+      "loss": 0.6134,
+      "step": 6290
+    },
+    {
+      "epoch": 1.7426592797783933,
+      "grad_norm": 0.6969537734985352,
+      "learning_rate": 4.04987954161957e-06,
+      "loss": 0.5537,
+      "step": 6291
+    },
+    {
+      "epoch": 1.7429362880886425,
+      "grad_norm": 0.628457248210907,
+      "learning_rate": 4.04959369273502e-06,
+      "loss": 0.6291,
+      "step": 6292
+    },
+    {
+      "epoch": 1.743213296398892,
+      "grad_norm": 0.7016260623931885,
+      "learning_rate": 4.049307810948636e-06,
+      "loss": 0.6569,
+      "step": 6293
+    },
+    {
+      "epoch": 1.7434903047091412,
+      "grad_norm": 0.7526865601539612,
+      "learning_rate": 4.049021896266493e-06,
+      "loss": 0.6649,
+      "step": 6294
+    },
+    {
+      "epoch": 1.7437673130193905,
+      "grad_norm": 0.7152197360992432,
+      "learning_rate": 4.048735948694658e-06,
+      "loss": 0.644,
+      "step": 6295
+    },
+    {
+      "epoch": 1.74404432132964,
+      "grad_norm": 0.6660102605819702,
+      "learning_rate": 4.048449968239204e-06,
+      "loss": 0.5963,
+      "step": 6296
+    },
+    {
+      "epoch": 1.7443213296398892,
+      "grad_norm": 0.6983559727668762,
+      "learning_rate": 4.048163954906203e-06,
+      "loss": 0.6267,
+      "step": 6297
+    },
+    {
+      "epoch": 1.7445983379501384,
+      "grad_norm": 0.677616536617279,
+      "learning_rate": 4.0478779087017276e-06,
+      "loss": 0.6113,
+      "step": 6298
+    },
+    {
+      "epoch": 1.7448753462603879,
+      "grad_norm": 0.6915004253387451,
+      "learning_rate": 4.047591829631852e-06,
+      "loss": 0.6493,
+      "step": 6299
+    },
+    {
+      "epoch": 1.745152354570637,
+      "grad_norm": 0.7380059957504272,
+      "learning_rate": 4.047305717702649e-06,
+      "loss": 0.6161,
+      "step": 6300
+    },
+    {
+      "epoch": 1.7454293628808863,
+      "grad_norm": 0.6928910613059998,
+      "learning_rate": 4.047019572920194e-06,
+      "loss": 0.6416,
+      "step": 6301
+    },
+    {
+      "epoch": 1.7457063711911358,
+      "grad_norm": 0.6989980340003967,
+      "learning_rate": 4.0467333952905635e-06,
+      "loss": 0.586,
+      "step": 6302
+    },
+    {
+      "epoch": 1.745983379501385,
+      "grad_norm": 0.6553187370300293,
+      "learning_rate": 4.046447184819833e-06,
+      "loss": 0.6156,
+      "step": 6303
+    },
+    {
+      "epoch": 1.7462603878116343,
+      "grad_norm": 0.6651337742805481,
+      "learning_rate": 4.046160941514079e-06,
+      "loss": 0.5633,
+      "step": 6304
+    },
+    {
+      "epoch": 1.7465373961218837,
+      "grad_norm": 0.6954764127731323,
+      "learning_rate": 4.0458746653793805e-06,
+      "loss": 0.6093,
+      "step": 6305
+    },
+    {
+      "epoch": 1.746814404432133,
+      "grad_norm": 0.7164486646652222,
+      "learning_rate": 4.045588356421814e-06,
+      "loss": 0.6432,
+      "step": 6306
+    },
+    {
+      "epoch": 1.7470914127423822,
+      "grad_norm": 0.6956983804702759,
+      "learning_rate": 4.045302014647461e-06,
+      "loss": 0.6357,
+      "step": 6307
+    },
+    {
+      "epoch": 1.7473684210526317,
+      "grad_norm": 0.6895577907562256,
+      "learning_rate": 4.0450156400623994e-06,
+      "loss": 0.6631,
+      "step": 6308
+    },
+    {
+      "epoch": 1.747645429362881,
+      "grad_norm": 0.7270070314407349,
+      "learning_rate": 4.044729232672711e-06,
+      "loss": 0.6687,
+      "step": 6309
+    },
+    {
+      "epoch": 1.7479224376731302,
+      "grad_norm": 0.6985939741134644,
+      "learning_rate": 4.044442792484475e-06,
+      "loss": 0.6499,
+      "step": 6310
+    },
+    {
+      "epoch": 1.7481994459833796,
+      "grad_norm": 0.7345278263092041,
+      "learning_rate": 4.044156319503774e-06,
+      "loss": 0.6322,
+      "step": 6311
+    },
+    {
+      "epoch": 1.7484764542936289,
+      "grad_norm": 0.7118226885795593,
+      "learning_rate": 4.0438698137366925e-06,
+      "loss": 0.6265,
+      "step": 6312
+    },
+    {
+      "epoch": 1.748753462603878,
+      "grad_norm": 0.685529887676239,
+      "learning_rate": 4.043583275189311e-06,
+      "loss": 0.6536,
+      "step": 6313
+    },
+    {
+      "epoch": 1.7490304709141276,
+      "grad_norm": 0.7099524736404419,
+      "learning_rate": 4.043296703867715e-06,
+      "loss": 0.6477,
+      "step": 6314
+    },
+    {
+      "epoch": 1.7493074792243766,
+      "grad_norm": 0.6660603880882263,
+      "learning_rate": 4.04301009977799e-06,
+      "loss": 0.593,
+      "step": 6315
+    },
+    {
+      "epoch": 1.749584487534626,
+      "grad_norm": 0.6841859221458435,
+      "learning_rate": 4.042723462926219e-06,
+      "loss": 0.6088,
+      "step": 6316
+    },
+    {
+      "epoch": 1.7498614958448755,
+      "grad_norm": 0.717239260673523,
+      "learning_rate": 4.042436793318489e-06,
+      "loss": 0.6679,
+      "step": 6317
+    },
+    {
+      "epoch": 1.7501385041551245,
+      "grad_norm": 0.6708807349205017,
+      "learning_rate": 4.042150090960888e-06,
+      "loss": 0.642,
+      "step": 6318
+    },
+    {
+      "epoch": 1.750415512465374,
+      "grad_norm": 0.7339285612106323,
+      "learning_rate": 4.041863355859502e-06,
+      "loss": 0.6353,
+      "step": 6319
+    },
+    {
+      "epoch": 1.7506925207756234,
+      "grad_norm": 0.6966876983642578,
+      "learning_rate": 4.041576588020419e-06,
+      "loss": 0.6094,
+      "step": 6320
+    },
+    {
+      "epoch": 1.7509695290858724,
+      "grad_norm": 0.6712494492530823,
+      "learning_rate": 4.041289787449728e-06,
+      "loss": 0.6468,
+      "step": 6321
+    },
+    {
+      "epoch": 1.751246537396122,
+      "grad_norm": 0.6606813669204712,
+      "learning_rate": 4.04100295415352e-06,
+      "loss": 0.5824,
+      "step": 6322
+    },
+    {
+      "epoch": 1.7515235457063711,
+      "grad_norm": 0.6949263215065002,
+      "learning_rate": 4.040716088137883e-06,
+      "loss": 0.6462,
+      "step": 6323
+    },
+    {
+      "epoch": 1.7518005540166204,
+      "grad_norm": 0.7004557251930237,
+      "learning_rate": 4.040429189408909e-06,
+      "loss": 0.6624,
+      "step": 6324
+    },
+    {
+      "epoch": 1.7520775623268698,
+      "grad_norm": 0.6793391704559326,
+      "learning_rate": 4.040142257972689e-06,
+      "loss": 0.6009,
+      "step": 6325
+    },
+    {
+      "epoch": 1.752354570637119,
+      "grad_norm": 0.6711333394050598,
+      "learning_rate": 4.039855293835316e-06,
+      "loss": 0.5837,
+      "step": 6326
+    },
+    {
+      "epoch": 1.7526315789473683,
+      "grad_norm": 0.7150370478630066,
+      "learning_rate": 4.039568297002884e-06,
+      "loss": 0.6158,
+      "step": 6327
+    },
+    {
+      "epoch": 1.7529085872576178,
+      "grad_norm": 0.7052631974220276,
+      "learning_rate": 4.039281267481484e-06,
+      "loss": 0.6346,
+      "step": 6328
+    },
+    {
+      "epoch": 1.753185595567867,
+      "grad_norm": 0.7503765225410461,
+      "learning_rate": 4.038994205277213e-06,
+      "loss": 0.6359,
+      "step": 6329
+    },
+    {
+      "epoch": 1.7534626038781163,
+      "grad_norm": 0.720745861530304,
+      "learning_rate": 4.038707110396164e-06,
+      "loss": 0.6278,
+      "step": 6330
+    },
+    {
+      "epoch": 1.7537396121883657,
+      "grad_norm": 0.6894276738166809,
+      "learning_rate": 4.038419982844433e-06,
+      "loss": 0.612,
+      "step": 6331
+    },
+    {
+      "epoch": 1.754016620498615,
+      "grad_norm": 0.6956282258033752,
+      "learning_rate": 4.038132822628118e-06,
+      "loss": 0.6444,
+      "step": 6332
+    },
+    {
+      "epoch": 1.7542936288088642,
+      "grad_norm": 0.6910063624382019,
+      "learning_rate": 4.0378456297533155e-06,
+      "loss": 0.6443,
+      "step": 6333
+    },
+    {
+      "epoch": 1.7545706371191137,
+      "grad_norm": 0.6689027547836304,
+      "learning_rate": 4.037558404226123e-06,
+      "loss": 0.6045,
+      "step": 6334
+    },
+    {
+      "epoch": 1.754847645429363,
+      "grad_norm": 0.6691604256629944,
+      "learning_rate": 4.037271146052639e-06,
+      "loss": 0.6199,
+      "step": 6335
+    },
+    {
+      "epoch": 1.7551246537396121,
+      "grad_norm": 0.7217824459075928,
+      "learning_rate": 4.036983855238962e-06,
+      "loss": 0.6079,
+      "step": 6336
+    },
+    {
+      "epoch": 1.7554016620498616,
+      "grad_norm": 0.6883172988891602,
+      "learning_rate": 4.036696531791193e-06,
+      "loss": 0.6142,
+      "step": 6337
+    },
+    {
+      "epoch": 1.7556786703601108,
+      "grad_norm": 0.6836881041526794,
+      "learning_rate": 4.036409175715434e-06,
+      "loss": 0.5989,
+      "step": 6338
+    },
+    {
+      "epoch": 1.75595567867036,
+      "grad_norm": 0.6560087203979492,
+      "learning_rate": 4.036121787017782e-06,
+      "loss": 0.6466,
+      "step": 6339
+    },
+    {
+      "epoch": 1.7562326869806095,
+      "grad_norm": 0.697339653968811,
+      "learning_rate": 4.035834365704343e-06,
+      "loss": 0.5821,
+      "step": 6340
+    },
+    {
+      "epoch": 1.7565096952908588,
+      "grad_norm": 0.6981000900268555,
+      "learning_rate": 4.035546911781219e-06,
+      "loss": 0.5742,
+      "step": 6341
+    },
+    {
+      "epoch": 1.756786703601108,
+      "grad_norm": 0.7049815654754639,
+      "learning_rate": 4.035259425254511e-06,
+      "loss": 0.6468,
+      "step": 6342
+    },
+    {
+      "epoch": 1.7570637119113575,
+      "grad_norm": 0.6798682808876038,
+      "learning_rate": 4.034971906130326e-06,
+      "loss": 0.6371,
+      "step": 6343
+    },
+    {
+      "epoch": 1.7573407202216067,
+      "grad_norm": 0.6624364852905273,
+      "learning_rate": 4.034684354414768e-06,
+      "loss": 0.6099,
+      "step": 6344
+    },
+    {
+      "epoch": 1.757617728531856,
+      "grad_norm": 0.7097635865211487,
+      "learning_rate": 4.03439677011394e-06,
+      "loss": 0.5877,
+      "step": 6345
+    },
+    {
+      "epoch": 1.7578947368421054,
+      "grad_norm": 0.7028236985206604,
+      "learning_rate": 4.034109153233952e-06,
+      "loss": 0.67,
+      "step": 6346
+    },
+    {
+      "epoch": 1.7581717451523544,
+      "grad_norm": 0.6697776913642883,
+      "learning_rate": 4.0338215037809085e-06,
+      "loss": 0.5897,
+      "step": 6347
+    },
+    {
+      "epoch": 1.7584487534626039,
+      "grad_norm": 0.7153249382972717,
+      "learning_rate": 4.033533821760917e-06,
+      "loss": 0.6275,
+      "step": 6348
+    },
+    {
+      "epoch": 1.7587257617728533,
+      "grad_norm": 0.7187476754188538,
+      "learning_rate": 4.033246107180086e-06,
+      "loss": 0.5889,
+      "step": 6349
+    },
+    {
+      "epoch": 1.7590027700831024,
+      "grad_norm": 0.7286702394485474,
+      "learning_rate": 4.032958360044526e-06,
+      "loss": 0.6521,
+      "step": 6350
+    },
+    {
+      "epoch": 1.7592797783933518,
+      "grad_norm": 0.7181779146194458,
+      "learning_rate": 4.032670580360345e-06,
+      "loss": 0.6534,
+      "step": 6351
+    },
+    {
+      "epoch": 1.7595567867036013,
+      "grad_norm": 0.7091637849807739,
+      "learning_rate": 4.032382768133652e-06,
+      "loss": 0.637,
+      "step": 6352
+    },
+    {
+      "epoch": 1.7598337950138503,
+      "grad_norm": 0.7320290803909302,
+      "learning_rate": 4.032094923370561e-06,
+      "loss": 0.6393,
+      "step": 6353
+    },
+    {
+      "epoch": 1.7601108033240997,
+      "grad_norm": 0.6917217373847961,
+      "learning_rate": 4.031807046077182e-06,
+      "loss": 0.6552,
+      "step": 6354
+    },
+    {
+      "epoch": 1.760387811634349,
+      "grad_norm": 0.6732197999954224,
+      "learning_rate": 4.031519136259628e-06,
+      "loss": 0.6087,
+      "step": 6355
+    },
+    {
+      "epoch": 1.7606648199445982,
+      "grad_norm": 0.7149708867073059,
+      "learning_rate": 4.031231193924011e-06,
+      "loss": 0.6239,
+      "step": 6356
+    },
+    {
+      "epoch": 1.7609418282548477,
+      "grad_norm": 0.6990062594413757,
+      "learning_rate": 4.0309432190764456e-06,
+      "loss": 0.6011,
+      "step": 6357
+    },
+    {
+      "epoch": 1.761218836565097,
+      "grad_norm": 0.7346669435501099,
+      "learning_rate": 4.030655211723046e-06,
+      "loss": 0.6554,
+      "step": 6358
+    },
+    {
+      "epoch": 1.7614958448753462,
+      "grad_norm": 0.7255174517631531,
+      "learning_rate": 4.030367171869927e-06,
+      "loss": 0.6156,
+      "step": 6359
+    },
+    {
+      "epoch": 1.7617728531855956,
+      "grad_norm": 0.7192071080207825,
+      "learning_rate": 4.030079099523205e-06,
+      "loss": 0.6886,
+      "step": 6360
+    },
+    {
+      "epoch": 1.7620498614958449,
+      "grad_norm": 0.7222265005111694,
+      "learning_rate": 4.029790994688997e-06,
+      "loss": 0.6347,
+      "step": 6361
+    },
+    {
+      "epoch": 1.762326869806094,
+      "grad_norm": 0.730577290058136,
+      "learning_rate": 4.029502857373419e-06,
+      "loss": 0.6324,
+      "step": 6362
+    },
+    {
+      "epoch": 1.7626038781163436,
+      "grad_norm": 0.6651086211204529,
+      "learning_rate": 4.02921468758259e-06,
+      "loss": 0.5205,
+      "step": 6363
+    },
+    {
+      "epoch": 1.7628808864265928,
+      "grad_norm": 0.7027130126953125,
+      "learning_rate": 4.028926485322627e-06,
+      "loss": 0.5416,
+      "step": 6364
+    },
+    {
+      "epoch": 1.763157894736842,
+      "grad_norm": 0.7234609723091125,
+      "learning_rate": 4.0286382505996505e-06,
+      "loss": 0.6434,
+      "step": 6365
+    },
+    {
+      "epoch": 1.7634349030470915,
+      "grad_norm": 0.6769989132881165,
+      "learning_rate": 4.02834998341978e-06,
+      "loss": 0.6248,
+      "step": 6366
+    },
+    {
+      "epoch": 1.7637119113573407,
+      "grad_norm": 0.7025700807571411,
+      "learning_rate": 4.028061683789137e-06,
+      "loss": 0.6469,
+      "step": 6367
+    },
+    {
+      "epoch": 1.76398891966759,
+      "grad_norm": 0.6738390326499939,
+      "learning_rate": 4.027773351713843e-06,
+      "loss": 0.6481,
+      "step": 6368
+    },
+    {
+      "epoch": 1.7642659279778394,
+      "grad_norm": 0.7171605229377747,
+      "learning_rate": 4.027484987200017e-06,
+      "loss": 0.6232,
+      "step": 6369
+    },
+    {
+      "epoch": 1.7645429362880887,
+      "grad_norm": 0.707499623298645,
+      "learning_rate": 4.027196590253786e-06,
+      "loss": 0.637,
+      "step": 6370
+    },
+    {
+      "epoch": 1.764819944598338,
+      "grad_norm": 0.6891076564788818,
+      "learning_rate": 4.026908160881271e-06,
+      "loss": 0.6388,
+      "step": 6371
+    },
+    {
+      "epoch": 1.7650969529085874,
+      "grad_norm": 0.7460578083992004,
+      "learning_rate": 4.0266196990885955e-06,
+      "loss": 0.6599,
+      "step": 6372
+    },
+    {
+      "epoch": 1.7653739612188366,
+      "grad_norm": 0.7125904560089111,
+      "learning_rate": 4.026331204881886e-06,
+      "loss": 0.6781,
+      "step": 6373
+    },
+    {
+      "epoch": 1.7656509695290858,
+      "grad_norm": 0.7035360932350159,
+      "learning_rate": 4.026042678267267e-06,
+      "loss": 0.6337,
+      "step": 6374
+    },
+    {
+      "epoch": 1.7659279778393353,
+      "grad_norm": 0.6471675038337708,
+      "learning_rate": 4.025754119250865e-06,
+      "loss": 0.6037,
+      "step": 6375
+    },
+    {
+      "epoch": 1.7662049861495843,
+      "grad_norm": 0.6844245195388794,
+      "learning_rate": 4.025465527838806e-06,
+      "loss": 0.6409,
+      "step": 6376
+    },
+    {
+      "epoch": 1.7664819944598338,
+      "grad_norm": 0.626818060874939,
+      "learning_rate": 4.025176904037219e-06,
+      "loss": 0.5556,
+      "step": 6377
+    },
+    {
+      "epoch": 1.7667590027700832,
+      "grad_norm": 0.7065606713294983,
+      "learning_rate": 4.024888247852231e-06,
+      "loss": 0.6036,
+      "step": 6378
+    },
+    {
+      "epoch": 1.7670360110803323,
+      "grad_norm": 0.7180307507514954,
+      "learning_rate": 4.024599559289972e-06,
+      "loss": 0.6045,
+      "step": 6379
+    },
+    {
+      "epoch": 1.7673130193905817,
+      "grad_norm": 0.7368504405021667,
+      "learning_rate": 4.02431083835657e-06,
+      "loss": 0.7031,
+      "step": 6380
+    },
+    {
+      "epoch": 1.7675900277008312,
+      "grad_norm": 0.7280147075653076,
+      "learning_rate": 4.024022085058157e-06,
+      "loss": 0.6692,
+      "step": 6381
+    },
+    {
+      "epoch": 1.7678670360110802,
+      "grad_norm": 0.6778074502944946,
+      "learning_rate": 4.0237332994008625e-06,
+      "loss": 0.5773,
+      "step": 6382
+    },
+    {
+      "epoch": 1.7681440443213297,
+      "grad_norm": 0.6655060052871704,
+      "learning_rate": 4.0234444813908195e-06,
+      "loss": 0.5874,
+      "step": 6383
+    },
+    {
+      "epoch": 1.768421052631579,
+      "grad_norm": 0.6540441513061523,
+      "learning_rate": 4.0231556310341595e-06,
+      "loss": 0.5792,
+      "step": 6384
+    },
+    {
+      "epoch": 1.7686980609418281,
+      "grad_norm": 0.6953973174095154,
+      "learning_rate": 4.022866748337016e-06,
+      "loss": 0.6148,
+      "step": 6385
+    },
+    {
+      "epoch": 1.7689750692520776,
+      "grad_norm": 0.707272469997406,
+      "learning_rate": 4.022577833305522e-06,
+      "loss": 0.589,
+      "step": 6386
+    },
+    {
+      "epoch": 1.7692520775623268,
+      "grad_norm": 0.6879422664642334,
+      "learning_rate": 4.022288885945813e-06,
+      "loss": 0.623,
+      "step": 6387
+    },
+    {
+      "epoch": 1.769529085872576,
+      "grad_norm": 0.6887927651405334,
+      "learning_rate": 4.021999906264023e-06,
+      "loss": 0.6404,
+      "step": 6388
+    },
+    {
+      "epoch": 1.7698060941828255,
+      "grad_norm": 0.7096054553985596,
+      "learning_rate": 4.021710894266288e-06,
+      "loss": 0.6306,
+      "step": 6389
+    },
+    {
+      "epoch": 1.7700831024930748,
+      "grad_norm": 0.6833407878875732,
+      "learning_rate": 4.021421849958745e-06,
+      "loss": 0.6155,
+      "step": 6390
+    },
+    {
+      "epoch": 1.770360110803324,
+      "grad_norm": 0.6665039658546448,
+      "learning_rate": 4.0211327733475305e-06,
+      "loss": 0.6208,
+      "step": 6391
+    },
+    {
+      "epoch": 1.7706371191135735,
+      "grad_norm": 0.661353588104248,
+      "learning_rate": 4.020843664438783e-06,
+      "loss": 0.5745,
+      "step": 6392
+    },
+    {
+      "epoch": 1.7709141274238227,
+      "grad_norm": 0.6623826622962952,
+      "learning_rate": 4.020554523238641e-06,
+      "loss": 0.5878,
+      "step": 6393
+    },
+    {
+      "epoch": 1.771191135734072,
+      "grad_norm": 0.6898151636123657,
+      "learning_rate": 4.0202653497532425e-06,
+      "loss": 0.6625,
+      "step": 6394
+    },
+    {
+      "epoch": 1.7714681440443214,
+      "grad_norm": 0.7333918809890747,
+      "learning_rate": 4.019976143988729e-06,
+      "loss": 0.5955,
+      "step": 6395
+    },
+    {
+      "epoch": 1.7717451523545706,
+      "grad_norm": 0.6648381948471069,
+      "learning_rate": 4.01968690595124e-06,
+      "loss": 0.6256,
+      "step": 6396
+    },
+    {
+      "epoch": 1.7720221606648199,
+      "grad_norm": 0.6879984140396118,
+      "learning_rate": 4.019397635646918e-06,
+      "loss": 0.6013,
+      "step": 6397
+    },
+    {
+      "epoch": 1.7722991689750693,
+      "grad_norm": 0.7033804655075073,
+      "learning_rate": 4.019108333081903e-06,
+      "loss": 0.6112,
+      "step": 6398
+    },
+    {
+      "epoch": 1.7725761772853186,
+      "grad_norm": 0.6876200437545776,
+      "learning_rate": 4.018818998262339e-06,
+      "loss": 0.6513,
+      "step": 6399
+    },
+    {
+      "epoch": 1.7728531855955678,
+      "grad_norm": 0.7130957841873169,
+      "learning_rate": 4.01852963119437e-06,
+      "loss": 0.6725,
+      "step": 6400
+    },
+    {
+      "epoch": 1.7731301939058173,
+      "grad_norm": 0.7180396318435669,
+      "learning_rate": 4.018240231884137e-06,
+      "loss": 0.6402,
+      "step": 6401
+    },
+    {
+      "epoch": 1.7734072022160665,
+      "grad_norm": 0.6846297383308411,
+      "learning_rate": 4.017950800337789e-06,
+      "loss": 0.5692,
+      "step": 6402
+    },
+    {
+      "epoch": 1.7736842105263158,
+      "grad_norm": 0.666314423084259,
+      "learning_rate": 4.017661336561467e-06,
+      "loss": 0.6093,
+      "step": 6403
+    },
+    {
+      "epoch": 1.7739612188365652,
+      "grad_norm": 0.703360915184021,
+      "learning_rate": 4.0173718405613205e-06,
+      "loss": 0.6181,
+      "step": 6404
+    },
+    {
+      "epoch": 1.7742382271468145,
+      "grad_norm": 0.7042622566223145,
+      "learning_rate": 4.017082312343494e-06,
+      "loss": 0.6297,
+      "step": 6405
+    },
+    {
+      "epoch": 1.7745152354570637,
+      "grad_norm": 0.6913483738899231,
+      "learning_rate": 4.016792751914137e-06,
+      "loss": 0.628,
+      "step": 6406
+    },
+    {
+      "epoch": 1.7747922437673131,
+      "grad_norm": 0.7046165466308594,
+      "learning_rate": 4.016503159279395e-06,
+      "loss": 0.6544,
+      "step": 6407
+    },
+    {
+      "epoch": 1.7750692520775622,
+      "grad_norm": 0.7160999178886414,
+      "learning_rate": 4.016213534445419e-06,
+      "loss": 0.5549,
+      "step": 6408
+    },
+    {
+      "epoch": 1.7753462603878116,
+      "grad_norm": 0.706745445728302,
+      "learning_rate": 4.015923877418357e-06,
+      "loss": 0.6263,
+      "step": 6409
+    },
+    {
+      "epoch": 1.775623268698061,
+      "grad_norm": 0.6807883977890015,
+      "learning_rate": 4.015634188204361e-06,
+      "loss": 0.5961,
+      "step": 6410
+    },
+    {
+      "epoch": 1.77590027700831,
+      "grad_norm": 0.737921953201294,
+      "learning_rate": 4.01534446680958e-06,
+      "loss": 0.6938,
+      "step": 6411
+    },
+    {
+      "epoch": 1.7761772853185596,
+      "grad_norm": 0.6510135531425476,
+      "learning_rate": 4.015054713240166e-06,
+      "loss": 0.5828,
+      "step": 6412
+    },
+    {
+      "epoch": 1.776454293628809,
+      "grad_norm": 0.6969490051269531,
+      "learning_rate": 4.014764927502272e-06,
+      "loss": 0.6384,
+      "step": 6413
+    },
+    {
+      "epoch": 1.776731301939058,
+      "grad_norm": 0.7064840197563171,
+      "learning_rate": 4.01447510960205e-06,
+      "loss": 0.6255,
+      "step": 6414
+    },
+    {
+      "epoch": 1.7770083102493075,
+      "grad_norm": 0.7026253938674927,
+      "learning_rate": 4.014185259545654e-06,
+      "loss": 0.639,
+      "step": 6415
+    },
+    {
+      "epoch": 1.7772853185595567,
+      "grad_norm": 0.683268129825592,
+      "learning_rate": 4.013895377339238e-06,
+      "loss": 0.588,
+      "step": 6416
+    },
+    {
+      "epoch": 1.777562326869806,
+      "grad_norm": 0.6896020174026489,
+      "learning_rate": 4.013605462988957e-06,
+      "loss": 0.6428,
+      "step": 6417
+    },
+    {
+      "epoch": 1.7778393351800554,
+      "grad_norm": 0.7973686456680298,
+      "learning_rate": 4.013315516500966e-06,
+      "loss": 0.6393,
+      "step": 6418
+    },
+    {
+      "epoch": 1.7781163434903047,
+      "grad_norm": 0.6856012344360352,
+      "learning_rate": 4.0130255378814235e-06,
+      "loss": 0.5807,
+      "step": 6419
+    },
+    {
+      "epoch": 1.778393351800554,
+      "grad_norm": 0.7140977382659912,
+      "learning_rate": 4.0127355271364845e-06,
+      "loss": 0.6148,
+      "step": 6420
+    },
+    {
+      "epoch": 1.7786703601108034,
+      "grad_norm": 0.6942720413208008,
+      "learning_rate": 4.012445484272307e-06,
+      "loss": 0.642,
+      "step": 6421
+    },
+    {
+      "epoch": 1.7789473684210526,
+      "grad_norm": 0.7332590818405151,
+      "learning_rate": 4.0121554092950485e-06,
+      "loss": 0.604,
+      "step": 6422
+    },
+    {
+      "epoch": 1.7792243767313018,
+      "grad_norm": 0.7232838273048401,
+      "learning_rate": 4.01186530221087e-06,
+      "loss": 0.6172,
+      "step": 6423
+    },
+    {
+      "epoch": 1.7795013850415513,
+      "grad_norm": 0.7233272194862366,
+      "learning_rate": 4.011575163025931e-06,
+      "loss": 0.6529,
+      "step": 6424
+    },
+    {
+      "epoch": 1.7797783933518005,
+      "grad_norm": 0.6849462389945984,
+      "learning_rate": 4.0112849917463905e-06,
+      "loss": 0.6532,
+      "step": 6425
+    },
+    {
+      "epoch": 1.7800554016620498,
+      "grad_norm": 0.7064739465713501,
+      "learning_rate": 4.01099478837841e-06,
+      "loss": 0.6258,
+      "step": 6426
+    },
+    {
+      "epoch": 1.7803324099722992,
+      "grad_norm": 0.691041886806488,
+      "learning_rate": 4.010704552928151e-06,
+      "loss": 0.5841,
+      "step": 6427
+    },
+    {
+      "epoch": 1.7806094182825485,
+      "grad_norm": 0.6790480017662048,
+      "learning_rate": 4.010414285401777e-06,
+      "loss": 0.6473,
+      "step": 6428
+    },
+    {
+      "epoch": 1.7808864265927977,
+      "grad_norm": 0.6644400954246521,
+      "learning_rate": 4.0101239858054504e-06,
+      "loss": 0.6176,
+      "step": 6429
+    },
+    {
+      "epoch": 1.7811634349030472,
+      "grad_norm": 0.6953248381614685,
+      "learning_rate": 4.009833654145335e-06,
+      "loss": 0.6312,
+      "step": 6430
+    },
+    {
+      "epoch": 1.7814404432132964,
+      "grad_norm": 0.6625707745552063,
+      "learning_rate": 4.009543290427596e-06,
+      "loss": 0.6244,
+      "step": 6431
+    },
+    {
+      "epoch": 1.7817174515235457,
+      "grad_norm": 0.6987664699554443,
+      "learning_rate": 4.009252894658398e-06,
+      "loss": 0.6033,
+      "step": 6432
+    },
+    {
+      "epoch": 1.7819944598337951,
+      "grad_norm": 0.6788783669471741,
+      "learning_rate": 4.008962466843907e-06,
+      "loss": 0.6241,
+      "step": 6433
+    },
+    {
+      "epoch": 1.7822714681440444,
+      "grad_norm": 0.7290467023849487,
+      "learning_rate": 4.008672006990289e-06,
+      "loss": 0.637,
+      "step": 6434
+    },
+    {
+      "epoch": 1.7825484764542936,
+      "grad_norm": 0.6927595734596252,
+      "learning_rate": 4.008381515103711e-06,
+      "loss": 0.608,
+      "step": 6435
+    },
+    {
+      "epoch": 1.782825484764543,
+      "grad_norm": 0.7004909515380859,
+      "learning_rate": 4.008090991190341e-06,
+      "loss": 0.6295,
+      "step": 6436
+    },
+    {
+      "epoch": 1.7831024930747923,
+      "grad_norm": 0.6821027994155884,
+      "learning_rate": 4.007800435256349e-06,
+      "loss": 0.6072,
+      "step": 6437
+    },
+    {
+      "epoch": 1.7833795013850415,
+      "grad_norm": 0.7391160130500793,
+      "learning_rate": 4.007509847307903e-06,
+      "loss": 0.6398,
+      "step": 6438
+    },
+    {
+      "epoch": 1.783656509695291,
+      "grad_norm": 0.6901292204856873,
+      "learning_rate": 4.007219227351173e-06,
+      "loss": 0.6062,
+      "step": 6439
+    },
+    {
+      "epoch": 1.78393351800554,
+      "grad_norm": 0.7144735455513,
+      "learning_rate": 4.00692857539233e-06,
+      "loss": 0.6049,
+      "step": 6440
+    },
+    {
+      "epoch": 1.7842105263157895,
+      "grad_norm": 0.7031808495521545,
+      "learning_rate": 4.006637891437546e-06,
+      "loss": 0.6131,
+      "step": 6441
+    },
+    {
+      "epoch": 1.784487534626039,
+      "grad_norm": 0.6666425466537476,
+      "learning_rate": 4.006347175492991e-06,
+      "loss": 0.6234,
+      "step": 6442
+    },
+    {
+      "epoch": 1.784764542936288,
+      "grad_norm": 0.7020919919013977,
+      "learning_rate": 4.00605642756484e-06,
+      "loss": 0.5821,
+      "step": 6443
+    },
+    {
+      "epoch": 1.7850415512465374,
+      "grad_norm": 0.7614548206329346,
+      "learning_rate": 4.005765647659263e-06,
+      "loss": 0.6563,
+      "step": 6444
+    },
+    {
+      "epoch": 1.7853185595567869,
+      "grad_norm": 0.6820983290672302,
+      "learning_rate": 4.005474835782437e-06,
+      "loss": 0.6032,
+      "step": 6445
+    },
+    {
+      "epoch": 1.7855955678670359,
+      "grad_norm": 0.6977035403251648,
+      "learning_rate": 4.0051839919405355e-06,
+      "loss": 0.6408,
+      "step": 6446
+    },
+    {
+      "epoch": 1.7858725761772853,
+      "grad_norm": 0.6919875741004944,
+      "learning_rate": 4.004893116139735e-06,
+      "loss": 0.6158,
+      "step": 6447
+    },
+    {
+      "epoch": 1.7861495844875346,
+      "grad_norm": 0.6647782921791077,
+      "learning_rate": 4.004602208386209e-06,
+      "loss": 0.6182,
+      "step": 6448
+    },
+    {
+      "epoch": 1.7864265927977838,
+      "grad_norm": 0.6721411347389221,
+      "learning_rate": 4.004311268686136e-06,
+      "loss": 0.65,
+      "step": 6449
+    },
+    {
+      "epoch": 1.7867036011080333,
+      "grad_norm": 0.7224037051200867,
+      "learning_rate": 4.0040202970456936e-06,
+      "loss": 0.6441,
+      "step": 6450
+    },
+    {
+      "epoch": 1.7869806094182825,
+      "grad_norm": 0.6760212182998657,
+      "learning_rate": 4.003729293471059e-06,
+      "loss": 0.597,
+      "step": 6451
+    },
+    {
+      "epoch": 1.7872576177285318,
+      "grad_norm": 0.6822620034217834,
+      "learning_rate": 4.003438257968413e-06,
+      "loss": 0.6238,
+      "step": 6452
+    },
+    {
+      "epoch": 1.7875346260387812,
+      "grad_norm": 0.7053095102310181,
+      "learning_rate": 4.003147190543931e-06,
+      "loss": 0.6508,
+      "step": 6453
+    },
+    {
+      "epoch": 1.7878116343490305,
+      "grad_norm": 0.6813586950302124,
+      "learning_rate": 4.002856091203797e-06,
+      "loss": 0.6425,
+      "step": 6454
+    },
+    {
+      "epoch": 1.7880886426592797,
+      "grad_norm": 0.693324863910675,
+      "learning_rate": 4.00256495995419e-06,
+      "loss": 0.6425,
+      "step": 6455
+    },
+    {
+      "epoch": 1.7883656509695292,
+      "grad_norm": 0.6755707263946533,
+      "learning_rate": 4.002273796801292e-06,
+      "loss": 0.6227,
+      "step": 6456
+    },
+    {
+      "epoch": 1.7886426592797784,
+      "grad_norm": 0.6835914850234985,
+      "learning_rate": 4.001982601751285e-06,
+      "loss": 0.5766,
+      "step": 6457
+    },
+    {
+      "epoch": 1.7889196675900276,
+      "grad_norm": 0.6911206841468811,
+      "learning_rate": 4.001691374810352e-06,
+      "loss": 0.615,
+      "step": 6458
+    },
+    {
+      "epoch": 1.789196675900277,
+      "grad_norm": 0.7716637253761292,
+      "learning_rate": 4.001400115984675e-06,
+      "loss": 0.6435,
+      "step": 6459
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.671084463596344,
+      "learning_rate": 4.001108825280441e-06,
+      "loss": 0.6313,
+      "step": 6460
+    },
+    {
+      "epoch": 1.7897506925207756,
+      "grad_norm": 0.6706188321113586,
+      "learning_rate": 4.000817502703832e-06,
+      "loss": 0.6186,
+      "step": 6461
+    },
+    {
+      "epoch": 1.790027700831025,
+      "grad_norm": 0.7258620858192444,
+      "learning_rate": 4.000526148261035e-06,
+      "loss": 0.6582,
+      "step": 6462
+    },
+    {
+      "epoch": 1.7903047091412743,
+      "grad_norm": 0.6809777617454529,
+      "learning_rate": 4.000234761958235e-06,
+      "loss": 0.6067,
+      "step": 6463
+    },
+    {
+      "epoch": 1.7905817174515235,
+      "grad_norm": 0.7288227677345276,
+      "learning_rate": 3.999943343801621e-06,
+      "loss": 0.6414,
+      "step": 6464
+    },
+    {
+      "epoch": 1.790858725761773,
+      "grad_norm": 0.7226935029029846,
+      "learning_rate": 3.99965189379738e-06,
+      "loss": 0.6005,
+      "step": 6465
+    },
+    {
+      "epoch": 1.7911357340720222,
+      "grad_norm": 0.716136634349823,
+      "learning_rate": 3.999360411951698e-06,
+      "loss": 0.6223,
+      "step": 6466
+    },
+    {
+      "epoch": 1.7914127423822714,
+      "grad_norm": 0.6857314109802246,
+      "learning_rate": 3.999068898270767e-06,
+      "loss": 0.6374,
+      "step": 6467
+    },
+    {
+      "epoch": 1.791689750692521,
+      "grad_norm": 0.7114225625991821,
+      "learning_rate": 3.998777352760774e-06,
+      "loss": 0.6016,
+      "step": 6468
+    },
+    {
+      "epoch": 1.7919667590027701,
+      "grad_norm": 0.7072104811668396,
+      "learning_rate": 3.9984857754279105e-06,
+      "loss": 0.5931,
+      "step": 6469
+    },
+    {
+      "epoch": 1.7922437673130194,
+      "grad_norm": 0.7062807679176331,
+      "learning_rate": 3.9981941662783675e-06,
+      "loss": 0.6769,
+      "step": 6470
+    },
+    {
+      "epoch": 1.7925207756232688,
+      "grad_norm": 0.6915931701660156,
+      "learning_rate": 3.997902525318337e-06,
+      "loss": 0.5963,
+      "step": 6471
+    },
+    {
+      "epoch": 1.7927977839335179,
+      "grad_norm": 0.6964583992958069,
+      "learning_rate": 3.997610852554009e-06,
+      "loss": 0.6054,
+      "step": 6472
+    },
+    {
+      "epoch": 1.7930747922437673,
+      "grad_norm": 0.7309070229530334,
+      "learning_rate": 3.99731914799158e-06,
+      "loss": 0.5954,
+      "step": 6473
+    },
+    {
+      "epoch": 1.7933518005540168,
+      "grad_norm": 0.6806193590164185,
+      "learning_rate": 3.997027411637241e-06,
+      "loss": 0.6184,
+      "step": 6474
+    },
+    {
+      "epoch": 1.7936288088642658,
+      "grad_norm": 0.7490919828414917,
+      "learning_rate": 3.996735643497187e-06,
+      "loss": 0.6083,
+      "step": 6475
+    },
+    {
+      "epoch": 1.7939058171745152,
+      "grad_norm": 0.6906395554542542,
+      "learning_rate": 3.996443843577612e-06,
+      "loss": 0.5904,
+      "step": 6476
+    },
+    {
+      "epoch": 1.7941828254847645,
+      "grad_norm": 0.7120215892791748,
+      "learning_rate": 3.9961520118847145e-06,
+      "loss": 0.61,
+      "step": 6477
+    },
+    {
+      "epoch": 1.7944598337950137,
+      "grad_norm": 0.7228304147720337,
+      "learning_rate": 3.995860148424689e-06,
+      "loss": 0.6304,
+      "step": 6478
+    },
+    {
+      "epoch": 1.7947368421052632,
+      "grad_norm": 0.7273550629615784,
+      "learning_rate": 3.995568253203731e-06,
+      "loss": 0.651,
+      "step": 6479
+    },
+    {
+      "epoch": 1.7950138504155124,
+      "grad_norm": 0.6840648055076599,
+      "learning_rate": 3.99527632622804e-06,
+      "loss": 0.5863,
+      "step": 6480
+    },
+    {
+      "epoch": 1.7952908587257617,
+      "grad_norm": 0.7078438997268677,
+      "learning_rate": 3.994984367503815e-06,
+      "loss": 0.66,
+      "step": 6481
+    },
+    {
+      "epoch": 1.7955678670360111,
+      "grad_norm": 0.7198485732078552,
+      "learning_rate": 3.994692377037254e-06,
+      "loss": 0.5683,
+      "step": 6482
+    },
+    {
+      "epoch": 1.7958448753462604,
+      "grad_norm": 0.7349383234977722,
+      "learning_rate": 3.994400354834556e-06,
+      "loss": 0.5938,
+      "step": 6483
+    },
+    {
+      "epoch": 1.7961218836565096,
+      "grad_norm": 0.6952342391014099,
+      "learning_rate": 3.994108300901923e-06,
+      "loss": 0.6079,
+      "step": 6484
+    },
+    {
+      "epoch": 1.796398891966759,
+      "grad_norm": 0.7039951682090759,
+      "learning_rate": 3.993816215245555e-06,
+      "loss": 0.6323,
+      "step": 6485
+    },
+    {
+      "epoch": 1.7966759002770083,
+      "grad_norm": 0.7002725005149841,
+      "learning_rate": 3.993524097871654e-06,
+      "loss": 0.6215,
+      "step": 6486
+    },
+    {
+      "epoch": 1.7969529085872575,
+      "grad_norm": 0.6925804615020752,
+      "learning_rate": 3.993231948786421e-06,
+      "loss": 0.6016,
+      "step": 6487
+    },
+    {
+      "epoch": 1.797229916897507,
+      "grad_norm": 0.6799862384796143,
+      "learning_rate": 3.992939767996063e-06,
+      "loss": 0.6471,
+      "step": 6488
+    },
+    {
+      "epoch": 1.7975069252077562,
+      "grad_norm": 0.6949800252914429,
+      "learning_rate": 3.992647555506779e-06,
+      "loss": 0.6095,
+      "step": 6489
+    },
+    {
+      "epoch": 1.7977839335180055,
+      "grad_norm": 0.707348108291626,
+      "learning_rate": 3.9923553113247764e-06,
+      "loss": 0.5838,
+      "step": 6490
+    },
+    {
+      "epoch": 1.798060941828255,
+      "grad_norm": 0.7468195557594299,
+      "learning_rate": 3.9920630354562595e-06,
+      "loss": 0.6265,
+      "step": 6491
+    },
+    {
+      "epoch": 1.7983379501385042,
+      "grad_norm": 0.6571547389030457,
+      "learning_rate": 3.991770727907434e-06,
+      "loss": 0.6172,
+      "step": 6492
+    },
+    {
+      "epoch": 1.7986149584487534,
+      "grad_norm": 0.6744096279144287,
+      "learning_rate": 3.9914783886845066e-06,
+      "loss": 0.5914,
+      "step": 6493
+    },
+    {
+      "epoch": 1.7988919667590029,
+      "grad_norm": 0.7042905688285828,
+      "learning_rate": 3.991186017793683e-06,
+      "loss": 0.6326,
+      "step": 6494
+    },
+    {
+      "epoch": 1.799168975069252,
+      "grad_norm": 0.6672032475471497,
+      "learning_rate": 3.990893615241174e-06,
+      "loss": 0.5663,
+      "step": 6495
+    },
+    {
+      "epoch": 1.7994459833795013,
+      "grad_norm": 0.6901209950447083,
+      "learning_rate": 3.990601181033185e-06,
+      "loss": 0.636,
+      "step": 6496
+    },
+    {
+      "epoch": 1.7997229916897508,
+      "grad_norm": 0.6690481901168823,
+      "learning_rate": 3.990308715175927e-06,
+      "loss": 0.6141,
+      "step": 6497
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7317468523979187,
+      "learning_rate": 3.9900162176756085e-06,
+      "loss": 0.6357,
+      "step": 6498
+    },
+    {
+      "epoch": 1.8002770083102493,
+      "grad_norm": 0.6796091198921204,
+      "learning_rate": 3.9897236885384415e-06,
+      "loss": 0.5713,
+      "step": 6499
+    },
+    {
+      "epoch": 1.8005540166204987,
+      "grad_norm": 0.7144341468811035,
+      "learning_rate": 3.989431127770635e-06,
+      "loss": 0.6112,
+      "step": 6500
+    },
+    {
+      "epoch": 1.8008310249307478,
+      "grad_norm": 0.6569699048995972,
+      "learning_rate": 3.989138535378404e-06,
+      "loss": 0.5959,
+      "step": 6501
+    },
+    {
+      "epoch": 1.8011080332409972,
+      "grad_norm": 0.7061322927474976,
+      "learning_rate": 3.988845911367957e-06,
+      "loss": 0.6334,
+      "step": 6502
+    },
+    {
+      "epoch": 1.8013850415512467,
+      "grad_norm": 0.6743340492248535,
+      "learning_rate": 3.988553255745511e-06,
+      "loss": 0.5976,
+      "step": 6503
+    },
+    {
+      "epoch": 1.8016620498614957,
+      "grad_norm": 0.6997945308685303,
+      "learning_rate": 3.988260568517277e-06,
+      "loss": 0.639,
+      "step": 6504
+    },
+    {
+      "epoch": 1.8019390581717452,
+      "grad_norm": 0.6599920988082886,
+      "learning_rate": 3.98796784968947e-06,
+      "loss": 0.5883,
+      "step": 6505
+    },
+    {
+      "epoch": 1.8022160664819946,
+      "grad_norm": 0.6996617317199707,
+      "learning_rate": 3.9876750992683065e-06,
+      "loss": 0.6091,
+      "step": 6506
+    },
+    {
+      "epoch": 1.8024930747922436,
+      "grad_norm": 0.709029495716095,
+      "learning_rate": 3.987382317260001e-06,
+      "loss": 0.5815,
+      "step": 6507
+    },
+    {
+      "epoch": 1.802770083102493,
+      "grad_norm": 0.7156533598899841,
+      "learning_rate": 3.987089503670771e-06,
+      "loss": 0.6297,
+      "step": 6508
+    },
+    {
+      "epoch": 1.8030470914127423,
+      "grad_norm": 0.7280424237251282,
+      "learning_rate": 3.986796658506833e-06,
+      "loss": 0.6483,
+      "step": 6509
+    },
+    {
+      "epoch": 1.8033240997229916,
+      "grad_norm": 0.7090050578117371,
+      "learning_rate": 3.986503781774406e-06,
+      "loss": 0.5907,
+      "step": 6510
+    },
+    {
+      "epoch": 1.803601108033241,
+      "grad_norm": 0.6697266101837158,
+      "learning_rate": 3.986210873479706e-06,
+      "loss": 0.6084,
+      "step": 6511
+    },
+    {
+      "epoch": 1.8038781163434903,
+      "grad_norm": 0.7092403173446655,
+      "learning_rate": 3.985917933628955e-06,
+      "loss": 0.6487,
+      "step": 6512
+    },
+    {
+      "epoch": 1.8041551246537395,
+      "grad_norm": 0.7133142948150635,
+      "learning_rate": 3.985624962228371e-06,
+      "loss": 0.6381,
+      "step": 6513
+    },
+    {
+      "epoch": 1.804432132963989,
+      "grad_norm": 0.6582397818565369,
+      "learning_rate": 3.9853319592841755e-06,
+      "loss": 0.596,
+      "step": 6514
+    },
+    {
+      "epoch": 1.8047091412742382,
+      "grad_norm": 0.7293874621391296,
+      "learning_rate": 3.985038924802589e-06,
+      "loss": 0.6,
+      "step": 6515
+    },
+    {
+      "epoch": 1.8049861495844874,
+      "grad_norm": 0.6582150459289551,
+      "learning_rate": 3.984745858789835e-06,
+      "loss": 0.6201,
+      "step": 6516
+    },
+    {
+      "epoch": 1.805263157894737,
+      "grad_norm": 0.7221015095710754,
+      "learning_rate": 3.9844527612521335e-06,
+      "loss": 0.5975,
+      "step": 6517
+    },
+    {
+      "epoch": 1.8055401662049861,
+      "grad_norm": 0.706571102142334,
+      "learning_rate": 3.984159632195709e-06,
+      "loss": 0.6357,
+      "step": 6518
+    },
+    {
+      "epoch": 1.8058171745152354,
+      "grad_norm": 0.6357813477516174,
+      "learning_rate": 3.983866471626786e-06,
+      "loss": 0.5697,
+      "step": 6519
+    },
+    {
+      "epoch": 1.8060941828254848,
+      "grad_norm": 0.7110602855682373,
+      "learning_rate": 3.983573279551588e-06,
+      "loss": 0.6401,
+      "step": 6520
+    },
+    {
+      "epoch": 1.806371191135734,
+      "grad_norm": 0.6931176781654358,
+      "learning_rate": 3.983280055976341e-06,
+      "loss": 0.6113,
+      "step": 6521
+    },
+    {
+      "epoch": 1.8066481994459833,
+      "grad_norm": 0.6891133785247803,
+      "learning_rate": 3.98298680090727e-06,
+      "loss": 0.6379,
+      "step": 6522
+    },
+    {
+      "epoch": 1.8069252077562328,
+      "grad_norm": 0.659149706363678,
+      "learning_rate": 3.982693514350602e-06,
+      "loss": 0.5731,
+      "step": 6523
+    },
+    {
+      "epoch": 1.807202216066482,
+      "grad_norm": 0.6716209650039673,
+      "learning_rate": 3.982400196312565e-06,
+      "loss": 0.6329,
+      "step": 6524
+    },
+    {
+      "epoch": 1.8074792243767313,
+      "grad_norm": 0.6645585894584656,
+      "learning_rate": 3.982106846799386e-06,
+      "loss": 0.6286,
+      "step": 6525
+    },
+    {
+      "epoch": 1.8077562326869807,
+      "grad_norm": 0.7084228992462158,
+      "learning_rate": 3.981813465817293e-06,
+      "loss": 0.6175,
+      "step": 6526
+    },
+    {
+      "epoch": 1.80803324099723,
+      "grad_norm": 0.6945014595985413,
+      "learning_rate": 3.981520053372517e-06,
+      "loss": 0.5685,
+      "step": 6527
+    },
+    {
+      "epoch": 1.8083102493074792,
+      "grad_norm": 0.6847140789031982,
+      "learning_rate": 3.981226609471286e-06,
+      "loss": 0.5825,
+      "step": 6528
+    },
+    {
+      "epoch": 1.8085872576177286,
+      "grad_norm": 0.7313349843025208,
+      "learning_rate": 3.980933134119832e-06,
+      "loss": 0.625,
+      "step": 6529
+    },
+    {
+      "epoch": 1.8088642659279779,
+      "grad_norm": 0.7253093719482422,
+      "learning_rate": 3.980639627324386e-06,
+      "loss": 0.6793,
+      "step": 6530
+    },
+    {
+      "epoch": 1.8091412742382271,
+      "grad_norm": 0.675082802772522,
+      "learning_rate": 3.980346089091179e-06,
+      "loss": 0.6159,
+      "step": 6531
+    },
+    {
+      "epoch": 1.8094182825484766,
+      "grad_norm": 0.7030022740364075,
+      "learning_rate": 3.980052519426444e-06,
+      "loss": 0.6353,
+      "step": 6532
+    },
+    {
+      "epoch": 1.8096952908587256,
+      "grad_norm": 0.6523541808128357,
+      "learning_rate": 3.979758918336415e-06,
+      "loss": 0.6172,
+      "step": 6533
+    },
+    {
+      "epoch": 1.809972299168975,
+      "grad_norm": 0.6885202527046204,
+      "learning_rate": 3.979465285827325e-06,
+      "loss": 0.6017,
+      "step": 6534
+    },
+    {
+      "epoch": 1.8102493074792245,
+      "grad_norm": 0.7288880944252014,
+      "learning_rate": 3.979171621905409e-06,
+      "loss": 0.6022,
+      "step": 6535
+    },
+    {
+      "epoch": 1.8105263157894735,
+      "grad_norm": 0.6946640610694885,
+      "learning_rate": 3.978877926576901e-06,
+      "loss": 0.645,
+      "step": 6536
+    },
+    {
+      "epoch": 1.810803324099723,
+      "grad_norm": 0.7339047193527222,
+      "learning_rate": 3.978584199848039e-06,
+      "loss": 0.6403,
+      "step": 6537
+    },
+    {
+      "epoch": 1.8110803324099725,
+      "grad_norm": 0.6518142223358154,
+      "learning_rate": 3.9782904417250586e-06,
+      "loss": 0.566,
+      "step": 6538
+    },
+    {
+      "epoch": 1.8113573407202215,
+      "grad_norm": 0.6858428120613098,
+      "learning_rate": 3.977996652214197e-06,
+      "loss": 0.6623,
+      "step": 6539
+    },
+    {
+      "epoch": 1.811634349030471,
+      "grad_norm": 0.7024521231651306,
+      "learning_rate": 3.977702831321692e-06,
+      "loss": 0.672,
+      "step": 6540
+    },
+    {
+      "epoch": 1.8119113573407202,
+      "grad_norm": 0.7442790865898132,
+      "learning_rate": 3.977408979053781e-06,
+      "loss": 0.6212,
+      "step": 6541
+    },
+    {
+      "epoch": 1.8121883656509694,
+      "grad_norm": 0.6858800053596497,
+      "learning_rate": 3.977115095416706e-06,
+      "loss": 0.6096,
+      "step": 6542
+    },
+    {
+      "epoch": 1.8124653739612189,
+      "grad_norm": 0.7015238404273987,
+      "learning_rate": 3.9768211804167054e-06,
+      "loss": 0.6182,
+      "step": 6543
+    },
+    {
+      "epoch": 1.8127423822714681,
+      "grad_norm": 0.6668867468833923,
+      "learning_rate": 3.976527234060019e-06,
+      "loss": 0.6471,
+      "step": 6544
+    },
+    {
+      "epoch": 1.8130193905817173,
+      "grad_norm": 0.6626103520393372,
+      "learning_rate": 3.97623325635289e-06,
+      "loss": 0.6262,
+      "step": 6545
+    },
+    {
+      "epoch": 1.8132963988919668,
+      "grad_norm": 0.7281420826911926,
+      "learning_rate": 3.975939247301558e-06,
+      "loss": 0.6242,
+      "step": 6546
+    },
+    {
+      "epoch": 1.813573407202216,
+      "grad_norm": 0.7330793142318726,
+      "learning_rate": 3.975645206912267e-06,
+      "loss": 0.6098,
+      "step": 6547
+    },
+    {
+      "epoch": 1.8138504155124653,
+      "grad_norm": 0.6654852628707886,
+      "learning_rate": 3.97535113519126e-06,
+      "loss": 0.6616,
+      "step": 6548
+    },
+    {
+      "epoch": 1.8141274238227147,
+      "grad_norm": 0.7465592622756958,
+      "learning_rate": 3.975057032144779e-06,
+      "loss": 0.6399,
+      "step": 6549
+    },
+    {
+      "epoch": 1.814404432132964,
+      "grad_norm": 0.6892106533050537,
+      "learning_rate": 3.974762897779072e-06,
+      "loss": 0.6272,
+      "step": 6550
+    },
+    {
+      "epoch": 1.8146814404432132,
+      "grad_norm": 0.6937387585639954,
+      "learning_rate": 3.974468732100383e-06,
+      "loss": 0.6166,
+      "step": 6551
+    },
+    {
+      "epoch": 1.8149584487534627,
+      "grad_norm": 0.7007038593292236,
+      "learning_rate": 3.974174535114956e-06,
+      "loss": 0.629,
+      "step": 6552
+    },
+    {
+      "epoch": 1.815235457063712,
+      "grad_norm": 0.6984164714813232,
+      "learning_rate": 3.97388030682904e-06,
+      "loss": 0.6131,
+      "step": 6553
+    },
+    {
+      "epoch": 1.8155124653739612,
+      "grad_norm": 0.6829065084457397,
+      "learning_rate": 3.9735860472488806e-06,
+      "loss": 0.6543,
+      "step": 6554
+    },
+    {
+      "epoch": 1.8157894736842106,
+      "grad_norm": 0.7143397331237793,
+      "learning_rate": 3.973291756380727e-06,
+      "loss": 0.6692,
+      "step": 6555
+    },
+    {
+      "epoch": 1.8160664819944599,
+      "grad_norm": 0.7546378374099731,
+      "learning_rate": 3.972997434230826e-06,
+      "loss": 0.5824,
+      "step": 6556
+    },
+    {
+      "epoch": 1.816343490304709,
+      "grad_norm": 0.6992888450622559,
+      "learning_rate": 3.972703080805429e-06,
+      "loss": 0.6236,
+      "step": 6557
+    },
+    {
+      "epoch": 1.8166204986149586,
+      "grad_norm": 0.7112694978713989,
+      "learning_rate": 3.972408696110785e-06,
+      "loss": 0.648,
+      "step": 6558
+    },
+    {
+      "epoch": 1.8168975069252078,
+      "grad_norm": 0.7382538914680481,
+      "learning_rate": 3.972114280153143e-06,
+      "loss": 0.6204,
+      "step": 6559
+    },
+    {
+      "epoch": 1.817174515235457,
+      "grad_norm": 0.6987351179122925,
+      "learning_rate": 3.971819832938756e-06,
+      "loss": 0.5919,
+      "step": 6560
+    },
+    {
+      "epoch": 1.8174515235457065,
+      "grad_norm": 0.6935312151908875,
+      "learning_rate": 3.971525354473875e-06,
+      "loss": 0.6464,
+      "step": 6561
+    },
+    {
+      "epoch": 1.8177285318559557,
+      "grad_norm": 0.6827068328857422,
+      "learning_rate": 3.971230844764753e-06,
+      "loss": 0.613,
+      "step": 6562
+    },
+    {
+      "epoch": 1.818005540166205,
+      "grad_norm": 0.7268776297569275,
+      "learning_rate": 3.970936303817643e-06,
+      "loss": 0.6078,
+      "step": 6563
+    },
+    {
+      "epoch": 1.8182825484764544,
+      "grad_norm": 0.695104718208313,
+      "learning_rate": 3.970641731638799e-06,
+      "loss": 0.6382,
+      "step": 6564
+    },
+    {
+      "epoch": 1.8185595567867034,
+      "grad_norm": 0.6960898041725159,
+      "learning_rate": 3.9703471282344754e-06,
+      "loss": 0.6293,
+      "step": 6565
+    },
+    {
+      "epoch": 1.818836565096953,
+      "grad_norm": 0.7311190962791443,
+      "learning_rate": 3.970052493610928e-06,
+      "loss": 0.6128,
+      "step": 6566
+    },
+    {
+      "epoch": 1.8191135734072024,
+      "grad_norm": 0.6929040551185608,
+      "learning_rate": 3.969757827774412e-06,
+      "loss": 0.6017,
+      "step": 6567
+    },
+    {
+      "epoch": 1.8193905817174514,
+      "grad_norm": 0.6849446892738342,
+      "learning_rate": 3.969463130731183e-06,
+      "loss": 0.6455,
+      "step": 6568
+    },
+    {
+      "epoch": 1.8196675900277008,
+      "grad_norm": 0.685980498790741,
+      "learning_rate": 3.9691684024875e-06,
+      "loss": 0.6173,
+      "step": 6569
+    },
+    {
+      "epoch": 1.8199445983379503,
+      "grad_norm": 0.7096467614173889,
+      "learning_rate": 3.96887364304962e-06,
+      "loss": 0.5953,
+      "step": 6570
+    },
+    {
+      "epoch": 1.8202216066481993,
+      "grad_norm": 0.7245126962661743,
+      "learning_rate": 3.968578852423801e-06,
+      "loss": 0.6379,
+      "step": 6571
+    },
+    {
+      "epoch": 1.8204986149584488,
+      "grad_norm": 0.708109974861145,
+      "learning_rate": 3.9682840306163025e-06,
+      "loss": 0.5782,
+      "step": 6572
+    },
+    {
+      "epoch": 1.820775623268698,
+      "grad_norm": 0.7031292915344238,
+      "learning_rate": 3.967989177633385e-06,
+      "loss": 0.6398,
+      "step": 6573
+    },
+    {
+      "epoch": 1.8210526315789473,
+      "grad_norm": 0.6725582480430603,
+      "learning_rate": 3.9676942934813075e-06,
+      "loss": 0.5867,
+      "step": 6574
+    },
+    {
+      "epoch": 1.8213296398891967,
+      "grad_norm": 0.7046939730644226,
+      "learning_rate": 3.967399378166333e-06,
+      "loss": 0.6597,
+      "step": 6575
+    },
+    {
+      "epoch": 1.821606648199446,
+      "grad_norm": 0.7320119142532349,
+      "learning_rate": 3.967104431694722e-06,
+      "loss": 0.6644,
+      "step": 6576
+    },
+    {
+      "epoch": 1.8218836565096952,
+      "grad_norm": 0.664082944393158,
+      "learning_rate": 3.966809454072738e-06,
+      "loss": 0.6144,
+      "step": 6577
+    },
+    {
+      "epoch": 1.8221606648199447,
+      "grad_norm": 0.6912119388580322,
+      "learning_rate": 3.966514445306642e-06,
+      "loss": 0.5829,
+      "step": 6578
+    },
+    {
+      "epoch": 1.822437673130194,
+      "grad_norm": 0.6881527304649353,
+      "learning_rate": 3.9662194054027e-06,
+      "loss": 0.6185,
+      "step": 6579
+    },
+    {
+      "epoch": 1.8227146814404431,
+      "grad_norm": 0.6926375031471252,
+      "learning_rate": 3.965924334367176e-06,
+      "loss": 0.6179,
+      "step": 6580
+    },
+    {
+      "epoch": 1.8229916897506926,
+      "grad_norm": 0.6690463423728943,
+      "learning_rate": 3.965629232206335e-06,
+      "loss": 0.6289,
+      "step": 6581
+    },
+    {
+      "epoch": 1.8232686980609418,
+      "grad_norm": 0.7154456377029419,
+      "learning_rate": 3.965334098926441e-06,
+      "loss": 0.6325,
+      "step": 6582
+    },
+    {
+      "epoch": 1.823545706371191,
+      "grad_norm": 0.6615016460418701,
+      "learning_rate": 3.965038934533764e-06,
+      "loss": 0.5962,
+      "step": 6583
+    },
+    {
+      "epoch": 1.8238227146814405,
+      "grad_norm": 0.6844783425331116,
+      "learning_rate": 3.964743739034569e-06,
+      "loss": 0.6,
+      "step": 6584
+    },
+    {
+      "epoch": 1.8240997229916898,
+      "grad_norm": 0.802054762840271,
+      "learning_rate": 3.964448512435123e-06,
+      "loss": 0.612,
+      "step": 6585
+    },
+    {
+      "epoch": 1.824376731301939,
+      "grad_norm": 0.671402096748352,
+      "learning_rate": 3.964153254741695e-06,
+      "loss": 0.6258,
+      "step": 6586
+    },
+    {
+      "epoch": 1.8246537396121885,
+      "grad_norm": 0.6834940314292908,
+      "learning_rate": 3.963857965960555e-06,
+      "loss": 0.6245,
+      "step": 6587
+    },
+    {
+      "epoch": 1.8249307479224377,
+      "grad_norm": 0.7301408052444458,
+      "learning_rate": 3.963562646097972e-06,
+      "loss": 0.6207,
+      "step": 6588
+    },
+    {
+      "epoch": 1.825207756232687,
+      "grad_norm": 0.6522077918052673,
+      "learning_rate": 3.9632672951602166e-06,
+      "loss": 0.6154,
+      "step": 6589
+    },
+    {
+      "epoch": 1.8254847645429364,
+      "grad_norm": 0.7064364552497864,
+      "learning_rate": 3.9629719131535595e-06,
+      "loss": 0.6165,
+      "step": 6590
+    },
+    {
+      "epoch": 1.8257617728531856,
+      "grad_norm": 0.7141890525817871,
+      "learning_rate": 3.962676500084272e-06,
+      "loss": 0.6153,
+      "step": 6591
+    },
+    {
+      "epoch": 1.8260387811634349,
+      "grad_norm": 0.7095588445663452,
+      "learning_rate": 3.962381055958629e-06,
+      "loss": 0.6256,
+      "step": 6592
+    },
+    {
+      "epoch": 1.8263157894736843,
+      "grad_norm": 0.6960277557373047,
+      "learning_rate": 3.9620855807829e-06,
+      "loss": 0.6223,
+      "step": 6593
+    },
+    {
+      "epoch": 1.8265927977839334,
+      "grad_norm": 0.7400093078613281,
+      "learning_rate": 3.961790074563362e-06,
+      "loss": 0.6138,
+      "step": 6594
+    },
+    {
+      "epoch": 1.8268698060941828,
+      "grad_norm": 0.6910619735717773,
+      "learning_rate": 3.961494537306286e-06,
+      "loss": 0.6275,
+      "step": 6595
+    },
+    {
+      "epoch": 1.8271468144044323,
+      "grad_norm": 0.6945144534111023,
+      "learning_rate": 3.96119896901795e-06,
+      "loss": 0.6323,
+      "step": 6596
+    },
+    {
+      "epoch": 1.8274238227146813,
+      "grad_norm": 0.6521016955375671,
+      "learning_rate": 3.9609033697046275e-06,
+      "loss": 0.6486,
+      "step": 6597
+    },
+    {
+      "epoch": 1.8277008310249307,
+      "grad_norm": 0.6771376132965088,
+      "learning_rate": 3.960607739372596e-06,
+      "loss": 0.6332,
+      "step": 6598
+    },
+    {
+      "epoch": 1.8279778393351802,
+      "grad_norm": 0.6685909628868103,
+      "learning_rate": 3.960312078028132e-06,
+      "loss": 0.6175,
+      "step": 6599
+    },
+    {
+      "epoch": 1.8282548476454292,
+      "grad_norm": 0.7308417558670044,
+      "learning_rate": 3.9600163856775135e-06,
+      "loss": 0.6134,
+      "step": 6600
+    },
+    {
+      "epoch": 1.8285318559556787,
+      "grad_norm": 0.670316219329834,
+      "learning_rate": 3.959720662327018e-06,
+      "loss": 0.5853,
+      "step": 6601
+    },
+    {
+      "epoch": 1.828808864265928,
+      "grad_norm": 0.7013362646102905,
+      "learning_rate": 3.959424907982927e-06,
+      "loss": 0.647,
+      "step": 6602
+    },
+    {
+      "epoch": 1.8290858725761772,
+      "grad_norm": 0.6852704286575317,
+      "learning_rate": 3.959129122651516e-06,
+      "loss": 0.5743,
+      "step": 6603
+    },
+    {
+      "epoch": 1.8293628808864266,
+      "grad_norm": 0.6822395920753479,
+      "learning_rate": 3.958833306339069e-06,
+      "loss": 0.6286,
+      "step": 6604
+    },
+    {
+      "epoch": 1.8296398891966759,
+      "grad_norm": 0.7460020184516907,
+      "learning_rate": 3.9585374590518654e-06,
+      "loss": 0.6033,
+      "step": 6605
+    },
+    {
+      "epoch": 1.829916897506925,
+      "grad_norm": 0.641464352607727,
+      "learning_rate": 3.958241580796186e-06,
+      "loss": 0.5939,
+      "step": 6606
+    },
+    {
+      "epoch": 1.8301939058171746,
+      "grad_norm": 0.7084401249885559,
+      "learning_rate": 3.957945671578315e-06,
+      "loss": 0.6111,
+      "step": 6607
+    },
+    {
+      "epoch": 1.8304709141274238,
+      "grad_norm": 0.717842161655426,
+      "learning_rate": 3.957649731404533e-06,
+      "loss": 0.6218,
+      "step": 6608
+    },
+    {
+      "epoch": 1.830747922437673,
+      "grad_norm": 0.6891061663627625,
+      "learning_rate": 3.9573537602811256e-06,
+      "loss": 0.631,
+      "step": 6609
+    },
+    {
+      "epoch": 1.8310249307479225,
+      "grad_norm": 0.6817660927772522,
+      "learning_rate": 3.957057758214376e-06,
+      "loss": 0.6099,
+      "step": 6610
+    },
+    {
+      "epoch": 1.8313019390581717,
+      "grad_norm": 0.6792852282524109,
+      "learning_rate": 3.95676172521057e-06,
+      "loss": 0.5927,
+      "step": 6611
+    },
+    {
+      "epoch": 1.831578947368421,
+      "grad_norm": 0.7312888503074646,
+      "learning_rate": 3.9564656612759904e-06,
+      "loss": 0.6568,
+      "step": 6612
+    },
+    {
+      "epoch": 1.8318559556786704,
+      "grad_norm": 0.6890841126441956,
+      "learning_rate": 3.956169566416928e-06,
+      "loss": 0.6552,
+      "step": 6613
+    },
+    {
+      "epoch": 1.8321329639889197,
+      "grad_norm": 0.688114583492279,
+      "learning_rate": 3.955873440639665e-06,
+      "loss": 0.6047,
+      "step": 6614
+    },
+    {
+      "epoch": 1.832409972299169,
+      "grad_norm": 0.7114209532737732,
+      "learning_rate": 3.955577283950491e-06,
+      "loss": 0.6245,
+      "step": 6615
+    },
+    {
+      "epoch": 1.8326869806094184,
+      "grad_norm": 0.7178516983985901,
+      "learning_rate": 3.9552810963556955e-06,
+      "loss": 0.6064,
+      "step": 6616
+    },
+    {
+      "epoch": 1.8329639889196676,
+      "grad_norm": 0.6884858012199402,
+      "learning_rate": 3.954984877861565e-06,
+      "loss": 0.6177,
+      "step": 6617
+    },
+    {
+      "epoch": 1.8332409972299168,
+      "grad_norm": 0.6996463537216187,
+      "learning_rate": 3.95468862847439e-06,
+      "loss": 0.6554,
+      "step": 6618
+    },
+    {
+      "epoch": 1.8335180055401663,
+      "grad_norm": 0.7320404648780823,
+      "learning_rate": 3.95439234820046e-06,
+      "loss": 0.632,
+      "step": 6619
+    },
+    {
+      "epoch": 1.8337950138504155,
+      "grad_norm": 0.6940857172012329,
+      "learning_rate": 3.954096037046068e-06,
+      "loss": 0.6096,
+      "step": 6620
+    },
+    {
+      "epoch": 1.8340720221606648,
+      "grad_norm": 0.7151522636413574,
+      "learning_rate": 3.953799695017503e-06,
+      "loss": 0.6366,
+      "step": 6621
+    },
+    {
+      "epoch": 1.8343490304709142,
+      "grad_norm": 0.7152765989303589,
+      "learning_rate": 3.9535033221210574e-06,
+      "loss": 0.619,
+      "step": 6622
+    },
+    {
+      "epoch": 1.8346260387811635,
+      "grad_norm": 0.6934623718261719,
+      "learning_rate": 3.953206918363024e-06,
+      "loss": 0.629,
+      "step": 6623
+    },
+    {
+      "epoch": 1.8349030470914127,
+      "grad_norm": 0.7444363236427307,
+      "learning_rate": 3.952910483749698e-06,
+      "loss": 0.6337,
+      "step": 6624
+    },
+    {
+      "epoch": 1.8351800554016622,
+      "grad_norm": 0.7064351439476013,
+      "learning_rate": 3.952614018287371e-06,
+      "loss": 0.6309,
+      "step": 6625
+    },
+    {
+      "epoch": 1.8354570637119112,
+      "grad_norm": 0.7176104784011841,
+      "learning_rate": 3.9523175219823394e-06,
+      "loss": 0.6072,
+      "step": 6626
+    },
+    {
+      "epoch": 1.8357340720221607,
+      "grad_norm": 0.7179607152938843,
+      "learning_rate": 3.952020994840898e-06,
+      "loss": 0.6372,
+      "step": 6627
+    },
+    {
+      "epoch": 1.8360110803324101,
+      "grad_norm": 0.7062599062919617,
+      "learning_rate": 3.951724436869343e-06,
+      "loss": 0.6095,
+      "step": 6628
+    },
+    {
+      "epoch": 1.8362880886426591,
+      "grad_norm": 0.6884281635284424,
+      "learning_rate": 3.95142784807397e-06,
+      "loss": 0.5208,
+      "step": 6629
+    },
+    {
+      "epoch": 1.8365650969529086,
+      "grad_norm": 0.7327911257743835,
+      "learning_rate": 3.9511312284610785e-06,
+      "loss": 0.6686,
+      "step": 6630
+    },
+    {
+      "epoch": 1.836842105263158,
+      "grad_norm": 0.6697911024093628,
+      "learning_rate": 3.950834578036964e-06,
+      "loss": 0.6197,
+      "step": 6631
+    },
+    {
+      "epoch": 1.837119113573407,
+      "grad_norm": 0.6918233036994934,
+      "learning_rate": 3.950537896807926e-06,
+      "loss": 0.6175,
+      "step": 6632
+    },
+    {
+      "epoch": 1.8373961218836565,
+      "grad_norm": 0.7282374501228333,
+      "learning_rate": 3.950241184780266e-06,
+      "loss": 0.6044,
+      "step": 6633
+    },
+    {
+      "epoch": 1.8376731301939058,
+      "grad_norm": 0.6883581876754761,
+      "learning_rate": 3.94994444196028e-06,
+      "loss": 0.6207,
+      "step": 6634
+    },
+    {
+      "epoch": 1.837950138504155,
+      "grad_norm": 0.658677339553833,
+      "learning_rate": 3.949647668354273e-06,
+      "loss": 0.564,
+      "step": 6635
+    },
+    {
+      "epoch": 1.8382271468144045,
+      "grad_norm": 0.7224382758140564,
+      "learning_rate": 3.949350863968543e-06,
+      "loss": 0.651,
+      "step": 6636
+    },
+    {
+      "epoch": 1.8385041551246537,
+      "grad_norm": 0.675486147403717,
+      "learning_rate": 3.949054028809393e-06,
+      "loss": 0.6152,
+      "step": 6637
+    },
+    {
+      "epoch": 1.838781163434903,
+      "grad_norm": 0.7212086319923401,
+      "learning_rate": 3.9487571628831255e-06,
+      "loss": 0.6426,
+      "step": 6638
+    },
+    {
+      "epoch": 1.8390581717451524,
+      "grad_norm": 0.6830087900161743,
+      "learning_rate": 3.948460266196043e-06,
+      "loss": 0.6306,
+      "step": 6639
+    },
+    {
+      "epoch": 1.8393351800554016,
+      "grad_norm": 0.7165666222572327,
+      "learning_rate": 3.948163338754452e-06,
+      "loss": 0.6318,
+      "step": 6640
+    },
+    {
+      "epoch": 1.8396121883656509,
+      "grad_norm": 0.689459502696991,
+      "learning_rate": 3.947866380564654e-06,
+      "loss": 0.6096,
+      "step": 6641
+    },
+    {
+      "epoch": 1.8398891966759003,
+      "grad_norm": 0.6928114295005798,
+      "learning_rate": 3.9475693916329536e-06,
+      "loss": 0.6445,
+      "step": 6642
+    },
+    {
+      "epoch": 1.8401662049861496,
+      "grad_norm": 0.6756209135055542,
+      "learning_rate": 3.94727237196566e-06,
+      "loss": 0.6097,
+      "step": 6643
+    },
+    {
+      "epoch": 1.8404432132963988,
+      "grad_norm": 0.7100651860237122,
+      "learning_rate": 3.9469753215690785e-06,
+      "loss": 0.6057,
+      "step": 6644
+    },
+    {
+      "epoch": 1.8407202216066483,
+      "grad_norm": 0.6669239401817322,
+      "learning_rate": 3.946678240449515e-06,
+      "loss": 0.5732,
+      "step": 6645
+    },
+    {
+      "epoch": 1.8409972299168975,
+      "grad_norm": 0.7072580456733704,
+      "learning_rate": 3.946381128613278e-06,
+      "loss": 0.5744,
+      "step": 6646
+    },
+    {
+      "epoch": 1.8412742382271468,
+      "grad_norm": 0.6962587833404541,
+      "learning_rate": 3.946083986066676e-06,
+      "loss": 0.6353,
+      "step": 6647
+    },
+    {
+      "epoch": 1.8415512465373962,
+      "grad_norm": 0.7108308672904968,
+      "learning_rate": 3.945786812816018e-06,
+      "loss": 0.5871,
+      "step": 6648
+    },
+    {
+      "epoch": 1.8418282548476455,
+      "grad_norm": 0.7100950479507446,
+      "learning_rate": 3.945489608867614e-06,
+      "loss": 0.6274,
+      "step": 6649
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.6926645040512085,
+      "learning_rate": 3.945192374227774e-06,
+      "loss": 0.5946,
+      "step": 6650
+    },
+    {
+      "epoch": 1.8423822714681442,
+      "grad_norm": 0.7518139481544495,
+      "learning_rate": 3.94489510890281e-06,
+      "loss": 0.6722,
+      "step": 6651
+    },
+    {
+      "epoch": 1.8426592797783934,
+      "grad_norm": 0.7210699915885925,
+      "learning_rate": 3.944597812899033e-06,
+      "loss": 0.6258,
+      "step": 6652
+    },
+    {
+      "epoch": 1.8429362880886426,
+      "grad_norm": 0.7152932286262512,
+      "learning_rate": 3.944300486222754e-06,
+      "loss": 0.5441,
+      "step": 6653
+    },
+    {
+      "epoch": 1.843213296398892,
+      "grad_norm": 0.7112598419189453,
+      "learning_rate": 3.9440031288802885e-06,
+      "loss": 0.6072,
+      "step": 6654
+    },
+    {
+      "epoch": 1.8434903047091413,
+      "grad_norm": 0.6846300959587097,
+      "learning_rate": 3.943705740877949e-06,
+      "loss": 0.6604,
+      "step": 6655
+    },
+    {
+      "epoch": 1.8437673130193906,
+      "grad_norm": 0.7109817266464233,
+      "learning_rate": 3.943408322222049e-06,
+      "loss": 0.5605,
+      "step": 6656
+    },
+    {
+      "epoch": 1.84404432132964,
+      "grad_norm": 0.669664740562439,
+      "learning_rate": 3.943110872918905e-06,
+      "loss": 0.5983,
+      "step": 6657
+    },
+    {
+      "epoch": 1.844321329639889,
+      "grad_norm": 0.6921775937080383,
+      "learning_rate": 3.942813392974832e-06,
+      "loss": 0.6385,
+      "step": 6658
+    },
+    {
+      "epoch": 1.8445983379501385,
+      "grad_norm": 0.6840718388557434,
+      "learning_rate": 3.942515882396145e-06,
+      "loss": 0.6221,
+      "step": 6659
+    },
+    {
+      "epoch": 1.844875346260388,
+      "grad_norm": 0.7287196516990662,
+      "learning_rate": 3.942218341189163e-06,
+      "loss": 0.6248,
+      "step": 6660
+    },
+    {
+      "epoch": 1.845152354570637,
+      "grad_norm": 0.7228674292564392,
+      "learning_rate": 3.941920769360202e-06,
+      "loss": 0.6277,
+      "step": 6661
+    },
+    {
+      "epoch": 1.8454293628808864,
+      "grad_norm": 0.688456654548645,
+      "learning_rate": 3.941623166915582e-06,
+      "loss": 0.5955,
+      "step": 6662
+    },
+    {
+      "epoch": 1.845706371191136,
+      "grad_norm": 0.6912646293640137,
+      "learning_rate": 3.941325533861619e-06,
+      "loss": 0.6411,
+      "step": 6663
+    },
+    {
+      "epoch": 1.845983379501385,
+      "grad_norm": 0.6859648823738098,
+      "learning_rate": 3.941027870204634e-06,
+      "loss": 0.6436,
+      "step": 6664
+    },
+    {
+      "epoch": 1.8462603878116344,
+      "grad_norm": 0.7074818015098572,
+      "learning_rate": 3.94073017595095e-06,
+      "loss": 0.6248,
+      "step": 6665
+    },
+    {
+      "epoch": 1.8465373961218836,
+      "grad_norm": 0.7277980446815491,
+      "learning_rate": 3.9404324511068825e-06,
+      "loss": 0.6328,
+      "step": 6666
+    },
+    {
+      "epoch": 1.8468144044321329,
+      "grad_norm": 0.6855519413948059,
+      "learning_rate": 3.940134695678757e-06,
+      "loss": 0.6123,
+      "step": 6667
+    },
+    {
+      "epoch": 1.8470914127423823,
+      "grad_norm": 0.7200990915298462,
+      "learning_rate": 3.939836909672893e-06,
+      "loss": 0.6347,
+      "step": 6668
+    },
+    {
+      "epoch": 1.8473684210526315,
+      "grad_norm": 0.73331218957901,
+      "learning_rate": 3.939539093095616e-06,
+      "loss": 0.6002,
+      "step": 6669
+    },
+    {
+      "epoch": 1.8476454293628808,
+      "grad_norm": 0.6926178932189941,
+      "learning_rate": 3.9392412459532466e-06,
+      "loss": 0.6748,
+      "step": 6670
+    },
+    {
+      "epoch": 1.8479224376731302,
+      "grad_norm": 0.704434871673584,
+      "learning_rate": 3.93894336825211e-06,
+      "loss": 0.6212,
+      "step": 6671
+    },
+    {
+      "epoch": 1.8481994459833795,
+      "grad_norm": 0.6933774948120117,
+      "learning_rate": 3.938645459998532e-06,
+      "loss": 0.6593,
+      "step": 6672
+    },
+    {
+      "epoch": 1.8484764542936287,
+      "grad_norm": 0.7267084121704102,
+      "learning_rate": 3.938347521198837e-06,
+      "loss": 0.646,
+      "step": 6673
+    },
+    {
+      "epoch": 1.8487534626038782,
+      "grad_norm": 0.6638674736022949,
+      "learning_rate": 3.938049551859351e-06,
+      "loss": 0.6442,
+      "step": 6674
+    },
+    {
+      "epoch": 1.8490304709141274,
+      "grad_norm": 0.7296227216720581,
+      "learning_rate": 3.9377515519864e-06,
+      "loss": 0.6231,
+      "step": 6675
+    },
+    {
+      "epoch": 1.8493074792243767,
+      "grad_norm": 0.6873810887336731,
+      "learning_rate": 3.937453521586312e-06,
+      "loss": 0.6157,
+      "step": 6676
+    },
+    {
+      "epoch": 1.8495844875346261,
+      "grad_norm": 0.6810946464538574,
+      "learning_rate": 3.937155460665415e-06,
+      "loss": 0.6375,
+      "step": 6677
+    },
+    {
+      "epoch": 1.8498614958448754,
+      "grad_norm": 0.6755537986755371,
+      "learning_rate": 3.936857369230037e-06,
+      "loss": 0.5826,
+      "step": 6678
+    },
+    {
+      "epoch": 1.8501385041551246,
+      "grad_norm": 0.6881328225135803,
+      "learning_rate": 3.936559247286509e-06,
+      "loss": 0.6105,
+      "step": 6679
+    },
+    {
+      "epoch": 1.850415512465374,
+      "grad_norm": 0.7300723791122437,
+      "learning_rate": 3.936261094841159e-06,
+      "loss": 0.622,
+      "step": 6680
+    },
+    {
+      "epoch": 1.8506925207756233,
+      "grad_norm": 0.7086427807807922,
+      "learning_rate": 3.935962911900317e-06,
+      "loss": 0.6495,
+      "step": 6681
+    },
+    {
+      "epoch": 1.8509695290858725,
+      "grad_norm": 0.6515028476715088,
+      "learning_rate": 3.9356646984703175e-06,
+      "loss": 0.6109,
+      "step": 6682
+    },
+    {
+      "epoch": 1.851246537396122,
+      "grad_norm": 0.72528076171875,
+      "learning_rate": 3.935366454557488e-06,
+      "loss": 0.6304,
+      "step": 6683
+    },
+    {
+      "epoch": 1.8515235457063712,
+      "grad_norm": 0.6935245394706726,
+      "learning_rate": 3.935068180168165e-06,
+      "loss": 0.6425,
+      "step": 6684
+    },
+    {
+      "epoch": 1.8518005540166205,
+      "grad_norm": 0.7298848628997803,
+      "learning_rate": 3.934769875308678e-06,
+      "loss": 0.6371,
+      "step": 6685
+    },
+    {
+      "epoch": 1.85207756232687,
+      "grad_norm": 0.7072490453720093,
+      "learning_rate": 3.934471539985364e-06,
+      "loss": 0.5982,
+      "step": 6686
+    },
+    {
+      "epoch": 1.8523545706371192,
+      "grad_norm": 0.7231565117835999,
+      "learning_rate": 3.934173174204555e-06,
+      "loss": 0.647,
+      "step": 6687
+    },
+    {
+      "epoch": 1.8526315789473684,
+      "grad_norm": 0.7302079796791077,
+      "learning_rate": 3.933874777972589e-06,
+      "loss": 0.6555,
+      "step": 6688
+    },
+    {
+      "epoch": 1.8529085872576179,
+      "grad_norm": 0.6899899840354919,
+      "learning_rate": 3.933576351295797e-06,
+      "loss": 0.5908,
+      "step": 6689
+    },
+    {
+      "epoch": 1.8531855955678669,
+      "grad_norm": 0.6647062301635742,
+      "learning_rate": 3.933277894180519e-06,
+      "loss": 0.6456,
+      "step": 6690
+    },
+    {
+      "epoch": 1.8534626038781163,
+      "grad_norm": 0.7126433253288269,
+      "learning_rate": 3.932979406633091e-06,
+      "loss": 0.6684,
+      "step": 6691
+    },
+    {
+      "epoch": 1.8537396121883658,
+      "grad_norm": 0.7348646521568298,
+      "learning_rate": 3.932680888659852e-06,
+      "loss": 0.6495,
+      "step": 6692
+    },
+    {
+      "epoch": 1.8540166204986148,
+      "grad_norm": 0.6930987238883972,
+      "learning_rate": 3.932382340267138e-06,
+      "loss": 0.6132,
+      "step": 6693
+    },
+    {
+      "epoch": 1.8542936288088643,
+      "grad_norm": 0.7253433465957642,
+      "learning_rate": 3.93208376146129e-06,
+      "loss": 0.644,
+      "step": 6694
+    },
+    {
+      "epoch": 1.8545706371191135,
+      "grad_norm": 0.679956316947937,
+      "learning_rate": 3.931785152248645e-06,
+      "loss": 0.6581,
+      "step": 6695
+    },
+    {
+      "epoch": 1.8548476454293628,
+      "grad_norm": 0.742157518863678,
+      "learning_rate": 3.931486512635546e-06,
+      "loss": 0.6561,
+      "step": 6696
+    },
+    {
+      "epoch": 1.8551246537396122,
+      "grad_norm": 0.7009865045547485,
+      "learning_rate": 3.931187842628332e-06,
+      "loss": 0.6122,
+      "step": 6697
+    },
+    {
+      "epoch": 1.8554016620498615,
+      "grad_norm": 0.704265832901001,
+      "learning_rate": 3.930889142233346e-06,
+      "loss": 0.6643,
+      "step": 6698
+    },
+    {
+      "epoch": 1.8556786703601107,
+      "grad_norm": 0.687735915184021,
+      "learning_rate": 3.9305904114569295e-06,
+      "loss": 0.6302,
+      "step": 6699
+    },
+    {
+      "epoch": 1.8559556786703602,
+      "grad_norm": 0.6906552910804749,
+      "learning_rate": 3.930291650305424e-06,
+      "loss": 0.6103,
+      "step": 6700
+    },
+    {
+      "epoch": 1.8562326869806094,
+      "grad_norm": 0.6972749829292297,
+      "learning_rate": 3.929992858785176e-06,
+      "loss": 0.624,
+      "step": 6701
+    },
+    {
+      "epoch": 1.8565096952908586,
+      "grad_norm": 0.74320387840271,
+      "learning_rate": 3.929694036902526e-06,
+      "loss": 0.691,
+      "step": 6702
+    },
+    {
+      "epoch": 1.856786703601108,
+      "grad_norm": 0.6657626032829285,
+      "learning_rate": 3.929395184663822e-06,
+      "loss": 0.616,
+      "step": 6703
+    },
+    {
+      "epoch": 1.8570637119113573,
+      "grad_norm": 0.681754469871521,
+      "learning_rate": 3.929096302075407e-06,
+      "loss": 0.6234,
+      "step": 6704
+    },
+    {
+      "epoch": 1.8573407202216066,
+      "grad_norm": 0.7159410119056702,
+      "learning_rate": 3.928797389143628e-06,
+      "loss": 0.6834,
+      "step": 6705
+    },
+    {
+      "epoch": 1.857617728531856,
+      "grad_norm": 0.6623358726501465,
+      "learning_rate": 3.9284984458748325e-06,
+      "loss": 0.5949,
+      "step": 6706
+    },
+    {
+      "epoch": 1.8578947368421053,
+      "grad_norm": 0.6965116858482361,
+      "learning_rate": 3.928199472275367e-06,
+      "loss": 0.6455,
+      "step": 6707
+    },
+    {
+      "epoch": 1.8581717451523545,
+      "grad_norm": 0.6975403428077698,
+      "learning_rate": 3.9279004683515785e-06,
+      "loss": 0.6341,
+      "step": 6708
+    },
+    {
+      "epoch": 1.858448753462604,
+      "grad_norm": 0.7415390014648438,
+      "learning_rate": 3.927601434109816e-06,
+      "loss": 0.6179,
+      "step": 6709
+    },
+    {
+      "epoch": 1.8587257617728532,
+      "grad_norm": 0.7149662971496582,
+      "learning_rate": 3.927302369556431e-06,
+      "loss": 0.6419,
+      "step": 6710
+    },
+    {
+      "epoch": 1.8590027700831024,
+      "grad_norm": 0.756695032119751,
+      "learning_rate": 3.927003274697772e-06,
+      "loss": 0.6325,
+      "step": 6711
+    },
+    {
+      "epoch": 1.859279778393352,
+      "grad_norm": 0.692173182964325,
+      "learning_rate": 3.926704149540188e-06,
+      "loss": 0.6093,
+      "step": 6712
+    },
+    {
+      "epoch": 1.8595567867036011,
+      "grad_norm": 0.7355810403823853,
+      "learning_rate": 3.9264049940900315e-06,
+      "loss": 0.6555,
+      "step": 6713
+    },
+    {
+      "epoch": 1.8598337950138504,
+      "grad_norm": 0.7171083688735962,
+      "learning_rate": 3.926105808353655e-06,
+      "loss": 0.6567,
+      "step": 6714
+    },
+    {
+      "epoch": 1.8601108033240998,
+      "grad_norm": 0.6942392587661743,
+      "learning_rate": 3.9258065923374104e-06,
+      "loss": 0.6598,
+      "step": 6715
+    },
+    {
+      "epoch": 1.860387811634349,
+      "grad_norm": 0.7160379886627197,
+      "learning_rate": 3.925507346047651e-06,
+      "loss": 0.6218,
+      "step": 6716
+    },
+    {
+      "epoch": 1.8606648199445983,
+      "grad_norm": 0.6698644757270813,
+      "learning_rate": 3.925208069490729e-06,
+      "loss": 0.6812,
+      "step": 6717
+    },
+    {
+      "epoch": 1.8609418282548478,
+      "grad_norm": 0.6921928524971008,
+      "learning_rate": 3.924908762673001e-06,
+      "loss": 0.6248,
+      "step": 6718
+    },
+    {
+      "epoch": 1.8612188365650968,
+      "grad_norm": 0.6881176233291626,
+      "learning_rate": 3.924609425600822e-06,
+      "loss": 0.5868,
+      "step": 6719
+    },
+    {
+      "epoch": 1.8614958448753463,
+      "grad_norm": 0.6716845035552979,
+      "learning_rate": 3.924310058280545e-06,
+      "loss": 0.6136,
+      "step": 6720
+    },
+    {
+      "epoch": 1.8617728531855957,
+      "grad_norm": 0.7087962031364441,
+      "learning_rate": 3.924010660718529e-06,
+      "loss": 0.6524,
+      "step": 6721
+    },
+    {
+      "epoch": 1.8620498614958447,
+      "grad_norm": 0.688025951385498,
+      "learning_rate": 3.92371123292113e-06,
+      "loss": 0.574,
+      "step": 6722
+    },
+    {
+      "epoch": 1.8623268698060942,
+      "grad_norm": 0.7003433108329773,
+      "learning_rate": 3.923411774894707e-06,
+      "loss": 0.6171,
+      "step": 6723
+    },
+    {
+      "epoch": 1.8626038781163436,
+      "grad_norm": 0.7128757238388062,
+      "learning_rate": 3.923112286645615e-06,
+      "loss": 0.6208,
+      "step": 6724
+    },
+    {
+      "epoch": 1.8628808864265927,
+      "grad_norm": 0.6693975925445557,
+      "learning_rate": 3.922812768180216e-06,
+      "loss": 0.5917,
+      "step": 6725
+    },
+    {
+      "epoch": 1.8631578947368421,
+      "grad_norm": 0.7504602074623108,
+      "learning_rate": 3.922513219504869e-06,
+      "loss": 0.6333,
+      "step": 6726
+    },
+    {
+      "epoch": 1.8634349030470914,
+      "grad_norm": 0.6776396632194519,
+      "learning_rate": 3.922213640625933e-06,
+      "loss": 0.6532,
+      "step": 6727
+    },
+    {
+      "epoch": 1.8637119113573406,
+      "grad_norm": 0.7138490080833435,
+      "learning_rate": 3.921914031549769e-06,
+      "loss": 0.6541,
+      "step": 6728
+    },
+    {
+      "epoch": 1.86398891966759,
+      "grad_norm": 0.6899004578590393,
+      "learning_rate": 3.92161439228274e-06,
+      "loss": 0.6395,
+      "step": 6729
+    },
+    {
+      "epoch": 1.8642659279778393,
+      "grad_norm": 0.6944865584373474,
+      "learning_rate": 3.921314722831205e-06,
+      "loss": 0.5886,
+      "step": 6730
+    },
+    {
+      "epoch": 1.8645429362880885,
+      "grad_norm": 0.7385538220405579,
+      "learning_rate": 3.9210150232015305e-06,
+      "loss": 0.6616,
+      "step": 6731
+    },
+    {
+      "epoch": 1.864819944598338,
+      "grad_norm": 0.7503032684326172,
+      "learning_rate": 3.920715293400079e-06,
+      "loss": 0.612,
+      "step": 6732
+    },
+    {
+      "epoch": 1.8650969529085872,
+      "grad_norm": 0.690791130065918,
+      "learning_rate": 3.920415533433212e-06,
+      "loss": 0.6164,
+      "step": 6733
+    },
+    {
+      "epoch": 1.8653739612188365,
+      "grad_norm": 0.7362712025642395,
+      "learning_rate": 3.920115743307297e-06,
+      "loss": 0.5862,
+      "step": 6734
+    },
+    {
+      "epoch": 1.865650969529086,
+      "grad_norm": 0.7012346982955933,
+      "learning_rate": 3.919815923028697e-06,
+      "loss": 0.6398,
+      "step": 6735
+    },
+    {
+      "epoch": 1.8659279778393352,
+      "grad_norm": 0.6703203916549683,
+      "learning_rate": 3.9195160726037805e-06,
+      "loss": 0.6499,
+      "step": 6736
+    },
+    {
+      "epoch": 1.8662049861495844,
+      "grad_norm": 0.7024924755096436,
+      "learning_rate": 3.919216192038912e-06,
+      "loss": 0.6364,
+      "step": 6737
+    },
+    {
+      "epoch": 1.8664819944598339,
+      "grad_norm": 0.6496801972389221,
+      "learning_rate": 3.918916281340459e-06,
+      "loss": 0.5743,
+      "step": 6738
+    },
+    {
+      "epoch": 1.866759002770083,
+      "grad_norm": 0.7026786804199219,
+      "learning_rate": 3.918616340514791e-06,
+      "loss": 0.6141,
+      "step": 6739
+    },
+    {
+      "epoch": 1.8670360110803323,
+      "grad_norm": 0.7047402262687683,
+      "learning_rate": 3.918316369568275e-06,
+      "loss": 0.6274,
+      "step": 6740
+    },
+    {
+      "epoch": 1.8673130193905818,
+      "grad_norm": 0.7136473655700684,
+      "learning_rate": 3.918016368507279e-06,
+      "loss": 0.6346,
+      "step": 6741
+    },
+    {
+      "epoch": 1.867590027700831,
+      "grad_norm": 0.6642279624938965,
+      "learning_rate": 3.917716337338176e-06,
+      "loss": 0.6038,
+      "step": 6742
+    },
+    {
+      "epoch": 1.8678670360110803,
+      "grad_norm": 0.6472945809364319,
+      "learning_rate": 3.917416276067334e-06,
+      "loss": 0.6186,
+      "step": 6743
+    },
+    {
+      "epoch": 1.8681440443213297,
+      "grad_norm": 0.7538015842437744,
+      "learning_rate": 3.917116184701125e-06,
+      "loss": 0.6696,
+      "step": 6744
+    },
+    {
+      "epoch": 1.868421052631579,
+      "grad_norm": 0.6811779141426086,
+      "learning_rate": 3.916816063245921e-06,
+      "loss": 0.5859,
+      "step": 6745
+    },
+    {
+      "epoch": 1.8686980609418282,
+      "grad_norm": 0.6205077767372131,
+      "learning_rate": 3.916515911708093e-06,
+      "loss": 0.4999,
+      "step": 6746
+    },
+    {
+      "epoch": 1.8689750692520777,
+      "grad_norm": 0.69630366563797,
+      "learning_rate": 3.916215730094015e-06,
+      "loss": 0.6043,
+      "step": 6747
+    },
+    {
+      "epoch": 1.869252077562327,
+      "grad_norm": 0.6904569268226624,
+      "learning_rate": 3.91591551841006e-06,
+      "loss": 0.6165,
+      "step": 6748
+    },
+    {
+      "epoch": 1.8695290858725762,
+      "grad_norm": 0.6998686790466309,
+      "learning_rate": 3.9156152766626036e-06,
+      "loss": 0.5871,
+      "step": 6749
+    },
+    {
+      "epoch": 1.8698060941828256,
+      "grad_norm": 0.6952171921730042,
+      "learning_rate": 3.915315004858019e-06,
+      "loss": 0.63,
+      "step": 6750
+    },
+    {
+      "epoch": 1.8700831024930746,
+      "grad_norm": 0.7425678372383118,
+      "learning_rate": 3.915014703002683e-06,
+      "loss": 0.634,
+      "step": 6751
+    },
+    {
+      "epoch": 1.870360110803324,
+      "grad_norm": 0.6801332831382751,
+      "learning_rate": 3.914714371102971e-06,
+      "loss": 0.6623,
+      "step": 6752
+    },
+    {
+      "epoch": 1.8706371191135736,
+      "grad_norm": 0.7020564079284668,
+      "learning_rate": 3.914414009165261e-06,
+      "loss": 0.5887,
+      "step": 6753
+    },
+    {
+      "epoch": 1.8709141274238226,
+      "grad_norm": 0.7230193018913269,
+      "learning_rate": 3.914113617195928e-06,
+      "loss": 0.5858,
+      "step": 6754
+    },
+    {
+      "epoch": 1.871191135734072,
+      "grad_norm": 0.6767334938049316,
+      "learning_rate": 3.9138131952013535e-06,
+      "loss": 0.6358,
+      "step": 6755
+    },
+    {
+      "epoch": 1.8714681440443215,
+      "grad_norm": 0.6952727437019348,
+      "learning_rate": 3.913512743187913e-06,
+      "loss": 0.618,
+      "step": 6756
+    },
+    {
+      "epoch": 1.8717451523545705,
+      "grad_norm": 0.7013217210769653,
+      "learning_rate": 3.913212261161988e-06,
+      "loss": 0.597,
+      "step": 6757
+    },
+    {
+      "epoch": 1.87202216066482,
+      "grad_norm": 0.6710230112075806,
+      "learning_rate": 3.912911749129957e-06,
+      "loss": 0.6258,
+      "step": 6758
+    },
+    {
+      "epoch": 1.8722991689750692,
+      "grad_norm": 0.6895468235015869,
+      "learning_rate": 3.912611207098202e-06,
+      "loss": 0.6536,
+      "step": 6759
+    },
+    {
+      "epoch": 1.8725761772853184,
+      "grad_norm": 0.7238779067993164,
+      "learning_rate": 3.9123106350731035e-06,
+      "loss": 0.6386,
+      "step": 6760
+    },
+    {
+      "epoch": 1.872853185595568,
+      "grad_norm": 0.705436110496521,
+      "learning_rate": 3.912010033061043e-06,
+      "loss": 0.6253,
+      "step": 6761
+    },
+    {
+      "epoch": 1.8731301939058171,
+      "grad_norm": 0.7319697737693787,
+      "learning_rate": 3.9117094010684046e-06,
+      "loss": 0.6386,
+      "step": 6762
+    },
+    {
+      "epoch": 1.8734072022160664,
+      "grad_norm": 0.6451613903045654,
+      "learning_rate": 3.911408739101569e-06,
+      "loss": 0.622,
+      "step": 6763
+    },
+    {
+      "epoch": 1.8736842105263158,
+      "grad_norm": 0.7184947729110718,
+      "learning_rate": 3.911108047166924e-06,
+      "loss": 0.6283,
+      "step": 6764
+    },
+    {
+      "epoch": 1.873961218836565,
+      "grad_norm": 0.7210145592689514,
+      "learning_rate": 3.91080732527085e-06,
+      "loss": 0.589,
+      "step": 6765
+    },
+    {
+      "epoch": 1.8742382271468143,
+      "grad_norm": 0.6852824091911316,
+      "learning_rate": 3.910506573419734e-06,
+      "loss": 0.6923,
+      "step": 6766
+    },
+    {
+      "epoch": 1.8745152354570638,
+      "grad_norm": 0.6745520234107971,
+      "learning_rate": 3.91020579161996e-06,
+      "loss": 0.6075,
+      "step": 6767
+    },
+    {
+      "epoch": 1.874792243767313,
+      "grad_norm": 0.6957927346229553,
+      "learning_rate": 3.9099049798779175e-06,
+      "loss": 0.6381,
+      "step": 6768
+    },
+    {
+      "epoch": 1.8750692520775623,
+      "grad_norm": 0.7385134100914001,
+      "learning_rate": 3.9096041381999915e-06,
+      "loss": 0.6347,
+      "step": 6769
+    },
+    {
+      "epoch": 1.8753462603878117,
+      "grad_norm": 0.6691296696662903,
+      "learning_rate": 3.90930326659257e-06,
+      "loss": 0.6456,
+      "step": 6770
+    },
+    {
+      "epoch": 1.875623268698061,
+      "grad_norm": 0.7090333700180054,
+      "learning_rate": 3.90900236506204e-06,
+      "loss": 0.6446,
+      "step": 6771
+    },
+    {
+      "epoch": 1.8759002770083102,
+      "grad_norm": 0.6893540024757385,
+      "learning_rate": 3.908701433614793e-06,
+      "loss": 0.5945,
+      "step": 6772
+    },
+    {
+      "epoch": 1.8761772853185597,
+      "grad_norm": 0.7020983695983887,
+      "learning_rate": 3.908400472257216e-06,
+      "loss": 0.565,
+      "step": 6773
+    },
+    {
+      "epoch": 1.876454293628809,
+      "grad_norm": 0.705759584903717,
+      "learning_rate": 3.908099480995701e-06,
+      "loss": 0.6329,
+      "step": 6774
+    },
+    {
+      "epoch": 1.8767313019390581,
+      "grad_norm": 0.7019346952438354,
+      "learning_rate": 3.907798459836638e-06,
+      "loss": 0.6426,
+      "step": 6775
+    },
+    {
+      "epoch": 1.8770083102493076,
+      "grad_norm": 0.8493272066116333,
+      "learning_rate": 3.907497408786418e-06,
+      "loss": 0.6011,
+      "step": 6776
+    },
+    {
+      "epoch": 1.8772853185595568,
+      "grad_norm": 0.6631448864936829,
+      "learning_rate": 3.9071963278514345e-06,
+      "loss": 0.6352,
+      "step": 6777
+    },
+    {
+      "epoch": 1.877562326869806,
+      "grad_norm": 0.7411801218986511,
+      "learning_rate": 3.906895217038079e-06,
+      "loss": 0.6191,
+      "step": 6778
+    },
+    {
+      "epoch": 1.8778393351800555,
+      "grad_norm": 0.7120534181594849,
+      "learning_rate": 3.906594076352746e-06,
+      "loss": 0.6346,
+      "step": 6779
+    },
+    {
+      "epoch": 1.8781163434903048,
+      "grad_norm": 0.6764026284217834,
+      "learning_rate": 3.906292905801828e-06,
+      "loss": 0.5605,
+      "step": 6780
+    },
+    {
+      "epoch": 1.878393351800554,
+      "grad_norm": 0.7547695636749268,
+      "learning_rate": 3.90599170539172e-06,
+      "loss": 0.6406,
+      "step": 6781
+    },
+    {
+      "epoch": 1.8786703601108035,
+      "grad_norm": 0.688042938709259,
+      "learning_rate": 3.9056904751288185e-06,
+      "loss": 0.6302,
+      "step": 6782
+    },
+    {
+      "epoch": 1.8789473684210525,
+      "grad_norm": 0.7339063882827759,
+      "learning_rate": 3.905389215019518e-06,
+      "loss": 0.6323,
+      "step": 6783
+    },
+    {
+      "epoch": 1.879224376731302,
+      "grad_norm": 0.7102497816085815,
+      "learning_rate": 3.905087925070214e-06,
+      "loss": 0.5828,
+      "step": 6784
+    },
+    {
+      "epoch": 1.8795013850415514,
+      "grad_norm": 0.7145693898200989,
+      "learning_rate": 3.9047866052873075e-06,
+      "loss": 0.6135,
+      "step": 6785
+    },
+    {
+      "epoch": 1.8797783933518004,
+      "grad_norm": 0.7017257213592529,
+      "learning_rate": 3.904485255677193e-06,
+      "loss": 0.6143,
+      "step": 6786
+    },
+    {
+      "epoch": 1.8800554016620499,
+      "grad_norm": 0.6663693785667419,
+      "learning_rate": 3.90418387624627e-06,
+      "loss": 0.6185,
+      "step": 6787
+    },
+    {
+      "epoch": 1.8803324099722993,
+      "grad_norm": 0.7595977783203125,
+      "learning_rate": 3.903882467000938e-06,
+      "loss": 0.6048,
+      "step": 6788
+    },
+    {
+      "epoch": 1.8806094182825484,
+      "grad_norm": 0.7022473812103271,
+      "learning_rate": 3.9035810279475945e-06,
+      "loss": 0.6308,
+      "step": 6789
+    },
+    {
+      "epoch": 1.8808864265927978,
+      "grad_norm": 0.7909603714942932,
+      "learning_rate": 3.903279559092642e-06,
+      "loss": 0.6012,
+      "step": 6790
+    },
+    {
+      "epoch": 1.881163434903047,
+      "grad_norm": 0.6895849704742432,
+      "learning_rate": 3.902978060442482e-06,
+      "loss": 0.634,
+      "step": 6791
+    },
+    {
+      "epoch": 1.8814404432132963,
+      "grad_norm": 0.7165695428848267,
+      "learning_rate": 3.902676532003513e-06,
+      "loss": 0.6076,
+      "step": 6792
+    },
+    {
+      "epoch": 1.8817174515235457,
+      "grad_norm": 0.7227752804756165,
+      "learning_rate": 3.902374973782141e-06,
+      "loss": 0.6158,
+      "step": 6793
+    },
+    {
+      "epoch": 1.881994459833795,
+      "grad_norm": 0.716238260269165,
+      "learning_rate": 3.9020733857847665e-06,
+      "loss": 0.6021,
+      "step": 6794
+    },
+    {
+      "epoch": 1.8822714681440442,
+      "grad_norm": 0.7465752959251404,
+      "learning_rate": 3.901771768017793e-06,
+      "loss": 0.6563,
+      "step": 6795
+    },
+    {
+      "epoch": 1.8825484764542937,
+      "grad_norm": 0.7316117882728577,
+      "learning_rate": 3.901470120487626e-06,
+      "loss": 0.6235,
+      "step": 6796
+    },
+    {
+      "epoch": 1.882825484764543,
+      "grad_norm": 0.7191451191902161,
+      "learning_rate": 3.901168443200669e-06,
+      "loss": 0.668,
+      "step": 6797
+    },
+    {
+      "epoch": 1.8831024930747922,
+      "grad_norm": 0.7118086218833923,
+      "learning_rate": 3.900866736163326e-06,
+      "loss": 0.5964,
+      "step": 6798
+    },
+    {
+      "epoch": 1.8833795013850416,
+      "grad_norm": 0.6637589931488037,
+      "learning_rate": 3.900564999382007e-06,
+      "loss": 0.6217,
+      "step": 6799
+    },
+    {
+      "epoch": 1.8836565096952909,
+      "grad_norm": 0.7318509221076965,
+      "learning_rate": 3.900263232863117e-06,
+      "loss": 0.5931,
+      "step": 6800
+    },
+    {
+      "epoch": 1.88393351800554,
+      "grad_norm": 0.6680594086647034,
+      "learning_rate": 3.899961436613061e-06,
+      "loss": 0.631,
+      "step": 6801
+    },
+    {
+      "epoch": 1.8842105263157896,
+      "grad_norm": 0.6916776299476624,
+      "learning_rate": 3.8996596106382486e-06,
+      "loss": 0.6443,
+      "step": 6802
+    },
+    {
+      "epoch": 1.8844875346260388,
+      "grad_norm": 0.6689826846122742,
+      "learning_rate": 3.899357754945089e-06,
+      "loss": 0.6351,
+      "step": 6803
+    },
+    {
+      "epoch": 1.884764542936288,
+      "grad_norm": 0.6716510057449341,
+      "learning_rate": 3.899055869539991e-06,
+      "loss": 0.5975,
+      "step": 6804
+    },
+    {
+      "epoch": 1.8850415512465375,
+      "grad_norm": 0.7200577259063721,
+      "learning_rate": 3.898753954429363e-06,
+      "loss": 0.6257,
+      "step": 6805
+    },
+    {
+      "epoch": 1.8853185595567867,
+      "grad_norm": 0.7464181780815125,
+      "learning_rate": 3.898452009619616e-06,
+      "loss": 0.6395,
+      "step": 6806
+    },
+    {
+      "epoch": 1.885595567867036,
+      "grad_norm": 0.6617922186851501,
+      "learning_rate": 3.8981500351171645e-06,
+      "loss": 0.5896,
+      "step": 6807
+    },
+    {
+      "epoch": 1.8858725761772854,
+      "grad_norm": 0.7187881469726562,
+      "learning_rate": 3.897848030928415e-06,
+      "loss": 0.6358,
+      "step": 6808
+    },
+    {
+      "epoch": 1.8861495844875347,
+      "grad_norm": 0.6788371205329895,
+      "learning_rate": 3.897545997059782e-06,
+      "loss": 0.5784,
+      "step": 6809
+    },
+    {
+      "epoch": 1.886426592797784,
+      "grad_norm": 0.7442015409469604,
+      "learning_rate": 3.897243933517679e-06,
+      "loss": 0.663,
+      "step": 6810
+    },
+    {
+      "epoch": 1.8867036011080334,
+      "grad_norm": 0.6612447500228882,
+      "learning_rate": 3.89694184030852e-06,
+      "loss": 0.6382,
+      "step": 6811
+    },
+    {
+      "epoch": 1.8869806094182824,
+      "grad_norm": 0.6605403423309326,
+      "learning_rate": 3.896639717438717e-06,
+      "loss": 0.5846,
+      "step": 6812
+    },
+    {
+      "epoch": 1.8872576177285318,
+      "grad_norm": 0.6534563302993774,
+      "learning_rate": 3.896337564914687e-06,
+      "loss": 0.5978,
+      "step": 6813
+    },
+    {
+      "epoch": 1.8875346260387813,
+      "grad_norm": 0.669242262840271,
+      "learning_rate": 3.896035382742844e-06,
+      "loss": 0.6266,
+      "step": 6814
+    },
+    {
+      "epoch": 1.8878116343490303,
+      "grad_norm": 0.6762076616287231,
+      "learning_rate": 3.895733170929606e-06,
+      "loss": 0.628,
+      "step": 6815
+    },
+    {
+      "epoch": 1.8880886426592798,
+      "grad_norm": 0.6946333646774292,
+      "learning_rate": 3.8954309294813865e-06,
+      "loss": 0.6718,
+      "step": 6816
+    },
+    {
+      "epoch": 1.8883656509695292,
+      "grad_norm": 0.6655415296554565,
+      "learning_rate": 3.895128658404605e-06,
+      "loss": 0.5833,
+      "step": 6817
+    },
+    {
+      "epoch": 1.8886426592797783,
+      "grad_norm": 0.6595086455345154,
+      "learning_rate": 3.894826357705681e-06,
+      "loss": 0.5797,
+      "step": 6818
+    },
+    {
+      "epoch": 1.8889196675900277,
+      "grad_norm": 0.6730854511260986,
+      "learning_rate": 3.894524027391031e-06,
+      "loss": 0.6075,
+      "step": 6819
+    },
+    {
+      "epoch": 1.889196675900277,
+      "grad_norm": 0.6942018866539001,
+      "learning_rate": 3.894221667467074e-06,
+      "loss": 0.627,
+      "step": 6820
+    },
+    {
+      "epoch": 1.8894736842105262,
+      "grad_norm": 0.7450907230377197,
+      "learning_rate": 3.893919277940231e-06,
+      "loss": 0.5793,
+      "step": 6821
+    },
+    {
+      "epoch": 1.8897506925207757,
+      "grad_norm": 0.6789368987083435,
+      "learning_rate": 3.893616858816921e-06,
+      "loss": 0.5696,
+      "step": 6822
+    },
+    {
+      "epoch": 1.890027700831025,
+      "grad_norm": 0.6660521626472473,
+      "learning_rate": 3.893314410103568e-06,
+      "loss": 0.5933,
+      "step": 6823
+    },
+    {
+      "epoch": 1.8903047091412741,
+      "grad_norm": 0.7321332693099976,
+      "learning_rate": 3.8930119318065905e-06,
+      "loss": 0.6425,
+      "step": 6824
+    },
+    {
+      "epoch": 1.8905817174515236,
+      "grad_norm": 0.7297572493553162,
+      "learning_rate": 3.892709423932412e-06,
+      "loss": 0.6211,
+      "step": 6825
+    },
+    {
+      "epoch": 1.8908587257617728,
+      "grad_norm": 0.6849077939987183,
+      "learning_rate": 3.892406886487457e-06,
+      "loss": 0.6179,
+      "step": 6826
+    },
+    {
+      "epoch": 1.891135734072022,
+      "grad_norm": 0.6857331395149231,
+      "learning_rate": 3.892104319478148e-06,
+      "loss": 0.6557,
+      "step": 6827
+    },
+    {
+      "epoch": 1.8914127423822715,
+      "grad_norm": 0.7214446663856506,
+      "learning_rate": 3.891801722910909e-06,
+      "loss": 0.6169,
+      "step": 6828
+    },
+    {
+      "epoch": 1.8916897506925208,
+      "grad_norm": 0.7125465869903564,
+      "learning_rate": 3.891499096792165e-06,
+      "loss": 0.5925,
+      "step": 6829
+    },
+    {
+      "epoch": 1.89196675900277,
+      "grad_norm": 0.7323399782180786,
+      "learning_rate": 3.891196441128342e-06,
+      "loss": 0.6084,
+      "step": 6830
+    },
+    {
+      "epoch": 1.8922437673130195,
+      "grad_norm": 0.7022059559822083,
+      "learning_rate": 3.890893755925865e-06,
+      "loss": 0.6626,
+      "step": 6831
+    },
+    {
+      "epoch": 1.8925207756232687,
+      "grad_norm": 0.7302187085151672,
+      "learning_rate": 3.890591041191162e-06,
+      "loss": 0.614,
+      "step": 6832
+    },
+    {
+      "epoch": 1.892797783933518,
+      "grad_norm": 0.7071456909179688,
+      "learning_rate": 3.89028829693066e-06,
+      "loss": 0.6617,
+      "step": 6833
+    },
+    {
+      "epoch": 1.8930747922437674,
+      "grad_norm": 0.7259766459465027,
+      "learning_rate": 3.889985523150788e-06,
+      "loss": 0.633,
+      "step": 6834
+    },
+    {
+      "epoch": 1.8933518005540166,
+      "grad_norm": 0.7103668451309204,
+      "learning_rate": 3.889682719857972e-06,
+      "loss": 0.641,
+      "step": 6835
+    },
+    {
+      "epoch": 1.8936288088642659,
+      "grad_norm": 0.7698994278907776,
+      "learning_rate": 3.889379887058644e-06,
+      "loss": 0.6379,
+      "step": 6836
+    },
+    {
+      "epoch": 1.8939058171745153,
+      "grad_norm": 0.7132614850997925,
+      "learning_rate": 3.889077024759233e-06,
+      "loss": 0.6053,
+      "step": 6837
+    },
+    {
+      "epoch": 1.8941828254847646,
+      "grad_norm": 0.7450546622276306,
+      "learning_rate": 3.8887741329661695e-06,
+      "loss": 0.6214,
+      "step": 6838
+    },
+    {
+      "epoch": 1.8944598337950138,
+      "grad_norm": 0.6779670715332031,
+      "learning_rate": 3.888471211685885e-06,
+      "loss": 0.6601,
+      "step": 6839
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.6860229969024658,
+      "learning_rate": 3.888168260924809e-06,
+      "loss": 0.6084,
+      "step": 6840
+    },
+    {
+      "epoch": 1.8950138504155125,
+      "grad_norm": 0.7304345369338989,
+      "learning_rate": 3.887865280689378e-06,
+      "loss": 0.619,
+      "step": 6841
+    },
+    {
+      "epoch": 1.8952908587257618,
+      "grad_norm": 0.6954903602600098,
+      "learning_rate": 3.887562270986021e-06,
+      "loss": 0.6084,
+      "step": 6842
+    },
+    {
+      "epoch": 1.8955678670360112,
+      "grad_norm": 0.6936541795730591,
+      "learning_rate": 3.887259231821176e-06,
+      "loss": 0.6134,
+      "step": 6843
+    },
+    {
+      "epoch": 1.8958448753462602,
+      "grad_norm": 0.7285112142562866,
+      "learning_rate": 3.886956163201272e-06,
+      "loss": 0.6549,
+      "step": 6844
+    },
+    {
+      "epoch": 1.8961218836565097,
+      "grad_norm": 0.7493306398391724,
+      "learning_rate": 3.886653065132749e-06,
+      "loss": 0.6692,
+      "step": 6845
+    },
+    {
+      "epoch": 1.8963988919667591,
+      "grad_norm": 0.7186040878295898,
+      "learning_rate": 3.886349937622039e-06,
+      "loss": 0.6429,
+      "step": 6846
+    },
+    {
+      "epoch": 1.8966759002770082,
+      "grad_norm": 0.6854330897331238,
+      "learning_rate": 3.8860467806755795e-06,
+      "loss": 0.6766,
+      "step": 6847
+    },
+    {
+      "epoch": 1.8969529085872576,
+      "grad_norm": 0.684589684009552,
+      "learning_rate": 3.885743594299808e-06,
+      "loss": 0.6185,
+      "step": 6848
+    },
+    {
+      "epoch": 1.897229916897507,
+      "grad_norm": 0.6987202167510986,
+      "learning_rate": 3.88544037850116e-06,
+      "loss": 0.5872,
+      "step": 6849
+    },
+    {
+      "epoch": 1.897506925207756,
+      "grad_norm": 0.6784823536872864,
+      "learning_rate": 3.885137133286076e-06,
+      "loss": 0.6039,
+      "step": 6850
+    },
+    {
+      "epoch": 1.8977839335180056,
+      "grad_norm": 0.6691726446151733,
+      "learning_rate": 3.8848338586609934e-06,
+      "loss": 0.6003,
+      "step": 6851
+    },
+    {
+      "epoch": 1.8980609418282548,
+      "grad_norm": 0.6847113966941833,
+      "learning_rate": 3.884530554632351e-06,
+      "loss": 0.6125,
+      "step": 6852
+    },
+    {
+      "epoch": 1.898337950138504,
+      "grad_norm": 0.667030394077301,
+      "learning_rate": 3.884227221206589e-06,
+      "loss": 0.5736,
+      "step": 6853
+    },
+    {
+      "epoch": 1.8986149584487535,
+      "grad_norm": 0.7349780201911926,
+      "learning_rate": 3.883923858390149e-06,
+      "loss": 0.6526,
+      "step": 6854
+    },
+    {
+      "epoch": 1.8988919667590027,
+      "grad_norm": 0.6886972188949585,
+      "learning_rate": 3.88362046618947e-06,
+      "loss": 0.6634,
+      "step": 6855
+    },
+    {
+      "epoch": 1.899168975069252,
+      "grad_norm": 0.7415412664413452,
+      "learning_rate": 3.883317044610996e-06,
+      "loss": 0.6238,
+      "step": 6856
+    },
+    {
+      "epoch": 1.8994459833795014,
+      "grad_norm": 0.7218773365020752,
+      "learning_rate": 3.883013593661168e-06,
+      "loss": 0.6392,
+      "step": 6857
+    },
+    {
+      "epoch": 1.8997229916897507,
+      "grad_norm": 0.741144061088562,
+      "learning_rate": 3.88271011334643e-06,
+      "loss": 0.6259,
+      "step": 6858
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7077538371086121,
+      "learning_rate": 3.882406603673226e-06,
+      "loss": 0.6401,
+      "step": 6859
+    },
+    {
+      "epoch": 1.9002770083102494,
+      "grad_norm": 0.6991947293281555,
+      "learning_rate": 3.882103064647999e-06,
+      "loss": 0.61,
+      "step": 6860
+    },
+    {
+      "epoch": 1.9005540166204986,
+      "grad_norm": 0.7192123532295227,
+      "learning_rate": 3.881799496277194e-06,
+      "loss": 0.6193,
+      "step": 6861
+    },
+    {
+      "epoch": 1.9008310249307478,
+      "grad_norm": 0.7007842659950256,
+      "learning_rate": 3.8814958985672564e-06,
+      "loss": 0.6135,
+      "step": 6862
+    },
+    {
+      "epoch": 1.9011080332409973,
+      "grad_norm": 0.7086727619171143,
+      "learning_rate": 3.881192271524634e-06,
+      "loss": 0.6553,
+      "step": 6863
+    },
+    {
+      "epoch": 1.9013850415512465,
+      "grad_norm": 0.7087380290031433,
+      "learning_rate": 3.880888615155772e-06,
+      "loss": 0.6604,
+      "step": 6864
+    },
+    {
+      "epoch": 1.9016620498614958,
+      "grad_norm": 0.693657398223877,
+      "learning_rate": 3.880584929467119e-06,
+      "loss": 0.641,
+      "step": 6865
+    },
+    {
+      "epoch": 1.9019390581717452,
+      "grad_norm": 0.6947059035301208,
+      "learning_rate": 3.880281214465122e-06,
+      "loss": 0.6528,
+      "step": 6866
+    },
+    {
+      "epoch": 1.9022160664819945,
+      "grad_norm": 0.7018727660179138,
+      "learning_rate": 3.879977470156229e-06,
+      "loss": 0.6823,
+      "step": 6867
+    },
+    {
+      "epoch": 1.9024930747922437,
+      "grad_norm": 0.7383748888969421,
+      "learning_rate": 3.879673696546891e-06,
+      "loss": 0.6329,
+      "step": 6868
+    },
+    {
+      "epoch": 1.9027700831024932,
+      "grad_norm": 0.7053066492080688,
+      "learning_rate": 3.8793698936435574e-06,
+      "loss": 0.5968,
+      "step": 6869
+    },
+    {
+      "epoch": 1.9030470914127424,
+      "grad_norm": 0.7110251784324646,
+      "learning_rate": 3.8790660614526776e-06,
+      "loss": 0.6363,
+      "step": 6870
+    },
+    {
+      "epoch": 1.9033240997229917,
+      "grad_norm": 0.6861403584480286,
+      "learning_rate": 3.878762199980705e-06,
+      "loss": 0.6327,
+      "step": 6871
+    },
+    {
+      "epoch": 1.9036011080332411,
+      "grad_norm": 0.7011222243309021,
+      "learning_rate": 3.878458309234089e-06,
+      "loss": 0.608,
+      "step": 6872
+    },
+    {
+      "epoch": 1.9038781163434904,
+      "grad_norm": 0.7192626595497131,
+      "learning_rate": 3.8781543892192825e-06,
+      "loss": 0.6199,
+      "step": 6873
+    },
+    {
+      "epoch": 1.9041551246537396,
+      "grad_norm": 0.6861461400985718,
+      "learning_rate": 3.877850439942739e-06,
+      "loss": 0.601,
+      "step": 6874
+    },
+    {
+      "epoch": 1.904432132963989,
+      "grad_norm": 0.6651129722595215,
+      "learning_rate": 3.877546461410912e-06,
+      "loss": 0.6312,
+      "step": 6875
+    },
+    {
+      "epoch": 1.904709141274238,
+      "grad_norm": 0.6886206269264221,
+      "learning_rate": 3.8772424536302565e-06,
+      "loss": 0.6376,
+      "step": 6876
+    },
+    {
+      "epoch": 1.9049861495844875,
+      "grad_norm": 0.7215523719787598,
+      "learning_rate": 3.876938416607227e-06,
+      "loss": 0.6144,
+      "step": 6877
+    },
+    {
+      "epoch": 1.905263157894737,
+      "grad_norm": 0.7637720108032227,
+      "learning_rate": 3.876634350348278e-06,
+      "loss": 0.6296,
+      "step": 6878
+    },
+    {
+      "epoch": 1.905540166204986,
+      "grad_norm": 0.7078676819801331,
+      "learning_rate": 3.876330254859865e-06,
+      "loss": 0.6593,
+      "step": 6879
+    },
+    {
+      "epoch": 1.9058171745152355,
+      "grad_norm": 0.6875743865966797,
+      "learning_rate": 3.876026130148447e-06,
+      "loss": 0.5992,
+      "step": 6880
+    },
+    {
+      "epoch": 1.906094182825485,
+      "grad_norm": 0.6959920525550842,
+      "learning_rate": 3.87572197622048e-06,
+      "loss": 0.6229,
+      "step": 6881
+    },
+    {
+      "epoch": 1.906371191135734,
+      "grad_norm": 0.7063423991203308,
+      "learning_rate": 3.875417793082423e-06,
+      "loss": 0.5705,
+      "step": 6882
+    },
+    {
+      "epoch": 1.9066481994459834,
+      "grad_norm": 0.6862561702728271,
+      "learning_rate": 3.875113580740732e-06,
+      "loss": 0.5985,
+      "step": 6883
+    },
+    {
+      "epoch": 1.9069252077562326,
+      "grad_norm": 0.6900488138198853,
+      "learning_rate": 3.874809339201871e-06,
+      "loss": 0.6148,
+      "step": 6884
+    },
+    {
+      "epoch": 1.9072022160664819,
+      "grad_norm": 0.7383355498313904,
+      "learning_rate": 3.874505068472294e-06,
+      "loss": 0.5925,
+      "step": 6885
+    },
+    {
+      "epoch": 1.9074792243767313,
+      "grad_norm": 0.7217558026313782,
+      "learning_rate": 3.874200768558466e-06,
+      "loss": 0.6417,
+      "step": 6886
+    },
+    {
+      "epoch": 1.9077562326869806,
+      "grad_norm": 0.6995415687561035,
+      "learning_rate": 3.873896439466845e-06,
+      "loss": 0.6171,
+      "step": 6887
+    },
+    {
+      "epoch": 1.9080332409972298,
+      "grad_norm": 0.6916835904121399,
+      "learning_rate": 3.873592081203895e-06,
+      "loss": 0.6121,
+      "step": 6888
+    },
+    {
+      "epoch": 1.9083102493074793,
+      "grad_norm": 0.6758975386619568,
+      "learning_rate": 3.873287693776077e-06,
+      "loss": 0.5845,
+      "step": 6889
+    },
+    {
+      "epoch": 1.9085872576177285,
+      "grad_norm": 0.6973399519920349,
+      "learning_rate": 3.872983277189854e-06,
+      "loss": 0.652,
+      "step": 6890
+    },
+    {
+      "epoch": 1.9088642659279778,
+      "grad_norm": 0.6787953972816467,
+      "learning_rate": 3.8726788314516915e-06,
+      "loss": 0.6738,
+      "step": 6891
+    },
+    {
+      "epoch": 1.9091412742382272,
+      "grad_norm": 0.6787337064743042,
+      "learning_rate": 3.87237435656805e-06,
+      "loss": 0.6276,
+      "step": 6892
+    },
+    {
+      "epoch": 1.9094182825484765,
+      "grad_norm": 0.6608779430389404,
+      "learning_rate": 3.872069852545397e-06,
+      "loss": 0.5507,
+      "step": 6893
+    },
+    {
+      "epoch": 1.9096952908587257,
+      "grad_norm": 0.7295418381690979,
+      "learning_rate": 3.871765319390197e-06,
+      "loss": 0.618,
+      "step": 6894
+    },
+    {
+      "epoch": 1.9099722991689752,
+      "grad_norm": 0.6533291935920715,
+      "learning_rate": 3.8714607571089165e-06,
+      "loss": 0.6097,
+      "step": 6895
+    },
+    {
+      "epoch": 1.9102493074792244,
+      "grad_norm": 0.6750076413154602,
+      "learning_rate": 3.871156165708021e-06,
+      "loss": 0.6211,
+      "step": 6896
+    },
+    {
+      "epoch": 1.9105263157894736,
+      "grad_norm": 0.7045161724090576,
+      "learning_rate": 3.870851545193979e-06,
+      "loss": 0.6372,
+      "step": 6897
+    },
+    {
+      "epoch": 1.910803324099723,
+      "grad_norm": 0.7043505907058716,
+      "learning_rate": 3.870546895573258e-06,
+      "loss": 0.6611,
+      "step": 6898
+    },
+    {
+      "epoch": 1.9110803324099723,
+      "grad_norm": 0.7087211012840271,
+      "learning_rate": 3.8702422168523265e-06,
+      "loss": 0.6516,
+      "step": 6899
+    },
+    {
+      "epoch": 1.9113573407202216,
+      "grad_norm": 0.6823852062225342,
+      "learning_rate": 3.869937509037654e-06,
+      "loss": 0.6354,
+      "step": 6900
+    },
+    {
+      "epoch": 1.911634349030471,
+      "grad_norm": 0.7137317657470703,
+      "learning_rate": 3.869632772135708e-06,
+      "loss": 0.6512,
+      "step": 6901
+    },
+    {
+      "epoch": 1.9119113573407203,
+      "grad_norm": 0.6764719486236572,
+      "learning_rate": 3.869328006152963e-06,
+      "loss": 0.6574,
+      "step": 6902
+    },
+    {
+      "epoch": 1.9121883656509695,
+      "grad_norm": 0.7081059217453003,
+      "learning_rate": 3.869023211095886e-06,
+      "loss": 0.612,
+      "step": 6903
+    },
+    {
+      "epoch": 1.912465373961219,
+      "grad_norm": 0.7032474279403687,
+      "learning_rate": 3.868718386970951e-06,
+      "loss": 0.6544,
+      "step": 6904
+    },
+    {
+      "epoch": 1.912742382271468,
+      "grad_norm": 0.7301380038261414,
+      "learning_rate": 3.868413533784629e-06,
+      "loss": 0.6419,
+      "step": 6905
+    },
+    {
+      "epoch": 1.9130193905817174,
+      "grad_norm": 0.7018072009086609,
+      "learning_rate": 3.8681086515433916e-06,
+      "loss": 0.6219,
+      "step": 6906
+    },
+    {
+      "epoch": 1.913296398891967,
+      "grad_norm": 0.7180678248405457,
+      "learning_rate": 3.867803740253715e-06,
+      "loss": 0.6432,
+      "step": 6907
+    },
+    {
+      "epoch": 1.913573407202216,
+      "grad_norm": 0.7180802226066589,
+      "learning_rate": 3.867498799922073e-06,
+      "loss": 0.646,
+      "step": 6908
+    },
+    {
+      "epoch": 1.9138504155124654,
+      "grad_norm": 0.654554009437561,
+      "learning_rate": 3.867193830554938e-06,
+      "loss": 0.612,
+      "step": 6909
+    },
+    {
+      "epoch": 1.9141274238227148,
+      "grad_norm": 0.6916167736053467,
+      "learning_rate": 3.866888832158787e-06,
+      "loss": 0.6228,
+      "step": 6910
+    },
+    {
+      "epoch": 1.9144044321329639,
+      "grad_norm": 0.7020214796066284,
+      "learning_rate": 3.866583804740095e-06,
+      "loss": 0.6184,
+      "step": 6911
+    },
+    {
+      "epoch": 1.9146814404432133,
+      "grad_norm": 0.7106627821922302,
+      "learning_rate": 3.866278748305339e-06,
+      "loss": 0.6465,
+      "step": 6912
+    },
+    {
+      "epoch": 1.9149584487534625,
+      "grad_norm": 0.6954221725463867,
+      "learning_rate": 3.865973662860997e-06,
+      "loss": 0.6271,
+      "step": 6913
+    },
+    {
+      "epoch": 1.9152354570637118,
+      "grad_norm": 0.6910799741744995,
+      "learning_rate": 3.865668548413545e-06,
+      "loss": 0.5818,
+      "step": 6914
+    },
+    {
+      "epoch": 1.9155124653739612,
+      "grad_norm": 0.6691675782203674,
+      "learning_rate": 3.865363404969462e-06,
+      "loss": 0.5956,
+      "step": 6915
+    },
+    {
+      "epoch": 1.9157894736842105,
+      "grad_norm": 0.7135936617851257,
+      "learning_rate": 3.865058232535228e-06,
+      "loss": 0.633,
+      "step": 6916
+    },
+    {
+      "epoch": 1.9160664819944597,
+      "grad_norm": 0.7141229510307312,
+      "learning_rate": 3.864753031117321e-06,
+      "loss": 0.6207,
+      "step": 6917
+    },
+    {
+      "epoch": 1.9163434903047092,
+      "grad_norm": 0.7615339756011963,
+      "learning_rate": 3.864447800722222e-06,
+      "loss": 0.6407,
+      "step": 6918
+    },
+    {
+      "epoch": 1.9166204986149584,
+      "grad_norm": 0.6961625814437866,
+      "learning_rate": 3.864142541356411e-06,
+      "loss": 0.618,
+      "step": 6919
+    },
+    {
+      "epoch": 1.9168975069252077,
+      "grad_norm": 0.6443533301353455,
+      "learning_rate": 3.863837253026372e-06,
+      "loss": 0.547,
+      "step": 6920
+    },
+    {
+      "epoch": 1.9171745152354571,
+      "grad_norm": 0.700282096862793,
+      "learning_rate": 3.863531935738585e-06,
+      "loss": 0.678,
+      "step": 6921
+    },
+    {
+      "epoch": 1.9174515235457064,
+      "grad_norm": 0.7155157923698425,
+      "learning_rate": 3.863226589499531e-06,
+      "loss": 0.6291,
+      "step": 6922
+    },
+    {
+      "epoch": 1.9177285318559556,
+      "grad_norm": 0.7124128341674805,
+      "learning_rate": 3.862921214315698e-06,
+      "loss": 0.6286,
+      "step": 6923
+    },
+    {
+      "epoch": 1.918005540166205,
+      "grad_norm": 0.7020770907402039,
+      "learning_rate": 3.862615810193565e-06,
+      "loss": 0.6362,
+      "step": 6924
+    },
+    {
+      "epoch": 1.9182825484764543,
+      "grad_norm": 0.703712522983551,
+      "learning_rate": 3.86231037713962e-06,
+      "loss": 0.6393,
+      "step": 6925
+    },
+    {
+      "epoch": 1.9185595567867035,
+      "grad_norm": 0.6949636340141296,
+      "learning_rate": 3.862004915160346e-06,
+      "loss": 0.645,
+      "step": 6926
+    },
+    {
+      "epoch": 1.918836565096953,
+      "grad_norm": 0.682137668132782,
+      "learning_rate": 3.861699424262229e-06,
+      "loss": 0.6339,
+      "step": 6927
+    },
+    {
+      "epoch": 1.9191135734072022,
+      "grad_norm": 0.7261094450950623,
+      "learning_rate": 3.861393904451756e-06,
+      "loss": 0.6081,
+      "step": 6928
+    },
+    {
+      "epoch": 1.9193905817174515,
+      "grad_norm": 0.7021178007125854,
+      "learning_rate": 3.861088355735414e-06,
+      "loss": 0.6645,
+      "step": 6929
+    },
+    {
+      "epoch": 1.919667590027701,
+      "grad_norm": 0.6526563763618469,
+      "learning_rate": 3.86078277811969e-06,
+      "loss": 0.5589,
+      "step": 6930
+    },
+    {
+      "epoch": 1.9199445983379502,
+      "grad_norm": 0.7094777226448059,
+      "learning_rate": 3.860477171611073e-06,
+      "loss": 0.6477,
+      "step": 6931
+    },
+    {
+      "epoch": 1.9202216066481994,
+      "grad_norm": 0.705771803855896,
+      "learning_rate": 3.860171536216052e-06,
+      "loss": 0.6519,
+      "step": 6932
+    },
+    {
+      "epoch": 1.9204986149584489,
+      "grad_norm": 0.6879554390907288,
+      "learning_rate": 3.859865871941115e-06,
+      "loss": 0.6272,
+      "step": 6933
+    },
+    {
+      "epoch": 1.920775623268698,
+      "grad_norm": 0.7064213156700134,
+      "learning_rate": 3.859560178792753e-06,
+      "loss": 0.6137,
+      "step": 6934
+    },
+    {
+      "epoch": 1.9210526315789473,
+      "grad_norm": 0.694358229637146,
+      "learning_rate": 3.8592544567774556e-06,
+      "loss": 0.644,
+      "step": 6935
+    },
+    {
+      "epoch": 1.9213296398891968,
+      "grad_norm": 0.6568576097488403,
+      "learning_rate": 3.858948705901716e-06,
+      "loss": 0.5892,
+      "step": 6936
+    },
+    {
+      "epoch": 1.9216066481994458,
+      "grad_norm": 0.6796807050704956,
+      "learning_rate": 3.858642926172025e-06,
+      "loss": 0.574,
+      "step": 6937
+    },
+    {
+      "epoch": 1.9218836565096953,
+      "grad_norm": 0.66985684633255,
+      "learning_rate": 3.8583371175948745e-06,
+      "loss": 0.6177,
+      "step": 6938
+    },
+    {
+      "epoch": 1.9221606648199447,
+      "grad_norm": 0.7518061399459839,
+      "learning_rate": 3.858031280176759e-06,
+      "loss": 0.6234,
+      "step": 6939
+    },
+    {
+      "epoch": 1.9224376731301938,
+      "grad_norm": 0.6907498240470886,
+      "learning_rate": 3.857725413924171e-06,
+      "loss": 0.6326,
+      "step": 6940
+    },
+    {
+      "epoch": 1.9227146814404432,
+      "grad_norm": 0.6684120893478394,
+      "learning_rate": 3.8574195188436046e-06,
+      "loss": 0.6297,
+      "step": 6941
+    },
+    {
+      "epoch": 1.9229916897506927,
+      "grad_norm": 0.6897550225257874,
+      "learning_rate": 3.857113594941556e-06,
+      "loss": 0.6339,
+      "step": 6942
+    },
+    {
+      "epoch": 1.9232686980609417,
+      "grad_norm": 0.7206911444664001,
+      "learning_rate": 3.85680764222452e-06,
+      "loss": 0.6439,
+      "step": 6943
+    },
+    {
+      "epoch": 1.9235457063711912,
+      "grad_norm": 0.7342230677604675,
+      "learning_rate": 3.856501660698993e-06,
+      "loss": 0.6193,
+      "step": 6944
+    },
+    {
+      "epoch": 1.9238227146814404,
+      "grad_norm": 0.6811208724975586,
+      "learning_rate": 3.856195650371471e-06,
+      "loss": 0.5903,
+      "step": 6945
+    },
+    {
+      "epoch": 1.9240997229916896,
+      "grad_norm": 0.6885085701942444,
+      "learning_rate": 3.855889611248452e-06,
+      "loss": 0.6008,
+      "step": 6946
+    },
+    {
+      "epoch": 1.924376731301939,
+      "grad_norm": 0.7090029716491699,
+      "learning_rate": 3.855583543336436e-06,
+      "loss": 0.6232,
+      "step": 6947
+    },
+    {
+      "epoch": 1.9246537396121883,
+      "grad_norm": 0.7263821959495544,
+      "learning_rate": 3.855277446641917e-06,
+      "loss": 0.6485,
+      "step": 6948
+    },
+    {
+      "epoch": 1.9249307479224376,
+      "grad_norm": 0.7115022540092468,
+      "learning_rate": 3.854971321171398e-06,
+      "loss": 0.6102,
+      "step": 6949
+    },
+    {
+      "epoch": 1.925207756232687,
+      "grad_norm": 0.6860071420669556,
+      "learning_rate": 3.854665166931377e-06,
+      "loss": 0.5767,
+      "step": 6950
+    },
+    {
+      "epoch": 1.9254847645429363,
+      "grad_norm": 0.6789618730545044,
+      "learning_rate": 3.854358983928356e-06,
+      "loss": 0.6297,
+      "step": 6951
+    },
+    {
+      "epoch": 1.9257617728531855,
+      "grad_norm": 0.688520073890686,
+      "learning_rate": 3.854052772168834e-06,
+      "loss": 0.6187,
+      "step": 6952
+    },
+    {
+      "epoch": 1.926038781163435,
+      "grad_norm": 0.7230477333068848,
+      "learning_rate": 3.853746531659315e-06,
+      "loss": 0.6277,
+      "step": 6953
+    },
+    {
+      "epoch": 1.9263157894736842,
+      "grad_norm": 0.687503457069397,
+      "learning_rate": 3.853440262406299e-06,
+      "loss": 0.6624,
+      "step": 6954
+    },
+    {
+      "epoch": 1.9265927977839334,
+      "grad_norm": 0.6723766922950745,
+      "learning_rate": 3.853133964416291e-06,
+      "loss": 0.574,
+      "step": 6955
+    },
+    {
+      "epoch": 1.926869806094183,
+      "grad_norm": 0.7176601886749268,
+      "learning_rate": 3.852827637695793e-06,
+      "loss": 0.6213,
+      "step": 6956
+    },
+    {
+      "epoch": 1.9271468144044321,
+      "grad_norm": 0.6974349021911621,
+      "learning_rate": 3.852521282251309e-06,
+      "loss": 0.6563,
+      "step": 6957
+    },
+    {
+      "epoch": 1.9274238227146814,
+      "grad_norm": 0.6568154692649841,
+      "learning_rate": 3.852214898089344e-06,
+      "loss": 0.6399,
+      "step": 6958
+    },
+    {
+      "epoch": 1.9277008310249308,
+      "grad_norm": 0.688548743724823,
+      "learning_rate": 3.851908485216404e-06,
+      "loss": 0.6376,
+      "step": 6959
+    },
+    {
+      "epoch": 1.92797783933518,
+      "grad_norm": 0.7333968281745911,
+      "learning_rate": 3.8516020436389945e-06,
+      "loss": 0.6203,
+      "step": 6960
+    },
+    {
+      "epoch": 1.9282548476454293,
+      "grad_norm": 0.6940326690673828,
+      "learning_rate": 3.851295573363622e-06,
+      "loss": 0.605,
+      "step": 6961
+    },
+    {
+      "epoch": 1.9285318559556788,
+      "grad_norm": 0.7149107456207275,
+      "learning_rate": 3.8509890743967935e-06,
+      "loss": 0.6015,
+      "step": 6962
+    },
+    {
+      "epoch": 1.928808864265928,
+      "grad_norm": 0.7581908106803894,
+      "learning_rate": 3.850682546745017e-06,
+      "loss": 0.6426,
+      "step": 6963
+    },
+    {
+      "epoch": 1.9290858725761773,
+      "grad_norm": 0.6629204750061035,
+      "learning_rate": 3.8503759904148005e-06,
+      "loss": 0.5729,
+      "step": 6964
+    },
+    {
+      "epoch": 1.9293628808864267,
+      "grad_norm": 0.6844896674156189,
+      "learning_rate": 3.850069405412654e-06,
+      "loss": 0.6621,
+      "step": 6965
+    },
+    {
+      "epoch": 1.929639889196676,
+      "grad_norm": 0.725421130657196,
+      "learning_rate": 3.849762791745085e-06,
+      "loss": 0.6268,
+      "step": 6966
+    },
+    {
+      "epoch": 1.9299168975069252,
+      "grad_norm": 0.7454162836074829,
+      "learning_rate": 3.8494561494186064e-06,
+      "loss": 0.6373,
+      "step": 6967
+    },
+    {
+      "epoch": 1.9301939058171746,
+      "grad_norm": 0.7131611108779907,
+      "learning_rate": 3.849149478439727e-06,
+      "loss": 0.6257,
+      "step": 6968
+    },
+    {
+      "epoch": 1.9304709141274237,
+      "grad_norm": 0.7022071480751038,
+      "learning_rate": 3.848842778814959e-06,
+      "loss": 0.6146,
+      "step": 6969
+    },
+    {
+      "epoch": 1.9307479224376731,
+      "grad_norm": 0.699440598487854,
+      "learning_rate": 3.848536050550814e-06,
+      "loss": 0.5895,
+      "step": 6970
+    },
+    {
+      "epoch": 1.9310249307479226,
+      "grad_norm": 0.6625146865844727,
+      "learning_rate": 3.8482292936538045e-06,
+      "loss": 0.5855,
+      "step": 6971
+    },
+    {
+      "epoch": 1.9313019390581716,
+      "grad_norm": 0.7169764041900635,
+      "learning_rate": 3.847922508130445e-06,
+      "loss": 0.6225,
+      "step": 6972
+    },
+    {
+      "epoch": 1.931578947368421,
+      "grad_norm": 0.6792388558387756,
+      "learning_rate": 3.847615693987248e-06,
+      "loss": 0.6174,
+      "step": 6973
+    },
+    {
+      "epoch": 1.9318559556786705,
+      "grad_norm": 0.7312890887260437,
+      "learning_rate": 3.847308851230728e-06,
+      "loss": 0.6168,
+      "step": 6974
+    },
+    {
+      "epoch": 1.9321329639889195,
+      "grad_norm": 0.6699267625808716,
+      "learning_rate": 3.847001979867401e-06,
+      "loss": 0.5454,
+      "step": 6975
+    },
+    {
+      "epoch": 1.932409972299169,
+      "grad_norm": 0.6831119060516357,
+      "learning_rate": 3.846695079903782e-06,
+      "loss": 0.5931,
+      "step": 6976
+    },
+    {
+      "epoch": 1.9326869806094182,
+      "grad_norm": 0.6708670854568481,
+      "learning_rate": 3.846388151346387e-06,
+      "loss": 0.5962,
+      "step": 6977
+    },
+    {
+      "epoch": 1.9329639889196675,
+      "grad_norm": 0.6922425627708435,
+      "learning_rate": 3.846081194201734e-06,
+      "loss": 0.6173,
+      "step": 6978
+    },
+    {
+      "epoch": 1.933240997229917,
+      "grad_norm": 0.6881274580955505,
+      "learning_rate": 3.845774208476339e-06,
+      "loss": 0.5995,
+      "step": 6979
+    },
+    {
+      "epoch": 1.9335180055401662,
+      "grad_norm": 0.6586433053016663,
+      "learning_rate": 3.845467194176721e-06,
+      "loss": 0.5384,
+      "step": 6980
+    },
+    {
+      "epoch": 1.9337950138504154,
+      "grad_norm": 0.7081720232963562,
+      "learning_rate": 3.845160151309399e-06,
+      "loss": 0.6032,
+      "step": 6981
+    },
+    {
+      "epoch": 1.9340720221606649,
+      "grad_norm": 0.681023120880127,
+      "learning_rate": 3.8448530798808905e-06,
+      "loss": 0.5974,
+      "step": 6982
+    },
+    {
+      "epoch": 1.934349030470914,
+      "grad_norm": 0.6930094361305237,
+      "learning_rate": 3.844545979897719e-06,
+      "loss": 0.6471,
+      "step": 6983
+    },
+    {
+      "epoch": 1.9346260387811633,
+      "grad_norm": 0.6955491900444031,
+      "learning_rate": 3.844238851366401e-06,
+      "loss": 0.59,
+      "step": 6984
+    },
+    {
+      "epoch": 1.9349030470914128,
+      "grad_norm": 0.6596637964248657,
+      "learning_rate": 3.843931694293459e-06,
+      "loss": 0.6269,
+      "step": 6985
+    },
+    {
+      "epoch": 1.935180055401662,
+      "grad_norm": 0.7362352013587952,
+      "learning_rate": 3.843624508685416e-06,
+      "loss": 0.6512,
+      "step": 6986
+    },
+    {
+      "epoch": 1.9354570637119113,
+      "grad_norm": 0.6971285343170166,
+      "learning_rate": 3.843317294548793e-06,
+      "loss": 0.618,
+      "step": 6987
+    },
+    {
+      "epoch": 1.9357340720221607,
+      "grad_norm": 0.7472339868545532,
+      "learning_rate": 3.843010051890114e-06,
+      "loss": 0.6397,
+      "step": 6988
+    },
+    {
+      "epoch": 1.93601108033241,
+      "grad_norm": 0.6800471544265747,
+      "learning_rate": 3.842702780715902e-06,
+      "loss": 0.5976,
+      "step": 6989
+    },
+    {
+      "epoch": 1.9362880886426592,
+      "grad_norm": 0.7063587307929993,
+      "learning_rate": 3.842395481032681e-06,
+      "loss": 0.6294,
+      "step": 6990
+    },
+    {
+      "epoch": 1.9365650969529087,
+      "grad_norm": 0.7280341386795044,
+      "learning_rate": 3.842088152846975e-06,
+      "loss": 0.6331,
+      "step": 6991
+    },
+    {
+      "epoch": 1.936842105263158,
+      "grad_norm": 0.6882317066192627,
+      "learning_rate": 3.841780796165312e-06,
+      "loss": 0.6022,
+      "step": 6992
+    },
+    {
+      "epoch": 1.9371191135734072,
+      "grad_norm": 0.6916040182113647,
+      "learning_rate": 3.8414734109942145e-06,
+      "loss": 0.6238,
+      "step": 6993
+    },
+    {
+      "epoch": 1.9373961218836566,
+      "grad_norm": 0.7397603988647461,
+      "learning_rate": 3.841165997340211e-06,
+      "loss": 0.6635,
+      "step": 6994
+    },
+    {
+      "epoch": 1.9376731301939059,
+      "grad_norm": 0.6798705458641052,
+      "learning_rate": 3.840858555209828e-06,
+      "loss": 0.6374,
+      "step": 6995
+    },
+    {
+      "epoch": 1.937950138504155,
+      "grad_norm": 0.7833229899406433,
+      "learning_rate": 3.8405510846095945e-06,
+      "loss": 0.6667,
+      "step": 6996
+    },
+    {
+      "epoch": 1.9382271468144046,
+      "grad_norm": 0.7077188491821289,
+      "learning_rate": 3.840243585546037e-06,
+      "loss": 0.5957,
+      "step": 6997
+    },
+    {
+      "epoch": 1.9385041551246538,
+      "grad_norm": 0.6946340799331665,
+      "learning_rate": 3.8399360580256876e-06,
+      "loss": 0.6245,
+      "step": 6998
+    },
+    {
+      "epoch": 1.938781163434903,
+      "grad_norm": 0.6952667832374573,
+      "learning_rate": 3.839628502055072e-06,
+      "loss": 0.6248,
+      "step": 6999
+    },
+    {
+      "epoch": 1.9390581717451525,
+      "grad_norm": 0.6928011775016785,
+      "learning_rate": 3.839320917640722e-06,
+      "loss": 0.6322,
+      "step": 7000
+    },
+    {
+      "epoch": 1.9393351800554015,
+      "grad_norm": 0.6780809164047241,
+      "learning_rate": 3.83901330478917e-06,
+      "loss": 0.5967,
+      "step": 7001
+    },
+    {
+      "epoch": 1.939612188365651,
+      "grad_norm": 0.6768255829811096,
+      "learning_rate": 3.838705663506947e-06,
+      "loss": 0.6307,
+      "step": 7002
+    },
+    {
+      "epoch": 1.9398891966759004,
+      "grad_norm": 0.6668100357055664,
+      "learning_rate": 3.838397993800582e-06,
+      "loss": 0.626,
+      "step": 7003
+    },
+    {
+      "epoch": 1.9401662049861494,
+      "grad_norm": 0.6606988906860352,
+      "learning_rate": 3.83809029567661e-06,
+      "loss": 0.5936,
+      "step": 7004
+    },
+    {
+      "epoch": 1.940443213296399,
+      "grad_norm": 0.7164947986602783,
+      "learning_rate": 3.837782569141565e-06,
+      "loss": 0.581,
+      "step": 7005
+    },
+    {
+      "epoch": 1.9407202216066484,
+      "grad_norm": 0.7346347570419312,
+      "learning_rate": 3.837474814201979e-06,
+      "loss": 0.6123,
+      "step": 7006
+    },
+    {
+      "epoch": 1.9409972299168974,
+      "grad_norm": 0.7159395813941956,
+      "learning_rate": 3.837167030864387e-06,
+      "loss": 0.6525,
+      "step": 7007
+    },
+    {
+      "epoch": 1.9412742382271468,
+      "grad_norm": 0.7122390866279602,
+      "learning_rate": 3.8368592191353246e-06,
+      "loss": 0.6176,
+      "step": 7008
+    },
+    {
+      "epoch": 1.941551246537396,
+      "grad_norm": 0.7087534666061401,
+      "learning_rate": 3.836551379021326e-06,
+      "loss": 0.6226,
+      "step": 7009
+    },
+    {
+      "epoch": 1.9418282548476453,
+      "grad_norm": 0.7045443058013916,
+      "learning_rate": 3.83624351052893e-06,
+      "loss": 0.6429,
+      "step": 7010
+    },
+    {
+      "epoch": 1.9421052631578948,
+      "grad_norm": 0.6841949820518494,
+      "learning_rate": 3.835935613664671e-06,
+      "loss": 0.6247,
+      "step": 7011
+    },
+    {
+      "epoch": 1.942382271468144,
+      "grad_norm": 0.7041621208190918,
+      "learning_rate": 3.835627688435088e-06,
+      "loss": 0.6402,
+      "step": 7012
+    },
+    {
+      "epoch": 1.9426592797783933,
+      "grad_norm": 0.7085411548614502,
+      "learning_rate": 3.835319734846717e-06,
+      "loss": 0.6208,
+      "step": 7013
+    },
+    {
+      "epoch": 1.9429362880886427,
+      "grad_norm": 0.7151727676391602,
+      "learning_rate": 3.835011752906099e-06,
+      "loss": 0.6507,
+      "step": 7014
+    },
+    {
+      "epoch": 1.943213296398892,
+      "grad_norm": 0.6990612149238586,
+      "learning_rate": 3.834703742619772e-06,
+      "loss": 0.5722,
+      "step": 7015
+    },
+    {
+      "epoch": 1.9434903047091412,
+      "grad_norm": 0.728863000869751,
+      "learning_rate": 3.8343957039942755e-06,
+      "loss": 0.6511,
+      "step": 7016
+    },
+    {
+      "epoch": 1.9437673130193907,
+      "grad_norm": 0.681682288646698,
+      "learning_rate": 3.834087637036152e-06,
+      "loss": 0.5959,
+      "step": 7017
+    },
+    {
+      "epoch": 1.94404432132964,
+      "grad_norm": 0.7117037177085876,
+      "learning_rate": 3.833779541751941e-06,
+      "loss": 0.6421,
+      "step": 7018
+    },
+    {
+      "epoch": 1.9443213296398891,
+      "grad_norm": 0.6765846610069275,
+      "learning_rate": 3.833471418148183e-06,
+      "loss": 0.5888,
+      "step": 7019
+    },
+    {
+      "epoch": 1.9445983379501386,
+      "grad_norm": 0.7212615013122559,
+      "learning_rate": 3.833163266231421e-06,
+      "loss": 0.6119,
+      "step": 7020
+    },
+    {
+      "epoch": 1.9448753462603878,
+      "grad_norm": 0.6902775168418884,
+      "learning_rate": 3.832855086008199e-06,
+      "loss": 0.6158,
+      "step": 7021
+    },
+    {
+      "epoch": 1.945152354570637,
+      "grad_norm": 0.7205889821052551,
+      "learning_rate": 3.832546877485061e-06,
+      "loss": 0.6345,
+      "step": 7022
+    },
+    {
+      "epoch": 1.9454293628808865,
+      "grad_norm": 0.6866292357444763,
+      "learning_rate": 3.832238640668549e-06,
+      "loss": 0.6169,
+      "step": 7023
+    },
+    {
+      "epoch": 1.9457063711911358,
+      "grad_norm": 0.7034695148468018,
+      "learning_rate": 3.8319303755652084e-06,
+      "loss": 0.6795,
+      "step": 7024
+    },
+    {
+      "epoch": 1.945983379501385,
+      "grad_norm": 0.7042098045349121,
+      "learning_rate": 3.8316220821815845e-06,
+      "loss": 0.629,
+      "step": 7025
+    },
+    {
+      "epoch": 1.9462603878116345,
+      "grad_norm": 0.685293972492218,
+      "learning_rate": 3.831313760524223e-06,
+      "loss": 0.6395,
+      "step": 7026
+    },
+    {
+      "epoch": 1.9465373961218837,
+      "grad_norm": 0.6994853019714355,
+      "learning_rate": 3.831005410599671e-06,
+      "loss": 0.5892,
+      "step": 7027
+    },
+    {
+      "epoch": 1.946814404432133,
+      "grad_norm": 0.714125394821167,
+      "learning_rate": 3.830697032414474e-06,
+      "loss": 0.6438,
+      "step": 7028
+    },
+    {
+      "epoch": 1.9470914127423824,
+      "grad_norm": 0.6867514848709106,
+      "learning_rate": 3.830388625975182e-06,
+      "loss": 0.6684,
+      "step": 7029
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.7274505496025085,
+      "learning_rate": 3.830080191288342e-06,
+      "loss": 0.6044,
+      "step": 7030
+    },
+    {
+      "epoch": 1.9476454293628809,
+      "grad_norm": 0.7258849143981934,
+      "learning_rate": 3.829771728360502e-06,
+      "loss": 0.5791,
+      "step": 7031
+    },
+    {
+      "epoch": 1.9479224376731303,
+      "grad_norm": 0.786751925945282,
+      "learning_rate": 3.829463237198213e-06,
+      "loss": 0.6339,
+      "step": 7032
+    },
+    {
+      "epoch": 1.9481994459833794,
+      "grad_norm": 0.7100234627723694,
+      "learning_rate": 3.829154717808025e-06,
+      "loss": 0.6027,
+      "step": 7033
+    },
+    {
+      "epoch": 1.9484764542936288,
+      "grad_norm": 0.7123561501502991,
+      "learning_rate": 3.828846170196487e-06,
+      "loss": 0.6743,
+      "step": 7034
+    },
+    {
+      "epoch": 1.9487534626038783,
+      "grad_norm": 0.6572225689888,
+      "learning_rate": 3.828537594370152e-06,
+      "loss": 0.6214,
+      "step": 7035
+    },
+    {
+      "epoch": 1.9490304709141273,
+      "grad_norm": 0.7247850894927979,
+      "learning_rate": 3.828228990335571e-06,
+      "loss": 0.634,
+      "step": 7036
+    },
+    {
+      "epoch": 1.9493074792243767,
+      "grad_norm": 0.6873048543930054,
+      "learning_rate": 3.827920358099297e-06,
+      "loss": 0.6139,
+      "step": 7037
+    },
+    {
+      "epoch": 1.949584487534626,
+      "grad_norm": 0.6822011470794678,
+      "learning_rate": 3.827611697667881e-06,
+      "loss": 0.5926,
+      "step": 7038
+    },
+    {
+      "epoch": 1.9498614958448752,
+      "grad_norm": 0.699113130569458,
+      "learning_rate": 3.827303009047879e-06,
+      "loss": 0.5752,
+      "step": 7039
+    },
+    {
+      "epoch": 1.9501385041551247,
+      "grad_norm": 0.6774667501449585,
+      "learning_rate": 3.826994292245846e-06,
+      "loss": 0.6233,
+      "step": 7040
+    },
+    {
+      "epoch": 1.950415512465374,
+      "grad_norm": 0.6731832027435303,
+      "learning_rate": 3.826685547268334e-06,
+      "loss": 0.6007,
+      "step": 7041
+    },
+    {
+      "epoch": 1.9506925207756232,
+      "grad_norm": 0.6880195736885071,
+      "learning_rate": 3.8263767741219e-06,
+      "loss": 0.5603,
+      "step": 7042
+    },
+    {
+      "epoch": 1.9509695290858726,
+      "grad_norm": 0.6659010052680969,
+      "learning_rate": 3.8260679728131e-06,
+      "loss": 0.6244,
+      "step": 7043
+    },
+    {
+      "epoch": 1.9512465373961219,
+      "grad_norm": 0.6810696125030518,
+      "learning_rate": 3.825759143348491e-06,
+      "loss": 0.6181,
+      "step": 7044
+    },
+    {
+      "epoch": 1.951523545706371,
+      "grad_norm": 0.7221637964248657,
+      "learning_rate": 3.825450285734629e-06,
+      "loss": 0.6393,
+      "step": 7045
+    },
+    {
+      "epoch": 1.9518005540166206,
+      "grad_norm": 0.6795384883880615,
+      "learning_rate": 3.825141399978073e-06,
+      "loss": 0.6128,
+      "step": 7046
+    },
+    {
+      "epoch": 1.9520775623268698,
+      "grad_norm": 0.666467547416687,
+      "learning_rate": 3.824832486085381e-06,
+      "loss": 0.6414,
+      "step": 7047
+    },
+    {
+      "epoch": 1.952354570637119,
+      "grad_norm": 0.6998277306556702,
+      "learning_rate": 3.824523544063113e-06,
+      "loss": 0.6423,
+      "step": 7048
+    },
+    {
+      "epoch": 1.9526315789473685,
+      "grad_norm": 0.7023756504058838,
+      "learning_rate": 3.824214573917826e-06,
+      "loss": 0.6358,
+      "step": 7049
+    },
+    {
+      "epoch": 1.9529085872576177,
+      "grad_norm": 0.6664944887161255,
+      "learning_rate": 3.8239055756560824e-06,
+      "loss": 0.6213,
+      "step": 7050
+    },
+    {
+      "epoch": 1.953185595567867,
+      "grad_norm": 0.7106117606163025,
+      "learning_rate": 3.823596549284443e-06,
+      "loss": 0.6264,
+      "step": 7051
+    },
+    {
+      "epoch": 1.9534626038781164,
+      "grad_norm": 0.673440158367157,
+      "learning_rate": 3.823287494809469e-06,
+      "loss": 0.6037,
+      "step": 7052
+    },
+    {
+      "epoch": 1.9537396121883657,
+      "grad_norm": 0.714493453502655,
+      "learning_rate": 3.822978412237721e-06,
+      "loss": 0.6296,
+      "step": 7053
+    },
+    {
+      "epoch": 1.954016620498615,
+      "grad_norm": 0.7021950483322144,
+      "learning_rate": 3.822669301575764e-06,
+      "loss": 0.5957,
+      "step": 7054
+    },
+    {
+      "epoch": 1.9542936288088644,
+      "grad_norm": 0.6818373799324036,
+      "learning_rate": 3.822360162830159e-06,
+      "loss": 0.576,
+      "step": 7055
+    },
+    {
+      "epoch": 1.9545706371191136,
+      "grad_norm": 0.6853843331336975,
+      "learning_rate": 3.8220509960074724e-06,
+      "loss": 0.6438,
+      "step": 7056
+    },
+    {
+      "epoch": 1.9548476454293628,
+      "grad_norm": 0.728827178478241,
+      "learning_rate": 3.821741801114266e-06,
+      "loss": 0.6795,
+      "step": 7057
+    },
+    {
+      "epoch": 1.9551246537396123,
+      "grad_norm": 0.7150498628616333,
+      "learning_rate": 3.821432578157105e-06,
+      "loss": 0.627,
+      "step": 7058
+    },
+    {
+      "epoch": 1.9554016620498615,
+      "grad_norm": 0.6961706876754761,
+      "learning_rate": 3.8211233271425575e-06,
+      "loss": 0.6272,
+      "step": 7059
+    },
+    {
+      "epoch": 1.9556786703601108,
+      "grad_norm": 0.7418909072875977,
+      "learning_rate": 3.820814048077186e-06,
+      "loss": 0.651,
+      "step": 7060
+    },
+    {
+      "epoch": 1.9559556786703602,
+      "grad_norm": 0.7025729417800903,
+      "learning_rate": 3.8205047409675614e-06,
+      "loss": 0.6511,
+      "step": 7061
+    },
+    {
+      "epoch": 1.9562326869806093,
+      "grad_norm": 0.7075372338294983,
+      "learning_rate": 3.820195405820248e-06,
+      "loss": 0.6339,
+      "step": 7062
+    },
+    {
+      "epoch": 1.9565096952908587,
+      "grad_norm": 0.7278872728347778,
+      "learning_rate": 3.819886042641814e-06,
+      "loss": 0.6402,
+      "step": 7063
+    },
+    {
+      "epoch": 1.9567867036011082,
+      "grad_norm": 0.7034096121788025,
+      "learning_rate": 3.819576651438829e-06,
+      "loss": 0.6072,
+      "step": 7064
+    },
+    {
+      "epoch": 1.9570637119113572,
+      "grad_norm": 0.6780850887298584,
+      "learning_rate": 3.8192672322178625e-06,
+      "loss": 0.6045,
+      "step": 7065
+    },
+    {
+      "epoch": 1.9573407202216067,
+      "grad_norm": 0.6713101863861084,
+      "learning_rate": 3.818957784985483e-06,
+      "loss": 0.6397,
+      "step": 7066
+    },
+    {
+      "epoch": 1.9576177285318561,
+      "grad_norm": 0.6640244722366333,
+      "learning_rate": 3.818648309748262e-06,
+      "loss": 0.5437,
+      "step": 7067
+    },
+    {
+      "epoch": 1.9578947368421051,
+      "grad_norm": 0.6953721642494202,
+      "learning_rate": 3.81833880651277e-06,
+      "loss": 0.6341,
+      "step": 7068
+    },
+    {
+      "epoch": 1.9581717451523546,
+      "grad_norm": 0.6671434640884399,
+      "learning_rate": 3.818029275285578e-06,
+      "loss": 0.5959,
+      "step": 7069
+    },
+    {
+      "epoch": 1.9584487534626038,
+      "grad_norm": 0.7431301474571228,
+      "learning_rate": 3.817719716073258e-06,
+      "loss": 0.5948,
+      "step": 7070
+    },
+    {
+      "epoch": 1.958725761772853,
+      "grad_norm": 0.677884578704834,
+      "learning_rate": 3.817410128882384e-06,
+      "loss": 0.5916,
+      "step": 7071
+    },
+    {
+      "epoch": 1.9590027700831025,
+      "grad_norm": 0.6750688552856445,
+      "learning_rate": 3.817100513719529e-06,
+      "loss": 0.6194,
+      "step": 7072
+    },
+    {
+      "epoch": 1.9592797783933518,
+      "grad_norm": 0.6718873381614685,
+      "learning_rate": 3.816790870591267e-06,
+      "loss": 0.6022,
+      "step": 7073
+    },
+    {
+      "epoch": 1.959556786703601,
+      "grad_norm": 0.7275941967964172,
+      "learning_rate": 3.816481199504171e-06,
+      "loss": 0.599,
+      "step": 7074
+    },
+    {
+      "epoch": 1.9598337950138505,
+      "grad_norm": 0.730496346950531,
+      "learning_rate": 3.816171500464818e-06,
+      "loss": 0.6444,
+      "step": 7075
+    },
+    {
+      "epoch": 1.9601108033240997,
+      "grad_norm": 0.7540499567985535,
+      "learning_rate": 3.815861773479782e-06,
+      "loss": 0.6436,
+      "step": 7076
+    },
+    {
+      "epoch": 1.960387811634349,
+      "grad_norm": 0.6974166631698608,
+      "learning_rate": 3.815552018555641e-06,
+      "loss": 0.5722,
+      "step": 7077
+    },
+    {
+      "epoch": 1.9606648199445984,
+      "grad_norm": 0.6711729764938354,
+      "learning_rate": 3.815242235698972e-06,
+      "loss": 0.6126,
+      "step": 7078
+    },
+    {
+      "epoch": 1.9609418282548476,
+      "grad_norm": 0.74052494764328,
+      "learning_rate": 3.814932424916349e-06,
+      "loss": 0.6235,
+      "step": 7079
+    },
+    {
+      "epoch": 1.9612188365650969,
+      "grad_norm": 0.692291796207428,
+      "learning_rate": 3.8146225862143542e-06,
+      "loss": 0.5607,
+      "step": 7080
+    },
+    {
+      "epoch": 1.9614958448753463,
+      "grad_norm": 0.7283056378364563,
+      "learning_rate": 3.8143127195995648e-06,
+      "loss": 0.6665,
+      "step": 7081
+    },
+    {
+      "epoch": 1.9617728531855956,
+      "grad_norm": 0.6930385828018188,
+      "learning_rate": 3.81400282507856e-06,
+      "loss": 0.5921,
+      "step": 7082
+    },
+    {
+      "epoch": 1.9620498614958448,
+      "grad_norm": 0.7250744700431824,
+      "learning_rate": 3.8136929026579188e-06,
+      "loss": 0.6461,
+      "step": 7083
+    },
+    {
+      "epoch": 1.9623268698060943,
+      "grad_norm": 0.6802274584770203,
+      "learning_rate": 3.8133829523442234e-06,
+      "loss": 0.572,
+      "step": 7084
+    },
+    {
+      "epoch": 1.9626038781163435,
+      "grad_norm": 0.6919212937355042,
+      "learning_rate": 3.8130729741440536e-06,
+      "loss": 0.5794,
+      "step": 7085
+    },
+    {
+      "epoch": 1.9628808864265928,
+      "grad_norm": 0.7192142605781555,
+      "learning_rate": 3.812762968063991e-06,
+      "loss": 0.6523,
+      "step": 7086
+    },
+    {
+      "epoch": 1.9631578947368422,
+      "grad_norm": 0.689793050289154,
+      "learning_rate": 3.8124529341106188e-06,
+      "loss": 0.5646,
+      "step": 7087
+    },
+    {
+      "epoch": 1.9634349030470915,
+      "grad_norm": 0.6961142420768738,
+      "learning_rate": 3.8121428722905183e-06,
+      "loss": 0.5875,
+      "step": 7088
+    },
+    {
+      "epoch": 1.9637119113573407,
+      "grad_norm": 0.7265340089797974,
+      "learning_rate": 3.8118327826102747e-06,
+      "loss": 0.6319,
+      "step": 7089
+    },
+    {
+      "epoch": 1.9639889196675901,
+      "grad_norm": 0.6598079204559326,
+      "learning_rate": 3.8115226650764708e-06,
+      "loss": 0.5717,
+      "step": 7090
+    },
+    {
+      "epoch": 1.9642659279778394,
+      "grad_norm": 0.7008779644966125,
+      "learning_rate": 3.8112125196956917e-06,
+      "loss": 0.6106,
+      "step": 7091
+    },
+    {
+      "epoch": 1.9645429362880886,
+      "grad_norm": 0.6735950708389282,
+      "learning_rate": 3.810902346474521e-06,
+      "loss": 0.6352,
+      "step": 7092
+    },
+    {
+      "epoch": 1.964819944598338,
+      "grad_norm": 0.7206198573112488,
+      "learning_rate": 3.810592145419547e-06,
+      "loss": 0.6003,
+      "step": 7093
+    },
+    {
+      "epoch": 1.965096952908587,
+      "grad_norm": 0.7294063568115234,
+      "learning_rate": 3.810281916537355e-06,
+      "loss": 0.6278,
+      "step": 7094
+    },
+    {
+      "epoch": 1.9653739612188366,
+      "grad_norm": 0.7309246063232422,
+      "learning_rate": 3.8099716598345317e-06,
+      "loss": 0.5992,
+      "step": 7095
+    },
+    {
+      "epoch": 1.965650969529086,
+      "grad_norm": 0.7022836804389954,
+      "learning_rate": 3.8096613753176635e-06,
+      "loss": 0.5967,
+      "step": 7096
+    },
+    {
+      "epoch": 1.965927977839335,
+      "grad_norm": 0.7502058148384094,
+      "learning_rate": 3.809351062993342e-06,
+      "loss": 0.607,
+      "step": 7097
+    },
+    {
+      "epoch": 1.9662049861495845,
+      "grad_norm": 0.7289679050445557,
+      "learning_rate": 3.809040722868152e-06,
+      "loss": 0.589,
+      "step": 7098
+    },
+    {
+      "epoch": 1.966481994459834,
+      "grad_norm": 0.6707630157470703,
+      "learning_rate": 3.8087303549486844e-06,
+      "loss": 0.5773,
+      "step": 7099
+    },
+    {
+      "epoch": 1.966759002770083,
+      "grad_norm": 0.7455605864524841,
+      "learning_rate": 3.8084199592415305e-06,
+      "loss": 0.6105,
+      "step": 7100
+    },
+    {
+      "epoch": 1.9670360110803324,
+      "grad_norm": 0.7032577395439148,
+      "learning_rate": 3.8081095357532787e-06,
+      "loss": 0.6293,
+      "step": 7101
+    },
+    {
+      "epoch": 1.9673130193905817,
+      "grad_norm": 0.6860726475715637,
+      "learning_rate": 3.807799084490521e-06,
+      "loss": 0.6318,
+      "step": 7102
+    },
+    {
+      "epoch": 1.967590027700831,
+      "grad_norm": 0.7498553991317749,
+      "learning_rate": 3.807488605459849e-06,
+      "loss": 0.6472,
+      "step": 7103
+    },
+    {
+      "epoch": 1.9678670360110804,
+      "grad_norm": 0.7227993607521057,
+      "learning_rate": 3.807178098667855e-06,
+      "loss": 0.594,
+      "step": 7104
+    },
+    {
+      "epoch": 1.9681440443213296,
+      "grad_norm": 0.721440851688385,
+      "learning_rate": 3.8068675641211316e-06,
+      "loss": 0.6256,
+      "step": 7105
+    },
+    {
+      "epoch": 1.9684210526315788,
+      "grad_norm": 0.7207381725311279,
+      "learning_rate": 3.8065570018262733e-06,
+      "loss": 0.582,
+      "step": 7106
+    },
+    {
+      "epoch": 1.9686980609418283,
+      "grad_norm": 0.676328718662262,
+      "learning_rate": 3.806246411789872e-06,
+      "loss": 0.6137,
+      "step": 7107
+    },
+    {
+      "epoch": 1.9689750692520775,
+      "grad_norm": 0.7162710428237915,
+      "learning_rate": 3.805935794018525e-06,
+      "loss": 0.6427,
+      "step": 7108
+    },
+    {
+      "epoch": 1.9692520775623268,
+      "grad_norm": 0.7399466037750244,
+      "learning_rate": 3.8056251485188244e-06,
+      "loss": 0.6157,
+      "step": 7109
+    },
+    {
+      "epoch": 1.9695290858725762,
+      "grad_norm": 0.6657085418701172,
+      "learning_rate": 3.80531447529737e-06,
+      "loss": 0.5755,
+      "step": 7110
+    },
+    {
+      "epoch": 1.9698060941828255,
+      "grad_norm": 0.6843543648719788,
+      "learning_rate": 3.805003774360754e-06,
+      "loss": 0.6252,
+      "step": 7111
+    },
+    {
+      "epoch": 1.9700831024930747,
+      "grad_norm": 0.7047277092933655,
+      "learning_rate": 3.804693045715576e-06,
+      "loss": 0.6521,
+      "step": 7112
+    },
+    {
+      "epoch": 1.9703601108033242,
+      "grad_norm": 0.7122973203659058,
+      "learning_rate": 3.804382289368432e-06,
+      "loss": 0.637,
+      "step": 7113
+    },
+    {
+      "epoch": 1.9706371191135734,
+      "grad_norm": 0.7243662476539612,
+      "learning_rate": 3.8040715053259213e-06,
+      "loss": 0.6483,
+      "step": 7114
+    },
+    {
+      "epoch": 1.9709141274238227,
+      "grad_norm": 0.6891406774520874,
+      "learning_rate": 3.8037606935946425e-06,
+      "loss": 0.6481,
+      "step": 7115
+    },
+    {
+      "epoch": 1.9711911357340721,
+      "grad_norm": 0.6714333295822144,
+      "learning_rate": 3.803449854181195e-06,
+      "loss": 0.6116,
+      "step": 7116
+    },
+    {
+      "epoch": 1.9714681440443214,
+      "grad_norm": 0.7190486788749695,
+      "learning_rate": 3.8031389870921785e-06,
+      "loss": 0.6091,
+      "step": 7117
+    },
+    {
+      "epoch": 1.9717451523545706,
+      "grad_norm": 0.7250654697418213,
+      "learning_rate": 3.8028280923341927e-06,
+      "loss": 0.6217,
+      "step": 7118
+    },
+    {
+      "epoch": 1.97202216066482,
+      "grad_norm": 0.738471508026123,
+      "learning_rate": 3.802517169913841e-06,
+      "loss": 0.6285,
+      "step": 7119
+    },
+    {
+      "epoch": 1.9722991689750693,
+      "grad_norm": 0.6984786987304688,
+      "learning_rate": 3.802206219837722e-06,
+      "loss": 0.633,
+      "step": 7120
+    },
+    {
+      "epoch": 1.9725761772853185,
+      "grad_norm": 0.747661292552948,
+      "learning_rate": 3.8018952421124393e-06,
+      "loss": 0.5987,
+      "step": 7121
+    },
+    {
+      "epoch": 1.972853185595568,
+      "grad_norm": 0.6906194090843201,
+      "learning_rate": 3.8015842367445967e-06,
+      "loss": 0.5797,
+      "step": 7122
+    },
+    {
+      "epoch": 1.973130193905817,
+      "grad_norm": 0.7184000611305237,
+      "learning_rate": 3.8012732037407975e-06,
+      "loss": 0.6018,
+      "step": 7123
+    },
+    {
+      "epoch": 1.9734072022160665,
+      "grad_norm": 0.6846181750297546,
+      "learning_rate": 3.8009621431076436e-06,
+      "loss": 0.6387,
+      "step": 7124
+    },
+    {
+      "epoch": 1.973684210526316,
+      "grad_norm": 0.699408233165741,
+      "learning_rate": 3.8006510548517416e-06,
+      "loss": 0.6615,
+      "step": 7125
+    },
+    {
+      "epoch": 1.973961218836565,
+      "grad_norm": 0.6856897473335266,
+      "learning_rate": 3.800339938979697e-06,
+      "loss": 0.6589,
+      "step": 7126
+    },
+    {
+      "epoch": 1.9742382271468144,
+      "grad_norm": 0.6751475930213928,
+      "learning_rate": 3.8000287954981134e-06,
+      "loss": 0.6341,
+      "step": 7127
+    },
+    {
+      "epoch": 1.9745152354570639,
+      "grad_norm": 0.7153899669647217,
+      "learning_rate": 3.799717624413599e-06,
+      "loss": 0.6402,
+      "step": 7128
+    },
+    {
+      "epoch": 1.9747922437673129,
+      "grad_norm": 0.6692858934402466,
+      "learning_rate": 3.7994064257327613e-06,
+      "loss": 0.6092,
+      "step": 7129
+    },
+    {
+      "epoch": 1.9750692520775623,
+      "grad_norm": 0.7093313932418823,
+      "learning_rate": 3.7990951994622062e-06,
+      "loss": 0.6155,
+      "step": 7130
+    },
+    {
+      "epoch": 1.9753462603878116,
+      "grad_norm": 0.7412492632865906,
+      "learning_rate": 3.7987839456085424e-06,
+      "loss": 0.6334,
+      "step": 7131
+    },
+    {
+      "epoch": 1.9756232686980608,
+      "grad_norm": 0.721476674079895,
+      "learning_rate": 3.7984726641783793e-06,
+      "loss": 0.6307,
+      "step": 7132
+    },
+    {
+      "epoch": 1.9759002770083103,
+      "grad_norm": 0.6939054727554321,
+      "learning_rate": 3.7981613551783246e-06,
+      "loss": 0.5683,
+      "step": 7133
+    },
+    {
+      "epoch": 1.9761772853185595,
+      "grad_norm": 0.7083145380020142,
+      "learning_rate": 3.7978500186149902e-06,
+      "loss": 0.5428,
+      "step": 7134
+    },
+    {
+      "epoch": 1.9764542936288088,
+      "grad_norm": 0.6860464811325073,
+      "learning_rate": 3.7975386544949848e-06,
+      "loss": 0.6463,
+      "step": 7135
+    },
+    {
+      "epoch": 1.9767313019390582,
+      "grad_norm": 0.6742307543754578,
+      "learning_rate": 3.7972272628249206e-06,
+      "loss": 0.606,
+      "step": 7136
+    },
+    {
+      "epoch": 1.9770083102493075,
+      "grad_norm": 0.6865066289901733,
+      "learning_rate": 3.7969158436114083e-06,
+      "loss": 0.667,
+      "step": 7137
+    },
+    {
+      "epoch": 1.9772853185595567,
+      "grad_norm": 0.716090202331543,
+      "learning_rate": 3.796604396861061e-06,
+      "loss": 0.6379,
+      "step": 7138
+    },
+    {
+      "epoch": 1.9775623268698062,
+      "grad_norm": 0.7939145565032959,
+      "learning_rate": 3.796292922580492e-06,
+      "loss": 0.6028,
+      "step": 7139
+    },
+    {
+      "epoch": 1.9778393351800554,
+      "grad_norm": 0.7300482988357544,
+      "learning_rate": 3.7959814207763134e-06,
+      "loss": 0.6127,
+      "step": 7140
+    },
+    {
+      "epoch": 1.9781163434903046,
+      "grad_norm": 0.6948409080505371,
+      "learning_rate": 3.795669891455139e-06,
+      "loss": 0.6754,
+      "step": 7141
+    },
+    {
+      "epoch": 1.978393351800554,
+      "grad_norm": 0.724706768989563,
+      "learning_rate": 3.7953583346235843e-06,
+      "loss": 0.6248,
+      "step": 7142
+    },
+    {
+      "epoch": 1.9786703601108033,
+      "grad_norm": 0.749442994594574,
+      "learning_rate": 3.795046750288265e-06,
+      "loss": 0.6157,
+      "step": 7143
+    },
+    {
+      "epoch": 1.9789473684210526,
+      "grad_norm": 0.6890953779220581,
+      "learning_rate": 3.7947351384557953e-06,
+      "loss": 0.6027,
+      "step": 7144
+    },
+    {
+      "epoch": 1.979224376731302,
+      "grad_norm": 0.6894163489341736,
+      "learning_rate": 3.794423499132792e-06,
+      "loss": 0.6555,
+      "step": 7145
+    },
+    {
+      "epoch": 1.9795013850415513,
+      "grad_norm": 0.7086030840873718,
+      "learning_rate": 3.7941118323258717e-06,
+      "loss": 0.6216,
+      "step": 7146
+    },
+    {
+      "epoch": 1.9797783933518005,
+      "grad_norm": 0.731605589389801,
+      "learning_rate": 3.7938001380416536e-06,
+      "loss": 0.614,
+      "step": 7147
+    },
+    {
+      "epoch": 1.98005540166205,
+      "grad_norm": 0.7541550397872925,
+      "learning_rate": 3.7934884162867546e-06,
+      "loss": 0.616,
+      "step": 7148
+    },
+    {
+      "epoch": 1.9803324099722992,
+      "grad_norm": 0.710583508014679,
+      "learning_rate": 3.793176667067793e-06,
+      "loss": 0.6436,
+      "step": 7149
+    },
+    {
+      "epoch": 1.9806094182825484,
+      "grad_norm": 0.7192587852478027,
+      "learning_rate": 3.7928648903913877e-06,
+      "loss": 0.6342,
+      "step": 7150
+    },
+    {
+      "epoch": 1.980886426592798,
+      "grad_norm": 0.6838364601135254,
+      "learning_rate": 3.79255308626416e-06,
+      "loss": 0.6229,
+      "step": 7151
+    },
+    {
+      "epoch": 1.9811634349030471,
+      "grad_norm": 0.6993123888969421,
+      "learning_rate": 3.7922412546927288e-06,
+      "loss": 0.6653,
+      "step": 7152
+    },
+    {
+      "epoch": 1.9814404432132964,
+      "grad_norm": 0.6950284838676453,
+      "learning_rate": 3.791929395683716e-06,
+      "loss": 0.6516,
+      "step": 7153
+    },
+    {
+      "epoch": 1.9817174515235458,
+      "grad_norm": 0.7114899754524231,
+      "learning_rate": 3.7916175092437436e-06,
+      "loss": 0.6277,
+      "step": 7154
+    },
+    {
+      "epoch": 1.9819944598337949,
+      "grad_norm": 0.7198684215545654,
+      "learning_rate": 3.791305595379432e-06,
+      "loss": 0.5895,
+      "step": 7155
+    },
+    {
+      "epoch": 1.9822714681440443,
+      "grad_norm": 0.7094388008117676,
+      "learning_rate": 3.7909936540974052e-06,
+      "loss": 0.6192,
+      "step": 7156
+    },
+    {
+      "epoch": 1.9825484764542938,
+      "grad_norm": 0.7088209390640259,
+      "learning_rate": 3.7906816854042866e-06,
+      "loss": 0.6554,
+      "step": 7157
+    },
+    {
+      "epoch": 1.9828254847645428,
+      "grad_norm": 0.7191725969314575,
+      "learning_rate": 3.7903696893067e-06,
+      "loss": 0.6065,
+      "step": 7158
+    },
+    {
+      "epoch": 1.9831024930747922,
+      "grad_norm": 0.6958176493644714,
+      "learning_rate": 3.7900576658112687e-06,
+      "loss": 0.5931,
+      "step": 7159
+    },
+    {
+      "epoch": 1.9833795013850417,
+      "grad_norm": 0.7286674380302429,
+      "learning_rate": 3.789745614924619e-06,
+      "loss": 0.5997,
+      "step": 7160
+    },
+    {
+      "epoch": 1.9836565096952907,
+      "grad_norm": 0.6912921071052551,
+      "learning_rate": 3.789433536653377e-06,
+      "loss": 0.5836,
+      "step": 7161
+    },
+    {
+      "epoch": 1.9839335180055402,
+      "grad_norm": 0.7034099698066711,
+      "learning_rate": 3.789121431004168e-06,
+      "loss": 0.6483,
+      "step": 7162
+    },
+    {
+      "epoch": 1.9842105263157894,
+      "grad_norm": 0.7091851830482483,
+      "learning_rate": 3.7888092979836176e-06,
+      "loss": 0.642,
+      "step": 7163
+    },
+    {
+      "epoch": 1.9844875346260387,
+      "grad_norm": 0.7511205077171326,
+      "learning_rate": 3.7884971375983563e-06,
+      "loss": 0.6504,
+      "step": 7164
+    },
+    {
+      "epoch": 1.9847645429362881,
+      "grad_norm": 0.6957712769508362,
+      "learning_rate": 3.7881849498550094e-06,
+      "loss": 0.5966,
+      "step": 7165
+    },
+    {
+      "epoch": 1.9850415512465374,
+      "grad_norm": 0.7258434891700745,
+      "learning_rate": 3.7878727347602067e-06,
+      "loss": 0.6163,
+      "step": 7166
+    },
+    {
+      "epoch": 1.9853185595567866,
+      "grad_norm": 0.6509557366371155,
+      "learning_rate": 3.7875604923205768e-06,
+      "loss": 0.5864,
+      "step": 7167
+    },
+    {
+      "epoch": 1.985595567867036,
+      "grad_norm": 0.682807445526123,
+      "learning_rate": 3.7872482225427496e-06,
+      "loss": 0.6128,
+      "step": 7168
+    },
+    {
+      "epoch": 1.9858725761772853,
+      "grad_norm": 0.6814695596694946,
+      "learning_rate": 3.7869359254333555e-06,
+      "loss": 0.5586,
+      "step": 7169
+    },
+    {
+      "epoch": 1.9861495844875345,
+      "grad_norm": 0.7576179504394531,
+      "learning_rate": 3.7866236009990253e-06,
+      "loss": 0.6315,
+      "step": 7170
+    },
+    {
+      "epoch": 1.986426592797784,
+      "grad_norm": 0.6948550343513489,
+      "learning_rate": 3.78631124924639e-06,
+      "loss": 0.6757,
+      "step": 7171
+    },
+    {
+      "epoch": 1.9867036011080332,
+      "grad_norm": 0.6997304558753967,
+      "learning_rate": 3.7859988701820818e-06,
+      "loss": 0.5999,
+      "step": 7172
+    },
+    {
+      "epoch": 1.9869806094182825,
+      "grad_norm": 0.7092477083206177,
+      "learning_rate": 3.7856864638127346e-06,
+      "loss": 0.595,
+      "step": 7173
+    },
+    {
+      "epoch": 1.987257617728532,
+      "grad_norm": 0.6907538175582886,
+      "learning_rate": 3.7853740301449792e-06,
+      "loss": 0.6108,
+      "step": 7174
+    },
+    {
+      "epoch": 1.9875346260387812,
+      "grad_norm": 0.6745933294296265,
+      "learning_rate": 3.7850615691854513e-06,
+      "loss": 0.6013,
+      "step": 7175
+    },
+    {
+      "epoch": 1.9878116343490304,
+      "grad_norm": 0.6901183724403381,
+      "learning_rate": 3.7847490809407848e-06,
+      "loss": 0.6337,
+      "step": 7176
+    },
+    {
+      "epoch": 1.9880886426592799,
+      "grad_norm": 0.680350124835968,
+      "learning_rate": 3.784436565417614e-06,
+      "loss": 0.6413,
+      "step": 7177
+    },
+    {
+      "epoch": 1.988365650969529,
+      "grad_norm": 0.6782408952713013,
+      "learning_rate": 3.784124022622575e-06,
+      "loss": 0.615,
+      "step": 7178
+    },
+    {
+      "epoch": 1.9886426592797783,
+      "grad_norm": 0.7361845374107361,
+      "learning_rate": 3.783811452562304e-06,
+      "loss": 0.588,
+      "step": 7179
+    },
+    {
+      "epoch": 1.9889196675900278,
+      "grad_norm": 0.6791418790817261,
+      "learning_rate": 3.7834988552434365e-06,
+      "loss": 0.5931,
+      "step": 7180
+    },
+    {
+      "epoch": 1.989196675900277,
+      "grad_norm": 0.6929588913917542,
+      "learning_rate": 3.7831862306726114e-06,
+      "loss": 0.6364,
+      "step": 7181
+    },
+    {
+      "epoch": 1.9894736842105263,
+      "grad_norm": 0.6642131805419922,
+      "learning_rate": 3.7828735788564654e-06,
+      "loss": 0.621,
+      "step": 7182
+    },
+    {
+      "epoch": 1.9897506925207757,
+      "grad_norm": 0.6879885792732239,
+      "learning_rate": 3.7825608998016372e-06,
+      "loss": 0.5924,
+      "step": 7183
+    },
+    {
+      "epoch": 1.990027700831025,
+      "grad_norm": 0.6691209673881531,
+      "learning_rate": 3.782248193514766e-06,
+      "loss": 0.6207,
+      "step": 7184
+    },
+    {
+      "epoch": 1.9903047091412742,
+      "grad_norm": 0.7109494209289551,
+      "learning_rate": 3.7819354600024903e-06,
+      "loss": 0.6008,
+      "step": 7185
+    },
+    {
+      "epoch": 1.9905817174515237,
+      "grad_norm": 0.7236234545707703,
+      "learning_rate": 3.7816226992714522e-06,
+      "loss": 0.638,
+      "step": 7186
+    },
+    {
+      "epoch": 1.9908587257617727,
+      "grad_norm": 0.7347348928451538,
+      "learning_rate": 3.7813099113282907e-06,
+      "loss": 0.6286,
+      "step": 7187
+    },
+    {
+      "epoch": 1.9911357340720222,
+      "grad_norm": 0.7098999619483948,
+      "learning_rate": 3.7809970961796484e-06,
+      "loss": 0.6073,
+      "step": 7188
+    },
+    {
+      "epoch": 1.9914127423822716,
+      "grad_norm": 0.7165220379829407,
+      "learning_rate": 3.7806842538321654e-06,
+      "loss": 0.6491,
+      "step": 7189
+    },
+    {
+      "epoch": 1.9916897506925206,
+      "grad_norm": 0.6704548597335815,
+      "learning_rate": 3.780371384292486e-06,
+      "loss": 0.5961,
+      "step": 7190
+    },
+    {
+      "epoch": 1.99196675900277,
+      "grad_norm": 0.6767783761024475,
+      "learning_rate": 3.780058487567252e-06,
+      "loss": 0.6109,
+      "step": 7191
+    },
+    {
+      "epoch": 1.9922437673130196,
+      "grad_norm": 0.6728404760360718,
+      "learning_rate": 3.7797455636631076e-06,
+      "loss": 0.613,
+      "step": 7192
+    },
+    {
+      "epoch": 1.9925207756232686,
+      "grad_norm": 0.7355282306671143,
+      "learning_rate": 3.7794326125866964e-06,
+      "loss": 0.6164,
+      "step": 7193
+    },
+    {
+      "epoch": 1.992797783933518,
+      "grad_norm": 0.7722780704498291,
+      "learning_rate": 3.7791196343446634e-06,
+      "loss": 0.6603,
+      "step": 7194
+    },
+    {
+      "epoch": 1.9930747922437673,
+      "grad_norm": 0.6995906233787537,
+      "learning_rate": 3.778806628943655e-06,
+      "loss": 0.6094,
+      "step": 7195
+    },
+    {
+      "epoch": 1.9933518005540165,
+      "grad_norm": 0.7133497595787048,
+      "learning_rate": 3.7784935963903157e-06,
+      "loss": 0.5914,
+      "step": 7196
+    },
+    {
+      "epoch": 1.993628808864266,
+      "grad_norm": 0.6836190223693848,
+      "learning_rate": 3.7781805366912918e-06,
+      "loss": 0.6134,
+      "step": 7197
+    },
+    {
+      "epoch": 1.9939058171745152,
+      "grad_norm": 0.7232199311256409,
+      "learning_rate": 3.777867449853232e-06,
+      "loss": 0.611,
+      "step": 7198
+    },
+    {
+      "epoch": 1.9941828254847644,
+      "grad_norm": 0.7082121968269348,
+      "learning_rate": 3.777554335882782e-06,
+      "loss": 0.6699,
+      "step": 7199
+    },
+    {
+      "epoch": 1.994459833795014,
+      "grad_norm": 0.6821137070655823,
+      "learning_rate": 3.777241194786591e-06,
+      "loss": 0.5794,
+      "step": 7200
+    },
+    {
+      "epoch": 1.9947368421052631,
+      "grad_norm": 0.7654339075088501,
+      "learning_rate": 3.776928026571308e-06,
+      "loss": 0.6198,
+      "step": 7201
+    },
+    {
+      "epoch": 1.9950138504155124,
+      "grad_norm": 0.6920021176338196,
+      "learning_rate": 3.776614831243582e-06,
+      "loss": 0.6145,
+      "step": 7202
+    },
+    {
+      "epoch": 1.9952908587257618,
+      "grad_norm": 0.7104975581169128,
+      "learning_rate": 3.7763016088100634e-06,
+      "loss": 0.6675,
+      "step": 7203
+    },
+    {
+      "epoch": 1.995567867036011,
+      "grad_norm": 0.7255128026008606,
+      "learning_rate": 3.7759883592774018e-06,
+      "loss": 0.6338,
+      "step": 7204
+    },
+    {
+      "epoch": 1.9958448753462603,
+      "grad_norm": 0.7115368843078613,
+      "learning_rate": 3.77567508265225e-06,
+      "loss": 0.6179,
+      "step": 7205
+    },
+    {
+      "epoch": 1.9961218836565098,
+      "grad_norm": 0.6810135245323181,
+      "learning_rate": 3.775361778941257e-06,
+      "loss": 0.5555,
+      "step": 7206
+    },
+    {
+      "epoch": 1.996398891966759,
+      "grad_norm": 0.7034701108932495,
+      "learning_rate": 3.7750484481510775e-06,
+      "loss": 0.6148,
+      "step": 7207
+    },
+    {
+      "epoch": 1.9966759002770083,
+      "grad_norm": 0.722679078578949,
+      "learning_rate": 3.7747350902883627e-06,
+      "loss": 0.5851,
+      "step": 7208
+    },
+    {
+      "epoch": 1.9969529085872577,
+      "grad_norm": 0.686889111995697,
+      "learning_rate": 3.7744217053597663e-06,
+      "loss": 0.596,
+      "step": 7209
+    },
+    {
+      "epoch": 1.997229916897507,
+      "grad_norm": 0.6772050261497498,
+      "learning_rate": 3.774108293371943e-06,
+      "loss": 0.6377,
+      "step": 7210
+    },
+    {
+      "epoch": 1.9975069252077562,
+      "grad_norm": 0.7226324677467346,
+      "learning_rate": 3.7737948543315473e-06,
+      "loss": 0.648,
+      "step": 7211
+    },
+    {
+      "epoch": 1.9977839335180057,
+      "grad_norm": 0.6985737681388855,
+      "learning_rate": 3.7734813882452336e-06,
+      "loss": 0.6217,
+      "step": 7212
+    },
+    {
+      "epoch": 1.9980609418282549,
+      "grad_norm": 0.6721072793006897,
+      "learning_rate": 3.7731678951196576e-06,
+      "loss": 0.6291,
+      "step": 7213
+    },
+    {
+      "epoch": 1.9983379501385041,
+      "grad_norm": 0.6800287961959839,
+      "learning_rate": 3.7728543749614766e-06,
+      "loss": 0.6149,
+      "step": 7214
+    },
+    {
+      "epoch": 1.9986149584487536,
+      "grad_norm": 0.6427153944969177,
+      "learning_rate": 3.772540827777346e-06,
+      "loss": 0.5985,
+      "step": 7215
+    },
+    {
+      "epoch": 1.9988919667590028,
+      "grad_norm": 0.7035733461380005,
+      "learning_rate": 3.7722272535739247e-06,
+      "loss": 0.6356,
+      "step": 7216
+    },
+    {
+      "epoch": 1.999168975069252,
+      "grad_norm": 0.6934232711791992,
+      "learning_rate": 3.7719136523578687e-06,
+      "loss": 0.6313,
+      "step": 7217
+    },
+    {
+      "epoch": 1.9994459833795015,
+      "grad_norm": 0.6813242435455322,
+      "learning_rate": 3.771600024135839e-06,
+      "loss": 0.602,
+      "step": 7218
+    },
+    {
+      "epoch": 1.9997229916897505,
+      "grad_norm": 0.7163416147232056,
+      "learning_rate": 3.771286368914493e-06,
+      "loss": 0.6378,
+      "step": 7219
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7049331665039062,
+      "learning_rate": 3.7709726867004905e-06,
+      "loss": 0.6352,
+      "step": 7220
+    },
+    {
+      "epoch": 2.0002770083102495,
+      "grad_norm": 0.6854837536811829,
+      "learning_rate": 3.7706589775004926e-06,
+      "loss": 0.6146,
+      "step": 7221
+    },
+    {
+      "epoch": 2.0005540166204985,
+      "grad_norm": 0.7240393161773682,
+      "learning_rate": 3.7703452413211596e-06,
+      "loss": 0.621,
+      "step": 7222
+    },
+    {
+      "epoch": 2.000831024930748,
+      "grad_norm": 0.6688227653503418,
+      "learning_rate": 3.770031478169153e-06,
+      "loss": 0.6052,
+      "step": 7223
+    },
+    {
+      "epoch": 2.0011080332409974,
+      "grad_norm": 0.6882873773574829,
+      "learning_rate": 3.7697176880511353e-06,
+      "loss": 0.5827,
+      "step": 7224
+    },
+    {
+      "epoch": 2.0013850415512464,
+      "grad_norm": 0.6904376745223999,
+      "learning_rate": 3.7694038709737674e-06,
+      "loss": 0.5836,
+      "step": 7225
+    },
+    {
+      "epoch": 2.001662049861496,
+      "grad_norm": 0.7087603211402893,
+      "learning_rate": 3.769090026943714e-06,
+      "loss": 0.5383,
+      "step": 7226
+    },
+    {
+      "epoch": 2.0019390581717453,
+      "grad_norm": 0.6635582447052002,
+      "learning_rate": 3.7687761559676397e-06,
+      "loss": 0.5963,
+      "step": 7227
+    },
+    {
+      "epoch": 2.0022160664819943,
+      "grad_norm": 0.6937400698661804,
+      "learning_rate": 3.7684622580522057e-06,
+      "loss": 0.5906,
+      "step": 7228
+    },
+    {
+      "epoch": 2.002493074792244,
+      "grad_norm": 0.7310560345649719,
+      "learning_rate": 3.7681483332040793e-06,
+      "loss": 0.5661,
+      "step": 7229
+    },
+    {
+      "epoch": 2.0027700831024933,
+      "grad_norm": 0.7068434357643127,
+      "learning_rate": 3.767834381429926e-06,
+      "loss": 0.6078,
+      "step": 7230
+    },
+    {
+      "epoch": 2.0030470914127423,
+      "grad_norm": 0.7256509065628052,
+      "learning_rate": 3.7675204027364105e-06,
+      "loss": 0.6001,
+      "step": 7231
+    },
+    {
+      "epoch": 2.0033240997229917,
+      "grad_norm": 0.711618959903717,
+      "learning_rate": 3.7672063971302e-06,
+      "loss": 0.6469,
+      "step": 7232
+    },
+    {
+      "epoch": 2.003601108033241,
+      "grad_norm": 0.7369129657745361,
+      "learning_rate": 3.766892364617961e-06,
+      "loss": 0.5762,
+      "step": 7233
+    },
+    {
+      "epoch": 2.0038781163434902,
+      "grad_norm": 0.7153844237327576,
+      "learning_rate": 3.766578305206363e-06,
+      "loss": 0.5663,
+      "step": 7234
+    },
+    {
+      "epoch": 2.0041551246537397,
+      "grad_norm": 0.7051813006401062,
+      "learning_rate": 3.7662642189020726e-06,
+      "loss": 0.583,
+      "step": 7235
+    },
+    {
+      "epoch": 2.0044321329639887,
+      "grad_norm": 0.7093333601951599,
+      "learning_rate": 3.7659501057117586e-06,
+      "loss": 0.6124,
+      "step": 7236
+    },
+    {
+      "epoch": 2.004709141274238,
+      "grad_norm": 0.717320442199707,
+      "learning_rate": 3.7656359656420915e-06,
+      "loss": 0.5999,
+      "step": 7237
+    },
+    {
+      "epoch": 2.0049861495844876,
+      "grad_norm": 0.7107758522033691,
+      "learning_rate": 3.7653217986997405e-06,
+      "loss": 0.592,
+      "step": 7238
+    },
+    {
+      "epoch": 2.0052631578947366,
+      "grad_norm": 0.6741863489151001,
+      "learning_rate": 3.765007604891377e-06,
+      "loss": 0.5523,
+      "step": 7239
+    },
+    {
+      "epoch": 2.005540166204986,
+      "grad_norm": 0.706916868686676,
+      "learning_rate": 3.764693384223671e-06,
+      "loss": 0.5652,
+      "step": 7240
+    },
+    {
+      "epoch": 2.0058171745152356,
+      "grad_norm": 0.740626335144043,
+      "learning_rate": 3.7643791367032945e-06,
+      "loss": 0.6423,
+      "step": 7241
+    },
+    {
+      "epoch": 2.0060941828254846,
+      "grad_norm": 0.649993896484375,
+      "learning_rate": 3.764064862336921e-06,
+      "loss": 0.5362,
+      "step": 7242
+    },
+    {
+      "epoch": 2.006371191135734,
+      "grad_norm": 0.7229255437850952,
+      "learning_rate": 3.763750561131222e-06,
+      "loss": 0.5756,
+      "step": 7243
+    },
+    {
+      "epoch": 2.0066481994459835,
+      "grad_norm": 0.7251301407814026,
+      "learning_rate": 3.7634362330928707e-06,
+      "loss": 0.6221,
+      "step": 7244
+    },
+    {
+      "epoch": 2.0069252077562325,
+      "grad_norm": 0.7756488919258118,
+      "learning_rate": 3.7631218782285426e-06,
+      "loss": 0.6022,
+      "step": 7245
+    },
+    {
+      "epoch": 2.007202216066482,
+      "grad_norm": 0.7517617344856262,
+      "learning_rate": 3.7628074965449114e-06,
+      "loss": 0.5917,
+      "step": 7246
+    },
+    {
+      "epoch": 2.0074792243767314,
+      "grad_norm": 0.7082700729370117,
+      "learning_rate": 3.7624930880486513e-06,
+      "loss": 0.5977,
+      "step": 7247
+    },
+    {
+      "epoch": 2.0077562326869804,
+      "grad_norm": 0.7056994438171387,
+      "learning_rate": 3.76217865274644e-06,
+      "loss": 0.5966,
+      "step": 7248
+    },
+    {
+      "epoch": 2.00803324099723,
+      "grad_norm": 0.685123085975647,
+      "learning_rate": 3.7618641906449515e-06,
+      "loss": 0.6168,
+      "step": 7249
+    },
+    {
+      "epoch": 2.0083102493074794,
+      "grad_norm": 0.7444612979888916,
+      "learning_rate": 3.761549701750865e-06,
+      "loss": 0.604,
+      "step": 7250
+    },
+    {
+      "epoch": 2.0085872576177284,
+      "grad_norm": 0.7341189384460449,
+      "learning_rate": 3.761235186070855e-06,
+      "loss": 0.5992,
+      "step": 7251
+    },
+    {
+      "epoch": 2.008864265927978,
+      "grad_norm": 0.694669246673584,
+      "learning_rate": 3.760920643611603e-06,
+      "loss": 0.5974,
+      "step": 7252
+    },
+    {
+      "epoch": 2.0091412742382273,
+      "grad_norm": 0.7314677834510803,
+      "learning_rate": 3.7606060743797844e-06,
+      "loss": 0.6183,
+      "step": 7253
+    },
+    {
+      "epoch": 2.0094182825484763,
+      "grad_norm": 0.7707086205482483,
+      "learning_rate": 3.76029147838208e-06,
+      "loss": 0.5845,
+      "step": 7254
+    },
+    {
+      "epoch": 2.009695290858726,
+      "grad_norm": 0.7177945375442505,
+      "learning_rate": 3.759976855625169e-06,
+      "loss": 0.6028,
+      "step": 7255
+    },
+    {
+      "epoch": 2.0099722991689752,
+      "grad_norm": 0.6916733384132385,
+      "learning_rate": 3.759662206115732e-06,
+      "loss": 0.6221,
+      "step": 7256
+    },
+    {
+      "epoch": 2.0102493074792243,
+      "grad_norm": 0.6647084951400757,
+      "learning_rate": 3.7593475298604496e-06,
+      "loss": 0.583,
+      "step": 7257
+    },
+    {
+      "epoch": 2.0105263157894737,
+      "grad_norm": 0.7277607917785645,
+      "learning_rate": 3.7590328268660025e-06,
+      "loss": 0.6652,
+      "step": 7258
+    },
+    {
+      "epoch": 2.010803324099723,
+      "grad_norm": 0.7275409698486328,
+      "learning_rate": 3.7587180971390748e-06,
+      "loss": 0.6171,
+      "step": 7259
+    },
+    {
+      "epoch": 2.011080332409972,
+      "grad_norm": 0.7116976976394653,
+      "learning_rate": 3.758403340686345e-06,
+      "loss": 0.5722,
+      "step": 7260
+    },
+    {
+      "epoch": 2.0113573407202217,
+      "grad_norm": 0.6645132899284363,
+      "learning_rate": 3.758088557514501e-06,
+      "loss": 0.5321,
+      "step": 7261
+    },
+    {
+      "epoch": 2.011634349030471,
+      "grad_norm": 0.680085301399231,
+      "learning_rate": 3.757773747630223e-06,
+      "loss": 0.5907,
+      "step": 7262
+    },
+    {
+      "epoch": 2.01191135734072,
+      "grad_norm": 0.6877182126045227,
+      "learning_rate": 3.7574589110401965e-06,
+      "loss": 0.5829,
+      "step": 7263
+    },
+    {
+      "epoch": 2.0121883656509696,
+      "grad_norm": 0.6982957720756531,
+      "learning_rate": 3.757144047751106e-06,
+      "loss": 0.5982,
+      "step": 7264
+    },
+    {
+      "epoch": 2.012465373961219,
+      "grad_norm": 0.6935186982154846,
+      "learning_rate": 3.756829157769637e-06,
+      "loss": 0.6095,
+      "step": 7265
+    },
+    {
+      "epoch": 2.012742382271468,
+      "grad_norm": 0.708821713924408,
+      "learning_rate": 3.7565142411024756e-06,
+      "loss": 0.5746,
+      "step": 7266
+    },
+    {
+      "epoch": 2.0130193905817175,
+      "grad_norm": 0.7320842146873474,
+      "learning_rate": 3.7561992977563077e-06,
+      "loss": 0.6074,
+      "step": 7267
+    },
+    {
+      "epoch": 2.0132963988919665,
+      "grad_norm": 0.6891639232635498,
+      "learning_rate": 3.755884327737821e-06,
+      "loss": 0.6069,
+      "step": 7268
+    },
+    {
+      "epoch": 2.013573407202216,
+      "grad_norm": 0.6643139123916626,
+      "learning_rate": 3.7555693310537026e-06,
+      "loss": 0.6329,
+      "step": 7269
+    },
+    {
+      "epoch": 2.0138504155124655,
+      "grad_norm": 0.6847847700119019,
+      "learning_rate": 3.7552543077106414e-06,
+      "loss": 0.6138,
+      "step": 7270
+    },
+    {
+      "epoch": 2.0141274238227145,
+      "grad_norm": 0.7067223787307739,
+      "learning_rate": 3.7549392577153255e-06,
+      "loss": 0.6273,
+      "step": 7271
+    },
+    {
+      "epoch": 2.014404432132964,
+      "grad_norm": 0.7176226377487183,
+      "learning_rate": 3.7546241810744444e-06,
+      "loss": 0.5945,
+      "step": 7272
+    },
+    {
+      "epoch": 2.0146814404432134,
+      "grad_norm": 0.6766518950462341,
+      "learning_rate": 3.7543090777946876e-06,
+      "loss": 0.5691,
+      "step": 7273
+    },
+    {
+      "epoch": 2.0149584487534624,
+      "grad_norm": 0.7263760566711426,
+      "learning_rate": 3.7539939478827457e-06,
+      "loss": 0.6014,
+      "step": 7274
+    },
+    {
+      "epoch": 2.015235457063712,
+      "grad_norm": 0.6988743543624878,
+      "learning_rate": 3.753678791345311e-06,
+      "loss": 0.5582,
+      "step": 7275
+    },
+    {
+      "epoch": 2.0155124653739613,
+      "grad_norm": 0.7526625990867615,
+      "learning_rate": 3.7533636081890735e-06,
+      "loss": 0.6624,
+      "step": 7276
+    },
+    {
+      "epoch": 2.0157894736842104,
+      "grad_norm": 0.709251344203949,
+      "learning_rate": 3.7530483984207254e-06,
+      "loss": 0.5901,
+      "step": 7277
+    },
+    {
+      "epoch": 2.01606648199446,
+      "grad_norm": 0.6591882705688477,
+      "learning_rate": 3.752733162046961e-06,
+      "loss": 0.5809,
+      "step": 7278
+    },
+    {
+      "epoch": 2.0163434903047093,
+      "grad_norm": 0.6839492321014404,
+      "learning_rate": 3.7524178990744712e-06,
+      "loss": 0.613,
+      "step": 7279
+    },
+    {
+      "epoch": 2.0166204986149583,
+      "grad_norm": 0.7151807546615601,
+      "learning_rate": 3.752102609509952e-06,
+      "loss": 0.6245,
+      "step": 7280
+    },
+    {
+      "epoch": 2.0168975069252078,
+      "grad_norm": 0.7240042686462402,
+      "learning_rate": 3.7517872933600963e-06,
+      "loss": 0.6042,
+      "step": 7281
+    },
+    {
+      "epoch": 2.017174515235457,
+      "grad_norm": 0.7010844945907593,
+      "learning_rate": 3.7514719506315995e-06,
+      "loss": 0.6244,
+      "step": 7282
+    },
+    {
+      "epoch": 2.0174515235457062,
+      "grad_norm": 0.7382633686065674,
+      "learning_rate": 3.751156581331158e-06,
+      "loss": 0.5639,
+      "step": 7283
+    },
+    {
+      "epoch": 2.0177285318559557,
+      "grad_norm": 0.684970498085022,
+      "learning_rate": 3.750841185465467e-06,
+      "loss": 0.5623,
+      "step": 7284
+    },
+    {
+      "epoch": 2.018005540166205,
+      "grad_norm": 0.7206228971481323,
+      "learning_rate": 3.750525763041223e-06,
+      "loss": 0.5525,
+      "step": 7285
+    },
+    {
+      "epoch": 2.018282548476454,
+      "grad_norm": 0.6992058753967285,
+      "learning_rate": 3.750210314065123e-06,
+      "loss": 0.5933,
+      "step": 7286
+    },
+    {
+      "epoch": 2.0185595567867036,
+      "grad_norm": 0.6930119395256042,
+      "learning_rate": 3.749894838543866e-06,
+      "loss": 0.5794,
+      "step": 7287
+    },
+    {
+      "epoch": 2.018836565096953,
+      "grad_norm": 0.6776397228240967,
+      "learning_rate": 3.7495793364841497e-06,
+      "loss": 0.5888,
+      "step": 7288
+    },
+    {
+      "epoch": 2.019113573407202,
+      "grad_norm": 0.6972619891166687,
+      "learning_rate": 3.7492638078926736e-06,
+      "loss": 0.6004,
+      "step": 7289
+    },
+    {
+      "epoch": 2.0193905817174516,
+      "grad_norm": 0.6977049708366394,
+      "learning_rate": 3.7489482527761358e-06,
+      "loss": 0.6138,
+      "step": 7290
+    },
+    {
+      "epoch": 2.019667590027701,
+      "grad_norm": 0.7075570225715637,
+      "learning_rate": 3.7486326711412373e-06,
+      "loss": 0.5772,
+      "step": 7291
+    },
+    {
+      "epoch": 2.01994459833795,
+      "grad_norm": 0.7177691459655762,
+      "learning_rate": 3.7483170629946784e-06,
+      "loss": 0.654,
+      "step": 7292
+    },
+    {
+      "epoch": 2.0202216066481995,
+      "grad_norm": 0.7144123911857605,
+      "learning_rate": 3.74800142834316e-06,
+      "loss": 0.5372,
+      "step": 7293
+    },
+    {
+      "epoch": 2.020498614958449,
+      "grad_norm": 0.7332642078399658,
+      "learning_rate": 3.747685767193385e-06,
+      "loss": 0.6061,
+      "step": 7294
+    },
+    {
+      "epoch": 2.020775623268698,
+      "grad_norm": 0.7368162870407104,
+      "learning_rate": 3.7473700795520553e-06,
+      "loss": 0.588,
+      "step": 7295
+    },
+    {
+      "epoch": 2.0210526315789474,
+      "grad_norm": 0.6972573399543762,
+      "learning_rate": 3.747054365425872e-06,
+      "loss": 0.6108,
+      "step": 7296
+    },
+    {
+      "epoch": 2.021329639889197,
+      "grad_norm": 0.7078856229782104,
+      "learning_rate": 3.746738624821541e-06,
+      "loss": 0.5558,
+      "step": 7297
+    },
+    {
+      "epoch": 2.021606648199446,
+      "grad_norm": 0.7175845503807068,
+      "learning_rate": 3.746422857745765e-06,
+      "loss": 0.5894,
+      "step": 7298
+    },
+    {
+      "epoch": 2.0218836565096954,
+      "grad_norm": 0.6819410920143127,
+      "learning_rate": 3.7461070642052477e-06,
+      "loss": 0.5635,
+      "step": 7299
+    },
+    {
+      "epoch": 2.0221606648199444,
+      "grad_norm": 0.6979035139083862,
+      "learning_rate": 3.745791244206697e-06,
+      "loss": 0.5714,
+      "step": 7300
+    },
+    {
+      "epoch": 2.022437673130194,
+      "grad_norm": 0.7359611988067627,
+      "learning_rate": 3.7454753977568154e-06,
+      "loss": 0.6479,
+      "step": 7301
+    },
+    {
+      "epoch": 2.0227146814404433,
+      "grad_norm": 0.724439799785614,
+      "learning_rate": 3.745159524862311e-06,
+      "loss": 0.5861,
+      "step": 7302
+    },
+    {
+      "epoch": 2.0229916897506923,
+      "grad_norm": 0.7226695418357849,
+      "learning_rate": 3.74484362552989e-06,
+      "loss": 0.5672,
+      "step": 7303
+    },
+    {
+      "epoch": 2.023268698060942,
+      "grad_norm": 0.7167572379112244,
+      "learning_rate": 3.7445276997662604e-06,
+      "loss": 0.6204,
+      "step": 7304
+    },
+    {
+      "epoch": 2.0235457063711912,
+      "grad_norm": 0.7843630313873291,
+      "learning_rate": 3.744211747578129e-06,
+      "loss": 0.5762,
+      "step": 7305
+    },
+    {
+      "epoch": 2.0238227146814403,
+      "grad_norm": 0.7135272026062012,
+      "learning_rate": 3.7438957689722043e-06,
+      "loss": 0.561,
+      "step": 7306
+    },
+    {
+      "epoch": 2.0240997229916897,
+      "grad_norm": 0.711733877658844,
+      "learning_rate": 3.743579763955197e-06,
+      "loss": 0.6362,
+      "step": 7307
+    },
+    {
+      "epoch": 2.024376731301939,
+      "grad_norm": 0.7069997191429138,
+      "learning_rate": 3.743263732533815e-06,
+      "loss": 0.5785,
+      "step": 7308
+    },
+    {
+      "epoch": 2.024653739612188,
+      "grad_norm": 0.6933404803276062,
+      "learning_rate": 3.7429476747147685e-06,
+      "loss": 0.5499,
+      "step": 7309
+    },
+    {
+      "epoch": 2.0249307479224377,
+      "grad_norm": 0.7288474440574646,
+      "learning_rate": 3.7426315905047694e-06,
+      "loss": 0.6466,
+      "step": 7310
+    },
+    {
+      "epoch": 2.025207756232687,
+      "grad_norm": 0.7310758829116821,
+      "learning_rate": 3.7423154799105283e-06,
+      "loss": 0.6243,
+      "step": 7311
+    },
+    {
+      "epoch": 2.025484764542936,
+      "grad_norm": 0.6922908425331116,
+      "learning_rate": 3.7419993429387563e-06,
+      "loss": 0.5676,
+      "step": 7312
+    },
+    {
+      "epoch": 2.0257617728531856,
+      "grad_norm": 0.6958523988723755,
+      "learning_rate": 3.741683179596168e-06,
+      "loss": 0.6155,
+      "step": 7313
+    },
+    {
+      "epoch": 2.026038781163435,
+      "grad_norm": 0.7387740612030029,
+      "learning_rate": 3.7413669898894734e-06,
+      "loss": 0.6175,
+      "step": 7314
+    },
+    {
+      "epoch": 2.026315789473684,
+      "grad_norm": 0.7025557160377502,
+      "learning_rate": 3.7410507738253876e-06,
+      "loss": 0.6047,
+      "step": 7315
+    },
+    {
+      "epoch": 2.0265927977839335,
+      "grad_norm": 0.6605740189552307,
+      "learning_rate": 3.740734531410626e-06,
+      "loss": 0.5722,
+      "step": 7316
+    },
+    {
+      "epoch": 2.026869806094183,
+      "grad_norm": 0.6882765889167786,
+      "learning_rate": 3.7404182626519005e-06,
+      "loss": 0.6053,
+      "step": 7317
+    },
+    {
+      "epoch": 2.027146814404432,
+      "grad_norm": 0.6996808648109436,
+      "learning_rate": 3.7401019675559277e-06,
+      "loss": 0.5809,
+      "step": 7318
+    },
+    {
+      "epoch": 2.0274238227146815,
+      "grad_norm": 0.7477633357048035,
+      "learning_rate": 3.7397856461294237e-06,
+      "loss": 0.6255,
+      "step": 7319
+    },
+    {
+      "epoch": 2.027700831024931,
+      "grad_norm": 0.6916236281394958,
+      "learning_rate": 3.7394692983791037e-06,
+      "loss": 0.587,
+      "step": 7320
+    },
+    {
+      "epoch": 2.02797783933518,
+      "grad_norm": 0.7061457633972168,
+      "learning_rate": 3.7391529243116857e-06,
+      "loss": 0.5835,
+      "step": 7321
+    },
+    {
+      "epoch": 2.0282548476454294,
+      "grad_norm": 0.7239747047424316,
+      "learning_rate": 3.7388365239338866e-06,
+      "loss": 0.5874,
+      "step": 7322
+    },
+    {
+      "epoch": 2.028531855955679,
+      "grad_norm": 0.7012773752212524,
+      "learning_rate": 3.738520097252425e-06,
+      "loss": 0.6072,
+      "step": 7323
+    },
+    {
+      "epoch": 2.028808864265928,
+      "grad_norm": 0.6934051513671875,
+      "learning_rate": 3.7382036442740183e-06,
+      "loss": 0.6076,
+      "step": 7324
+    },
+    {
+      "epoch": 2.0290858725761773,
+      "grad_norm": 0.6960504651069641,
+      "learning_rate": 3.737887165005386e-06,
+      "loss": 0.5879,
+      "step": 7325
+    },
+    {
+      "epoch": 2.029362880886427,
+      "grad_norm": 0.777347981929779,
+      "learning_rate": 3.737570659453249e-06,
+      "loss": 0.594,
+      "step": 7326
+    },
+    {
+      "epoch": 2.029639889196676,
+      "grad_norm": 0.7199022173881531,
+      "learning_rate": 3.737254127624325e-06,
+      "loss": 0.5921,
+      "step": 7327
+    },
+    {
+      "epoch": 2.0299168975069253,
+      "grad_norm": 0.7547945976257324,
+      "learning_rate": 3.7369375695253374e-06,
+      "loss": 0.5838,
+      "step": 7328
+    },
+    {
+      "epoch": 2.0301939058171747,
+      "grad_norm": 0.754098117351532,
+      "learning_rate": 3.7366209851630057e-06,
+      "loss": 0.5948,
+      "step": 7329
+    },
+    {
+      "epoch": 2.0304709141274238,
+      "grad_norm": 0.7216700911521912,
+      "learning_rate": 3.736304374544053e-06,
+      "loss": 0.5521,
+      "step": 7330
+    },
+    {
+      "epoch": 2.030747922437673,
+      "grad_norm": 0.7701409459114075,
+      "learning_rate": 3.7359877376752006e-06,
+      "loss": 0.5908,
+      "step": 7331
+    },
+    {
+      "epoch": 2.0310249307479222,
+      "grad_norm": 0.7693707942962646,
+      "learning_rate": 3.7356710745631735e-06,
+      "loss": 0.5769,
+      "step": 7332
+    },
+    {
+      "epoch": 2.0313019390581717,
+      "grad_norm": 0.7225122451782227,
+      "learning_rate": 3.7353543852146926e-06,
+      "loss": 0.6152,
+      "step": 7333
+    },
+    {
+      "epoch": 2.031578947368421,
+      "grad_norm": 0.7436615824699402,
+      "learning_rate": 3.7350376696364843e-06,
+      "loss": 0.619,
+      "step": 7334
+    },
+    {
+      "epoch": 2.03185595567867,
+      "grad_norm": 0.735950231552124,
+      "learning_rate": 3.734720927835272e-06,
+      "loss": 0.6075,
+      "step": 7335
+    },
+    {
+      "epoch": 2.0321329639889196,
+      "grad_norm": 0.7073866724967957,
+      "learning_rate": 3.7344041598177812e-06,
+      "loss": 0.6224,
+      "step": 7336
+    },
+    {
+      "epoch": 2.032409972299169,
+      "grad_norm": 0.6516287326812744,
+      "learning_rate": 3.734087365590737e-06,
+      "loss": 0.5226,
+      "step": 7337
+    },
+    {
+      "epoch": 2.032686980609418,
+      "grad_norm": 0.7546796202659607,
+      "learning_rate": 3.7337705451608676e-06,
+      "loss": 0.6169,
+      "step": 7338
+    },
+    {
+      "epoch": 2.0329639889196676,
+      "grad_norm": 0.7578698396682739,
+      "learning_rate": 3.7334536985348986e-06,
+      "loss": 0.587,
+      "step": 7339
+    },
+    {
+      "epoch": 2.033240997229917,
+      "grad_norm": 0.6956138014793396,
+      "learning_rate": 3.733136825719557e-06,
+      "loss": 0.6144,
+      "step": 7340
+    },
+    {
+      "epoch": 2.033518005540166,
+      "grad_norm": 0.716972291469574,
+      "learning_rate": 3.732819926721572e-06,
+      "loss": 0.5942,
+      "step": 7341
+    },
+    {
+      "epoch": 2.0337950138504155,
+      "grad_norm": 0.7351236939430237,
+      "learning_rate": 3.732503001547672e-06,
+      "loss": 0.5588,
+      "step": 7342
+    },
+    {
+      "epoch": 2.034072022160665,
+      "grad_norm": 0.6928535103797913,
+      "learning_rate": 3.732186050204585e-06,
+      "loss": 0.613,
+      "step": 7343
+    },
+    {
+      "epoch": 2.034349030470914,
+      "grad_norm": 0.7150267362594604,
+      "learning_rate": 3.731869072699042e-06,
+      "loss": 0.6011,
+      "step": 7344
+    },
+    {
+      "epoch": 2.0346260387811634,
+      "grad_norm": 0.7005348205566406,
+      "learning_rate": 3.731552069037774e-06,
+      "loss": 0.5847,
+      "step": 7345
+    },
+    {
+      "epoch": 2.034903047091413,
+      "grad_norm": 0.6965834498405457,
+      "learning_rate": 3.731235039227508e-06,
+      "loss": 0.5559,
+      "step": 7346
+    },
+    {
+      "epoch": 2.035180055401662,
+      "grad_norm": 0.7243964076042175,
+      "learning_rate": 3.7309179832749798e-06,
+      "loss": 0.6314,
+      "step": 7347
+    },
+    {
+      "epoch": 2.0354570637119114,
+      "grad_norm": 0.7027679681777954,
+      "learning_rate": 3.7306009011869187e-06,
+      "loss": 0.5905,
+      "step": 7348
+    },
+    {
+      "epoch": 2.035734072022161,
+      "grad_norm": 0.7306333780288696,
+      "learning_rate": 3.730283792970058e-06,
+      "loss": 0.5806,
+      "step": 7349
+    },
+    {
+      "epoch": 2.03601108033241,
+      "grad_norm": 0.7483391761779785,
+      "learning_rate": 3.72996665863113e-06,
+      "loss": 0.6087,
+      "step": 7350
+    },
+    {
+      "epoch": 2.0362880886426593,
+      "grad_norm": 0.7386797070503235,
+      "learning_rate": 3.7296494981768694e-06,
+      "loss": 0.6034,
+      "step": 7351
+    },
+    {
+      "epoch": 2.0365650969529088,
+      "grad_norm": 0.7524106502532959,
+      "learning_rate": 3.72933231161401e-06,
+      "loss": 0.5895,
+      "step": 7352
+    },
+    {
+      "epoch": 2.036842105263158,
+      "grad_norm": 0.7452194690704346,
+      "learning_rate": 3.7290150989492856e-06,
+      "loss": 0.5701,
+      "step": 7353
+    },
+    {
+      "epoch": 2.0371191135734072,
+      "grad_norm": 0.724443793296814,
+      "learning_rate": 3.728697860189433e-06,
+      "loss": 0.6122,
+      "step": 7354
+    },
+    {
+      "epoch": 2.0373961218836567,
+      "grad_norm": 0.737069845199585,
+      "learning_rate": 3.7283805953411863e-06,
+      "loss": 0.5866,
+      "step": 7355
+    },
+    {
+      "epoch": 2.0376731301939057,
+      "grad_norm": 0.7455108761787415,
+      "learning_rate": 3.7280633044112826e-06,
+      "loss": 0.6072,
+      "step": 7356
+    },
+    {
+      "epoch": 2.037950138504155,
+      "grad_norm": 0.7549130916595459,
+      "learning_rate": 3.7277459874064593e-06,
+      "loss": 0.5652,
+      "step": 7357
+    },
+    {
+      "epoch": 2.0382271468144046,
+      "grad_norm": 0.7499793767929077,
+      "learning_rate": 3.727428644333453e-06,
+      "loss": 0.5898,
+      "step": 7358
+    },
+    {
+      "epoch": 2.0385041551246537,
+      "grad_norm": 0.7160870432853699,
+      "learning_rate": 3.7271112751990017e-06,
+      "loss": 0.5952,
+      "step": 7359
+    },
+    {
+      "epoch": 2.038781163434903,
+      "grad_norm": 0.7001029253005981,
+      "learning_rate": 3.7267938800098454e-06,
+      "loss": 0.5834,
+      "step": 7360
+    },
+    {
+      "epoch": 2.039058171745152,
+      "grad_norm": 0.7471390962600708,
+      "learning_rate": 3.7264764587727217e-06,
+      "loss": 0.6395,
+      "step": 7361
+    },
+    {
+      "epoch": 2.0393351800554016,
+      "grad_norm": 0.7129092812538147,
+      "learning_rate": 3.72615901149437e-06,
+      "loss": 0.6213,
+      "step": 7362
+    },
+    {
+      "epoch": 2.039612188365651,
+      "grad_norm": 0.7496341466903687,
+      "learning_rate": 3.7258415381815315e-06,
+      "loss": 0.6129,
+      "step": 7363
+    },
+    {
+      "epoch": 2.0398891966759,
+      "grad_norm": 0.7098085284233093,
+      "learning_rate": 3.7255240388409473e-06,
+      "loss": 0.5844,
+      "step": 7364
+    },
+    {
+      "epoch": 2.0401662049861495,
+      "grad_norm": 0.7387219071388245,
+      "learning_rate": 3.725206513479358e-06,
+      "loss": 0.6032,
+      "step": 7365
+    },
+    {
+      "epoch": 2.040443213296399,
+      "grad_norm": 0.712770402431488,
+      "learning_rate": 3.724888962103505e-06,
+      "loss": 0.5806,
+      "step": 7366
+    },
+    {
+      "epoch": 2.040720221606648,
+      "grad_norm": 0.7392038106918335,
+      "learning_rate": 3.724571384720132e-06,
+      "loss": 0.6242,
+      "step": 7367
+    },
+    {
+      "epoch": 2.0409972299168975,
+      "grad_norm": 0.6714391112327576,
+      "learning_rate": 3.7242537813359803e-06,
+      "loss": 0.6208,
+      "step": 7368
+    },
+    {
+      "epoch": 2.041274238227147,
+      "grad_norm": 0.6978964805603027,
+      "learning_rate": 3.7239361519577956e-06,
+      "loss": 0.5974,
+      "step": 7369
+    },
+    {
+      "epoch": 2.041551246537396,
+      "grad_norm": 0.7669302821159363,
+      "learning_rate": 3.72361849659232e-06,
+      "loss": 0.6248,
+      "step": 7370
+    },
+    {
+      "epoch": 2.0418282548476454,
+      "grad_norm": 0.6692224740982056,
+      "learning_rate": 3.723300815246299e-06,
+      "loss": 0.586,
+      "step": 7371
+    },
+    {
+      "epoch": 2.042105263157895,
+      "grad_norm": 0.6983182430267334,
+      "learning_rate": 3.722983107926478e-06,
+      "loss": 0.5746,
+      "step": 7372
+    },
+    {
+      "epoch": 2.042382271468144,
+      "grad_norm": 0.6932432651519775,
+      "learning_rate": 3.7226653746396025e-06,
+      "loss": 0.5787,
+      "step": 7373
+    },
+    {
+      "epoch": 2.0426592797783933,
+      "grad_norm": 0.7647563815116882,
+      "learning_rate": 3.722347615392419e-06,
+      "loss": 0.6347,
+      "step": 7374
+    },
+    {
+      "epoch": 2.042936288088643,
+      "grad_norm": 0.7473055720329285,
+      "learning_rate": 3.7220298301916735e-06,
+      "loss": 0.6685,
+      "step": 7375
+    },
+    {
+      "epoch": 2.043213296398892,
+      "grad_norm": 0.7167715430259705,
+      "learning_rate": 3.7217120190441134e-06,
+      "loss": 0.6559,
+      "step": 7376
+    },
+    {
+      "epoch": 2.0434903047091413,
+      "grad_norm": 0.7501154541969299,
+      "learning_rate": 3.721394181956489e-06,
+      "loss": 0.5932,
+      "step": 7377
+    },
+    {
+      "epoch": 2.0437673130193907,
+      "grad_norm": 0.729622483253479,
+      "learning_rate": 3.7210763189355455e-06,
+      "loss": 0.5582,
+      "step": 7378
+    },
+    {
+      "epoch": 2.0440443213296398,
+      "grad_norm": 0.7165427803993225,
+      "learning_rate": 3.7207584299880337e-06,
+      "loss": 0.6243,
+      "step": 7379
+    },
+    {
+      "epoch": 2.044321329639889,
+      "grad_norm": 0.6978292465209961,
+      "learning_rate": 3.7204405151207036e-06,
+      "loss": 0.6019,
+      "step": 7380
+    },
+    {
+      "epoch": 2.0445983379501387,
+      "grad_norm": 0.7220966219902039,
+      "learning_rate": 3.720122574340304e-06,
+      "loss": 0.6171,
+      "step": 7381
+    },
+    {
+      "epoch": 2.0448753462603877,
+      "grad_norm": 0.7531928420066833,
+      "learning_rate": 3.7198046076535865e-06,
+      "loss": 0.5751,
+      "step": 7382
+    },
+    {
+      "epoch": 2.045152354570637,
+      "grad_norm": 0.6982653737068176,
+      "learning_rate": 3.7194866150673027e-06,
+      "loss": 0.5655,
+      "step": 7383
+    },
+    {
+      "epoch": 2.0454293628808866,
+      "grad_norm": 0.6789435148239136,
+      "learning_rate": 3.7191685965882033e-06,
+      "loss": 0.5926,
+      "step": 7384
+    },
+    {
+      "epoch": 2.0457063711911356,
+      "grad_norm": 0.7709596753120422,
+      "learning_rate": 3.718850552223041e-06,
+      "loss": 0.6178,
+      "step": 7385
+    },
+    {
+      "epoch": 2.045983379501385,
+      "grad_norm": 0.71182781457901,
+      "learning_rate": 3.71853248197857e-06,
+      "loss": 0.6004,
+      "step": 7386
+    },
+    {
+      "epoch": 2.0462603878116346,
+      "grad_norm": 0.6998679637908936,
+      "learning_rate": 3.7182143858615416e-06,
+      "loss": 0.5959,
+      "step": 7387
+    },
+    {
+      "epoch": 2.0465373961218836,
+      "grad_norm": 0.721081554889679,
+      "learning_rate": 3.717896263878711e-06,
+      "loss": 0.5927,
+      "step": 7388
+    },
+    {
+      "epoch": 2.046814404432133,
+      "grad_norm": 0.7610851526260376,
+      "learning_rate": 3.7175781160368324e-06,
+      "loss": 0.6316,
+      "step": 7389
+    },
+    {
+      "epoch": 2.0470914127423825,
+      "grad_norm": 0.7433952689170837,
+      "learning_rate": 3.7172599423426614e-06,
+      "loss": 0.6325,
+      "step": 7390
+    },
+    {
+      "epoch": 2.0473684210526315,
+      "grad_norm": 0.7426512837409973,
+      "learning_rate": 3.7169417428029525e-06,
+      "loss": 0.6334,
+      "step": 7391
+    },
+    {
+      "epoch": 2.047645429362881,
+      "grad_norm": 0.7442209124565125,
+      "learning_rate": 3.7166235174244637e-06,
+      "loss": 0.5711,
+      "step": 7392
+    },
+    {
+      "epoch": 2.04792243767313,
+      "grad_norm": 0.7405110597610474,
+      "learning_rate": 3.7163052662139497e-06,
+      "loss": 0.5779,
+      "step": 7393
+    },
+    {
+      "epoch": 2.0481994459833794,
+      "grad_norm": 0.7343849539756775,
+      "learning_rate": 3.7159869891781698e-06,
+      "loss": 0.6144,
+      "step": 7394
+    },
+    {
+      "epoch": 2.048476454293629,
+      "grad_norm": 0.6877952814102173,
+      "learning_rate": 3.71566868632388e-06,
+      "loss": 0.5983,
+      "step": 7395
+    },
+    {
+      "epoch": 2.048753462603878,
+      "grad_norm": 0.7006337642669678,
+      "learning_rate": 3.71535035765784e-06,
+      "loss": 0.6,
+      "step": 7396
+    },
+    {
+      "epoch": 2.0490304709141274,
+      "grad_norm": 0.7534791827201843,
+      "learning_rate": 3.715032003186808e-06,
+      "loss": 0.6124,
+      "step": 7397
+    },
+    {
+      "epoch": 2.049307479224377,
+      "grad_norm": 0.7348729968070984,
+      "learning_rate": 3.7147136229175436e-06,
+      "loss": 0.6252,
+      "step": 7398
+    },
+    {
+      "epoch": 2.049584487534626,
+      "grad_norm": 0.6931565999984741,
+      "learning_rate": 3.7143952168568077e-06,
+      "loss": 0.5888,
+      "step": 7399
+    },
+    {
+      "epoch": 2.0498614958448753,
+      "grad_norm": 0.7033185362815857,
+      "learning_rate": 3.7140767850113592e-06,
+      "loss": 0.5964,
+      "step": 7400
+    },
+    {
+      "epoch": 2.0501385041551248,
+      "grad_norm": 0.7246971130371094,
+      "learning_rate": 3.7137583273879606e-06,
+      "loss": 0.6151,
+      "step": 7401
+    },
+    {
+      "epoch": 2.050415512465374,
+      "grad_norm": 0.7070119380950928,
+      "learning_rate": 3.713439843993373e-06,
+      "loss": 0.5908,
+      "step": 7402
+    },
+    {
+      "epoch": 2.0506925207756233,
+      "grad_norm": 0.7307522296905518,
+      "learning_rate": 3.7131213348343585e-06,
+      "loss": 0.624,
+      "step": 7403
+    },
+    {
+      "epoch": 2.0509695290858727,
+      "grad_norm": 0.7054348587989807,
+      "learning_rate": 3.71280279991768e-06,
+      "loss": 0.5466,
+      "step": 7404
+    },
+    {
+      "epoch": 2.0512465373961217,
+      "grad_norm": 0.7453206181526184,
+      "learning_rate": 3.7124842392501014e-06,
+      "loss": 0.6172,
+      "step": 7405
+    },
+    {
+      "epoch": 2.051523545706371,
+      "grad_norm": 0.7176256775856018,
+      "learning_rate": 3.7121656528383866e-06,
+      "loss": 0.6353,
+      "step": 7406
+    },
+    {
+      "epoch": 2.0518005540166206,
+      "grad_norm": 0.7283518314361572,
+      "learning_rate": 3.7118470406892986e-06,
+      "loss": 0.6067,
+      "step": 7407
+    },
+    {
+      "epoch": 2.0520775623268697,
+      "grad_norm": 0.7188484072685242,
+      "learning_rate": 3.7115284028096033e-06,
+      "loss": 0.5842,
+      "step": 7408
+    },
+    {
+      "epoch": 2.052354570637119,
+      "grad_norm": 0.7040556073188782,
+      "learning_rate": 3.7112097392060658e-06,
+      "loss": 0.6007,
+      "step": 7409
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.700164794921875,
+      "learning_rate": 3.7108910498854526e-06,
+      "loss": 0.5744,
+      "step": 7410
+    },
+    {
+      "epoch": 2.0529085872576176,
+      "grad_norm": 0.7129940986633301,
+      "learning_rate": 3.7105723348545308e-06,
+      "loss": 0.5953,
+      "step": 7411
+    },
+    {
+      "epoch": 2.053185595567867,
+      "grad_norm": 0.7359551191329956,
+      "learning_rate": 3.710253594120067e-06,
+      "loss": 0.5892,
+      "step": 7412
+    },
+    {
+      "epoch": 2.0534626038781165,
+      "grad_norm": 0.7358061671257019,
+      "learning_rate": 3.709934827688827e-06,
+      "loss": 0.5844,
+      "step": 7413
+    },
+    {
+      "epoch": 2.0537396121883655,
+      "grad_norm": 0.6994933485984802,
+      "learning_rate": 3.7096160355675827e-06,
+      "loss": 0.5767,
+      "step": 7414
+    },
+    {
+      "epoch": 2.054016620498615,
+      "grad_norm": 0.734293520450592,
+      "learning_rate": 3.7092972177630998e-06,
+      "loss": 0.6198,
+      "step": 7415
+    },
+    {
+      "epoch": 2.0542936288088645,
+      "grad_norm": 0.7511576414108276,
+      "learning_rate": 3.7089783742821493e-06,
+      "loss": 0.6362,
+      "step": 7416
+    },
+    {
+      "epoch": 2.0545706371191135,
+      "grad_norm": 0.7923803925514221,
+      "learning_rate": 3.7086595051315e-06,
+      "loss": 0.5829,
+      "step": 7417
+    },
+    {
+      "epoch": 2.054847645429363,
+      "grad_norm": 0.7466977834701538,
+      "learning_rate": 3.7083406103179235e-06,
+      "loss": 0.621,
+      "step": 7418
+    },
+    {
+      "epoch": 2.0551246537396124,
+      "grad_norm": 0.7417007088661194,
+      "learning_rate": 3.7080216898481897e-06,
+      "loss": 0.6103,
+      "step": 7419
+    },
+    {
+      "epoch": 2.0554016620498614,
+      "grad_norm": 0.69034743309021,
+      "learning_rate": 3.70770274372907e-06,
+      "loss": 0.6059,
+      "step": 7420
+    },
+    {
+      "epoch": 2.055678670360111,
+      "grad_norm": 0.6910015344619751,
+      "learning_rate": 3.707383771967338e-06,
+      "loss": 0.5914,
+      "step": 7421
+    },
+    {
+      "epoch": 2.05595567867036,
+      "grad_norm": 0.7586956024169922,
+      "learning_rate": 3.7070647745697647e-06,
+      "loss": 0.6004,
+      "step": 7422
+    },
+    {
+      "epoch": 2.0562326869806093,
+      "grad_norm": 0.7114026546478271,
+      "learning_rate": 3.706745751543123e-06,
+      "loss": 0.6295,
+      "step": 7423
+    },
+    {
+      "epoch": 2.056509695290859,
+      "grad_norm": 0.7049426436424255,
+      "learning_rate": 3.706426702894188e-06,
+      "loss": 0.6064,
+      "step": 7424
+    },
+    {
+      "epoch": 2.056786703601108,
+      "grad_norm": 0.7011547684669495,
+      "learning_rate": 3.7061076286297335e-06,
+      "loss": 0.5655,
+      "step": 7425
+    },
+    {
+      "epoch": 2.0570637119113573,
+      "grad_norm": 0.7338706254959106,
+      "learning_rate": 3.705788528756533e-06,
+      "loss": 0.5801,
+      "step": 7426
+    },
+    {
+      "epoch": 2.0573407202216067,
+      "grad_norm": 0.7272680401802063,
+      "learning_rate": 3.7054694032813644e-06,
+      "loss": 0.5601,
+      "step": 7427
+    },
+    {
+      "epoch": 2.0576177285318558,
+      "grad_norm": 0.744717001914978,
+      "learning_rate": 3.7051502522110006e-06,
+      "loss": 0.5771,
+      "step": 7428
+    },
+    {
+      "epoch": 2.057894736842105,
+      "grad_norm": 0.7375504970550537,
+      "learning_rate": 3.70483107555222e-06,
+      "loss": 0.6157,
+      "step": 7429
+    },
+    {
+      "epoch": 2.0581717451523547,
+      "grad_norm": 0.7337304353713989,
+      "learning_rate": 3.704511873311798e-06,
+      "loss": 0.5899,
+      "step": 7430
+    },
+    {
+      "epoch": 2.0584487534626037,
+      "grad_norm": 0.720333456993103,
+      "learning_rate": 3.7041926454965137e-06,
+      "loss": 0.628,
+      "step": 7431
+    },
+    {
+      "epoch": 2.058725761772853,
+      "grad_norm": 0.6853072643280029,
+      "learning_rate": 3.7038733921131444e-06,
+      "loss": 0.5362,
+      "step": 7432
+    },
+    {
+      "epoch": 2.0590027700831026,
+      "grad_norm": 0.6991992592811584,
+      "learning_rate": 3.7035541131684684e-06,
+      "loss": 0.5874,
+      "step": 7433
+    },
+    {
+      "epoch": 2.0592797783933516,
+      "grad_norm": 0.7000507712364197,
+      "learning_rate": 3.703234808669265e-06,
+      "loss": 0.6198,
+      "step": 7434
+    },
+    {
+      "epoch": 2.059556786703601,
+      "grad_norm": 0.7110688090324402,
+      "learning_rate": 3.7029154786223144e-06,
+      "loss": 0.609,
+      "step": 7435
+    },
+    {
+      "epoch": 2.0598337950138506,
+      "grad_norm": 0.7329934239387512,
+      "learning_rate": 3.7025961230343954e-06,
+      "loss": 0.6258,
+      "step": 7436
+    },
+    {
+      "epoch": 2.0601108033240996,
+      "grad_norm": 0.741564929485321,
+      "learning_rate": 3.70227674191229e-06,
+      "loss": 0.6174,
+      "step": 7437
+    },
+    {
+      "epoch": 2.060387811634349,
+      "grad_norm": 0.7025269865989685,
+      "learning_rate": 3.701957335262779e-06,
+      "loss": 0.5482,
+      "step": 7438
+    },
+    {
+      "epoch": 2.0606648199445985,
+      "grad_norm": 0.7336528301239014,
+      "learning_rate": 3.701637903092644e-06,
+      "loss": 0.6085,
+      "step": 7439
+    },
+    {
+      "epoch": 2.0609418282548475,
+      "grad_norm": 0.6992329955101013,
+      "learning_rate": 3.7013184454086688e-06,
+      "loss": 0.5614,
+      "step": 7440
+    },
+    {
+      "epoch": 2.061218836565097,
+      "grad_norm": 0.7023910284042358,
+      "learning_rate": 3.700998962217634e-06,
+      "loss": 0.5856,
+      "step": 7441
+    },
+    {
+      "epoch": 2.0614958448753464,
+      "grad_norm": 0.727231502532959,
+      "learning_rate": 3.7006794535263246e-06,
+      "loss": 0.6333,
+      "step": 7442
+    },
+    {
+      "epoch": 2.0617728531855954,
+      "grad_norm": 0.7130199670791626,
+      "learning_rate": 3.7003599193415247e-06,
+      "loss": 0.5509,
+      "step": 7443
+    },
+    {
+      "epoch": 2.062049861495845,
+      "grad_norm": 0.6830217242240906,
+      "learning_rate": 3.700040359670018e-06,
+      "loss": 0.5795,
+      "step": 7444
+    },
+    {
+      "epoch": 2.0623268698060944,
+      "grad_norm": 0.7464885115623474,
+      "learning_rate": 3.6997207745185892e-06,
+      "loss": 0.63,
+      "step": 7445
+    },
+    {
+      "epoch": 2.0626038781163434,
+      "grad_norm": 0.749214768409729,
+      "learning_rate": 3.6994011638940253e-06,
+      "loss": 0.6179,
+      "step": 7446
+    },
+    {
+      "epoch": 2.062880886426593,
+      "grad_norm": 0.7185910940170288,
+      "learning_rate": 3.6990815278031116e-06,
+      "loss": 0.5941,
+      "step": 7447
+    },
+    {
+      "epoch": 2.0631578947368423,
+      "grad_norm": 0.7367299795150757,
+      "learning_rate": 3.698761866252635e-06,
+      "loss": 0.6337,
+      "step": 7448
+    },
+    {
+      "epoch": 2.0634349030470913,
+      "grad_norm": 0.7037851810455322,
+      "learning_rate": 3.6984421792493824e-06,
+      "loss": 0.5475,
+      "step": 7449
+    },
+    {
+      "epoch": 2.063711911357341,
+      "grad_norm": 0.7410505414009094,
+      "learning_rate": 3.6981224668001427e-06,
+      "loss": 0.5965,
+      "step": 7450
+    },
+    {
+      "epoch": 2.0639889196675902,
+      "grad_norm": 0.702735185623169,
+      "learning_rate": 3.6978027289117024e-06,
+      "loss": 0.561,
+      "step": 7451
+    },
+    {
+      "epoch": 2.0642659279778393,
+      "grad_norm": 0.7526459693908691,
+      "learning_rate": 3.6974829655908518e-06,
+      "loss": 0.6003,
+      "step": 7452
+    },
+    {
+      "epoch": 2.0645429362880887,
+      "grad_norm": 0.6807567477226257,
+      "learning_rate": 3.69716317684438e-06,
+      "loss": 0.5654,
+      "step": 7453
+    },
+    {
+      "epoch": 2.064819944598338,
+      "grad_norm": 0.7194676995277405,
+      "learning_rate": 3.6968433626790755e-06,
+      "loss": 0.6455,
+      "step": 7454
+    },
+    {
+      "epoch": 2.065096952908587,
+      "grad_norm": 0.7026492953300476,
+      "learning_rate": 3.6965235231017316e-06,
+      "loss": 0.6082,
+      "step": 7455
+    },
+    {
+      "epoch": 2.0653739612188367,
+      "grad_norm": 0.6919288635253906,
+      "learning_rate": 3.6962036581191367e-06,
+      "loss": 0.6148,
+      "step": 7456
+    },
+    {
+      "epoch": 2.0656509695290857,
+      "grad_norm": 0.7587344646453857,
+      "learning_rate": 3.6958837677380843e-06,
+      "loss": 0.6202,
+      "step": 7457
+    },
+    {
+      "epoch": 2.065927977839335,
+      "grad_norm": 0.7068833708763123,
+      "learning_rate": 3.695563851965364e-06,
+      "loss": 0.6204,
+      "step": 7458
+    },
+    {
+      "epoch": 2.0662049861495846,
+      "grad_norm": 0.771920382976532,
+      "learning_rate": 3.695243910807772e-06,
+      "loss": 0.6288,
+      "step": 7459
+    },
+    {
+      "epoch": 2.0664819944598336,
+      "grad_norm": 0.6968458294868469,
+      "learning_rate": 3.694923944272098e-06,
+      "loss": 0.5845,
+      "step": 7460
+    },
+    {
+      "epoch": 2.066759002770083,
+      "grad_norm": 0.6951239705085754,
+      "learning_rate": 3.694603952365138e-06,
+      "loss": 0.6345,
+      "step": 7461
+    },
+    {
+      "epoch": 2.0670360110803325,
+      "grad_norm": 0.7480777502059937,
+      "learning_rate": 3.694283935093684e-06,
+      "loss": 0.6369,
+      "step": 7462
+    },
+    {
+      "epoch": 2.0673130193905815,
+      "grad_norm": 0.7199879288673401,
+      "learning_rate": 3.6939638924645336e-06,
+      "loss": 0.6139,
+      "step": 7463
+    },
+    {
+      "epoch": 2.067590027700831,
+      "grad_norm": 0.6921995878219604,
+      "learning_rate": 3.6936438244844802e-06,
+      "loss": 0.5678,
+      "step": 7464
+    },
+    {
+      "epoch": 2.0678670360110805,
+      "grad_norm": 0.7843019366264343,
+      "learning_rate": 3.6933237311603205e-06,
+      "loss": 0.6417,
+      "step": 7465
+    },
+    {
+      "epoch": 2.0681440443213295,
+      "grad_norm": 0.7345570921897888,
+      "learning_rate": 3.69300361249885e-06,
+      "loss": 0.6187,
+      "step": 7466
+    },
+    {
+      "epoch": 2.068421052631579,
+      "grad_norm": 0.7142504453659058,
+      "learning_rate": 3.692683468506866e-06,
+      "loss": 0.6057,
+      "step": 7467
+    },
+    {
+      "epoch": 2.0686980609418284,
+      "grad_norm": 0.690905749797821,
+      "learning_rate": 3.6923632991911672e-06,
+      "loss": 0.6299,
+      "step": 7468
+    },
+    {
+      "epoch": 2.0689750692520774,
+      "grad_norm": 0.7089608311653137,
+      "learning_rate": 3.69204310455855e-06,
+      "loss": 0.6054,
+      "step": 7469
+    },
+    {
+      "epoch": 2.069252077562327,
+      "grad_norm": 0.757316529750824,
+      "learning_rate": 3.691722884615814e-06,
+      "loss": 0.6013,
+      "step": 7470
+    },
+    {
+      "epoch": 2.0695290858725763,
+      "grad_norm": 0.7615708708763123,
+      "learning_rate": 3.6914026393697573e-06,
+      "loss": 0.5864,
+      "step": 7471
+    },
+    {
+      "epoch": 2.0698060941828254,
+      "grad_norm": 0.7319211959838867,
+      "learning_rate": 3.6910823688271803e-06,
+      "loss": 0.5808,
+      "step": 7472
+    },
+    {
+      "epoch": 2.070083102493075,
+      "grad_norm": 0.7206616401672363,
+      "learning_rate": 3.690762072994883e-06,
+      "loss": 0.582,
+      "step": 7473
+    },
+    {
+      "epoch": 2.0703601108033243,
+      "grad_norm": 0.7332830429077148,
+      "learning_rate": 3.6904417518796655e-06,
+      "loss": 0.6,
+      "step": 7474
+    },
+    {
+      "epoch": 2.0706371191135733,
+      "grad_norm": 0.7284979820251465,
+      "learning_rate": 3.69012140548833e-06,
+      "loss": 0.5562,
+      "step": 7475
+    },
+    {
+      "epoch": 2.0709141274238227,
+      "grad_norm": 0.7407473921775818,
+      "learning_rate": 3.689801033827678e-06,
+      "loss": 0.5966,
+      "step": 7476
+    },
+    {
+      "epoch": 2.071191135734072,
+      "grad_norm": 0.7004116177558899,
+      "learning_rate": 3.689480636904511e-06,
+      "loss": 0.6051,
+      "step": 7477
+    },
+    {
+      "epoch": 2.0714681440443212,
+      "grad_norm": 0.7138201594352722,
+      "learning_rate": 3.689160214725633e-06,
+      "loss": 0.5829,
+      "step": 7478
+    },
+    {
+      "epoch": 2.0717451523545707,
+      "grad_norm": 0.7637184858322144,
+      "learning_rate": 3.6888397672978465e-06,
+      "loss": 0.6599,
+      "step": 7479
+    },
+    {
+      "epoch": 2.07202216066482,
+      "grad_norm": 0.6777431964874268,
+      "learning_rate": 3.6885192946279557e-06,
+      "loss": 0.6032,
+      "step": 7480
+    },
+    {
+      "epoch": 2.072299168975069,
+      "grad_norm": 0.7209843397140503,
+      "learning_rate": 3.6881987967227656e-06,
+      "loss": 0.6343,
+      "step": 7481
+    },
+    {
+      "epoch": 2.0725761772853186,
+      "grad_norm": 0.6750618815422058,
+      "learning_rate": 3.6878782735890807e-06,
+      "loss": 0.6436,
+      "step": 7482
+    },
+    {
+      "epoch": 2.072853185595568,
+      "grad_norm": 0.7013235688209534,
+      "learning_rate": 3.6875577252337065e-06,
+      "loss": 0.6335,
+      "step": 7483
+    },
+    {
+      "epoch": 2.073130193905817,
+      "grad_norm": 0.7054664492607117,
+      "learning_rate": 3.6872371516634485e-06,
+      "loss": 0.5663,
+      "step": 7484
+    },
+    {
+      "epoch": 2.0734072022160666,
+      "grad_norm": 0.7187961339950562,
+      "learning_rate": 3.686916552885115e-06,
+      "loss": 0.6126,
+      "step": 7485
+    },
+    {
+      "epoch": 2.0736842105263156,
+      "grad_norm": 0.6890035271644592,
+      "learning_rate": 3.686595928905511e-06,
+      "loss": 0.5854,
+      "step": 7486
+    },
+    {
+      "epoch": 2.073961218836565,
+      "grad_norm": 0.7117055654525757,
+      "learning_rate": 3.6862752797314455e-06,
+      "loss": 0.5917,
+      "step": 7487
+    },
+    {
+      "epoch": 2.0742382271468145,
+      "grad_norm": 0.7330850958824158,
+      "learning_rate": 3.6859546053697264e-06,
+      "loss": 0.6316,
+      "step": 7488
+    },
+    {
+      "epoch": 2.0745152354570635,
+      "grad_norm": 0.7155464291572571,
+      "learning_rate": 3.685633905827163e-06,
+      "loss": 0.6043,
+      "step": 7489
+    },
+    {
+      "epoch": 2.074792243767313,
+      "grad_norm": 0.752101719379425,
+      "learning_rate": 3.685313181110563e-06,
+      "loss": 0.5926,
+      "step": 7490
+    },
+    {
+      "epoch": 2.0750692520775624,
+      "grad_norm": 0.695989727973938,
+      "learning_rate": 3.684992431226738e-06,
+      "loss": 0.5759,
+      "step": 7491
+    },
+    {
+      "epoch": 2.0753462603878114,
+      "grad_norm": 0.7123996019363403,
+      "learning_rate": 3.684671656182497e-06,
+      "loss": 0.6116,
+      "step": 7492
+    },
+    {
+      "epoch": 2.075623268698061,
+      "grad_norm": 0.7395370006561279,
+      "learning_rate": 3.684350855984652e-06,
+      "loss": 0.5756,
+      "step": 7493
+    },
+    {
+      "epoch": 2.0759002770083104,
+      "grad_norm": 0.7672297954559326,
+      "learning_rate": 3.6840300306400135e-06,
+      "loss": 0.58,
+      "step": 7494
+    },
+    {
+      "epoch": 2.0761772853185594,
+      "grad_norm": 0.7145938277244568,
+      "learning_rate": 3.683709180155393e-06,
+      "loss": 0.6158,
+      "step": 7495
+    },
+    {
+      "epoch": 2.076454293628809,
+      "grad_norm": 0.6776156425476074,
+      "learning_rate": 3.683388304537605e-06,
+      "loss": 0.6006,
+      "step": 7496
+    },
+    {
+      "epoch": 2.0767313019390583,
+      "grad_norm": 0.7504327297210693,
+      "learning_rate": 3.683067403793461e-06,
+      "loss": 0.5999,
+      "step": 7497
+    },
+    {
+      "epoch": 2.0770083102493073,
+      "grad_norm": 0.7060835957527161,
+      "learning_rate": 3.6827464779297744e-06,
+      "loss": 0.6208,
+      "step": 7498
+    },
+    {
+      "epoch": 2.077285318559557,
+      "grad_norm": 0.7075418829917908,
+      "learning_rate": 3.6824255269533595e-06,
+      "loss": 0.5903,
+      "step": 7499
+    },
+    {
+      "epoch": 2.0775623268698062,
+      "grad_norm": 0.7271056771278381,
+      "learning_rate": 3.682104550871031e-06,
+      "loss": 0.6038,
+      "step": 7500
+    },
+    {
+      "epoch": 2.0778393351800553,
+      "grad_norm": 0.7134436964988708,
+      "learning_rate": 3.6817835496896047e-06,
+      "loss": 0.5851,
+      "step": 7501
+    },
+    {
+      "epoch": 2.0781163434903047,
+      "grad_norm": 0.7162055373191833,
+      "learning_rate": 3.681462523415896e-06,
+      "loss": 0.6299,
+      "step": 7502
+    },
+    {
+      "epoch": 2.078393351800554,
+      "grad_norm": 0.6949959397315979,
+      "learning_rate": 3.6811414720567197e-06,
+      "loss": 0.5712,
+      "step": 7503
+    },
+    {
+      "epoch": 2.078670360110803,
+      "grad_norm": 0.6959918141365051,
+      "learning_rate": 3.6808203956188947e-06,
+      "loss": 0.6075,
+      "step": 7504
+    },
+    {
+      "epoch": 2.0789473684210527,
+      "grad_norm": 0.7247087955474854,
+      "learning_rate": 3.6804992941092367e-06,
+      "loss": 0.6493,
+      "step": 7505
+    },
+    {
+      "epoch": 2.079224376731302,
+      "grad_norm": 0.7352406978607178,
+      "learning_rate": 3.680178167534564e-06,
+      "loss": 0.6091,
+      "step": 7506
+    },
+    {
+      "epoch": 2.079501385041551,
+      "grad_norm": 0.7071592211723328,
+      "learning_rate": 3.6798570159016955e-06,
+      "loss": 0.5718,
+      "step": 7507
+    },
+    {
+      "epoch": 2.0797783933518006,
+      "grad_norm": 0.739302933216095,
+      "learning_rate": 3.6795358392174484e-06,
+      "loss": 0.5814,
+      "step": 7508
+    },
+    {
+      "epoch": 2.08005540166205,
+      "grad_norm": 0.7068384289741516,
+      "learning_rate": 3.679214637488644e-06,
+      "loss": 0.6191,
+      "step": 7509
+    },
+    {
+      "epoch": 2.080332409972299,
+      "grad_norm": 0.7154803276062012,
+      "learning_rate": 3.6788934107221016e-06,
+      "loss": 0.6188,
+      "step": 7510
+    },
+    {
+      "epoch": 2.0806094182825485,
+      "grad_norm": 0.7316164374351501,
+      "learning_rate": 3.6785721589246413e-06,
+      "loss": 0.5853,
+      "step": 7511
+    },
+    {
+      "epoch": 2.080886426592798,
+      "grad_norm": 0.7105570435523987,
+      "learning_rate": 3.678250882103084e-06,
+      "loss": 0.6135,
+      "step": 7512
+    },
+    {
+      "epoch": 2.081163434903047,
+      "grad_norm": 0.6962291598320007,
+      "learning_rate": 3.6779295802642524e-06,
+      "loss": 0.5665,
+      "step": 7513
+    },
+    {
+      "epoch": 2.0814404432132965,
+      "grad_norm": 0.7440493702888489,
+      "learning_rate": 3.6776082534149664e-06,
+      "loss": 0.6088,
+      "step": 7514
+    },
+    {
+      "epoch": 2.081717451523546,
+      "grad_norm": 0.7357923984527588,
+      "learning_rate": 3.677286901562051e-06,
+      "loss": 0.5822,
+      "step": 7515
+    },
+    {
+      "epoch": 2.081994459833795,
+      "grad_norm": 0.7124137878417969,
+      "learning_rate": 3.676965524712328e-06,
+      "loss": 0.6255,
+      "step": 7516
+    },
+    {
+      "epoch": 2.0822714681440444,
+      "grad_norm": 0.7291326522827148,
+      "learning_rate": 3.6766441228726212e-06,
+      "loss": 0.5945,
+      "step": 7517
+    },
+    {
+      "epoch": 2.0825484764542934,
+      "grad_norm": 0.7190305590629578,
+      "learning_rate": 3.6763226960497543e-06,
+      "loss": 0.5883,
+      "step": 7518
+    },
+    {
+      "epoch": 2.082825484764543,
+      "grad_norm": 0.7270352244377136,
+      "learning_rate": 3.676001244250553e-06,
+      "loss": 0.6422,
+      "step": 7519
+    },
+    {
+      "epoch": 2.0831024930747923,
+      "grad_norm": 0.7007784843444824,
+      "learning_rate": 3.675679767481842e-06,
+      "loss": 0.5702,
+      "step": 7520
+    },
+    {
+      "epoch": 2.0833795013850414,
+      "grad_norm": 0.7307643890380859,
+      "learning_rate": 3.6753582657504476e-06,
+      "loss": 0.5961,
+      "step": 7521
+    },
+    {
+      "epoch": 2.083656509695291,
+      "grad_norm": 0.7441365122795105,
+      "learning_rate": 3.6750367390631954e-06,
+      "loss": 0.6231,
+      "step": 7522
+    },
+    {
+      "epoch": 2.0839335180055403,
+      "grad_norm": 0.7360666394233704,
+      "learning_rate": 3.6747151874269125e-06,
+      "loss": 0.5898,
+      "step": 7523
+    },
+    {
+      "epoch": 2.0842105263157893,
+      "grad_norm": 0.7180540561676025,
+      "learning_rate": 3.674393610848426e-06,
+      "loss": 0.6219,
+      "step": 7524
+    },
+    {
+      "epoch": 2.0844875346260388,
+      "grad_norm": 0.7242913246154785,
+      "learning_rate": 3.6740720093345646e-06,
+      "loss": 0.6056,
+      "step": 7525
+    },
+    {
+      "epoch": 2.084764542936288,
+      "grad_norm": 0.7217580080032349,
+      "learning_rate": 3.673750382892156e-06,
+      "loss": 0.6093,
+      "step": 7526
+    },
+    {
+      "epoch": 2.0850415512465372,
+      "grad_norm": 0.6987811923027039,
+      "learning_rate": 3.6734287315280293e-06,
+      "loss": 0.625,
+      "step": 7527
+    },
+    {
+      "epoch": 2.0853185595567867,
+      "grad_norm": 0.7222145199775696,
+      "learning_rate": 3.673107055249014e-06,
+      "loss": 0.6025,
+      "step": 7528
+    },
+    {
+      "epoch": 2.085595567867036,
+      "grad_norm": 0.7623053789138794,
+      "learning_rate": 3.67278535406194e-06,
+      "loss": 0.6376,
+      "step": 7529
+    },
+    {
+      "epoch": 2.085872576177285,
+      "grad_norm": 0.7541896104812622,
+      "learning_rate": 3.672463627973639e-06,
+      "loss": 0.6107,
+      "step": 7530
+    },
+    {
+      "epoch": 2.0861495844875346,
+      "grad_norm": 0.7208195328712463,
+      "learning_rate": 3.672141876990939e-06,
+      "loss": 0.597,
+      "step": 7531
+    },
+    {
+      "epoch": 2.086426592797784,
+      "grad_norm": 0.7163652777671814,
+      "learning_rate": 3.671820101120675e-06,
+      "loss": 0.6422,
+      "step": 7532
+    },
+    {
+      "epoch": 2.086703601108033,
+      "grad_norm": 0.7693494558334351,
+      "learning_rate": 3.671498300369678e-06,
+      "loss": 0.6053,
+      "step": 7533
+    },
+    {
+      "epoch": 2.0869806094182826,
+      "grad_norm": 0.6839724779129028,
+      "learning_rate": 3.6711764747447797e-06,
+      "loss": 0.5143,
+      "step": 7534
+    },
+    {
+      "epoch": 2.087257617728532,
+      "grad_norm": 0.7062249183654785,
+      "learning_rate": 3.6708546242528135e-06,
+      "loss": 0.5237,
+      "step": 7535
+    },
+    {
+      "epoch": 2.087534626038781,
+      "grad_norm": 0.7552582025527954,
+      "learning_rate": 3.670532748900615e-06,
+      "loss": 0.5947,
+      "step": 7536
+    },
+    {
+      "epoch": 2.0878116343490305,
+      "grad_norm": 0.738476037979126,
+      "learning_rate": 3.670210848695016e-06,
+      "loss": 0.6166,
+      "step": 7537
+    },
+    {
+      "epoch": 2.08808864265928,
+      "grad_norm": 0.7047578692436218,
+      "learning_rate": 3.6698889236428526e-06,
+      "loss": 0.6175,
+      "step": 7538
+    },
+    {
+      "epoch": 2.088365650969529,
+      "grad_norm": 0.7480363845825195,
+      "learning_rate": 3.66956697375096e-06,
+      "loss": 0.6328,
+      "step": 7539
+    },
+    {
+      "epoch": 2.0886426592797784,
+      "grad_norm": 0.7898052930831909,
+      "learning_rate": 3.6692449990261736e-06,
+      "loss": 0.6133,
+      "step": 7540
+    },
+    {
+      "epoch": 2.088919667590028,
+      "grad_norm": 0.7435026168823242,
+      "learning_rate": 3.6689229994753293e-06,
+      "loss": 0.5961,
+      "step": 7541
+    },
+    {
+      "epoch": 2.089196675900277,
+      "grad_norm": 0.716643750667572,
+      "learning_rate": 3.668600975105266e-06,
+      "loss": 0.5999,
+      "step": 7542
+    },
+    {
+      "epoch": 2.0894736842105264,
+      "grad_norm": 0.7211320400238037,
+      "learning_rate": 3.668278925922819e-06,
+      "loss": 0.6316,
+      "step": 7543
+    },
+    {
+      "epoch": 2.089750692520776,
+      "grad_norm": 0.7620773911476135,
+      "learning_rate": 3.6679568519348264e-06,
+      "loss": 0.5709,
+      "step": 7544
+    },
+    {
+      "epoch": 2.090027700831025,
+      "grad_norm": 0.7366681694984436,
+      "learning_rate": 3.6676347531481283e-06,
+      "loss": 0.6344,
+      "step": 7545
+    },
+    {
+      "epoch": 2.0903047091412743,
+      "grad_norm": 0.7111081480979919,
+      "learning_rate": 3.6673126295695615e-06,
+      "loss": 0.5959,
+      "step": 7546
+    },
+    {
+      "epoch": 2.0905817174515233,
+      "grad_norm": 0.74237060546875,
+      "learning_rate": 3.6669904812059674e-06,
+      "loss": 0.6095,
+      "step": 7547
+    },
+    {
+      "epoch": 2.090858725761773,
+      "grad_norm": 0.7371329665184021,
+      "learning_rate": 3.6666683080641846e-06,
+      "loss": 0.5726,
+      "step": 7548
+    },
+    {
+      "epoch": 2.0911357340720222,
+      "grad_norm": 0.7270698547363281,
+      "learning_rate": 3.666346110151055e-06,
+      "loss": 0.6043,
+      "step": 7549
+    },
+    {
+      "epoch": 2.0914127423822713,
+      "grad_norm": 0.738433301448822,
+      "learning_rate": 3.666023887473418e-06,
+      "loss": 0.6094,
+      "step": 7550
+    },
+    {
+      "epoch": 2.0916897506925207,
+      "grad_norm": 0.6856642961502075,
+      "learning_rate": 3.665701640038117e-06,
+      "loss": 0.5672,
+      "step": 7551
+    },
+    {
+      "epoch": 2.09196675900277,
+      "grad_norm": 0.741007924079895,
+      "learning_rate": 3.6653793678519926e-06,
+      "loss": 0.6287,
+      "step": 7552
+    },
+    {
+      "epoch": 2.092243767313019,
+      "grad_norm": 0.703421413898468,
+      "learning_rate": 3.6650570709218884e-06,
+      "loss": 0.6057,
+      "step": 7553
+    },
+    {
+      "epoch": 2.0925207756232687,
+      "grad_norm": 0.7339208722114563,
+      "learning_rate": 3.6647347492546477e-06,
+      "loss": 0.6702,
+      "step": 7554
+    },
+    {
+      "epoch": 2.092797783933518,
+      "grad_norm": 0.7148478627204895,
+      "learning_rate": 3.664412402857114e-06,
+      "loss": 0.6141,
+      "step": 7555
+    },
+    {
+      "epoch": 2.093074792243767,
+      "grad_norm": 0.7352780699729919,
+      "learning_rate": 3.6640900317361306e-06,
+      "loss": 0.6118,
+      "step": 7556
+    },
+    {
+      "epoch": 2.0933518005540166,
+      "grad_norm": 0.7099614143371582,
+      "learning_rate": 3.6637676358985435e-06,
+      "loss": 0.6144,
+      "step": 7557
+    },
+    {
+      "epoch": 2.093628808864266,
+      "grad_norm": 0.7019182443618774,
+      "learning_rate": 3.663445215351198e-06,
+      "loss": 0.6303,
+      "step": 7558
+    },
+    {
+      "epoch": 2.093905817174515,
+      "grad_norm": 0.6563045382499695,
+      "learning_rate": 3.6631227701009387e-06,
+      "loss": 0.5652,
+      "step": 7559
+    },
+    {
+      "epoch": 2.0941828254847645,
+      "grad_norm": 0.7421273589134216,
+      "learning_rate": 3.662800300154613e-06,
+      "loss": 0.608,
+      "step": 7560
+    },
+    {
+      "epoch": 2.094459833795014,
+      "grad_norm": 0.71537846326828,
+      "learning_rate": 3.662477805519068e-06,
+      "loss": 0.6085,
+      "step": 7561
+    },
+    {
+      "epoch": 2.094736842105263,
+      "grad_norm": 0.7306783199310303,
+      "learning_rate": 3.66215528620115e-06,
+      "loss": 0.5915,
+      "step": 7562
+    },
+    {
+      "epoch": 2.0950138504155125,
+      "grad_norm": 0.7352632284164429,
+      "learning_rate": 3.661832742207707e-06,
+      "loss": 0.6423,
+      "step": 7563
+    },
+    {
+      "epoch": 2.095290858725762,
+      "grad_norm": 0.7389764189720154,
+      "learning_rate": 3.6615101735455883e-06,
+      "loss": 0.6159,
+      "step": 7564
+    },
+    {
+      "epoch": 2.095567867036011,
+      "grad_norm": 0.7337745428085327,
+      "learning_rate": 3.6611875802216427e-06,
+      "loss": 0.5833,
+      "step": 7565
+    },
+    {
+      "epoch": 2.0958448753462604,
+      "grad_norm": 0.6928842663764954,
+      "learning_rate": 3.660864962242719e-06,
+      "loss": 0.5738,
+      "step": 7566
+    },
+    {
+      "epoch": 2.09612188365651,
+      "grad_norm": 0.6791226863861084,
+      "learning_rate": 3.6605423196156687e-06,
+      "loss": 0.5377,
+      "step": 7567
+    },
+    {
+      "epoch": 2.096398891966759,
+      "grad_norm": 0.7218356132507324,
+      "learning_rate": 3.66021965234734e-06,
+      "loss": 0.5785,
+      "step": 7568
+    },
+    {
+      "epoch": 2.0966759002770083,
+      "grad_norm": 0.700420618057251,
+      "learning_rate": 3.6598969604445854e-06,
+      "loss": 0.585,
+      "step": 7569
+    },
+    {
+      "epoch": 2.096952908587258,
+      "grad_norm": 0.7712434530258179,
+      "learning_rate": 3.659574243914257e-06,
+      "loss": 0.6434,
+      "step": 7570
+    },
+    {
+      "epoch": 2.097229916897507,
+      "grad_norm": 0.7493064999580383,
+      "learning_rate": 3.659251502763206e-06,
+      "loss": 0.6556,
+      "step": 7571
+    },
+    {
+      "epoch": 2.0975069252077563,
+      "grad_norm": 0.7280117273330688,
+      "learning_rate": 3.658928736998284e-06,
+      "loss": 0.596,
+      "step": 7572
+    },
+    {
+      "epoch": 2.0977839335180057,
+      "grad_norm": 0.7482536435127258,
+      "learning_rate": 3.6586059466263463e-06,
+      "loss": 0.5978,
+      "step": 7573
+    },
+    {
+      "epoch": 2.0980609418282548,
+      "grad_norm": 0.690613329410553,
+      "learning_rate": 3.6582831316542454e-06,
+      "loss": 0.6279,
+      "step": 7574
+    },
+    {
+      "epoch": 2.098337950138504,
+      "grad_norm": 0.6842516660690308,
+      "learning_rate": 3.657960292088836e-06,
+      "loss": 0.5736,
+      "step": 7575
+    },
+    {
+      "epoch": 2.0986149584487537,
+      "grad_norm": 0.7524868845939636,
+      "learning_rate": 3.657637427936972e-06,
+      "loss": 0.671,
+      "step": 7576
+    },
+    {
+      "epoch": 2.0988919667590027,
+      "grad_norm": 0.708136796951294,
+      "learning_rate": 3.6573145392055097e-06,
+      "loss": 0.6255,
+      "step": 7577
+    },
+    {
+      "epoch": 2.099168975069252,
+      "grad_norm": 0.7181639075279236,
+      "learning_rate": 3.656991625901304e-06,
+      "loss": 0.6044,
+      "step": 7578
+    },
+    {
+      "epoch": 2.0994459833795016,
+      "grad_norm": 0.7263873815536499,
+      "learning_rate": 3.6566686880312106e-06,
+      "loss": 0.6201,
+      "step": 7579
+    },
+    {
+      "epoch": 2.0997229916897506,
+      "grad_norm": 0.7212371826171875,
+      "learning_rate": 3.656345725602089e-06,
+      "loss": 0.591,
+      "step": 7580
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7355701923370361,
+      "learning_rate": 3.656022738620793e-06,
+      "loss": 0.6131,
+      "step": 7581
+    },
+    {
+      "epoch": 2.100277008310249,
+      "grad_norm": 0.6883821487426758,
+      "learning_rate": 3.655699727094183e-06,
+      "loss": 0.5995,
+      "step": 7582
+    },
+    {
+      "epoch": 2.1005540166204986,
+      "grad_norm": 0.716107189655304,
+      "learning_rate": 3.655376691029116e-06,
+      "loss": 0.6251,
+      "step": 7583
+    },
+    {
+      "epoch": 2.100831024930748,
+      "grad_norm": 0.7313331961631775,
+      "learning_rate": 3.655053630432452e-06,
+      "loss": 0.6354,
+      "step": 7584
+    },
+    {
+      "epoch": 2.101108033240997,
+      "grad_norm": 0.7299684882164001,
+      "learning_rate": 3.654730545311049e-06,
+      "loss": 0.6014,
+      "step": 7585
+    },
+    {
+      "epoch": 2.1013850415512465,
+      "grad_norm": 0.7288849949836731,
+      "learning_rate": 3.6544074356717675e-06,
+      "loss": 0.5749,
+      "step": 7586
+    },
+    {
+      "epoch": 2.101662049861496,
+      "grad_norm": 0.724434494972229,
+      "learning_rate": 3.654084301521469e-06,
+      "loss": 0.575,
+      "step": 7587
+    },
+    {
+      "epoch": 2.101939058171745,
+      "grad_norm": 0.7240490913391113,
+      "learning_rate": 3.6537611428670127e-06,
+      "loss": 0.6028,
+      "step": 7588
+    },
+    {
+      "epoch": 2.1022160664819944,
+      "grad_norm": 0.7254226803779602,
+      "learning_rate": 3.6534379597152615e-06,
+      "loss": 0.641,
+      "step": 7589
+    },
+    {
+      "epoch": 2.102493074792244,
+      "grad_norm": 0.7376781702041626,
+      "learning_rate": 3.653114752073077e-06,
+      "loss": 0.6108,
+      "step": 7590
+    },
+    {
+      "epoch": 2.102770083102493,
+      "grad_norm": 0.7038312554359436,
+      "learning_rate": 3.6527915199473206e-06,
+      "loss": 0.609,
+      "step": 7591
+    },
+    {
+      "epoch": 2.1030470914127424,
+      "grad_norm": 0.7228493094444275,
+      "learning_rate": 3.652468263344857e-06,
+      "loss": 0.6033,
+      "step": 7592
+    },
+    {
+      "epoch": 2.103324099722992,
+      "grad_norm": 0.7285172343254089,
+      "learning_rate": 3.65214498227255e-06,
+      "loss": 0.6323,
+      "step": 7593
+    },
+    {
+      "epoch": 2.103601108033241,
+      "grad_norm": 0.7170820236206055,
+      "learning_rate": 3.651821676737261e-06,
+      "loss": 0.5395,
+      "step": 7594
+    },
+    {
+      "epoch": 2.1038781163434903,
+      "grad_norm": 0.7217197418212891,
+      "learning_rate": 3.6514983467458577e-06,
+      "loss": 0.6448,
+      "step": 7595
+    },
+    {
+      "epoch": 2.1041551246537398,
+      "grad_norm": 0.7080284953117371,
+      "learning_rate": 3.651174992305203e-06,
+      "loss": 0.6125,
+      "step": 7596
+    },
+    {
+      "epoch": 2.104432132963989,
+      "grad_norm": 0.7334601879119873,
+      "learning_rate": 3.650851613422164e-06,
+      "loss": 0.5814,
+      "step": 7597
+    },
+    {
+      "epoch": 2.1047091412742382,
+      "grad_norm": 0.7425627708435059,
+      "learning_rate": 3.650528210103606e-06,
+      "loss": 0.5994,
+      "step": 7598
+    },
+    {
+      "epoch": 2.1049861495844877,
+      "grad_norm": 0.720252275466919,
+      "learning_rate": 3.6502047823563967e-06,
+      "loss": 0.5901,
+      "step": 7599
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.6992184519767761,
+      "learning_rate": 3.6498813301874016e-06,
+      "loss": 0.5599,
+      "step": 7600
+    },
+    {
+      "epoch": 2.105540166204986,
+      "grad_norm": 0.7782900333404541,
+      "learning_rate": 3.64955785360349e-06,
+      "loss": 0.6099,
+      "step": 7601
+    },
+    {
+      "epoch": 2.1058171745152356,
+      "grad_norm": 0.7176836729049683,
+      "learning_rate": 3.6492343526115292e-06,
+      "loss": 0.5997,
+      "step": 7602
+    },
+    {
+      "epoch": 2.1060941828254847,
+      "grad_norm": 0.6862257719039917,
+      "learning_rate": 3.6489108272183883e-06,
+      "loss": 0.5779,
+      "step": 7603
+    },
+    {
+      "epoch": 2.106371191135734,
+      "grad_norm": 0.7442038059234619,
+      "learning_rate": 3.648587277430936e-06,
+      "loss": 0.6,
+      "step": 7604
+    },
+    {
+      "epoch": 2.1066481994459836,
+      "grad_norm": 0.7054752111434937,
+      "learning_rate": 3.6482637032560437e-06,
+      "loss": 0.533,
+      "step": 7605
+    },
+    {
+      "epoch": 2.1069252077562326,
+      "grad_norm": 0.7280328869819641,
+      "learning_rate": 3.64794010470058e-06,
+      "loss": 0.666,
+      "step": 7606
+    },
+    {
+      "epoch": 2.107202216066482,
+      "grad_norm": 0.8324413299560547,
+      "learning_rate": 3.6476164817714166e-06,
+      "loss": 0.6287,
+      "step": 7607
+    },
+    {
+      "epoch": 2.107479224376731,
+      "grad_norm": 0.725425660610199,
+      "learning_rate": 3.647292834475424e-06,
+      "loss": 0.6133,
+      "step": 7608
+    },
+    {
+      "epoch": 2.1077562326869805,
+      "grad_norm": 0.6995035409927368,
+      "learning_rate": 3.6469691628194753e-06,
+      "loss": 0.6095,
+      "step": 7609
+    },
+    {
+      "epoch": 2.10803324099723,
+      "grad_norm": 0.6647850871086121,
+      "learning_rate": 3.6466454668104423e-06,
+      "loss": 0.5523,
+      "step": 7610
+    },
+    {
+      "epoch": 2.108310249307479,
+      "grad_norm": 0.7180917263031006,
+      "learning_rate": 3.6463217464551977e-06,
+      "loss": 0.5811,
+      "step": 7611
+    },
+    {
+      "epoch": 2.1085872576177285,
+      "grad_norm": 0.7054855823516846,
+      "learning_rate": 3.645998001760615e-06,
+      "loss": 0.5845,
+      "step": 7612
+    },
+    {
+      "epoch": 2.108864265927978,
+      "grad_norm": 0.7367815375328064,
+      "learning_rate": 3.645674232733568e-06,
+      "loss": 0.6051,
+      "step": 7613
+    },
+    {
+      "epoch": 2.109141274238227,
+      "grad_norm": 0.6971177458763123,
+      "learning_rate": 3.6453504393809312e-06,
+      "loss": 0.5849,
+      "step": 7614
+    },
+    {
+      "epoch": 2.1094182825484764,
+      "grad_norm": 0.6860442161560059,
+      "learning_rate": 3.6450266217095804e-06,
+      "loss": 0.545,
+      "step": 7615
+    },
+    {
+      "epoch": 2.109695290858726,
+      "grad_norm": 0.7291378974914551,
+      "learning_rate": 3.64470277972639e-06,
+      "loss": 0.5841,
+      "step": 7616
+    },
+    {
+      "epoch": 2.109972299168975,
+      "grad_norm": 0.686470627784729,
+      "learning_rate": 3.6443789134382356e-06,
+      "loss": 0.5594,
+      "step": 7617
+    },
+    {
+      "epoch": 2.1102493074792243,
+      "grad_norm": 0.7135543823242188,
+      "learning_rate": 3.6440550228519954e-06,
+      "loss": 0.5797,
+      "step": 7618
+    },
+    {
+      "epoch": 2.110526315789474,
+      "grad_norm": 0.7279857397079468,
+      "learning_rate": 3.643731107974545e-06,
+      "loss": 0.5673,
+      "step": 7619
+    },
+    {
+      "epoch": 2.110803324099723,
+      "grad_norm": 0.6739892363548279,
+      "learning_rate": 3.6434071688127627e-06,
+      "loss": 0.5728,
+      "step": 7620
+    },
+    {
+      "epoch": 2.1110803324099723,
+      "grad_norm": 0.7875750660896301,
+      "learning_rate": 3.643083205373526e-06,
+      "loss": 0.611,
+      "step": 7621
+    },
+    {
+      "epoch": 2.1113573407202217,
+      "grad_norm": 0.7165523767471313,
+      "learning_rate": 3.6427592176637133e-06,
+      "loss": 0.5847,
+      "step": 7622
+    },
+    {
+      "epoch": 2.1116343490304708,
+      "grad_norm": 0.7262118458747864,
+      "learning_rate": 3.6424352056902046e-06,
+      "loss": 0.6097,
+      "step": 7623
+    },
+    {
+      "epoch": 2.11191135734072,
+      "grad_norm": 0.6838838458061218,
+      "learning_rate": 3.642111169459879e-06,
+      "loss": 0.6121,
+      "step": 7624
+    },
+    {
+      "epoch": 2.1121883656509697,
+      "grad_norm": 0.7182348966598511,
+      "learning_rate": 3.641787108979617e-06,
+      "loss": 0.5979,
+      "step": 7625
+    },
+    {
+      "epoch": 2.1124653739612187,
+      "grad_norm": 0.7114434242248535,
+      "learning_rate": 3.641463024256298e-06,
+      "loss": 0.6148,
+      "step": 7626
+    },
+    {
+      "epoch": 2.112742382271468,
+      "grad_norm": 0.7483696937561035,
+      "learning_rate": 3.6411389152968046e-06,
+      "loss": 0.6394,
+      "step": 7627
+    },
+    {
+      "epoch": 2.1130193905817176,
+      "grad_norm": 0.7226950526237488,
+      "learning_rate": 3.6408147821080183e-06,
+      "loss": 0.6132,
+      "step": 7628
+    },
+    {
+      "epoch": 2.1132963988919666,
+      "grad_norm": 0.7192668914794922,
+      "learning_rate": 3.64049062469682e-06,
+      "loss": 0.6113,
+      "step": 7629
+    },
+    {
+      "epoch": 2.113573407202216,
+      "grad_norm": 0.6802382469177246,
+      "learning_rate": 3.6401664430700935e-06,
+      "loss": 0.5291,
+      "step": 7630
+    },
+    {
+      "epoch": 2.1138504155124656,
+      "grad_norm": 0.7091238498687744,
+      "learning_rate": 3.6398422372347227e-06,
+      "loss": 0.6269,
+      "step": 7631
+    },
+    {
+      "epoch": 2.1141274238227146,
+      "grad_norm": 0.7575234770774841,
+      "learning_rate": 3.639518007197589e-06,
+      "loss": 0.5931,
+      "step": 7632
+    },
+    {
+      "epoch": 2.114404432132964,
+      "grad_norm": 0.7375851273536682,
+      "learning_rate": 3.6391937529655786e-06,
+      "loss": 0.6294,
+      "step": 7633
+    },
+    {
+      "epoch": 2.1146814404432135,
+      "grad_norm": 0.7400649785995483,
+      "learning_rate": 3.6388694745455756e-06,
+      "loss": 0.623,
+      "step": 7634
+    },
+    {
+      "epoch": 2.1149584487534625,
+      "grad_norm": 0.7256798148155212,
+      "learning_rate": 3.6385451719444655e-06,
+      "loss": 0.5864,
+      "step": 7635
+    },
+    {
+      "epoch": 2.115235457063712,
+      "grad_norm": 0.7284665107727051,
+      "learning_rate": 3.6382208451691333e-06,
+      "loss": 0.5948,
+      "step": 7636
+    },
+    {
+      "epoch": 2.1155124653739614,
+      "grad_norm": 0.7431206107139587,
+      "learning_rate": 3.6378964942264662e-06,
+      "loss": 0.6077,
+      "step": 7637
+    },
+    {
+      "epoch": 2.1157894736842104,
+      "grad_norm": 0.727976381778717,
+      "learning_rate": 3.6375721191233508e-06,
+      "loss": 0.6108,
+      "step": 7638
+    },
+    {
+      "epoch": 2.11606648199446,
+      "grad_norm": 0.7988527417182922,
+      "learning_rate": 3.637247719866674e-06,
+      "loss": 0.6067,
+      "step": 7639
+    },
+    {
+      "epoch": 2.1163434903047094,
+      "grad_norm": 0.7804294228553772,
+      "learning_rate": 3.6369232964633237e-06,
+      "loss": 0.5959,
+      "step": 7640
+    },
+    {
+      "epoch": 2.1166204986149584,
+      "grad_norm": 0.7260686159133911,
+      "learning_rate": 3.6365988489201887e-06,
+      "loss": 0.6191,
+      "step": 7641
+    },
+    {
+      "epoch": 2.116897506925208,
+      "grad_norm": 0.712604820728302,
+      "learning_rate": 3.6362743772441576e-06,
+      "loss": 0.6136,
+      "step": 7642
+    },
+    {
+      "epoch": 2.117174515235457,
+      "grad_norm": 0.7114692330360413,
+      "learning_rate": 3.63594988144212e-06,
+      "loss": 0.5637,
+      "step": 7643
+    },
+    {
+      "epoch": 2.1174515235457063,
+      "grad_norm": 0.751819372177124,
+      "learning_rate": 3.635625361520966e-06,
+      "loss": 0.5996,
+      "step": 7644
+    },
+    {
+      "epoch": 2.1177285318559558,
+      "grad_norm": 0.7611382603645325,
+      "learning_rate": 3.6353008174875844e-06,
+      "loss": 0.6446,
+      "step": 7645
+    },
+    {
+      "epoch": 2.118005540166205,
+      "grad_norm": 0.7532554268836975,
+      "learning_rate": 3.634976249348867e-06,
+      "loss": 0.6,
+      "step": 7646
+    },
+    {
+      "epoch": 2.1182825484764543,
+      "grad_norm": 0.7395018935203552,
+      "learning_rate": 3.6346516571117063e-06,
+      "loss": 0.599,
+      "step": 7647
+    },
+    {
+      "epoch": 2.1185595567867037,
+      "grad_norm": 0.7286550998687744,
+      "learning_rate": 3.6343270407829934e-06,
+      "loss": 0.6025,
+      "step": 7648
+    },
+    {
+      "epoch": 2.1188365650969527,
+      "grad_norm": 0.7371105551719666,
+      "learning_rate": 3.63400240036962e-06,
+      "loss": 0.6017,
+      "step": 7649
+    },
+    {
+      "epoch": 2.119113573407202,
+      "grad_norm": 0.7373232841491699,
+      "learning_rate": 3.63367773587848e-06,
+      "loss": 0.587,
+      "step": 7650
+    },
+    {
+      "epoch": 2.1193905817174516,
+      "grad_norm": 0.7241517901420593,
+      "learning_rate": 3.633353047316467e-06,
+      "loss": 0.61,
+      "step": 7651
+    },
+    {
+      "epoch": 2.1196675900277007,
+      "grad_norm": 0.7286142110824585,
+      "learning_rate": 3.633028334690473e-06,
+      "loss": 0.6041,
+      "step": 7652
+    },
+    {
+      "epoch": 2.11994459833795,
+      "grad_norm": 0.7213249206542969,
+      "learning_rate": 3.632703598007396e-06,
+      "loss": 0.6292,
+      "step": 7653
+    },
+    {
+      "epoch": 2.1202216066481996,
+      "grad_norm": 0.7063266634941101,
+      "learning_rate": 3.632378837274128e-06,
+      "loss": 0.5628,
+      "step": 7654
+    },
+    {
+      "epoch": 2.1204986149584486,
+      "grad_norm": 0.7615188956260681,
+      "learning_rate": 3.6320540524975655e-06,
+      "loss": 0.6539,
+      "step": 7655
+    },
+    {
+      "epoch": 2.120775623268698,
+      "grad_norm": 0.7243799567222595,
+      "learning_rate": 3.6317292436846053e-06,
+      "loss": 0.598,
+      "step": 7656
+    },
+    {
+      "epoch": 2.1210526315789475,
+      "grad_norm": 0.716153085231781,
+      "learning_rate": 3.631404410842142e-06,
+      "loss": 0.6016,
+      "step": 7657
+    },
+    {
+      "epoch": 2.1213296398891965,
+      "grad_norm": 0.7011458277702332,
+      "learning_rate": 3.6310795539770738e-06,
+      "loss": 0.5904,
+      "step": 7658
+    },
+    {
+      "epoch": 2.121606648199446,
+      "grad_norm": 0.7134310603141785,
+      "learning_rate": 3.630754673096299e-06,
+      "loss": 0.6095,
+      "step": 7659
+    },
+    {
+      "epoch": 2.1218836565096955,
+      "grad_norm": 0.6938473582267761,
+      "learning_rate": 3.6304297682067146e-06,
+      "loss": 0.5916,
+      "step": 7660
+    },
+    {
+      "epoch": 2.1221606648199445,
+      "grad_norm": 0.7490888833999634,
+      "learning_rate": 3.6301048393152196e-06,
+      "loss": 0.6128,
+      "step": 7661
+    },
+    {
+      "epoch": 2.122437673130194,
+      "grad_norm": 0.6940358877182007,
+      "learning_rate": 3.6297798864287122e-06,
+      "loss": 0.5638,
+      "step": 7662
+    },
+    {
+      "epoch": 2.1227146814404434,
+      "grad_norm": 0.7214860916137695,
+      "learning_rate": 3.6294549095540933e-06,
+      "loss": 0.636,
+      "step": 7663
+    },
+    {
+      "epoch": 2.1229916897506924,
+      "grad_norm": 0.7091872096061707,
+      "learning_rate": 3.6291299086982624e-06,
+      "loss": 0.612,
+      "step": 7664
+    },
+    {
+      "epoch": 2.123268698060942,
+      "grad_norm": 0.7438629269599915,
+      "learning_rate": 3.628804883868119e-06,
+      "loss": 0.6278,
+      "step": 7665
+    },
+    {
+      "epoch": 2.1235457063711913,
+      "grad_norm": 0.7003153562545776,
+      "learning_rate": 3.628479835070567e-06,
+      "loss": 0.567,
+      "step": 7666
+    },
+    {
+      "epoch": 2.1238227146814403,
+      "grad_norm": 0.7519192695617676,
+      "learning_rate": 3.6281547623125046e-06,
+      "loss": 0.6302,
+      "step": 7667
+    },
+    {
+      "epoch": 2.12409972299169,
+      "grad_norm": 0.753014087677002,
+      "learning_rate": 3.6278296656008366e-06,
+      "loss": 0.6041,
+      "step": 7668
+    },
+    {
+      "epoch": 2.1243767313019393,
+      "grad_norm": 0.7036569118499756,
+      "learning_rate": 3.6275045449424643e-06,
+      "loss": 0.5714,
+      "step": 7669
+    },
+    {
+      "epoch": 2.1246537396121883,
+      "grad_norm": 0.805932343006134,
+      "learning_rate": 3.6271794003442918e-06,
+      "loss": 0.6174,
+      "step": 7670
+    },
+    {
+      "epoch": 2.1249307479224377,
+      "grad_norm": 0.7357967495918274,
+      "learning_rate": 3.626854231813221e-06,
+      "loss": 0.6006,
+      "step": 7671
+    },
+    {
+      "epoch": 2.1252077562326868,
+      "grad_norm": 0.7303361296653748,
+      "learning_rate": 3.6265290393561586e-06,
+      "loss": 0.5773,
+      "step": 7672
+    },
+    {
+      "epoch": 2.1254847645429362,
+      "grad_norm": 0.7263323664665222,
+      "learning_rate": 3.6262038229800067e-06,
+      "loss": 0.6136,
+      "step": 7673
+    },
+    {
+      "epoch": 2.1257617728531857,
+      "grad_norm": 0.7003388404846191,
+      "learning_rate": 3.625878582691672e-06,
+      "loss": 0.6143,
+      "step": 7674
+    },
+    {
+      "epoch": 2.1260387811634347,
+      "grad_norm": 0.7387146949768066,
+      "learning_rate": 3.6255533184980595e-06,
+      "loss": 0.548,
+      "step": 7675
+    },
+    {
+      "epoch": 2.126315789473684,
+      "grad_norm": 0.7310213446617126,
+      "learning_rate": 3.6252280304060768e-06,
+      "loss": 0.6148,
+      "step": 7676
+    },
+    {
+      "epoch": 2.1265927977839336,
+      "grad_norm": 0.7344486713409424,
+      "learning_rate": 3.624902718422628e-06,
+      "loss": 0.6146,
+      "step": 7677
+    },
+    {
+      "epoch": 2.1268698060941826,
+      "grad_norm": 0.7056968808174133,
+      "learning_rate": 3.6245773825546227e-06,
+      "loss": 0.57,
+      "step": 7678
+    },
+    {
+      "epoch": 2.127146814404432,
+      "grad_norm": 0.7251293659210205,
+      "learning_rate": 3.6242520228089673e-06,
+      "loss": 0.5487,
+      "step": 7679
+    },
+    {
+      "epoch": 2.1274238227146816,
+      "grad_norm": 0.706130862236023,
+      "learning_rate": 3.6239266391925703e-06,
+      "loss": 0.5979,
+      "step": 7680
+    },
+    {
+      "epoch": 2.1277008310249306,
+      "grad_norm": 0.6742924451828003,
+      "learning_rate": 3.623601231712341e-06,
+      "loss": 0.5492,
+      "step": 7681
+    },
+    {
+      "epoch": 2.12797783933518,
+      "grad_norm": 0.7342013716697693,
+      "learning_rate": 3.623275800375188e-06,
+      "loss": 0.6248,
+      "step": 7682
+    },
+    {
+      "epoch": 2.1282548476454295,
+      "grad_norm": 0.7537729144096375,
+      "learning_rate": 3.6229503451880215e-06,
+      "loss": 0.5866,
+      "step": 7683
+    },
+    {
+      "epoch": 2.1285318559556785,
+      "grad_norm": 0.7483304142951965,
+      "learning_rate": 3.622624866157751e-06,
+      "loss": 0.6452,
+      "step": 7684
+    },
+    {
+      "epoch": 2.128808864265928,
+      "grad_norm": 0.7003677487373352,
+      "learning_rate": 3.6222993632912882e-06,
+      "loss": 0.6525,
+      "step": 7685
+    },
+    {
+      "epoch": 2.1290858725761774,
+      "grad_norm": 0.7310715317726135,
+      "learning_rate": 3.6219738365955435e-06,
+      "loss": 0.5519,
+      "step": 7686
+    },
+    {
+      "epoch": 2.1293628808864264,
+      "grad_norm": 0.7531241178512573,
+      "learning_rate": 3.621648286077429e-06,
+      "loss": 0.555,
+      "step": 7687
+    },
+    {
+      "epoch": 2.129639889196676,
+      "grad_norm": 0.7238945960998535,
+      "learning_rate": 3.621322711743858e-06,
+      "loss": 0.5715,
+      "step": 7688
+    },
+    {
+      "epoch": 2.1299168975069254,
+      "grad_norm": 0.7214088439941406,
+      "learning_rate": 3.6209971136017413e-06,
+      "loss": 0.5603,
+      "step": 7689
+    },
+    {
+      "epoch": 2.1301939058171744,
+      "grad_norm": 0.7460694313049316,
+      "learning_rate": 3.6206714916579925e-06,
+      "loss": 0.5563,
+      "step": 7690
+    },
+    {
+      "epoch": 2.130470914127424,
+      "grad_norm": 0.7241212725639343,
+      "learning_rate": 3.620345845919527e-06,
+      "loss": 0.5997,
+      "step": 7691
+    },
+    {
+      "epoch": 2.1307479224376733,
+      "grad_norm": 0.7266137003898621,
+      "learning_rate": 3.6200201763932583e-06,
+      "loss": 0.5617,
+      "step": 7692
+    },
+    {
+      "epoch": 2.1310249307479223,
+      "grad_norm": 0.7282004952430725,
+      "learning_rate": 3.6196944830861003e-06,
+      "loss": 0.5975,
+      "step": 7693
+    },
+    {
+      "epoch": 2.131301939058172,
+      "grad_norm": 0.7729506492614746,
+      "learning_rate": 3.61936876600497e-06,
+      "loss": 0.6256,
+      "step": 7694
+    },
+    {
+      "epoch": 2.1315789473684212,
+      "grad_norm": 0.721755862236023,
+      "learning_rate": 3.6190430251567817e-06,
+      "loss": 0.6396,
+      "step": 7695
+    },
+    {
+      "epoch": 2.1318559556786703,
+      "grad_norm": 0.7221630811691284,
+      "learning_rate": 3.6187172605484523e-06,
+      "loss": 0.6348,
+      "step": 7696
+    },
+    {
+      "epoch": 2.1321329639889197,
+      "grad_norm": 0.6905040144920349,
+      "learning_rate": 3.6183914721868986e-06,
+      "loss": 0.5733,
+      "step": 7697
+    },
+    {
+      "epoch": 2.132409972299169,
+      "grad_norm": 0.7066609859466553,
+      "learning_rate": 3.618065660079038e-06,
+      "loss": 0.5561,
+      "step": 7698
+    },
+    {
+      "epoch": 2.132686980609418,
+      "grad_norm": 0.7088309526443481,
+      "learning_rate": 3.6177398242317878e-06,
+      "loss": 0.5973,
+      "step": 7699
+    },
+    {
+      "epoch": 2.1329639889196677,
+      "grad_norm": 0.7220036387443542,
+      "learning_rate": 3.617413964652067e-06,
+      "loss": 0.5885,
+      "step": 7700
+    },
+    {
+      "epoch": 2.133240997229917,
+      "grad_norm": 0.7076956629753113,
+      "learning_rate": 3.6170880813467947e-06,
+      "loss": 0.5444,
+      "step": 7701
+    },
+    {
+      "epoch": 2.133518005540166,
+      "grad_norm": 0.7317376136779785,
+      "learning_rate": 3.6167621743228886e-06,
+      "loss": 0.6535,
+      "step": 7702
+    },
+    {
+      "epoch": 2.1337950138504156,
+      "grad_norm": 0.6807712316513062,
+      "learning_rate": 3.61643624358727e-06,
+      "loss": 0.5867,
+      "step": 7703
+    },
+    {
+      "epoch": 2.134072022160665,
+      "grad_norm": 0.7171390056610107,
+      "learning_rate": 3.6161102891468598e-06,
+      "loss": 0.5741,
+      "step": 7704
+    },
+    {
+      "epoch": 2.134349030470914,
+      "grad_norm": 0.7733952403068542,
+      "learning_rate": 3.6157843110085768e-06,
+      "loss": 0.5826,
+      "step": 7705
+    },
+    {
+      "epoch": 2.1346260387811635,
+      "grad_norm": 0.7412691712379456,
+      "learning_rate": 3.6154583091793434e-06,
+      "loss": 0.6361,
+      "step": 7706
+    },
+    {
+      "epoch": 2.1349030470914125,
+      "grad_norm": 0.6959484219551086,
+      "learning_rate": 3.615132283666083e-06,
+      "loss": 0.6095,
+      "step": 7707
+    },
+    {
+      "epoch": 2.135180055401662,
+      "grad_norm": 0.742755651473999,
+      "learning_rate": 3.6148062344757145e-06,
+      "loss": 0.6281,
+      "step": 7708
+    },
+    {
+      "epoch": 2.1354570637119115,
+      "grad_norm": 0.767560601234436,
+      "learning_rate": 3.6144801616151642e-06,
+      "loss": 0.6221,
+      "step": 7709
+    },
+    {
+      "epoch": 2.1357340720221605,
+      "grad_norm": 0.7019870281219482,
+      "learning_rate": 3.614154065091353e-06,
+      "loss": 0.5758,
+      "step": 7710
+    },
+    {
+      "epoch": 2.13601108033241,
+      "grad_norm": 0.7854132056236267,
+      "learning_rate": 3.613827944911207e-06,
+      "loss": 0.5995,
+      "step": 7711
+    },
+    {
+      "epoch": 2.1362880886426594,
+      "grad_norm": 0.7392581105232239,
+      "learning_rate": 3.613501801081648e-06,
+      "loss": 0.6336,
+      "step": 7712
+    },
+    {
+      "epoch": 2.1365650969529084,
+      "grad_norm": 0.7203420400619507,
+      "learning_rate": 3.613175633609603e-06,
+      "loss": 0.5868,
+      "step": 7713
+    },
+    {
+      "epoch": 2.136842105263158,
+      "grad_norm": 0.7442747950553894,
+      "learning_rate": 3.6128494425019966e-06,
+      "loss": 0.608,
+      "step": 7714
+    },
+    {
+      "epoch": 2.1371191135734073,
+      "grad_norm": 0.7600914239883423,
+      "learning_rate": 3.612523227765754e-06,
+      "loss": 0.6087,
+      "step": 7715
+    },
+    {
+      "epoch": 2.1373961218836564,
+      "grad_norm": 0.6972166895866394,
+      "learning_rate": 3.612196989407802e-06,
+      "loss": 0.6376,
+      "step": 7716
+    },
+    {
+      "epoch": 2.137673130193906,
+      "grad_norm": 0.7419806122779846,
+      "learning_rate": 3.6118707274350695e-06,
+      "loss": 0.6143,
+      "step": 7717
+    },
+    {
+      "epoch": 2.1379501385041553,
+      "grad_norm": 0.690531849861145,
+      "learning_rate": 3.6115444418544803e-06,
+      "loss": 0.5944,
+      "step": 7718
+    },
+    {
+      "epoch": 2.1382271468144043,
+      "grad_norm": 0.7530912160873413,
+      "learning_rate": 3.6112181326729647e-06,
+      "loss": 0.6317,
+      "step": 7719
+    },
+    {
+      "epoch": 2.1385041551246537,
+      "grad_norm": 0.7327772378921509,
+      "learning_rate": 3.61089179989745e-06,
+      "loss": 0.6371,
+      "step": 7720
+    },
+    {
+      "epoch": 2.138781163434903,
+      "grad_norm": 0.7340015172958374,
+      "learning_rate": 3.610565443534865e-06,
+      "loss": 0.5727,
+      "step": 7721
+    },
+    {
+      "epoch": 2.1390581717451522,
+      "grad_norm": 0.7161919474601746,
+      "learning_rate": 3.610239063592141e-06,
+      "loss": 0.5914,
+      "step": 7722
+    },
+    {
+      "epoch": 2.1393351800554017,
+      "grad_norm": 0.721517026424408,
+      "learning_rate": 3.6099126600762056e-06,
+      "loss": 0.6026,
+      "step": 7723
+    },
+    {
+      "epoch": 2.139612188365651,
+      "grad_norm": 0.7476277351379395,
+      "learning_rate": 3.60958623299399e-06,
+      "loss": 0.6114,
+      "step": 7724
+    },
+    {
+      "epoch": 2.1398891966759,
+      "grad_norm": 0.7581577897071838,
+      "learning_rate": 3.609259782352425e-06,
+      "loss": 0.5867,
+      "step": 7725
+    },
+    {
+      "epoch": 2.1401662049861496,
+      "grad_norm": 0.7624351978302002,
+      "learning_rate": 3.6089333081584434e-06,
+      "loss": 0.5996,
+      "step": 7726
+    },
+    {
+      "epoch": 2.140443213296399,
+      "grad_norm": 0.7194927334785461,
+      "learning_rate": 3.608606810418974e-06,
+      "loss": 0.5811,
+      "step": 7727
+    },
+    {
+      "epoch": 2.140720221606648,
+      "grad_norm": 0.752798318862915,
+      "learning_rate": 3.608280289140952e-06,
+      "loss": 0.6557,
+      "step": 7728
+    },
+    {
+      "epoch": 2.1409972299168976,
+      "grad_norm": 0.7455686330795288,
+      "learning_rate": 3.6079537443313084e-06,
+      "loss": 0.6307,
+      "step": 7729
+    },
+    {
+      "epoch": 2.141274238227147,
+      "grad_norm": 0.7468121647834778,
+      "learning_rate": 3.6076271759969785e-06,
+      "loss": 0.6192,
+      "step": 7730
+    },
+    {
+      "epoch": 2.141551246537396,
+      "grad_norm": 0.7016478180885315,
+      "learning_rate": 3.6073005841448936e-06,
+      "loss": 0.5576,
+      "step": 7731
+    },
+    {
+      "epoch": 2.1418282548476455,
+      "grad_norm": 0.6816582083702087,
+      "learning_rate": 3.6069739687819905e-06,
+      "loss": 0.5899,
+      "step": 7732
+    },
+    {
+      "epoch": 2.1421052631578945,
+      "grad_norm": 0.6748509407043457,
+      "learning_rate": 3.6066473299152025e-06,
+      "loss": 0.6112,
+      "step": 7733
+    },
+    {
+      "epoch": 2.142382271468144,
+      "grad_norm": 0.7195847034454346,
+      "learning_rate": 3.606320667551466e-06,
+      "loss": 0.5903,
+      "step": 7734
+    },
+    {
+      "epoch": 2.1426592797783934,
+      "grad_norm": 0.7213793396949768,
+      "learning_rate": 3.6059939816977167e-06,
+      "loss": 0.6142,
+      "step": 7735
+    },
+    {
+      "epoch": 2.1429362880886424,
+      "grad_norm": 0.7996098399162292,
+      "learning_rate": 3.605667272360891e-06,
+      "loss": 0.6043,
+      "step": 7736
+    },
+    {
+      "epoch": 2.143213296398892,
+      "grad_norm": 0.7544435858726501,
+      "learning_rate": 3.605340539547925e-06,
+      "loss": 0.6007,
+      "step": 7737
+    },
+    {
+      "epoch": 2.1434903047091414,
+      "grad_norm": 0.7508959770202637,
+      "learning_rate": 3.6050137832657565e-06,
+      "loss": 0.6343,
+      "step": 7738
+    },
+    {
+      "epoch": 2.1437673130193904,
+      "grad_norm": 0.7130975127220154,
+      "learning_rate": 3.6046870035213246e-06,
+      "loss": 0.6139,
+      "step": 7739
+    },
+    {
+      "epoch": 2.14404432132964,
+      "grad_norm": 0.7473478317260742,
+      "learning_rate": 3.6043602003215657e-06,
+      "loss": 0.6321,
+      "step": 7740
+    },
+    {
+      "epoch": 2.1443213296398893,
+      "grad_norm": 0.7028003334999084,
+      "learning_rate": 3.6040333736734195e-06,
+      "loss": 0.5854,
+      "step": 7741
+    },
+    {
+      "epoch": 2.1445983379501383,
+      "grad_norm": 0.7346315383911133,
+      "learning_rate": 3.6037065235838255e-06,
+      "loss": 0.596,
+      "step": 7742
+    },
+    {
+      "epoch": 2.144875346260388,
+      "grad_norm": 0.6991076469421387,
+      "learning_rate": 3.6033796500597236e-06,
+      "loss": 0.564,
+      "step": 7743
+    },
+    {
+      "epoch": 2.1451523545706372,
+      "grad_norm": 0.7503802180290222,
+      "learning_rate": 3.6030527531080533e-06,
+      "loss": 0.6233,
+      "step": 7744
+    },
+    {
+      "epoch": 2.1454293628808863,
+      "grad_norm": 0.718914270401001,
+      "learning_rate": 3.6027258327357573e-06,
+      "loss": 0.6037,
+      "step": 7745
+    },
+    {
+      "epoch": 2.1457063711911357,
+      "grad_norm": 0.7305037975311279,
+      "learning_rate": 3.6023988889497753e-06,
+      "loss": 0.5828,
+      "step": 7746
+    },
+    {
+      "epoch": 2.145983379501385,
+      "grad_norm": 0.7057591080665588,
+      "learning_rate": 3.602071921757049e-06,
+      "loss": 0.6097,
+      "step": 7747
+    },
+    {
+      "epoch": 2.146260387811634,
+      "grad_norm": 0.6981334090232849,
+      "learning_rate": 3.601744931164522e-06,
+      "loss": 0.5359,
+      "step": 7748
+    },
+    {
+      "epoch": 2.1465373961218837,
+      "grad_norm": 0.7055646777153015,
+      "learning_rate": 3.6014179171791365e-06,
+      "loss": 0.5552,
+      "step": 7749
+    },
+    {
+      "epoch": 2.146814404432133,
+      "grad_norm": 0.7171012163162231,
+      "learning_rate": 3.601090879807836e-06,
+      "loss": 0.619,
+      "step": 7750
+    },
+    {
+      "epoch": 2.147091412742382,
+      "grad_norm": 0.7078838348388672,
+      "learning_rate": 3.6007638190575644e-06,
+      "loss": 0.5707,
+      "step": 7751
+    },
+    {
+      "epoch": 2.1473684210526316,
+      "grad_norm": 0.731663167476654,
+      "learning_rate": 3.6004367349352658e-06,
+      "loss": 0.6335,
+      "step": 7752
+    },
+    {
+      "epoch": 2.147645429362881,
+      "grad_norm": 0.6805864572525024,
+      "learning_rate": 3.6001096274478847e-06,
+      "loss": 0.5926,
+      "step": 7753
+    },
+    {
+      "epoch": 2.14792243767313,
+      "grad_norm": 0.7225302457809448,
+      "learning_rate": 3.599782496602367e-06,
+      "loss": 0.6214,
+      "step": 7754
+    },
+    {
+      "epoch": 2.1481994459833795,
+      "grad_norm": 0.7431020736694336,
+      "learning_rate": 3.599455342405659e-06,
+      "loss": 0.598,
+      "step": 7755
+    },
+    {
+      "epoch": 2.148476454293629,
+      "grad_norm": 0.7166742086410522,
+      "learning_rate": 3.599128164864706e-06,
+      "loss": 0.6303,
+      "step": 7756
+    },
+    {
+      "epoch": 2.148753462603878,
+      "grad_norm": 0.7254209518432617,
+      "learning_rate": 3.598800963986454e-06,
+      "loss": 0.556,
+      "step": 7757
+    },
+    {
+      "epoch": 2.1490304709141275,
+      "grad_norm": 0.7554706931114197,
+      "learning_rate": 3.5984737397778537e-06,
+      "loss": 0.608,
+      "step": 7758
+    },
+    {
+      "epoch": 2.149307479224377,
+      "grad_norm": 0.7507380247116089,
+      "learning_rate": 3.5981464922458488e-06,
+      "loss": 0.5717,
+      "step": 7759
+    },
+    {
+      "epoch": 2.149584487534626,
+      "grad_norm": 0.7267554402351379,
+      "learning_rate": 3.5978192213973904e-06,
+      "loss": 0.6567,
+      "step": 7760
+    },
+    {
+      "epoch": 2.1498614958448754,
+      "grad_norm": 0.692969560623169,
+      "learning_rate": 3.5974919272394265e-06,
+      "loss": 0.6093,
+      "step": 7761
+    },
+    {
+      "epoch": 2.150138504155125,
+      "grad_norm": 0.7091807126998901,
+      "learning_rate": 3.597164609778906e-06,
+      "loss": 0.5593,
+      "step": 7762
+    },
+    {
+      "epoch": 2.150415512465374,
+      "grad_norm": 0.7118282914161682,
+      "learning_rate": 3.596837269022779e-06,
+      "loss": 0.6712,
+      "step": 7763
+    },
+    {
+      "epoch": 2.1506925207756233,
+      "grad_norm": 0.7038248777389526,
+      "learning_rate": 3.5965099049779954e-06,
+      "loss": 0.5974,
+      "step": 7764
+    },
+    {
+      "epoch": 2.150969529085873,
+      "grad_norm": 0.7268661260604858,
+      "learning_rate": 3.596182517651507e-06,
+      "loss": 0.6084,
+      "step": 7765
+    },
+    {
+      "epoch": 2.151246537396122,
+      "grad_norm": 0.7075693607330322,
+      "learning_rate": 3.595855107050264e-06,
+      "loss": 0.6143,
+      "step": 7766
+    },
+    {
+      "epoch": 2.1515235457063713,
+      "grad_norm": 0.7023481130599976,
+      "learning_rate": 3.595527673181219e-06,
+      "loss": 0.5946,
+      "step": 7767
+    },
+    {
+      "epoch": 2.1518005540166203,
+      "grad_norm": 0.7299481630325317,
+      "learning_rate": 3.5952002160513233e-06,
+      "loss": 0.6691,
+      "step": 7768
+    },
+    {
+      "epoch": 2.1520775623268698,
+      "grad_norm": 0.697287380695343,
+      "learning_rate": 3.5948727356675305e-06,
+      "loss": 0.609,
+      "step": 7769
+    },
+    {
+      "epoch": 2.152354570637119,
+      "grad_norm": 0.7602940201759338,
+      "learning_rate": 3.594545232036793e-06,
+      "loss": 0.643,
+      "step": 7770
+    },
+    {
+      "epoch": 2.1526315789473682,
+      "grad_norm": 0.6859704852104187,
+      "learning_rate": 3.5942177051660663e-06,
+      "loss": 0.6061,
+      "step": 7771
+    },
+    {
+      "epoch": 2.1529085872576177,
+      "grad_norm": 0.7591689825057983,
+      "learning_rate": 3.5938901550623022e-06,
+      "loss": 0.629,
+      "step": 7772
+    },
+    {
+      "epoch": 2.153185595567867,
+      "grad_norm": 0.7486093044281006,
+      "learning_rate": 3.5935625817324572e-06,
+      "loss": 0.6165,
+      "step": 7773
+    },
+    {
+      "epoch": 2.153462603878116,
+      "grad_norm": 0.7580902576446533,
+      "learning_rate": 3.593234985183486e-06,
+      "loss": 0.6549,
+      "step": 7774
+    },
+    {
+      "epoch": 2.1537396121883656,
+      "grad_norm": 0.717231273651123,
+      "learning_rate": 3.592907365422344e-06,
+      "loss": 0.5984,
+      "step": 7775
+    },
+    {
+      "epoch": 2.154016620498615,
+      "grad_norm": 0.7410253286361694,
+      "learning_rate": 3.592579722455987e-06,
+      "loss": 0.5982,
+      "step": 7776
+    },
+    {
+      "epoch": 2.154293628808864,
+      "grad_norm": 0.7209257483482361,
+      "learning_rate": 3.592252056291373e-06,
+      "loss": 0.5881,
+      "step": 7777
+    },
+    {
+      "epoch": 2.1545706371191136,
+      "grad_norm": 0.7834051251411438,
+      "learning_rate": 3.5919243669354585e-06,
+      "loss": 0.6453,
+      "step": 7778
+    },
+    {
+      "epoch": 2.154847645429363,
+      "grad_norm": 0.7775634527206421,
+      "learning_rate": 3.5915966543952006e-06,
+      "loss": 0.6243,
+      "step": 7779
+    },
+    {
+      "epoch": 2.155124653739612,
+      "grad_norm": 0.7410075068473816,
+      "learning_rate": 3.5912689186775594e-06,
+      "loss": 0.6137,
+      "step": 7780
+    },
+    {
+      "epoch": 2.1554016620498615,
+      "grad_norm": 0.7614990472793579,
+      "learning_rate": 3.5909411597894916e-06,
+      "loss": 0.5606,
+      "step": 7781
+    },
+    {
+      "epoch": 2.155678670360111,
+      "grad_norm": 0.7304420471191406,
+      "learning_rate": 3.5906133777379572e-06,
+      "loss": 0.5241,
+      "step": 7782
+    },
+    {
+      "epoch": 2.15595567867036,
+      "grad_norm": 0.7032648921012878,
+      "learning_rate": 3.5902855725299156e-06,
+      "loss": 0.625,
+      "step": 7783
+    },
+    {
+      "epoch": 2.1562326869806094,
+      "grad_norm": 0.7103816866874695,
+      "learning_rate": 3.589957744172327e-06,
+      "loss": 0.5811,
+      "step": 7784
+    },
+    {
+      "epoch": 2.156509695290859,
+      "grad_norm": 0.7315899133682251,
+      "learning_rate": 3.589629892672152e-06,
+      "loss": 0.6132,
+      "step": 7785
+    },
+    {
+      "epoch": 2.156786703601108,
+      "grad_norm": 0.7375378608703613,
+      "learning_rate": 3.5893020180363513e-06,
+      "loss": 0.6179,
+      "step": 7786
+    },
+    {
+      "epoch": 2.1570637119113574,
+      "grad_norm": 0.7213611602783203,
+      "learning_rate": 3.5889741202718874e-06,
+      "loss": 0.571,
+      "step": 7787
+    },
+    {
+      "epoch": 2.157340720221607,
+      "grad_norm": 0.7085563540458679,
+      "learning_rate": 3.5886461993857225e-06,
+      "loss": 0.6045,
+      "step": 7788
+    },
+    {
+      "epoch": 2.157617728531856,
+      "grad_norm": 0.7195646166801453,
+      "learning_rate": 3.588318255384817e-06,
+      "loss": 0.5944,
+      "step": 7789
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.7459489107131958,
+      "learning_rate": 3.5879902882761375e-06,
+      "loss": 0.5727,
+      "step": 7790
+    },
+    {
+      "epoch": 2.1581717451523548,
+      "grad_norm": 0.7280673980712891,
+      "learning_rate": 3.587662298066645e-06,
+      "loss": 0.5796,
+      "step": 7791
+    },
+    {
+      "epoch": 2.158448753462604,
+      "grad_norm": 0.7652393579483032,
+      "learning_rate": 3.587334284763304e-06,
+      "loss": 0.5919,
+      "step": 7792
+    },
+    {
+      "epoch": 2.1587257617728532,
+      "grad_norm": 0.6975575685501099,
+      "learning_rate": 3.5870062483730807e-06,
+      "loss": 0.5909,
+      "step": 7793
+    },
+    {
+      "epoch": 2.1590027700831023,
+      "grad_norm": 0.7558895945549011,
+      "learning_rate": 3.586678188902937e-06,
+      "loss": 0.5739,
+      "step": 7794
+    },
+    {
+      "epoch": 2.1592797783933517,
+      "grad_norm": 0.6866469383239746,
+      "learning_rate": 3.5863501063598418e-06,
+      "loss": 0.554,
+      "step": 7795
+    },
+    {
+      "epoch": 2.159556786703601,
+      "grad_norm": 0.7125341892242432,
+      "learning_rate": 3.586022000750759e-06,
+      "loss": 0.6129,
+      "step": 7796
+    },
+    {
+      "epoch": 2.15983379501385,
+      "grad_norm": 0.735459566116333,
+      "learning_rate": 3.585693872082656e-06,
+      "loss": 0.5721,
+      "step": 7797
+    },
+    {
+      "epoch": 2.1601108033240997,
+      "grad_norm": 0.7106569409370422,
+      "learning_rate": 3.5853657203624985e-06,
+      "loss": 0.6055,
+      "step": 7798
+    },
+    {
+      "epoch": 2.160387811634349,
+      "grad_norm": 0.7532612085342407,
+      "learning_rate": 3.585037545597256e-06,
+      "loss": 0.5598,
+      "step": 7799
+    },
+    {
+      "epoch": 2.160664819944598,
+      "grad_norm": 0.7423479557037354,
+      "learning_rate": 3.5847093477938955e-06,
+      "loss": 0.6258,
+      "step": 7800
+    },
+    {
+      "epoch": 2.1609418282548476,
+      "grad_norm": 0.7185900807380676,
+      "learning_rate": 3.5843811269593854e-06,
+      "loss": 0.5885,
+      "step": 7801
+    },
+    {
+      "epoch": 2.161218836565097,
+      "grad_norm": 0.7055425047874451,
+      "learning_rate": 3.584052883100695e-06,
+      "loss": 0.6391,
+      "step": 7802
+    },
+    {
+      "epoch": 2.161495844875346,
+      "grad_norm": 0.7540387511253357,
+      "learning_rate": 3.5837246162247937e-06,
+      "loss": 0.6234,
+      "step": 7803
+    },
+    {
+      "epoch": 2.1617728531855955,
+      "grad_norm": 0.6934828758239746,
+      "learning_rate": 3.58339632633865e-06,
+      "loss": 0.5873,
+      "step": 7804
+    },
+    {
+      "epoch": 2.162049861495845,
+      "grad_norm": 0.7087048292160034,
+      "learning_rate": 3.583068013449237e-06,
+      "loss": 0.6268,
+      "step": 7805
+    },
+    {
+      "epoch": 2.162326869806094,
+      "grad_norm": 0.7429408431053162,
+      "learning_rate": 3.5827396775635232e-06,
+      "loss": 0.6136,
+      "step": 7806
+    },
+    {
+      "epoch": 2.1626038781163435,
+      "grad_norm": 0.7649708986282349,
+      "learning_rate": 3.582411318688481e-06,
+      "loss": 0.6156,
+      "step": 7807
+    },
+    {
+      "epoch": 2.162880886426593,
+      "grad_norm": 0.724372386932373,
+      "learning_rate": 3.582082936831083e-06,
+      "loss": 0.5765,
+      "step": 7808
+    },
+    {
+      "epoch": 2.163157894736842,
+      "grad_norm": 0.7201005220413208,
+      "learning_rate": 3.581754531998301e-06,
+      "loss": 0.591,
+      "step": 7809
+    },
+    {
+      "epoch": 2.1634349030470914,
+      "grad_norm": 0.6978209018707275,
+      "learning_rate": 3.5814261041971075e-06,
+      "loss": 0.5794,
+      "step": 7810
+    },
+    {
+      "epoch": 2.163711911357341,
+      "grad_norm": 0.727821409702301,
+      "learning_rate": 3.5810976534344756e-06,
+      "loss": 0.612,
+      "step": 7811
+    },
+    {
+      "epoch": 2.16398891966759,
+      "grad_norm": 0.6700218319892883,
+      "learning_rate": 3.580769179717381e-06,
+      "loss": 0.567,
+      "step": 7812
+    },
+    {
+      "epoch": 2.1642659279778393,
+      "grad_norm": 0.6753674745559692,
+      "learning_rate": 3.5804406830527955e-06,
+      "loss": 0.5484,
+      "step": 7813
+    },
+    {
+      "epoch": 2.164542936288089,
+      "grad_norm": 0.6982012391090393,
+      "learning_rate": 3.5801121634476955e-06,
+      "loss": 0.5868,
+      "step": 7814
+    },
+    {
+      "epoch": 2.164819944598338,
+      "grad_norm": 0.7340361475944519,
+      "learning_rate": 3.5797836209090564e-06,
+      "loss": 0.5949,
+      "step": 7815
+    },
+    {
+      "epoch": 2.1650969529085873,
+      "grad_norm": 0.7365831732749939,
+      "learning_rate": 3.5794550554438532e-06,
+      "loss": 0.5934,
+      "step": 7816
+    },
+    {
+      "epoch": 2.1653739612188367,
+      "grad_norm": 0.7187327146530151,
+      "learning_rate": 3.579126467059062e-06,
+      "loss": 0.5934,
+      "step": 7817
+    },
+    {
+      "epoch": 2.1656509695290858,
+      "grad_norm": 0.7340055108070374,
+      "learning_rate": 3.578797855761661e-06,
+      "loss": 0.6309,
+      "step": 7818
+    },
+    {
+      "epoch": 2.165927977839335,
+      "grad_norm": 0.7367265224456787,
+      "learning_rate": 3.5784692215586265e-06,
+      "loss": 0.6118,
+      "step": 7819
+    },
+    {
+      "epoch": 2.1662049861495847,
+      "grad_norm": 0.7496496438980103,
+      "learning_rate": 3.5781405644569356e-06,
+      "loss": 0.5916,
+      "step": 7820
+    },
+    {
+      "epoch": 2.1664819944598337,
+      "grad_norm": 0.7447252869606018,
+      "learning_rate": 3.5778118844635678e-06,
+      "loss": 0.5563,
+      "step": 7821
+    },
+    {
+      "epoch": 2.166759002770083,
+      "grad_norm": 0.7554126381874084,
+      "learning_rate": 3.5774831815855017e-06,
+      "loss": 0.5793,
+      "step": 7822
+    },
+    {
+      "epoch": 2.1670360110803326,
+      "grad_norm": 0.7430973052978516,
+      "learning_rate": 3.5771544558297157e-06,
+      "loss": 0.5782,
+      "step": 7823
+    },
+    {
+      "epoch": 2.1673130193905816,
+      "grad_norm": 0.7045478820800781,
+      "learning_rate": 3.5768257072031888e-06,
+      "loss": 0.618,
+      "step": 7824
+    },
+    {
+      "epoch": 2.167590027700831,
+      "grad_norm": 0.7048739194869995,
+      "learning_rate": 3.5764969357129036e-06,
+      "loss": 0.6288,
+      "step": 7825
+    },
+    {
+      "epoch": 2.1678670360110806,
+      "grad_norm": 0.7383768558502197,
+      "learning_rate": 3.5761681413658387e-06,
+      "loss": 0.5781,
+      "step": 7826
+    },
+    {
+      "epoch": 2.1681440443213296,
+      "grad_norm": 0.7174773216247559,
+      "learning_rate": 3.575839324168976e-06,
+      "loss": 0.6336,
+      "step": 7827
+    },
+    {
+      "epoch": 2.168421052631579,
+      "grad_norm": 0.713789701461792,
+      "learning_rate": 3.5755104841292977e-06,
+      "loss": 0.607,
+      "step": 7828
+    },
+    {
+      "epoch": 2.1686980609418285,
+      "grad_norm": 0.7295719981193542,
+      "learning_rate": 3.575181621253785e-06,
+      "loss": 0.6057,
+      "step": 7829
+    },
+    {
+      "epoch": 2.1689750692520775,
+      "grad_norm": 0.7278553247451782,
+      "learning_rate": 3.57485273554942e-06,
+      "loss": 0.5967,
+      "step": 7830
+    },
+    {
+      "epoch": 2.169252077562327,
+      "grad_norm": 0.7679740786552429,
+      "learning_rate": 3.5745238270231875e-06,
+      "loss": 0.6474,
+      "step": 7831
+    },
+    {
+      "epoch": 2.169529085872576,
+      "grad_norm": 0.6951683759689331,
+      "learning_rate": 3.5741948956820693e-06,
+      "loss": 0.5941,
+      "step": 7832
+    },
+    {
+      "epoch": 2.1698060941828254,
+      "grad_norm": 0.76551353931427,
+      "learning_rate": 3.57386594153305e-06,
+      "loss": 0.5924,
+      "step": 7833
+    },
+    {
+      "epoch": 2.170083102493075,
+      "grad_norm": 0.7162626385688782,
+      "learning_rate": 3.5735369645831157e-06,
+      "loss": 0.6041,
+      "step": 7834
+    },
+    {
+      "epoch": 2.170360110803324,
+      "grad_norm": 0.6925835013389587,
+      "learning_rate": 3.573207964839249e-06,
+      "loss": 0.6008,
+      "step": 7835
+    },
+    {
+      "epoch": 2.1706371191135734,
+      "grad_norm": 0.7028840184211731,
+      "learning_rate": 3.5728789423084375e-06,
+      "loss": 0.6002,
+      "step": 7836
+    },
+    {
+      "epoch": 2.170914127423823,
+      "grad_norm": 0.7359599471092224,
+      "learning_rate": 3.5725498969976657e-06,
+      "loss": 0.5797,
+      "step": 7837
+    },
+    {
+      "epoch": 2.171191135734072,
+      "grad_norm": 0.724312961101532,
+      "learning_rate": 3.57222082891392e-06,
+      "loss": 0.6589,
+      "step": 7838
+    },
+    {
+      "epoch": 2.1714681440443213,
+      "grad_norm": 0.7298066020011902,
+      "learning_rate": 3.571891738064189e-06,
+      "loss": 0.5953,
+      "step": 7839
+    },
+    {
+      "epoch": 2.1717451523545708,
+      "grad_norm": 0.6635059118270874,
+      "learning_rate": 3.5715626244554587e-06,
+      "loss": 0.5261,
+      "step": 7840
+    },
+    {
+      "epoch": 2.17202216066482,
+      "grad_norm": 0.7057240009307861,
+      "learning_rate": 3.5712334880947175e-06,
+      "loss": 0.5618,
+      "step": 7841
+    },
+    {
+      "epoch": 2.1722991689750693,
+      "grad_norm": 0.819150984287262,
+      "learning_rate": 3.5709043289889538e-06,
+      "loss": 0.625,
+      "step": 7842
+    },
+    {
+      "epoch": 2.1725761772853187,
+      "grad_norm": 0.7225316762924194,
+      "learning_rate": 3.5705751471451555e-06,
+      "loss": 0.6125,
+      "step": 7843
+    },
+    {
+      "epoch": 2.1728531855955677,
+      "grad_norm": 0.7002236843109131,
+      "learning_rate": 3.5702459425703146e-06,
+      "loss": 0.5824,
+      "step": 7844
+    },
+    {
+      "epoch": 2.173130193905817,
+      "grad_norm": 0.6795487999916077,
+      "learning_rate": 3.5699167152714174e-06,
+      "loss": 0.56,
+      "step": 7845
+    },
+    {
+      "epoch": 2.1734072022160666,
+      "grad_norm": 0.6594259142875671,
+      "learning_rate": 3.5695874652554573e-06,
+      "loss": 0.5474,
+      "step": 7846
+    },
+    {
+      "epoch": 2.1736842105263157,
+      "grad_norm": 0.7530267834663391,
+      "learning_rate": 3.5692581925294234e-06,
+      "loss": 0.5675,
+      "step": 7847
+    },
+    {
+      "epoch": 2.173961218836565,
+      "grad_norm": 0.7334387898445129,
+      "learning_rate": 3.5689288971003073e-06,
+      "loss": 0.6163,
+      "step": 7848
+    },
+    {
+      "epoch": 2.1742382271468146,
+      "grad_norm": 0.7736994624137878,
+      "learning_rate": 3.568599578975101e-06,
+      "loss": 0.6707,
+      "step": 7849
+    },
+    {
+      "epoch": 2.1745152354570636,
+      "grad_norm": 0.7506792545318604,
+      "learning_rate": 3.568270238160797e-06,
+      "loss": 0.5819,
+      "step": 7850
+    },
+    {
+      "epoch": 2.174792243767313,
+      "grad_norm": 0.7608615159988403,
+      "learning_rate": 3.567940874664388e-06,
+      "loss": 0.5939,
+      "step": 7851
+    },
+    {
+      "epoch": 2.1750692520775625,
+      "grad_norm": 0.7625887989997864,
+      "learning_rate": 3.5676114884928666e-06,
+      "loss": 0.571,
+      "step": 7852
+    },
+    {
+      "epoch": 2.1753462603878115,
+      "grad_norm": 0.7180339694023132,
+      "learning_rate": 3.567282079653227e-06,
+      "loss": 0.5816,
+      "step": 7853
+    },
+    {
+      "epoch": 2.175623268698061,
+      "grad_norm": 0.723455548286438,
+      "learning_rate": 3.566952648152463e-06,
+      "loss": 0.619,
+      "step": 7854
+    },
+    {
+      "epoch": 2.1759002770083105,
+      "grad_norm": 0.7430705428123474,
+      "learning_rate": 3.5666231939975702e-06,
+      "loss": 0.6143,
+      "step": 7855
+    },
+    {
+      "epoch": 2.1761772853185595,
+      "grad_norm": 0.7321606278419495,
+      "learning_rate": 3.566293717195543e-06,
+      "loss": 0.593,
+      "step": 7856
+    },
+    {
+      "epoch": 2.176454293628809,
+      "grad_norm": 0.7337496280670166,
+      "learning_rate": 3.5659642177533773e-06,
+      "loss": 0.5753,
+      "step": 7857
+    },
+    {
+      "epoch": 2.176731301939058,
+      "grad_norm": 0.7165916562080383,
+      "learning_rate": 3.5656346956780685e-06,
+      "loss": 0.6012,
+      "step": 7858
+    },
+    {
+      "epoch": 2.1770083102493074,
+      "grad_norm": 0.7456101179122925,
+      "learning_rate": 3.565305150976614e-06,
+      "loss": 0.6037,
+      "step": 7859
+    },
+    {
+      "epoch": 2.177285318559557,
+      "grad_norm": 0.7195954322814941,
+      "learning_rate": 3.5649755836560107e-06,
+      "loss": 0.6076,
+      "step": 7860
+    },
+    {
+      "epoch": 2.177562326869806,
+      "grad_norm": 0.7319379448890686,
+      "learning_rate": 3.5646459937232556e-06,
+      "loss": 0.5677,
+      "step": 7861
+    },
+    {
+      "epoch": 2.1778393351800553,
+      "grad_norm": 0.7634261846542358,
+      "learning_rate": 3.564316381185348e-06,
+      "loss": 0.5846,
+      "step": 7862
+    },
+    {
+      "epoch": 2.178116343490305,
+      "grad_norm": 0.7687897086143494,
+      "learning_rate": 3.563986746049286e-06,
+      "loss": 0.6219,
+      "step": 7863
+    },
+    {
+      "epoch": 2.178393351800554,
+      "grad_norm": 0.6961368918418884,
+      "learning_rate": 3.563657088322068e-06,
+      "loss": 0.6413,
+      "step": 7864
+    },
+    {
+      "epoch": 2.1786703601108033,
+      "grad_norm": 0.7764820456504822,
+      "learning_rate": 3.5633274080106927e-06,
+      "loss": 0.62,
+      "step": 7865
+    },
+    {
+      "epoch": 2.1789473684210527,
+      "grad_norm": 0.6963434219360352,
+      "learning_rate": 3.562997705122162e-06,
+      "loss": 0.572,
+      "step": 7866
+    },
+    {
+      "epoch": 2.1792243767313018,
+      "grad_norm": 0.7819361090660095,
+      "learning_rate": 3.562667979663475e-06,
+      "loss": 0.6084,
+      "step": 7867
+    },
+    {
+      "epoch": 2.179501385041551,
+      "grad_norm": 0.6757632493972778,
+      "learning_rate": 3.5623382316416334e-06,
+      "loss": 0.5393,
+      "step": 7868
+    },
+    {
+      "epoch": 2.1797783933518007,
+      "grad_norm": 0.690197765827179,
+      "learning_rate": 3.5620084610636385e-06,
+      "loss": 0.5816,
+      "step": 7869
+    },
+    {
+      "epoch": 2.1800554016620497,
+      "grad_norm": 0.7298179864883423,
+      "learning_rate": 3.561678667936491e-06,
+      "loss": 0.6169,
+      "step": 7870
+    },
+    {
+      "epoch": 2.180332409972299,
+      "grad_norm": 0.6915467381477356,
+      "learning_rate": 3.5613488522671947e-06,
+      "loss": 0.6357,
+      "step": 7871
+    },
+    {
+      "epoch": 2.1806094182825486,
+      "grad_norm": 0.7635214328765869,
+      "learning_rate": 3.561019014062752e-06,
+      "loss": 0.6146,
+      "step": 7872
+    },
+    {
+      "epoch": 2.1808864265927976,
+      "grad_norm": 0.737816572189331,
+      "learning_rate": 3.560689153330166e-06,
+      "loss": 0.6044,
+      "step": 7873
+    },
+    {
+      "epoch": 2.181163434903047,
+      "grad_norm": 0.6996750831604004,
+      "learning_rate": 3.5603592700764406e-06,
+      "loss": 0.5811,
+      "step": 7874
+    },
+    {
+      "epoch": 2.1814404432132966,
+      "grad_norm": 0.7186986804008484,
+      "learning_rate": 3.560029364308579e-06,
+      "loss": 0.6217,
+      "step": 7875
+    },
+    {
+      "epoch": 2.1817174515235456,
+      "grad_norm": 0.740057110786438,
+      "learning_rate": 3.5596994360335883e-06,
+      "loss": 0.5861,
+      "step": 7876
+    },
+    {
+      "epoch": 2.181994459833795,
+      "grad_norm": 0.7343260049819946,
+      "learning_rate": 3.559369485258472e-06,
+      "loss": 0.5851,
+      "step": 7877
+    },
+    {
+      "epoch": 2.1822714681440445,
+      "grad_norm": 0.7151075005531311,
+      "learning_rate": 3.559039511990236e-06,
+      "loss": 0.563,
+      "step": 7878
+    },
+    {
+      "epoch": 2.1825484764542935,
+      "grad_norm": 0.7891567945480347,
+      "learning_rate": 3.5587095162358874e-06,
+      "loss": 0.6416,
+      "step": 7879
+    },
+    {
+      "epoch": 2.182825484764543,
+      "grad_norm": 0.7027353644371033,
+      "learning_rate": 3.558379498002431e-06,
+      "loss": 0.6098,
+      "step": 7880
+    },
+    {
+      "epoch": 2.1831024930747924,
+      "grad_norm": 0.7250735759735107,
+      "learning_rate": 3.558049457296876e-06,
+      "loss": 0.5759,
+      "step": 7881
+    },
+    {
+      "epoch": 2.1833795013850414,
+      "grad_norm": 0.6830952763557434,
+      "learning_rate": 3.5577193941262288e-06,
+      "loss": 0.5817,
+      "step": 7882
+    },
+    {
+      "epoch": 2.183656509695291,
+      "grad_norm": 0.7518918514251709,
+      "learning_rate": 3.5573893084974974e-06,
+      "loss": 0.5948,
+      "step": 7883
+    },
+    {
+      "epoch": 2.1839335180055404,
+      "grad_norm": 0.6851316094398499,
+      "learning_rate": 3.5570592004176906e-06,
+      "loss": 0.6175,
+      "step": 7884
+    },
+    {
+      "epoch": 2.1842105263157894,
+      "grad_norm": 0.7158825397491455,
+      "learning_rate": 3.5567290698938185e-06,
+      "loss": 0.6003,
+      "step": 7885
+    },
+    {
+      "epoch": 2.184487534626039,
+      "grad_norm": 0.7178775072097778,
+      "learning_rate": 3.556398916932889e-06,
+      "loss": 0.5882,
+      "step": 7886
+    },
+    {
+      "epoch": 2.1847645429362883,
+      "grad_norm": 0.7214950323104858,
+      "learning_rate": 3.5560687415419128e-06,
+      "loss": 0.6367,
+      "step": 7887
+    },
+    {
+      "epoch": 2.1850415512465373,
+      "grad_norm": 0.73500657081604,
+      "learning_rate": 3.5557385437279e-06,
+      "loss": 0.5479,
+      "step": 7888
+    },
+    {
+      "epoch": 2.1853185595567868,
+      "grad_norm": 0.7137157320976257,
+      "learning_rate": 3.555408323497862e-06,
+      "loss": 0.5845,
+      "step": 7889
+    },
+    {
+      "epoch": 2.1855955678670362,
+      "grad_norm": 0.65468829870224,
+      "learning_rate": 3.5550780808588103e-06,
+      "loss": 0.5309,
+      "step": 7890
+    },
+    {
+      "epoch": 2.1858725761772853,
+      "grad_norm": 0.71401047706604,
+      "learning_rate": 3.554747815817756e-06,
+      "loss": 0.5621,
+      "step": 7891
+    },
+    {
+      "epoch": 2.1861495844875347,
+      "grad_norm": 0.753894567489624,
+      "learning_rate": 3.5544175283817127e-06,
+      "loss": 0.6106,
+      "step": 7892
+    },
+    {
+      "epoch": 2.1864265927977837,
+      "grad_norm": 0.7078069448471069,
+      "learning_rate": 3.5540872185576924e-06,
+      "loss": 0.577,
+      "step": 7893
+    },
+    {
+      "epoch": 2.186703601108033,
+      "grad_norm": 0.7022084593772888,
+      "learning_rate": 3.553756886352708e-06,
+      "loss": 0.5592,
+      "step": 7894
+    },
+    {
+      "epoch": 2.1869806094182827,
+      "grad_norm": 0.7479102611541748,
+      "learning_rate": 3.5534265317737755e-06,
+      "loss": 0.6316,
+      "step": 7895
+    },
+    {
+      "epoch": 2.1872576177285317,
+      "grad_norm": 0.7682097554206848,
+      "learning_rate": 3.5530961548279062e-06,
+      "loss": 0.6092,
+      "step": 7896
+    },
+    {
+      "epoch": 2.187534626038781,
+      "grad_norm": 0.7278096675872803,
+      "learning_rate": 3.5527657555221163e-06,
+      "loss": 0.5541,
+      "step": 7897
+    },
+    {
+      "epoch": 2.1878116343490306,
+      "grad_norm": 0.7436990141868591,
+      "learning_rate": 3.552435333863422e-06,
+      "loss": 0.5878,
+      "step": 7898
+    },
+    {
+      "epoch": 2.1880886426592796,
+      "grad_norm": 0.7513325810432434,
+      "learning_rate": 3.552104889858837e-06,
+      "loss": 0.6403,
+      "step": 7899
+    },
+    {
+      "epoch": 2.188365650969529,
+      "grad_norm": 0.7407289147377014,
+      "learning_rate": 3.5517744235153786e-06,
+      "loss": 0.5961,
+      "step": 7900
+    },
+    {
+      "epoch": 2.1886426592797785,
+      "grad_norm": 0.7015832662582397,
+      "learning_rate": 3.551443934840063e-06,
+      "loss": 0.6063,
+      "step": 7901
+    },
+    {
+      "epoch": 2.1889196675900275,
+      "grad_norm": 0.7211142778396606,
+      "learning_rate": 3.551113423839908e-06,
+      "loss": 0.5823,
+      "step": 7902
+    },
+    {
+      "epoch": 2.189196675900277,
+      "grad_norm": 0.7457058429718018,
+      "learning_rate": 3.5507828905219298e-06,
+      "loss": 0.6282,
+      "step": 7903
+    },
+    {
+      "epoch": 2.1894736842105265,
+      "grad_norm": 0.7473145723342896,
+      "learning_rate": 3.5504523348931486e-06,
+      "loss": 0.5861,
+      "step": 7904
+    },
+    {
+      "epoch": 2.1897506925207755,
+      "grad_norm": 0.7545790672302246,
+      "learning_rate": 3.550121756960581e-06,
+      "loss": 0.5869,
+      "step": 7905
+    },
+    {
+      "epoch": 2.190027700831025,
+      "grad_norm": 0.6933561563491821,
+      "learning_rate": 3.549791156731247e-06,
+      "loss": 0.5913,
+      "step": 7906
+    },
+    {
+      "epoch": 2.1903047091412744,
+      "grad_norm": 0.7041760087013245,
+      "learning_rate": 3.5494605342121663e-06,
+      "loss": 0.5895,
+      "step": 7907
+    },
+    {
+      "epoch": 2.1905817174515234,
+      "grad_norm": 0.7071885466575623,
+      "learning_rate": 3.549129889410358e-06,
+      "loss": 0.581,
+      "step": 7908
+    },
+    {
+      "epoch": 2.190858725761773,
+      "grad_norm": 0.7134894728660583,
+      "learning_rate": 3.5487992223328434e-06,
+      "loss": 0.6123,
+      "step": 7909
+    },
+    {
+      "epoch": 2.1911357340720223,
+      "grad_norm": 0.7097256779670715,
+      "learning_rate": 3.5484685329866424e-06,
+      "loss": 0.6277,
+      "step": 7910
+    },
+    {
+      "epoch": 2.1914127423822714,
+      "grad_norm": 0.7199460864067078,
+      "learning_rate": 3.5481378213787777e-06,
+      "loss": 0.6123,
+      "step": 7911
+    },
+    {
+      "epoch": 2.191689750692521,
+      "grad_norm": 0.724058985710144,
+      "learning_rate": 3.5478070875162697e-06,
+      "loss": 0.6141,
+      "step": 7912
+    },
+    {
+      "epoch": 2.1919667590027703,
+      "grad_norm": 0.7432721853256226,
+      "learning_rate": 3.547476331406141e-06,
+      "loss": 0.5756,
+      "step": 7913
+    },
+    {
+      "epoch": 2.1922437673130193,
+      "grad_norm": 0.7083652019500732,
+      "learning_rate": 3.5471455530554157e-06,
+      "loss": 0.5764,
+      "step": 7914
+    },
+    {
+      "epoch": 2.1925207756232687,
+      "grad_norm": 0.7218777537345886,
+      "learning_rate": 3.546814752471116e-06,
+      "loss": 0.6246,
+      "step": 7915
+    },
+    {
+      "epoch": 2.192797783933518,
+      "grad_norm": 0.7310037016868591,
+      "learning_rate": 3.5464839296602655e-06,
+      "loss": 0.6346,
+      "step": 7916
+    },
+    {
+      "epoch": 2.1930747922437672,
+      "grad_norm": 0.7280022501945496,
+      "learning_rate": 3.5461530846298894e-06,
+      "loss": 0.6128,
+      "step": 7917
+    },
+    {
+      "epoch": 2.1933518005540167,
+      "grad_norm": 0.7664794325828552,
+      "learning_rate": 3.545822217387011e-06,
+      "loss": 0.6016,
+      "step": 7918
+    },
+    {
+      "epoch": 2.1936288088642657,
+      "grad_norm": 0.7563583254814148,
+      "learning_rate": 3.545491327938656e-06,
+      "loss": 0.6326,
+      "step": 7919
+    },
+    {
+      "epoch": 2.193905817174515,
+      "grad_norm": 0.7406309843063354,
+      "learning_rate": 3.545160416291852e-06,
+      "loss": 0.612,
+      "step": 7920
+    },
+    {
+      "epoch": 2.1941828254847646,
+      "grad_norm": 0.7212445735931396,
+      "learning_rate": 3.5448294824536212e-06,
+      "loss": 0.6482,
+      "step": 7921
+    },
+    {
+      "epoch": 2.1944598337950136,
+      "grad_norm": 0.7566403150558472,
+      "learning_rate": 3.5444985264309933e-06,
+      "loss": 0.6009,
+      "step": 7922
+    },
+    {
+      "epoch": 2.194736842105263,
+      "grad_norm": 0.6928979158401489,
+      "learning_rate": 3.5441675482309947e-06,
+      "loss": 0.5822,
+      "step": 7923
+    },
+    {
+      "epoch": 2.1950138504155126,
+      "grad_norm": 0.7452521324157715,
+      "learning_rate": 3.543836547860652e-06,
+      "loss": 0.6114,
+      "step": 7924
+    },
+    {
+      "epoch": 2.1952908587257616,
+      "grad_norm": 0.7450066804885864,
+      "learning_rate": 3.543505525326993e-06,
+      "loss": 0.6043,
+      "step": 7925
+    },
+    {
+      "epoch": 2.195567867036011,
+      "grad_norm": 0.7499051690101624,
+      "learning_rate": 3.5431744806370475e-06,
+      "loss": 0.6264,
+      "step": 7926
+    },
+    {
+      "epoch": 2.1958448753462605,
+      "grad_norm": 0.7499408721923828,
+      "learning_rate": 3.542843413797844e-06,
+      "loss": 0.6628,
+      "step": 7927
+    },
+    {
+      "epoch": 2.1961218836565095,
+      "grad_norm": 0.7410886287689209,
+      "learning_rate": 3.5425123248164116e-06,
+      "loss": 0.6215,
+      "step": 7928
+    },
+    {
+      "epoch": 2.196398891966759,
+      "grad_norm": 0.6964012980461121,
+      "learning_rate": 3.54218121369978e-06,
+      "loss": 0.6098,
+      "step": 7929
+    },
+    {
+      "epoch": 2.1966759002770084,
+      "grad_norm": 0.7218595743179321,
+      "learning_rate": 3.5418500804549805e-06,
+      "loss": 0.6061,
+      "step": 7930
+    },
+    {
+      "epoch": 2.1969529085872574,
+      "grad_norm": 0.750812292098999,
+      "learning_rate": 3.5415189250890425e-06,
+      "loss": 0.6302,
+      "step": 7931
+    },
+    {
+      "epoch": 2.197229916897507,
+      "grad_norm": 0.7358490228652954,
+      "learning_rate": 3.541187747608998e-06,
+      "loss": 0.6109,
+      "step": 7932
+    },
+    {
+      "epoch": 2.1975069252077564,
+      "grad_norm": 0.7339686155319214,
+      "learning_rate": 3.5408565480218795e-06,
+      "loss": 0.6392,
+      "step": 7933
+    },
+    {
+      "epoch": 2.1977839335180054,
+      "grad_norm": 0.7436063885688782,
+      "learning_rate": 3.5405253263347173e-06,
+      "loss": 0.619,
+      "step": 7934
+    },
+    {
+      "epoch": 2.198060941828255,
+      "grad_norm": 0.8030242919921875,
+      "learning_rate": 3.5401940825545456e-06,
+      "loss": 0.6468,
+      "step": 7935
+    },
+    {
+      "epoch": 2.1983379501385043,
+      "grad_norm": 0.7203686833381653,
+      "learning_rate": 3.5398628166883976e-06,
+      "loss": 0.6196,
+      "step": 7936
+    },
+    {
+      "epoch": 2.1986149584487533,
+      "grad_norm": 0.7480223178863525,
+      "learning_rate": 3.5395315287433063e-06,
+      "loss": 0.6097,
+      "step": 7937
+    },
+    {
+      "epoch": 2.198891966759003,
+      "grad_norm": 0.6992270946502686,
+      "learning_rate": 3.5392002187263054e-06,
+      "loss": 0.5731,
+      "step": 7938
+    },
+    {
+      "epoch": 2.1991689750692522,
+      "grad_norm": 0.7402212023735046,
+      "learning_rate": 3.538868886644431e-06,
+      "loss": 0.5997,
+      "step": 7939
+    },
+    {
+      "epoch": 2.1994459833795013,
+      "grad_norm": 0.6975589990615845,
+      "learning_rate": 3.5385375325047167e-06,
+      "loss": 0.5753,
+      "step": 7940
+    },
+    {
+      "epoch": 2.1997229916897507,
+      "grad_norm": 0.6758665442466736,
+      "learning_rate": 3.5382061563141986e-06,
+      "loss": 0.6344,
+      "step": 7941
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7077716588973999,
+      "learning_rate": 3.5378747580799122e-06,
+      "loss": 0.692,
+      "step": 7942
+    },
+    {
+      "epoch": 2.200277008310249,
+      "grad_norm": 0.74322110414505,
+      "learning_rate": 3.537543337808894e-06,
+      "loss": 0.5733,
+      "step": 7943
+    },
+    {
+      "epoch": 2.2005540166204987,
+      "grad_norm": 0.7803390622138977,
+      "learning_rate": 3.5372118955081815e-06,
+      "loss": 0.6034,
+      "step": 7944
+    },
+    {
+      "epoch": 2.200831024930748,
+      "grad_norm": 0.7345275282859802,
+      "learning_rate": 3.5368804311848114e-06,
+      "loss": 0.6302,
+      "step": 7945
+    },
+    {
+      "epoch": 2.201108033240997,
+      "grad_norm": 0.7228006720542908,
+      "learning_rate": 3.536548944845822e-06,
+      "loss": 0.5631,
+      "step": 7946
+    },
+    {
+      "epoch": 2.2013850415512466,
+      "grad_norm": 0.7092156410217285,
+      "learning_rate": 3.5362174364982517e-06,
+      "loss": 0.5946,
+      "step": 7947
+    },
+    {
+      "epoch": 2.201662049861496,
+      "grad_norm": 0.7122862339019775,
+      "learning_rate": 3.535885906149139e-06,
+      "loss": 0.5755,
+      "step": 7948
+    },
+    {
+      "epoch": 2.201939058171745,
+      "grad_norm": 0.737549364566803,
+      "learning_rate": 3.5355543538055232e-06,
+      "loss": 0.6109,
+      "step": 7949
+    },
+    {
+      "epoch": 2.2022160664819945,
+      "grad_norm": 0.7759289741516113,
+      "learning_rate": 3.5352227794744437e-06,
+      "loss": 0.6329,
+      "step": 7950
+    },
+    {
+      "epoch": 2.202493074792244,
+      "grad_norm": 0.7205516695976257,
+      "learning_rate": 3.5348911831629405e-06,
+      "loss": 0.5825,
+      "step": 7951
+    },
+    {
+      "epoch": 2.202770083102493,
+      "grad_norm": 0.7226518988609314,
+      "learning_rate": 3.5345595648780562e-06,
+      "loss": 0.5624,
+      "step": 7952
+    },
+    {
+      "epoch": 2.2030470914127425,
+      "grad_norm": 0.7641230821609497,
+      "learning_rate": 3.5342279246268286e-06,
+      "loss": 0.6224,
+      "step": 7953
+    },
+    {
+      "epoch": 2.203324099722992,
+      "grad_norm": 0.7313389182090759,
+      "learning_rate": 3.533896262416302e-06,
+      "loss": 0.6215,
+      "step": 7954
+    },
+    {
+      "epoch": 2.203601108033241,
+      "grad_norm": 0.7629680633544922,
+      "learning_rate": 3.5335645782535174e-06,
+      "loss": 0.6266,
+      "step": 7955
+    },
+    {
+      "epoch": 2.2038781163434904,
+      "grad_norm": 0.7637547850608826,
+      "learning_rate": 3.5332328721455174e-06,
+      "loss": 0.6103,
+      "step": 7956
+    },
+    {
+      "epoch": 2.2041551246537394,
+      "grad_norm": 0.750389575958252,
+      "learning_rate": 3.5329011440993443e-06,
+      "loss": 0.5681,
+      "step": 7957
+    },
+    {
+      "epoch": 2.204432132963989,
+      "grad_norm": 0.7237061262130737,
+      "learning_rate": 3.532569394122043e-06,
+      "loss": 0.5848,
+      "step": 7958
+    },
+    {
+      "epoch": 2.2047091412742383,
+      "grad_norm": 0.7263107895851135,
+      "learning_rate": 3.5322376222206556e-06,
+      "loss": 0.6069,
+      "step": 7959
+    },
+    {
+      "epoch": 2.2049861495844874,
+      "grad_norm": 0.7003798484802246,
+      "learning_rate": 3.531905828402228e-06,
+      "loss": 0.6112,
+      "step": 7960
+    },
+    {
+      "epoch": 2.205263157894737,
+      "grad_norm": 0.7772372364997864,
+      "learning_rate": 3.5315740126738053e-06,
+      "loss": 0.593,
+      "step": 7961
+    },
+    {
+      "epoch": 2.2055401662049863,
+      "grad_norm": 0.7660946846008301,
+      "learning_rate": 3.5312421750424304e-06,
+      "loss": 0.5922,
+      "step": 7962
+    },
+    {
+      "epoch": 2.2058171745152353,
+      "grad_norm": 0.7145115733146667,
+      "learning_rate": 3.5309103155151512e-06,
+      "loss": 0.6184,
+      "step": 7963
+    },
+    {
+      "epoch": 2.2060941828254848,
+      "grad_norm": 0.7356007099151611,
+      "learning_rate": 3.530578434099014e-06,
+      "loss": 0.5603,
+      "step": 7964
+    },
+    {
+      "epoch": 2.206371191135734,
+      "grad_norm": 0.6965721249580383,
+      "learning_rate": 3.530246530801064e-06,
+      "loss": 0.5912,
+      "step": 7965
+    },
+    {
+      "epoch": 2.2066481994459832,
+      "grad_norm": 0.7097307443618774,
+      "learning_rate": 3.5299146056283496e-06,
+      "loss": 0.6278,
+      "step": 7966
+    },
+    {
+      "epoch": 2.2069252077562327,
+      "grad_norm": 0.7481800317764282,
+      "learning_rate": 3.529582658587918e-06,
+      "loss": 0.6229,
+      "step": 7967
+    },
+    {
+      "epoch": 2.207202216066482,
+      "grad_norm": 0.7750024795532227,
+      "learning_rate": 3.529250689686817e-06,
+      "loss": 0.5928,
+      "step": 7968
+    },
+    {
+      "epoch": 2.207479224376731,
+      "grad_norm": 0.6997442841529846,
+      "learning_rate": 3.5289186989320954e-06,
+      "loss": 0.5417,
+      "step": 7969
+    },
+    {
+      "epoch": 2.2077562326869806,
+      "grad_norm": 0.712722897529602,
+      "learning_rate": 3.5285866863308025e-06,
+      "loss": 0.595,
+      "step": 7970
+    },
+    {
+      "epoch": 2.20803324099723,
+      "grad_norm": 0.76949542760849,
+      "learning_rate": 3.5282546518899878e-06,
+      "loss": 0.6026,
+      "step": 7971
+    },
+    {
+      "epoch": 2.208310249307479,
+      "grad_norm": 0.8288016319274902,
+      "learning_rate": 3.5279225956167006e-06,
+      "loss": 0.5849,
+      "step": 7972
+    },
+    {
+      "epoch": 2.2085872576177286,
+      "grad_norm": 0.7565370202064514,
+      "learning_rate": 3.527590517517992e-06,
+      "loss": 0.5859,
+      "step": 7973
+    },
+    {
+      "epoch": 2.208864265927978,
+      "grad_norm": 0.7121152877807617,
+      "learning_rate": 3.527258417600912e-06,
+      "loss": 0.5988,
+      "step": 7974
+    },
+    {
+      "epoch": 2.209141274238227,
+      "grad_norm": 0.7364125847816467,
+      "learning_rate": 3.5269262958725124e-06,
+      "loss": 0.6228,
+      "step": 7975
+    },
+    {
+      "epoch": 2.2094182825484765,
+      "grad_norm": 0.7307944297790527,
+      "learning_rate": 3.5265941523398455e-06,
+      "loss": 0.6114,
+      "step": 7976
+    },
+    {
+      "epoch": 2.209695290858726,
+      "grad_norm": 0.7665871381759644,
+      "learning_rate": 3.5262619870099635e-06,
+      "loss": 0.6285,
+      "step": 7977
+    },
+    {
+      "epoch": 2.209972299168975,
+      "grad_norm": 0.7392860054969788,
+      "learning_rate": 3.5259297998899184e-06,
+      "loss": 0.5715,
+      "step": 7978
+    },
+    {
+      "epoch": 2.2102493074792244,
+      "grad_norm": 0.7718192338943481,
+      "learning_rate": 3.525597590986763e-06,
+      "loss": 0.616,
+      "step": 7979
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.7106596231460571,
+      "learning_rate": 3.5252653603075522e-06,
+      "loss": 0.6153,
+      "step": 7980
+    },
+    {
+      "epoch": 2.210803324099723,
+      "grad_norm": 0.7141216993331909,
+      "learning_rate": 3.5249331078593397e-06,
+      "loss": 0.6069,
+      "step": 7981
+    },
+    {
+      "epoch": 2.2110803324099724,
+      "grad_norm": 0.7057957649230957,
+      "learning_rate": 3.52460083364918e-06,
+      "loss": 0.5425,
+      "step": 7982
+    },
+    {
+      "epoch": 2.2113573407202214,
+      "grad_norm": 0.7577681541442871,
+      "learning_rate": 3.5242685376841273e-06,
+      "loss": 0.621,
+      "step": 7983
+    },
+    {
+      "epoch": 2.211634349030471,
+      "grad_norm": 0.72541743516922,
+      "learning_rate": 3.523936219971239e-06,
+      "loss": 0.5972,
+      "step": 7984
+    },
+    {
+      "epoch": 2.2119113573407203,
+      "grad_norm": 0.7325984239578247,
+      "learning_rate": 3.5236038805175686e-06,
+      "loss": 0.6067,
+      "step": 7985
+    },
+    {
+      "epoch": 2.2121883656509693,
+      "grad_norm": 0.752688467502594,
+      "learning_rate": 3.5232715193301746e-06,
+      "loss": 0.6298,
+      "step": 7986
+    },
+    {
+      "epoch": 2.212465373961219,
+      "grad_norm": 0.6812304258346558,
+      "learning_rate": 3.5229391364161137e-06,
+      "loss": 0.5075,
+      "step": 7987
+    },
+    {
+      "epoch": 2.2127423822714682,
+      "grad_norm": 0.7274405360221863,
+      "learning_rate": 3.5226067317824415e-06,
+      "loss": 0.5995,
+      "step": 7988
+    },
+    {
+      "epoch": 2.2130193905817173,
+      "grad_norm": 0.7464661598205566,
+      "learning_rate": 3.5222743054362173e-06,
+      "loss": 0.6198,
+      "step": 7989
+    },
+    {
+      "epoch": 2.2132963988919667,
+      "grad_norm": 0.761583149433136,
+      "learning_rate": 3.521941857384499e-06,
+      "loss": 0.56,
+      "step": 7990
+    },
+    {
+      "epoch": 2.213573407202216,
+      "grad_norm": 0.7192556262016296,
+      "learning_rate": 3.5216093876343452e-06,
+      "loss": 0.6106,
+      "step": 7991
+    },
+    {
+      "epoch": 2.213850415512465,
+      "grad_norm": 0.7213461995124817,
+      "learning_rate": 3.5212768961928153e-06,
+      "loss": 0.5608,
+      "step": 7992
+    },
+    {
+      "epoch": 2.2141274238227147,
+      "grad_norm": 0.7017572522163391,
+      "learning_rate": 3.52094438306697e-06,
+      "loss": 0.5734,
+      "step": 7993
+    },
+    {
+      "epoch": 2.214404432132964,
+      "grad_norm": 0.7463569641113281,
+      "learning_rate": 3.5206118482638664e-06,
+      "loss": 0.6096,
+      "step": 7994
+    },
+    {
+      "epoch": 2.214681440443213,
+      "grad_norm": 0.754392683506012,
+      "learning_rate": 3.5202792917905676e-06,
+      "loss": 0.608,
+      "step": 7995
+    },
+    {
+      "epoch": 2.2149584487534626,
+      "grad_norm": 0.7312411069869995,
+      "learning_rate": 3.5199467136541342e-06,
+      "loss": 0.5546,
+      "step": 7996
+    },
+    {
+      "epoch": 2.215235457063712,
+      "grad_norm": 0.7295164465904236,
+      "learning_rate": 3.519614113861627e-06,
+      "loss": 0.5974,
+      "step": 7997
+    },
+    {
+      "epoch": 2.215512465373961,
+      "grad_norm": 0.7471133470535278,
+      "learning_rate": 3.519281492420108e-06,
+      "loss": 0.6558,
+      "step": 7998
+    },
+    {
+      "epoch": 2.2157894736842105,
+      "grad_norm": 0.7575737237930298,
+      "learning_rate": 3.518948849336641e-06,
+      "loss": 0.6012,
+      "step": 7999
+    },
+    {
+      "epoch": 2.21606648199446,
+      "grad_norm": 0.6775148510932922,
+      "learning_rate": 3.5186161846182874e-06,
+      "loss": 0.5475,
+      "step": 8000
+    },
+    {
+      "epoch": 2.216343490304709,
+      "grad_norm": 0.7134628891944885,
+      "learning_rate": 3.518283498272111e-06,
+      "loss": 0.5482,
+      "step": 8001
+    },
+    {
+      "epoch": 2.2166204986149585,
+      "grad_norm": 0.7109372019767761,
+      "learning_rate": 3.5179507903051746e-06,
+      "loss": 0.6066,
+      "step": 8002
+    },
+    {
+      "epoch": 2.216897506925208,
+      "grad_norm": 0.7727257013320923,
+      "learning_rate": 3.5176180607245447e-06,
+      "loss": 0.6076,
+      "step": 8003
+    },
+    {
+      "epoch": 2.217174515235457,
+      "grad_norm": 0.7703292965888977,
+      "learning_rate": 3.5172853095372837e-06,
+      "loss": 0.6273,
+      "step": 8004
+    },
+    {
+      "epoch": 2.2174515235457064,
+      "grad_norm": 0.7161400318145752,
+      "learning_rate": 3.5169525367504574e-06,
+      "loss": 0.6014,
+      "step": 8005
+    },
+    {
+      "epoch": 2.217728531855956,
+      "grad_norm": 0.8021954298019409,
+      "learning_rate": 3.5166197423711336e-06,
+      "loss": 0.6468,
+      "step": 8006
+    },
+    {
+      "epoch": 2.218005540166205,
+      "grad_norm": 0.725152313709259,
+      "learning_rate": 3.516286926406375e-06,
+      "loss": 0.6367,
+      "step": 8007
+    },
+    {
+      "epoch": 2.2182825484764543,
+      "grad_norm": 0.7513183951377869,
+      "learning_rate": 3.51595408886325e-06,
+      "loss": 0.6373,
+      "step": 8008
+    },
+    {
+      "epoch": 2.218559556786704,
+      "grad_norm": 0.719607412815094,
+      "learning_rate": 3.515621229748825e-06,
+      "loss": 0.6255,
+      "step": 8009
+    },
+    {
+      "epoch": 2.218836565096953,
+      "grad_norm": 0.7222772240638733,
+      "learning_rate": 3.5152883490701682e-06,
+      "loss": 0.5828,
+      "step": 8010
+    },
+    {
+      "epoch": 2.2191135734072023,
+      "grad_norm": 0.7790684700012207,
+      "learning_rate": 3.5149554468343463e-06,
+      "loss": 0.5909,
+      "step": 8011
+    },
+    {
+      "epoch": 2.2193905817174517,
+      "grad_norm": 0.7023915648460388,
+      "learning_rate": 3.514622523048429e-06,
+      "loss": 0.5937,
+      "step": 8012
+    },
+    {
+      "epoch": 2.2196675900277008,
+      "grad_norm": 0.7843884229660034,
+      "learning_rate": 3.514289577719484e-06,
+      "loss": 0.6214,
+      "step": 8013
+    },
+    {
+      "epoch": 2.21994459833795,
+      "grad_norm": 0.7272974252700806,
+      "learning_rate": 3.5139566108545814e-06,
+      "loss": 0.6052,
+      "step": 8014
+    },
+    {
+      "epoch": 2.2202216066481997,
+      "grad_norm": 0.7483271956443787,
+      "learning_rate": 3.5136236224607906e-06,
+      "loss": 0.5648,
+      "step": 8015
+    },
+    {
+      "epoch": 2.2204986149584487,
+      "grad_norm": 0.7253241539001465,
+      "learning_rate": 3.513290612545181e-06,
+      "loss": 0.5772,
+      "step": 8016
+    },
+    {
+      "epoch": 2.220775623268698,
+      "grad_norm": 0.7203364372253418,
+      "learning_rate": 3.512957581114825e-06,
+      "loss": 0.5419,
+      "step": 8017
+    },
+    {
+      "epoch": 2.221052631578947,
+      "grad_norm": 0.7568876147270203,
+      "learning_rate": 3.5126245281767924e-06,
+      "loss": 0.645,
+      "step": 8018
+    },
+    {
+      "epoch": 2.2213296398891966,
+      "grad_norm": 0.7632686495780945,
+      "learning_rate": 3.5122914537381554e-06,
+      "loss": 0.606,
+      "step": 8019
+    },
+    {
+      "epoch": 2.221606648199446,
+      "grad_norm": 0.7100051045417786,
+      "learning_rate": 3.5119583578059845e-06,
+      "loss": 0.5833,
+      "step": 8020
+    },
+    {
+      "epoch": 2.221883656509695,
+      "grad_norm": 0.704388439655304,
+      "learning_rate": 3.5116252403873546e-06,
+      "loss": 0.5662,
+      "step": 8021
+    },
+    {
+      "epoch": 2.2221606648199446,
+      "grad_norm": 0.7191629409790039,
+      "learning_rate": 3.5112921014893375e-06,
+      "loss": 0.5969,
+      "step": 8022
+    },
+    {
+      "epoch": 2.222437673130194,
+      "grad_norm": 0.7262014746665955,
+      "learning_rate": 3.510958941119006e-06,
+      "loss": 0.608,
+      "step": 8023
+    },
+    {
+      "epoch": 2.222714681440443,
+      "grad_norm": 0.7569392323493958,
+      "learning_rate": 3.510625759283434e-06,
+      "loss": 0.6573,
+      "step": 8024
+    },
+    {
+      "epoch": 2.2229916897506925,
+      "grad_norm": 0.7174189686775208,
+      "learning_rate": 3.5102925559896977e-06,
+      "loss": 0.5955,
+      "step": 8025
+    },
+    {
+      "epoch": 2.223268698060942,
+      "grad_norm": 0.7140632271766663,
+      "learning_rate": 3.5099593312448694e-06,
+      "loss": 0.5963,
+      "step": 8026
+    },
+    {
+      "epoch": 2.223545706371191,
+      "grad_norm": 0.739309549331665,
+      "learning_rate": 3.509626085056026e-06,
+      "loss": 0.6161,
+      "step": 8027
+    },
+    {
+      "epoch": 2.2238227146814404,
+      "grad_norm": 0.7275558710098267,
+      "learning_rate": 3.509292817430242e-06,
+      "loss": 0.6273,
+      "step": 8028
+    },
+    {
+      "epoch": 2.22409972299169,
+      "grad_norm": 0.6667734384536743,
+      "learning_rate": 3.5089595283745943e-06,
+      "loss": 0.5515,
+      "step": 8029
+    },
+    {
+      "epoch": 2.224376731301939,
+      "grad_norm": 0.7046604752540588,
+      "learning_rate": 3.508626217896158e-06,
+      "loss": 0.6296,
+      "step": 8030
+    },
+    {
+      "epoch": 2.2246537396121884,
+      "grad_norm": 0.6949466466903687,
+      "learning_rate": 3.508292886002013e-06,
+      "loss": 0.59,
+      "step": 8031
+    },
+    {
+      "epoch": 2.224930747922438,
+      "grad_norm": 0.7113898992538452,
+      "learning_rate": 3.5079595326992345e-06,
+      "loss": 0.6243,
+      "step": 8032
+    },
+    {
+      "epoch": 2.225207756232687,
+      "grad_norm": 0.6967516541481018,
+      "learning_rate": 3.507626157994901e-06,
+      "loss": 0.5638,
+      "step": 8033
+    },
+    {
+      "epoch": 2.2254847645429363,
+      "grad_norm": 0.7007837891578674,
+      "learning_rate": 3.507292761896091e-06,
+      "loss": 0.5835,
+      "step": 8034
+    },
+    {
+      "epoch": 2.2257617728531858,
+      "grad_norm": 0.7281460762023926,
+      "learning_rate": 3.506959344409884e-06,
+      "loss": 0.605,
+      "step": 8035
+    },
+    {
+      "epoch": 2.226038781163435,
+      "grad_norm": 0.6967830061912537,
+      "learning_rate": 3.5066259055433575e-06,
+      "loss": 0.6225,
+      "step": 8036
+    },
+    {
+      "epoch": 2.2263157894736842,
+      "grad_norm": 0.7038334608078003,
+      "learning_rate": 3.506292445303593e-06,
+      "loss": 0.5902,
+      "step": 8037
+    },
+    {
+      "epoch": 2.2265927977839337,
+      "grad_norm": 0.7103922367095947,
+      "learning_rate": 3.5059589636976704e-06,
+      "loss": 0.5674,
+      "step": 8038
+    },
+    {
+      "epoch": 2.2268698060941827,
+      "grad_norm": 0.759024977684021,
+      "learning_rate": 3.505625460732669e-06,
+      "loss": 0.5504,
+      "step": 8039
+    },
+    {
+      "epoch": 2.227146814404432,
+      "grad_norm": 0.6931019425392151,
+      "learning_rate": 3.5052919364156718e-06,
+      "loss": 0.5769,
+      "step": 8040
+    },
+    {
+      "epoch": 2.2274238227146816,
+      "grad_norm": 0.7260349988937378,
+      "learning_rate": 3.5049583907537593e-06,
+      "loss": 0.5885,
+      "step": 8041
+    },
+    {
+      "epoch": 2.2277008310249307,
+      "grad_norm": 0.7508696913719177,
+      "learning_rate": 3.504624823754014e-06,
+      "loss": 0.5971,
+      "step": 8042
+    },
+    {
+      "epoch": 2.22797783933518,
+      "grad_norm": 0.7547364234924316,
+      "learning_rate": 3.5042912354235183e-06,
+      "loss": 0.6605,
+      "step": 8043
+    },
+    {
+      "epoch": 2.228254847645429,
+      "grad_norm": 0.7311235070228577,
+      "learning_rate": 3.5039576257693543e-06,
+      "loss": 0.6323,
+      "step": 8044
+    },
+    {
+      "epoch": 2.2285318559556786,
+      "grad_norm": 0.7337801456451416,
+      "learning_rate": 3.5036239947986066e-06,
+      "loss": 0.604,
+      "step": 8045
+    },
+    {
+      "epoch": 2.228808864265928,
+      "grad_norm": 0.7613157629966736,
+      "learning_rate": 3.503290342518358e-06,
+      "loss": 0.6399,
+      "step": 8046
+    },
+    {
+      "epoch": 2.229085872576177,
+      "grad_norm": 0.7410513162612915,
+      "learning_rate": 3.502956668935694e-06,
+      "loss": 0.5878,
+      "step": 8047
+    },
+    {
+      "epoch": 2.2293628808864265,
+      "grad_norm": 0.7026098370552063,
+      "learning_rate": 3.5026229740576985e-06,
+      "loss": 0.5976,
+      "step": 8048
+    },
+    {
+      "epoch": 2.229639889196676,
+      "grad_norm": 0.7590800523757935,
+      "learning_rate": 3.502289257891457e-06,
+      "loss": 0.5861,
+      "step": 8049
+    },
+    {
+      "epoch": 2.229916897506925,
+      "grad_norm": 0.700961172580719,
+      "learning_rate": 3.5019555204440546e-06,
+      "loss": 0.5721,
+      "step": 8050
+    },
+    {
+      "epoch": 2.2301939058171745,
+      "grad_norm": 0.7313873767852783,
+      "learning_rate": 3.5016217617225777e-06,
+      "loss": 0.6313,
+      "step": 8051
+    },
+    {
+      "epoch": 2.230470914127424,
+      "grad_norm": 0.7244826555252075,
+      "learning_rate": 3.5012879817341126e-06,
+      "loss": 0.5764,
+      "step": 8052
+    },
+    {
+      "epoch": 2.230747922437673,
+      "grad_norm": 0.7611440420150757,
+      "learning_rate": 3.5009541804857474e-06,
+      "loss": 0.6375,
+      "step": 8053
+    },
+    {
+      "epoch": 2.2310249307479224,
+      "grad_norm": 0.7172898650169373,
+      "learning_rate": 3.5006203579845687e-06,
+      "loss": 0.5634,
+      "step": 8054
+    },
+    {
+      "epoch": 2.231301939058172,
+      "grad_norm": 0.7191401720046997,
+      "learning_rate": 3.500286514237664e-06,
+      "loss": 0.6085,
+      "step": 8055
+    },
+    {
+      "epoch": 2.231578947368421,
+      "grad_norm": 0.7275636196136475,
+      "learning_rate": 3.4999526492521223e-06,
+      "loss": 0.6298,
+      "step": 8056
+    },
+    {
+      "epoch": 2.2318559556786703,
+      "grad_norm": 0.7126155495643616,
+      "learning_rate": 3.499618763035033e-06,
+      "loss": 0.6263,
+      "step": 8057
+    },
+    {
+      "epoch": 2.23213296398892,
+      "grad_norm": 0.7015401721000671,
+      "learning_rate": 3.4992848555934837e-06,
+      "loss": 0.5872,
+      "step": 8058
+    },
+    {
+      "epoch": 2.232409972299169,
+      "grad_norm": 0.8176736235618591,
+      "learning_rate": 3.498950926934565e-06,
+      "loss": 0.6324,
+      "step": 8059
+    },
+    {
+      "epoch": 2.2326869806094183,
+      "grad_norm": 0.6858984231948853,
+      "learning_rate": 3.4986169770653684e-06,
+      "loss": 0.5711,
+      "step": 8060
+    },
+    {
+      "epoch": 2.2329639889196677,
+      "grad_norm": 0.7461230158805847,
+      "learning_rate": 3.498283005992982e-06,
+      "loss": 0.6037,
+      "step": 8061
+    },
+    {
+      "epoch": 2.2332409972299168,
+      "grad_norm": 0.7021551728248596,
+      "learning_rate": 3.497949013724498e-06,
+      "loss": 0.5787,
+      "step": 8062
+    },
+    {
+      "epoch": 2.233518005540166,
+      "grad_norm": 0.7063918113708496,
+      "learning_rate": 3.4976150002670085e-06,
+      "loss": 0.5894,
+      "step": 8063
+    },
+    {
+      "epoch": 2.2337950138504157,
+      "grad_norm": 0.7009927034378052,
+      "learning_rate": 3.4972809656276047e-06,
+      "loss": 0.6006,
+      "step": 8064
+    },
+    {
+      "epoch": 2.2340720221606647,
+      "grad_norm": 0.7303755879402161,
+      "learning_rate": 3.4969469098133784e-06,
+      "loss": 0.6307,
+      "step": 8065
+    },
+    {
+      "epoch": 2.234349030470914,
+      "grad_norm": 0.717224657535553,
+      "learning_rate": 3.4966128328314243e-06,
+      "loss": 0.6233,
+      "step": 8066
+    },
+    {
+      "epoch": 2.2346260387811636,
+      "grad_norm": 0.7022543549537659,
+      "learning_rate": 3.4962787346888348e-06,
+      "loss": 0.5937,
+      "step": 8067
+    },
+    {
+      "epoch": 2.2349030470914126,
+      "grad_norm": 0.747253954410553,
+      "learning_rate": 3.4959446153927035e-06,
+      "loss": 0.603,
+      "step": 8068
+    },
+    {
+      "epoch": 2.235180055401662,
+      "grad_norm": 0.6907825469970703,
+      "learning_rate": 3.495610474950124e-06,
+      "loss": 0.5969,
+      "step": 8069
+    },
+    {
+      "epoch": 2.2354570637119116,
+      "grad_norm": 0.6939720511436462,
+      "learning_rate": 3.4952763133681923e-06,
+      "loss": 0.6142,
+      "step": 8070
+    },
+    {
+      "epoch": 2.2357340720221606,
+      "grad_norm": 0.7356911897659302,
+      "learning_rate": 3.494942130654002e-06,
+      "loss": 0.5927,
+      "step": 8071
+    },
+    {
+      "epoch": 2.23601108033241,
+      "grad_norm": 0.7271432280540466,
+      "learning_rate": 3.4946079268146497e-06,
+      "loss": 0.5765,
+      "step": 8072
+    },
+    {
+      "epoch": 2.2362880886426595,
+      "grad_norm": 0.7113369107246399,
+      "learning_rate": 3.4942737018572316e-06,
+      "loss": 0.6001,
+      "step": 8073
+    },
+    {
+      "epoch": 2.2365650969529085,
+      "grad_norm": 0.6899865865707397,
+      "learning_rate": 3.493939455788843e-06,
+      "loss": 0.5519,
+      "step": 8074
+    },
+    {
+      "epoch": 2.236842105263158,
+      "grad_norm": 0.7123438119888306,
+      "learning_rate": 3.4936051886165825e-06,
+      "loss": 0.5608,
+      "step": 8075
+    },
+    {
+      "epoch": 2.2371191135734074,
+      "grad_norm": 0.7510248422622681,
+      "learning_rate": 3.4932709003475456e-06,
+      "loss": 0.6301,
+      "step": 8076
+    },
+    {
+      "epoch": 2.2373961218836564,
+      "grad_norm": 0.704402506351471,
+      "learning_rate": 3.4929365909888314e-06,
+      "loss": 0.622,
+      "step": 8077
+    },
+    {
+      "epoch": 2.237673130193906,
+      "grad_norm": 0.7204563617706299,
+      "learning_rate": 3.4926022605475373e-06,
+      "loss": 0.6035,
+      "step": 8078
+    },
+    {
+      "epoch": 2.237950138504155,
+      "grad_norm": 0.7577924132347107,
+      "learning_rate": 3.4922679090307633e-06,
+      "loss": 0.5979,
+      "step": 8079
+    },
+    {
+      "epoch": 2.2382271468144044,
+      "grad_norm": 0.7190003991127014,
+      "learning_rate": 3.4919335364456063e-06,
+      "loss": 0.6026,
+      "step": 8080
+    },
+    {
+      "epoch": 2.238504155124654,
+      "grad_norm": 0.7183365821838379,
+      "learning_rate": 3.491599142799168e-06,
+      "loss": 0.6029,
+      "step": 8081
+    },
+    {
+      "epoch": 2.238781163434903,
+      "grad_norm": 0.6923686861991882,
+      "learning_rate": 3.491264728098548e-06,
+      "loss": 0.5583,
+      "step": 8082
+    },
+    {
+      "epoch": 2.2390581717451523,
+      "grad_norm": 0.6836802363395691,
+      "learning_rate": 3.490930292350846e-06,
+      "loss": 0.5992,
+      "step": 8083
+    },
+    {
+      "epoch": 2.2393351800554018,
+      "grad_norm": 0.73431795835495,
+      "learning_rate": 3.490595835563163e-06,
+      "loss": 0.5559,
+      "step": 8084
+    },
+    {
+      "epoch": 2.239612188365651,
+      "grad_norm": 0.7008590698242188,
+      "learning_rate": 3.4902613577426007e-06,
+      "loss": 0.5306,
+      "step": 8085
+    },
+    {
+      "epoch": 2.2398891966759003,
+      "grad_norm": 0.7099395394325256,
+      "learning_rate": 3.4899268588962613e-06,
+      "loss": 0.613,
+      "step": 8086
+    },
+    {
+      "epoch": 2.2401662049861497,
+      "grad_norm": 0.7241997122764587,
+      "learning_rate": 3.4895923390312465e-06,
+      "loss": 0.6184,
+      "step": 8087
+    },
+    {
+      "epoch": 2.2404432132963987,
+      "grad_norm": 0.7413530349731445,
+      "learning_rate": 3.4892577981546595e-06,
+      "loss": 0.589,
+      "step": 8088
+    },
+    {
+      "epoch": 2.240720221606648,
+      "grad_norm": 0.736506462097168,
+      "learning_rate": 3.4889232362736024e-06,
+      "loss": 0.5764,
+      "step": 8089
+    },
+    {
+      "epoch": 2.2409972299168976,
+      "grad_norm": 0.777686595916748,
+      "learning_rate": 3.48858865339518e-06,
+      "loss": 0.622,
+      "step": 8090
+    },
+    {
+      "epoch": 2.2412742382271467,
+      "grad_norm": 0.7260578274726868,
+      "learning_rate": 3.488254049526496e-06,
+      "loss": 0.6277,
+      "step": 8091
+    },
+    {
+      "epoch": 2.241551246537396,
+      "grad_norm": 0.6801887154579163,
+      "learning_rate": 3.4879194246746552e-06,
+      "loss": 0.5615,
+      "step": 8092
+    },
+    {
+      "epoch": 2.2418282548476456,
+      "grad_norm": 0.7225388884544373,
+      "learning_rate": 3.4875847788467614e-06,
+      "loss": 0.5972,
+      "step": 8093
+    },
+    {
+      "epoch": 2.2421052631578946,
+      "grad_norm": 0.7416013479232788,
+      "learning_rate": 3.4872501120499215e-06,
+      "loss": 0.6084,
+      "step": 8094
+    },
+    {
+      "epoch": 2.242382271468144,
+      "grad_norm": 0.7532913684844971,
+      "learning_rate": 3.48691542429124e-06,
+      "loss": 0.6234,
+      "step": 8095
+    },
+    {
+      "epoch": 2.2426592797783935,
+      "grad_norm": 0.701535701751709,
+      "learning_rate": 3.486580715577824e-06,
+      "loss": 0.6049,
+      "step": 8096
+    },
+    {
+      "epoch": 2.2429362880886425,
+      "grad_norm": 0.7444953918457031,
+      "learning_rate": 3.486245985916779e-06,
+      "loss": 0.6411,
+      "step": 8097
+    },
+    {
+      "epoch": 2.243213296398892,
+      "grad_norm": 0.7297351956367493,
+      "learning_rate": 3.4859112353152146e-06,
+      "loss": 0.5702,
+      "step": 8098
+    },
+    {
+      "epoch": 2.2434903047091415,
+      "grad_norm": 0.7168240547180176,
+      "learning_rate": 3.485576463780236e-06,
+      "loss": 0.6272,
+      "step": 8099
+    },
+    {
+      "epoch": 2.2437673130193905,
+      "grad_norm": 0.7194172143936157,
+      "learning_rate": 3.485241671318953e-06,
+      "loss": 0.5861,
+      "step": 8100
+    },
+    {
+      "epoch": 2.24404432132964,
+      "grad_norm": 0.7189883589744568,
+      "learning_rate": 3.4849068579384728e-06,
+      "loss": 0.5831,
+      "step": 8101
+    },
+    {
+      "epoch": 2.2443213296398894,
+      "grad_norm": 0.7073516845703125,
+      "learning_rate": 3.484572023645904e-06,
+      "loss": 0.588,
+      "step": 8102
+    },
+    {
+      "epoch": 2.2445983379501384,
+      "grad_norm": 0.738373875617981,
+      "learning_rate": 3.4842371684483576e-06,
+      "loss": 0.6412,
+      "step": 8103
+    },
+    {
+      "epoch": 2.244875346260388,
+      "grad_norm": 0.7670702934265137,
+      "learning_rate": 3.483902292352943e-06,
+      "loss": 0.6184,
+      "step": 8104
+    },
+    {
+      "epoch": 2.245152354570637,
+      "grad_norm": 0.757293701171875,
+      "learning_rate": 3.48356739536677e-06,
+      "loss": 0.6134,
+      "step": 8105
+    },
+    {
+      "epoch": 2.2454293628808863,
+      "grad_norm": 0.7449289560317993,
+      "learning_rate": 3.4832324774969482e-06,
+      "loss": 0.6422,
+      "step": 8106
+    },
+    {
+      "epoch": 2.245706371191136,
+      "grad_norm": 0.7168107032775879,
+      "learning_rate": 3.4828975387505914e-06,
+      "loss": 0.6189,
+      "step": 8107
+    },
+    {
+      "epoch": 2.245983379501385,
+      "grad_norm": 0.7923652529716492,
+      "learning_rate": 3.4825625791348093e-06,
+      "loss": 0.5577,
+      "step": 8108
+    },
+    {
+      "epoch": 2.2462603878116343,
+      "grad_norm": 0.7246668934822083,
+      "learning_rate": 3.4822275986567146e-06,
+      "loss": 0.555,
+      "step": 8109
+    },
+    {
+      "epoch": 2.2465373961218837,
+      "grad_norm": 0.7271319031715393,
+      "learning_rate": 3.48189259732342e-06,
+      "loss": 0.5997,
+      "step": 8110
+    },
+    {
+      "epoch": 2.2468144044321328,
+      "grad_norm": 0.7270683646202087,
+      "learning_rate": 3.481557575142038e-06,
+      "loss": 0.609,
+      "step": 8111
+    },
+    {
+      "epoch": 2.247091412742382,
+      "grad_norm": 0.7465324401855469,
+      "learning_rate": 3.481222532119681e-06,
+      "loss": 0.5592,
+      "step": 8112
+    },
+    {
+      "epoch": 2.2473684210526317,
+      "grad_norm": 0.7218868732452393,
+      "learning_rate": 3.480887468263465e-06,
+      "loss": 0.6101,
+      "step": 8113
+    },
+    {
+      "epoch": 2.2476454293628807,
+      "grad_norm": 0.7048141360282898,
+      "learning_rate": 3.480552383580502e-06,
+      "loss": 0.5884,
+      "step": 8114
+    },
+    {
+      "epoch": 2.24792243767313,
+      "grad_norm": 0.7366784811019897,
+      "learning_rate": 3.4802172780779088e-06,
+      "loss": 0.6199,
+      "step": 8115
+    },
+    {
+      "epoch": 2.2481994459833796,
+      "grad_norm": 0.7352172136306763,
+      "learning_rate": 3.479882151762799e-06,
+      "loss": 0.6353,
+      "step": 8116
+    },
+    {
+      "epoch": 2.2484764542936286,
+      "grad_norm": 0.7354872226715088,
+      "learning_rate": 3.4795470046422896e-06,
+      "loss": 0.6304,
+      "step": 8117
+    },
+    {
+      "epoch": 2.248753462603878,
+      "grad_norm": 0.6982522010803223,
+      "learning_rate": 3.479211836723495e-06,
+      "loss": 0.5632,
+      "step": 8118
+    },
+    {
+      "epoch": 2.2490304709141276,
+      "grad_norm": 0.7308842539787292,
+      "learning_rate": 3.478876648013533e-06,
+      "loss": 0.5846,
+      "step": 8119
+    },
+    {
+      "epoch": 2.2493074792243766,
+      "grad_norm": 0.7680553793907166,
+      "learning_rate": 3.4785414385195205e-06,
+      "loss": 0.6225,
+      "step": 8120
+    },
+    {
+      "epoch": 2.249584487534626,
+      "grad_norm": 0.7662866711616516,
+      "learning_rate": 3.478206208248573e-06,
+      "loss": 0.5734,
+      "step": 8121
+    },
+    {
+      "epoch": 2.2498614958448755,
+      "grad_norm": 0.7630404233932495,
+      "learning_rate": 3.47787095720781e-06,
+      "loss": 0.5667,
+      "step": 8122
+    },
+    {
+      "epoch": 2.2501385041551245,
+      "grad_norm": 0.7009580731391907,
+      "learning_rate": 3.4775356854043497e-06,
+      "loss": 0.5763,
+      "step": 8123
+    },
+    {
+      "epoch": 2.250415512465374,
+      "grad_norm": 0.7273560762405396,
+      "learning_rate": 3.4772003928453106e-06,
+      "loss": 0.58,
+      "step": 8124
+    },
+    {
+      "epoch": 2.2506925207756234,
+      "grad_norm": 0.7440486550331116,
+      "learning_rate": 3.476865079537811e-06,
+      "loss": 0.6165,
+      "step": 8125
+    },
+    {
+      "epoch": 2.2509695290858724,
+      "grad_norm": 0.712805449962616,
+      "learning_rate": 3.4765297454889717e-06,
+      "loss": 0.6104,
+      "step": 8126
+    },
+    {
+      "epoch": 2.251246537396122,
+      "grad_norm": 0.7313037514686584,
+      "learning_rate": 3.4761943907059122e-06,
+      "loss": 0.573,
+      "step": 8127
+    },
+    {
+      "epoch": 2.2515235457063714,
+      "grad_norm": 0.7176337838172913,
+      "learning_rate": 3.475859015195753e-06,
+      "loss": 0.6145,
+      "step": 8128
+    },
+    {
+      "epoch": 2.2518005540166204,
+      "grad_norm": 0.7878891825675964,
+      "learning_rate": 3.4755236189656137e-06,
+      "loss": 0.588,
+      "step": 8129
+    },
+    {
+      "epoch": 2.25207756232687,
+      "grad_norm": 0.7895088195800781,
+      "learning_rate": 3.4751882020226174e-06,
+      "loss": 0.602,
+      "step": 8130
+    },
+    {
+      "epoch": 2.2523545706371193,
+      "grad_norm": 0.7553879618644714,
+      "learning_rate": 3.4748527643738856e-06,
+      "loss": 0.6034,
+      "step": 8131
+    },
+    {
+      "epoch": 2.2526315789473683,
+      "grad_norm": 0.6995738744735718,
+      "learning_rate": 3.474517306026539e-06,
+      "loss": 0.5861,
+      "step": 8132
+    },
+    {
+      "epoch": 2.252908587257618,
+      "grad_norm": 0.7691447138786316,
+      "learning_rate": 3.474181826987703e-06,
+      "loss": 0.5564,
+      "step": 8133
+    },
+    {
+      "epoch": 2.2531855955678672,
+      "grad_norm": 0.6974400877952576,
+      "learning_rate": 3.473846327264498e-06,
+      "loss": 0.6081,
+      "step": 8134
+    },
+    {
+      "epoch": 2.2534626038781163,
+      "grad_norm": 0.7364979386329651,
+      "learning_rate": 3.473510806864049e-06,
+      "loss": 0.6534,
+      "step": 8135
+    },
+    {
+      "epoch": 2.2537396121883657,
+      "grad_norm": 0.7088637948036194,
+      "learning_rate": 3.4731752657934793e-06,
+      "loss": 0.5524,
+      "step": 8136
+    },
+    {
+      "epoch": 2.254016620498615,
+      "grad_norm": 0.7505431175231934,
+      "learning_rate": 3.4728397040599137e-06,
+      "loss": 0.5805,
+      "step": 8137
+    },
+    {
+      "epoch": 2.254293628808864,
+      "grad_norm": 0.7170282006263733,
+      "learning_rate": 3.472504121670476e-06,
+      "loss": 0.5365,
+      "step": 8138
+    },
+    {
+      "epoch": 2.2545706371191137,
+      "grad_norm": 0.713196873664856,
+      "learning_rate": 3.472168518632293e-06,
+      "loss": 0.6709,
+      "step": 8139
+    },
+    {
+      "epoch": 2.254847645429363,
+      "grad_norm": 0.7076315879821777,
+      "learning_rate": 3.4718328949524894e-06,
+      "loss": 0.5865,
+      "step": 8140
+    },
+    {
+      "epoch": 2.255124653739612,
+      "grad_norm": 0.7079249620437622,
+      "learning_rate": 3.4714972506381916e-06,
+      "loss": 0.5976,
+      "step": 8141
+    },
+    {
+      "epoch": 2.2554016620498616,
+      "grad_norm": 0.7071895003318787,
+      "learning_rate": 3.4711615856965264e-06,
+      "loss": 0.5831,
+      "step": 8142
+    },
+    {
+      "epoch": 2.255678670360111,
+      "grad_norm": 0.7474883198738098,
+      "learning_rate": 3.4708259001346204e-06,
+      "loss": 0.6041,
+      "step": 8143
+    },
+    {
+      "epoch": 2.25595567867036,
+      "grad_norm": 0.7047725319862366,
+      "learning_rate": 3.470490193959602e-06,
+      "loss": 0.5933,
+      "step": 8144
+    },
+    {
+      "epoch": 2.2562326869806095,
+      "grad_norm": 0.7273043990135193,
+      "learning_rate": 3.470154467178598e-06,
+      "loss": 0.6108,
+      "step": 8145
+    },
+    {
+      "epoch": 2.2565096952908585,
+      "grad_norm": 0.7017418146133423,
+      "learning_rate": 3.4698187197987377e-06,
+      "loss": 0.5158,
+      "step": 8146
+    },
+    {
+      "epoch": 2.256786703601108,
+      "grad_norm": 0.7130430936813354,
+      "learning_rate": 3.4694829518271485e-06,
+      "loss": 0.6265,
+      "step": 8147
+    },
+    {
+      "epoch": 2.2570637119113575,
+      "grad_norm": 0.6907879710197449,
+      "learning_rate": 3.4691471632709604e-06,
+      "loss": 0.6218,
+      "step": 8148
+    },
+    {
+      "epoch": 2.2573407202216065,
+      "grad_norm": 0.7310351729393005,
+      "learning_rate": 3.4688113541373037e-06,
+      "loss": 0.6313,
+      "step": 8149
+    },
+    {
+      "epoch": 2.257617728531856,
+      "grad_norm": 0.7051122188568115,
+      "learning_rate": 3.468475524433308e-06,
+      "loss": 0.6325,
+      "step": 8150
+    },
+    {
+      "epoch": 2.2578947368421054,
+      "grad_norm": 0.7112586498260498,
+      "learning_rate": 3.4681396741661028e-06,
+      "loss": 0.6221,
+      "step": 8151
+    },
+    {
+      "epoch": 2.2581717451523544,
+      "grad_norm": 0.7428193092346191,
+      "learning_rate": 3.467803803342821e-06,
+      "loss": 0.6333,
+      "step": 8152
+    },
+    {
+      "epoch": 2.258448753462604,
+      "grad_norm": 0.7509081363677979,
+      "learning_rate": 3.467467911970592e-06,
+      "loss": 0.5948,
+      "step": 8153
+    },
+    {
+      "epoch": 2.2587257617728533,
+      "grad_norm": 0.7133442163467407,
+      "learning_rate": 3.467132000056549e-06,
+      "loss": 0.5938,
+      "step": 8154
+    },
+    {
+      "epoch": 2.2590027700831024,
+      "grad_norm": 0.73454350233078,
+      "learning_rate": 3.466796067607824e-06,
+      "loss": 0.6126,
+      "step": 8155
+    },
+    {
+      "epoch": 2.259279778393352,
+      "grad_norm": 0.7535995244979858,
+      "learning_rate": 3.4664601146315496e-06,
+      "loss": 0.5605,
+      "step": 8156
+    },
+    {
+      "epoch": 2.2595567867036013,
+      "grad_norm": 0.7269196510314941,
+      "learning_rate": 3.4661241411348586e-06,
+      "loss": 0.6224,
+      "step": 8157
+    },
+    {
+      "epoch": 2.2598337950138503,
+      "grad_norm": 0.7485670447349548,
+      "learning_rate": 3.465788147124885e-06,
+      "loss": 0.6162,
+      "step": 8158
+    },
+    {
+      "epoch": 2.2601108033240997,
+      "grad_norm": 0.7506015300750732,
+      "learning_rate": 3.4654521326087627e-06,
+      "loss": 0.6221,
+      "step": 8159
+    },
+    {
+      "epoch": 2.260387811634349,
+      "grad_norm": 0.7276700139045715,
+      "learning_rate": 3.4651160975936256e-06,
+      "loss": 0.5293,
+      "step": 8160
+    },
+    {
+      "epoch": 2.2606648199445982,
+      "grad_norm": 0.7210519909858704,
+      "learning_rate": 3.4647800420866096e-06,
+      "loss": 0.5598,
+      "step": 8161
+    },
+    {
+      "epoch": 2.2609418282548477,
+      "grad_norm": 0.6977429389953613,
+      "learning_rate": 3.4644439660948497e-06,
+      "loss": 0.5866,
+      "step": 8162
+    },
+    {
+      "epoch": 2.261218836565097,
+      "grad_norm": 0.7661787271499634,
+      "learning_rate": 3.4641078696254814e-06,
+      "loss": 0.5967,
+      "step": 8163
+    },
+    {
+      "epoch": 2.261495844875346,
+      "grad_norm": 0.7270045280456543,
+      "learning_rate": 3.4637717526856408e-06,
+      "loss": 0.6014,
+      "step": 8164
+    },
+    {
+      "epoch": 2.2617728531855956,
+      "grad_norm": 0.7071644067764282,
+      "learning_rate": 3.4634356152824654e-06,
+      "loss": 0.6252,
+      "step": 8165
+    },
+    {
+      "epoch": 2.2620498614958446,
+      "grad_norm": 0.7300014495849609,
+      "learning_rate": 3.4630994574230904e-06,
+      "loss": 0.5668,
+      "step": 8166
+    },
+    {
+      "epoch": 2.262326869806094,
+      "grad_norm": 0.8083431124687195,
+      "learning_rate": 3.462763279114655e-06,
+      "loss": 0.6159,
+      "step": 8167
+    },
+    {
+      "epoch": 2.2626038781163436,
+      "grad_norm": 0.7473114132881165,
+      "learning_rate": 3.4624270803642967e-06,
+      "loss": 0.5696,
+      "step": 8168
+    },
+    {
+      "epoch": 2.2628808864265926,
+      "grad_norm": 0.7719151973724365,
+      "learning_rate": 3.462090861179154e-06,
+      "loss": 0.5858,
+      "step": 8169
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.7300695180892944,
+      "learning_rate": 3.461754621566364e-06,
+      "loss": 0.573,
+      "step": 8170
+    },
+    {
+      "epoch": 2.2634349030470915,
+      "grad_norm": 0.7490740418434143,
+      "learning_rate": 3.4614183615330683e-06,
+      "loss": 0.5772,
+      "step": 8171
+    },
+    {
+      "epoch": 2.2637119113573405,
+      "grad_norm": 0.6871481537818909,
+      "learning_rate": 3.4610820810864053e-06,
+      "loss": 0.5572,
+      "step": 8172
+    },
+    {
+      "epoch": 2.26398891966759,
+      "grad_norm": 0.7448837161064148,
+      "learning_rate": 3.4607457802335153e-06,
+      "loss": 0.5986,
+      "step": 8173
+    },
+    {
+      "epoch": 2.2642659279778394,
+      "grad_norm": 0.7582398056983948,
+      "learning_rate": 3.4604094589815402e-06,
+      "loss": 0.6259,
+      "step": 8174
+    },
+    {
+      "epoch": 2.2645429362880884,
+      "grad_norm": 0.7317702770233154,
+      "learning_rate": 3.4600731173376184e-06,
+      "loss": 0.5945,
+      "step": 8175
+    },
+    {
+      "epoch": 2.264819944598338,
+      "grad_norm": 0.6793874502182007,
+      "learning_rate": 3.4597367553088927e-06,
+      "loss": 0.5673,
+      "step": 8176
+    },
+    {
+      "epoch": 2.2650969529085874,
+      "grad_norm": 0.745205283164978,
+      "learning_rate": 3.459400372902505e-06,
+      "loss": 0.6109,
+      "step": 8177
+    },
+    {
+      "epoch": 2.2653739612188364,
+      "grad_norm": 0.7162503004074097,
+      "learning_rate": 3.459063970125597e-06,
+      "loss": 0.6033,
+      "step": 8178
+    },
+    {
+      "epoch": 2.265650969529086,
+      "grad_norm": 0.7062127590179443,
+      "learning_rate": 3.458727546985312e-06,
+      "loss": 0.5657,
+      "step": 8179
+    },
+    {
+      "epoch": 2.2659279778393353,
+      "grad_norm": 0.7209470272064209,
+      "learning_rate": 3.458391103488793e-06,
+      "loss": 0.5944,
+      "step": 8180
+    },
+    {
+      "epoch": 2.2662049861495843,
+      "grad_norm": 0.7374285459518433,
+      "learning_rate": 3.4580546396431837e-06,
+      "loss": 0.6177,
+      "step": 8181
+    },
+    {
+      "epoch": 2.266481994459834,
+      "grad_norm": 0.7219187021255493,
+      "learning_rate": 3.4577181554556273e-06,
+      "loss": 0.6365,
+      "step": 8182
+    },
+    {
+      "epoch": 2.2667590027700832,
+      "grad_norm": 0.7407369613647461,
+      "learning_rate": 3.457381650933268e-06,
+      "loss": 0.5901,
+      "step": 8183
+    },
+    {
+      "epoch": 2.2670360110803323,
+      "grad_norm": 0.7512839436531067,
+      "learning_rate": 3.4570451260832533e-06,
+      "loss": 0.5955,
+      "step": 8184
+    },
+    {
+      "epoch": 2.2673130193905817,
+      "grad_norm": 0.7049588561058044,
+      "learning_rate": 3.4567085809127247e-06,
+      "loss": 0.6389,
+      "step": 8185
+    },
+    {
+      "epoch": 2.267590027700831,
+      "grad_norm": 0.7323600053787231,
+      "learning_rate": 3.4563720154288306e-06,
+      "loss": 0.5776,
+      "step": 8186
+    },
+    {
+      "epoch": 2.26786703601108,
+      "grad_norm": 0.7171778082847595,
+      "learning_rate": 3.4560354296387167e-06,
+      "loss": 0.6096,
+      "step": 8187
+    },
+    {
+      "epoch": 2.2681440443213297,
+      "grad_norm": 0.7620019316673279,
+      "learning_rate": 3.455698823549528e-06,
+      "loss": 0.6409,
+      "step": 8188
+    },
+    {
+      "epoch": 2.268421052631579,
+      "grad_norm": 0.7481417059898376,
+      "learning_rate": 3.455362197168413e-06,
+      "loss": 0.6072,
+      "step": 8189
+    },
+    {
+      "epoch": 2.268698060941828,
+      "grad_norm": 0.7183117270469666,
+      "learning_rate": 3.4550255505025198e-06,
+      "loss": 0.5921,
+      "step": 8190
+    },
+    {
+      "epoch": 2.2689750692520776,
+      "grad_norm": 0.7216724753379822,
+      "learning_rate": 3.454688883558994e-06,
+      "loss": 0.5779,
+      "step": 8191
+    },
+    {
+      "epoch": 2.269252077562327,
+      "grad_norm": 0.7133140563964844,
+      "learning_rate": 3.454352196344986e-06,
+      "loss": 0.6287,
+      "step": 8192
+    },
+    {
+      "epoch": 2.269529085872576,
+      "grad_norm": 0.698267936706543,
+      "learning_rate": 3.4540154888676435e-06,
+      "loss": 0.5707,
+      "step": 8193
+    },
+    {
+      "epoch": 2.2698060941828255,
+      "grad_norm": 0.6946624517440796,
+      "learning_rate": 3.4536787611341162e-06,
+      "loss": 0.5806,
+      "step": 8194
+    },
+    {
+      "epoch": 2.270083102493075,
+      "grad_norm": 0.7460806369781494,
+      "learning_rate": 3.453342013151553e-06,
+      "loss": 0.5734,
+      "step": 8195
+    },
+    {
+      "epoch": 2.270360110803324,
+      "grad_norm": 0.6981709599494934,
+      "learning_rate": 3.4530052449271044e-06,
+      "loss": 0.5999,
+      "step": 8196
+    },
+    {
+      "epoch": 2.2706371191135735,
+      "grad_norm": 0.744199812412262,
+      "learning_rate": 3.4526684564679212e-06,
+      "loss": 0.6382,
+      "step": 8197
+    },
+    {
+      "epoch": 2.270914127423823,
+      "grad_norm": 0.7204859256744385,
+      "learning_rate": 3.452331647781153e-06,
+      "loss": 0.599,
+      "step": 8198
+    },
+    {
+      "epoch": 2.271191135734072,
+      "grad_norm": 0.7273908257484436,
+      "learning_rate": 3.4519948188739523e-06,
+      "loss": 0.6032,
+      "step": 8199
+    },
+    {
+      "epoch": 2.2714681440443214,
+      "grad_norm": 0.7486979961395264,
+      "learning_rate": 3.4516579697534708e-06,
+      "loss": 0.6435,
+      "step": 8200
+    },
+    {
+      "epoch": 2.271745152354571,
+      "grad_norm": 0.7121773362159729,
+      "learning_rate": 3.4513211004268595e-06,
+      "loss": 0.5527,
+      "step": 8201
+    },
+    {
+      "epoch": 2.27202216066482,
+      "grad_norm": 0.736427903175354,
+      "learning_rate": 3.4509842109012725e-06,
+      "loss": 0.6071,
+      "step": 8202
+    },
+    {
+      "epoch": 2.2722991689750693,
+      "grad_norm": 0.7484080791473389,
+      "learning_rate": 3.450647301183862e-06,
+      "loss": 0.6488,
+      "step": 8203
+    },
+    {
+      "epoch": 2.272576177285319,
+      "grad_norm": 0.705027163028717,
+      "learning_rate": 3.4503103712817815e-06,
+      "loss": 0.6334,
+      "step": 8204
+    },
+    {
+      "epoch": 2.272853185595568,
+      "grad_norm": 0.7610204815864563,
+      "learning_rate": 3.4499734212021847e-06,
+      "loss": 0.5988,
+      "step": 8205
+    },
+    {
+      "epoch": 2.2731301939058173,
+      "grad_norm": 0.7779908776283264,
+      "learning_rate": 3.4496364509522275e-06,
+      "loss": 0.5966,
+      "step": 8206
+    },
+    {
+      "epoch": 2.2734072022160663,
+      "grad_norm": 0.7502352595329285,
+      "learning_rate": 3.449299460539062e-06,
+      "loss": 0.5684,
+      "step": 8207
+    },
+    {
+      "epoch": 2.2736842105263158,
+      "grad_norm": 0.8014580011367798,
+      "learning_rate": 3.4489624499698447e-06,
+      "loss": 0.5943,
+      "step": 8208
+    },
+    {
+      "epoch": 2.273961218836565,
+      "grad_norm": 0.7266554832458496,
+      "learning_rate": 3.448625419251732e-06,
+      "loss": 0.6069,
+      "step": 8209
+    },
+    {
+      "epoch": 2.2742382271468142,
+      "grad_norm": 0.764426052570343,
+      "learning_rate": 3.4482883683918788e-06,
+      "loss": 0.636,
+      "step": 8210
+    },
+    {
+      "epoch": 2.2745152354570637,
+      "grad_norm": 0.7291103005409241,
+      "learning_rate": 3.4479512973974415e-06,
+      "loss": 0.6315,
+      "step": 8211
+    },
+    {
+      "epoch": 2.274792243767313,
+      "grad_norm": 0.7153321504592896,
+      "learning_rate": 3.447614206275578e-06,
+      "loss": 0.5893,
+      "step": 8212
+    },
+    {
+      "epoch": 2.275069252077562,
+      "grad_norm": 0.7283423542976379,
+      "learning_rate": 3.4472770950334443e-06,
+      "loss": 0.5989,
+      "step": 8213
+    },
+    {
+      "epoch": 2.2753462603878116,
+      "grad_norm": 0.7300357818603516,
+      "learning_rate": 3.4469399636781987e-06,
+      "loss": 0.547,
+      "step": 8214
+    },
+    {
+      "epoch": 2.275623268698061,
+      "grad_norm": 0.751139223575592,
+      "learning_rate": 3.4466028122170003e-06,
+      "loss": 0.579,
+      "step": 8215
+    },
+    {
+      "epoch": 2.27590027700831,
+      "grad_norm": 0.7210396528244019,
+      "learning_rate": 3.446265640657006e-06,
+      "loss": 0.6054,
+      "step": 8216
+    },
+    {
+      "epoch": 2.2761772853185596,
+      "grad_norm": 0.7154489755630493,
+      "learning_rate": 3.445928449005376e-06,
+      "loss": 0.5572,
+      "step": 8217
+    },
+    {
+      "epoch": 2.276454293628809,
+      "grad_norm": 0.7469511032104492,
+      "learning_rate": 3.4455912372692696e-06,
+      "loss": 0.5857,
+      "step": 8218
+    },
+    {
+      "epoch": 2.276731301939058,
+      "grad_norm": 0.7889577150344849,
+      "learning_rate": 3.445254005455847e-06,
+      "loss": 0.5819,
+      "step": 8219
+    },
+    {
+      "epoch": 2.2770083102493075,
+      "grad_norm": 0.7772889733314514,
+      "learning_rate": 3.444916753572267e-06,
+      "loss": 0.5903,
+      "step": 8220
+    },
+    {
+      "epoch": 2.277285318559557,
+      "grad_norm": 0.7089385390281677,
+      "learning_rate": 3.444579481625691e-06,
+      "loss": 0.5651,
+      "step": 8221
+    },
+    {
+      "epoch": 2.277562326869806,
+      "grad_norm": 0.7505936026573181,
+      "learning_rate": 3.444242189623281e-06,
+      "loss": 0.66,
+      "step": 8222
+    },
+    {
+      "epoch": 2.2778393351800554,
+      "grad_norm": 0.7459689378738403,
+      "learning_rate": 3.4439048775721983e-06,
+      "loss": 0.5863,
+      "step": 8223
+    },
+    {
+      "epoch": 2.278116343490305,
+      "grad_norm": 0.6946021318435669,
+      "learning_rate": 3.4435675454796035e-06,
+      "loss": 0.6156,
+      "step": 8224
+    },
+    {
+      "epoch": 2.278393351800554,
+      "grad_norm": 0.7873076796531677,
+      "learning_rate": 3.4432301933526614e-06,
+      "loss": 0.6069,
+      "step": 8225
+    },
+    {
+      "epoch": 2.2786703601108034,
+      "grad_norm": 0.7463105916976929,
+      "learning_rate": 3.442892821198533e-06,
+      "loss": 0.6001,
+      "step": 8226
+    },
+    {
+      "epoch": 2.2789473684210524,
+      "grad_norm": 0.7991135716438293,
+      "learning_rate": 3.442555429024382e-06,
+      "loss": 0.611,
+      "step": 8227
+    },
+    {
+      "epoch": 2.279224376731302,
+      "grad_norm": 0.7371905446052551,
+      "learning_rate": 3.4422180168373725e-06,
+      "loss": 0.655,
+      "step": 8228
+    },
+    {
+      "epoch": 2.2795013850415513,
+      "grad_norm": 0.7202360033988953,
+      "learning_rate": 3.4418805846446678e-06,
+      "loss": 0.6133,
+      "step": 8229
+    },
+    {
+      "epoch": 2.2797783933518003,
+      "grad_norm": 0.7565407156944275,
+      "learning_rate": 3.441543132453433e-06,
+      "loss": 0.622,
+      "step": 8230
+    },
+    {
+      "epoch": 2.28005540166205,
+      "grad_norm": 0.7521671652793884,
+      "learning_rate": 3.4412056602708337e-06,
+      "loss": 0.6133,
+      "step": 8231
+    },
+    {
+      "epoch": 2.2803324099722992,
+      "grad_norm": 0.709683895111084,
+      "learning_rate": 3.4408681681040344e-06,
+      "loss": 0.5736,
+      "step": 8232
+    },
+    {
+      "epoch": 2.2806094182825483,
+      "grad_norm": 0.7759260535240173,
+      "learning_rate": 3.440530655960201e-06,
+      "loss": 0.5931,
+      "step": 8233
+    },
+    {
+      "epoch": 2.2808864265927977,
+      "grad_norm": 0.722997784614563,
+      "learning_rate": 3.4401931238464996e-06,
+      "loss": 0.5617,
+      "step": 8234
+    },
+    {
+      "epoch": 2.281163434903047,
+      "grad_norm": 0.7048691511154175,
+      "learning_rate": 3.4398555717700975e-06,
+      "loss": 0.5994,
+      "step": 8235
+    },
+    {
+      "epoch": 2.281440443213296,
+      "grad_norm": 0.7193622589111328,
+      "learning_rate": 3.439517999738162e-06,
+      "loss": 0.5688,
+      "step": 8236
+    },
+    {
+      "epoch": 2.2817174515235457,
+      "grad_norm": 0.7562454342842102,
+      "learning_rate": 3.4391804077578592e-06,
+      "loss": 0.6113,
+      "step": 8237
+    },
+    {
+      "epoch": 2.281994459833795,
+      "grad_norm": 0.737359344959259,
+      "learning_rate": 3.438842795836359e-06,
+      "loss": 0.5539,
+      "step": 8238
+    },
+    {
+      "epoch": 2.282271468144044,
+      "grad_norm": 0.7290983200073242,
+      "learning_rate": 3.4385051639808277e-06,
+      "loss": 0.645,
+      "step": 8239
+    },
+    {
+      "epoch": 2.2825484764542936,
+      "grad_norm": 0.7329986095428467,
+      "learning_rate": 3.438167512198436e-06,
+      "loss": 0.5951,
+      "step": 8240
+    },
+    {
+      "epoch": 2.282825484764543,
+      "grad_norm": 0.7410658001899719,
+      "learning_rate": 3.4378298404963517e-06,
+      "loss": 0.6091,
+      "step": 8241
+    },
+    {
+      "epoch": 2.283102493074792,
+      "grad_norm": 0.7463988661766052,
+      "learning_rate": 3.437492148881745e-06,
+      "loss": 0.6081,
+      "step": 8242
+    },
+    {
+      "epoch": 2.2833795013850415,
+      "grad_norm": 0.7464751601219177,
+      "learning_rate": 3.4371544373617857e-06,
+      "loss": 0.5989,
+      "step": 8243
+    },
+    {
+      "epoch": 2.283656509695291,
+      "grad_norm": 0.744561493396759,
+      "learning_rate": 3.436816705943644e-06,
+      "loss": 0.5968,
+      "step": 8244
+    },
+    {
+      "epoch": 2.28393351800554,
+      "grad_norm": 0.7255566716194153,
+      "learning_rate": 3.436478954634492e-06,
+      "loss": 0.5382,
+      "step": 8245
+    },
+    {
+      "epoch": 2.2842105263157895,
+      "grad_norm": 0.749146580696106,
+      "learning_rate": 3.4361411834415e-06,
+      "loss": 0.6063,
+      "step": 8246
+    },
+    {
+      "epoch": 2.284487534626039,
+      "grad_norm": 0.7449796795845032,
+      "learning_rate": 3.4358033923718405e-06,
+      "loss": 0.6034,
+      "step": 8247
+    },
+    {
+      "epoch": 2.284764542936288,
+      "grad_norm": 0.7429720163345337,
+      "learning_rate": 3.4354655814326847e-06,
+      "loss": 0.5836,
+      "step": 8248
+    },
+    {
+      "epoch": 2.2850415512465374,
+      "grad_norm": 0.6745700836181641,
+      "learning_rate": 3.435127750631206e-06,
+      "loss": 0.5858,
+      "step": 8249
+    },
+    {
+      "epoch": 2.285318559556787,
+      "grad_norm": 0.7240549325942993,
+      "learning_rate": 3.4347898999745766e-06,
+      "loss": 0.5872,
+      "step": 8250
+    },
+    {
+      "epoch": 2.285595567867036,
+      "grad_norm": 0.7113307118415833,
+      "learning_rate": 3.4344520294699714e-06,
+      "loss": 0.6165,
+      "step": 8251
+    },
+    {
+      "epoch": 2.2858725761772853,
+      "grad_norm": 0.7669567465782166,
+      "learning_rate": 3.4341141391245626e-06,
+      "loss": 0.5966,
+      "step": 8252
+    },
+    {
+      "epoch": 2.286149584487535,
+      "grad_norm": 0.7530611157417297,
+      "learning_rate": 3.433776228945525e-06,
+      "loss": 0.5964,
+      "step": 8253
+    },
+    {
+      "epoch": 2.286426592797784,
+      "grad_norm": 0.7283420562744141,
+      "learning_rate": 3.433438298940034e-06,
+      "loss": 0.6044,
+      "step": 8254
+    },
+    {
+      "epoch": 2.2867036011080333,
+      "grad_norm": 0.6976724863052368,
+      "learning_rate": 3.4331003491152635e-06,
+      "loss": 0.5674,
+      "step": 8255
+    },
+    {
+      "epoch": 2.2869806094182827,
+      "grad_norm": 0.6952914595603943,
+      "learning_rate": 3.43276237947839e-06,
+      "loss": 0.5729,
+      "step": 8256
+    },
+    {
+      "epoch": 2.2872576177285318,
+      "grad_norm": 0.7647460103034973,
+      "learning_rate": 3.4324243900365894e-06,
+      "loss": 0.5718,
+      "step": 8257
+    },
+    {
+      "epoch": 2.287534626038781,
+      "grad_norm": 0.7178311347961426,
+      "learning_rate": 3.432086380797037e-06,
+      "loss": 0.5855,
+      "step": 8258
+    },
+    {
+      "epoch": 2.2878116343490307,
+      "grad_norm": 0.7695008516311646,
+      "learning_rate": 3.4317483517669115e-06,
+      "loss": 0.6255,
+      "step": 8259
+    },
+    {
+      "epoch": 2.2880886426592797,
+      "grad_norm": 0.7568122744560242,
+      "learning_rate": 3.431410302953389e-06,
+      "loss": 0.5432,
+      "step": 8260
+    },
+    {
+      "epoch": 2.288365650969529,
+      "grad_norm": 0.6739785075187683,
+      "learning_rate": 3.4310722343636464e-06,
+      "loss": 0.5778,
+      "step": 8261
+    },
+    {
+      "epoch": 2.2886426592797786,
+      "grad_norm": 0.7452985644340515,
+      "learning_rate": 3.4307341460048633e-06,
+      "loss": 0.6284,
+      "step": 8262
+    },
+    {
+      "epoch": 2.2889196675900276,
+      "grad_norm": 0.7170669436454773,
+      "learning_rate": 3.430396037884217e-06,
+      "loss": 0.6227,
+      "step": 8263
+    },
+    {
+      "epoch": 2.289196675900277,
+      "grad_norm": 0.7811469435691833,
+      "learning_rate": 3.4300579100088875e-06,
+      "loss": 0.5934,
+      "step": 8264
+    },
+    {
+      "epoch": 2.2894736842105265,
+      "grad_norm": 0.7597870826721191,
+      "learning_rate": 3.4297197623860525e-06,
+      "loss": 0.6306,
+      "step": 8265
+    },
+    {
+      "epoch": 2.2897506925207756,
+      "grad_norm": 0.7523090839385986,
+      "learning_rate": 3.4293815950228936e-06,
+      "loss": 0.5915,
+      "step": 8266
+    },
+    {
+      "epoch": 2.290027700831025,
+      "grad_norm": 0.7411076426506042,
+      "learning_rate": 3.4290434079265894e-06,
+      "loss": 0.6061,
+      "step": 8267
+    },
+    {
+      "epoch": 2.2903047091412745,
+      "grad_norm": 0.7374351620674133,
+      "learning_rate": 3.4287052011043214e-06,
+      "loss": 0.5945,
+      "step": 8268
+    },
+    {
+      "epoch": 2.2905817174515235,
+      "grad_norm": 0.7511986494064331,
+      "learning_rate": 3.4283669745632705e-06,
+      "loss": 0.5952,
+      "step": 8269
+    },
+    {
+      "epoch": 2.290858725761773,
+      "grad_norm": 0.7723880410194397,
+      "learning_rate": 3.4280287283106172e-06,
+      "loss": 0.6109,
+      "step": 8270
+    },
+    {
+      "epoch": 2.291135734072022,
+      "grad_norm": 0.694105327129364,
+      "learning_rate": 3.427690462353545e-06,
+      "loss": 0.6328,
+      "step": 8271
+    },
+    {
+      "epoch": 2.2914127423822714,
+      "grad_norm": 0.7096155285835266,
+      "learning_rate": 3.4273521766992348e-06,
+      "loss": 0.589,
+      "step": 8272
+    },
+    {
+      "epoch": 2.291689750692521,
+      "grad_norm": 0.7510867714881897,
+      "learning_rate": 3.42701387135487e-06,
+      "loss": 0.5721,
+      "step": 8273
+    },
+    {
+      "epoch": 2.29196675900277,
+      "grad_norm": 0.7638363838195801,
+      "learning_rate": 3.4266755463276325e-06,
+      "loss": 0.6358,
+      "step": 8274
+    },
+    {
+      "epoch": 2.2922437673130194,
+      "grad_norm": 0.7557123899459839,
+      "learning_rate": 3.4263372016247074e-06,
+      "loss": 0.6244,
+      "step": 8275
+    },
+    {
+      "epoch": 2.292520775623269,
+      "grad_norm": 0.725536048412323,
+      "learning_rate": 3.4259988372532783e-06,
+      "loss": 0.6014,
+      "step": 8276
+    },
+    {
+      "epoch": 2.292797783933518,
+      "grad_norm": 0.7189280390739441,
+      "learning_rate": 3.4256604532205285e-06,
+      "loss": 0.5749,
+      "step": 8277
+    },
+    {
+      "epoch": 2.2930747922437673,
+      "grad_norm": 0.745522677898407,
+      "learning_rate": 3.425322049533643e-06,
+      "loss": 0.6433,
+      "step": 8278
+    },
+    {
+      "epoch": 2.2933518005540168,
+      "grad_norm": 0.7054515480995178,
+      "learning_rate": 3.4249836261998086e-06,
+      "loss": 0.5802,
+      "step": 8279
+    },
+    {
+      "epoch": 2.293628808864266,
+      "grad_norm": 0.75422602891922,
+      "learning_rate": 3.4246451832262085e-06,
+      "loss": 0.595,
+      "step": 8280
+    },
+    {
+      "epoch": 2.2939058171745152,
+      "grad_norm": 0.6765782237052917,
+      "learning_rate": 3.424306720620031e-06,
+      "loss": 0.5849,
+      "step": 8281
+    },
+    {
+      "epoch": 2.2941828254847647,
+      "grad_norm": 0.7536196112632751,
+      "learning_rate": 3.4239682383884603e-06,
+      "loss": 0.6326,
+      "step": 8282
+    },
+    {
+      "epoch": 2.2944598337950137,
+      "grad_norm": 0.7260246872901917,
+      "learning_rate": 3.423629736538685e-06,
+      "loss": 0.555,
+      "step": 8283
+    },
+    {
+      "epoch": 2.294736842105263,
+      "grad_norm": 0.7225943207740784,
+      "learning_rate": 3.4232912150778914e-06,
+      "loss": 0.6272,
+      "step": 8284
+    },
+    {
+      "epoch": 2.2950138504155126,
+      "grad_norm": 0.7318356037139893,
+      "learning_rate": 3.4229526740132672e-06,
+      "loss": 0.6487,
+      "step": 8285
+    },
+    {
+      "epoch": 2.2952908587257617,
+      "grad_norm": 0.6813390851020813,
+      "learning_rate": 3.4226141133520014e-06,
+      "loss": 0.5994,
+      "step": 8286
+    },
+    {
+      "epoch": 2.295567867036011,
+      "grad_norm": 0.7114197015762329,
+      "learning_rate": 3.4222755331012813e-06,
+      "loss": 0.5947,
+      "step": 8287
+    },
+    {
+      "epoch": 2.29584487534626,
+      "grad_norm": 0.7054829597473145,
+      "learning_rate": 3.4219369332682967e-06,
+      "loss": 0.6042,
+      "step": 8288
+    },
+    {
+      "epoch": 2.2961218836565096,
+      "grad_norm": 0.7095068693161011,
+      "learning_rate": 3.421598313860237e-06,
+      "loss": 0.6166,
+      "step": 8289
+    },
+    {
+      "epoch": 2.296398891966759,
+      "grad_norm": 0.7199954390525818,
+      "learning_rate": 3.421259674884291e-06,
+      "loss": 0.6092,
+      "step": 8290
+    },
+    {
+      "epoch": 2.296675900277008,
+      "grad_norm": 0.7267475128173828,
+      "learning_rate": 3.420921016347649e-06,
+      "loss": 0.6098,
+      "step": 8291
+    },
+    {
+      "epoch": 2.2969529085872575,
+      "grad_norm": 0.7270237803459167,
+      "learning_rate": 3.420582338257504e-06,
+      "loss": 0.6621,
+      "step": 8292
+    },
+    {
+      "epoch": 2.297229916897507,
+      "grad_norm": 0.7058791518211365,
+      "learning_rate": 3.420243640621043e-06,
+      "loss": 0.58,
+      "step": 8293
+    },
+    {
+      "epoch": 2.297506925207756,
+      "grad_norm": 0.6881886124610901,
+      "learning_rate": 3.419904923445461e-06,
+      "loss": 0.5873,
+      "step": 8294
+    },
+    {
+      "epoch": 2.2977839335180055,
+      "grad_norm": 0.70926433801651,
+      "learning_rate": 3.419566186737947e-06,
+      "loss": 0.6075,
+      "step": 8295
+    },
+    {
+      "epoch": 2.298060941828255,
+      "grad_norm": 0.7379501461982727,
+      "learning_rate": 3.4192274305056956e-06,
+      "loss": 0.5708,
+      "step": 8296
+    },
+    {
+      "epoch": 2.298337950138504,
+      "grad_norm": 0.7627331614494324,
+      "learning_rate": 3.4188886547558974e-06,
+      "loss": 0.6277,
+      "step": 8297
+    },
+    {
+      "epoch": 2.2986149584487534,
+      "grad_norm": 0.7369683384895325,
+      "learning_rate": 3.418549859495747e-06,
+      "loss": 0.6135,
+      "step": 8298
+    },
+    {
+      "epoch": 2.298891966759003,
+      "grad_norm": 0.7661909461021423,
+      "learning_rate": 3.4182110447324374e-06,
+      "loss": 0.603,
+      "step": 8299
+    },
+    {
+      "epoch": 2.299168975069252,
+      "grad_norm": 0.7049823999404907,
+      "learning_rate": 3.417872210473162e-06,
+      "loss": 0.6006,
+      "step": 8300
+    },
+    {
+      "epoch": 2.2994459833795013,
+      "grad_norm": 0.7327120304107666,
+      "learning_rate": 3.4175333567251163e-06,
+      "loss": 0.5925,
+      "step": 8301
+    },
+    {
+      "epoch": 2.299722991689751,
+      "grad_norm": 0.7394639253616333,
+      "learning_rate": 3.417194483495494e-06,
+      "loss": 0.6356,
+      "step": 8302
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7379699349403381,
+      "learning_rate": 3.4168555907914904e-06,
+      "loss": 0.6067,
+      "step": 8303
+    },
+    {
+      "epoch": 2.3002770083102493,
+      "grad_norm": 0.7700881958007812,
+      "learning_rate": 3.4165166786203012e-06,
+      "loss": 0.6182,
+      "step": 8304
+    },
+    {
+      "epoch": 2.3005540166204987,
+      "grad_norm": 0.7343998551368713,
+      "learning_rate": 3.4161777469891227e-06,
+      "loss": 0.5753,
+      "step": 8305
+    },
+    {
+      "epoch": 2.3008310249307478,
+      "grad_norm": 0.7086063623428345,
+      "learning_rate": 3.415838795905151e-06,
+      "loss": 0.5989,
+      "step": 8306
+    },
+    {
+      "epoch": 2.301108033240997,
+      "grad_norm": 0.7197765111923218,
+      "learning_rate": 3.4154998253755823e-06,
+      "loss": 0.6181,
+      "step": 8307
+    },
+    {
+      "epoch": 2.3013850415512467,
+      "grad_norm": 0.7595135569572449,
+      "learning_rate": 3.415160835407615e-06,
+      "loss": 0.6289,
+      "step": 8308
+    },
+    {
+      "epoch": 2.3016620498614957,
+      "grad_norm": 0.7638220191001892,
+      "learning_rate": 3.414821826008446e-06,
+      "loss": 0.6284,
+      "step": 8309
+    },
+    {
+      "epoch": 2.301939058171745,
+      "grad_norm": 0.7343319058418274,
+      "learning_rate": 3.414482797185273e-06,
+      "loss": 0.5967,
+      "step": 8310
+    },
+    {
+      "epoch": 2.3022160664819946,
+      "grad_norm": 0.7228246331214905,
+      "learning_rate": 3.4141437489452955e-06,
+      "loss": 0.6157,
+      "step": 8311
+    },
+    {
+      "epoch": 2.3024930747922436,
+      "grad_norm": 0.7245225310325623,
+      "learning_rate": 3.413804681295711e-06,
+      "loss": 0.5779,
+      "step": 8312
+    },
+    {
+      "epoch": 2.302770083102493,
+      "grad_norm": 0.749394416809082,
+      "learning_rate": 3.4134655942437197e-06,
+      "loss": 0.6285,
+      "step": 8313
+    },
+    {
+      "epoch": 2.3030470914127426,
+      "grad_norm": 0.6916159987449646,
+      "learning_rate": 3.413126487796522e-06,
+      "loss": 0.6378,
+      "step": 8314
+    },
+    {
+      "epoch": 2.3033240997229916,
+      "grad_norm": 0.7362995147705078,
+      "learning_rate": 3.4127873619613154e-06,
+      "loss": 0.6086,
+      "step": 8315
+    },
+    {
+      "epoch": 2.303601108033241,
+      "grad_norm": 0.758251965045929,
+      "learning_rate": 3.412448216745303e-06,
+      "loss": 0.5945,
+      "step": 8316
+    },
+    {
+      "epoch": 2.3038781163434905,
+      "grad_norm": 0.7599399089813232,
+      "learning_rate": 3.4121090521556856e-06,
+      "loss": 0.636,
+      "step": 8317
+    },
+    {
+      "epoch": 2.3041551246537395,
+      "grad_norm": 0.7391661405563354,
+      "learning_rate": 3.411769868199663e-06,
+      "loss": 0.6042,
+      "step": 8318
+    },
+    {
+      "epoch": 2.304432132963989,
+      "grad_norm": 0.7565715312957764,
+      "learning_rate": 3.4114306648844374e-06,
+      "loss": 0.6037,
+      "step": 8319
+    },
+    {
+      "epoch": 2.3047091412742384,
+      "grad_norm": 0.7810177206993103,
+      "learning_rate": 3.411091442217212e-06,
+      "loss": 0.5512,
+      "step": 8320
+    },
+    {
+      "epoch": 2.3049861495844874,
+      "grad_norm": 0.7262973785400391,
+      "learning_rate": 3.410752200205188e-06,
+      "loss": 0.5859,
+      "step": 8321
+    },
+    {
+      "epoch": 2.305263157894737,
+      "grad_norm": 0.7462362051010132,
+      "learning_rate": 3.4104129388555695e-06,
+      "loss": 0.6192,
+      "step": 8322
+    },
+    {
+      "epoch": 2.3055401662049864,
+      "grad_norm": 0.7291772365570068,
+      "learning_rate": 3.4100736581755587e-06,
+      "loss": 0.5888,
+      "step": 8323
+    },
+    {
+      "epoch": 2.3058171745152354,
+      "grad_norm": 0.7448986172676086,
+      "learning_rate": 3.4097343581723607e-06,
+      "loss": 0.5408,
+      "step": 8324
+    },
+    {
+      "epoch": 2.306094182825485,
+      "grad_norm": 0.6898923516273499,
+      "learning_rate": 3.409395038853179e-06,
+      "loss": 0.542,
+      "step": 8325
+    },
+    {
+      "epoch": 2.3063711911357343,
+      "grad_norm": 0.7469924092292786,
+      "learning_rate": 3.409055700225218e-06,
+      "loss": 0.6154,
+      "step": 8326
+    },
+    {
+      "epoch": 2.3066481994459833,
+      "grad_norm": 0.7666891813278198,
+      "learning_rate": 3.408716342295683e-06,
+      "loss": 0.6176,
+      "step": 8327
+    },
+    {
+      "epoch": 2.3069252077562328,
+      "grad_norm": 0.7476541996002197,
+      "learning_rate": 3.408376965071779e-06,
+      "loss": 0.5989,
+      "step": 8328
+    },
+    {
+      "epoch": 2.3072022160664822,
+      "grad_norm": 0.7600588798522949,
+      "learning_rate": 3.408037568560713e-06,
+      "loss": 0.5709,
+      "step": 8329
+    },
+    {
+      "epoch": 2.3074792243767313,
+      "grad_norm": 0.7915143370628357,
+      "learning_rate": 3.4076981527696904e-06,
+      "loss": 0.5905,
+      "step": 8330
+    },
+    {
+      "epoch": 2.3077562326869807,
+      "grad_norm": 0.7339158654212952,
+      "learning_rate": 3.407358717705918e-06,
+      "loss": 0.6444,
+      "step": 8331
+    },
+    {
+      "epoch": 2.3080332409972297,
+      "grad_norm": 0.7408339381217957,
+      "learning_rate": 3.4070192633766025e-06,
+      "loss": 0.6151,
+      "step": 8332
+    },
+    {
+      "epoch": 2.308310249307479,
+      "grad_norm": 0.7135486006736755,
+      "learning_rate": 3.4066797897889525e-06,
+      "loss": 0.5706,
+      "step": 8333
+    },
+    {
+      "epoch": 2.3085872576177286,
+      "grad_norm": 0.7659527659416199,
+      "learning_rate": 3.406340296950175e-06,
+      "loss": 0.618,
+      "step": 8334
+    },
+    {
+      "epoch": 2.3088642659279777,
+      "grad_norm": 0.7388715744018555,
+      "learning_rate": 3.406000784867478e-06,
+      "loss": 0.6056,
+      "step": 8335
+    },
+    {
+      "epoch": 2.309141274238227,
+      "grad_norm": 0.6996129155158997,
+      "learning_rate": 3.4056612535480708e-06,
+      "loss": 0.5716,
+      "step": 8336
+    },
+    {
+      "epoch": 2.3094182825484766,
+      "grad_norm": 0.7552042007446289,
+      "learning_rate": 3.4053217029991626e-06,
+      "loss": 0.6436,
+      "step": 8337
+    },
+    {
+      "epoch": 2.3096952908587256,
+      "grad_norm": 0.7340453863143921,
+      "learning_rate": 3.4049821332279627e-06,
+      "loss": 0.6117,
+      "step": 8338
+    },
+    {
+      "epoch": 2.309972299168975,
+      "grad_norm": 0.7310197949409485,
+      "learning_rate": 3.4046425442416807e-06,
+      "loss": 0.5781,
+      "step": 8339
+    },
+    {
+      "epoch": 2.3102493074792245,
+      "grad_norm": 0.7644363641738892,
+      "learning_rate": 3.4043029360475277e-06,
+      "loss": 0.568,
+      "step": 8340
+    },
+    {
+      "epoch": 2.3105263157894735,
+      "grad_norm": 0.7341253161430359,
+      "learning_rate": 3.4039633086527137e-06,
+      "loss": 0.6255,
+      "step": 8341
+    },
+    {
+      "epoch": 2.310803324099723,
+      "grad_norm": 0.7020862698554993,
+      "learning_rate": 3.4036236620644503e-06,
+      "loss": 0.5814,
+      "step": 8342
+    },
+    {
+      "epoch": 2.3110803324099725,
+      "grad_norm": 0.7140727639198303,
+      "learning_rate": 3.4032839962899488e-06,
+      "loss": 0.5942,
+      "step": 8343
+    },
+    {
+      "epoch": 2.3113573407202215,
+      "grad_norm": 0.7498226165771484,
+      "learning_rate": 3.402944311336421e-06,
+      "loss": 0.6625,
+      "step": 8344
+    },
+    {
+      "epoch": 2.311634349030471,
+      "grad_norm": 0.7233909964561462,
+      "learning_rate": 3.4026046072110803e-06,
+      "loss": 0.5978,
+      "step": 8345
+    },
+    {
+      "epoch": 2.3119113573407204,
+      "grad_norm": 0.6893860697746277,
+      "learning_rate": 3.4022648839211392e-06,
+      "loss": 0.5497,
+      "step": 8346
+    },
+    {
+      "epoch": 2.3121883656509694,
+      "grad_norm": 0.7233559489250183,
+      "learning_rate": 3.4019251414738095e-06,
+      "loss": 0.5462,
+      "step": 8347
+    },
+    {
+      "epoch": 2.312465373961219,
+      "grad_norm": 0.7121790051460266,
+      "learning_rate": 3.401585379876306e-06,
+      "loss": 0.6133,
+      "step": 8348
+    },
+    {
+      "epoch": 2.3127423822714683,
+      "grad_norm": 0.7205917835235596,
+      "learning_rate": 3.401245599135843e-06,
+      "loss": 0.5382,
+      "step": 8349
+    },
+    {
+      "epoch": 2.3130193905817173,
+      "grad_norm": 0.7295979261398315,
+      "learning_rate": 3.400905799259634e-06,
+      "loss": 0.601,
+      "step": 8350
+    },
+    {
+      "epoch": 2.313296398891967,
+      "grad_norm": 0.7754128575325012,
+      "learning_rate": 3.400565980254894e-06,
+      "loss": 0.6109,
+      "step": 8351
+    },
+    {
+      "epoch": 2.313573407202216,
+      "grad_norm": 0.7137840390205383,
+      "learning_rate": 3.4002261421288386e-06,
+      "loss": 0.5897,
+      "step": 8352
+    },
+    {
+      "epoch": 2.3138504155124653,
+      "grad_norm": 0.7415006756782532,
+      "learning_rate": 3.3998862848886837e-06,
+      "loss": 0.6249,
+      "step": 8353
+    },
+    {
+      "epoch": 2.3141274238227147,
+      "grad_norm": 0.7502437233924866,
+      "learning_rate": 3.3995464085416446e-06,
+      "loss": 0.6291,
+      "step": 8354
+    },
+    {
+      "epoch": 2.3144044321329638,
+      "grad_norm": 0.7403385043144226,
+      "learning_rate": 3.3992065130949374e-06,
+      "loss": 0.6393,
+      "step": 8355
+    },
+    {
+      "epoch": 2.3146814404432132,
+      "grad_norm": 0.7676597833633423,
+      "learning_rate": 3.3988665985557806e-06,
+      "loss": 0.5729,
+      "step": 8356
+    },
+    {
+      "epoch": 2.3149584487534627,
+      "grad_norm": 0.7246792912483215,
+      "learning_rate": 3.39852666493139e-06,
+      "loss": 0.6401,
+      "step": 8357
+    },
+    {
+      "epoch": 2.3152354570637117,
+      "grad_norm": 0.7207006216049194,
+      "learning_rate": 3.398186712228984e-06,
+      "loss": 0.6507,
+      "step": 8358
+    },
+    {
+      "epoch": 2.315512465373961,
+      "grad_norm": 0.736382007598877,
+      "learning_rate": 3.3978467404557804e-06,
+      "loss": 0.622,
+      "step": 8359
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.7351194620132446,
+      "learning_rate": 3.3975067496189967e-06,
+      "loss": 0.568,
+      "step": 8360
+    },
+    {
+      "epoch": 2.3160664819944596,
+      "grad_norm": 0.7116756439208984,
+      "learning_rate": 3.397166739725854e-06,
+      "loss": 0.6029,
+      "step": 8361
+    },
+    {
+      "epoch": 2.316343490304709,
+      "grad_norm": 0.7074264883995056,
+      "learning_rate": 3.3968267107835694e-06,
+      "loss": 0.5407,
+      "step": 8362
+    },
+    {
+      "epoch": 2.3166204986149586,
+      "grad_norm": 0.7448917627334595,
+      "learning_rate": 3.396486662799364e-06,
+      "loss": 0.5873,
+      "step": 8363
+    },
+    {
+      "epoch": 2.3168975069252076,
+      "grad_norm": 0.7695679068565369,
+      "learning_rate": 3.3961465957804576e-06,
+      "loss": 0.6348,
+      "step": 8364
+    },
+    {
+      "epoch": 2.317174515235457,
+      "grad_norm": 0.7828521728515625,
+      "learning_rate": 3.3958065097340705e-06,
+      "loss": 0.6235,
+      "step": 8365
+    },
+    {
+      "epoch": 2.3174515235457065,
+      "grad_norm": 0.7109050154685974,
+      "learning_rate": 3.395466404667422e-06,
+      "loss": 0.6131,
+      "step": 8366
+    },
+    {
+      "epoch": 2.3177285318559555,
+      "grad_norm": 0.7334344983100891,
+      "learning_rate": 3.3951262805877366e-06,
+      "loss": 0.5505,
+      "step": 8367
+    },
+    {
+      "epoch": 2.318005540166205,
+      "grad_norm": 0.7734163999557495,
+      "learning_rate": 3.394786137502234e-06,
+      "loss": 0.5752,
+      "step": 8368
+    },
+    {
+      "epoch": 2.3182825484764544,
+      "grad_norm": 0.7226410508155823,
+      "learning_rate": 3.3944459754181358e-06,
+      "loss": 0.5992,
+      "step": 8369
+    },
+    {
+      "epoch": 2.3185595567867034,
+      "grad_norm": 0.7585699558258057,
+      "learning_rate": 3.3941057943426662e-06,
+      "loss": 0.5907,
+      "step": 8370
+    },
+    {
+      "epoch": 2.318836565096953,
+      "grad_norm": 0.7304117679595947,
+      "learning_rate": 3.393765594283048e-06,
+      "loss": 0.574,
+      "step": 8371
+    },
+    {
+      "epoch": 2.3191135734072024,
+      "grad_norm": 0.7299443483352661,
+      "learning_rate": 3.393425375246503e-06,
+      "loss": 0.5973,
+      "step": 8372
+    },
+    {
+      "epoch": 2.3193905817174514,
+      "grad_norm": 0.7117907404899597,
+      "learning_rate": 3.3930851372402553e-06,
+      "loss": 0.5968,
+      "step": 8373
+    },
+    {
+      "epoch": 2.319667590027701,
+      "grad_norm": 0.6997157335281372,
+      "learning_rate": 3.3927448802715294e-06,
+      "loss": 0.5612,
+      "step": 8374
+    },
+    {
+      "epoch": 2.3199445983379503,
+      "grad_norm": 0.7486206889152527,
+      "learning_rate": 3.3924046043475504e-06,
+      "loss": 0.6317,
+      "step": 8375
+    },
+    {
+      "epoch": 2.3202216066481993,
+      "grad_norm": 0.7381830811500549,
+      "learning_rate": 3.392064309475543e-06,
+      "loss": 0.6116,
+      "step": 8376
+    },
+    {
+      "epoch": 2.320498614958449,
+      "grad_norm": 0.7718245387077332,
+      "learning_rate": 3.3917239956627313e-06,
+      "loss": 0.6098,
+      "step": 8377
+    },
+    {
+      "epoch": 2.3207756232686982,
+      "grad_norm": 0.7560763359069824,
+      "learning_rate": 3.391383662916343e-06,
+      "loss": 0.6764,
+      "step": 8378
+    },
+    {
+      "epoch": 2.3210526315789473,
+      "grad_norm": 0.7376003861427307,
+      "learning_rate": 3.3910433112436023e-06,
+      "loss": 0.6091,
+      "step": 8379
+    },
+    {
+      "epoch": 2.3213296398891967,
+      "grad_norm": 0.7103112936019897,
+      "learning_rate": 3.3907029406517373e-06,
+      "loss": 0.6179,
+      "step": 8380
+    },
+    {
+      "epoch": 2.321606648199446,
+      "grad_norm": 0.7458259463310242,
+      "learning_rate": 3.3903625511479745e-06,
+      "loss": 0.6336,
+      "step": 8381
+    },
+    {
+      "epoch": 2.321883656509695,
+      "grad_norm": 0.763077437877655,
+      "learning_rate": 3.39002214273954e-06,
+      "loss": 0.6306,
+      "step": 8382
+    },
+    {
+      "epoch": 2.3221606648199447,
+      "grad_norm": 0.7884553670883179,
+      "learning_rate": 3.3896817154336633e-06,
+      "loss": 0.6052,
+      "step": 8383
+    },
+    {
+      "epoch": 2.322437673130194,
+      "grad_norm": 0.7037105560302734,
+      "learning_rate": 3.3893412692375716e-06,
+      "loss": 0.5635,
+      "step": 8384
+    },
+    {
+      "epoch": 2.322714681440443,
+      "grad_norm": 0.7140675783157349,
+      "learning_rate": 3.389000804158493e-06,
+      "loss": 0.617,
+      "step": 8385
+    },
+    {
+      "epoch": 2.3229916897506926,
+      "grad_norm": 0.7219676375389099,
+      "learning_rate": 3.388660320203657e-06,
+      "loss": 0.5544,
+      "step": 8386
+    },
+    {
+      "epoch": 2.323268698060942,
+      "grad_norm": 0.7285830974578857,
+      "learning_rate": 3.3883198173802945e-06,
+      "loss": 0.5882,
+      "step": 8387
+    },
+    {
+      "epoch": 2.323545706371191,
+      "grad_norm": 0.71585613489151,
+      "learning_rate": 3.3879792956956325e-06,
+      "loss": 0.6073,
+      "step": 8388
+    },
+    {
+      "epoch": 2.3238227146814405,
+      "grad_norm": 0.7456919550895691,
+      "learning_rate": 3.3876387551569027e-06,
+      "loss": 0.6009,
+      "step": 8389
+    },
+    {
+      "epoch": 2.32409972299169,
+      "grad_norm": 0.7833950519561768,
+      "learning_rate": 3.387298195771336e-06,
+      "loss": 0.5607,
+      "step": 8390
+    },
+    {
+      "epoch": 2.324376731301939,
+      "grad_norm": 0.7337169051170349,
+      "learning_rate": 3.386957617546162e-06,
+      "loss": 0.5897,
+      "step": 8391
+    },
+    {
+      "epoch": 2.3246537396121885,
+      "grad_norm": 0.7572849988937378,
+      "learning_rate": 3.386617020488613e-06,
+      "loss": 0.6423,
+      "step": 8392
+    },
+    {
+      "epoch": 2.3249307479224375,
+      "grad_norm": 0.7516975402832031,
+      "learning_rate": 3.38627640460592e-06,
+      "loss": 0.657,
+      "step": 8393
+    },
+    {
+      "epoch": 2.325207756232687,
+      "grad_norm": 0.764487624168396,
+      "learning_rate": 3.3859357699053165e-06,
+      "loss": 0.6049,
+      "step": 8394
+    },
+    {
+      "epoch": 2.3254847645429364,
+      "grad_norm": 0.717101514339447,
+      "learning_rate": 3.3855951163940333e-06,
+      "loss": 0.5653,
+      "step": 8395
+    },
+    {
+      "epoch": 2.3257617728531854,
+      "grad_norm": 0.6801470518112183,
+      "learning_rate": 3.385254444079304e-06,
+      "loss": 0.5768,
+      "step": 8396
+    },
+    {
+      "epoch": 2.326038781163435,
+      "grad_norm": 0.7541027665138245,
+      "learning_rate": 3.3849137529683627e-06,
+      "loss": 0.587,
+      "step": 8397
+    },
+    {
+      "epoch": 2.3263157894736843,
+      "grad_norm": 0.7836506962776184,
+      "learning_rate": 3.384573043068443e-06,
+      "loss": 0.5845,
+      "step": 8398
+    },
+    {
+      "epoch": 2.3265927977839334,
+      "grad_norm": 0.6979066729545593,
+      "learning_rate": 3.384232314386778e-06,
+      "loss": 0.5293,
+      "step": 8399
+    },
+    {
+      "epoch": 2.326869806094183,
+      "grad_norm": 0.749910295009613,
+      "learning_rate": 3.3838915669306038e-06,
+      "loss": 0.5824,
+      "step": 8400
+    },
+    {
+      "epoch": 2.3271468144044323,
+      "grad_norm": 0.71835857629776,
+      "learning_rate": 3.3835508007071534e-06,
+      "loss": 0.5732,
+      "step": 8401
+    },
+    {
+      "epoch": 2.3274238227146813,
+      "grad_norm": 0.742140531539917,
+      "learning_rate": 3.3832100157236635e-06,
+      "loss": 0.5952,
+      "step": 8402
+    },
+    {
+      "epoch": 2.3277008310249307,
+      "grad_norm": 0.7104704976081848,
+      "learning_rate": 3.3828692119873694e-06,
+      "loss": 0.5742,
+      "step": 8403
+    },
+    {
+      "epoch": 2.32797783933518,
+      "grad_norm": 0.7173059582710266,
+      "learning_rate": 3.3825283895055073e-06,
+      "loss": 0.6068,
+      "step": 8404
+    },
+    {
+      "epoch": 2.3282548476454292,
+      "grad_norm": 0.7052407264709473,
+      "learning_rate": 3.382187548285314e-06,
+      "loss": 0.5935,
+      "step": 8405
+    },
+    {
+      "epoch": 2.3285318559556787,
+      "grad_norm": 0.7657210826873779,
+      "learning_rate": 3.381846688334026e-06,
+      "loss": 0.6105,
+      "step": 8406
+    },
+    {
+      "epoch": 2.328808864265928,
+      "grad_norm": 0.7314922213554382,
+      "learning_rate": 3.381505809658881e-06,
+      "loss": 0.5613,
+      "step": 8407
+    },
+    {
+      "epoch": 2.329085872576177,
+      "grad_norm": 0.7196123600006104,
+      "learning_rate": 3.3811649122671166e-06,
+      "loss": 0.6182,
+      "step": 8408
+    },
+    {
+      "epoch": 2.3293628808864266,
+      "grad_norm": 0.7536889910697937,
+      "learning_rate": 3.38082399616597e-06,
+      "loss": 0.6235,
+      "step": 8409
+    },
+    {
+      "epoch": 2.329639889196676,
+      "grad_norm": 0.7460631132125854,
+      "learning_rate": 3.380483061362682e-06,
+      "loss": 0.6351,
+      "step": 8410
+    },
+    {
+      "epoch": 2.329916897506925,
+      "grad_norm": 0.7032564878463745,
+      "learning_rate": 3.380142107864489e-06,
+      "loss": 0.5941,
+      "step": 8411
+    },
+    {
+      "epoch": 2.3301939058171746,
+      "grad_norm": 0.7191113233566284,
+      "learning_rate": 3.379801135678632e-06,
+      "loss": 0.6226,
+      "step": 8412
+    },
+    {
+      "epoch": 2.3304709141274236,
+      "grad_norm": 0.7230985760688782,
+      "learning_rate": 3.3794601448123498e-06,
+      "loss": 0.601,
+      "step": 8413
+    },
+    {
+      "epoch": 2.330747922437673,
+      "grad_norm": 0.7542722225189209,
+      "learning_rate": 3.3791191352728823e-06,
+      "loss": 0.6325,
+      "step": 8414
+    },
+    {
+      "epoch": 2.3310249307479225,
+      "grad_norm": 0.7551451325416565,
+      "learning_rate": 3.3787781070674713e-06,
+      "loss": 0.6039,
+      "step": 8415
+    },
+    {
+      "epoch": 2.3313019390581715,
+      "grad_norm": 0.7307731509208679,
+      "learning_rate": 3.3784370602033572e-06,
+      "loss": 0.6207,
+      "step": 8416
+    },
+    {
+      "epoch": 2.331578947368421,
+      "grad_norm": 0.7245236039161682,
+      "learning_rate": 3.3780959946877806e-06,
+      "loss": 0.6177,
+      "step": 8417
+    },
+    {
+      "epoch": 2.3318559556786704,
+      "grad_norm": 0.7293622493743896,
+      "learning_rate": 3.3777549105279834e-06,
+      "loss": 0.6165,
+      "step": 8418
+    },
+    {
+      "epoch": 2.3321329639889194,
+      "grad_norm": 0.7345945239067078,
+      "learning_rate": 3.3774138077312087e-06,
+      "loss": 0.5866,
+      "step": 8419
+    },
+    {
+      "epoch": 2.332409972299169,
+      "grad_norm": 0.7424420118331909,
+      "learning_rate": 3.3770726863046973e-06,
+      "loss": 0.6303,
+      "step": 8420
+    },
+    {
+      "epoch": 2.3326869806094184,
+      "grad_norm": 0.7340677976608276,
+      "learning_rate": 3.376731546255693e-06,
+      "loss": 0.6184,
+      "step": 8421
+    },
+    {
+      "epoch": 2.3329639889196674,
+      "grad_norm": 0.6973327398300171,
+      "learning_rate": 3.37639038759144e-06,
+      "loss": 0.5948,
+      "step": 8422
+    },
+    {
+      "epoch": 2.333240997229917,
+      "grad_norm": 0.7196833491325378,
+      "learning_rate": 3.3760492103191806e-06,
+      "loss": 0.5884,
+      "step": 8423
+    },
+    {
+      "epoch": 2.3335180055401663,
+      "grad_norm": 0.7527026534080505,
+      "learning_rate": 3.3757080144461583e-06,
+      "loss": 0.5772,
+      "step": 8424
+    },
+    {
+      "epoch": 2.3337950138504153,
+      "grad_norm": 0.7155172228813171,
+      "learning_rate": 3.3753667999796193e-06,
+      "loss": 0.6154,
+      "step": 8425
+    },
+    {
+      "epoch": 2.334072022160665,
+      "grad_norm": 0.7029452323913574,
+      "learning_rate": 3.375025566926808e-06,
+      "loss": 0.6057,
+      "step": 8426
+    },
+    {
+      "epoch": 2.3343490304709142,
+      "grad_norm": 0.7369312644004822,
+      "learning_rate": 3.3746843152949687e-06,
+      "loss": 0.542,
+      "step": 8427
+    },
+    {
+      "epoch": 2.3346260387811633,
+      "grad_norm": 0.6887011528015137,
+      "learning_rate": 3.3743430450913485e-06,
+      "loss": 0.5605,
+      "step": 8428
+    },
+    {
+      "epoch": 2.3349030470914127,
+      "grad_norm": 0.7363905310630798,
+      "learning_rate": 3.374001756323192e-06,
+      "loss": 0.5791,
+      "step": 8429
+    },
+    {
+      "epoch": 2.335180055401662,
+      "grad_norm": 0.6778358817100525,
+      "learning_rate": 3.3736604489977465e-06,
+      "loss": 0.5713,
+      "step": 8430
+    },
+    {
+      "epoch": 2.335457063711911,
+      "grad_norm": 0.7384731769561768,
+      "learning_rate": 3.3733191231222577e-06,
+      "loss": 0.6085,
+      "step": 8431
+    },
+    {
+      "epoch": 2.3357340720221607,
+      "grad_norm": 0.7593647241592407,
+      "learning_rate": 3.3729777787039752e-06,
+      "loss": 0.6202,
+      "step": 8432
+    },
+    {
+      "epoch": 2.33601108033241,
+      "grad_norm": 0.7375938296318054,
+      "learning_rate": 3.372636415750144e-06,
+      "loss": 0.6186,
+      "step": 8433
+    },
+    {
+      "epoch": 2.336288088642659,
+      "grad_norm": 0.7485685348510742,
+      "learning_rate": 3.372295034268014e-06,
+      "loss": 0.5942,
+      "step": 8434
+    },
+    {
+      "epoch": 2.3365650969529086,
+      "grad_norm": 0.7516508102416992,
+      "learning_rate": 3.3719536342648323e-06,
+      "loss": 0.5925,
+      "step": 8435
+    },
+    {
+      "epoch": 2.336842105263158,
+      "grad_norm": 0.7607292532920837,
+      "learning_rate": 3.3716122157478488e-06,
+      "loss": 0.606,
+      "step": 8436
+    },
+    {
+      "epoch": 2.337119113573407,
+      "grad_norm": 0.7966906428337097,
+      "learning_rate": 3.371270778724311e-06,
+      "loss": 0.6508,
+      "step": 8437
+    },
+    {
+      "epoch": 2.3373961218836565,
+      "grad_norm": 0.7265411019325256,
+      "learning_rate": 3.3709293232014705e-06,
+      "loss": 0.6312,
+      "step": 8438
+    },
+    {
+      "epoch": 2.337673130193906,
+      "grad_norm": 0.7127956748008728,
+      "learning_rate": 3.370587849186576e-06,
+      "loss": 0.5349,
+      "step": 8439
+    },
+    {
+      "epoch": 2.337950138504155,
+      "grad_norm": 0.7615678310394287,
+      "learning_rate": 3.370246356686878e-06,
+      "loss": 0.6082,
+      "step": 8440
+    },
+    {
+      "epoch": 2.3382271468144045,
+      "grad_norm": 0.7224056124687195,
+      "learning_rate": 3.3699048457096283e-06,
+      "loss": 0.5952,
+      "step": 8441
+    },
+    {
+      "epoch": 2.338504155124654,
+      "grad_norm": 0.7433934807777405,
+      "learning_rate": 3.3695633162620765e-06,
+      "loss": 0.604,
+      "step": 8442
+    },
+    {
+      "epoch": 2.338781163434903,
+      "grad_norm": 0.767772912979126,
+      "learning_rate": 3.369221768351475e-06,
+      "loss": 0.6144,
+      "step": 8443
+    },
+    {
+      "epoch": 2.3390581717451524,
+      "grad_norm": 0.754507839679718,
+      "learning_rate": 3.3688802019850757e-06,
+      "loss": 0.6141,
+      "step": 8444
+    },
+    {
+      "epoch": 2.339335180055402,
+      "grad_norm": 0.7405969500541687,
+      "learning_rate": 3.368538617170131e-06,
+      "loss": 0.6131,
+      "step": 8445
+    },
+    {
+      "epoch": 2.339612188365651,
+      "grad_norm": 0.6835383176803589,
+      "learning_rate": 3.368197013913893e-06,
+      "loss": 0.5526,
+      "step": 8446
+    },
+    {
+      "epoch": 2.3398891966759003,
+      "grad_norm": 0.7214337587356567,
+      "learning_rate": 3.367855392223616e-06,
+      "loss": 0.6244,
+      "step": 8447
+    },
+    {
+      "epoch": 2.34016620498615,
+      "grad_norm": 0.714373767375946,
+      "learning_rate": 3.367513752106552e-06,
+      "loss": 0.5685,
+      "step": 8448
+    },
+    {
+      "epoch": 2.340443213296399,
+      "grad_norm": 0.7374280095100403,
+      "learning_rate": 3.3671720935699566e-06,
+      "loss": 0.6242,
+      "step": 8449
+    },
+    {
+      "epoch": 2.3407202216066483,
+      "grad_norm": 0.8040729761123657,
+      "learning_rate": 3.3668304166210817e-06,
+      "loss": 0.5767,
+      "step": 8450
+    },
+    {
+      "epoch": 2.3409972299168977,
+      "grad_norm": 0.765451967716217,
+      "learning_rate": 3.366488721267185e-06,
+      "loss": 0.6042,
+      "step": 8451
+    },
+    {
+      "epoch": 2.3412742382271468,
+      "grad_norm": 0.751152515411377,
+      "learning_rate": 3.366147007515519e-06,
+      "loss": 0.6059,
+      "step": 8452
+    },
+    {
+      "epoch": 2.341551246537396,
+      "grad_norm": 0.7184852361679077,
+      "learning_rate": 3.3658052753733406e-06,
+      "loss": 0.6025,
+      "step": 8453
+    },
+    {
+      "epoch": 2.3418282548476457,
+      "grad_norm": 0.7253008484840393,
+      "learning_rate": 3.3654635248479046e-06,
+      "loss": 0.6071,
+      "step": 8454
+    },
+    {
+      "epoch": 2.3421052631578947,
+      "grad_norm": 0.7163746953010559,
+      "learning_rate": 3.365121755946468e-06,
+      "loss": 0.6088,
+      "step": 8455
+    },
+    {
+      "epoch": 2.342382271468144,
+      "grad_norm": 0.7567439675331116,
+      "learning_rate": 3.3647799686762873e-06,
+      "loss": 0.5979,
+      "step": 8456
+    },
+    {
+      "epoch": 2.342659279778393,
+      "grad_norm": 0.7153251767158508,
+      "learning_rate": 3.36443816304462e-06,
+      "loss": 0.612,
+      "step": 8457
+    },
+    {
+      "epoch": 2.3429362880886426,
+      "grad_norm": 0.6974598169326782,
+      "learning_rate": 3.3640963390587223e-06,
+      "loss": 0.5867,
+      "step": 8458
+    },
+    {
+      "epoch": 2.343213296398892,
+      "grad_norm": 0.7750906348228455,
+      "learning_rate": 3.363754496725852e-06,
+      "loss": 0.5887,
+      "step": 8459
+    },
+    {
+      "epoch": 2.343490304709141,
+      "grad_norm": 0.69064861536026,
+      "learning_rate": 3.3634126360532694e-06,
+      "loss": 0.5824,
+      "step": 8460
+    },
+    {
+      "epoch": 2.3437673130193906,
+      "grad_norm": 0.7157204151153564,
+      "learning_rate": 3.3630707570482315e-06,
+      "loss": 0.5771,
+      "step": 8461
+    },
+    {
+      "epoch": 2.34404432132964,
+      "grad_norm": 0.7043836712837219,
+      "learning_rate": 3.362728859717996e-06,
+      "loss": 0.581,
+      "step": 8462
+    },
+    {
+      "epoch": 2.344321329639889,
+      "grad_norm": 0.7589282989501953,
+      "learning_rate": 3.3623869440698254e-06,
+      "loss": 0.5845,
+      "step": 8463
+    },
+    {
+      "epoch": 2.3445983379501385,
+      "grad_norm": 0.7066418528556824,
+      "learning_rate": 3.3620450101109767e-06,
+      "loss": 0.6094,
+      "step": 8464
+    },
+    {
+      "epoch": 2.344875346260388,
+      "grad_norm": 0.7088079452514648,
+      "learning_rate": 3.361703057848711e-06,
+      "loss": 0.5995,
+      "step": 8465
+    },
+    {
+      "epoch": 2.345152354570637,
+      "grad_norm": 0.7533881664276123,
+      "learning_rate": 3.361361087290289e-06,
+      "loss": 0.591,
+      "step": 8466
+    },
+    {
+      "epoch": 2.3454293628808864,
+      "grad_norm": 0.7128363251686096,
+      "learning_rate": 3.3610190984429713e-06,
+      "loss": 0.5998,
+      "step": 8467
+    },
+    {
+      "epoch": 2.345706371191136,
+      "grad_norm": 0.7468311190605164,
+      "learning_rate": 3.360677091314019e-06,
+      "loss": 0.6157,
+      "step": 8468
+    },
+    {
+      "epoch": 2.345983379501385,
+      "grad_norm": 0.7234650254249573,
+      "learning_rate": 3.360335065910695e-06,
+      "loss": 0.6295,
+      "step": 8469
+    },
+    {
+      "epoch": 2.3462603878116344,
+      "grad_norm": 0.7158196568489075,
+      "learning_rate": 3.35999302224026e-06,
+      "loss": 0.5866,
+      "step": 8470
+    },
+    {
+      "epoch": 2.346537396121884,
+      "grad_norm": 0.7302791476249695,
+      "learning_rate": 3.359650960309977e-06,
+      "loss": 0.5765,
+      "step": 8471
+    },
+    {
+      "epoch": 2.346814404432133,
+      "grad_norm": 0.7104723453521729,
+      "learning_rate": 3.3593088801271085e-06,
+      "loss": 0.5691,
+      "step": 8472
+    },
+    {
+      "epoch": 2.3470914127423823,
+      "grad_norm": 0.766150176525116,
+      "learning_rate": 3.358966781698919e-06,
+      "loss": 0.5465,
+      "step": 8473
+    },
+    {
+      "epoch": 2.3473684210526318,
+      "grad_norm": 0.7038397192955017,
+      "learning_rate": 3.3586246650326704e-06,
+      "loss": 0.588,
+      "step": 8474
+    },
+    {
+      "epoch": 2.347645429362881,
+      "grad_norm": 0.7408996820449829,
+      "learning_rate": 3.3582825301356274e-06,
+      "loss": 0.5938,
+      "step": 8475
+    },
+    {
+      "epoch": 2.3479224376731302,
+      "grad_norm": 0.7109339237213135,
+      "learning_rate": 3.3579403770150546e-06,
+      "loss": 0.5421,
+      "step": 8476
+    },
+    {
+      "epoch": 2.3481994459833793,
+      "grad_norm": 0.7604177594184875,
+      "learning_rate": 3.3575982056782167e-06,
+      "loss": 0.5848,
+      "step": 8477
+    },
+    {
+      "epoch": 2.3484764542936287,
+      "grad_norm": 0.7184967398643494,
+      "learning_rate": 3.3572560161323787e-06,
+      "loss": 0.5767,
+      "step": 8478
+    },
+    {
+      "epoch": 2.348753462603878,
+      "grad_norm": 0.7625906467437744,
+      "learning_rate": 3.3569138083848068e-06,
+      "loss": 0.5969,
+      "step": 8479
+    },
+    {
+      "epoch": 2.349030470914127,
+      "grad_norm": 0.6968985795974731,
+      "learning_rate": 3.356571582442766e-06,
+      "loss": 0.5782,
+      "step": 8480
+    },
+    {
+      "epoch": 2.3493074792243767,
+      "grad_norm": 0.7008188962936401,
+      "learning_rate": 3.356229338313524e-06,
+      "loss": 0.606,
+      "step": 8481
+    },
+    {
+      "epoch": 2.349584487534626,
+      "grad_norm": 0.7138946056365967,
+      "learning_rate": 3.355887076004345e-06,
+      "loss": 0.5827,
+      "step": 8482
+    },
+    {
+      "epoch": 2.349861495844875,
+      "grad_norm": 0.7512415051460266,
+      "learning_rate": 3.355544795522498e-06,
+      "loss": 0.6008,
+      "step": 8483
+    },
+    {
+      "epoch": 2.3501385041551246,
+      "grad_norm": 0.7460567355155945,
+      "learning_rate": 3.3552024968752504e-06,
+      "loss": 0.5968,
+      "step": 8484
+    },
+    {
+      "epoch": 2.350415512465374,
+      "grad_norm": 0.7085126638412476,
+      "learning_rate": 3.35486018006987e-06,
+      "loss": 0.6042,
+      "step": 8485
+    },
+    {
+      "epoch": 2.350692520775623,
+      "grad_norm": 0.7720244526863098,
+      "learning_rate": 3.3545178451136253e-06,
+      "loss": 0.6363,
+      "step": 8486
+    },
+    {
+      "epoch": 2.3509695290858725,
+      "grad_norm": 0.705949068069458,
+      "learning_rate": 3.354175492013784e-06,
+      "loss": 0.6077,
+      "step": 8487
+    },
+    {
+      "epoch": 2.351246537396122,
+      "grad_norm": 0.7648314833641052,
+      "learning_rate": 3.353833120777616e-06,
+      "loss": 0.5903,
+      "step": 8488
+    },
+    {
+      "epoch": 2.351523545706371,
+      "grad_norm": 0.7231563925743103,
+      "learning_rate": 3.35349073141239e-06,
+      "loss": 0.607,
+      "step": 8489
+    },
+    {
+      "epoch": 2.3518005540166205,
+      "grad_norm": 0.7560357451438904,
+      "learning_rate": 3.3531483239253765e-06,
+      "loss": 0.6346,
+      "step": 8490
+    },
+    {
+      "epoch": 2.35207756232687,
+      "grad_norm": 0.7295835018157959,
+      "learning_rate": 3.3528058983238447e-06,
+      "loss": 0.6109,
+      "step": 8491
+    },
+    {
+      "epoch": 2.352354570637119,
+      "grad_norm": 0.7257558703422546,
+      "learning_rate": 3.3524634546150673e-06,
+      "loss": 0.6057,
+      "step": 8492
+    },
+    {
+      "epoch": 2.3526315789473684,
+      "grad_norm": 0.7374874353408813,
+      "learning_rate": 3.3521209928063127e-06,
+      "loss": 0.5959,
+      "step": 8493
+    },
+    {
+      "epoch": 2.352908587257618,
+      "grad_norm": 0.7184003591537476,
+      "learning_rate": 3.3517785129048538e-06,
+      "loss": 0.562,
+      "step": 8494
+    },
+    {
+      "epoch": 2.353185595567867,
+      "grad_norm": 0.7235756516456604,
+      "learning_rate": 3.3514360149179616e-06,
+      "loss": 0.6034,
+      "step": 8495
+    },
+    {
+      "epoch": 2.3534626038781163,
+      "grad_norm": 0.7779785394668579,
+      "learning_rate": 3.3510934988529083e-06,
+      "loss": 0.6005,
+      "step": 8496
+    },
+    {
+      "epoch": 2.353739612188366,
+      "grad_norm": 0.7211990356445312,
+      "learning_rate": 3.3507509647169666e-06,
+      "loss": 0.5829,
+      "step": 8497
+    },
+    {
+      "epoch": 2.354016620498615,
+      "grad_norm": 0.7217147350311279,
+      "learning_rate": 3.35040841251741e-06,
+      "loss": 0.6246,
+      "step": 8498
+    },
+    {
+      "epoch": 2.3542936288088643,
+      "grad_norm": 0.7280838489532471,
+      "learning_rate": 3.3500658422615107e-06,
+      "loss": 0.6213,
+      "step": 8499
+    },
+    {
+      "epoch": 2.3545706371191137,
+      "grad_norm": 0.7369555234909058,
+      "learning_rate": 3.3497232539565414e-06,
+      "loss": 0.6067,
+      "step": 8500
+    },
+    {
+      "epoch": 2.3548476454293628,
+      "grad_norm": 0.71979159116745,
+      "learning_rate": 3.3493806476097797e-06,
+      "loss": 0.6007,
+      "step": 8501
+    },
+    {
+      "epoch": 2.355124653739612,
+      "grad_norm": 0.7194019556045532,
+      "learning_rate": 3.349038023228497e-06,
+      "loss": 0.5802,
+      "step": 8502
+    },
+    {
+      "epoch": 2.3554016620498617,
+      "grad_norm": 0.7042348980903625,
+      "learning_rate": 3.3486953808199685e-06,
+      "loss": 0.6007,
+      "step": 8503
+    },
+    {
+      "epoch": 2.3556786703601107,
+      "grad_norm": 0.7309350371360779,
+      "learning_rate": 3.3483527203914694e-06,
+      "loss": 0.6101,
+      "step": 8504
+    },
+    {
+      "epoch": 2.35595567867036,
+      "grad_norm": 0.796105146408081,
+      "learning_rate": 3.3480100419502764e-06,
+      "loss": 0.6098,
+      "step": 8505
+    },
+    {
+      "epoch": 2.3562326869806096,
+      "grad_norm": 0.7181329727172852,
+      "learning_rate": 3.347667345503664e-06,
+      "loss": 0.5815,
+      "step": 8506
+    },
+    {
+      "epoch": 2.3565096952908586,
+      "grad_norm": 0.7414490580558777,
+      "learning_rate": 3.3473246310589097e-06,
+      "loss": 0.5826,
+      "step": 8507
+    },
+    {
+      "epoch": 2.356786703601108,
+      "grad_norm": 0.7326338291168213,
+      "learning_rate": 3.3469818986232893e-06,
+      "loss": 0.6284,
+      "step": 8508
+    },
+    {
+      "epoch": 2.3570637119113576,
+      "grad_norm": 0.7372599244117737,
+      "learning_rate": 3.3466391482040807e-06,
+      "loss": 0.6585,
+      "step": 8509
+    },
+    {
+      "epoch": 2.3573407202216066,
+      "grad_norm": 0.7290738224983215,
+      "learning_rate": 3.34629637980856e-06,
+      "loss": 0.619,
+      "step": 8510
+    },
+    {
+      "epoch": 2.357617728531856,
+      "grad_norm": 0.7647110819816589,
+      "learning_rate": 3.345953593444007e-06,
+      "loss": 0.6146,
+      "step": 8511
+    },
+    {
+      "epoch": 2.3578947368421055,
+      "grad_norm": 0.7436761856079102,
+      "learning_rate": 3.345610789117698e-06,
+      "loss": 0.6048,
+      "step": 8512
+    },
+    {
+      "epoch": 2.3581717451523545,
+      "grad_norm": 0.7494994401931763,
+      "learning_rate": 3.3452679668369126e-06,
+      "loss": 0.6417,
+      "step": 8513
+    },
+    {
+      "epoch": 2.358448753462604,
+      "grad_norm": 0.7640016674995422,
+      "learning_rate": 3.34492512660893e-06,
+      "loss": 0.606,
+      "step": 8514
+    },
+    {
+      "epoch": 2.3587257617728534,
+      "grad_norm": 0.6900897026062012,
+      "learning_rate": 3.3445822684410295e-06,
+      "loss": 0.6021,
+      "step": 8515
+    },
+    {
+      "epoch": 2.3590027700831024,
+      "grad_norm": 0.7223372459411621,
+      "learning_rate": 3.34423939234049e-06,
+      "loss": 0.6225,
+      "step": 8516
+    },
+    {
+      "epoch": 2.359279778393352,
+      "grad_norm": 0.6969015598297119,
+      "learning_rate": 3.3438964983145928e-06,
+      "loss": 0.6112,
+      "step": 8517
+    },
+    {
+      "epoch": 2.359556786703601,
+      "grad_norm": 0.747728705406189,
+      "learning_rate": 3.343553586370618e-06,
+      "loss": 0.6208,
+      "step": 8518
+    },
+    {
+      "epoch": 2.3598337950138504,
+      "grad_norm": 0.7828081250190735,
+      "learning_rate": 3.3432106565158457e-06,
+      "loss": 0.596,
+      "step": 8519
+    },
+    {
+      "epoch": 2.3601108033241,
+      "grad_norm": 0.7545477747917175,
+      "learning_rate": 3.342867708757558e-06,
+      "loss": 0.6248,
+      "step": 8520
+    },
+    {
+      "epoch": 2.360387811634349,
+      "grad_norm": 0.7424569725990295,
+      "learning_rate": 3.3425247431030367e-06,
+      "loss": 0.596,
+      "step": 8521
+    },
+    {
+      "epoch": 2.3606648199445983,
+      "grad_norm": 0.7445389032363892,
+      "learning_rate": 3.3421817595595635e-06,
+      "loss": 0.6547,
+      "step": 8522
+    },
+    {
+      "epoch": 2.3609418282548478,
+      "grad_norm": 0.7122578620910645,
+      "learning_rate": 3.3418387581344204e-06,
+      "loss": 0.6121,
+      "step": 8523
+    },
+    {
+      "epoch": 2.361218836565097,
+      "grad_norm": 0.6893017888069153,
+      "learning_rate": 3.341495738834891e-06,
+      "loss": 0.5736,
+      "step": 8524
+    },
+    {
+      "epoch": 2.3614958448753463,
+      "grad_norm": 0.7213153839111328,
+      "learning_rate": 3.3411527016682575e-06,
+      "loss": 0.6046,
+      "step": 8525
+    },
+    {
+      "epoch": 2.3617728531855957,
+      "grad_norm": 0.7515587210655212,
+      "learning_rate": 3.340809646641805e-06,
+      "loss": 0.587,
+      "step": 8526
+    },
+    {
+      "epoch": 2.3620498614958447,
+      "grad_norm": 0.7319797873497009,
+      "learning_rate": 3.3404665737628165e-06,
+      "loss": 0.5814,
+      "step": 8527
+    },
+    {
+      "epoch": 2.362326869806094,
+      "grad_norm": 0.7438617944717407,
+      "learning_rate": 3.3401234830385753e-06,
+      "loss": 0.672,
+      "step": 8528
+    },
+    {
+      "epoch": 2.3626038781163436,
+      "grad_norm": 0.7243683338165283,
+      "learning_rate": 3.339780374476368e-06,
+      "loss": 0.5919,
+      "step": 8529
+    },
+    {
+      "epoch": 2.3628808864265927,
+      "grad_norm": 0.6928403377532959,
+      "learning_rate": 3.339437248083479e-06,
+      "loss": 0.5576,
+      "step": 8530
+    },
+    {
+      "epoch": 2.363157894736842,
+      "grad_norm": 0.7498994469642639,
+      "learning_rate": 3.339094103867193e-06,
+      "loss": 0.6324,
+      "step": 8531
+    },
+    {
+      "epoch": 2.3634349030470916,
+      "grad_norm": 0.7130890488624573,
+      "learning_rate": 3.3387509418347962e-06,
+      "loss": 0.589,
+      "step": 8532
+    },
+    {
+      "epoch": 2.3637119113573406,
+      "grad_norm": 0.7753434777259827,
+      "learning_rate": 3.338407761993575e-06,
+      "loss": 0.6037,
+      "step": 8533
+    },
+    {
+      "epoch": 2.36398891966759,
+      "grad_norm": 0.7679981589317322,
+      "learning_rate": 3.3380645643508165e-06,
+      "loss": 0.6238,
+      "step": 8534
+    },
+    {
+      "epoch": 2.3642659279778395,
+      "grad_norm": 0.7225174903869629,
+      "learning_rate": 3.3377213489138073e-06,
+      "loss": 0.5852,
+      "step": 8535
+    },
+    {
+      "epoch": 2.3645429362880885,
+      "grad_norm": 0.7681257724761963,
+      "learning_rate": 3.3373781156898333e-06,
+      "loss": 0.5804,
+      "step": 8536
+    },
+    {
+      "epoch": 2.364819944598338,
+      "grad_norm": 0.726750373840332,
+      "learning_rate": 3.337034864686185e-06,
+      "loss": 0.6111,
+      "step": 8537
+    },
+    {
+      "epoch": 2.365096952908587,
+      "grad_norm": 0.7270079255104065,
+      "learning_rate": 3.3366915959101478e-06,
+      "loss": 0.5785,
+      "step": 8538
+    },
+    {
+      "epoch": 2.3653739612188365,
+      "grad_norm": 0.7376360893249512,
+      "learning_rate": 3.3363483093690107e-06,
+      "loss": 0.6256,
+      "step": 8539
+    },
+    {
+      "epoch": 2.365650969529086,
+      "grad_norm": 0.7967438101768494,
+      "learning_rate": 3.3360050050700655e-06,
+      "loss": 0.6157,
+      "step": 8540
+    },
+    {
+      "epoch": 2.365927977839335,
+      "grad_norm": 0.7275079488754272,
+      "learning_rate": 3.3356616830205974e-06,
+      "loss": 0.6113,
+      "step": 8541
+    },
+    {
+      "epoch": 2.3662049861495844,
+      "grad_norm": 0.7312076091766357,
+      "learning_rate": 3.335318343227898e-06,
+      "loss": 0.5816,
+      "step": 8542
+    },
+    {
+      "epoch": 2.366481994459834,
+      "grad_norm": 0.7246550917625427,
+      "learning_rate": 3.3349749856992575e-06,
+      "loss": 0.574,
+      "step": 8543
+    },
+    {
+      "epoch": 2.366759002770083,
+      "grad_norm": 0.7144362926483154,
+      "learning_rate": 3.334631610441965e-06,
+      "loss": 0.6176,
+      "step": 8544
+    },
+    {
+      "epoch": 2.3670360110803323,
+      "grad_norm": 0.7541229724884033,
+      "learning_rate": 3.334288217463312e-06,
+      "loss": 0.5949,
+      "step": 8545
+    },
+    {
+      "epoch": 2.367313019390582,
+      "grad_norm": 0.7368307113647461,
+      "learning_rate": 3.333944806770591e-06,
+      "loss": 0.6445,
+      "step": 8546
+    },
+    {
+      "epoch": 2.367590027700831,
+      "grad_norm": 0.7364919781684875,
+      "learning_rate": 3.333601378371091e-06,
+      "loss": 0.5953,
+      "step": 8547
+    },
+    {
+      "epoch": 2.3678670360110803,
+      "grad_norm": 0.72370445728302,
+      "learning_rate": 3.333257932272105e-06,
+      "loss": 0.6163,
+      "step": 8548
+    },
+    {
+      "epoch": 2.3681440443213297,
+      "grad_norm": 0.711737871170044,
+      "learning_rate": 3.3329144684809257e-06,
+      "loss": 0.5909,
+      "step": 8549
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.7610693573951721,
+      "learning_rate": 3.3325709870048446e-06,
+      "loss": 0.6281,
+      "step": 8550
+    },
+    {
+      "epoch": 2.368698060941828,
+      "grad_norm": 0.7502447366714478,
+      "learning_rate": 3.332227487851155e-06,
+      "loss": 0.6048,
+      "step": 8551
+    },
+    {
+      "epoch": 2.3689750692520777,
+      "grad_norm": 0.7391875386238098,
+      "learning_rate": 3.331883971027151e-06,
+      "loss": 0.5574,
+      "step": 8552
+    },
+    {
+      "epoch": 2.3692520775623267,
+      "grad_norm": 0.7459149956703186,
+      "learning_rate": 3.331540436540126e-06,
+      "loss": 0.5775,
+      "step": 8553
+    },
+    {
+      "epoch": 2.369529085872576,
+      "grad_norm": 0.7106911540031433,
+      "learning_rate": 3.3311968843973738e-06,
+      "loss": 0.5932,
+      "step": 8554
+    },
+    {
+      "epoch": 2.3698060941828256,
+      "grad_norm": 0.7712510824203491,
+      "learning_rate": 3.330853314606189e-06,
+      "loss": 0.6059,
+      "step": 8555
+    },
+    {
+      "epoch": 2.3700831024930746,
+      "grad_norm": 0.7939676642417908,
+      "learning_rate": 3.3305097271738666e-06,
+      "loss": 0.5826,
+      "step": 8556
+    },
+    {
+      "epoch": 2.370360110803324,
+      "grad_norm": 0.7157027125358582,
+      "learning_rate": 3.3301661221077024e-06,
+      "loss": 0.5834,
+      "step": 8557
+    },
+    {
+      "epoch": 2.3706371191135736,
+      "grad_norm": 0.757570743560791,
+      "learning_rate": 3.329822499414991e-06,
+      "loss": 0.6023,
+      "step": 8558
+    },
+    {
+      "epoch": 2.3709141274238226,
+      "grad_norm": 0.7631760835647583,
+      "learning_rate": 3.329478859103029e-06,
+      "loss": 0.6508,
+      "step": 8559
+    },
+    {
+      "epoch": 2.371191135734072,
+      "grad_norm": 0.7067531943321228,
+      "learning_rate": 3.3291352011791116e-06,
+      "loss": 0.608,
+      "step": 8560
+    },
+    {
+      "epoch": 2.3714681440443215,
+      "grad_norm": 0.719627857208252,
+      "learning_rate": 3.3287915256505377e-06,
+      "loss": 0.6121,
+      "step": 8561
+    },
+    {
+      "epoch": 2.3717451523545705,
+      "grad_norm": 0.7045274376869202,
+      "learning_rate": 3.3284478325246027e-06,
+      "loss": 0.6026,
+      "step": 8562
+    },
+    {
+      "epoch": 2.37202216066482,
+      "grad_norm": 0.7519272565841675,
+      "learning_rate": 3.328104121808605e-06,
+      "loss": 0.5889,
+      "step": 8563
+    },
+    {
+      "epoch": 2.3722991689750694,
+      "grad_norm": 0.7602554559707642,
+      "learning_rate": 3.3277603935098413e-06,
+      "loss": 0.5731,
+      "step": 8564
+    },
+    {
+      "epoch": 2.3725761772853184,
+      "grad_norm": 0.7468898296356201,
+      "learning_rate": 3.3274166476356106e-06,
+      "loss": 0.5784,
+      "step": 8565
+    },
+    {
+      "epoch": 2.372853185595568,
+      "grad_norm": 0.7231388688087463,
+      "learning_rate": 3.3270728841932122e-06,
+      "loss": 0.6044,
+      "step": 8566
+    },
+    {
+      "epoch": 2.3731301939058174,
+      "grad_norm": 0.742412269115448,
+      "learning_rate": 3.326729103189944e-06,
+      "loss": 0.6142,
+      "step": 8567
+    },
+    {
+      "epoch": 2.3734072022160664,
+      "grad_norm": 0.7619256377220154,
+      "learning_rate": 3.326385304633105e-06,
+      "loss": 0.6047,
+      "step": 8568
+    },
+    {
+      "epoch": 2.373684210526316,
+      "grad_norm": 0.7784067392349243,
+      "learning_rate": 3.3260414885299967e-06,
+      "loss": 0.6268,
+      "step": 8569
+    },
+    {
+      "epoch": 2.3739612188365653,
+      "grad_norm": 0.7004120349884033,
+      "learning_rate": 3.3256976548879183e-06,
+      "loss": 0.5829,
+      "step": 8570
+    },
+    {
+      "epoch": 2.3742382271468143,
+      "grad_norm": 0.7318438291549683,
+      "learning_rate": 3.3253538037141697e-06,
+      "loss": 0.5994,
+      "step": 8571
+    },
+    {
+      "epoch": 2.3745152354570638,
+      "grad_norm": 0.755072832107544,
+      "learning_rate": 3.3250099350160518e-06,
+      "loss": 0.6282,
+      "step": 8572
+    },
+    {
+      "epoch": 2.3747922437673132,
+      "grad_norm": 0.7150866389274597,
+      "learning_rate": 3.324666048800866e-06,
+      "loss": 0.6095,
+      "step": 8573
+    },
+    {
+      "epoch": 2.3750692520775623,
+      "grad_norm": 0.7552684545516968,
+      "learning_rate": 3.3243221450759143e-06,
+      "loss": 0.6406,
+      "step": 8574
+    },
+    {
+      "epoch": 2.3753462603878117,
+      "grad_norm": 0.7518280148506165,
+      "learning_rate": 3.3239782238484985e-06,
+      "loss": 0.5867,
+      "step": 8575
+    },
+    {
+      "epoch": 2.375623268698061,
+      "grad_norm": 0.74309241771698,
+      "learning_rate": 3.3236342851259208e-06,
+      "loss": 0.5983,
+      "step": 8576
+    },
+    {
+      "epoch": 2.37590027700831,
+      "grad_norm": 0.7382872700691223,
+      "learning_rate": 3.3232903289154835e-06,
+      "loss": 0.609,
+      "step": 8577
+    },
+    {
+      "epoch": 2.3761772853185597,
+      "grad_norm": 0.7038781642913818,
+      "learning_rate": 3.3229463552244905e-06,
+      "loss": 0.5662,
+      "step": 8578
+    },
+    {
+      "epoch": 2.376454293628809,
+      "grad_norm": 0.7423188090324402,
+      "learning_rate": 3.3226023640602446e-06,
+      "loss": 0.5867,
+      "step": 8579
+    },
+    {
+      "epoch": 2.376731301939058,
+      "grad_norm": 0.7664845585823059,
+      "learning_rate": 3.32225835543005e-06,
+      "loss": 0.5634,
+      "step": 8580
+    },
+    {
+      "epoch": 2.3770083102493076,
+      "grad_norm": 0.7739835977554321,
+      "learning_rate": 3.3219143293412104e-06,
+      "loss": 0.617,
+      "step": 8581
+    },
+    {
+      "epoch": 2.3772853185595566,
+      "grad_norm": 0.727225661277771,
+      "learning_rate": 3.3215702858010308e-06,
+      "loss": 0.6212,
+      "step": 8582
+    },
+    {
+      "epoch": 2.377562326869806,
+      "grad_norm": 0.7422259449958801,
+      "learning_rate": 3.321226224816816e-06,
+      "loss": 0.596,
+      "step": 8583
+    },
+    {
+      "epoch": 2.3778393351800555,
+      "grad_norm": 0.788818359375,
+      "learning_rate": 3.320882146395871e-06,
+      "loss": 0.6313,
+      "step": 8584
+    },
+    {
+      "epoch": 2.3781163434903045,
+      "grad_norm": 0.7638080716133118,
+      "learning_rate": 3.320538050545502e-06,
+      "loss": 0.5853,
+      "step": 8585
+    },
+    {
+      "epoch": 2.378393351800554,
+      "grad_norm": 0.6938989162445068,
+      "learning_rate": 3.3201939372730147e-06,
+      "loss": 0.5933,
+      "step": 8586
+    },
+    {
+      "epoch": 2.3786703601108035,
+      "grad_norm": 0.7522276639938354,
+      "learning_rate": 3.3198498065857152e-06,
+      "loss": 0.6793,
+      "step": 8587
+    },
+    {
+      "epoch": 2.3789473684210525,
+      "grad_norm": 0.7383003234863281,
+      "learning_rate": 3.319505658490911e-06,
+      "loss": 0.5879,
+      "step": 8588
+    },
+    {
+      "epoch": 2.379224376731302,
+      "grad_norm": 0.7505053281784058,
+      "learning_rate": 3.319161492995909e-06,
+      "loss": 0.5962,
+      "step": 8589
+    },
+    {
+      "epoch": 2.3795013850415514,
+      "grad_norm": 0.7599852085113525,
+      "learning_rate": 3.318817310108017e-06,
+      "loss": 0.6705,
+      "step": 8590
+    },
+    {
+      "epoch": 2.3797783933518004,
+      "grad_norm": 0.7736787796020508,
+      "learning_rate": 3.3184731098345424e-06,
+      "loss": 0.6063,
+      "step": 8591
+    },
+    {
+      "epoch": 2.38005540166205,
+      "grad_norm": 0.7720979452133179,
+      "learning_rate": 3.3181288921827925e-06,
+      "loss": 0.6008,
+      "step": 8592
+    },
+    {
+      "epoch": 2.3803324099722993,
+      "grad_norm": 0.7370125651359558,
+      "learning_rate": 3.3177846571600773e-06,
+      "loss": 0.6201,
+      "step": 8593
+    },
+    {
+      "epoch": 2.3806094182825484,
+      "grad_norm": 0.7008771896362305,
+      "learning_rate": 3.317440404773706e-06,
+      "loss": 0.5812,
+      "step": 8594
+    },
+    {
+      "epoch": 2.380886426592798,
+      "grad_norm": 0.770255446434021,
+      "learning_rate": 3.3170961350309866e-06,
+      "loss": 0.5996,
+      "step": 8595
+    },
+    {
+      "epoch": 2.3811634349030473,
+      "grad_norm": 0.6982191801071167,
+      "learning_rate": 3.31675184793923e-06,
+      "loss": 0.515,
+      "step": 8596
+    },
+    {
+      "epoch": 2.3814404432132963,
+      "grad_norm": 0.7257846593856812,
+      "learning_rate": 3.3164075435057465e-06,
+      "loss": 0.628,
+      "step": 8597
+    },
+    {
+      "epoch": 2.3817174515235457,
+      "grad_norm": 0.7340461611747742,
+      "learning_rate": 3.316063221737845e-06,
+      "loss": 0.5919,
+      "step": 8598
+    },
+    {
+      "epoch": 2.381994459833795,
+      "grad_norm": 0.7259990572929382,
+      "learning_rate": 3.315718882642838e-06,
+      "loss": 0.6031,
+      "step": 8599
+    },
+    {
+      "epoch": 2.3822714681440442,
+      "grad_norm": 0.713935911655426,
+      "learning_rate": 3.3153745262280363e-06,
+      "loss": 0.6512,
+      "step": 8600
+    },
+    {
+      "epoch": 2.3825484764542937,
+      "grad_norm": 0.7421038746833801,
+      "learning_rate": 3.31503015250075e-06,
+      "loss": 0.6127,
+      "step": 8601
+    },
+    {
+      "epoch": 2.3828254847645427,
+      "grad_norm": 0.7182706594467163,
+      "learning_rate": 3.3146857614682933e-06,
+      "loss": 0.6394,
+      "step": 8602
+    },
+    {
+      "epoch": 2.383102493074792,
+      "grad_norm": 0.7398244738578796,
+      "learning_rate": 3.3143413531379763e-06,
+      "loss": 0.5603,
+      "step": 8603
+    },
+    {
+      "epoch": 2.3833795013850416,
+      "grad_norm": 0.717022716999054,
+      "learning_rate": 3.313996927517114e-06,
+      "loss": 0.6001,
+      "step": 8604
+    },
+    {
+      "epoch": 2.3836565096952906,
+      "grad_norm": 0.7280988097190857,
+      "learning_rate": 3.313652484613017e-06,
+      "loss": 0.625,
+      "step": 8605
+    },
+    {
+      "epoch": 2.38393351800554,
+      "grad_norm": 0.7831609845161438,
+      "learning_rate": 3.3133080244330007e-06,
+      "loss": 0.6018,
+      "step": 8606
+    },
+    {
+      "epoch": 2.3842105263157896,
+      "grad_norm": 0.7226575016975403,
+      "learning_rate": 3.312963546984378e-06,
+      "loss": 0.5894,
+      "step": 8607
+    },
+    {
+      "epoch": 2.3844875346260386,
+      "grad_norm": 0.7334833145141602,
+      "learning_rate": 3.312619052274463e-06,
+      "loss": 0.6033,
+      "step": 8608
+    },
+    {
+      "epoch": 2.384764542936288,
+      "grad_norm": 0.6648973226547241,
+      "learning_rate": 3.3122745403105706e-06,
+      "loss": 0.5626,
+      "step": 8609
+    },
+    {
+      "epoch": 2.3850415512465375,
+      "grad_norm": 0.7012839317321777,
+      "learning_rate": 3.3119300111000152e-06,
+      "loss": 0.5786,
+      "step": 8610
+    },
+    {
+      "epoch": 2.3853185595567865,
+      "grad_norm": 0.7229165434837341,
+      "learning_rate": 3.311585464650112e-06,
+      "loss": 0.5846,
+      "step": 8611
+    },
+    {
+      "epoch": 2.385595567867036,
+      "grad_norm": 0.6839649677276611,
+      "learning_rate": 3.3112409009681767e-06,
+      "loss": 0.6118,
+      "step": 8612
+    },
+    {
+      "epoch": 2.3858725761772854,
+      "grad_norm": 0.751091718673706,
+      "learning_rate": 3.3108963200615265e-06,
+      "loss": 0.5951,
+      "step": 8613
+    },
+    {
+      "epoch": 2.3861495844875344,
+      "grad_norm": 0.7311224341392517,
+      "learning_rate": 3.310551721937475e-06,
+      "loss": 0.6278,
+      "step": 8614
+    },
+    {
+      "epoch": 2.386426592797784,
+      "grad_norm": 0.7401997447013855,
+      "learning_rate": 3.3102071066033414e-06,
+      "loss": 0.6383,
+      "step": 8615
+    },
+    {
+      "epoch": 2.3867036011080334,
+      "grad_norm": 0.7039486169815063,
+      "learning_rate": 3.309862474066442e-06,
+      "loss": 0.6233,
+      "step": 8616
+    },
+    {
+      "epoch": 2.3869806094182824,
+      "grad_norm": 0.7550800442695618,
+      "learning_rate": 3.309517824334094e-06,
+      "loss": 0.6055,
+      "step": 8617
+    },
+    {
+      "epoch": 2.387257617728532,
+      "grad_norm": 0.7891804575920105,
+      "learning_rate": 3.309173157413615e-06,
+      "loss": 0.6289,
+      "step": 8618
+    },
+    {
+      "epoch": 2.3875346260387813,
+      "grad_norm": 0.7476217150688171,
+      "learning_rate": 3.3088284733123233e-06,
+      "loss": 0.6219,
+      "step": 8619
+    },
+    {
+      "epoch": 2.3878116343490303,
+      "grad_norm": 0.7575064897537231,
+      "learning_rate": 3.3084837720375386e-06,
+      "loss": 0.5688,
+      "step": 8620
+    },
+    {
+      "epoch": 2.38808864265928,
+      "grad_norm": 0.7405401468276978,
+      "learning_rate": 3.308139053596578e-06,
+      "loss": 0.6082,
+      "step": 8621
+    },
+    {
+      "epoch": 2.3883656509695292,
+      "grad_norm": 0.7421559691429138,
+      "learning_rate": 3.307794317996762e-06,
+      "loss": 0.5862,
+      "step": 8622
+    },
+    {
+      "epoch": 2.3886426592797783,
+      "grad_norm": 0.7410457134246826,
+      "learning_rate": 3.3074495652454087e-06,
+      "loss": 0.6485,
+      "step": 8623
+    },
+    {
+      "epoch": 2.3889196675900277,
+      "grad_norm": 0.7029555439949036,
+      "learning_rate": 3.3071047953498396e-06,
+      "loss": 0.5818,
+      "step": 8624
+    },
+    {
+      "epoch": 2.389196675900277,
+      "grad_norm": 0.7657647132873535,
+      "learning_rate": 3.3067600083173752e-06,
+      "loss": 0.6682,
+      "step": 8625
+    },
+    {
+      "epoch": 2.389473684210526,
+      "grad_norm": 0.7279041409492493,
+      "learning_rate": 3.3064152041553356e-06,
+      "loss": 0.5996,
+      "step": 8626
+    },
+    {
+      "epoch": 2.3897506925207757,
+      "grad_norm": 0.7268471717834473,
+      "learning_rate": 3.3060703828710407e-06,
+      "loss": 0.5638,
+      "step": 8627
+    },
+    {
+      "epoch": 2.390027700831025,
+      "grad_norm": 0.7710117697715759,
+      "learning_rate": 3.3057255444718137e-06,
+      "loss": 0.5929,
+      "step": 8628
+    },
+    {
+      "epoch": 2.390304709141274,
+      "grad_norm": 0.7515334486961365,
+      "learning_rate": 3.3053806889649763e-06,
+      "loss": 0.6153,
+      "step": 8629
+    },
+    {
+      "epoch": 2.3905817174515236,
+      "grad_norm": 0.7906890511512756,
+      "learning_rate": 3.30503581635785e-06,
+      "loss": 0.603,
+      "step": 8630
+    },
+    {
+      "epoch": 2.390858725761773,
+      "grad_norm": 0.743398904800415,
+      "learning_rate": 3.304690926657757e-06,
+      "loss": 0.602,
+      "step": 8631
+    },
+    {
+      "epoch": 2.391135734072022,
+      "grad_norm": 0.7293583154678345,
+      "learning_rate": 3.3043460198720213e-06,
+      "loss": 0.5593,
+      "step": 8632
+    },
+    {
+      "epoch": 2.3914127423822715,
+      "grad_norm": 0.7299348711967468,
+      "learning_rate": 3.304001096007965e-06,
+      "loss": 0.6329,
+      "step": 8633
+    },
+    {
+      "epoch": 2.391689750692521,
+      "grad_norm": 0.7271016240119934,
+      "learning_rate": 3.3036561550729123e-06,
+      "loss": 0.5929,
+      "step": 8634
+    },
+    {
+      "epoch": 2.39196675900277,
+      "grad_norm": 0.7364197373390198,
+      "learning_rate": 3.303311197074188e-06,
+      "loss": 0.6214,
+      "step": 8635
+    },
+    {
+      "epoch": 2.3922437673130195,
+      "grad_norm": 0.7587757110595703,
+      "learning_rate": 3.3029662220191146e-06,
+      "loss": 0.6121,
+      "step": 8636
+    },
+    {
+      "epoch": 2.392520775623269,
+      "grad_norm": 0.7601677775382996,
+      "learning_rate": 3.3026212299150173e-06,
+      "loss": 0.6149,
+      "step": 8637
+    },
+    {
+      "epoch": 2.392797783933518,
+      "grad_norm": 0.7155644297599792,
+      "learning_rate": 3.3022762207692222e-06,
+      "loss": 0.576,
+      "step": 8638
+    },
+    {
+      "epoch": 2.3930747922437674,
+      "grad_norm": 0.7605409622192383,
+      "learning_rate": 3.301931194589054e-06,
+      "loss": 0.6194,
+      "step": 8639
+    },
+    {
+      "epoch": 2.393351800554017,
+      "grad_norm": 0.7717528939247131,
+      "learning_rate": 3.3015861513818388e-06,
+      "loss": 0.5927,
+      "step": 8640
+    },
+    {
+      "epoch": 2.393628808864266,
+      "grad_norm": 0.7234730124473572,
+      "learning_rate": 3.3012410911549023e-06,
+      "loss": 0.6205,
+      "step": 8641
+    },
+    {
+      "epoch": 2.3939058171745153,
+      "grad_norm": 0.7156244516372681,
+      "learning_rate": 3.300896013915572e-06,
+      "loss": 0.6189,
+      "step": 8642
+    },
+    {
+      "epoch": 2.3941828254847644,
+      "grad_norm": 0.7031002044677734,
+      "learning_rate": 3.300550919671174e-06,
+      "loss": 0.6049,
+      "step": 8643
+    },
+    {
+      "epoch": 2.394459833795014,
+      "grad_norm": 0.7399904131889343,
+      "learning_rate": 3.3002058084290345e-06,
+      "loss": 0.6305,
+      "step": 8644
+    },
+    {
+      "epoch": 2.3947368421052633,
+      "grad_norm": 0.7451629638671875,
+      "learning_rate": 3.2998606801964837e-06,
+      "loss": 0.6022,
+      "step": 8645
+    },
+    {
+      "epoch": 2.3950138504155123,
+      "grad_norm": 0.7191906571388245,
+      "learning_rate": 3.2995155349808474e-06,
+      "loss": 0.636,
+      "step": 8646
+    },
+    {
+      "epoch": 2.3952908587257618,
+      "grad_norm": 0.6923609972000122,
+      "learning_rate": 3.299170372789454e-06,
+      "loss": 0.5295,
+      "step": 8647
+    },
+    {
+      "epoch": 2.395567867036011,
+      "grad_norm": 0.7502071857452393,
+      "learning_rate": 3.2988251936296334e-06,
+      "loss": 0.6359,
+      "step": 8648
+    },
+    {
+      "epoch": 2.3958448753462602,
+      "grad_norm": 0.7677696943283081,
+      "learning_rate": 3.298479997508713e-06,
+      "loss": 0.6039,
+      "step": 8649
+    },
+    {
+      "epoch": 2.3961218836565097,
+      "grad_norm": 0.7665891051292419,
+      "learning_rate": 3.2981347844340237e-06,
+      "loss": 0.5768,
+      "step": 8650
+    },
+    {
+      "epoch": 2.396398891966759,
+      "grad_norm": 0.7692534327507019,
+      "learning_rate": 3.2977895544128955e-06,
+      "loss": 0.6114,
+      "step": 8651
+    },
+    {
+      "epoch": 2.396675900277008,
+      "grad_norm": 0.7409048676490784,
+      "learning_rate": 3.2974443074526572e-06,
+      "loss": 0.5829,
+      "step": 8652
+    },
+    {
+      "epoch": 2.3969529085872576,
+      "grad_norm": 0.7364702224731445,
+      "learning_rate": 3.2970990435606394e-06,
+      "loss": 0.6205,
+      "step": 8653
+    },
+    {
+      "epoch": 2.397229916897507,
+      "grad_norm": 0.7530790567398071,
+      "learning_rate": 3.296753762744174e-06,
+      "loss": 0.5833,
+      "step": 8654
+    },
+    {
+      "epoch": 2.397506925207756,
+      "grad_norm": 0.7629604339599609,
+      "learning_rate": 3.296408465010591e-06,
+      "loss": 0.5572,
+      "step": 8655
+    },
+    {
+      "epoch": 2.3977839335180056,
+      "grad_norm": 0.7859547734260559,
+      "learning_rate": 3.2960631503672223e-06,
+      "loss": 0.647,
+      "step": 8656
+    },
+    {
+      "epoch": 2.398060941828255,
+      "grad_norm": 0.7414343953132629,
+      "learning_rate": 3.2957178188214006e-06,
+      "loss": 0.5865,
+      "step": 8657
+    },
+    {
+      "epoch": 2.398337950138504,
+      "grad_norm": 0.6949269771575928,
+      "learning_rate": 3.2953724703804572e-06,
+      "loss": 0.5801,
+      "step": 8658
+    },
+    {
+      "epoch": 2.3986149584487535,
+      "grad_norm": 0.7935546636581421,
+      "learning_rate": 3.295027105051725e-06,
+      "loss": 0.6344,
+      "step": 8659
+    },
+    {
+      "epoch": 2.398891966759003,
+      "grad_norm": 0.767785906791687,
+      "learning_rate": 3.2946817228425377e-06,
+      "loss": 0.6063,
+      "step": 8660
+    },
+    {
+      "epoch": 2.399168975069252,
+      "grad_norm": 0.7153326272964478,
+      "learning_rate": 3.294336323760228e-06,
+      "loss": 0.5846,
+      "step": 8661
+    },
+    {
+      "epoch": 2.3994459833795014,
+      "grad_norm": 0.7636194229125977,
+      "learning_rate": 3.293990907812129e-06,
+      "loss": 0.6472,
+      "step": 8662
+    },
+    {
+      "epoch": 2.3997229916897505,
+      "grad_norm": 0.7240091562271118,
+      "learning_rate": 3.293645475005575e-06,
+      "loss": 0.5596,
+      "step": 8663
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7160698771476746,
+      "learning_rate": 3.293300025347902e-06,
+      "loss": 0.5677,
+      "step": 8664
+    },
+    {
+      "epoch": 2.4002770083102494,
+      "grad_norm": 0.7526683807373047,
+      "learning_rate": 3.2929545588464428e-06,
+      "loss": 0.5953,
+      "step": 8665
+    },
+    {
+      "epoch": 2.4005540166204984,
+      "grad_norm": 0.7347311973571777,
+      "learning_rate": 3.292609075508533e-06,
+      "loss": 0.6312,
+      "step": 8666
+    },
+    {
+      "epoch": 2.400831024930748,
+      "grad_norm": 0.6851674914360046,
+      "learning_rate": 3.29226357534151e-06,
+      "loss": 0.5971,
+      "step": 8667
+    },
+    {
+      "epoch": 2.4011080332409973,
+      "grad_norm": 0.7400670051574707,
+      "learning_rate": 3.2919180583527065e-06,
+      "loss": 0.6272,
+      "step": 8668
+    },
+    {
+      "epoch": 2.4013850415512463,
+      "grad_norm": 0.7210569977760315,
+      "learning_rate": 3.291572524549461e-06,
+      "loss": 0.6248,
+      "step": 8669
+    },
+    {
+      "epoch": 2.401662049861496,
+      "grad_norm": 0.7295917272567749,
+      "learning_rate": 3.291226973939109e-06,
+      "loss": 0.575,
+      "step": 8670
+    },
+    {
+      "epoch": 2.4019390581717452,
+      "grad_norm": 0.7628531455993652,
+      "learning_rate": 3.2908814065289884e-06,
+      "loss": 0.6255,
+      "step": 8671
+    },
+    {
+      "epoch": 2.4022160664819943,
+      "grad_norm": 0.7158731818199158,
+      "learning_rate": 3.290535822326435e-06,
+      "loss": 0.6073,
+      "step": 8672
+    },
+    {
+      "epoch": 2.4024930747922437,
+      "grad_norm": 0.7516884207725525,
+      "learning_rate": 3.2901902213387876e-06,
+      "loss": 0.6121,
+      "step": 8673
+    },
+    {
+      "epoch": 2.402770083102493,
+      "grad_norm": 0.7457286715507507,
+      "learning_rate": 3.2898446035733845e-06,
+      "loss": 0.6163,
+      "step": 8674
+    },
+    {
+      "epoch": 2.403047091412742,
+      "grad_norm": 0.7692743539810181,
+      "learning_rate": 3.289498969037563e-06,
+      "loss": 0.6117,
+      "step": 8675
+    },
+    {
+      "epoch": 2.4033240997229917,
+      "grad_norm": 0.7491902709007263,
+      "learning_rate": 3.289153317738662e-06,
+      "loss": 0.6537,
+      "step": 8676
+    },
+    {
+      "epoch": 2.403601108033241,
+      "grad_norm": 0.730591893196106,
+      "learning_rate": 3.288807649684021e-06,
+      "loss": 0.6004,
+      "step": 8677
+    },
+    {
+      "epoch": 2.40387811634349,
+      "grad_norm": 0.7152724862098694,
+      "learning_rate": 3.288461964880979e-06,
+      "loss": 0.5616,
+      "step": 8678
+    },
+    {
+      "epoch": 2.4041551246537396,
+      "grad_norm": 0.7206489443778992,
+      "learning_rate": 3.2881162633368757e-06,
+      "loss": 0.5998,
+      "step": 8679
+    },
+    {
+      "epoch": 2.404432132963989,
+      "grad_norm": 0.7845282554626465,
+      "learning_rate": 3.2877705450590525e-06,
+      "loss": 0.5912,
+      "step": 8680
+    },
+    {
+      "epoch": 2.404709141274238,
+      "grad_norm": 0.7430789470672607,
+      "learning_rate": 3.287424810054848e-06,
+      "loss": 0.619,
+      "step": 8681
+    },
+    {
+      "epoch": 2.4049861495844875,
+      "grad_norm": 0.7434074878692627,
+      "learning_rate": 3.2870790583316035e-06,
+      "loss": 0.6069,
+      "step": 8682
+    },
+    {
+      "epoch": 2.405263157894737,
+      "grad_norm": 0.7171652913093567,
+      "learning_rate": 3.286733289896662e-06,
+      "loss": 0.6173,
+      "step": 8683
+    },
+    {
+      "epoch": 2.405540166204986,
+      "grad_norm": 0.7580751180648804,
+      "learning_rate": 3.286387504757363e-06,
+      "loss": 0.59,
+      "step": 8684
+    },
+    {
+      "epoch": 2.4058171745152355,
+      "grad_norm": 0.7650940418243408,
+      "learning_rate": 3.286041702921048e-06,
+      "loss": 0.589,
+      "step": 8685
+    },
+    {
+      "epoch": 2.406094182825485,
+      "grad_norm": 0.7545360922813416,
+      "learning_rate": 3.2856958843950626e-06,
+      "loss": 0.5474,
+      "step": 8686
+    },
+    {
+      "epoch": 2.406371191135734,
+      "grad_norm": 0.7186492681503296,
+      "learning_rate": 3.2853500491867452e-06,
+      "loss": 0.6071,
+      "step": 8687
+    },
+    {
+      "epoch": 2.4066481994459834,
+      "grad_norm": 0.742397665977478,
+      "learning_rate": 3.2850041973034417e-06,
+      "loss": 0.6075,
+      "step": 8688
+    },
+    {
+      "epoch": 2.406925207756233,
+      "grad_norm": 0.6936323046684265,
+      "learning_rate": 3.284658328752494e-06,
+      "loss": 0.5361,
+      "step": 8689
+    },
+    {
+      "epoch": 2.407202216066482,
+      "grad_norm": 0.7993597388267517,
+      "learning_rate": 3.2843124435412464e-06,
+      "loss": 0.598,
+      "step": 8690
+    },
+    {
+      "epoch": 2.4074792243767313,
+      "grad_norm": 0.738192617893219,
+      "learning_rate": 3.2839665416770422e-06,
+      "loss": 0.6012,
+      "step": 8691
+    },
+    {
+      "epoch": 2.407756232686981,
+      "grad_norm": 0.7682150602340698,
+      "learning_rate": 3.283620623167227e-06,
+      "loss": 0.5988,
+      "step": 8692
+    },
+    {
+      "epoch": 2.40803324099723,
+      "grad_norm": 0.6933428049087524,
+      "learning_rate": 3.2832746880191447e-06,
+      "loss": 0.6201,
+      "step": 8693
+    },
+    {
+      "epoch": 2.4083102493074793,
+      "grad_norm": 0.7243341207504272,
+      "learning_rate": 3.2829287362401398e-06,
+      "loss": 0.5975,
+      "step": 8694
+    },
+    {
+      "epoch": 2.4085872576177287,
+      "grad_norm": 0.7616162300109863,
+      "learning_rate": 3.282582767837559e-06,
+      "loss": 0.5779,
+      "step": 8695
+    },
+    {
+      "epoch": 2.4088642659279778,
+      "grad_norm": 0.7368990182876587,
+      "learning_rate": 3.2822367828187475e-06,
+      "loss": 0.5882,
+      "step": 8696
+    },
+    {
+      "epoch": 2.409141274238227,
+      "grad_norm": 0.7216224670410156,
+      "learning_rate": 3.281890781191051e-06,
+      "loss": 0.5886,
+      "step": 8697
+    },
+    {
+      "epoch": 2.4094182825484767,
+      "grad_norm": 0.739101767539978,
+      "learning_rate": 3.2815447629618167e-06,
+      "loss": 0.6104,
+      "step": 8698
+    },
+    {
+      "epoch": 2.4096952908587257,
+      "grad_norm": 0.7256513833999634,
+      "learning_rate": 3.281198728138392e-06,
+      "loss": 0.5413,
+      "step": 8699
+    },
+    {
+      "epoch": 2.409972299168975,
+      "grad_norm": 0.7555835247039795,
+      "learning_rate": 3.2808526767281223e-06,
+      "loss": 0.6377,
+      "step": 8700
+    },
+    {
+      "epoch": 2.4102493074792246,
+      "grad_norm": 0.7214328050613403,
+      "learning_rate": 3.280506608738357e-06,
+      "loss": 0.5582,
+      "step": 8701
+    },
+    {
+      "epoch": 2.4105263157894736,
+      "grad_norm": 0.7244924306869507,
+      "learning_rate": 3.2801605241764432e-06,
+      "loss": 0.6037,
+      "step": 8702
+    },
+    {
+      "epoch": 2.410803324099723,
+      "grad_norm": 0.7230759859085083,
+      "learning_rate": 3.2798144230497287e-06,
+      "loss": 0.6108,
+      "step": 8703
+    },
+    {
+      "epoch": 2.4110803324099725,
+      "grad_norm": 0.7757418751716614,
+      "learning_rate": 3.2794683053655623e-06,
+      "loss": 0.6112,
+      "step": 8704
+    },
+    {
+      "epoch": 2.4113573407202216,
+      "grad_norm": 0.756781816482544,
+      "learning_rate": 3.2791221711312945e-06,
+      "loss": 0.5969,
+      "step": 8705
+    },
+    {
+      "epoch": 2.411634349030471,
+      "grad_norm": 0.7342284917831421,
+      "learning_rate": 3.2787760203542716e-06,
+      "loss": 0.6036,
+      "step": 8706
+    },
+    {
+      "epoch": 2.41191135734072,
+      "grad_norm": 0.7624217867851257,
+      "learning_rate": 3.2784298530418464e-06,
+      "loss": 0.6183,
+      "step": 8707
+    },
+    {
+      "epoch": 2.4121883656509695,
+      "grad_norm": 0.7700241208076477,
+      "learning_rate": 3.278083669201367e-06,
+      "loss": 0.5969,
+      "step": 8708
+    },
+    {
+      "epoch": 2.412465373961219,
+      "grad_norm": 0.7754845023155212,
+      "learning_rate": 3.277737468840184e-06,
+      "loss": 0.63,
+      "step": 8709
+    },
+    {
+      "epoch": 2.412742382271468,
+      "grad_norm": 0.7279965877532959,
+      "learning_rate": 3.277391251965649e-06,
+      "loss": 0.601,
+      "step": 8710
+    },
+    {
+      "epoch": 2.4130193905817174,
+      "grad_norm": 0.7578972578048706,
+      "learning_rate": 3.2770450185851123e-06,
+      "loss": 0.6033,
+      "step": 8711
+    },
+    {
+      "epoch": 2.413296398891967,
+      "grad_norm": 0.7431296110153198,
+      "learning_rate": 3.2766987687059254e-06,
+      "loss": 0.6556,
+      "step": 8712
+    },
+    {
+      "epoch": 2.413573407202216,
+      "grad_norm": 0.7281125783920288,
+      "learning_rate": 3.27635250233544e-06,
+      "loss": 0.6182,
+      "step": 8713
+    },
+    {
+      "epoch": 2.4138504155124654,
+      "grad_norm": 0.7503303289413452,
+      "learning_rate": 3.2760062194810083e-06,
+      "loss": 0.6364,
+      "step": 8714
+    },
+    {
+      "epoch": 2.414127423822715,
+      "grad_norm": 0.7167266607284546,
+      "learning_rate": 3.2756599201499834e-06,
+      "loss": 0.635,
+      "step": 8715
+    },
+    {
+      "epoch": 2.414404432132964,
+      "grad_norm": 0.7465100288391113,
+      "learning_rate": 3.2753136043497175e-06,
+      "loss": 0.5931,
+      "step": 8716
+    },
+    {
+      "epoch": 2.4146814404432133,
+      "grad_norm": 0.7193043828010559,
+      "learning_rate": 3.2749672720875636e-06,
+      "loss": 0.6033,
+      "step": 8717
+    },
+    {
+      "epoch": 2.4149584487534628,
+      "grad_norm": 0.7016171216964722,
+      "learning_rate": 3.274620923370876e-06,
+      "loss": 0.5477,
+      "step": 8718
+    },
+    {
+      "epoch": 2.415235457063712,
+      "grad_norm": 0.7321090698242188,
+      "learning_rate": 3.2742745582070068e-06,
+      "loss": 0.588,
+      "step": 8719
+    },
+    {
+      "epoch": 2.4155124653739612,
+      "grad_norm": 0.7451436519622803,
+      "learning_rate": 3.2739281766033126e-06,
+      "loss": 0.6086,
+      "step": 8720
+    },
+    {
+      "epoch": 2.4157894736842107,
+      "grad_norm": 0.7388331294059753,
+      "learning_rate": 3.2735817785671466e-06,
+      "loss": 0.5931,
+      "step": 8721
+    },
+    {
+      "epoch": 2.4160664819944597,
+      "grad_norm": 0.7376699447631836,
+      "learning_rate": 3.2732353641058633e-06,
+      "loss": 0.5939,
+      "step": 8722
+    },
+    {
+      "epoch": 2.416343490304709,
+      "grad_norm": 0.771159827709198,
+      "learning_rate": 3.272888933226819e-06,
+      "loss": 0.5893,
+      "step": 8723
+    },
+    {
+      "epoch": 2.416620498614958,
+      "grad_norm": 0.7518683671951294,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.6219,
+      "step": 8724
+    },
+    {
+      "epoch": 2.4168975069252077,
+      "grad_norm": 0.7549920082092285,
+      "learning_rate": 3.272196022244869e-06,
+      "loss": 0.5993,
+      "step": 8725
+    },
+    {
+      "epoch": 2.417174515235457,
+      "grad_norm": 0.7153391242027283,
+      "learning_rate": 3.2718495421566754e-06,
+      "loss": 0.612,
+      "step": 8726
+    },
+    {
+      "epoch": 2.417451523545706,
+      "grad_norm": 0.747154951095581,
+      "learning_rate": 3.271503045680145e-06,
+      "loss": 0.5827,
+      "step": 8727
+    },
+    {
+      "epoch": 2.4177285318559556,
+      "grad_norm": 0.7624956369400024,
+      "learning_rate": 3.2711565328226354e-06,
+      "loss": 0.6285,
+      "step": 8728
+    },
+    {
+      "epoch": 2.418005540166205,
+      "grad_norm": 0.731533408164978,
+      "learning_rate": 3.2708100035915027e-06,
+      "loss": 0.5908,
+      "step": 8729
+    },
+    {
+      "epoch": 2.418282548476454,
+      "grad_norm": 0.7518516182899475,
+      "learning_rate": 3.270463457994105e-06,
+      "loss": 0.618,
+      "step": 8730
+    },
+    {
+      "epoch": 2.4185595567867035,
+      "grad_norm": 0.7448464035987854,
+      "learning_rate": 3.270116896037801e-06,
+      "loss": 0.5555,
+      "step": 8731
+    },
+    {
+      "epoch": 2.418836565096953,
+      "grad_norm": 0.7565663456916809,
+      "learning_rate": 3.269770317729948e-06,
+      "loss": 0.6204,
+      "step": 8732
+    },
+    {
+      "epoch": 2.419113573407202,
+      "grad_norm": 0.730437695980072,
+      "learning_rate": 3.2694237230779056e-06,
+      "loss": 0.5792,
+      "step": 8733
+    },
+    {
+      "epoch": 2.4193905817174515,
+      "grad_norm": 0.7168110609054565,
+      "learning_rate": 3.269077112089033e-06,
+      "loss": 0.5732,
+      "step": 8734
+    },
+    {
+      "epoch": 2.419667590027701,
+      "grad_norm": 0.725572407245636,
+      "learning_rate": 3.268730484770689e-06,
+      "loss": 0.5353,
+      "step": 8735
+    },
+    {
+      "epoch": 2.41994459833795,
+      "grad_norm": 0.7277836799621582,
+      "learning_rate": 3.268383841130234e-06,
+      "loss": 0.6352,
+      "step": 8736
+    },
+    {
+      "epoch": 2.4202216066481994,
+      "grad_norm": 0.7155618667602539,
+      "learning_rate": 3.268037181175028e-06,
+      "loss": 0.5453,
+      "step": 8737
+    },
+    {
+      "epoch": 2.420498614958449,
+      "grad_norm": 0.7514188289642334,
+      "learning_rate": 3.2676905049124306e-06,
+      "loss": 0.5491,
+      "step": 8738
+    },
+    {
+      "epoch": 2.420775623268698,
+      "grad_norm": 0.776049792766571,
+      "learning_rate": 3.267343812349803e-06,
+      "loss": 0.5975,
+      "step": 8739
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.7240673303604126,
+      "learning_rate": 3.266997103494508e-06,
+      "loss": 0.6213,
+      "step": 8740
+    },
+    {
+      "epoch": 2.421329639889197,
+      "grad_norm": 0.721198320388794,
+      "learning_rate": 3.266650378353905e-06,
+      "loss": 0.5853,
+      "step": 8741
+    },
+    {
+      "epoch": 2.421606648199446,
+      "grad_norm": 0.7547255754470825,
+      "learning_rate": 3.266303636935357e-06,
+      "loss": 0.589,
+      "step": 8742
+    },
+    {
+      "epoch": 2.4218836565096953,
+      "grad_norm": 0.7153939604759216,
+      "learning_rate": 3.2659568792462254e-06,
+      "loss": 0.576,
+      "step": 8743
+    },
+    {
+      "epoch": 2.4221606648199447,
+      "grad_norm": 0.7191527485847473,
+      "learning_rate": 3.2656101052938737e-06,
+      "loss": 0.6107,
+      "step": 8744
+    },
+    {
+      "epoch": 2.4224376731301938,
+      "grad_norm": 0.6901411414146423,
+      "learning_rate": 3.2652633150856637e-06,
+      "loss": 0.5964,
+      "step": 8745
+    },
+    {
+      "epoch": 2.422714681440443,
+      "grad_norm": 0.7672993540763855,
+      "learning_rate": 3.2649165086289597e-06,
+      "loss": 0.6053,
+      "step": 8746
+    },
+    {
+      "epoch": 2.4229916897506927,
+      "grad_norm": 0.7194650769233704,
+      "learning_rate": 3.2645696859311248e-06,
+      "loss": 0.5777,
+      "step": 8747
+    },
+    {
+      "epoch": 2.4232686980609417,
+      "grad_norm": 0.7330414652824402,
+      "learning_rate": 3.264222846999523e-06,
+      "loss": 0.6037,
+      "step": 8748
+    },
+    {
+      "epoch": 2.423545706371191,
+      "grad_norm": 0.6991344094276428,
+      "learning_rate": 3.263875991841518e-06,
+      "loss": 0.5679,
+      "step": 8749
+    },
+    {
+      "epoch": 2.4238227146814406,
+      "grad_norm": 0.7010491490364075,
+      "learning_rate": 3.2635291204644755e-06,
+      "loss": 0.5512,
+      "step": 8750
+    },
+    {
+      "epoch": 2.4240997229916896,
+      "grad_norm": 0.7519541382789612,
+      "learning_rate": 3.26318223287576e-06,
+      "loss": 0.6287,
+      "step": 8751
+    },
+    {
+      "epoch": 2.424376731301939,
+      "grad_norm": 0.7270293235778809,
+      "learning_rate": 3.2628353290827365e-06,
+      "loss": 0.5173,
+      "step": 8752
+    },
+    {
+      "epoch": 2.4246537396121886,
+      "grad_norm": 0.7756695747375488,
+      "learning_rate": 3.262488409092771e-06,
+      "loss": 0.57,
+      "step": 8753
+    },
+    {
+      "epoch": 2.4249307479224376,
+      "grad_norm": 0.7207167744636536,
+      "learning_rate": 3.2621414729132296e-06,
+      "loss": 0.6317,
+      "step": 8754
+    },
+    {
+      "epoch": 2.425207756232687,
+      "grad_norm": 0.699981153011322,
+      "learning_rate": 3.261794520551478e-06,
+      "loss": 0.5664,
+      "step": 8755
+    },
+    {
+      "epoch": 2.4254847645429365,
+      "grad_norm": 0.7614565491676331,
+      "learning_rate": 3.2614475520148836e-06,
+      "loss": 0.592,
+      "step": 8756
+    },
+    {
+      "epoch": 2.4257617728531855,
+      "grad_norm": 0.7084278464317322,
+      "learning_rate": 3.2611005673108133e-06,
+      "loss": 0.6096,
+      "step": 8757
+    },
+    {
+      "epoch": 2.426038781163435,
+      "grad_norm": 0.6804847121238708,
+      "learning_rate": 3.2607535664466334e-06,
+      "loss": 0.5873,
+      "step": 8758
+    },
+    {
+      "epoch": 2.4263157894736844,
+      "grad_norm": 0.7468960881233215,
+      "learning_rate": 3.260406549429714e-06,
+      "loss": 0.6394,
+      "step": 8759
+    },
+    {
+      "epoch": 2.4265927977839334,
+      "grad_norm": 0.7546526789665222,
+      "learning_rate": 3.26005951626742e-06,
+      "loss": 0.6261,
+      "step": 8760
+    },
+    {
+      "epoch": 2.426869806094183,
+      "grad_norm": 0.751997172832489,
+      "learning_rate": 3.259712466967123e-06,
+      "loss": 0.6334,
+      "step": 8761
+    },
+    {
+      "epoch": 2.4271468144044324,
+      "grad_norm": 0.7271677255630493,
+      "learning_rate": 3.25936540153619e-06,
+      "loss": 0.5969,
+      "step": 8762
+    },
+    {
+      "epoch": 2.4274238227146814,
+      "grad_norm": 0.7059834599494934,
+      "learning_rate": 3.2590183199819896e-06,
+      "loss": 0.5882,
+      "step": 8763
+    },
+    {
+      "epoch": 2.427700831024931,
+      "grad_norm": 0.7132251858711243,
+      "learning_rate": 3.258671222311892e-06,
+      "loss": 0.5784,
+      "step": 8764
+    },
+    {
+      "epoch": 2.4279778393351803,
+      "grad_norm": 0.7054438591003418,
+      "learning_rate": 3.2583241085332674e-06,
+      "loss": 0.6,
+      "step": 8765
+    },
+    {
+      "epoch": 2.4282548476454293,
+      "grad_norm": 0.7704301476478577,
+      "learning_rate": 3.257976978653485e-06,
+      "loss": 0.5941,
+      "step": 8766
+    },
+    {
+      "epoch": 2.4285318559556788,
+      "grad_norm": 0.7076848745346069,
+      "learning_rate": 3.257629832679916e-06,
+      "loss": 0.6003,
+      "step": 8767
+    },
+    {
+      "epoch": 2.428808864265928,
+      "grad_norm": 0.6868513226509094,
+      "learning_rate": 3.2572826706199304e-06,
+      "loss": 0.5847,
+      "step": 8768
+    },
+    {
+      "epoch": 2.4290858725761773,
+      "grad_norm": 0.7200565338134766,
+      "learning_rate": 3.2569354924809005e-06,
+      "loss": 0.5691,
+      "step": 8769
+    },
+    {
+      "epoch": 2.4293628808864267,
+      "grad_norm": 0.722827672958374,
+      "learning_rate": 3.2565882982701967e-06,
+      "loss": 0.5868,
+      "step": 8770
+    },
+    {
+      "epoch": 2.4296398891966757,
+      "grad_norm": 0.7694891691207886,
+      "learning_rate": 3.256241087995191e-06,
+      "loss": 0.6318,
+      "step": 8771
+    },
+    {
+      "epoch": 2.429916897506925,
+      "grad_norm": 0.7697635293006897,
+      "learning_rate": 3.255893861663257e-06,
+      "loss": 0.5825,
+      "step": 8772
+    },
+    {
+      "epoch": 2.4301939058171746,
+      "grad_norm": 0.7312008142471313,
+      "learning_rate": 3.255546619281765e-06,
+      "loss": 0.5826,
+      "step": 8773
+    },
+    {
+      "epoch": 2.4304709141274237,
+      "grad_norm": 0.7236773371696472,
+      "learning_rate": 3.2551993608580884e-06,
+      "loss": 0.5953,
+      "step": 8774
+    },
+    {
+      "epoch": 2.430747922437673,
+      "grad_norm": 0.7374091148376465,
+      "learning_rate": 3.2548520863996013e-06,
+      "loss": 0.6046,
+      "step": 8775
+    },
+    {
+      "epoch": 2.4310249307479226,
+      "grad_norm": 0.746926486492157,
+      "learning_rate": 3.254504795913677e-06,
+      "loss": 0.6044,
+      "step": 8776
+    },
+    {
+      "epoch": 2.4313019390581716,
+      "grad_norm": 0.7748494148254395,
+      "learning_rate": 3.2541574894076884e-06,
+      "loss": 0.5798,
+      "step": 8777
+    },
+    {
+      "epoch": 2.431578947368421,
+      "grad_norm": 0.7435228228569031,
+      "learning_rate": 3.253810166889011e-06,
+      "loss": 0.5884,
+      "step": 8778
+    },
+    {
+      "epoch": 2.4318559556786705,
+      "grad_norm": 0.756584882736206,
+      "learning_rate": 3.2534628283650184e-06,
+      "loss": 0.5781,
+      "step": 8779
+    },
+    {
+      "epoch": 2.4321329639889195,
+      "grad_norm": 0.7341037392616272,
+      "learning_rate": 3.2531154738430857e-06,
+      "loss": 0.5717,
+      "step": 8780
+    },
+    {
+      "epoch": 2.432409972299169,
+      "grad_norm": 0.7264748811721802,
+      "learning_rate": 3.252768103330589e-06,
+      "loss": 0.6186,
+      "step": 8781
+    },
+    {
+      "epoch": 2.4326869806094185,
+      "grad_norm": 0.715746283531189,
+      "learning_rate": 3.252420716834902e-06,
+      "loss": 0.5697,
+      "step": 8782
+    },
+    {
+      "epoch": 2.4329639889196675,
+      "grad_norm": 0.8117491006851196,
+      "learning_rate": 3.2520733143634027e-06,
+      "loss": 0.6284,
+      "step": 8783
+    },
+    {
+      "epoch": 2.433240997229917,
+      "grad_norm": 0.7050234079360962,
+      "learning_rate": 3.251725895923466e-06,
+      "loss": 0.625,
+      "step": 8784
+    },
+    {
+      "epoch": 2.4335180055401664,
+      "grad_norm": 0.7163100838661194,
+      "learning_rate": 3.2513784615224686e-06,
+      "loss": 0.6722,
+      "step": 8785
+    },
+    {
+      "epoch": 2.4337950138504154,
+      "grad_norm": 0.7334553003311157,
+      "learning_rate": 3.2510310111677877e-06,
+      "loss": 0.6164,
+      "step": 8786
+    },
+    {
+      "epoch": 2.434072022160665,
+      "grad_norm": 0.7212862968444824,
+      "learning_rate": 3.2506835448668007e-06,
+      "loss": 0.6006,
+      "step": 8787
+    },
+    {
+      "epoch": 2.434349030470914,
+      "grad_norm": 0.7204136252403259,
+      "learning_rate": 3.2503360626268855e-06,
+      "loss": 0.6021,
+      "step": 8788
+    },
+    {
+      "epoch": 2.4346260387811633,
+      "grad_norm": 0.7287476062774658,
+      "learning_rate": 3.2499885644554185e-06,
+      "loss": 0.6392,
+      "step": 8789
+    },
+    {
+      "epoch": 2.434903047091413,
+      "grad_norm": 0.7399448156356812,
+      "learning_rate": 3.249641050359779e-06,
+      "loss": 0.6185,
+      "step": 8790
+    },
+    {
+      "epoch": 2.435180055401662,
+      "grad_norm": 0.7647336721420288,
+      "learning_rate": 3.2492935203473468e-06,
+      "loss": 0.635,
+      "step": 8791
+    },
+    {
+      "epoch": 2.4354570637119113,
+      "grad_norm": 0.7237324118614197,
+      "learning_rate": 3.2489459744254986e-06,
+      "loss": 0.612,
+      "step": 8792
+    },
+    {
+      "epoch": 2.4357340720221607,
+      "grad_norm": 0.7611182332038879,
+      "learning_rate": 3.248598412601615e-06,
+      "loss": 0.6438,
+      "step": 8793
+    },
+    {
+      "epoch": 2.4360110803324098,
+      "grad_norm": 0.7452676296234131,
+      "learning_rate": 3.248250834883076e-06,
+      "loss": 0.5965,
+      "step": 8794
+    },
+    {
+      "epoch": 2.436288088642659,
+      "grad_norm": 0.7790424227714539,
+      "learning_rate": 3.24790324127726e-06,
+      "loss": 0.5996,
+      "step": 8795
+    },
+    {
+      "epoch": 2.4365650969529087,
+      "grad_norm": 0.7863867878913879,
+      "learning_rate": 3.247555631791549e-06,
+      "loss": 0.5955,
+      "step": 8796
+    },
+    {
+      "epoch": 2.4368421052631577,
+      "grad_norm": 0.7386239171028137,
+      "learning_rate": 3.2472080064333226e-06,
+      "loss": 0.586,
+      "step": 8797
+    },
+    {
+      "epoch": 2.437119113573407,
+      "grad_norm": 0.7536293864250183,
+      "learning_rate": 3.2468603652099624e-06,
+      "loss": 0.6069,
+      "step": 8798
+    },
+    {
+      "epoch": 2.4373961218836566,
+      "grad_norm": 0.7368284463882446,
+      "learning_rate": 3.246512708128849e-06,
+      "loss": 0.6208,
+      "step": 8799
+    },
+    {
+      "epoch": 2.4376731301939056,
+      "grad_norm": 0.7683560252189636,
+      "learning_rate": 3.2461650351973644e-06,
+      "loss": 0.5284,
+      "step": 8800
+    },
+    {
+      "epoch": 2.437950138504155,
+      "grad_norm": 0.7497748732566833,
+      "learning_rate": 3.2458173464228905e-06,
+      "loss": 0.6377,
+      "step": 8801
+    },
+    {
+      "epoch": 2.4382271468144046,
+      "grad_norm": 0.763282060623169,
+      "learning_rate": 3.2454696418128103e-06,
+      "loss": 0.602,
+      "step": 8802
+    },
+    {
+      "epoch": 2.4385041551246536,
+      "grad_norm": 0.6915131211280823,
+      "learning_rate": 3.245121921374505e-06,
+      "loss": 0.6292,
+      "step": 8803
+    },
+    {
+      "epoch": 2.438781163434903,
+      "grad_norm": 0.7452968955039978,
+      "learning_rate": 3.2447741851153596e-06,
+      "loss": 0.5869,
+      "step": 8804
+    },
+    {
+      "epoch": 2.4390581717451525,
+      "grad_norm": 0.7962399125099182,
+      "learning_rate": 3.2444264330427555e-06,
+      "loss": 0.627,
+      "step": 8805
+    },
+    {
+      "epoch": 2.4393351800554015,
+      "grad_norm": 0.7612031698226929,
+      "learning_rate": 3.244078665164077e-06,
+      "loss": 0.6268,
+      "step": 8806
+    },
+    {
+      "epoch": 2.439612188365651,
+      "grad_norm": 0.6910748481750488,
+      "learning_rate": 3.243730881486709e-06,
+      "loss": 0.5159,
+      "step": 8807
+    },
+    {
+      "epoch": 2.4398891966759004,
+      "grad_norm": 0.7467472553253174,
+      "learning_rate": 3.2433830820180346e-06,
+      "loss": 0.5723,
+      "step": 8808
+    },
+    {
+      "epoch": 2.4401662049861494,
+      "grad_norm": 0.7672812938690186,
+      "learning_rate": 3.243035266765439e-06,
+      "loss": 0.6108,
+      "step": 8809
+    },
+    {
+      "epoch": 2.440443213296399,
+      "grad_norm": 0.7103387117385864,
+      "learning_rate": 3.2426874357363077e-06,
+      "loss": 0.593,
+      "step": 8810
+    },
+    {
+      "epoch": 2.4407202216066484,
+      "grad_norm": 0.7184141278266907,
+      "learning_rate": 3.242339588938025e-06,
+      "loss": 0.6287,
+      "step": 8811
+    },
+    {
+      "epoch": 2.4409972299168974,
+      "grad_norm": 0.7140890955924988,
+      "learning_rate": 3.2419917263779765e-06,
+      "loss": 0.6144,
+      "step": 8812
+    },
+    {
+      "epoch": 2.441274238227147,
+      "grad_norm": 0.7000736594200134,
+      "learning_rate": 3.24164384806355e-06,
+      "loss": 0.5956,
+      "step": 8813
+    },
+    {
+      "epoch": 2.4415512465373963,
+      "grad_norm": 0.709296703338623,
+      "learning_rate": 3.24129595400213e-06,
+      "loss": 0.5779,
+      "step": 8814
+    },
+    {
+      "epoch": 2.4418282548476453,
+      "grad_norm": 0.7582641839981079,
+      "learning_rate": 3.2409480442011042e-06,
+      "loss": 0.6205,
+      "step": 8815
+    },
+    {
+      "epoch": 2.442105263157895,
+      "grad_norm": 0.7473920583724976,
+      "learning_rate": 3.2406001186678594e-06,
+      "loss": 0.6211,
+      "step": 8816
+    },
+    {
+      "epoch": 2.4423822714681442,
+      "grad_norm": 0.700025200843811,
+      "learning_rate": 3.2402521774097824e-06,
+      "loss": 0.5868,
+      "step": 8817
+    },
+    {
+      "epoch": 2.4426592797783933,
+      "grad_norm": 0.7863488793373108,
+      "learning_rate": 3.239904220434261e-06,
+      "loss": 0.7038,
+      "step": 8818
+    },
+    {
+      "epoch": 2.4429362880886427,
+      "grad_norm": 0.7783229351043701,
+      "learning_rate": 3.2395562477486836e-06,
+      "loss": 0.5894,
+      "step": 8819
+    },
+    {
+      "epoch": 2.443213296398892,
+      "grad_norm": 0.7442140579223633,
+      "learning_rate": 3.239208259360439e-06,
+      "loss": 0.5693,
+      "step": 8820
+    },
+    {
+      "epoch": 2.443490304709141,
+      "grad_norm": 0.6942600011825562,
+      "learning_rate": 3.2388602552769145e-06,
+      "loss": 0.5629,
+      "step": 8821
+    },
+    {
+      "epoch": 2.4437673130193907,
+      "grad_norm": 0.7239096760749817,
+      "learning_rate": 3.2385122355055004e-06,
+      "loss": 0.5711,
+      "step": 8822
+    },
+    {
+      "epoch": 2.44404432132964,
+      "grad_norm": 0.7572760581970215,
+      "learning_rate": 3.2381642000535858e-06,
+      "loss": 0.6027,
+      "step": 8823
+    },
+    {
+      "epoch": 2.444321329639889,
+      "grad_norm": 0.6888085603713989,
+      "learning_rate": 3.2378161489285597e-06,
+      "loss": 0.5529,
+      "step": 8824
+    },
+    {
+      "epoch": 2.4445983379501386,
+      "grad_norm": 0.7307559251785278,
+      "learning_rate": 3.2374680821378122e-06,
+      "loss": 0.6143,
+      "step": 8825
+    },
+    {
+      "epoch": 2.444875346260388,
+      "grad_norm": 0.7402287125587463,
+      "learning_rate": 3.2371199996887357e-06,
+      "loss": 0.6132,
+      "step": 8826
+    },
+    {
+      "epoch": 2.445152354570637,
+      "grad_norm": 0.7784111499786377,
+      "learning_rate": 3.2367719015887177e-06,
+      "loss": 0.5761,
+      "step": 8827
+    },
+    {
+      "epoch": 2.4454293628808865,
+      "grad_norm": 0.7317233681678772,
+      "learning_rate": 3.2364237878451517e-06,
+      "loss": 0.5782,
+      "step": 8828
+    },
+    {
+      "epoch": 2.4457063711911355,
+      "grad_norm": 0.7439118027687073,
+      "learning_rate": 3.2360756584654275e-06,
+      "loss": 0.5967,
+      "step": 8829
+    },
+    {
+      "epoch": 2.445983379501385,
+      "grad_norm": 0.7780766487121582,
+      "learning_rate": 3.235727513456938e-06,
+      "loss": 0.5966,
+      "step": 8830
+    },
+    {
+      "epoch": 2.4462603878116345,
+      "grad_norm": 0.7411060929298401,
+      "learning_rate": 3.2353793528270727e-06,
+      "loss": 0.604,
+      "step": 8831
+    },
+    {
+      "epoch": 2.4465373961218835,
+      "grad_norm": 0.7590367197990417,
+      "learning_rate": 3.2350311765832275e-06,
+      "loss": 0.5924,
+      "step": 8832
+    },
+    {
+      "epoch": 2.446814404432133,
+      "grad_norm": 0.7714030742645264,
+      "learning_rate": 3.234682984732793e-06,
+      "loss": 0.6387,
+      "step": 8833
+    },
+    {
+      "epoch": 2.4470914127423824,
+      "grad_norm": 0.7148301005363464,
+      "learning_rate": 3.234334777283162e-06,
+      "loss": 0.5936,
+      "step": 8834
+    },
+    {
+      "epoch": 2.4473684210526314,
+      "grad_norm": 0.7669836282730103,
+      "learning_rate": 3.2339865542417292e-06,
+      "loss": 0.6355,
+      "step": 8835
+    },
+    {
+      "epoch": 2.447645429362881,
+      "grad_norm": 0.7615107297897339,
+      "learning_rate": 3.233638315615887e-06,
+      "loss": 0.6287,
+      "step": 8836
+    },
+    {
+      "epoch": 2.4479224376731303,
+      "grad_norm": 0.7076432108879089,
+      "learning_rate": 3.2332900614130296e-06,
+      "loss": 0.6353,
+      "step": 8837
+    },
+    {
+      "epoch": 2.4481994459833794,
+      "grad_norm": 0.7279842495918274,
+      "learning_rate": 3.2329417916405527e-06,
+      "loss": 0.5932,
+      "step": 8838
+    },
+    {
+      "epoch": 2.448476454293629,
+      "grad_norm": 0.7167196869850159,
+      "learning_rate": 3.232593506305849e-06,
+      "loss": 0.5794,
+      "step": 8839
+    },
+    {
+      "epoch": 2.4487534626038783,
+      "grad_norm": 0.7963582277297974,
+      "learning_rate": 3.232245205416314e-06,
+      "loss": 0.6319,
+      "step": 8840
+    },
+    {
+      "epoch": 2.4490304709141273,
+      "grad_norm": 0.7460035681724548,
+      "learning_rate": 3.2318968889793434e-06,
+      "loss": 0.5953,
+      "step": 8841
+    },
+    {
+      "epoch": 2.4493074792243767,
+      "grad_norm": 0.7167977690696716,
+      "learning_rate": 3.231548557002333e-06,
+      "loss": 0.5879,
+      "step": 8842
+    },
+    {
+      "epoch": 2.449584487534626,
+      "grad_norm": 0.752393901348114,
+      "learning_rate": 3.231200209492679e-06,
+      "loss": 0.6175,
+      "step": 8843
+    },
+    {
+      "epoch": 2.4498614958448752,
+      "grad_norm": 0.7277638912200928,
+      "learning_rate": 3.2308518464577767e-06,
+      "loss": 0.5695,
+      "step": 8844
+    },
+    {
+      "epoch": 2.4501385041551247,
+      "grad_norm": 0.7235784530639648,
+      "learning_rate": 3.230503467905024e-06,
+      "loss": 0.5852,
+      "step": 8845
+    },
+    {
+      "epoch": 2.450415512465374,
+      "grad_norm": 0.7358263731002808,
+      "learning_rate": 3.2301550738418165e-06,
+      "loss": 0.5507,
+      "step": 8846
+    },
+    {
+      "epoch": 2.450692520775623,
+      "grad_norm": 0.7292805910110474,
+      "learning_rate": 3.229806664275552e-06,
+      "loss": 0.6298,
+      "step": 8847
+    },
+    {
+      "epoch": 2.4509695290858726,
+      "grad_norm": 0.7070525288581848,
+      "learning_rate": 3.2294582392136293e-06,
+      "loss": 0.5415,
+      "step": 8848
+    },
+    {
+      "epoch": 2.4512465373961216,
+      "grad_norm": 0.7506101727485657,
+      "learning_rate": 3.229109798663444e-06,
+      "loss": 0.599,
+      "step": 8849
+    },
+    {
+      "epoch": 2.451523545706371,
+      "grad_norm": 0.7339098453521729,
+      "learning_rate": 3.2287613426323966e-06,
+      "loss": 0.5921,
+      "step": 8850
+    },
+    {
+      "epoch": 2.4518005540166206,
+      "grad_norm": 0.8022693991661072,
+      "learning_rate": 3.228412871127885e-06,
+      "loss": 0.5857,
+      "step": 8851
+    },
+    {
+      "epoch": 2.4520775623268696,
+      "grad_norm": 0.7564173340797424,
+      "learning_rate": 3.228064384157308e-06,
+      "loss": 0.6334,
+      "step": 8852
+    },
+    {
+      "epoch": 2.452354570637119,
+      "grad_norm": 0.7244262099266052,
+      "learning_rate": 3.227715881728064e-06,
+      "loss": 0.6374,
+      "step": 8853
+    },
+    {
+      "epoch": 2.4526315789473685,
+      "grad_norm": 0.7423130869865417,
+      "learning_rate": 3.227367363847554e-06,
+      "loss": 0.6213,
+      "step": 8854
+    },
+    {
+      "epoch": 2.4529085872576175,
+      "grad_norm": 0.7089356780052185,
+      "learning_rate": 3.2270188305231776e-06,
+      "loss": 0.5893,
+      "step": 8855
+    },
+    {
+      "epoch": 2.453185595567867,
+      "grad_norm": 0.7374478578567505,
+      "learning_rate": 3.2266702817623348e-06,
+      "loss": 0.5772,
+      "step": 8856
+    },
+    {
+      "epoch": 2.4534626038781164,
+      "grad_norm": 0.7225603461265564,
+      "learning_rate": 3.2263217175724255e-06,
+      "loss": 0.5922,
+      "step": 8857
+    },
+    {
+      "epoch": 2.4537396121883654,
+      "grad_norm": 0.7521680593490601,
+      "learning_rate": 3.225973137960852e-06,
+      "loss": 0.6135,
+      "step": 8858
+    },
+    {
+      "epoch": 2.454016620498615,
+      "grad_norm": 0.7221606373786926,
+      "learning_rate": 3.2256245429350142e-06,
+      "loss": 0.6049,
+      "step": 8859
+    },
+    {
+      "epoch": 2.4542936288088644,
+      "grad_norm": 0.7255619168281555,
+      "learning_rate": 3.2252759325023153e-06,
+      "loss": 0.605,
+      "step": 8860
+    },
+    {
+      "epoch": 2.4545706371191134,
+      "grad_norm": 0.7641960382461548,
+      "learning_rate": 3.224927306670156e-06,
+      "loss": 0.5981,
+      "step": 8861
+    },
+    {
+      "epoch": 2.454847645429363,
+      "grad_norm": 0.7062257528305054,
+      "learning_rate": 3.224578665445938e-06,
+      "loss": 0.5689,
+      "step": 8862
+    },
+    {
+      "epoch": 2.4551246537396123,
+      "grad_norm": 0.6948105096817017,
+      "learning_rate": 3.2242300088370646e-06,
+      "loss": 0.5594,
+      "step": 8863
+    },
+    {
+      "epoch": 2.4554016620498613,
+      "grad_norm": 0.6900063157081604,
+      "learning_rate": 3.2238813368509393e-06,
+      "loss": 0.5439,
+      "step": 8864
+    },
+    {
+      "epoch": 2.455678670360111,
+      "grad_norm": 0.7369881272315979,
+      "learning_rate": 3.2235326494949644e-06,
+      "loss": 0.5593,
+      "step": 8865
+    },
+    {
+      "epoch": 2.4559556786703602,
+      "grad_norm": 0.69451904296875,
+      "learning_rate": 3.2231839467765423e-06,
+      "loss": 0.5791,
+      "step": 8866
+    },
+    {
+      "epoch": 2.4562326869806093,
+      "grad_norm": 0.7775591611862183,
+      "learning_rate": 3.2228352287030807e-06,
+      "loss": 0.5762,
+      "step": 8867
+    },
+    {
+      "epoch": 2.4565096952908587,
+      "grad_norm": 0.7555267214775085,
+      "learning_rate": 3.2224864952819797e-06,
+      "loss": 0.5774,
+      "step": 8868
+    },
+    {
+      "epoch": 2.456786703601108,
+      "grad_norm": 0.7609298229217529,
+      "learning_rate": 3.2221377465206457e-06,
+      "loss": 0.5823,
+      "step": 8869
+    },
+    {
+      "epoch": 2.457063711911357,
+      "grad_norm": 0.7338734269142151,
+      "learning_rate": 3.221788982426483e-06,
+      "loss": 0.5866,
+      "step": 8870
+    },
+    {
+      "epoch": 2.4573407202216067,
+      "grad_norm": 0.7334607243537903,
+      "learning_rate": 3.2214402030068974e-06,
+      "loss": 0.6116,
+      "step": 8871
+    },
+    {
+      "epoch": 2.457617728531856,
+      "grad_norm": 0.7804694771766663,
+      "learning_rate": 3.2210914082692935e-06,
+      "loss": 0.6157,
+      "step": 8872
+    },
+    {
+      "epoch": 2.457894736842105,
+      "grad_norm": 0.682929515838623,
+      "learning_rate": 3.220742598221078e-06,
+      "loss": 0.5643,
+      "step": 8873
+    },
+    {
+      "epoch": 2.4581717451523546,
+      "grad_norm": 0.7627580761909485,
+      "learning_rate": 3.220393772869656e-06,
+      "loss": 0.5662,
+      "step": 8874
+    },
+    {
+      "epoch": 2.458448753462604,
+      "grad_norm": 0.741961658000946,
+      "learning_rate": 3.2200449322224347e-06,
+      "loss": 0.6184,
+      "step": 8875
+    },
+    {
+      "epoch": 2.458725761772853,
+      "grad_norm": 0.7630350589752197,
+      "learning_rate": 3.2196960762868206e-06,
+      "loss": 0.5638,
+      "step": 8876
+    },
+    {
+      "epoch": 2.4590027700831025,
+      "grad_norm": 0.7240167856216431,
+      "learning_rate": 3.2193472050702214e-06,
+      "loss": 0.5876,
+      "step": 8877
+    },
+    {
+      "epoch": 2.459279778393352,
+      "grad_norm": 0.7393755912780762,
+      "learning_rate": 3.218998318580043e-06,
+      "loss": 0.5972,
+      "step": 8878
+    },
+    {
+      "epoch": 2.459556786703601,
+      "grad_norm": 0.7540607452392578,
+      "learning_rate": 3.2186494168236947e-06,
+      "loss": 0.5818,
+      "step": 8879
+    },
+    {
+      "epoch": 2.4598337950138505,
+      "grad_norm": 0.6882795095443726,
+      "learning_rate": 3.2183004998085852e-06,
+      "loss": 0.5603,
+      "step": 8880
+    },
+    {
+      "epoch": 2.4601108033241,
+      "grad_norm": 0.7658278346061707,
+      "learning_rate": 3.2179515675421203e-06,
+      "loss": 0.6163,
+      "step": 8881
+    },
+    {
+      "epoch": 2.460387811634349,
+      "grad_norm": 0.7589587569236755,
+      "learning_rate": 3.2176026200317105e-06,
+      "loss": 0.6216,
+      "step": 8882
+    },
+    {
+      "epoch": 2.4606648199445984,
+      "grad_norm": 0.733665943145752,
+      "learning_rate": 3.2172536572847653e-06,
+      "loss": 0.5882,
+      "step": 8883
+    },
+    {
+      "epoch": 2.460941828254848,
+      "grad_norm": 0.7620385885238647,
+      "learning_rate": 3.216904679308692e-06,
+      "loss": 0.5885,
+      "step": 8884
+    },
+    {
+      "epoch": 2.461218836565097,
+      "grad_norm": 0.7346259355545044,
+      "learning_rate": 3.2165556861109027e-06,
+      "loss": 0.5972,
+      "step": 8885
+    },
+    {
+      "epoch": 2.4614958448753463,
+      "grad_norm": 0.7112541794776917,
+      "learning_rate": 3.216206677698806e-06,
+      "loss": 0.5705,
+      "step": 8886
+    },
+    {
+      "epoch": 2.461772853185596,
+      "grad_norm": 0.7470067739486694,
+      "learning_rate": 3.215857654079812e-06,
+      "loss": 0.6043,
+      "step": 8887
+    },
+    {
+      "epoch": 2.462049861495845,
+      "grad_norm": 0.7251924872398376,
+      "learning_rate": 3.2155086152613325e-06,
+      "loss": 0.5956,
+      "step": 8888
+    },
+    {
+      "epoch": 2.4623268698060943,
+      "grad_norm": 0.772384762763977,
+      "learning_rate": 3.215159561250778e-06,
+      "loss": 0.594,
+      "step": 8889
+    },
+    {
+      "epoch": 2.4626038781163437,
+      "grad_norm": 0.9380196332931519,
+      "learning_rate": 3.214810492055559e-06,
+      "loss": 0.6098,
+      "step": 8890
+    },
+    {
+      "epoch": 2.4628808864265928,
+      "grad_norm": 0.7411641478538513,
+      "learning_rate": 3.214461407683088e-06,
+      "loss": 0.5922,
+      "step": 8891
+    },
+    {
+      "epoch": 2.463157894736842,
+      "grad_norm": 0.7432118654251099,
+      "learning_rate": 3.2141123081407773e-06,
+      "loss": 0.6242,
+      "step": 8892
+    },
+    {
+      "epoch": 2.4634349030470912,
+      "grad_norm": 0.7496050000190735,
+      "learning_rate": 3.213763193436038e-06,
+      "loss": 0.6358,
+      "step": 8893
+    },
+    {
+      "epoch": 2.4637119113573407,
+      "grad_norm": 0.7449604868888855,
+      "learning_rate": 3.2134140635762834e-06,
+      "loss": 0.6158,
+      "step": 8894
+    },
+    {
+      "epoch": 2.46398891966759,
+      "grad_norm": 0.6991225481033325,
+      "learning_rate": 3.213064918568927e-06,
+      "loss": 0.5943,
+      "step": 8895
+    },
+    {
+      "epoch": 2.464265927977839,
+      "grad_norm": 0.7275388240814209,
+      "learning_rate": 3.2127157584213806e-06,
+      "loss": 0.6057,
+      "step": 8896
+    },
+    {
+      "epoch": 2.4645429362880886,
+      "grad_norm": 0.6911985278129578,
+      "learning_rate": 3.21236658314106e-06,
+      "loss": 0.5863,
+      "step": 8897
+    },
+    {
+      "epoch": 2.464819944598338,
+      "grad_norm": 0.7673735618591309,
+      "learning_rate": 3.212017392735376e-06,
+      "loss": 0.6264,
+      "step": 8898
+    },
+    {
+      "epoch": 2.465096952908587,
+      "grad_norm": 0.7564421892166138,
+      "learning_rate": 3.2116681872117455e-06,
+      "loss": 0.6232,
+      "step": 8899
+    },
+    {
+      "epoch": 2.4653739612188366,
+      "grad_norm": 0.7045862078666687,
+      "learning_rate": 3.2113189665775812e-06,
+      "loss": 0.5353,
+      "step": 8900
+    },
+    {
+      "epoch": 2.465650969529086,
+      "grad_norm": 0.7565885782241821,
+      "learning_rate": 3.2109697308402994e-06,
+      "loss": 0.5873,
+      "step": 8901
+    },
+    {
+      "epoch": 2.465927977839335,
+      "grad_norm": 0.7609415054321289,
+      "learning_rate": 3.210620480007315e-06,
+      "loss": 0.6316,
+      "step": 8902
+    },
+    {
+      "epoch": 2.4662049861495845,
+      "grad_norm": 0.7758066654205322,
+      "learning_rate": 3.210271214086042e-06,
+      "loss": 0.6035,
+      "step": 8903
+    },
+    {
+      "epoch": 2.466481994459834,
+      "grad_norm": 0.7386062741279602,
+      "learning_rate": 3.209921933083897e-06,
+      "loss": 0.6171,
+      "step": 8904
+    },
+    {
+      "epoch": 2.466759002770083,
+      "grad_norm": 0.7325729131698608,
+      "learning_rate": 3.2095726370082974e-06,
+      "loss": 0.606,
+      "step": 8905
+    },
+    {
+      "epoch": 2.4670360110803324,
+      "grad_norm": 0.8654530048370361,
+      "learning_rate": 3.209223325866659e-06,
+      "loss": 0.6013,
+      "step": 8906
+    },
+    {
+      "epoch": 2.467313019390582,
+      "grad_norm": 0.7589396238327026,
+      "learning_rate": 3.208873999666397e-06,
+      "loss": 0.5761,
+      "step": 8907
+    },
+    {
+      "epoch": 2.467590027700831,
+      "grad_norm": 0.7238092422485352,
+      "learning_rate": 3.2085246584149303e-06,
+      "loss": 0.6005,
+      "step": 8908
+    },
+    {
+      "epoch": 2.4678670360110804,
+      "grad_norm": 0.7196837067604065,
+      "learning_rate": 3.208175302119676e-06,
+      "loss": 0.5775,
+      "step": 8909
+    },
+    {
+      "epoch": 2.46814404432133,
+      "grad_norm": 0.740156888961792,
+      "learning_rate": 3.2078259307880518e-06,
+      "loss": 0.6192,
+      "step": 8910
+    },
+    {
+      "epoch": 2.468421052631579,
+      "grad_norm": 0.7320531010627747,
+      "learning_rate": 3.207476544427474e-06,
+      "loss": 0.5975,
+      "step": 8911
+    },
+    {
+      "epoch": 2.4686980609418283,
+      "grad_norm": 0.7081652879714966,
+      "learning_rate": 3.207127143045364e-06,
+      "loss": 0.6177,
+      "step": 8912
+    },
+    {
+      "epoch": 2.4689750692520773,
+      "grad_norm": 0.7383638620376587,
+      "learning_rate": 3.2067777266491383e-06,
+      "loss": 0.5823,
+      "step": 8913
+    },
+    {
+      "epoch": 2.469252077562327,
+      "grad_norm": 0.7729959487915039,
+      "learning_rate": 3.2064282952462166e-06,
+      "loss": 0.6159,
+      "step": 8914
+    },
+    {
+      "epoch": 2.4695290858725762,
+      "grad_norm": 0.7924338579177856,
+      "learning_rate": 3.206078848844019e-06,
+      "loss": 0.6329,
+      "step": 8915
+    },
+    {
+      "epoch": 2.4698060941828253,
+      "grad_norm": 0.7191593647003174,
+      "learning_rate": 3.2057293874499634e-06,
+      "loss": 0.5837,
+      "step": 8916
+    },
+    {
+      "epoch": 2.4700831024930747,
+      "grad_norm": 0.7165887355804443,
+      "learning_rate": 3.2053799110714704e-06,
+      "loss": 0.5895,
+      "step": 8917
+    },
+    {
+      "epoch": 2.470360110803324,
+      "grad_norm": 0.7286561131477356,
+      "learning_rate": 3.205030419715962e-06,
+      "loss": 0.5786,
+      "step": 8918
+    },
+    {
+      "epoch": 2.470637119113573,
+      "grad_norm": 0.7367630004882812,
+      "learning_rate": 3.204680913390856e-06,
+      "loss": 0.5704,
+      "step": 8919
+    },
+    {
+      "epoch": 2.4709141274238227,
+      "grad_norm": 0.7660843729972839,
+      "learning_rate": 3.2043313921035747e-06,
+      "loss": 0.6413,
+      "step": 8920
+    },
+    {
+      "epoch": 2.471191135734072,
+      "grad_norm": 0.7649071216583252,
+      "learning_rate": 3.2039818558615397e-06,
+      "loss": 0.57,
+      "step": 8921
+    },
+    {
+      "epoch": 2.471468144044321,
+      "grad_norm": 0.7629241347312927,
+      "learning_rate": 3.203632304672172e-06,
+      "loss": 0.639,
+      "step": 8922
+    },
+    {
+      "epoch": 2.4717451523545706,
+      "grad_norm": 0.7380210161209106,
+      "learning_rate": 3.2032827385428932e-06,
+      "loss": 0.585,
+      "step": 8923
+    },
+    {
+      "epoch": 2.47202216066482,
+      "grad_norm": 0.7167991995811462,
+      "learning_rate": 3.2029331574811263e-06,
+      "loss": 0.5925,
+      "step": 8924
+    },
+    {
+      "epoch": 2.472299168975069,
+      "grad_norm": 0.7697342038154602,
+      "learning_rate": 3.2025835614942933e-06,
+      "loss": 0.5853,
+      "step": 8925
+    },
+    {
+      "epoch": 2.4725761772853185,
+      "grad_norm": 0.7063153982162476,
+      "learning_rate": 3.202233950589817e-06,
+      "loss": 0.5851,
+      "step": 8926
+    },
+    {
+      "epoch": 2.472853185595568,
+      "grad_norm": 0.7513769268989563,
+      "learning_rate": 3.201884324775121e-06,
+      "loss": 0.5711,
+      "step": 8927
+    },
+    {
+      "epoch": 2.473130193905817,
+      "grad_norm": 0.7194318771362305,
+      "learning_rate": 3.2015346840576277e-06,
+      "loss": 0.5436,
+      "step": 8928
+    },
+    {
+      "epoch": 2.4734072022160665,
+      "grad_norm": 0.7521985769271851,
+      "learning_rate": 3.201185028444762e-06,
+      "loss": 0.5963,
+      "step": 8929
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.7365816235542297,
+      "learning_rate": 3.2008353579439477e-06,
+      "loss": 0.595,
+      "step": 8930
+    },
+    {
+      "epoch": 2.473961218836565,
+      "grad_norm": 0.7061067819595337,
+      "learning_rate": 3.200485672562609e-06,
+      "loss": 0.5693,
+      "step": 8931
+    },
+    {
+      "epoch": 2.4742382271468144,
+      "grad_norm": 0.7480983138084412,
+      "learning_rate": 3.2001359723081698e-06,
+      "loss": 0.5948,
+      "step": 8932
+    },
+    {
+      "epoch": 2.474515235457064,
+      "grad_norm": 0.7211030721664429,
+      "learning_rate": 3.199786257188057e-06,
+      "loss": 0.6069,
+      "step": 8933
+    },
+    {
+      "epoch": 2.474792243767313,
+      "grad_norm": 0.7504557371139526,
+      "learning_rate": 3.1994365272096943e-06,
+      "loss": 0.5919,
+      "step": 8934
+    },
+    {
+      "epoch": 2.4750692520775623,
+      "grad_norm": 0.7413660883903503,
+      "learning_rate": 3.1990867823805076e-06,
+      "loss": 0.6089,
+      "step": 8935
+    },
+    {
+      "epoch": 2.475346260387812,
+      "grad_norm": 0.74480140209198,
+      "learning_rate": 3.198737022707924e-06,
+      "loss": 0.6138,
+      "step": 8936
+    },
+    {
+      "epoch": 2.475623268698061,
+      "grad_norm": 0.7864667773246765,
+      "learning_rate": 3.1983872481993687e-06,
+      "loss": 0.595,
+      "step": 8937
+    },
+    {
+      "epoch": 2.4759002770083103,
+      "grad_norm": 0.73521888256073,
+      "learning_rate": 3.1980374588622692e-06,
+      "loss": 0.6243,
+      "step": 8938
+    },
+    {
+      "epoch": 2.4761772853185597,
+      "grad_norm": 0.7209713459014893,
+      "learning_rate": 3.197687654704051e-06,
+      "loss": 0.6126,
+      "step": 8939
+    },
+    {
+      "epoch": 2.4764542936288088,
+      "grad_norm": 0.7106951475143433,
+      "learning_rate": 3.1973378357321427e-06,
+      "loss": 0.5393,
+      "step": 8940
+    },
+    {
+      "epoch": 2.476731301939058,
+      "grad_norm": 0.7514813542366028,
+      "learning_rate": 3.1969880019539713e-06,
+      "loss": 0.598,
+      "step": 8941
+    },
+    {
+      "epoch": 2.4770083102493077,
+      "grad_norm": 0.7596821188926697,
+      "learning_rate": 3.1966381533769642e-06,
+      "loss": 0.6087,
+      "step": 8942
+    },
+    {
+      "epoch": 2.4772853185595567,
+      "grad_norm": 0.7739108204841614,
+      "learning_rate": 3.1962882900085507e-06,
+      "loss": 0.6094,
+      "step": 8943
+    },
+    {
+      "epoch": 2.477562326869806,
+      "grad_norm": 0.7667077779769897,
+      "learning_rate": 3.1959384118561587e-06,
+      "loss": 0.591,
+      "step": 8944
+    },
+    {
+      "epoch": 2.4778393351800556,
+      "grad_norm": 0.7343790531158447,
+      "learning_rate": 3.195588518927216e-06,
+      "loss": 0.5652,
+      "step": 8945
+    },
+    {
+      "epoch": 2.4781163434903046,
+      "grad_norm": 0.7375099062919617,
+      "learning_rate": 3.195238611229154e-06,
+      "loss": 0.5776,
+      "step": 8946
+    },
+    {
+      "epoch": 2.478393351800554,
+      "grad_norm": 0.7456963062286377,
+      "learning_rate": 3.1948886887694004e-06,
+      "loss": 0.6226,
+      "step": 8947
+    },
+    {
+      "epoch": 2.4786703601108035,
+      "grad_norm": 0.7197065353393555,
+      "learning_rate": 3.194538751555385e-06,
+      "loss": 0.598,
+      "step": 8948
+    },
+    {
+      "epoch": 2.4789473684210526,
+      "grad_norm": 0.715035617351532,
+      "learning_rate": 3.194188799594538e-06,
+      "loss": 0.5993,
+      "step": 8949
+    },
+    {
+      "epoch": 2.479224376731302,
+      "grad_norm": 0.763393759727478,
+      "learning_rate": 3.193838832894291e-06,
+      "loss": 0.5925,
+      "step": 8950
+    },
+    {
+      "epoch": 2.4795013850415515,
+      "grad_norm": 0.7357152104377747,
+      "learning_rate": 3.1934888514620733e-06,
+      "loss": 0.6367,
+      "step": 8951
+    },
+    {
+      "epoch": 2.4797783933518005,
+      "grad_norm": 0.746059000492096,
+      "learning_rate": 3.1931388553053156e-06,
+      "loss": 0.6413,
+      "step": 8952
+    },
+    {
+      "epoch": 2.48005540166205,
+      "grad_norm": 0.7911154627799988,
+      "learning_rate": 3.1927888444314514e-06,
+      "loss": 0.6082,
+      "step": 8953
+    },
+    {
+      "epoch": 2.480332409972299,
+      "grad_norm": 0.7596408128738403,
+      "learning_rate": 3.1924388188479093e-06,
+      "loss": 0.574,
+      "step": 8954
+    },
+    {
+      "epoch": 2.4806094182825484,
+      "grad_norm": 0.6973130702972412,
+      "learning_rate": 3.1920887785621233e-06,
+      "loss": 0.5997,
+      "step": 8955
+    },
+    {
+      "epoch": 2.480886426592798,
+      "grad_norm": 0.7324879765510559,
+      "learning_rate": 3.191738723581526e-06,
+      "loss": 0.5778,
+      "step": 8956
+    },
+    {
+      "epoch": 2.481163434903047,
+      "grad_norm": 0.7372378706932068,
+      "learning_rate": 3.1913886539135486e-06,
+      "loss": 0.592,
+      "step": 8957
+    },
+    {
+      "epoch": 2.4814404432132964,
+      "grad_norm": 0.7409850358963013,
+      "learning_rate": 3.1910385695656237e-06,
+      "loss": 0.6055,
+      "step": 8958
+    },
+    {
+      "epoch": 2.481717451523546,
+      "grad_norm": 0.7677171230316162,
+      "learning_rate": 3.1906884705451857e-06,
+      "loss": 0.6248,
+      "step": 8959
+    },
+    {
+      "epoch": 2.481994459833795,
+      "grad_norm": 0.7217593193054199,
+      "learning_rate": 3.190338356859668e-06,
+      "loss": 0.5707,
+      "step": 8960
+    },
+    {
+      "epoch": 2.4822714681440443,
+      "grad_norm": 0.704062819480896,
+      "learning_rate": 3.189988228516504e-06,
+      "loss": 0.5925,
+      "step": 8961
+    },
+    {
+      "epoch": 2.4825484764542938,
+      "grad_norm": 0.7098830938339233,
+      "learning_rate": 3.1896380855231274e-06,
+      "loss": 0.5572,
+      "step": 8962
+    },
+    {
+      "epoch": 2.482825484764543,
+      "grad_norm": 0.7203060984611511,
+      "learning_rate": 3.189287927886973e-06,
+      "loss": 0.5544,
+      "step": 8963
+    },
+    {
+      "epoch": 2.4831024930747922,
+      "grad_norm": 0.6978738903999329,
+      "learning_rate": 3.1889377556154765e-06,
+      "loss": 0.6224,
+      "step": 8964
+    },
+    {
+      "epoch": 2.4833795013850417,
+      "grad_norm": 0.7589989900588989,
+      "learning_rate": 3.1885875687160716e-06,
+      "loss": 0.6335,
+      "step": 8965
+    },
+    {
+      "epoch": 2.4836565096952907,
+      "grad_norm": 0.7540053725242615,
+      "learning_rate": 3.188237367196194e-06,
+      "loss": 0.594,
+      "step": 8966
+    },
+    {
+      "epoch": 2.48393351800554,
+      "grad_norm": 0.7392476201057434,
+      "learning_rate": 3.18788715106328e-06,
+      "loss": 0.6004,
+      "step": 8967
+    },
+    {
+      "epoch": 2.4842105263157896,
+      "grad_norm": 0.7759470343589783,
+      "learning_rate": 3.1875369203247648e-06,
+      "loss": 0.5797,
+      "step": 8968
+    },
+    {
+      "epoch": 2.4844875346260387,
+      "grad_norm": 0.7690457701683044,
+      "learning_rate": 3.1871866749880846e-06,
+      "loss": 0.6422,
+      "step": 8969
+    },
+    {
+      "epoch": 2.484764542936288,
+      "grad_norm": 0.7595497369766235,
+      "learning_rate": 3.1868364150606774e-06,
+      "loss": 0.6317,
+      "step": 8970
+    },
+    {
+      "epoch": 2.4850415512465376,
+      "grad_norm": 0.7643975615501404,
+      "learning_rate": 3.186486140549978e-06,
+      "loss": 0.6209,
+      "step": 8971
+    },
+    {
+      "epoch": 2.4853185595567866,
+      "grad_norm": 0.7429293394088745,
+      "learning_rate": 3.1861358514634257e-06,
+      "loss": 0.65,
+      "step": 8972
+    },
+    {
+      "epoch": 2.485595567867036,
+      "grad_norm": 0.6742926836013794,
+      "learning_rate": 3.1857855478084564e-06,
+      "loss": 0.5493,
+      "step": 8973
+    },
+    {
+      "epoch": 2.485872576177285,
+      "grad_norm": 0.7188777327537537,
+      "learning_rate": 3.185435229592509e-06,
+      "loss": 0.6283,
+      "step": 8974
+    },
+    {
+      "epoch": 2.4861495844875345,
+      "grad_norm": 0.7193275094032288,
+      "learning_rate": 3.1850848968230213e-06,
+      "loss": 0.6176,
+      "step": 8975
+    },
+    {
+      "epoch": 2.486426592797784,
+      "grad_norm": 0.738929033279419,
+      "learning_rate": 3.1847345495074312e-06,
+      "loss": 0.5802,
+      "step": 8976
+    },
+    {
+      "epoch": 2.486703601108033,
+      "grad_norm": 0.7202889919281006,
+      "learning_rate": 3.1843841876531783e-06,
+      "loss": 0.6494,
+      "step": 8977
+    },
+    {
+      "epoch": 2.4869806094182825,
+      "grad_norm": 0.720928430557251,
+      "learning_rate": 3.1840338112677017e-06,
+      "loss": 0.5937,
+      "step": 8978
+    },
+    {
+      "epoch": 2.487257617728532,
+      "grad_norm": 0.7134568691253662,
+      "learning_rate": 3.18368342035844e-06,
+      "loss": 0.5703,
+      "step": 8979
+    },
+    {
+      "epoch": 2.487534626038781,
+      "grad_norm": 0.7694113850593567,
+      "learning_rate": 3.1833330149328333e-06,
+      "loss": 0.5544,
+      "step": 8980
+    },
+    {
+      "epoch": 2.4878116343490304,
+      "grad_norm": 0.7342956066131592,
+      "learning_rate": 3.182982594998322e-06,
+      "loss": 0.6107,
+      "step": 8981
+    },
+    {
+      "epoch": 2.48808864265928,
+      "grad_norm": 0.7699294090270996,
+      "learning_rate": 3.1826321605623456e-06,
+      "loss": 0.5856,
+      "step": 8982
+    },
+    {
+      "epoch": 2.488365650969529,
+      "grad_norm": 0.7589293122291565,
+      "learning_rate": 3.1822817116323455e-06,
+      "loss": 0.6173,
+      "step": 8983
+    },
+    {
+      "epoch": 2.4886426592797783,
+      "grad_norm": 0.715154767036438,
+      "learning_rate": 3.181931248215762e-06,
+      "loss": 0.6067,
+      "step": 8984
+    },
+    {
+      "epoch": 2.488919667590028,
+      "grad_norm": 0.7071780562400818,
+      "learning_rate": 3.181580770320038e-06,
+      "loss": 0.5683,
+      "step": 8985
+    },
+    {
+      "epoch": 2.489196675900277,
+      "grad_norm": 0.760847806930542,
+      "learning_rate": 3.1812302779526124e-06,
+      "loss": 0.5814,
+      "step": 8986
+    },
+    {
+      "epoch": 2.4894736842105263,
+      "grad_norm": 0.7390440106391907,
+      "learning_rate": 3.1808797711209282e-06,
+      "loss": 0.5837,
+      "step": 8987
+    },
+    {
+      "epoch": 2.4897506925207757,
+      "grad_norm": 0.7559418678283691,
+      "learning_rate": 3.180529249832428e-06,
+      "loss": 0.6051,
+      "step": 8988
+    },
+    {
+      "epoch": 2.4900277008310248,
+      "grad_norm": 0.7289136648178101,
+      "learning_rate": 3.1801787140945544e-06,
+      "loss": 0.6074,
+      "step": 8989
+    },
+    {
+      "epoch": 2.490304709141274,
+      "grad_norm": 0.7169703841209412,
+      "learning_rate": 3.179828163914749e-06,
+      "loss": 0.5931,
+      "step": 8990
+    },
+    {
+      "epoch": 2.4905817174515237,
+      "grad_norm": 0.7092621922492981,
+      "learning_rate": 3.1794775993004557e-06,
+      "loss": 0.6474,
+      "step": 8991
+    },
+    {
+      "epoch": 2.4908587257617727,
+      "grad_norm": 0.7630212903022766,
+      "learning_rate": 3.1791270202591183e-06,
+      "loss": 0.6145,
+      "step": 8992
+    },
+    {
+      "epoch": 2.491135734072022,
+      "grad_norm": 0.6899539232254028,
+      "learning_rate": 3.1787764267981793e-06,
+      "loss": 0.5043,
+      "step": 8993
+    },
+    {
+      "epoch": 2.4914127423822716,
+      "grad_norm": 0.7707846760749817,
+      "learning_rate": 3.178425818925084e-06,
+      "loss": 0.6317,
+      "step": 8994
+    },
+    {
+      "epoch": 2.4916897506925206,
+      "grad_norm": 0.7168049216270447,
+      "learning_rate": 3.1780751966472754e-06,
+      "loss": 0.6151,
+      "step": 8995
+    },
+    {
+      "epoch": 2.49196675900277,
+      "grad_norm": 0.7072448134422302,
+      "learning_rate": 3.1777245599721993e-06,
+      "loss": 0.5583,
+      "step": 8996
+    },
+    {
+      "epoch": 2.4922437673130196,
+      "grad_norm": 0.744780957698822,
+      "learning_rate": 3.1773739089073e-06,
+      "loss": 0.5981,
+      "step": 8997
+    },
+    {
+      "epoch": 2.4925207756232686,
+      "grad_norm": 0.7633280158042908,
+      "learning_rate": 3.1770232434600223e-06,
+      "loss": 0.5987,
+      "step": 8998
+    },
+    {
+      "epoch": 2.492797783933518,
+      "grad_norm": 0.7523921132087708,
+      "learning_rate": 3.1766725636378125e-06,
+      "loss": 0.6084,
+      "step": 8999
+    },
+    {
+      "epoch": 2.4930747922437675,
+      "grad_norm": 0.7315114140510559,
+      "learning_rate": 3.1763218694481164e-06,
+      "loss": 0.6382,
+      "step": 9000
+    },
+    {
+      "epoch": 2.4933518005540165,
+      "grad_norm": 0.7941367030143738,
+      "learning_rate": 3.1759711608983794e-06,
+      "loss": 0.6669,
+      "step": 9001
+    },
+    {
+      "epoch": 2.493628808864266,
+      "grad_norm": 0.7714728116989136,
+      "learning_rate": 3.1756204379960494e-06,
+      "loss": 0.6014,
+      "step": 9002
+    },
+    {
+      "epoch": 2.4939058171745154,
+      "grad_norm": 0.7583754062652588,
+      "learning_rate": 3.1752697007485706e-06,
+      "loss": 0.5995,
+      "step": 9003
+    },
+    {
+      "epoch": 2.4941828254847644,
+      "grad_norm": 0.7508134841918945,
+      "learning_rate": 3.1749189491633926e-06,
+      "loss": 0.614,
+      "step": 9004
+    },
+    {
+      "epoch": 2.494459833795014,
+      "grad_norm": 0.7639035582542419,
+      "learning_rate": 3.174568183247962e-06,
+      "loss": 0.6044,
+      "step": 9005
+    },
+    {
+      "epoch": 2.4947368421052634,
+      "grad_norm": 0.7041856646537781,
+      "learning_rate": 3.174217403009725e-06,
+      "loss": 0.5785,
+      "step": 9006
+    },
+    {
+      "epoch": 2.4950138504155124,
+      "grad_norm": 0.7462358474731445,
+      "learning_rate": 3.173866608456132e-06,
+      "loss": 0.617,
+      "step": 9007
+    },
+    {
+      "epoch": 2.495290858725762,
+      "grad_norm": 0.7660013437271118,
+      "learning_rate": 3.17351579959463e-06,
+      "loss": 0.6425,
+      "step": 9008
+    },
+    {
+      "epoch": 2.4955678670360113,
+      "grad_norm": 0.7612264156341553,
+      "learning_rate": 3.1731649764326665e-06,
+      "loss": 0.6364,
+      "step": 9009
+    },
+    {
+      "epoch": 2.4958448753462603,
+      "grad_norm": 0.7336184978485107,
+      "learning_rate": 3.172814138977692e-06,
+      "loss": 0.6149,
+      "step": 9010
+    },
+    {
+      "epoch": 2.4961218836565098,
+      "grad_norm": 0.7028403878211975,
+      "learning_rate": 3.1724632872371566e-06,
+      "loss": 0.5825,
+      "step": 9011
+    },
+    {
+      "epoch": 2.4963988919667592,
+      "grad_norm": 0.7467078566551208,
+      "learning_rate": 3.1721124212185063e-06,
+      "loss": 0.5607,
+      "step": 9012
+    },
+    {
+      "epoch": 2.4966759002770083,
+      "grad_norm": 0.7464378476142883,
+      "learning_rate": 3.1717615409291936e-06,
+      "loss": 0.5978,
+      "step": 9013
+    },
+    {
+      "epoch": 2.4969529085872577,
+      "grad_norm": 0.7410820722579956,
+      "learning_rate": 3.171410646376668e-06,
+      "loss": 0.5892,
+      "step": 9014
+    },
+    {
+      "epoch": 2.497229916897507,
+      "grad_norm": 0.737551212310791,
+      "learning_rate": 3.17105973756838e-06,
+      "loss": 0.5904,
+      "step": 9015
+    },
+    {
+      "epoch": 2.497506925207756,
+      "grad_norm": 0.7431277632713318,
+      "learning_rate": 3.1707088145117794e-06,
+      "loss": 0.6267,
+      "step": 9016
+    },
+    {
+      "epoch": 2.4977839335180057,
+      "grad_norm": 0.7089725136756897,
+      "learning_rate": 3.170357877214319e-06,
+      "loss": 0.5503,
+      "step": 9017
+    },
+    {
+      "epoch": 2.4980609418282547,
+      "grad_norm": 0.7238456606864929,
+      "learning_rate": 3.1700069256834478e-06,
+      "loss": 0.5906,
+      "step": 9018
+    },
+    {
+      "epoch": 2.498337950138504,
+      "grad_norm": 0.7766544818878174,
+      "learning_rate": 3.1696559599266192e-06,
+      "loss": 0.578,
+      "step": 9019
+    },
+    {
+      "epoch": 2.4986149584487536,
+      "grad_norm": 0.7459467649459839,
+      "learning_rate": 3.1693049799512843e-06,
+      "loss": 0.6228,
+      "step": 9020
+    },
+    {
+      "epoch": 2.4988919667590026,
+      "grad_norm": 0.7257552146911621,
+      "learning_rate": 3.168953985764895e-06,
+      "loss": 0.6254,
+      "step": 9021
+    },
+    {
+      "epoch": 2.499168975069252,
+      "grad_norm": 0.7231073379516602,
+      "learning_rate": 3.1686029773749045e-06,
+      "loss": 0.5736,
+      "step": 9022
+    },
+    {
+      "epoch": 2.4994459833795015,
+      "grad_norm": 0.6939492225646973,
+      "learning_rate": 3.1682519547887658e-06,
+      "loss": 0.5798,
+      "step": 9023
+    },
+    {
+      "epoch": 2.4997229916897505,
+      "grad_norm": 0.7410045862197876,
+      "learning_rate": 3.1679009180139315e-06,
+      "loss": 0.612,
+      "step": 9024
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7503851652145386,
+      "learning_rate": 3.1675498670578544e-06,
+      "loss": 0.6032,
+      "step": 9025
+    },
+    {
+      "epoch": 2.5002770083102495,
+      "grad_norm": 0.7446761727333069,
+      "learning_rate": 3.16719880192799e-06,
+      "loss": 0.6122,
+      "step": 9026
+    },
+    {
+      "epoch": 2.5005540166204985,
+      "grad_norm": 0.7108980417251587,
+      "learning_rate": 3.16684772263179e-06,
+      "loss": 0.6324,
+      "step": 9027
+    },
+    {
+      "epoch": 2.500831024930748,
+      "grad_norm": 0.7736824750900269,
+      "learning_rate": 3.1664966291767106e-06,
+      "loss": 0.6475,
+      "step": 9028
+    },
+    {
+      "epoch": 2.501108033240997,
+      "grad_norm": 0.7655525803565979,
+      "learning_rate": 3.166145521570206e-06,
+      "loss": 0.5897,
+      "step": 9029
+    },
+    {
+      "epoch": 2.5013850415512464,
+      "grad_norm": 0.7319421172142029,
+      "learning_rate": 3.1657943998197305e-06,
+      "loss": 0.6392,
+      "step": 9030
+    },
+    {
+      "epoch": 2.501662049861496,
+      "grad_norm": 0.7983322143554688,
+      "learning_rate": 3.165443263932739e-06,
+      "loss": 0.5927,
+      "step": 9031
+    },
+    {
+      "epoch": 2.501939058171745,
+      "grad_norm": 0.7041431665420532,
+      "learning_rate": 3.165092113916688e-06,
+      "loss": 0.5515,
+      "step": 9032
+    },
+    {
+      "epoch": 2.5022160664819943,
+      "grad_norm": 0.709709882736206,
+      "learning_rate": 3.1647409497790333e-06,
+      "loss": 0.6061,
+      "step": 9033
+    },
+    {
+      "epoch": 2.502493074792244,
+      "grad_norm": 0.7182940244674683,
+      "learning_rate": 3.1643897715272306e-06,
+      "loss": 0.6129,
+      "step": 9034
+    },
+    {
+      "epoch": 2.502770083102493,
+      "grad_norm": 0.7012434601783752,
+      "learning_rate": 3.164038579168736e-06,
+      "loss": 0.5421,
+      "step": 9035
+    },
+    {
+      "epoch": 2.5030470914127423,
+      "grad_norm": 0.7288820743560791,
+      "learning_rate": 3.1636873727110073e-06,
+      "loss": 0.6085,
+      "step": 9036
+    },
+    {
+      "epoch": 2.5033240997229917,
+      "grad_norm": 0.7410088777542114,
+      "learning_rate": 3.1633361521615e-06,
+      "loss": 0.5921,
+      "step": 9037
+    },
+    {
+      "epoch": 2.5036011080332408,
+      "grad_norm": 0.6875979900360107,
+      "learning_rate": 3.1629849175276726e-06,
+      "loss": 0.5737,
+      "step": 9038
+    },
+    {
+      "epoch": 2.5038781163434902,
+      "grad_norm": 0.7268178462982178,
+      "learning_rate": 3.162633668816983e-06,
+      "loss": 0.6238,
+      "step": 9039
+    },
+    {
+      "epoch": 2.5041551246537397,
+      "grad_norm": 0.7688652276992798,
+      "learning_rate": 3.162282406036887e-06,
+      "loss": 0.6311,
+      "step": 9040
+    },
+    {
+      "epoch": 2.5044321329639887,
+      "grad_norm": 0.7122343182563782,
+      "learning_rate": 3.161931129194845e-06,
+      "loss": 0.5949,
+      "step": 9041
+    },
+    {
+      "epoch": 2.504709141274238,
+      "grad_norm": 0.7419912219047546,
+      "learning_rate": 3.1615798382983155e-06,
+      "loss": 0.5822,
+      "step": 9042
+    },
+    {
+      "epoch": 2.5049861495844876,
+      "grad_norm": 0.7522318959236145,
+      "learning_rate": 3.1612285333547556e-06,
+      "loss": 0.6055,
+      "step": 9043
+    },
+    {
+      "epoch": 2.5052631578947366,
+      "grad_norm": 0.7779367566108704,
+      "learning_rate": 3.160877214371625e-06,
+      "loss": 0.6163,
+      "step": 9044
+    },
+    {
+      "epoch": 2.505540166204986,
+      "grad_norm": 0.6923019289970398,
+      "learning_rate": 3.160525881356385e-06,
+      "loss": 0.5621,
+      "step": 9045
+    },
+    {
+      "epoch": 2.5058171745152356,
+      "grad_norm": 0.7429254055023193,
+      "learning_rate": 3.1601745343164917e-06,
+      "loss": 0.6012,
+      "step": 9046
+    },
+    {
+      "epoch": 2.5060941828254846,
+      "grad_norm": 0.713860034942627,
+      "learning_rate": 3.159823173259408e-06,
+      "loss": 0.579,
+      "step": 9047
+    },
+    {
+      "epoch": 2.506371191135734,
+      "grad_norm": 0.7542449235916138,
+      "learning_rate": 3.1594717981925945e-06,
+      "loss": 0.5522,
+      "step": 9048
+    },
+    {
+      "epoch": 2.5066481994459835,
+      "grad_norm": 0.7367417216300964,
+      "learning_rate": 3.159120409123509e-06,
+      "loss": 0.5463,
+      "step": 9049
+    },
+    {
+      "epoch": 2.5069252077562325,
+      "grad_norm": 0.714651882648468,
+      "learning_rate": 3.1587690060596143e-06,
+      "loss": 0.606,
+      "step": 9050
+    },
+    {
+      "epoch": 2.507202216066482,
+      "grad_norm": 0.7805426120758057,
+      "learning_rate": 3.1584175890083724e-06,
+      "loss": 0.6347,
+      "step": 9051
+    },
+    {
+      "epoch": 2.5074792243767314,
+      "grad_norm": 0.68819260597229,
+      "learning_rate": 3.1580661579772425e-06,
+      "loss": 0.6517,
+      "step": 9052
+    },
+    {
+      "epoch": 2.5077562326869804,
+      "grad_norm": 0.7432016730308533,
+      "learning_rate": 3.1577147129736875e-06,
+      "loss": 0.5857,
+      "step": 9053
+    },
+    {
+      "epoch": 2.50803324099723,
+      "grad_norm": 0.7592369914054871,
+      "learning_rate": 3.1573632540051702e-06,
+      "loss": 0.6106,
+      "step": 9054
+    },
+    {
+      "epoch": 2.5083102493074794,
+      "grad_norm": 0.7354898452758789,
+      "learning_rate": 3.157011781079152e-06,
+      "loss": 0.5963,
+      "step": 9055
+    },
+    {
+      "epoch": 2.5085872576177284,
+      "grad_norm": 0.7453470826148987,
+      "learning_rate": 3.156660294203096e-06,
+      "loss": 0.6188,
+      "step": 9056
+    },
+    {
+      "epoch": 2.508864265927978,
+      "grad_norm": 0.70748370885849,
+      "learning_rate": 3.156308793384465e-06,
+      "loss": 0.6119,
+      "step": 9057
+    },
+    {
+      "epoch": 2.5091412742382273,
+      "grad_norm": 0.7526542544364929,
+      "learning_rate": 3.1559572786307225e-06,
+      "loss": 0.5699,
+      "step": 9058
+    },
+    {
+      "epoch": 2.5094182825484763,
+      "grad_norm": 0.7344613075256348,
+      "learning_rate": 3.1556057499493307e-06,
+      "loss": 0.56,
+      "step": 9059
+    },
+    {
+      "epoch": 2.509695290858726,
+      "grad_norm": 0.7244347333908081,
+      "learning_rate": 3.1552542073477554e-06,
+      "loss": 0.6521,
+      "step": 9060
+    },
+    {
+      "epoch": 2.5099722991689752,
+      "grad_norm": 0.8829190731048584,
+      "learning_rate": 3.1549026508334607e-06,
+      "loss": 0.627,
+      "step": 9061
+    },
+    {
+      "epoch": 2.5102493074792243,
+      "grad_norm": 0.7521383762359619,
+      "learning_rate": 3.154551080413909e-06,
+      "loss": 0.5615,
+      "step": 9062
+    },
+    {
+      "epoch": 2.5105263157894737,
+      "grad_norm": 0.7531755566596985,
+      "learning_rate": 3.1541994960965674e-06,
+      "loss": 0.5935,
+      "step": 9063
+    },
+    {
+      "epoch": 2.510803324099723,
+      "grad_norm": 0.7439650893211365,
+      "learning_rate": 3.153847897888899e-06,
+      "loss": 0.5532,
+      "step": 9064
+    },
+    {
+      "epoch": 2.511080332409972,
+      "grad_norm": 0.746944010257721,
+      "learning_rate": 3.153496285798371e-06,
+      "loss": 0.5943,
+      "step": 9065
+    },
+    {
+      "epoch": 2.5113573407202217,
+      "grad_norm": 0.7312531471252441,
+      "learning_rate": 3.153144659832447e-06,
+      "loss": 0.6373,
+      "step": 9066
+    },
+    {
+      "epoch": 2.511634349030471,
+      "grad_norm": 0.741093099117279,
+      "learning_rate": 3.152793019998594e-06,
+      "loss": 0.5853,
+      "step": 9067
+    },
+    {
+      "epoch": 2.51191135734072,
+      "grad_norm": 0.7265978455543518,
+      "learning_rate": 3.1524413663042784e-06,
+      "loss": 0.5412,
+      "step": 9068
+    },
+    {
+      "epoch": 2.5121883656509696,
+      "grad_norm": 0.7318276166915894,
+      "learning_rate": 3.1520896987569666e-06,
+      "loss": 0.5864,
+      "step": 9069
+    },
+    {
+      "epoch": 2.512465373961219,
+      "grad_norm": 0.6997889280319214,
+      "learning_rate": 3.151738017364125e-06,
+      "loss": 0.5453,
+      "step": 9070
+    },
+    {
+      "epoch": 2.512742382271468,
+      "grad_norm": 0.7524303197860718,
+      "learning_rate": 3.1513863221332207e-06,
+      "loss": 0.6075,
+      "step": 9071
+    },
+    {
+      "epoch": 2.5130193905817175,
+      "grad_norm": 0.7480543255805969,
+      "learning_rate": 3.1510346130717217e-06,
+      "loss": 0.6009,
+      "step": 9072
+    },
+    {
+      "epoch": 2.513296398891967,
+      "grad_norm": 0.7288585901260376,
+      "learning_rate": 3.1506828901870958e-06,
+      "loss": 0.5811,
+      "step": 9073
+    },
+    {
+      "epoch": 2.513573407202216,
+      "grad_norm": 0.6908002495765686,
+      "learning_rate": 3.1503311534868097e-06,
+      "loss": 0.5221,
+      "step": 9074
+    },
+    {
+      "epoch": 2.5138504155124655,
+      "grad_norm": 0.7216440439224243,
+      "learning_rate": 3.149979402978332e-06,
+      "loss": 0.5988,
+      "step": 9075
+    },
+    {
+      "epoch": 2.514127423822715,
+      "grad_norm": 0.7588446140289307,
+      "learning_rate": 3.1496276386691327e-06,
+      "loss": 0.6346,
+      "step": 9076
+    },
+    {
+      "epoch": 2.514404432132964,
+      "grad_norm": 0.7544839382171631,
+      "learning_rate": 3.1492758605666796e-06,
+      "loss": 0.5734,
+      "step": 9077
+    },
+    {
+      "epoch": 2.5146814404432134,
+      "grad_norm": 0.7451032996177673,
+      "learning_rate": 3.1489240686784417e-06,
+      "loss": 0.6202,
+      "step": 9078
+    },
+    {
+      "epoch": 2.514958448753463,
+      "grad_norm": 0.7149298787117004,
+      "learning_rate": 3.1485722630118874e-06,
+      "loss": 0.5596,
+      "step": 9079
+    },
+    {
+      "epoch": 2.515235457063712,
+      "grad_norm": 0.7400791049003601,
+      "learning_rate": 3.1482204435744903e-06,
+      "loss": 0.6237,
+      "step": 9080
+    },
+    {
+      "epoch": 2.5155124653739613,
+      "grad_norm": 0.7374815940856934,
+      "learning_rate": 3.1478686103737154e-06,
+      "loss": 0.6092,
+      "step": 9081
+    },
+    {
+      "epoch": 2.515789473684211,
+      "grad_norm": 0.8207628726959229,
+      "learning_rate": 3.1475167634170365e-06,
+      "loss": 0.6153,
+      "step": 9082
+    },
+    {
+      "epoch": 2.51606648199446,
+      "grad_norm": 0.6817259788513184,
+      "learning_rate": 3.1471649027119232e-06,
+      "loss": 0.5391,
+      "step": 9083
+    },
+    {
+      "epoch": 2.5163434903047093,
+      "grad_norm": 0.737742006778717,
+      "learning_rate": 3.1468130282658467e-06,
+      "loss": 0.59,
+      "step": 9084
+    },
+    {
+      "epoch": 2.5166204986149583,
+      "grad_norm": 0.7260866761207581,
+      "learning_rate": 3.1464611400862766e-06,
+      "loss": 0.6319,
+      "step": 9085
+    },
+    {
+      "epoch": 2.5168975069252078,
+      "grad_norm": 0.6829667687416077,
+      "learning_rate": 3.146109238180686e-06,
+      "loss": 0.5478,
+      "step": 9086
+    },
+    {
+      "epoch": 2.517174515235457,
+      "grad_norm": 0.7200605869293213,
+      "learning_rate": 3.1457573225565472e-06,
+      "loss": 0.5868,
+      "step": 9087
+    },
+    {
+      "epoch": 2.5174515235457062,
+      "grad_norm": 0.7318903207778931,
+      "learning_rate": 3.145405393221331e-06,
+      "loss": 0.5995,
+      "step": 9088
+    },
+    {
+      "epoch": 2.5177285318559557,
+      "grad_norm": 0.6964850425720215,
+      "learning_rate": 3.145053450182509e-06,
+      "loss": 0.5842,
+      "step": 9089
+    },
+    {
+      "epoch": 2.518005540166205,
+      "grad_norm": 0.67149817943573,
+      "learning_rate": 3.144701493447556e-06,
+      "loss": 0.5927,
+      "step": 9090
+    },
+    {
+      "epoch": 2.518282548476454,
+      "grad_norm": 0.7514351606369019,
+      "learning_rate": 3.1443495230239442e-06,
+      "loss": 0.6169,
+      "step": 9091
+    },
+    {
+      "epoch": 2.5185595567867036,
+      "grad_norm": 0.7472149729728699,
+      "learning_rate": 3.1439975389191456e-06,
+      "loss": 0.6059,
+      "step": 9092
+    },
+    {
+      "epoch": 2.5188365650969526,
+      "grad_norm": 0.7219860553741455,
+      "learning_rate": 3.1436455411406356e-06,
+      "loss": 0.5874,
+      "step": 9093
+    },
+    {
+      "epoch": 2.519113573407202,
+      "grad_norm": 0.7607834935188293,
+      "learning_rate": 3.1432935296958865e-06,
+      "loss": 0.5884,
+      "step": 9094
+    },
+    {
+      "epoch": 2.5193905817174516,
+      "grad_norm": 0.7781618237495422,
+      "learning_rate": 3.1429415045923727e-06,
+      "loss": 0.6411,
+      "step": 9095
+    },
+    {
+      "epoch": 2.5196675900277006,
+      "grad_norm": 0.7018707394599915,
+      "learning_rate": 3.14258946583757e-06,
+      "loss": 0.5536,
+      "step": 9096
+    },
+    {
+      "epoch": 2.51994459833795,
+      "grad_norm": 0.7120254039764404,
+      "learning_rate": 3.142237413438951e-06,
+      "loss": 0.618,
+      "step": 9097
+    },
+    {
+      "epoch": 2.5202216066481995,
+      "grad_norm": 0.734156608581543,
+      "learning_rate": 3.1418853474039913e-06,
+      "loss": 0.5684,
+      "step": 9098
+    },
+    {
+      "epoch": 2.5204986149584485,
+      "grad_norm": 0.7341066002845764,
+      "learning_rate": 3.141533267740168e-06,
+      "loss": 0.5693,
+      "step": 9099
+    },
+    {
+      "epoch": 2.520775623268698,
+      "grad_norm": 0.7034329771995544,
+      "learning_rate": 3.1411811744549535e-06,
+      "loss": 0.5534,
+      "step": 9100
+    },
+    {
+      "epoch": 2.5210526315789474,
+      "grad_norm": 0.7042141556739807,
+      "learning_rate": 3.140829067555826e-06,
+      "loss": 0.5826,
+      "step": 9101
+    },
+    {
+      "epoch": 2.5213296398891965,
+      "grad_norm": 0.742813229560852,
+      "learning_rate": 3.1404769470502606e-06,
+      "loss": 0.6049,
+      "step": 9102
+    },
+    {
+      "epoch": 2.521606648199446,
+      "grad_norm": 0.7536957859992981,
+      "learning_rate": 3.140124812945734e-06,
+      "loss": 0.6111,
+      "step": 9103
+    },
+    {
+      "epoch": 2.5218836565096954,
+      "grad_norm": 0.7326040863990784,
+      "learning_rate": 3.139772665249723e-06,
+      "loss": 0.5774,
+      "step": 9104
+    },
+    {
+      "epoch": 2.5221606648199444,
+      "grad_norm": 0.7297189831733704,
+      "learning_rate": 3.1394205039697052e-06,
+      "loss": 0.6173,
+      "step": 9105
+    },
+    {
+      "epoch": 2.522437673130194,
+      "grad_norm": 0.7193797826766968,
+      "learning_rate": 3.1390683291131563e-06,
+      "loss": 0.5445,
+      "step": 9106
+    },
+    {
+      "epoch": 2.5227146814404433,
+      "grad_norm": 0.7528589367866516,
+      "learning_rate": 3.138716140687555e-06,
+      "loss": 0.5849,
+      "step": 9107
+    },
+    {
+      "epoch": 2.5229916897506923,
+      "grad_norm": 0.7352349758148193,
+      "learning_rate": 3.138363938700379e-06,
+      "loss": 0.6135,
+      "step": 9108
+    },
+    {
+      "epoch": 2.523268698060942,
+      "grad_norm": 0.7288463115692139,
+      "learning_rate": 3.138011723159107e-06,
+      "loss": 0.592,
+      "step": 9109
+    },
+    {
+      "epoch": 2.5235457063711912,
+      "grad_norm": 0.7227898836135864,
+      "learning_rate": 3.137659494071216e-06,
+      "loss": 0.5576,
+      "step": 9110
+    },
+    {
+      "epoch": 2.5238227146814403,
+      "grad_norm": 0.7009102702140808,
+      "learning_rate": 3.137307251444185e-06,
+      "loss": 0.568,
+      "step": 9111
+    },
+    {
+      "epoch": 2.5240997229916897,
+      "grad_norm": 0.7480496168136597,
+      "learning_rate": 3.136954995285495e-06,
+      "loss": 0.6032,
+      "step": 9112
+    },
+    {
+      "epoch": 2.524376731301939,
+      "grad_norm": 0.7623210549354553,
+      "learning_rate": 3.1366027256026226e-06,
+      "loss": 0.6047,
+      "step": 9113
+    },
+    {
+      "epoch": 2.524653739612188,
+      "grad_norm": 0.730629563331604,
+      "learning_rate": 3.136250442403049e-06,
+      "loss": 0.6057,
+      "step": 9114
+    },
+    {
+      "epoch": 2.5249307479224377,
+      "grad_norm": 0.7473331689834595,
+      "learning_rate": 3.1358981456942543e-06,
+      "loss": 0.5614,
+      "step": 9115
+    },
+    {
+      "epoch": 2.525207756232687,
+      "grad_norm": 0.7419679164886475,
+      "learning_rate": 3.1355458354837183e-06,
+      "loss": 0.609,
+      "step": 9116
+    },
+    {
+      "epoch": 2.525484764542936,
+      "grad_norm": 0.751901388168335,
+      "learning_rate": 3.13519351177892e-06,
+      "loss": 0.5983,
+      "step": 9117
+    },
+    {
+      "epoch": 2.5257617728531856,
+      "grad_norm": 0.7652584314346313,
+      "learning_rate": 3.134841174587342e-06,
+      "loss": 0.5973,
+      "step": 9118
+    },
+    {
+      "epoch": 2.526038781163435,
+      "grad_norm": 0.7338992953300476,
+      "learning_rate": 3.1344888239164647e-06,
+      "loss": 0.5932,
+      "step": 9119
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.6915652751922607,
+      "learning_rate": 3.1341364597737684e-06,
+      "loss": 0.5841,
+      "step": 9120
+    },
+    {
+      "epoch": 2.5265927977839335,
+      "grad_norm": 0.7430443167686462,
+      "learning_rate": 3.1337840821667363e-06,
+      "loss": 0.6248,
+      "step": 9121
+    },
+    {
+      "epoch": 2.526869806094183,
+      "grad_norm": 0.7357264161109924,
+      "learning_rate": 3.1334316911028508e-06,
+      "loss": 0.5925,
+      "step": 9122
+    },
+    {
+      "epoch": 2.527146814404432,
+      "grad_norm": 0.6949090361595154,
+      "learning_rate": 3.1330792865895914e-06,
+      "loss": 0.602,
+      "step": 9123
+    },
+    {
+      "epoch": 2.5274238227146815,
+      "grad_norm": 0.7181215286254883,
+      "learning_rate": 3.1327268686344426e-06,
+      "loss": 0.5912,
+      "step": 9124
+    },
+    {
+      "epoch": 2.527700831024931,
+      "grad_norm": 0.7383121252059937,
+      "learning_rate": 3.1323744372448867e-06,
+      "loss": 0.588,
+      "step": 9125
+    },
+    {
+      "epoch": 2.52797783933518,
+      "grad_norm": 0.7196062803268433,
+      "learning_rate": 3.1320219924284055e-06,
+      "loss": 0.5496,
+      "step": 9126
+    },
+    {
+      "epoch": 2.5282548476454294,
+      "grad_norm": 0.7350928783416748,
+      "learning_rate": 3.1316695341924847e-06,
+      "loss": 0.577,
+      "step": 9127
+    },
+    {
+      "epoch": 2.528531855955679,
+      "grad_norm": 0.6696033477783203,
+      "learning_rate": 3.1313170625446064e-06,
+      "loss": 0.5497,
+      "step": 9128
+    },
+    {
+      "epoch": 2.528808864265928,
+      "grad_norm": 0.7050279974937439,
+      "learning_rate": 3.1309645774922544e-06,
+      "loss": 0.5604,
+      "step": 9129
+    },
+    {
+      "epoch": 2.5290858725761773,
+      "grad_norm": 0.7316223382949829,
+      "learning_rate": 3.130612079042913e-06,
+      "loss": 0.5885,
+      "step": 9130
+    },
+    {
+      "epoch": 2.529362880886427,
+      "grad_norm": 0.7581227421760559,
+      "learning_rate": 3.130259567204067e-06,
+      "loss": 0.6254,
+      "step": 9131
+    },
+    {
+      "epoch": 2.529639889196676,
+      "grad_norm": 0.7304221391677856,
+      "learning_rate": 3.1299070419832006e-06,
+      "loss": 0.6362,
+      "step": 9132
+    },
+    {
+      "epoch": 2.5299168975069253,
+      "grad_norm": 0.7422270178794861,
+      "learning_rate": 3.1295545033877987e-06,
+      "loss": 0.5861,
+      "step": 9133
+    },
+    {
+      "epoch": 2.5301939058171747,
+      "grad_norm": 0.7402244210243225,
+      "learning_rate": 3.1292019514253478e-06,
+      "loss": 0.5923,
+      "step": 9134
+    },
+    {
+      "epoch": 2.5304709141274238,
+      "grad_norm": 0.7696836590766907,
+      "learning_rate": 3.1288493861033327e-06,
+      "loss": 0.5708,
+      "step": 9135
+    },
+    {
+      "epoch": 2.530747922437673,
+      "grad_norm": 0.7355846762657166,
+      "learning_rate": 3.1284968074292388e-06,
+      "loss": 0.5911,
+      "step": 9136
+    },
+    {
+      "epoch": 2.5310249307479227,
+      "grad_norm": 0.7229421734809875,
+      "learning_rate": 3.1281442154105528e-06,
+      "loss": 0.6072,
+      "step": 9137
+    },
+    {
+      "epoch": 2.5313019390581717,
+      "grad_norm": 0.7466177344322205,
+      "learning_rate": 3.1277916100547613e-06,
+      "loss": 0.6076,
+      "step": 9138
+    },
+    {
+      "epoch": 2.531578947368421,
+      "grad_norm": 0.7761610746383667,
+      "learning_rate": 3.12743899136935e-06,
+      "loss": 0.5949,
+      "step": 9139
+    },
+    {
+      "epoch": 2.5318559556786706,
+      "grad_norm": 0.7328271269798279,
+      "learning_rate": 3.1270863593618066e-06,
+      "loss": 0.5951,
+      "step": 9140
+    },
+    {
+      "epoch": 2.5321329639889196,
+      "grad_norm": 0.7607014775276184,
+      "learning_rate": 3.1267337140396193e-06,
+      "loss": 0.5789,
+      "step": 9141
+    },
+    {
+      "epoch": 2.532409972299169,
+      "grad_norm": 0.7491313815116882,
+      "learning_rate": 3.126381055410274e-06,
+      "loss": 0.635,
+      "step": 9142
+    },
+    {
+      "epoch": 2.5326869806094185,
+      "grad_norm": 0.7202014327049255,
+      "learning_rate": 3.1260283834812595e-06,
+      "loss": 0.6184,
+      "step": 9143
+    },
+    {
+      "epoch": 2.5329639889196676,
+      "grad_norm": 0.7408089637756348,
+      "learning_rate": 3.1256756982600642e-06,
+      "loss": 0.5675,
+      "step": 9144
+    },
+    {
+      "epoch": 2.533240997229917,
+      "grad_norm": 0.7064687013626099,
+      "learning_rate": 3.125322999754175e-06,
+      "loss": 0.5815,
+      "step": 9145
+    },
+    {
+      "epoch": 2.533518005540166,
+      "grad_norm": 0.7416212558746338,
+      "learning_rate": 3.124970287971082e-06,
+      "loss": 0.6292,
+      "step": 9146
+    },
+    {
+      "epoch": 2.5337950138504155,
+      "grad_norm": 0.7549963593482971,
+      "learning_rate": 3.1246175629182744e-06,
+      "loss": 0.592,
+      "step": 9147
+    },
+    {
+      "epoch": 2.534072022160665,
+      "grad_norm": 0.6834939122200012,
+      "learning_rate": 3.12426482460324e-06,
+      "loss": 0.5995,
+      "step": 9148
+    },
+    {
+      "epoch": 2.534349030470914,
+      "grad_norm": 0.7434266209602356,
+      "learning_rate": 3.12391207303347e-06,
+      "loss": 0.6189,
+      "step": 9149
+    },
+    {
+      "epoch": 2.5346260387811634,
+      "grad_norm": 0.718361496925354,
+      "learning_rate": 3.1235593082164527e-06,
+      "loss": 0.5777,
+      "step": 9150
+    },
+    {
+      "epoch": 2.534903047091413,
+      "grad_norm": 0.7535122036933899,
+      "learning_rate": 3.123206530159679e-06,
+      "loss": 0.6064,
+      "step": 9151
+    },
+    {
+      "epoch": 2.535180055401662,
+      "grad_norm": 0.7786555290222168,
+      "learning_rate": 3.1228537388706392e-06,
+      "loss": 0.5837,
+      "step": 9152
+    },
+    {
+      "epoch": 2.5354570637119114,
+      "grad_norm": 0.7389187812805176,
+      "learning_rate": 3.122500934356825e-06,
+      "loss": 0.5584,
+      "step": 9153
+    },
+    {
+      "epoch": 2.5357340720221604,
+      "grad_norm": 0.7314004898071289,
+      "learning_rate": 3.122148116625725e-06,
+      "loss": 0.5931,
+      "step": 9154
+    },
+    {
+      "epoch": 2.53601108033241,
+      "grad_norm": 0.7248198390007019,
+      "learning_rate": 3.121795285684832e-06,
+      "loss": 0.6388,
+      "step": 9155
+    },
+    {
+      "epoch": 2.5362880886426593,
+      "grad_norm": 0.6863377690315247,
+      "learning_rate": 3.1214424415416373e-06,
+      "loss": 0.5441,
+      "step": 9156
+    },
+    {
+      "epoch": 2.5365650969529083,
+      "grad_norm": 0.7189471125602722,
+      "learning_rate": 3.1210895842036326e-06,
+      "loss": 0.5627,
+      "step": 9157
+    },
+    {
+      "epoch": 2.536842105263158,
+      "grad_norm": 0.736508309841156,
+      "learning_rate": 3.1207367136783096e-06,
+      "loss": 0.588,
+      "step": 9158
+    },
+    {
+      "epoch": 2.5371191135734072,
+      "grad_norm": 0.7444504499435425,
+      "learning_rate": 3.1203838299731607e-06,
+      "loss": 0.5952,
+      "step": 9159
+    },
+    {
+      "epoch": 2.5373961218836563,
+      "grad_norm": 0.718096911907196,
+      "learning_rate": 3.12003093309568e-06,
+      "loss": 0.5898,
+      "step": 9160
+    },
+    {
+      "epoch": 2.5376731301939057,
+      "grad_norm": 0.7805988192558289,
+      "learning_rate": 3.119678023053358e-06,
+      "loss": 0.5813,
+      "step": 9161
+    },
+    {
+      "epoch": 2.537950138504155,
+      "grad_norm": 0.7315322756767273,
+      "learning_rate": 3.1193250998536894e-06,
+      "loss": 0.5954,
+      "step": 9162
+    },
+    {
+      "epoch": 2.538227146814404,
+      "grad_norm": 0.744376540184021,
+      "learning_rate": 3.118972163504168e-06,
+      "loss": 0.638,
+      "step": 9163
+    },
+    {
+      "epoch": 2.5385041551246537,
+      "grad_norm": 0.7060050964355469,
+      "learning_rate": 3.1186192140122863e-06,
+      "loss": 0.5638,
+      "step": 9164
+    },
+    {
+      "epoch": 2.538781163434903,
+      "grad_norm": 0.7389654517173767,
+      "learning_rate": 3.118266251385539e-06,
+      "loss": 0.5786,
+      "step": 9165
+    },
+    {
+      "epoch": 2.539058171745152,
+      "grad_norm": 0.8068351745605469,
+      "learning_rate": 3.1179132756314205e-06,
+      "loss": 0.649,
+      "step": 9166
+    },
+    {
+      "epoch": 2.5393351800554016,
+      "grad_norm": 0.7584858536720276,
+      "learning_rate": 3.117560286757425e-06,
+      "loss": 0.5953,
+      "step": 9167
+    },
+    {
+      "epoch": 2.539612188365651,
+      "grad_norm": 0.703183114528656,
+      "learning_rate": 3.117207284771048e-06,
+      "loss": 0.6066,
+      "step": 9168
+    },
+    {
+      "epoch": 2.5398891966759,
+      "grad_norm": 0.6991854906082153,
+      "learning_rate": 3.116854269679783e-06,
+      "loss": 0.545,
+      "step": 9169
+    },
+    {
+      "epoch": 2.5401662049861495,
+      "grad_norm": 0.7469720244407654,
+      "learning_rate": 3.116501241491128e-06,
+      "loss": 0.5961,
+      "step": 9170
+    },
+    {
+      "epoch": 2.540443213296399,
+      "grad_norm": 0.7148388028144836,
+      "learning_rate": 3.116148200212576e-06,
+      "loss": 0.5065,
+      "step": 9171
+    },
+    {
+      "epoch": 2.540720221606648,
+      "grad_norm": 0.7173408269882202,
+      "learning_rate": 3.115795145851625e-06,
+      "loss": 0.5957,
+      "step": 9172
+    },
+    {
+      "epoch": 2.5409972299168975,
+      "grad_norm": 0.7517775893211365,
+      "learning_rate": 3.11544207841577e-06,
+      "loss": 0.5684,
+      "step": 9173
+    },
+    {
+      "epoch": 2.541274238227147,
+      "grad_norm": 0.743894636631012,
+      "learning_rate": 3.115088997912508e-06,
+      "loss": 0.5854,
+      "step": 9174
+    },
+    {
+      "epoch": 2.541551246537396,
+      "grad_norm": 0.7796741724014282,
+      "learning_rate": 3.1147359043493365e-06,
+      "loss": 0.6541,
+      "step": 9175
+    },
+    {
+      "epoch": 2.5418282548476454,
+      "grad_norm": 0.7304288744926453,
+      "learning_rate": 3.1143827977337516e-06,
+      "loss": 0.5804,
+      "step": 9176
+    },
+    {
+      "epoch": 2.542105263157895,
+      "grad_norm": 0.722074031829834,
+      "learning_rate": 3.114029678073251e-06,
+      "loss": 0.596,
+      "step": 9177
+    },
+    {
+      "epoch": 2.542382271468144,
+      "grad_norm": 0.7742172479629517,
+      "learning_rate": 3.113676545375332e-06,
+      "loss": 0.5868,
+      "step": 9178
+    },
+    {
+      "epoch": 2.5426592797783933,
+      "grad_norm": 0.7246047854423523,
+      "learning_rate": 3.1133233996474933e-06,
+      "loss": 0.6096,
+      "step": 9179
+    },
+    {
+      "epoch": 2.542936288088643,
+      "grad_norm": 0.8202196359634399,
+      "learning_rate": 3.1129702408972316e-06,
+      "loss": 0.5459,
+      "step": 9180
+    },
+    {
+      "epoch": 2.543213296398892,
+      "grad_norm": 0.7334191799163818,
+      "learning_rate": 3.1126170691320466e-06,
+      "loss": 0.598,
+      "step": 9181
+    },
+    {
+      "epoch": 2.5434903047091413,
+      "grad_norm": 0.7115674614906311,
+      "learning_rate": 3.1122638843594377e-06,
+      "loss": 0.6247,
+      "step": 9182
+    },
+    {
+      "epoch": 2.5437673130193907,
+      "grad_norm": 0.6812747716903687,
+      "learning_rate": 3.1119106865869026e-06,
+      "loss": 0.5424,
+      "step": 9183
+    },
+    {
+      "epoch": 2.5440443213296398,
+      "grad_norm": 0.760249137878418,
+      "learning_rate": 3.1115574758219404e-06,
+      "loss": 0.6014,
+      "step": 9184
+    },
+    {
+      "epoch": 2.544321329639889,
+      "grad_norm": 0.7050676345825195,
+      "learning_rate": 3.111204252072052e-06,
+      "loss": 0.5766,
+      "step": 9185
+    },
+    {
+      "epoch": 2.5445983379501387,
+      "grad_norm": 0.7156957983970642,
+      "learning_rate": 3.1108510153447352e-06,
+      "loss": 0.5683,
+      "step": 9186
+    },
+    {
+      "epoch": 2.5448753462603877,
+      "grad_norm": 0.744422197341919,
+      "learning_rate": 3.1104977656474923e-06,
+      "loss": 0.587,
+      "step": 9187
+    },
+    {
+      "epoch": 2.545152354570637,
+      "grad_norm": 0.7329248785972595,
+      "learning_rate": 3.110144502987823e-06,
+      "loss": 0.5842,
+      "step": 9188
+    },
+    {
+      "epoch": 2.5454293628808866,
+      "grad_norm": 0.7504559755325317,
+      "learning_rate": 3.109791227373227e-06,
+      "loss": 0.614,
+      "step": 9189
+    },
+    {
+      "epoch": 2.5457063711911356,
+      "grad_norm": 0.7142932415008545,
+      "learning_rate": 3.109437938811207e-06,
+      "loss": 0.5663,
+      "step": 9190
+    },
+    {
+      "epoch": 2.545983379501385,
+      "grad_norm": 0.7647702693939209,
+      "learning_rate": 3.1090846373092624e-06,
+      "loss": 0.6442,
+      "step": 9191
+    },
+    {
+      "epoch": 2.5462603878116346,
+      "grad_norm": 0.7035728096961975,
+      "learning_rate": 3.108731322874896e-06,
+      "loss": 0.6074,
+      "step": 9192
+    },
+    {
+      "epoch": 2.5465373961218836,
+      "grad_norm": 0.7685317397117615,
+      "learning_rate": 3.108377995515609e-06,
+      "loss": 0.6127,
+      "step": 9193
+    },
+    {
+      "epoch": 2.546814404432133,
+      "grad_norm": 0.7221921682357788,
+      "learning_rate": 3.108024655238903e-06,
+      "loss": 0.5644,
+      "step": 9194
+    },
+    {
+      "epoch": 2.5470914127423825,
+      "grad_norm": 0.7649849653244019,
+      "learning_rate": 3.107671302052282e-06,
+      "loss": 0.6145,
+      "step": 9195
+    },
+    {
+      "epoch": 2.5473684210526315,
+      "grad_norm": 0.7878473401069641,
+      "learning_rate": 3.1073179359632465e-06,
+      "loss": 0.5651,
+      "step": 9196
+    },
+    {
+      "epoch": 2.547645429362881,
+      "grad_norm": 0.7036198377609253,
+      "learning_rate": 3.1069645569793004e-06,
+      "loss": 0.5796,
+      "step": 9197
+    },
+    {
+      "epoch": 2.5479224376731304,
+      "grad_norm": 0.7540354132652283,
+      "learning_rate": 3.1066111651079473e-06,
+      "loss": 0.6188,
+      "step": 9198
+    },
+    {
+      "epoch": 2.5481994459833794,
+      "grad_norm": 0.7185037732124329,
+      "learning_rate": 3.106257760356689e-06,
+      "loss": 0.5986,
+      "step": 9199
+    },
+    {
+      "epoch": 2.548476454293629,
+      "grad_norm": 0.7159909605979919,
+      "learning_rate": 3.1059043427330317e-06,
+      "loss": 0.5796,
+      "step": 9200
+    },
+    {
+      "epoch": 2.5487534626038784,
+      "grad_norm": 0.7194716930389404,
+      "learning_rate": 3.1055509122444768e-06,
+      "loss": 0.5852,
+      "step": 9201
+    },
+    {
+      "epoch": 2.5490304709141274,
+      "grad_norm": 0.7099069356918335,
+      "learning_rate": 3.1051974688985292e-06,
+      "loss": 0.5592,
+      "step": 9202
+    },
+    {
+      "epoch": 2.549307479224377,
+      "grad_norm": 0.7702147960662842,
+      "learning_rate": 3.1048440127026946e-06,
+      "loss": 0.5597,
+      "step": 9203
+    },
+    {
+      "epoch": 2.5495844875346263,
+      "grad_norm": 0.742659866809845,
+      "learning_rate": 3.1044905436644772e-06,
+      "loss": 0.5495,
+      "step": 9204
+    },
+    {
+      "epoch": 2.5498614958448753,
+      "grad_norm": 0.7312626242637634,
+      "learning_rate": 3.1041370617913818e-06,
+      "loss": 0.6404,
+      "step": 9205
+    },
+    {
+      "epoch": 2.5501385041551248,
+      "grad_norm": 0.7318844795227051,
+      "learning_rate": 3.1037835670909134e-06,
+      "loss": 0.6033,
+      "step": 9206
+    },
+    {
+      "epoch": 2.5504155124653742,
+      "grad_norm": 0.7092210650444031,
+      "learning_rate": 3.1034300595705784e-06,
+      "loss": 0.5582,
+      "step": 9207
+    },
+    {
+      "epoch": 2.5506925207756233,
+      "grad_norm": 0.7274510860443115,
+      "learning_rate": 3.1030765392378813e-06,
+      "loss": 0.6148,
+      "step": 9208
+    },
+    {
+      "epoch": 2.5509695290858727,
+      "grad_norm": 0.7525596022605896,
+      "learning_rate": 3.10272300610033e-06,
+      "loss": 0.5924,
+      "step": 9209
+    },
+    {
+      "epoch": 2.5512465373961217,
+      "grad_norm": 0.7422453761100769,
+      "learning_rate": 3.10236946016543e-06,
+      "loss": 0.5386,
+      "step": 9210
+    },
+    {
+      "epoch": 2.551523545706371,
+      "grad_norm": 0.7085767984390259,
+      "learning_rate": 3.1020159014406885e-06,
+      "loss": 0.5287,
+      "step": 9211
+    },
+    {
+      "epoch": 2.5518005540166206,
+      "grad_norm": 0.7217498421669006,
+      "learning_rate": 3.101662329933611e-06,
+      "loss": 0.5654,
+      "step": 9212
+    },
+    {
+      "epoch": 2.5520775623268697,
+      "grad_norm": 0.7595999836921692,
+      "learning_rate": 3.1013087456517064e-06,
+      "loss": 0.6237,
+      "step": 9213
+    },
+    {
+      "epoch": 2.552354570637119,
+      "grad_norm": 0.7279402017593384,
+      "learning_rate": 3.1009551486024814e-06,
+      "loss": 0.6141,
+      "step": 9214
+    },
+    {
+      "epoch": 2.5526315789473686,
+      "grad_norm": 0.7226688861846924,
+      "learning_rate": 3.100601538793444e-06,
+      "loss": 0.6513,
+      "step": 9215
+    },
+    {
+      "epoch": 2.5529085872576176,
+      "grad_norm": 0.741881787776947,
+      "learning_rate": 3.100247916232102e-06,
+      "loss": 0.603,
+      "step": 9216
+    },
+    {
+      "epoch": 2.553185595567867,
+      "grad_norm": 0.7893657684326172,
+      "learning_rate": 3.0998942809259643e-06,
+      "loss": 0.6009,
+      "step": 9217
+    },
+    {
+      "epoch": 2.553462603878116,
+      "grad_norm": 0.7880337238311768,
+      "learning_rate": 3.099540632882539e-06,
+      "loss": 0.6172,
+      "step": 9218
+    },
+    {
+      "epoch": 2.5537396121883655,
+      "grad_norm": 0.7472859621047974,
+      "learning_rate": 3.0991869721093342e-06,
+      "loss": 0.5933,
+      "step": 9219
+    },
+    {
+      "epoch": 2.554016620498615,
+      "grad_norm": 0.7323741316795349,
+      "learning_rate": 3.098833298613861e-06,
+      "loss": 0.6109,
+      "step": 9220
+    },
+    {
+      "epoch": 2.554293628808864,
+      "grad_norm": 0.7406541109085083,
+      "learning_rate": 3.0984796124036267e-06,
+      "loss": 0.5547,
+      "step": 9221
+    },
+    {
+      "epoch": 2.5545706371191135,
+      "grad_norm": 0.7627723813056946,
+      "learning_rate": 3.0981259134861426e-06,
+      "loss": 0.6178,
+      "step": 9222
+    },
+    {
+      "epoch": 2.554847645429363,
+      "grad_norm": 0.7222011089324951,
+      "learning_rate": 3.097772201868917e-06,
+      "loss": 0.5889,
+      "step": 9223
+    },
+    {
+      "epoch": 2.555124653739612,
+      "grad_norm": 0.7353097200393677,
+      "learning_rate": 3.0974184775594616e-06,
+      "loss": 0.5875,
+      "step": 9224
+    },
+    {
+      "epoch": 2.5554016620498614,
+      "grad_norm": 0.7108656167984009,
+      "learning_rate": 3.097064740565286e-06,
+      "loss": 0.5561,
+      "step": 9225
+    },
+    {
+      "epoch": 2.555678670360111,
+      "grad_norm": 0.7538282871246338,
+      "learning_rate": 3.096710990893902e-06,
+      "loss": 0.6066,
+      "step": 9226
+    },
+    {
+      "epoch": 2.55595567867036,
+      "grad_norm": 0.7404365539550781,
+      "learning_rate": 3.0963572285528186e-06,
+      "loss": 0.6072,
+      "step": 9227
+    },
+    {
+      "epoch": 2.5562326869806093,
+      "grad_norm": 0.7235316038131714,
+      "learning_rate": 3.0960034535495493e-06,
+      "loss": 0.632,
+      "step": 9228
+    },
+    {
+      "epoch": 2.556509695290859,
+      "grad_norm": 0.744694173336029,
+      "learning_rate": 3.0956496658916048e-06,
+      "loss": 0.6258,
+      "step": 9229
+    },
+    {
+      "epoch": 2.556786703601108,
+      "grad_norm": 0.7643329501152039,
+      "learning_rate": 3.0952958655864957e-06,
+      "loss": 0.6022,
+      "step": 9230
+    },
+    {
+      "epoch": 2.5570637119113573,
+      "grad_norm": 0.7612898349761963,
+      "learning_rate": 3.0949420526417357e-06,
+      "loss": 0.6229,
+      "step": 9231
+    },
+    {
+      "epoch": 2.5573407202216067,
+      "grad_norm": 0.7228426933288574,
+      "learning_rate": 3.094588227064837e-06,
+      "loss": 0.6418,
+      "step": 9232
+    },
+    {
+      "epoch": 2.5576177285318558,
+      "grad_norm": 0.7141867280006409,
+      "learning_rate": 3.0942343888633114e-06,
+      "loss": 0.5661,
+      "step": 9233
+    },
+    {
+      "epoch": 2.557894736842105,
+      "grad_norm": 0.7086227536201477,
+      "learning_rate": 3.0938805380446717e-06,
+      "loss": 0.582,
+      "step": 9234
+    },
+    {
+      "epoch": 2.5581717451523547,
+      "grad_norm": 0.6948062181472778,
+      "learning_rate": 3.0935266746164326e-06,
+      "loss": 0.5512,
+      "step": 9235
+    },
+    {
+      "epoch": 2.5584487534626037,
+      "grad_norm": 0.7016825675964355,
+      "learning_rate": 3.093172798586106e-06,
+      "loss": 0.6331,
+      "step": 9236
+    },
+    {
+      "epoch": 2.558725761772853,
+      "grad_norm": 0.774544358253479,
+      "learning_rate": 3.092818909961206e-06,
+      "loss": 0.5849,
+      "step": 9237
+    },
+    {
+      "epoch": 2.5590027700831026,
+      "grad_norm": 0.7048248052597046,
+      "learning_rate": 3.092465008749247e-06,
+      "loss": 0.5755,
+      "step": 9238
+    },
+    {
+      "epoch": 2.5592797783933516,
+      "grad_norm": 0.7674547433853149,
+      "learning_rate": 3.0921110949577425e-06,
+      "loss": 0.6052,
+      "step": 9239
+    },
+    {
+      "epoch": 2.559556786703601,
+      "grad_norm": 0.7589477896690369,
+      "learning_rate": 3.091757168594208e-06,
+      "loss": 0.6178,
+      "step": 9240
+    },
+    {
+      "epoch": 2.5598337950138506,
+      "grad_norm": 0.7594172358512878,
+      "learning_rate": 3.0914032296661566e-06,
+      "loss": 0.6464,
+      "step": 9241
+    },
+    {
+      "epoch": 2.5601108033240996,
+      "grad_norm": 0.7501422166824341,
+      "learning_rate": 3.0910492781811053e-06,
+      "loss": 0.623,
+      "step": 9242
+    },
+    {
+      "epoch": 2.560387811634349,
+      "grad_norm": 0.7276908755302429,
+      "learning_rate": 3.0906953141465677e-06,
+      "loss": 0.5813,
+      "step": 9243
+    },
+    {
+      "epoch": 2.5606648199445985,
+      "grad_norm": 0.7989957928657532,
+      "learning_rate": 3.09034133757006e-06,
+      "loss": 0.6081,
+      "step": 9244
+    },
+    {
+      "epoch": 2.5609418282548475,
+      "grad_norm": 0.754078209400177,
+      "learning_rate": 3.0899873484590994e-06,
+      "loss": 0.5448,
+      "step": 9245
+    },
+    {
+      "epoch": 2.561218836565097,
+      "grad_norm": 0.7445813417434692,
+      "learning_rate": 3.0896333468212e-06,
+      "loss": 0.5775,
+      "step": 9246
+    },
+    {
+      "epoch": 2.5614958448753464,
+      "grad_norm": 0.7109984159469604,
+      "learning_rate": 3.089279332663878e-06,
+      "loss": 0.5937,
+      "step": 9247
+    },
+    {
+      "epoch": 2.5617728531855954,
+      "grad_norm": 0.712151050567627,
+      "learning_rate": 3.088925305994652e-06,
+      "loss": 0.6057,
+      "step": 9248
+    },
+    {
+      "epoch": 2.562049861495845,
+      "grad_norm": 0.7844370603561401,
+      "learning_rate": 3.0885712668210376e-06,
+      "loss": 0.5883,
+      "step": 9249
+    },
+    {
+      "epoch": 2.5623268698060944,
+      "grad_norm": 0.747381329536438,
+      "learning_rate": 3.0882172151505523e-06,
+      "loss": 0.589,
+      "step": 9250
+    },
+    {
+      "epoch": 2.5626038781163434,
+      "grad_norm": 0.7583556175231934,
+      "learning_rate": 3.0878631509907127e-06,
+      "loss": 0.5968,
+      "step": 9251
+    },
+    {
+      "epoch": 2.562880886426593,
+      "grad_norm": 0.7440918684005737,
+      "learning_rate": 3.0875090743490383e-06,
+      "loss": 0.5754,
+      "step": 9252
+    },
+    {
+      "epoch": 2.5631578947368423,
+      "grad_norm": 0.7422070503234863,
+      "learning_rate": 3.087154985233045e-06,
+      "loss": 0.5908,
+      "step": 9253
+    },
+    {
+      "epoch": 2.5634349030470913,
+      "grad_norm": 0.8005231618881226,
+      "learning_rate": 3.086800883650252e-06,
+      "loss": 0.6444,
+      "step": 9254
+    },
+    {
+      "epoch": 2.563711911357341,
+      "grad_norm": 0.7100370526313782,
+      "learning_rate": 3.0864467696081785e-06,
+      "loss": 0.5453,
+      "step": 9255
+    },
+    {
+      "epoch": 2.5639889196675902,
+      "grad_norm": 0.6949264407157898,
+      "learning_rate": 3.086092643114342e-06,
+      "loss": 0.5595,
+      "step": 9256
+    },
+    {
+      "epoch": 2.5642659279778393,
+      "grad_norm": 0.7308492064476013,
+      "learning_rate": 3.085738504176261e-06,
+      "loss": 0.5984,
+      "step": 9257
+    },
+    {
+      "epoch": 2.5645429362880887,
+      "grad_norm": 0.7675203680992126,
+      "learning_rate": 3.085384352801457e-06,
+      "loss": 0.6014,
+      "step": 9258
+    },
+    {
+      "epoch": 2.564819944598338,
+      "grad_norm": 0.7616082429885864,
+      "learning_rate": 3.0850301889974477e-06,
+      "loss": 0.5905,
+      "step": 9259
+    },
+    {
+      "epoch": 2.565096952908587,
+      "grad_norm": 0.7435650825500488,
+      "learning_rate": 3.0846760127717534e-06,
+      "loss": 0.5649,
+      "step": 9260
+    },
+    {
+      "epoch": 2.5653739612188367,
+      "grad_norm": 0.7471524477005005,
+      "learning_rate": 3.084321824131895e-06,
+      "loss": 0.5982,
+      "step": 9261
+    },
+    {
+      "epoch": 2.565650969529086,
+      "grad_norm": 0.7317099571228027,
+      "learning_rate": 3.083967623085391e-06,
+      "loss": 0.5763,
+      "step": 9262
+    },
+    {
+      "epoch": 2.565927977839335,
+      "grad_norm": 0.7465954422950745,
+      "learning_rate": 3.0836134096397642e-06,
+      "loss": 0.6037,
+      "step": 9263
+    },
+    {
+      "epoch": 2.5662049861495846,
+      "grad_norm": 0.7400816082954407,
+      "learning_rate": 3.0832591838025337e-06,
+      "loss": 0.603,
+      "step": 9264
+    },
+    {
+      "epoch": 2.566481994459834,
+      "grad_norm": 0.7448139190673828,
+      "learning_rate": 3.0829049455812215e-06,
+      "loss": 0.5512,
+      "step": 9265
+    },
+    {
+      "epoch": 2.566759002770083,
+      "grad_norm": 0.7737401723861694,
+      "learning_rate": 3.0825506949833477e-06,
+      "loss": 0.6541,
+      "step": 9266
+    },
+    {
+      "epoch": 2.5670360110803325,
+      "grad_norm": 0.7257832884788513,
+      "learning_rate": 3.082196432016436e-06,
+      "loss": 0.5674,
+      "step": 9267
+    },
+    {
+      "epoch": 2.567313019390582,
+      "grad_norm": 0.7949074506759644,
+      "learning_rate": 3.0818421566880075e-06,
+      "loss": 0.6083,
+      "step": 9268
+    },
+    {
+      "epoch": 2.567590027700831,
+      "grad_norm": 0.703048050403595,
+      "learning_rate": 3.0814878690055836e-06,
+      "loss": 0.5917,
+      "step": 9269
+    },
+    {
+      "epoch": 2.5678670360110805,
+      "grad_norm": 0.7743631601333618,
+      "learning_rate": 3.081133568976687e-06,
+      "loss": 0.605,
+      "step": 9270
+    },
+    {
+      "epoch": 2.5681440443213295,
+      "grad_norm": 0.7685450911521912,
+      "learning_rate": 3.0807792566088417e-06,
+      "loss": 0.6056,
+      "step": 9271
+    },
+    {
+      "epoch": 2.568421052631579,
+      "grad_norm": 0.6992116570472717,
+      "learning_rate": 3.0804249319095686e-06,
+      "loss": 0.5336,
+      "step": 9272
+    },
+    {
+      "epoch": 2.5686980609418284,
+      "grad_norm": 0.7721379995346069,
+      "learning_rate": 3.0800705948863922e-06,
+      "loss": 0.5926,
+      "step": 9273
+    },
+    {
+      "epoch": 2.5689750692520774,
+      "grad_norm": 0.6962541937828064,
+      "learning_rate": 3.0797162455468367e-06,
+      "loss": 0.5648,
+      "step": 9274
+    },
+    {
+      "epoch": 2.569252077562327,
+      "grad_norm": 0.7683770656585693,
+      "learning_rate": 3.0793618838984233e-06,
+      "loss": 0.5939,
+      "step": 9275
+    },
+    {
+      "epoch": 2.5695290858725763,
+      "grad_norm": 0.709267795085907,
+      "learning_rate": 3.0790075099486787e-06,
+      "loss": 0.5971,
+      "step": 9276
+    },
+    {
+      "epoch": 2.5698060941828254,
+      "grad_norm": 0.7617084383964539,
+      "learning_rate": 3.078653123705126e-06,
+      "loss": 0.5788,
+      "step": 9277
+    },
+    {
+      "epoch": 2.570083102493075,
+      "grad_norm": 0.7246497273445129,
+      "learning_rate": 3.0782987251752894e-06,
+      "loss": 0.6082,
+      "step": 9278
+    },
+    {
+      "epoch": 2.570360110803324,
+      "grad_norm": 0.7413263916969299,
+      "learning_rate": 3.077944314366694e-06,
+      "loss": 0.6204,
+      "step": 9279
+    },
+    {
+      "epoch": 2.5706371191135733,
+      "grad_norm": 0.7515942454338074,
+      "learning_rate": 3.0775898912868653e-06,
+      "loss": 0.5992,
+      "step": 9280
+    },
+    {
+      "epoch": 2.5709141274238227,
+      "grad_norm": 0.7281467318534851,
+      "learning_rate": 3.077235455943328e-06,
+      "loss": 0.6147,
+      "step": 9281
+    },
+    {
+      "epoch": 2.5711911357340718,
+      "grad_norm": 0.746329128742218,
+      "learning_rate": 3.0768810083436072e-06,
+      "loss": 0.5948,
+      "step": 9282
+    },
+    {
+      "epoch": 2.5714681440443212,
+      "grad_norm": 0.7518007755279541,
+      "learning_rate": 3.076526548495231e-06,
+      "loss": 0.572,
+      "step": 9283
+    },
+    {
+      "epoch": 2.5717451523545707,
+      "grad_norm": 0.7037748098373413,
+      "learning_rate": 3.0761720764057222e-06,
+      "loss": 0.5791,
+      "step": 9284
+    },
+    {
+      "epoch": 2.5720221606648197,
+      "grad_norm": 0.7124174237251282,
+      "learning_rate": 3.0758175920826096e-06,
+      "loss": 0.558,
+      "step": 9285
+    },
+    {
+      "epoch": 2.572299168975069,
+      "grad_norm": 0.7646774649620056,
+      "learning_rate": 3.0754630955334187e-06,
+      "loss": 0.5765,
+      "step": 9286
+    },
+    {
+      "epoch": 2.5725761772853186,
+      "grad_norm": 0.7965149283409119,
+      "learning_rate": 3.075108586765677e-06,
+      "loss": 0.5765,
+      "step": 9287
+    },
+    {
+      "epoch": 2.5728531855955676,
+      "grad_norm": 0.7802436351776123,
+      "learning_rate": 3.0747540657869113e-06,
+      "loss": 0.598,
+      "step": 9288
+    },
+    {
+      "epoch": 2.573130193905817,
+      "grad_norm": 0.742306649684906,
+      "learning_rate": 3.0743995326046487e-06,
+      "loss": 0.564,
+      "step": 9289
+    },
+    {
+      "epoch": 2.5734072022160666,
+      "grad_norm": 0.7256410717964172,
+      "learning_rate": 3.0740449872264178e-06,
+      "loss": 0.6033,
+      "step": 9290
+    },
+    {
+      "epoch": 2.5736842105263156,
+      "grad_norm": 0.7125851511955261,
+      "learning_rate": 3.0736904296597457e-06,
+      "loss": 0.5585,
+      "step": 9291
+    },
+    {
+      "epoch": 2.573961218836565,
+      "grad_norm": 0.7355688810348511,
+      "learning_rate": 3.07333585991216e-06,
+      "loss": 0.5999,
+      "step": 9292
+    },
+    {
+      "epoch": 2.5742382271468145,
+      "grad_norm": 0.7512667775154114,
+      "learning_rate": 3.072981277991191e-06,
+      "loss": 0.5896,
+      "step": 9293
+    },
+    {
+      "epoch": 2.5745152354570635,
+      "grad_norm": 0.7280569076538086,
+      "learning_rate": 3.0726266839043656e-06,
+      "loss": 0.5661,
+      "step": 9294
+    },
+    {
+      "epoch": 2.574792243767313,
+      "grad_norm": 0.707859218120575,
+      "learning_rate": 3.0722720776592136e-06,
+      "loss": 0.5933,
+      "step": 9295
+    },
+    {
+      "epoch": 2.5750692520775624,
+      "grad_norm": 0.6893876194953918,
+      "learning_rate": 3.071917459263264e-06,
+      "loss": 0.5667,
+      "step": 9296
+    },
+    {
+      "epoch": 2.5753462603878114,
+      "grad_norm": 0.7232053875923157,
+      "learning_rate": 3.071562828724046e-06,
+      "loss": 0.6296,
+      "step": 9297
+    },
+    {
+      "epoch": 2.575623268698061,
+      "grad_norm": 0.7339138984680176,
+      "learning_rate": 3.0712081860490887e-06,
+      "loss": 0.6231,
+      "step": 9298
+    },
+    {
+      "epoch": 2.5759002770083104,
+      "grad_norm": 0.7207698822021484,
+      "learning_rate": 3.0708535312459244e-06,
+      "loss": 0.6124,
+      "step": 9299
+    },
+    {
+      "epoch": 2.5761772853185594,
+      "grad_norm": 0.7031732797622681,
+      "learning_rate": 3.070498864322081e-06,
+      "loss": 0.5907,
+      "step": 9300
+    },
+    {
+      "epoch": 2.576454293628809,
+      "grad_norm": 0.7313705086708069,
+      "learning_rate": 3.07014418528509e-06,
+      "loss": 0.6229,
+      "step": 9301
+    },
+    {
+      "epoch": 2.5767313019390583,
+      "grad_norm": 0.7237061262130737,
+      "learning_rate": 3.069789494142482e-06,
+      "loss": 0.6196,
+      "step": 9302
+    },
+    {
+      "epoch": 2.5770083102493073,
+      "grad_norm": 0.7333438992500305,
+      "learning_rate": 3.069434790901788e-06,
+      "loss": 0.575,
+      "step": 9303
+    },
+    {
+      "epoch": 2.577285318559557,
+      "grad_norm": 0.7455644011497498,
+      "learning_rate": 3.069080075570539e-06,
+      "loss": 0.5693,
+      "step": 9304
+    },
+    {
+      "epoch": 2.5775623268698062,
+      "grad_norm": 0.7772599458694458,
+      "learning_rate": 3.068725348156267e-06,
+      "loss": 0.5994,
+      "step": 9305
+    },
+    {
+      "epoch": 2.5778393351800553,
+      "grad_norm": 0.7441378831863403,
+      "learning_rate": 3.0683706086665038e-06,
+      "loss": 0.6199,
+      "step": 9306
+    },
+    {
+      "epoch": 2.5781163434903047,
+      "grad_norm": 0.7142274379730225,
+      "learning_rate": 3.0680158571087805e-06,
+      "loss": 0.6052,
+      "step": 9307
+    },
+    {
+      "epoch": 2.578393351800554,
+      "grad_norm": 0.7183087468147278,
+      "learning_rate": 3.06766109349063e-06,
+      "loss": 0.5608,
+      "step": 9308
+    },
+    {
+      "epoch": 2.578670360110803,
+      "grad_norm": 0.7334474325180054,
+      "learning_rate": 3.067306317819585e-06,
+      "loss": 0.5904,
+      "step": 9309
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.7224025726318359,
+      "learning_rate": 3.066951530103178e-06,
+      "loss": 0.5993,
+      "step": 9310
+    },
+    {
+      "epoch": 2.579224376731302,
+      "grad_norm": 0.7499139308929443,
+      "learning_rate": 3.0665967303489426e-06,
+      "loss": 0.6156,
+      "step": 9311
+    },
+    {
+      "epoch": 2.579501385041551,
+      "grad_norm": 0.7297726273536682,
+      "learning_rate": 3.0662419185644117e-06,
+      "loss": 0.6286,
+      "step": 9312
+    },
+    {
+      "epoch": 2.5797783933518006,
+      "grad_norm": 0.7476484775543213,
+      "learning_rate": 3.0658870947571184e-06,
+      "loss": 0.5896,
+      "step": 9313
+    },
+    {
+      "epoch": 2.58005540166205,
+      "grad_norm": 0.7497901916503906,
+      "learning_rate": 3.0655322589345963e-06,
+      "loss": 0.6288,
+      "step": 9314
+    },
+    {
+      "epoch": 2.580332409972299,
+      "grad_norm": 0.73838871717453,
+      "learning_rate": 3.0651774111043818e-06,
+      "loss": 0.5779,
+      "step": 9315
+    },
+    {
+      "epoch": 2.5806094182825485,
+      "grad_norm": 0.7194763422012329,
+      "learning_rate": 3.0648225512740067e-06,
+      "loss": 0.5829,
+      "step": 9316
+    },
+    {
+      "epoch": 2.580886426592798,
+      "grad_norm": 0.7514410614967346,
+      "learning_rate": 3.0644676794510063e-06,
+      "loss": 0.6221,
+      "step": 9317
+    },
+    {
+      "epoch": 2.581163434903047,
+      "grad_norm": 0.7325095534324646,
+      "learning_rate": 3.0641127956429157e-06,
+      "loss": 0.5811,
+      "step": 9318
+    },
+    {
+      "epoch": 2.5814404432132965,
+      "grad_norm": 0.7283647656440735,
+      "learning_rate": 3.06375789985727e-06,
+      "loss": 0.5841,
+      "step": 9319
+    },
+    {
+      "epoch": 2.581717451523546,
+      "grad_norm": 0.7078214287757874,
+      "learning_rate": 3.063402992101604e-06,
+      "loss": 0.6195,
+      "step": 9320
+    },
+    {
+      "epoch": 2.581994459833795,
+      "grad_norm": 0.7164523601531982,
+      "learning_rate": 3.063048072383454e-06,
+      "loss": 0.6172,
+      "step": 9321
+    },
+    {
+      "epoch": 2.5822714681440444,
+      "grad_norm": 0.7628939747810364,
+      "learning_rate": 3.0626931407103553e-06,
+      "loss": 0.6604,
+      "step": 9322
+    },
+    {
+      "epoch": 2.582548476454294,
+      "grad_norm": 0.7950290441513062,
+      "learning_rate": 3.0623381970898446e-06,
+      "loss": 0.6,
+      "step": 9323
+    },
+    {
+      "epoch": 2.582825484764543,
+      "grad_norm": 0.7289746403694153,
+      "learning_rate": 3.0619832415294576e-06,
+      "loss": 0.5875,
+      "step": 9324
+    },
+    {
+      "epoch": 2.5831024930747923,
+      "grad_norm": 0.7681859731674194,
+      "learning_rate": 3.061628274036732e-06,
+      "loss": 0.5503,
+      "step": 9325
+    },
+    {
+      "epoch": 2.583379501385042,
+      "grad_norm": 0.7553620338439941,
+      "learning_rate": 3.0612732946192032e-06,
+      "loss": 0.5748,
+      "step": 9326
+    },
+    {
+      "epoch": 2.583656509695291,
+      "grad_norm": 0.7522404789924622,
+      "learning_rate": 3.0609183032844093e-06,
+      "loss": 0.5743,
+      "step": 9327
+    },
+    {
+      "epoch": 2.5839335180055403,
+      "grad_norm": 0.740872859954834,
+      "learning_rate": 3.060563300039887e-06,
+      "loss": 0.5524,
+      "step": 9328
+    },
+    {
+      "epoch": 2.5842105263157897,
+      "grad_norm": 0.745161235332489,
+      "learning_rate": 3.0602082848931746e-06,
+      "loss": 0.6281,
+      "step": 9329
+    },
+    {
+      "epoch": 2.5844875346260388,
+      "grad_norm": 0.7245493531227112,
+      "learning_rate": 3.0598532578518097e-06,
+      "loss": 0.6164,
+      "step": 9330
+    },
+    {
+      "epoch": 2.584764542936288,
+      "grad_norm": 0.7285045981407166,
+      "learning_rate": 3.05949821892333e-06,
+      "loss": 0.5903,
+      "step": 9331
+    },
+    {
+      "epoch": 2.5850415512465377,
+      "grad_norm": 0.7477935552597046,
+      "learning_rate": 3.0591431681152756e-06,
+      "loss": 0.6193,
+      "step": 9332
+    },
+    {
+      "epoch": 2.5853185595567867,
+      "grad_norm": 0.7486900687217712,
+      "learning_rate": 3.058788105435182e-06,
+      "loss": 0.6019,
+      "step": 9333
+    },
+    {
+      "epoch": 2.585595567867036,
+      "grad_norm": 0.7068623900413513,
+      "learning_rate": 3.058433030890591e-06,
+      "loss": 0.5654,
+      "step": 9334
+    },
+    {
+      "epoch": 2.585872576177285,
+      "grad_norm": 0.7580755949020386,
+      "learning_rate": 3.058077944489041e-06,
+      "loss": 0.613,
+      "step": 9335
+    },
+    {
+      "epoch": 2.5861495844875346,
+      "grad_norm": 0.7915595173835754,
+      "learning_rate": 3.057722846238071e-06,
+      "loss": 0.6224,
+      "step": 9336
+    },
+    {
+      "epoch": 2.586426592797784,
+      "grad_norm": 0.772760272026062,
+      "learning_rate": 3.0573677361452204e-06,
+      "loss": 0.6323,
+      "step": 9337
+    },
+    {
+      "epoch": 2.586703601108033,
+      "grad_norm": 0.7564798593521118,
+      "learning_rate": 3.057012614218029e-06,
+      "loss": 0.5997,
+      "step": 9338
+    },
+    {
+      "epoch": 2.5869806094182826,
+      "grad_norm": 0.7556514143943787,
+      "learning_rate": 3.0566574804640375e-06,
+      "loss": 0.5907,
+      "step": 9339
+    },
+    {
+      "epoch": 2.587257617728532,
+      "grad_norm": 0.7544880509376526,
+      "learning_rate": 3.056302334890786e-06,
+      "loss": 0.5859,
+      "step": 9340
+    },
+    {
+      "epoch": 2.587534626038781,
+      "grad_norm": 0.8055458068847656,
+      "learning_rate": 3.0559471775058156e-06,
+      "loss": 0.6009,
+      "step": 9341
+    },
+    {
+      "epoch": 2.5878116343490305,
+      "grad_norm": 0.7201597690582275,
+      "learning_rate": 3.0555920083166673e-06,
+      "loss": 0.5794,
+      "step": 9342
+    },
+    {
+      "epoch": 2.5880886426592795,
+      "grad_norm": 0.7319862842559814,
+      "learning_rate": 3.0552368273308814e-06,
+      "loss": 0.5893,
+      "step": 9343
+    },
+    {
+      "epoch": 2.588365650969529,
+      "grad_norm": 0.7258013486862183,
+      "learning_rate": 3.0548816345559995e-06,
+      "loss": 0.5683,
+      "step": 9344
+    },
+    {
+      "epoch": 2.5886426592797784,
+      "grad_norm": 0.7262520790100098,
+      "learning_rate": 3.0545264299995643e-06,
+      "loss": 0.5766,
+      "step": 9345
+    },
+    {
+      "epoch": 2.5889196675900275,
+      "grad_norm": 0.7422440052032471,
+      "learning_rate": 3.0541712136691158e-06,
+      "loss": 0.6025,
+      "step": 9346
+    },
+    {
+      "epoch": 2.589196675900277,
+      "grad_norm": 0.7541458606719971,
+      "learning_rate": 3.0538159855721984e-06,
+      "loss": 0.5908,
+      "step": 9347
+    },
+    {
+      "epoch": 2.5894736842105264,
+      "grad_norm": 0.7304244637489319,
+      "learning_rate": 3.053460745716352e-06,
+      "loss": 0.581,
+      "step": 9348
+    },
+    {
+      "epoch": 2.5897506925207754,
+      "grad_norm": 0.7334722280502319,
+      "learning_rate": 3.0531054941091214e-06,
+      "loss": 0.5827,
+      "step": 9349
+    },
+    {
+      "epoch": 2.590027700831025,
+      "grad_norm": 0.7334140539169312,
+      "learning_rate": 3.0527502307580486e-06,
+      "loss": 0.6097,
+      "step": 9350
+    },
+    {
+      "epoch": 2.5903047091412743,
+      "grad_norm": 0.7113667130470276,
+      "learning_rate": 3.052394955670677e-06,
+      "loss": 0.6159,
+      "step": 9351
+    },
+    {
+      "epoch": 2.5905817174515233,
+      "grad_norm": 0.7234874367713928,
+      "learning_rate": 3.0520396688545496e-06,
+      "loss": 0.6341,
+      "step": 9352
+    },
+    {
+      "epoch": 2.590858725761773,
+      "grad_norm": 0.7299479842185974,
+      "learning_rate": 3.0516843703172107e-06,
+      "loss": 0.5955,
+      "step": 9353
+    },
+    {
+      "epoch": 2.5911357340720222,
+      "grad_norm": 0.7276083827018738,
+      "learning_rate": 3.051329060066203e-06,
+      "loss": 0.5695,
+      "step": 9354
+    },
+    {
+      "epoch": 2.5914127423822713,
+      "grad_norm": 0.6600974798202515,
+      "learning_rate": 3.050973738109072e-06,
+      "loss": 0.5695,
+      "step": 9355
+    },
+    {
+      "epoch": 2.5916897506925207,
+      "grad_norm": 0.7582259178161621,
+      "learning_rate": 3.050618404453361e-06,
+      "loss": 0.625,
+      "step": 9356
+    },
+    {
+      "epoch": 2.59196675900277,
+      "grad_norm": 0.8048502206802368,
+      "learning_rate": 3.050263059106615e-06,
+      "loss": 0.6124,
+      "step": 9357
+    },
+    {
+      "epoch": 2.592243767313019,
+      "grad_norm": 0.7244232296943665,
+      "learning_rate": 3.04990770207638e-06,
+      "loss": 0.5737,
+      "step": 9358
+    },
+    {
+      "epoch": 2.5925207756232687,
+      "grad_norm": 0.750846803188324,
+      "learning_rate": 3.0495523333701993e-06,
+      "loss": 0.5174,
+      "step": 9359
+    },
+    {
+      "epoch": 2.592797783933518,
+      "grad_norm": 0.7714466452598572,
+      "learning_rate": 3.0491969529956202e-06,
+      "loss": 0.593,
+      "step": 9360
+    },
+    {
+      "epoch": 2.593074792243767,
+      "grad_norm": 0.7113026976585388,
+      "learning_rate": 3.0488415609601863e-06,
+      "loss": 0.5794,
+      "step": 9361
+    },
+    {
+      "epoch": 2.5933518005540166,
+      "grad_norm": 0.7664023637771606,
+      "learning_rate": 3.0484861572714446e-06,
+      "loss": 0.6544,
+      "step": 9362
+    },
+    {
+      "epoch": 2.593628808864266,
+      "grad_norm": 0.7229735851287842,
+      "learning_rate": 3.048130741936942e-06,
+      "loss": 0.578,
+      "step": 9363
+    },
+    {
+      "epoch": 2.593905817174515,
+      "grad_norm": 0.7353859543800354,
+      "learning_rate": 3.0477753149642232e-06,
+      "loss": 0.6325,
+      "step": 9364
+    },
+    {
+      "epoch": 2.5941828254847645,
+      "grad_norm": 0.7365750670433044,
+      "learning_rate": 3.0474198763608347e-06,
+      "loss": 0.6022,
+      "step": 9365
+    },
+    {
+      "epoch": 2.594459833795014,
+      "grad_norm": 0.7197463512420654,
+      "learning_rate": 3.0470644261343262e-06,
+      "loss": 0.5467,
+      "step": 9366
+    },
+    {
+      "epoch": 2.594736842105263,
+      "grad_norm": 0.7440528273582458,
+      "learning_rate": 3.0467089642922415e-06,
+      "loss": 0.6267,
+      "step": 9367
+    },
+    {
+      "epoch": 2.5950138504155125,
+      "grad_norm": 0.7311400175094604,
+      "learning_rate": 3.04635349084213e-06,
+      "loss": 0.5595,
+      "step": 9368
+    },
+    {
+      "epoch": 2.595290858725762,
+      "grad_norm": 0.7559048533439636,
+      "learning_rate": 3.0459980057915383e-06,
+      "loss": 0.5791,
+      "step": 9369
+    },
+    {
+      "epoch": 2.595567867036011,
+      "grad_norm": 0.7741128206253052,
+      "learning_rate": 3.045642509148015e-06,
+      "loss": 0.6215,
+      "step": 9370
+    },
+    {
+      "epoch": 2.5958448753462604,
+      "grad_norm": 0.7354574799537659,
+      "learning_rate": 3.0452870009191067e-06,
+      "loss": 0.5868,
+      "step": 9371
+    },
+    {
+      "epoch": 2.59612188365651,
+      "grad_norm": 0.7495721578598022,
+      "learning_rate": 3.0449314811123636e-06,
+      "loss": 0.587,
+      "step": 9372
+    },
+    {
+      "epoch": 2.596398891966759,
+      "grad_norm": 0.7539669275283813,
+      "learning_rate": 3.044575949735334e-06,
+      "loss": 0.6182,
+      "step": 9373
+    },
+    {
+      "epoch": 2.5966759002770083,
+      "grad_norm": 0.736501157283783,
+      "learning_rate": 3.0442204067955657e-06,
+      "loss": 0.6197,
+      "step": 9374
+    },
+    {
+      "epoch": 2.596952908587258,
+      "grad_norm": 0.7483559250831604,
+      "learning_rate": 3.0438648523006088e-06,
+      "loss": 0.5887,
+      "step": 9375
+    },
+    {
+      "epoch": 2.597229916897507,
+      "grad_norm": 0.7232039570808411,
+      "learning_rate": 3.043509286258012e-06,
+      "loss": 0.579,
+      "step": 9376
+    },
+    {
+      "epoch": 2.5975069252077563,
+      "grad_norm": 0.7397487759590149,
+      "learning_rate": 3.0431537086753247e-06,
+      "loss": 0.5991,
+      "step": 9377
+    },
+    {
+      "epoch": 2.5977839335180057,
+      "grad_norm": 0.782886266708374,
+      "learning_rate": 3.0427981195600975e-06,
+      "loss": 0.6576,
+      "step": 9378
+    },
+    {
+      "epoch": 2.5980609418282548,
+      "grad_norm": 0.8020239472389221,
+      "learning_rate": 3.042442518919881e-06,
+      "loss": 0.5827,
+      "step": 9379
+    },
+    {
+      "epoch": 2.598337950138504,
+      "grad_norm": 0.6829942464828491,
+      "learning_rate": 3.0420869067622227e-06,
+      "loss": 0.5674,
+      "step": 9380
+    },
+    {
+      "epoch": 2.5986149584487537,
+      "grad_norm": 0.7796286940574646,
+      "learning_rate": 3.041731283094676e-06,
+      "loss": 0.6177,
+      "step": 9381
+    },
+    {
+      "epoch": 2.5988919667590027,
+      "grad_norm": 0.7237001657485962,
+      "learning_rate": 3.0413756479247915e-06,
+      "loss": 0.5829,
+      "step": 9382
+    },
+    {
+      "epoch": 2.599168975069252,
+      "grad_norm": 0.7349637746810913,
+      "learning_rate": 3.041020001260118e-06,
+      "loss": 0.6106,
+      "step": 9383
+    },
+    {
+      "epoch": 2.5994459833795016,
+      "grad_norm": 0.7296071648597717,
+      "learning_rate": 3.0406643431082088e-06,
+      "loss": 0.5722,
+      "step": 9384
+    },
+    {
+      "epoch": 2.5997229916897506,
+      "grad_norm": 0.7325433492660522,
+      "learning_rate": 3.0403086734766152e-06,
+      "loss": 0.5514,
+      "step": 9385
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7508443593978882,
+      "learning_rate": 3.0399529923728887e-06,
+      "loss": 0.5739,
+      "step": 9386
+    },
+    {
+      "epoch": 2.6002770083102495,
+      "grad_norm": 0.7678627371788025,
+      "learning_rate": 3.0395972998045802e-06,
+      "loss": 0.5893,
+      "step": 9387
+    },
+    {
+      "epoch": 2.6005540166204986,
+      "grad_norm": 0.9463872909545898,
+      "learning_rate": 3.0392415957792448e-06,
+      "loss": 0.5723,
+      "step": 9388
+    },
+    {
+      "epoch": 2.600831024930748,
+      "grad_norm": 0.7246379256248474,
+      "learning_rate": 3.038885880304432e-06,
+      "loss": 0.5698,
+      "step": 9389
+    },
+    {
+      "epoch": 2.6011080332409975,
+      "grad_norm": 0.7395390868186951,
+      "learning_rate": 3.0385301533876965e-06,
+      "loss": 0.6199,
+      "step": 9390
+    },
+    {
+      "epoch": 2.6013850415512465,
+      "grad_norm": 0.748449444770813,
+      "learning_rate": 3.0381744150365897e-06,
+      "loss": 0.6023,
+      "step": 9391
+    },
+    {
+      "epoch": 2.601662049861496,
+      "grad_norm": 0.7404671907424927,
+      "learning_rate": 3.0378186652586667e-06,
+      "loss": 0.5641,
+      "step": 9392
+    },
+    {
+      "epoch": 2.6019390581717454,
+      "grad_norm": 0.7181746363639832,
+      "learning_rate": 3.0374629040614788e-06,
+      "loss": 0.597,
+      "step": 9393
+    },
+    {
+      "epoch": 2.6022160664819944,
+      "grad_norm": 0.7396783232688904,
+      "learning_rate": 3.0371071314525822e-06,
+      "loss": 0.5861,
+      "step": 9394
+    },
+    {
+      "epoch": 2.602493074792244,
+      "grad_norm": 0.7946869730949402,
+      "learning_rate": 3.0367513474395293e-06,
+      "loss": 0.5751,
+      "step": 9395
+    },
+    {
+      "epoch": 2.602770083102493,
+      "grad_norm": 0.730797529220581,
+      "learning_rate": 3.0363955520298742e-06,
+      "loss": 0.604,
+      "step": 9396
+    },
+    {
+      "epoch": 2.6030470914127424,
+      "grad_norm": 0.7194575667381287,
+      "learning_rate": 3.0360397452311713e-06,
+      "loss": 0.5726,
+      "step": 9397
+    },
+    {
+      "epoch": 2.603324099722992,
+      "grad_norm": 0.727661669254303,
+      "learning_rate": 3.035683927050977e-06,
+      "loss": 0.6121,
+      "step": 9398
+    },
+    {
+      "epoch": 2.603601108033241,
+      "grad_norm": 0.7224810719490051,
+      "learning_rate": 3.0353280974968443e-06,
+      "loss": 0.618,
+      "step": 9399
+    },
+    {
+      "epoch": 2.6038781163434903,
+      "grad_norm": 0.7539339661598206,
+      "learning_rate": 3.034972256576328e-06,
+      "loss": 0.6185,
+      "step": 9400
+    },
+    {
+      "epoch": 2.6041551246537398,
+      "grad_norm": 0.7253374457359314,
+      "learning_rate": 3.034616404296986e-06,
+      "loss": 0.5817,
+      "step": 9401
+    },
+    {
+      "epoch": 2.604432132963989,
+      "grad_norm": 0.7376832365989685,
+      "learning_rate": 3.0342605406663717e-06,
+      "loss": 0.6031,
+      "step": 9402
+    },
+    {
+      "epoch": 2.6047091412742382,
+      "grad_norm": 0.7316452860832214,
+      "learning_rate": 3.033904665692042e-06,
+      "loss": 0.5959,
+      "step": 9403
+    },
+    {
+      "epoch": 2.6049861495844873,
+      "grad_norm": 0.717576265335083,
+      "learning_rate": 3.033548779381553e-06,
+      "loss": 0.5744,
+      "step": 9404
+    },
+    {
+      "epoch": 2.6052631578947367,
+      "grad_norm": 0.6863055229187012,
+      "learning_rate": 3.0331928817424603e-06,
+      "loss": 0.5895,
+      "step": 9405
+    },
+    {
+      "epoch": 2.605540166204986,
+      "grad_norm": 0.7601767778396606,
+      "learning_rate": 3.0328369727823216e-06,
+      "loss": 0.6063,
+      "step": 9406
+    },
+    {
+      "epoch": 2.605817174515235,
+      "grad_norm": 0.7255947589874268,
+      "learning_rate": 3.0324810525086933e-06,
+      "loss": 0.5959,
+      "step": 9407
+    },
+    {
+      "epoch": 2.6060941828254847,
+      "grad_norm": 0.6987281441688538,
+      "learning_rate": 3.0321251209291323e-06,
+      "loss": 0.5663,
+      "step": 9408
+    },
+    {
+      "epoch": 2.606371191135734,
+      "grad_norm": 0.7383803725242615,
+      "learning_rate": 3.031769178051196e-06,
+      "loss": 0.5965,
+      "step": 9409
+    },
+    {
+      "epoch": 2.606648199445983,
+      "grad_norm": 0.7562844753265381,
+      "learning_rate": 3.0314132238824416e-06,
+      "loss": 0.5968,
+      "step": 9410
+    },
+    {
+      "epoch": 2.6069252077562326,
+      "grad_norm": 0.7574225664138794,
+      "learning_rate": 3.031057258430429e-06,
+      "loss": 0.5961,
+      "step": 9411
+    },
+    {
+      "epoch": 2.607202216066482,
+      "grad_norm": 0.6951255798339844,
+      "learning_rate": 3.0307012817027124e-06,
+      "loss": 0.5847,
+      "step": 9412
+    },
+    {
+      "epoch": 2.607479224376731,
+      "grad_norm": 0.7327682971954346,
+      "learning_rate": 3.030345293706854e-06,
+      "loss": 0.5954,
+      "step": 9413
+    },
+    {
+      "epoch": 2.6077562326869805,
+      "grad_norm": 0.7225331664085388,
+      "learning_rate": 3.02998929445041e-06,
+      "loss": 0.5196,
+      "step": 9414
+    },
+    {
+      "epoch": 2.60803324099723,
+      "grad_norm": 0.7270516157150269,
+      "learning_rate": 3.02963328394094e-06,
+      "loss": 0.5885,
+      "step": 9415
+    },
+    {
+      "epoch": 2.608310249307479,
+      "grad_norm": 0.7316294312477112,
+      "learning_rate": 3.0292772621860023e-06,
+      "loss": 0.5801,
+      "step": 9416
+    },
+    {
+      "epoch": 2.6085872576177285,
+      "grad_norm": 0.76264488697052,
+      "learning_rate": 3.0289212291931576e-06,
+      "loss": 0.597,
+      "step": 9417
+    },
+    {
+      "epoch": 2.608864265927978,
+      "grad_norm": 0.7400818467140198,
+      "learning_rate": 3.028565184969964e-06,
+      "loss": 0.6227,
+      "step": 9418
+    },
+    {
+      "epoch": 2.609141274238227,
+      "grad_norm": 0.7234652042388916,
+      "learning_rate": 3.0282091295239814e-06,
+      "loss": 0.56,
+      "step": 9419
+    },
+    {
+      "epoch": 2.6094182825484764,
+      "grad_norm": 0.75355064868927,
+      "learning_rate": 3.0278530628627713e-06,
+      "loss": 0.5878,
+      "step": 9420
+    },
+    {
+      "epoch": 2.609695290858726,
+      "grad_norm": 0.6999250054359436,
+      "learning_rate": 3.0274969849938914e-06,
+      "loss": 0.5948,
+      "step": 9421
+    },
+    {
+      "epoch": 2.609972299168975,
+      "grad_norm": 0.7420984506607056,
+      "learning_rate": 3.0271408959249037e-06,
+      "loss": 0.5955,
+      "step": 9422
+    },
+    {
+      "epoch": 2.6102493074792243,
+      "grad_norm": 0.7665450572967529,
+      "learning_rate": 3.026784795663369e-06,
+      "loss": 0.6236,
+      "step": 9423
+    },
+    {
+      "epoch": 2.610526315789474,
+      "grad_norm": 0.7386185526847839,
+      "learning_rate": 3.026428684216848e-06,
+      "loss": 0.6541,
+      "step": 9424
+    },
+    {
+      "epoch": 2.610803324099723,
+      "grad_norm": 0.7392768263816833,
+      "learning_rate": 3.0260725615929e-06,
+      "loss": 0.62,
+      "step": 9425
+    },
+    {
+      "epoch": 2.6110803324099723,
+      "grad_norm": 0.7245485186576843,
+      "learning_rate": 3.02571642779909e-06,
+      "loss": 0.6487,
+      "step": 9426
+    },
+    {
+      "epoch": 2.6113573407202217,
+      "grad_norm": 0.7127525806427002,
+      "learning_rate": 3.0253602828429774e-06,
+      "loss": 0.5901,
+      "step": 9427
+    },
+    {
+      "epoch": 2.6116343490304708,
+      "grad_norm": 0.7617471814155579,
+      "learning_rate": 3.0250041267321234e-06,
+      "loss": 0.6056,
+      "step": 9428
+    },
+    {
+      "epoch": 2.61191135734072,
+      "grad_norm": 0.7574014067649841,
+      "learning_rate": 3.0246479594740918e-06,
+      "loss": 0.6256,
+      "step": 9429
+    },
+    {
+      "epoch": 2.6121883656509697,
+      "grad_norm": 0.6903671026229858,
+      "learning_rate": 3.0242917810764444e-06,
+      "loss": 0.5489,
+      "step": 9430
+    },
+    {
+      "epoch": 2.6124653739612187,
+      "grad_norm": 0.6944217085838318,
+      "learning_rate": 3.0239355915467433e-06,
+      "loss": 0.5895,
+      "step": 9431
+    },
+    {
+      "epoch": 2.612742382271468,
+      "grad_norm": 0.7788039445877075,
+      "learning_rate": 3.0235793908925514e-06,
+      "loss": 0.5921,
+      "step": 9432
+    },
+    {
+      "epoch": 2.6130193905817176,
+      "grad_norm": 0.7304327487945557,
+      "learning_rate": 3.023223179121433e-06,
+      "loss": 0.5859,
+      "step": 9433
+    },
+    {
+      "epoch": 2.6132963988919666,
+      "grad_norm": 0.7489771842956543,
+      "learning_rate": 3.022866956240949e-06,
+      "loss": 0.5972,
+      "step": 9434
+    },
+    {
+      "epoch": 2.613573407202216,
+      "grad_norm": 0.7207261323928833,
+      "learning_rate": 3.022510722258665e-06,
+      "loss": 0.627,
+      "step": 9435
+    },
+    {
+      "epoch": 2.6138504155124656,
+      "grad_norm": 0.7087815403938293,
+      "learning_rate": 3.022154477182144e-06,
+      "loss": 0.5543,
+      "step": 9436
+    },
+    {
+      "epoch": 2.6141274238227146,
+      "grad_norm": 0.7843639254570007,
+      "learning_rate": 3.0217982210189496e-06,
+      "loss": 0.5903,
+      "step": 9437
+    },
+    {
+      "epoch": 2.614404432132964,
+      "grad_norm": 0.7218259572982788,
+      "learning_rate": 3.021441953776646e-06,
+      "loss": 0.5908,
+      "step": 9438
+    },
+    {
+      "epoch": 2.6146814404432135,
+      "grad_norm": 0.7581648230552673,
+      "learning_rate": 3.0210856754627993e-06,
+      "loss": 0.5855,
+      "step": 9439
+    },
+    {
+      "epoch": 2.6149584487534625,
+      "grad_norm": 0.752878725528717,
+      "learning_rate": 3.0207293860849725e-06,
+      "loss": 0.5708,
+      "step": 9440
+    },
+    {
+      "epoch": 2.615235457063712,
+      "grad_norm": 0.6773368716239929,
+      "learning_rate": 3.02037308565073e-06,
+      "loss": 0.5578,
+      "step": 9441
+    },
+    {
+      "epoch": 2.6155124653739614,
+      "grad_norm": 0.875789225101471,
+      "learning_rate": 3.02001677416764e-06,
+      "loss": 0.5907,
+      "step": 9442
+    },
+    {
+      "epoch": 2.6157894736842104,
+      "grad_norm": 0.7268041372299194,
+      "learning_rate": 3.0196604516432646e-06,
+      "loss": 0.6047,
+      "step": 9443
+    },
+    {
+      "epoch": 2.61606648199446,
+      "grad_norm": 0.7566415667533875,
+      "learning_rate": 3.0193041180851712e-06,
+      "loss": 0.5801,
+      "step": 9444
+    },
+    {
+      "epoch": 2.6163434903047094,
+      "grad_norm": 0.6960709095001221,
+      "learning_rate": 3.0189477735009255e-06,
+      "loss": 0.562,
+      "step": 9445
+    },
+    {
+      "epoch": 2.6166204986149584,
+      "grad_norm": 0.7445232272148132,
+      "learning_rate": 3.0185914178980924e-06,
+      "loss": 0.6094,
+      "step": 9446
+    },
+    {
+      "epoch": 2.616897506925208,
+      "grad_norm": 0.7602598667144775,
+      "learning_rate": 3.0182350512842396e-06,
+      "loss": 0.6036,
+      "step": 9447
+    },
+    {
+      "epoch": 2.6171745152354573,
+      "grad_norm": 0.7294239401817322,
+      "learning_rate": 3.017878673666933e-06,
+      "loss": 0.575,
+      "step": 9448
+    },
+    {
+      "epoch": 2.6174515235457063,
+      "grad_norm": 0.7791168689727783,
+      "learning_rate": 3.01752228505374e-06,
+      "loss": 0.6117,
+      "step": 9449
+    },
+    {
+      "epoch": 2.6177285318559558,
+      "grad_norm": 0.7321231365203857,
+      "learning_rate": 3.0171658854522274e-06,
+      "loss": 0.582,
+      "step": 9450
+    },
+    {
+      "epoch": 2.6180055401662052,
+      "grad_norm": 0.7250687479972839,
+      "learning_rate": 3.0168094748699616e-06,
+      "loss": 0.5873,
+      "step": 9451
+    },
+    {
+      "epoch": 2.6182825484764543,
+      "grad_norm": 0.7413502931594849,
+      "learning_rate": 3.0164530533145128e-06,
+      "loss": 0.6043,
+      "step": 9452
+    },
+    {
+      "epoch": 2.6185595567867037,
+      "grad_norm": 0.6866628527641296,
+      "learning_rate": 3.0160966207934444e-06,
+      "loss": 0.5894,
+      "step": 9453
+    },
+    {
+      "epoch": 2.618836565096953,
+      "grad_norm": 0.7689675092697144,
+      "learning_rate": 3.015740177314328e-06,
+      "loss": 0.6129,
+      "step": 9454
+    },
+    {
+      "epoch": 2.619113573407202,
+      "grad_norm": 0.7070875763893127,
+      "learning_rate": 3.0153837228847306e-06,
+      "loss": 0.5956,
+      "step": 9455
+    },
+    {
+      "epoch": 2.6193905817174516,
+      "grad_norm": 0.7495928406715393,
+      "learning_rate": 3.01502725751222e-06,
+      "loss": 0.5748,
+      "step": 9456
+    },
+    {
+      "epoch": 2.6196675900277007,
+      "grad_norm": 0.7458643913269043,
+      "learning_rate": 3.0146707812043657e-06,
+      "loss": 0.6489,
+      "step": 9457
+    },
+    {
+      "epoch": 2.61994459833795,
+      "grad_norm": 0.7419530153274536,
+      "learning_rate": 3.014314293968737e-06,
+      "loss": 0.6332,
+      "step": 9458
+    },
+    {
+      "epoch": 2.6202216066481996,
+      "grad_norm": 0.6879041194915771,
+      "learning_rate": 3.013957795812902e-06,
+      "loss": 0.5733,
+      "step": 9459
+    },
+    {
+      "epoch": 2.6204986149584486,
+      "grad_norm": 0.7298100590705872,
+      "learning_rate": 3.0136012867444298e-06,
+      "loss": 0.6445,
+      "step": 9460
+    },
+    {
+      "epoch": 2.620775623268698,
+      "grad_norm": 0.7113451957702637,
+      "learning_rate": 3.013244766770892e-06,
+      "loss": 0.581,
+      "step": 9461
+    },
+    {
+      "epoch": 2.6210526315789475,
+      "grad_norm": 0.7236753702163696,
+      "learning_rate": 3.0128882358998562e-06,
+      "loss": 0.5927,
+      "step": 9462
+    },
+    {
+      "epoch": 2.6213296398891965,
+      "grad_norm": 0.7375840544700623,
+      "learning_rate": 3.012531694138894e-06,
+      "loss": 0.6359,
+      "step": 9463
+    },
+    {
+      "epoch": 2.621606648199446,
+      "grad_norm": 0.7284663915634155,
+      "learning_rate": 3.012175141495574e-06,
+      "loss": 0.581,
+      "step": 9464
+    },
+    {
+      "epoch": 2.621883656509695,
+      "grad_norm": 0.7280388474464417,
+      "learning_rate": 3.0118185779774693e-06,
+      "loss": 0.6284,
+      "step": 9465
+    },
+    {
+      "epoch": 2.6221606648199445,
+      "grad_norm": 0.6828910708427429,
+      "learning_rate": 3.011462003592148e-06,
+      "loss": 0.5252,
+      "step": 9466
+    },
+    {
+      "epoch": 2.622437673130194,
+      "grad_norm": 0.7209635972976685,
+      "learning_rate": 3.0111054183471825e-06,
+      "loss": 0.6168,
+      "step": 9467
+    },
+    {
+      "epoch": 2.622714681440443,
+      "grad_norm": 0.7128496766090393,
+      "learning_rate": 3.010748822250144e-06,
+      "loss": 0.5724,
+      "step": 9468
+    },
+    {
+      "epoch": 2.6229916897506924,
+      "grad_norm": 0.7473496198654175,
+      "learning_rate": 3.010392215308604e-06,
+      "loss": 0.6042,
+      "step": 9469
+    },
+    {
+      "epoch": 2.623268698060942,
+      "grad_norm": 0.7299935817718506,
+      "learning_rate": 3.0100355975301328e-06,
+      "loss": 0.529,
+      "step": 9470
+    },
+    {
+      "epoch": 2.623545706371191,
+      "grad_norm": 0.7437604069709778,
+      "learning_rate": 3.009678968922304e-06,
+      "loss": 0.5975,
+      "step": 9471
+    },
+    {
+      "epoch": 2.6238227146814403,
+      "grad_norm": 0.7657999396324158,
+      "learning_rate": 3.009322329492689e-06,
+      "loss": 0.6125,
+      "step": 9472
+    },
+    {
+      "epoch": 2.62409972299169,
+      "grad_norm": 0.7673649191856384,
+      "learning_rate": 3.0089656792488604e-06,
+      "loss": 0.6358,
+      "step": 9473
+    },
+    {
+      "epoch": 2.624376731301939,
+      "grad_norm": 0.7584307789802551,
+      "learning_rate": 3.0086090181983914e-06,
+      "loss": 0.5712,
+      "step": 9474
+    },
+    {
+      "epoch": 2.6246537396121883,
+      "grad_norm": 0.7290375232696533,
+      "learning_rate": 3.008252346348853e-06,
+      "loss": 0.5979,
+      "step": 9475
+    },
+    {
+      "epoch": 2.6249307479224377,
+      "grad_norm": 0.7436773777008057,
+      "learning_rate": 3.0078956637078193e-06,
+      "loss": 0.6078,
+      "step": 9476
+    },
+    {
+      "epoch": 2.6252077562326868,
+      "grad_norm": 0.7523717880249023,
+      "learning_rate": 3.0075389702828647e-06,
+      "loss": 0.6113,
+      "step": 9477
+    },
+    {
+      "epoch": 2.6254847645429362,
+      "grad_norm": 0.722774863243103,
+      "learning_rate": 3.007182266081561e-06,
+      "loss": 0.6151,
+      "step": 9478
+    },
+    {
+      "epoch": 2.6257617728531857,
+      "grad_norm": 0.7217817902565002,
+      "learning_rate": 3.006825551111482e-06,
+      "loss": 0.5692,
+      "step": 9479
+    },
+    {
+      "epoch": 2.6260387811634347,
+      "grad_norm": 0.7214882373809814,
+      "learning_rate": 3.006468825380203e-06,
+      "loss": 0.5631,
+      "step": 9480
+    },
+    {
+      "epoch": 2.626315789473684,
+      "grad_norm": 0.7373204827308655,
+      "learning_rate": 3.006112088895297e-06,
+      "loss": 0.5715,
+      "step": 9481
+    },
+    {
+      "epoch": 2.6265927977839336,
+      "grad_norm": 0.7106059789657593,
+      "learning_rate": 3.005755341664339e-06,
+      "loss": 0.5559,
+      "step": 9482
+    },
+    {
+      "epoch": 2.6268698060941826,
+      "grad_norm": 0.7502145767211914,
+      "learning_rate": 3.005398583694904e-06,
+      "loss": 0.5861,
+      "step": 9483
+    },
+    {
+      "epoch": 2.627146814404432,
+      "grad_norm": 0.7245887517929077,
+      "learning_rate": 3.005041814994566e-06,
+      "loss": 0.5801,
+      "step": 9484
+    },
+    {
+      "epoch": 2.6274238227146816,
+      "grad_norm": 0.7328919172286987,
+      "learning_rate": 3.0046850355709005e-06,
+      "loss": 0.5702,
+      "step": 9485
+    },
+    {
+      "epoch": 2.6277008310249306,
+      "grad_norm": 0.7673099040985107,
+      "learning_rate": 3.0043282454314824e-06,
+      "loss": 0.5663,
+      "step": 9486
+    },
+    {
+      "epoch": 2.62797783933518,
+      "grad_norm": 0.7004749774932861,
+      "learning_rate": 3.0039714445838896e-06,
+      "loss": 0.5603,
+      "step": 9487
+    },
+    {
+      "epoch": 2.6282548476454295,
+      "grad_norm": 0.7447711229324341,
+      "learning_rate": 3.003614633035694e-06,
+      "loss": 0.6117,
+      "step": 9488
+    },
+    {
+      "epoch": 2.6285318559556785,
+      "grad_norm": 0.7462581992149353,
+      "learning_rate": 3.0032578107944747e-06,
+      "loss": 0.6043,
+      "step": 9489
+    },
+    {
+      "epoch": 2.628808864265928,
+      "grad_norm": 0.7413774132728577,
+      "learning_rate": 3.002900977867807e-06,
+      "loss": 0.5636,
+      "step": 9490
+    },
+    {
+      "epoch": 2.6290858725761774,
+      "grad_norm": 0.7752082347869873,
+      "learning_rate": 3.0025441342632666e-06,
+      "loss": 0.595,
+      "step": 9491
+    },
+    {
+      "epoch": 2.6293628808864264,
+      "grad_norm": 0.7633326053619385,
+      "learning_rate": 3.002187279988431e-06,
+      "loss": 0.5867,
+      "step": 9492
+    },
+    {
+      "epoch": 2.629639889196676,
+      "grad_norm": 0.714028000831604,
+      "learning_rate": 3.001830415050877e-06,
+      "loss": 0.5907,
+      "step": 9493
+    },
+    {
+      "epoch": 2.6299168975069254,
+      "grad_norm": 0.7516353726387024,
+      "learning_rate": 3.0014735394581824e-06,
+      "loss": 0.5741,
+      "step": 9494
+    },
+    {
+      "epoch": 2.6301939058171744,
+      "grad_norm": 0.7380667924880981,
+      "learning_rate": 3.0011166532179235e-06,
+      "loss": 0.6013,
+      "step": 9495
+    },
+    {
+      "epoch": 2.630470914127424,
+      "grad_norm": 0.7714182138442993,
+      "learning_rate": 3.0007597563376784e-06,
+      "loss": 0.6277,
+      "step": 9496
+    },
+    {
+      "epoch": 2.6307479224376733,
+      "grad_norm": 0.7297613024711609,
+      "learning_rate": 3.0004028488250248e-06,
+      "loss": 0.5951,
+      "step": 9497
+    },
+    {
+      "epoch": 2.6310249307479223,
+      "grad_norm": 0.6872463822364807,
+      "learning_rate": 3.0000459306875405e-06,
+      "loss": 0.5985,
+      "step": 9498
+    },
+    {
+      "epoch": 2.631301939058172,
+      "grad_norm": 0.7453889846801758,
+      "learning_rate": 2.9996890019328044e-06,
+      "loss": 0.6077,
+      "step": 9499
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.7232484817504883,
+      "learning_rate": 2.9993320625683954e-06,
+      "loss": 0.6089,
+      "step": 9500
+    },
+    {
+      "epoch": 2.6318559556786703,
+      "grad_norm": 0.7161245942115784,
+      "learning_rate": 2.99897511260189e-06,
+      "loss": 0.5707,
+      "step": 9501
+    },
+    {
+      "epoch": 2.6321329639889197,
+      "grad_norm": 0.7165181040763855,
+      "learning_rate": 2.9986181520408695e-06,
+      "loss": 0.6155,
+      "step": 9502
+    },
+    {
+      "epoch": 2.632409972299169,
+      "grad_norm": 0.7250982522964478,
+      "learning_rate": 2.998261180892913e-06,
+      "loss": 0.6123,
+      "step": 9503
+    },
+    {
+      "epoch": 2.632686980609418,
+      "grad_norm": 0.6966186761856079,
+      "learning_rate": 2.997904199165599e-06,
+      "loss": 0.5974,
+      "step": 9504
+    },
+    {
+      "epoch": 2.6329639889196677,
+      "grad_norm": 0.7354041934013367,
+      "learning_rate": 2.9975472068665063e-06,
+      "loss": 0.6189,
+      "step": 9505
+    },
+    {
+      "epoch": 2.633240997229917,
+      "grad_norm": 0.7224618792533875,
+      "learning_rate": 2.997190204003217e-06,
+      "loss": 0.5636,
+      "step": 9506
+    },
+    {
+      "epoch": 2.633518005540166,
+      "grad_norm": 0.7574952840805054,
+      "learning_rate": 2.9968331905833094e-06,
+      "loss": 0.5655,
+      "step": 9507
+    },
+    {
+      "epoch": 2.6337950138504156,
+      "grad_norm": 0.720216691493988,
+      "learning_rate": 2.9964761666143638e-06,
+      "loss": 0.5666,
+      "step": 9508
+    },
+    {
+      "epoch": 2.634072022160665,
+      "grad_norm": 0.7723554968833923,
+      "learning_rate": 2.9961191321039622e-06,
+      "loss": 0.6007,
+      "step": 9509
+    },
+    {
+      "epoch": 2.634349030470914,
+      "grad_norm": 0.709218442440033,
+      "learning_rate": 2.9957620870596845e-06,
+      "loss": 0.6237,
+      "step": 9510
+    },
+    {
+      "epoch": 2.6346260387811635,
+      "grad_norm": 0.7179205417633057,
+      "learning_rate": 2.995405031489111e-06,
+      "loss": 0.6205,
+      "step": 9511
+    },
+    {
+      "epoch": 2.634903047091413,
+      "grad_norm": 0.6865183711051941,
+      "learning_rate": 2.9950479653998245e-06,
+      "loss": 0.6088,
+      "step": 9512
+    },
+    {
+      "epoch": 2.635180055401662,
+      "grad_norm": 0.6995043158531189,
+      "learning_rate": 2.9946908887994046e-06,
+      "loss": 0.5564,
+      "step": 9513
+    },
+    {
+      "epoch": 2.6354570637119115,
+      "grad_norm": 0.7671514749526978,
+      "learning_rate": 2.994333801695434e-06,
+      "loss": 0.6239,
+      "step": 9514
+    },
+    {
+      "epoch": 2.635734072022161,
+      "grad_norm": 0.7148752808570862,
+      "learning_rate": 2.9939767040954943e-06,
+      "loss": 0.616,
+      "step": 9515
+    },
+    {
+      "epoch": 2.63601108033241,
+      "grad_norm": 0.7045048475265503,
+      "learning_rate": 2.993619596007168e-06,
+      "loss": 0.5905,
+      "step": 9516
+    },
+    {
+      "epoch": 2.6362880886426594,
+      "grad_norm": 0.7336992621421814,
+      "learning_rate": 2.9932624774380375e-06,
+      "loss": 0.61,
+      "step": 9517
+    },
+    {
+      "epoch": 2.636565096952909,
+      "grad_norm": 0.7365102767944336,
+      "learning_rate": 2.992905348395684e-06,
+      "loss": 0.6127,
+      "step": 9518
+    },
+    {
+      "epoch": 2.636842105263158,
+      "grad_norm": 0.7184882760047913,
+      "learning_rate": 2.9925482088876923e-06,
+      "loss": 0.5515,
+      "step": 9519
+    },
+    {
+      "epoch": 2.6371191135734073,
+      "grad_norm": 0.7389319539070129,
+      "learning_rate": 2.992191058921643e-06,
+      "loss": 0.6074,
+      "step": 9520
+    },
+    {
+      "epoch": 2.6373961218836564,
+      "grad_norm": 0.7419624924659729,
+      "learning_rate": 2.9918338985051216e-06,
+      "loss": 0.5779,
+      "step": 9521
+    },
+    {
+      "epoch": 2.637673130193906,
+      "grad_norm": 0.681654155254364,
+      "learning_rate": 2.99147672764571e-06,
+      "loss": 0.5927,
+      "step": 9522
+    },
+    {
+      "epoch": 2.6379501385041553,
+      "grad_norm": 0.7966417670249939,
+      "learning_rate": 2.9911195463509935e-06,
+      "loss": 0.6108,
+      "step": 9523
+    },
+    {
+      "epoch": 2.6382271468144043,
+      "grad_norm": 0.7548315525054932,
+      "learning_rate": 2.990762354628553e-06,
+      "loss": 0.568,
+      "step": 9524
+    },
+    {
+      "epoch": 2.6385041551246537,
+      "grad_norm": 0.7525621652603149,
+      "learning_rate": 2.990405152485976e-06,
+      "loss": 0.6276,
+      "step": 9525
+    },
+    {
+      "epoch": 2.638781163434903,
+      "grad_norm": 0.7998759150505066,
+      "learning_rate": 2.9900479399308444e-06,
+      "loss": 0.5655,
+      "step": 9526
+    },
+    {
+      "epoch": 2.6390581717451522,
+      "grad_norm": 0.7348849773406982,
+      "learning_rate": 2.9896907169707435e-06,
+      "loss": 0.604,
+      "step": 9527
+    },
+    {
+      "epoch": 2.6393351800554017,
+      "grad_norm": 0.7512754797935486,
+      "learning_rate": 2.9893334836132594e-06,
+      "loss": 0.5545,
+      "step": 9528
+    },
+    {
+      "epoch": 2.6396121883656507,
+      "grad_norm": 0.7238155007362366,
+      "learning_rate": 2.9889762398659743e-06,
+      "loss": 0.5913,
+      "step": 9529
+    },
+    {
+      "epoch": 2.6398891966759,
+      "grad_norm": 0.7376219630241394,
+      "learning_rate": 2.9886189857364755e-06,
+      "loss": 0.6672,
+      "step": 9530
+    },
+    {
+      "epoch": 2.6401662049861496,
+      "grad_norm": 0.767388641834259,
+      "learning_rate": 2.9882617212323485e-06,
+      "loss": 0.5675,
+      "step": 9531
+    },
+    {
+      "epoch": 2.6404432132963986,
+      "grad_norm": 0.7074344754219055,
+      "learning_rate": 2.9879044463611774e-06,
+      "loss": 0.5669,
+      "step": 9532
+    },
+    {
+      "epoch": 2.640720221606648,
+      "grad_norm": 0.6814509630203247,
+      "learning_rate": 2.9875471611305486e-06,
+      "loss": 0.5579,
+      "step": 9533
+    },
+    {
+      "epoch": 2.6409972299168976,
+      "grad_norm": 0.7133084535598755,
+      "learning_rate": 2.9871898655480493e-06,
+      "loss": 0.5783,
+      "step": 9534
+    },
+    {
+      "epoch": 2.6412742382271466,
+      "grad_norm": 0.7370125651359558,
+      "learning_rate": 2.9868325596212648e-06,
+      "loss": 0.584,
+      "step": 9535
+    },
+    {
+      "epoch": 2.641551246537396,
+      "grad_norm": 0.7439171671867371,
+      "learning_rate": 2.986475243357782e-06,
+      "loss": 0.5746,
+      "step": 9536
+    },
+    {
+      "epoch": 2.6418282548476455,
+      "grad_norm": 0.7310712933540344,
+      "learning_rate": 2.9861179167651866e-06,
+      "loss": 0.6293,
+      "step": 9537
+    },
+    {
+      "epoch": 2.6421052631578945,
+      "grad_norm": 0.7177421450614929,
+      "learning_rate": 2.985760579851068e-06,
+      "loss": 0.6244,
+      "step": 9538
+    },
+    {
+      "epoch": 2.642382271468144,
+      "grad_norm": 0.7210702896118164,
+      "learning_rate": 2.9854032326230102e-06,
+      "loss": 0.6147,
+      "step": 9539
+    },
+    {
+      "epoch": 2.6426592797783934,
+      "grad_norm": 0.7226083278656006,
+      "learning_rate": 2.9850458750886025e-06,
+      "loss": 0.5809,
+      "step": 9540
+    },
+    {
+      "epoch": 2.6429362880886424,
+      "grad_norm": 0.7365908026695251,
+      "learning_rate": 2.9846885072554333e-06,
+      "loss": 0.5798,
+      "step": 9541
+    },
+    {
+      "epoch": 2.643213296398892,
+      "grad_norm": 0.6953509449958801,
+      "learning_rate": 2.984331129131088e-06,
+      "loss": 0.5839,
+      "step": 9542
+    },
+    {
+      "epoch": 2.6434903047091414,
+      "grad_norm": 0.7429455518722534,
+      "learning_rate": 2.983973740723157e-06,
+      "loss": 0.5836,
+      "step": 9543
+    },
+    {
+      "epoch": 2.6437673130193904,
+      "grad_norm": 0.6981019377708435,
+      "learning_rate": 2.9836163420392273e-06,
+      "loss": 0.5913,
+      "step": 9544
+    },
+    {
+      "epoch": 2.64404432132964,
+      "grad_norm": 0.7662050724029541,
+      "learning_rate": 2.9832589330868877e-06,
+      "loss": 0.5759,
+      "step": 9545
+    },
+    {
+      "epoch": 2.6443213296398893,
+      "grad_norm": 0.6959735155105591,
+      "learning_rate": 2.982901513873726e-06,
+      "loss": 0.6031,
+      "step": 9546
+    },
+    {
+      "epoch": 2.6445983379501383,
+      "grad_norm": 0.7445115447044373,
+      "learning_rate": 2.982544084407333e-06,
+      "loss": 0.6363,
+      "step": 9547
+    },
+    {
+      "epoch": 2.644875346260388,
+      "grad_norm": 0.7081512212753296,
+      "learning_rate": 2.982186644695296e-06,
+      "loss": 0.5881,
+      "step": 9548
+    },
+    {
+      "epoch": 2.6451523545706372,
+      "grad_norm": 0.7388466000556946,
+      "learning_rate": 2.9818291947452062e-06,
+      "loss": 0.5806,
+      "step": 9549
+    },
+    {
+      "epoch": 2.6454293628808863,
+      "grad_norm": 0.6682338714599609,
+      "learning_rate": 2.9814717345646516e-06,
+      "loss": 0.5682,
+      "step": 9550
+    },
+    {
+      "epoch": 2.6457063711911357,
+      "grad_norm": 0.7200564742088318,
+      "learning_rate": 2.9811142641612223e-06,
+      "loss": 0.5747,
+      "step": 9551
+    },
+    {
+      "epoch": 2.645983379501385,
+      "grad_norm": 0.7652223110198975,
+      "learning_rate": 2.9807567835425085e-06,
+      "loss": 0.6063,
+      "step": 9552
+    },
+    {
+      "epoch": 2.646260387811634,
+      "grad_norm": 0.7285156846046448,
+      "learning_rate": 2.9803992927161e-06,
+      "loss": 0.5795,
+      "step": 9553
+    },
+    {
+      "epoch": 2.6465373961218837,
+      "grad_norm": 0.7372853755950928,
+      "learning_rate": 2.9800417916895885e-06,
+      "loss": 0.5766,
+      "step": 9554
+    },
+    {
+      "epoch": 2.646814404432133,
+      "grad_norm": 0.7120565176010132,
+      "learning_rate": 2.979684280470564e-06,
+      "loss": 0.5672,
+      "step": 9555
+    },
+    {
+      "epoch": 2.647091412742382,
+      "grad_norm": 0.7326462268829346,
+      "learning_rate": 2.9793267590666164e-06,
+      "loss": 0.5471,
+      "step": 9556
+    },
+    {
+      "epoch": 2.6473684210526316,
+      "grad_norm": 0.7396538853645325,
+      "learning_rate": 2.978969227485339e-06,
+      "loss": 0.5931,
+      "step": 9557
+    },
+    {
+      "epoch": 2.647645429362881,
+      "grad_norm": 0.7358024716377258,
+      "learning_rate": 2.9786116857343205e-06,
+      "loss": 0.618,
+      "step": 9558
+    },
+    {
+      "epoch": 2.64792243767313,
+      "grad_norm": 0.7099702954292297,
+      "learning_rate": 2.9782541338211534e-06,
+      "loss": 0.601,
+      "step": 9559
+    },
+    {
+      "epoch": 2.6481994459833795,
+      "grad_norm": 0.7943007349967957,
+      "learning_rate": 2.9778965717534314e-06,
+      "loss": 0.5557,
+      "step": 9560
+    },
+    {
+      "epoch": 2.648476454293629,
+      "grad_norm": 0.7590210437774658,
+      "learning_rate": 2.9775389995387434e-06,
+      "loss": 0.6195,
+      "step": 9561
+    },
+    {
+      "epoch": 2.648753462603878,
+      "grad_norm": 0.742610514163971,
+      "learning_rate": 2.9771814171846836e-06,
+      "loss": 0.5486,
+      "step": 9562
+    },
+    {
+      "epoch": 2.6490304709141275,
+      "grad_norm": 0.7486650347709656,
+      "learning_rate": 2.9768238246988433e-06,
+      "loss": 0.5742,
+      "step": 9563
+    },
+    {
+      "epoch": 2.649307479224377,
+      "grad_norm": 0.7409623861312866,
+      "learning_rate": 2.976466222088816e-06,
+      "loss": 0.5885,
+      "step": 9564
+    },
+    {
+      "epoch": 2.649584487534626,
+      "grad_norm": 0.7396969199180603,
+      "learning_rate": 2.9761086093621934e-06,
+      "loss": 0.6073,
+      "step": 9565
+    },
+    {
+      "epoch": 2.6498614958448754,
+      "grad_norm": 0.700952410697937,
+      "learning_rate": 2.9757509865265693e-06,
+      "loss": 0.588,
+      "step": 9566
+    },
+    {
+      "epoch": 2.650138504155125,
+      "grad_norm": 0.7877334356307983,
+      "learning_rate": 2.9753933535895373e-06,
+      "loss": 0.6178,
+      "step": 9567
+    },
+    {
+      "epoch": 2.650415512465374,
+      "grad_norm": 0.7633309960365295,
+      "learning_rate": 2.97503571055869e-06,
+      "loss": 0.6016,
+      "step": 9568
+    },
+    {
+      "epoch": 2.6506925207756233,
+      "grad_norm": 0.7346745729446411,
+      "learning_rate": 2.9746780574416214e-06,
+      "loss": 0.5777,
+      "step": 9569
+    },
+    {
+      "epoch": 2.650969529085873,
+      "grad_norm": 0.6854615807533264,
+      "learning_rate": 2.974320394245925e-06,
+      "loss": 0.6255,
+      "step": 9570
+    },
+    {
+      "epoch": 2.651246537396122,
+      "grad_norm": 0.743980348110199,
+      "learning_rate": 2.9739627209791965e-06,
+      "loss": 0.5989,
+      "step": 9571
+    },
+    {
+      "epoch": 2.6515235457063713,
+      "grad_norm": 0.6915601491928101,
+      "learning_rate": 2.973605037649029e-06,
+      "loss": 0.5674,
+      "step": 9572
+    },
+    {
+      "epoch": 2.6518005540166207,
+      "grad_norm": 0.7698721885681152,
+      "learning_rate": 2.973247344263016e-06,
+      "loss": 0.6151,
+      "step": 9573
+    },
+    {
+      "epoch": 2.6520775623268698,
+      "grad_norm": 0.7377426028251648,
+      "learning_rate": 2.972889640828754e-06,
+      "loss": 0.6186,
+      "step": 9574
+    },
+    {
+      "epoch": 2.652354570637119,
+      "grad_norm": 0.7789843082427979,
+      "learning_rate": 2.9725319273538372e-06,
+      "loss": 0.5894,
+      "step": 9575
+    },
+    {
+      "epoch": 2.6526315789473687,
+      "grad_norm": 0.7466744184494019,
+      "learning_rate": 2.9721742038458613e-06,
+      "loss": 0.6251,
+      "step": 9576
+    },
+    {
+      "epoch": 2.6529085872576177,
+      "grad_norm": 0.708580493927002,
+      "learning_rate": 2.971816470312421e-06,
+      "loss": 0.6142,
+      "step": 9577
+    },
+    {
+      "epoch": 2.653185595567867,
+      "grad_norm": 0.7652707695960999,
+      "learning_rate": 2.971458726761111e-06,
+      "loss": 0.611,
+      "step": 9578
+    },
+    {
+      "epoch": 2.6534626038781166,
+      "grad_norm": 0.7291722893714905,
+      "learning_rate": 2.97110097319953e-06,
+      "loss": 0.6017,
+      "step": 9579
+    },
+    {
+      "epoch": 2.6537396121883656,
+      "grad_norm": 0.7659996747970581,
+      "learning_rate": 2.970743209635271e-06,
+      "loss": 0.5777,
+      "step": 9580
+    },
+    {
+      "epoch": 2.654016620498615,
+      "grad_norm": 0.7659127116203308,
+      "learning_rate": 2.9703854360759328e-06,
+      "loss": 0.5903,
+      "step": 9581
+    },
+    {
+      "epoch": 2.654293628808864,
+      "grad_norm": 0.7328753471374512,
+      "learning_rate": 2.9700276525291096e-06,
+      "loss": 0.5977,
+      "step": 9582
+    },
+    {
+      "epoch": 2.6545706371191136,
+      "grad_norm": 0.7215525507926941,
+      "learning_rate": 2.9696698590023985e-06,
+      "loss": 0.5738,
+      "step": 9583
+    },
+    {
+      "epoch": 2.654847645429363,
+      "grad_norm": 0.7107459902763367,
+      "learning_rate": 2.969312055503398e-06,
+      "loss": 0.6168,
+      "step": 9584
+    },
+    {
+      "epoch": 2.655124653739612,
+      "grad_norm": 0.6944848299026489,
+      "learning_rate": 2.968954242039704e-06,
+      "loss": 0.591,
+      "step": 9585
+    },
+    {
+      "epoch": 2.6554016620498615,
+      "grad_norm": 0.7375234961509705,
+      "learning_rate": 2.968596418618913e-06,
+      "loss": 0.5599,
+      "step": 9586
+    },
+    {
+      "epoch": 2.655678670360111,
+      "grad_norm": 0.7887413501739502,
+      "learning_rate": 2.968238585248624e-06,
+      "loss": 0.5592,
+      "step": 9587
+    },
+    {
+      "epoch": 2.65595567867036,
+      "grad_norm": 0.7620272040367126,
+      "learning_rate": 2.9678807419364342e-06,
+      "loss": 0.6038,
+      "step": 9588
+    },
+    {
+      "epoch": 2.6562326869806094,
+      "grad_norm": 0.7225857377052307,
+      "learning_rate": 2.967522888689941e-06,
+      "loss": 0.6041,
+      "step": 9589
+    },
+    {
+      "epoch": 2.6565096952908585,
+      "grad_norm": 0.700484037399292,
+      "learning_rate": 2.9671650255167433e-06,
+      "loss": 0.5717,
+      "step": 9590
+    },
+    {
+      "epoch": 2.656786703601108,
+      "grad_norm": 0.7240809798240662,
+      "learning_rate": 2.9668071524244387e-06,
+      "loss": 0.5635,
+      "step": 9591
+    },
+    {
+      "epoch": 2.6570637119113574,
+      "grad_norm": 0.7192869782447815,
+      "learning_rate": 2.9664492694206272e-06,
+      "loss": 0.6109,
+      "step": 9592
+    },
+    {
+      "epoch": 2.6573407202216064,
+      "grad_norm": 0.7375945448875427,
+      "learning_rate": 2.966091376512905e-06,
+      "loss": 0.6293,
+      "step": 9593
+    },
+    {
+      "epoch": 2.657617728531856,
+      "grad_norm": 0.7899526953697205,
+      "learning_rate": 2.9657334737088734e-06,
+      "loss": 0.613,
+      "step": 9594
+    },
+    {
+      "epoch": 2.6578947368421053,
+      "grad_norm": 0.7093481421470642,
+      "learning_rate": 2.965375561016131e-06,
+      "loss": 0.562,
+      "step": 9595
+    },
+    {
+      "epoch": 2.6581717451523543,
+      "grad_norm": 0.7345967292785645,
+      "learning_rate": 2.9650176384422764e-06,
+      "loss": 0.5783,
+      "step": 9596
+    },
+    {
+      "epoch": 2.658448753462604,
+      "grad_norm": 0.7613131403923035,
+      "learning_rate": 2.96465970599491e-06,
+      "loss": 0.6016,
+      "step": 9597
+    },
+    {
+      "epoch": 2.6587257617728532,
+      "grad_norm": 0.7190057635307312,
+      "learning_rate": 2.964301763681632e-06,
+      "loss": 0.5851,
+      "step": 9598
+    },
+    {
+      "epoch": 2.6590027700831023,
+      "grad_norm": 0.7403254508972168,
+      "learning_rate": 2.9639438115100416e-06,
+      "loss": 0.577,
+      "step": 9599
+    },
+    {
+      "epoch": 2.6592797783933517,
+      "grad_norm": 0.7889108061790466,
+      "learning_rate": 2.9635858494877386e-06,
+      "loss": 0.6024,
+      "step": 9600
+    },
+    {
+      "epoch": 2.659556786703601,
+      "grad_norm": 0.750576913356781,
+      "learning_rate": 2.9632278776223253e-06,
+      "loss": 0.6004,
+      "step": 9601
+    },
+    {
+      "epoch": 2.65983379501385,
+      "grad_norm": 0.7538860440254211,
+      "learning_rate": 2.9628698959214e-06,
+      "loss": 0.5941,
+      "step": 9602
+    },
+    {
+      "epoch": 2.6601108033240997,
+      "grad_norm": 0.7576056122779846,
+      "learning_rate": 2.9625119043925655e-06,
+      "loss": 0.5905,
+      "step": 9603
+    },
+    {
+      "epoch": 2.660387811634349,
+      "grad_norm": 0.7413800954818726,
+      "learning_rate": 2.9621539030434223e-06,
+      "loss": 0.5935,
+      "step": 9604
+    },
+    {
+      "epoch": 2.660664819944598,
+      "grad_norm": 0.7557872533798218,
+      "learning_rate": 2.9617958918815713e-06,
+      "loss": 0.5932,
+      "step": 9605
+    },
+    {
+      "epoch": 2.6609418282548476,
+      "grad_norm": 0.7386396527290344,
+      "learning_rate": 2.9614378709146136e-06,
+      "loss": 0.5809,
+      "step": 9606
+    },
+    {
+      "epoch": 2.661218836565097,
+      "grad_norm": 0.7721570134162903,
+      "learning_rate": 2.9610798401501523e-06,
+      "loss": 0.6371,
+      "step": 9607
+    },
+    {
+      "epoch": 2.661495844875346,
+      "grad_norm": 0.7065140604972839,
+      "learning_rate": 2.960721799595788e-06,
+      "loss": 0.6045,
+      "step": 9608
+    },
+    {
+      "epoch": 2.6617728531855955,
+      "grad_norm": 0.7167521715164185,
+      "learning_rate": 2.960363749259124e-06,
+      "loss": 0.6045,
+      "step": 9609
+    },
+    {
+      "epoch": 2.662049861495845,
+      "grad_norm": 0.7369183301925659,
+      "learning_rate": 2.9600056891477613e-06,
+      "loss": 0.6044,
+      "step": 9610
+    },
+    {
+      "epoch": 2.662326869806094,
+      "grad_norm": 0.7165054082870483,
+      "learning_rate": 2.9596476192693036e-06,
+      "loss": 0.5877,
+      "step": 9611
+    },
+    {
+      "epoch": 2.6626038781163435,
+      "grad_norm": 0.733930230140686,
+      "learning_rate": 2.959289539631353e-06,
+      "loss": 0.5752,
+      "step": 9612
+    },
+    {
+      "epoch": 2.662880886426593,
+      "grad_norm": 0.7349205017089844,
+      "learning_rate": 2.9589314502415125e-06,
+      "loss": 0.5882,
+      "step": 9613
+    },
+    {
+      "epoch": 2.663157894736842,
+      "grad_norm": 0.7827674746513367,
+      "learning_rate": 2.9585733511073855e-06,
+      "loss": 0.5944,
+      "step": 9614
+    },
+    {
+      "epoch": 2.6634349030470914,
+      "grad_norm": 0.7822999954223633,
+      "learning_rate": 2.9582152422365745e-06,
+      "loss": 0.6061,
+      "step": 9615
+    },
+    {
+      "epoch": 2.663711911357341,
+      "grad_norm": 0.719627857208252,
+      "learning_rate": 2.9578571236366843e-06,
+      "loss": 0.6213,
+      "step": 9616
+    },
+    {
+      "epoch": 2.66398891966759,
+      "grad_norm": 0.7425439357757568,
+      "learning_rate": 2.957498995315318e-06,
+      "loss": 0.6105,
+      "step": 9617
+    },
+    {
+      "epoch": 2.6642659279778393,
+      "grad_norm": 0.7512136697769165,
+      "learning_rate": 2.9571408572800802e-06,
+      "loss": 0.5737,
+      "step": 9618
+    },
+    {
+      "epoch": 2.664542936288089,
+      "grad_norm": 0.7477335333824158,
+      "learning_rate": 2.9567827095385737e-06,
+      "loss": 0.6346,
+      "step": 9619
+    },
+    {
+      "epoch": 2.664819944598338,
+      "grad_norm": 0.7497116327285767,
+      "learning_rate": 2.956424552098405e-06,
+      "loss": 0.5831,
+      "step": 9620
+    },
+    {
+      "epoch": 2.6650969529085873,
+      "grad_norm": 0.7096310257911682,
+      "learning_rate": 2.956066384967176e-06,
+      "loss": 0.5804,
+      "step": 9621
+    },
+    {
+      "epoch": 2.6653739612188367,
+      "grad_norm": 0.7228270769119263,
+      "learning_rate": 2.955708208152494e-06,
+      "loss": 0.5566,
+      "step": 9622
+    },
+    {
+      "epoch": 2.6656509695290858,
+      "grad_norm": 0.7391560077667236,
+      "learning_rate": 2.9553500216619626e-06,
+      "loss": 0.5742,
+      "step": 9623
+    },
+    {
+      "epoch": 2.665927977839335,
+      "grad_norm": 0.7600433826446533,
+      "learning_rate": 2.954991825503188e-06,
+      "loss": 0.6008,
+      "step": 9624
+    },
+    {
+      "epoch": 2.6662049861495847,
+      "grad_norm": 0.7527140974998474,
+      "learning_rate": 2.9546336196837743e-06,
+      "loss": 0.5938,
+      "step": 9625
+    },
+    {
+      "epoch": 2.6664819944598337,
+      "grad_norm": 0.727566659450531,
+      "learning_rate": 2.954275404211328e-06,
+      "loss": 0.6062,
+      "step": 9626
+    },
+    {
+      "epoch": 2.666759002770083,
+      "grad_norm": 0.7642120122909546,
+      "learning_rate": 2.9539171790934552e-06,
+      "loss": 0.5261,
+      "step": 9627
+    },
+    {
+      "epoch": 2.6670360110803326,
+      "grad_norm": 0.7786373496055603,
+      "learning_rate": 2.9535589443377604e-06,
+      "loss": 0.6142,
+      "step": 9628
+    },
+    {
+      "epoch": 2.6673130193905816,
+      "grad_norm": 0.7548345327377319,
+      "learning_rate": 2.953200699951852e-06,
+      "loss": 0.6039,
+      "step": 9629
+    },
+    {
+      "epoch": 2.667590027700831,
+      "grad_norm": 0.7462404370307922,
+      "learning_rate": 2.952842445943335e-06,
+      "loss": 0.61,
+      "step": 9630
+    },
+    {
+      "epoch": 2.6678670360110806,
+      "grad_norm": 0.7438084483146667,
+      "learning_rate": 2.952484182319817e-06,
+      "loss": 0.6228,
+      "step": 9631
+    },
+    {
+      "epoch": 2.6681440443213296,
+      "grad_norm": 0.7126626968383789,
+      "learning_rate": 2.9521259090889032e-06,
+      "loss": 0.5762,
+      "step": 9632
+    },
+    {
+      "epoch": 2.668421052631579,
+      "grad_norm": 0.7101565003395081,
+      "learning_rate": 2.951767626258203e-06,
+      "loss": 0.5764,
+      "step": 9633
+    },
+    {
+      "epoch": 2.6686980609418285,
+      "grad_norm": 0.8123554587364197,
+      "learning_rate": 2.951409333835321e-06,
+      "loss": 0.5824,
+      "step": 9634
+    },
+    {
+      "epoch": 2.6689750692520775,
+      "grad_norm": 0.7582353353500366,
+      "learning_rate": 2.9510510318278675e-06,
+      "loss": 0.5428,
+      "step": 9635
+    },
+    {
+      "epoch": 2.669252077562327,
+      "grad_norm": 0.691589891910553,
+      "learning_rate": 2.950692720243448e-06,
+      "loss": 0.5842,
+      "step": 9636
+    },
+    {
+      "epoch": 2.6695290858725764,
+      "grad_norm": 0.7699835896492004,
+      "learning_rate": 2.9503343990896715e-06,
+      "loss": 0.599,
+      "step": 9637
+    },
+    {
+      "epoch": 2.6698060941828254,
+      "grad_norm": 0.753537118434906,
+      "learning_rate": 2.9499760683741448e-06,
+      "loss": 0.5947,
+      "step": 9638
+    },
+    {
+      "epoch": 2.670083102493075,
+      "grad_norm": 0.7336054444313049,
+      "learning_rate": 2.949617728104478e-06,
+      "loss": 0.5759,
+      "step": 9639
+    },
+    {
+      "epoch": 2.6703601108033244,
+      "grad_norm": 0.7584792375564575,
+      "learning_rate": 2.9492593782882785e-06,
+      "loss": 0.6344,
+      "step": 9640
+    },
+    {
+      "epoch": 2.6706371191135734,
+      "grad_norm": 0.7488454580307007,
+      "learning_rate": 2.948901018933154e-06,
+      "loss": 0.6385,
+      "step": 9641
+    },
+    {
+      "epoch": 2.670914127423823,
+      "grad_norm": 0.7217627167701721,
+      "learning_rate": 2.9485426500467157e-06,
+      "loss": 0.5351,
+      "step": 9642
+    },
+    {
+      "epoch": 2.6711911357340723,
+      "grad_norm": 0.716277003288269,
+      "learning_rate": 2.948184271636571e-06,
+      "loss": 0.5781,
+      "step": 9643
+    },
+    {
+      "epoch": 2.6714681440443213,
+      "grad_norm": 0.7645633220672607,
+      "learning_rate": 2.9478258837103302e-06,
+      "loss": 0.6118,
+      "step": 9644
+    },
+    {
+      "epoch": 2.6717451523545708,
+      "grad_norm": 0.7127770781517029,
+      "learning_rate": 2.9474674862756014e-06,
+      "loss": 0.6011,
+      "step": 9645
+    },
+    {
+      "epoch": 2.67202216066482,
+      "grad_norm": 0.7803115248680115,
+      "learning_rate": 2.9471090793399966e-06,
+      "loss": 0.6041,
+      "step": 9646
+    },
+    {
+      "epoch": 2.6722991689750693,
+      "grad_norm": 0.7439555525779724,
+      "learning_rate": 2.9467506629111227e-06,
+      "loss": 0.5994,
+      "step": 9647
+    },
+    {
+      "epoch": 2.6725761772853187,
+      "grad_norm": 0.7263895869255066,
+      "learning_rate": 2.946392236996592e-06,
+      "loss": 0.5833,
+      "step": 9648
+    },
+    {
+      "epoch": 2.6728531855955677,
+      "grad_norm": 0.7334993481636047,
+      "learning_rate": 2.946033801604014e-06,
+      "loss": 0.5808,
+      "step": 9649
+    },
+    {
+      "epoch": 2.673130193905817,
+      "grad_norm": 0.7821438312530518,
+      "learning_rate": 2.9456753567409997e-06,
+      "loss": 0.6225,
+      "step": 9650
+    },
+    {
+      "epoch": 2.6734072022160666,
+      "grad_norm": 0.7657595276832581,
+      "learning_rate": 2.945316902415159e-06,
+      "loss": 0.6224,
+      "step": 9651
+    },
+    {
+      "epoch": 2.6736842105263157,
+      "grad_norm": 0.7557787299156189,
+      "learning_rate": 2.9449584386341037e-06,
+      "loss": 0.6104,
+      "step": 9652
+    },
+    {
+      "epoch": 2.673961218836565,
+      "grad_norm": 0.7655667066574097,
+      "learning_rate": 2.944599965405444e-06,
+      "loss": 0.5885,
+      "step": 9653
+    },
+    {
+      "epoch": 2.674238227146814,
+      "grad_norm": 0.7425467371940613,
+      "learning_rate": 2.9442414827367916e-06,
+      "loss": 0.6347,
+      "step": 9654
+    },
+    {
+      "epoch": 2.6745152354570636,
+      "grad_norm": 0.7573320865631104,
+      "learning_rate": 2.943882990635759e-06,
+      "loss": 0.5746,
+      "step": 9655
+    },
+    {
+      "epoch": 2.674792243767313,
+      "grad_norm": 0.7560545802116394,
+      "learning_rate": 2.9435244891099555e-06,
+      "loss": 0.5833,
+      "step": 9656
+    },
+    {
+      "epoch": 2.675069252077562,
+      "grad_norm": 0.761762261390686,
+      "learning_rate": 2.943165978166995e-06,
+      "loss": 0.5782,
+      "step": 9657
+    },
+    {
+      "epoch": 2.6753462603878115,
+      "grad_norm": 0.7574995160102844,
+      "learning_rate": 2.94280745781449e-06,
+      "loss": 0.6128,
+      "step": 9658
+    },
+    {
+      "epoch": 2.675623268698061,
+      "grad_norm": 0.7315826416015625,
+      "learning_rate": 2.942448928060051e-06,
+      "loss": 0.5502,
+      "step": 9659
+    },
+    {
+      "epoch": 2.67590027700831,
+      "grad_norm": 0.7283254861831665,
+      "learning_rate": 2.942090388911291e-06,
+      "loss": 0.5925,
+      "step": 9660
+    },
+    {
+      "epoch": 2.6761772853185595,
+      "grad_norm": 0.7325502038002014,
+      "learning_rate": 2.941731840375824e-06,
+      "loss": 0.5873,
+      "step": 9661
+    },
+    {
+      "epoch": 2.676454293628809,
+      "grad_norm": 0.7269346714019775,
+      "learning_rate": 2.9413732824612617e-06,
+      "loss": 0.6004,
+      "step": 9662
+    },
+    {
+      "epoch": 2.676731301939058,
+      "grad_norm": 0.7535871863365173,
+      "learning_rate": 2.9410147151752172e-06,
+      "loss": 0.655,
+      "step": 9663
+    },
+    {
+      "epoch": 2.6770083102493074,
+      "grad_norm": 0.7521666884422302,
+      "learning_rate": 2.940656138525304e-06,
+      "loss": 0.6068,
+      "step": 9664
+    },
+    {
+      "epoch": 2.677285318559557,
+      "grad_norm": 0.7384536862373352,
+      "learning_rate": 2.9402975525191364e-06,
+      "loss": 0.6126,
+      "step": 9665
+    },
+    {
+      "epoch": 2.677562326869806,
+      "grad_norm": 0.6997430920600891,
+      "learning_rate": 2.9399389571643263e-06,
+      "loss": 0.5439,
+      "step": 9666
+    },
+    {
+      "epoch": 2.6778393351800553,
+      "grad_norm": 0.7200013399124146,
+      "learning_rate": 2.9395803524684885e-06,
+      "loss": 0.5884,
+      "step": 9667
+    },
+    {
+      "epoch": 2.678116343490305,
+      "grad_norm": 0.7792138457298279,
+      "learning_rate": 2.939221738439239e-06,
+      "loss": 0.5921,
+      "step": 9668
+    },
+    {
+      "epoch": 2.678393351800554,
+      "grad_norm": 0.7050516605377197,
+      "learning_rate": 2.9388631150841884e-06,
+      "loss": 0.59,
+      "step": 9669
+    },
+    {
+      "epoch": 2.6786703601108033,
+      "grad_norm": 0.7874724268913269,
+      "learning_rate": 2.9385044824109544e-06,
+      "loss": 0.615,
+      "step": 9670
+    },
+    {
+      "epoch": 2.6789473684210527,
+      "grad_norm": 0.7635147571563721,
+      "learning_rate": 2.93814584042715e-06,
+      "loss": 0.5562,
+      "step": 9671
+    },
+    {
+      "epoch": 2.6792243767313018,
+      "grad_norm": 0.7510080337524414,
+      "learning_rate": 2.9377871891403907e-06,
+      "loss": 0.6612,
+      "step": 9672
+    },
+    {
+      "epoch": 2.679501385041551,
+      "grad_norm": 0.7657120227813721,
+      "learning_rate": 2.937428528558291e-06,
+      "loss": 0.6431,
+      "step": 9673
+    },
+    {
+      "epoch": 2.6797783933518007,
+      "grad_norm": 0.7359679937362671,
+      "learning_rate": 2.9370698586884666e-06,
+      "loss": 0.5634,
+      "step": 9674
+    },
+    {
+      "epoch": 2.6800554016620497,
+      "grad_norm": 0.7201767563819885,
+      "learning_rate": 2.936711179538533e-06,
+      "loss": 0.5771,
+      "step": 9675
+    },
+    {
+      "epoch": 2.680332409972299,
+      "grad_norm": 0.7329818606376648,
+      "learning_rate": 2.9363524911161063e-06,
+      "loss": 0.6105,
+      "step": 9676
+    },
+    {
+      "epoch": 2.6806094182825486,
+      "grad_norm": 0.7106482982635498,
+      "learning_rate": 2.9359937934288013e-06,
+      "loss": 0.5773,
+      "step": 9677
+    },
+    {
+      "epoch": 2.6808864265927976,
+      "grad_norm": 0.705254852771759,
+      "learning_rate": 2.9356350864842355e-06,
+      "loss": 0.516,
+      "step": 9678
+    },
+    {
+      "epoch": 2.681163434903047,
+      "grad_norm": 0.7572067975997925,
+      "learning_rate": 2.9352763702900234e-06,
+      "loss": 0.6003,
+      "step": 9679
+    },
+    {
+      "epoch": 2.6814404432132966,
+      "grad_norm": 0.7427892088890076,
+      "learning_rate": 2.9349176448537826e-06,
+      "loss": 0.6282,
+      "step": 9680
+    },
+    {
+      "epoch": 2.6817174515235456,
+      "grad_norm": 0.7307045459747314,
+      "learning_rate": 2.9345589101831294e-06,
+      "loss": 0.5686,
+      "step": 9681
+    },
+    {
+      "epoch": 2.681994459833795,
+      "grad_norm": 0.7494450211524963,
+      "learning_rate": 2.934200166285681e-06,
+      "loss": 0.6109,
+      "step": 9682
+    },
+    {
+      "epoch": 2.6822714681440445,
+      "grad_norm": 0.7409537434577942,
+      "learning_rate": 2.9338414131690546e-06,
+      "loss": 0.6004,
+      "step": 9683
+    },
+    {
+      "epoch": 2.6825484764542935,
+      "grad_norm": 0.7529603242874146,
+      "learning_rate": 2.9334826508408665e-06,
+      "loss": 0.6336,
+      "step": 9684
+    },
+    {
+      "epoch": 2.682825484764543,
+      "grad_norm": 0.7535399198532104,
+      "learning_rate": 2.9331238793087353e-06,
+      "loss": 0.582,
+      "step": 9685
+    },
+    {
+      "epoch": 2.6831024930747924,
+      "grad_norm": 0.7496642470359802,
+      "learning_rate": 2.9327650985802776e-06,
+      "loss": 0.5911,
+      "step": 9686
+    },
+    {
+      "epoch": 2.6833795013850414,
+      "grad_norm": 0.7197352647781372,
+      "learning_rate": 2.932406308663112e-06,
+      "loss": 0.5887,
+      "step": 9687
+    },
+    {
+      "epoch": 2.683656509695291,
+      "grad_norm": 0.7457350492477417,
+      "learning_rate": 2.932047509564856e-06,
+      "loss": 0.5508,
+      "step": 9688
+    },
+    {
+      "epoch": 2.6839335180055404,
+      "grad_norm": 0.7327432632446289,
+      "learning_rate": 2.9316887012931283e-06,
+      "loss": 0.5757,
+      "step": 9689
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.6928367614746094,
+      "learning_rate": 2.931329883855547e-06,
+      "loss": 0.586,
+      "step": 9690
+    },
+    {
+      "epoch": 2.684487534626039,
+      "grad_norm": 0.7441751956939697,
+      "learning_rate": 2.9309710572597307e-06,
+      "loss": 0.5658,
+      "step": 9691
+    },
+    {
+      "epoch": 2.6847645429362883,
+      "grad_norm": 0.7539411783218384,
+      "learning_rate": 2.9306122215132974e-06,
+      "loss": 0.6482,
+      "step": 9692
+    },
+    {
+      "epoch": 2.6850415512465373,
+      "grad_norm": 0.7249196171760559,
+      "learning_rate": 2.930253376623868e-06,
+      "loss": 0.6555,
+      "step": 9693
+    },
+    {
+      "epoch": 2.6853185595567868,
+      "grad_norm": 0.7135176062583923,
+      "learning_rate": 2.9298945225990606e-06,
+      "loss": 0.5994,
+      "step": 9694
+    },
+    {
+      "epoch": 2.6855955678670362,
+      "grad_norm": 0.7672198414802551,
+      "learning_rate": 2.9295356594464943e-06,
+      "loss": 0.6298,
+      "step": 9695
+    },
+    {
+      "epoch": 2.6858725761772853,
+      "grad_norm": 0.746395468711853,
+      "learning_rate": 2.929176787173789e-06,
+      "loss": 0.6163,
+      "step": 9696
+    },
+    {
+      "epoch": 2.6861495844875347,
+      "grad_norm": 0.6857293844223022,
+      "learning_rate": 2.928817905788565e-06,
+      "loss": 0.5453,
+      "step": 9697
+    },
+    {
+      "epoch": 2.686426592797784,
+      "grad_norm": 0.7374250292778015,
+      "learning_rate": 2.9284590152984416e-06,
+      "loss": 0.5891,
+      "step": 9698
+    },
+    {
+      "epoch": 2.686703601108033,
+      "grad_norm": 0.7343308925628662,
+      "learning_rate": 2.928100115711039e-06,
+      "loss": 0.5837,
+      "step": 9699
+    },
+    {
+      "epoch": 2.6869806094182827,
+      "grad_norm": 0.7549000978469849,
+      "learning_rate": 2.9277412070339783e-06,
+      "loss": 0.5832,
+      "step": 9700
+    },
+    {
+      "epoch": 2.687257617728532,
+      "grad_norm": 0.7739076614379883,
+      "learning_rate": 2.927382289274878e-06,
+      "loss": 0.6404,
+      "step": 9701
+    },
+    {
+      "epoch": 2.687534626038781,
+      "grad_norm": 0.7524203062057495,
+      "learning_rate": 2.9270233624413617e-06,
+      "loss": 0.5872,
+      "step": 9702
+    },
+    {
+      "epoch": 2.6878116343490306,
+      "grad_norm": 0.7455158829689026,
+      "learning_rate": 2.926664426541049e-06,
+      "loss": 0.6154,
+      "step": 9703
+    },
+    {
+      "epoch": 2.68808864265928,
+      "grad_norm": 0.7223373055458069,
+      "learning_rate": 2.92630548158156e-06,
+      "loss": 0.6114,
+      "step": 9704
+    },
+    {
+      "epoch": 2.688365650969529,
+      "grad_norm": 0.738690972328186,
+      "learning_rate": 2.9259465275705166e-06,
+      "loss": 0.5711,
+      "step": 9705
+    },
+    {
+      "epoch": 2.6886426592797785,
+      "grad_norm": 0.7432278394699097,
+      "learning_rate": 2.9255875645155426e-06,
+      "loss": 0.5757,
+      "step": 9706
+    },
+    {
+      "epoch": 2.6889196675900275,
+      "grad_norm": 0.7743191123008728,
+      "learning_rate": 2.925228592424256e-06,
+      "loss": 0.6445,
+      "step": 9707
+    },
+    {
+      "epoch": 2.689196675900277,
+      "grad_norm": 0.7368574142456055,
+      "learning_rate": 2.9248696113042815e-06,
+      "loss": 0.5934,
+      "step": 9708
+    },
+    {
+      "epoch": 2.6894736842105265,
+      "grad_norm": 0.7550312876701355,
+      "learning_rate": 2.9245106211632398e-06,
+      "loss": 0.5848,
+      "step": 9709
+    },
+    {
+      "epoch": 2.6897506925207755,
+      "grad_norm": 0.7899466156959534,
+      "learning_rate": 2.9241516220087523e-06,
+      "loss": 0.5965,
+      "step": 9710
+    },
+    {
+      "epoch": 2.690027700831025,
+      "grad_norm": 0.7387313842773438,
+      "learning_rate": 2.923792613848444e-06,
+      "loss": 0.5383,
+      "step": 9711
+    },
+    {
+      "epoch": 2.6903047091412744,
+      "grad_norm": 0.7411960363388062,
+      "learning_rate": 2.9234335966899364e-06,
+      "loss": 0.5996,
+      "step": 9712
+    },
+    {
+      "epoch": 2.6905817174515234,
+      "grad_norm": 0.7664662003517151,
+      "learning_rate": 2.923074570540852e-06,
+      "loss": 0.6016,
+      "step": 9713
+    },
+    {
+      "epoch": 2.690858725761773,
+      "grad_norm": 0.7731584906578064,
+      "learning_rate": 2.9227155354088134e-06,
+      "loss": 0.616,
+      "step": 9714
+    },
+    {
+      "epoch": 2.691135734072022,
+      "grad_norm": 0.7891464829444885,
+      "learning_rate": 2.922356491301445e-06,
+      "loss": 0.6266,
+      "step": 9715
+    },
+    {
+      "epoch": 2.6914127423822714,
+      "grad_norm": 0.7034754157066345,
+      "learning_rate": 2.92199743822637e-06,
+      "loss": 0.5748,
+      "step": 9716
+    },
+    {
+      "epoch": 2.691689750692521,
+      "grad_norm": 0.716183066368103,
+      "learning_rate": 2.9216383761912113e-06,
+      "loss": 0.559,
+      "step": 9717
+    },
+    {
+      "epoch": 2.69196675900277,
+      "grad_norm": 0.7393279075622559,
+      "learning_rate": 2.9212793052035934e-06,
+      "loss": 0.5972,
+      "step": 9718
+    },
+    {
+      "epoch": 2.6922437673130193,
+      "grad_norm": 0.7390798330307007,
+      "learning_rate": 2.9209202252711406e-06,
+      "loss": 0.6116,
+      "step": 9719
+    },
+    {
+      "epoch": 2.6925207756232687,
+      "grad_norm": 0.7001668214797974,
+      "learning_rate": 2.920561136401475e-06,
+      "loss": 0.5541,
+      "step": 9720
+    },
+    {
+      "epoch": 2.6927977839335178,
+      "grad_norm": 0.7521222829818726,
+      "learning_rate": 2.9202020386022235e-06,
+      "loss": 0.5988,
+      "step": 9721
+    },
+    {
+      "epoch": 2.6930747922437672,
+      "grad_norm": 0.7328142523765564,
+      "learning_rate": 2.9198429318810095e-06,
+      "loss": 0.63,
+      "step": 9722
+    },
+    {
+      "epoch": 2.6933518005540167,
+      "grad_norm": 0.7463355660438538,
+      "learning_rate": 2.9194838162454573e-06,
+      "loss": 0.61,
+      "step": 9723
+    },
+    {
+      "epoch": 2.6936288088642657,
+      "grad_norm": 0.7568442225456238,
+      "learning_rate": 2.9191246917031936e-06,
+      "loss": 0.5795,
+      "step": 9724
+    },
+    {
+      "epoch": 2.693905817174515,
+      "grad_norm": 0.7941131591796875,
+      "learning_rate": 2.9187655582618413e-06,
+      "loss": 0.6286,
+      "step": 9725
+    },
+    {
+      "epoch": 2.6941828254847646,
+      "grad_norm": 0.7539636492729187,
+      "learning_rate": 2.918406415929028e-06,
+      "loss": 0.5493,
+      "step": 9726
+    },
+    {
+      "epoch": 2.6944598337950136,
+      "grad_norm": 0.7329300045967102,
+      "learning_rate": 2.9180472647123765e-06,
+      "loss": 0.6389,
+      "step": 9727
+    },
+    {
+      "epoch": 2.694736842105263,
+      "grad_norm": 0.7334020733833313,
+      "learning_rate": 2.9176881046195145e-06,
+      "loss": 0.6247,
+      "step": 9728
+    },
+    {
+      "epoch": 2.6950138504155126,
+      "grad_norm": 0.7126622796058655,
+      "learning_rate": 2.9173289356580676e-06,
+      "loss": 0.5517,
+      "step": 9729
+    },
+    {
+      "epoch": 2.6952908587257616,
+      "grad_norm": 0.7421119213104248,
+      "learning_rate": 2.916969757835662e-06,
+      "loss": 0.5809,
+      "step": 9730
+    },
+    {
+      "epoch": 2.695567867036011,
+      "grad_norm": 0.6701324582099915,
+      "learning_rate": 2.9166105711599237e-06,
+      "loss": 0.5741,
+      "step": 9731
+    },
+    {
+      "epoch": 2.6958448753462605,
+      "grad_norm": 0.7376656532287598,
+      "learning_rate": 2.916251375638478e-06,
+      "loss": 0.5799,
+      "step": 9732
+    },
+    {
+      "epoch": 2.6961218836565095,
+      "grad_norm": 0.7610093355178833,
+      "learning_rate": 2.915892171278953e-06,
+      "loss": 0.5926,
+      "step": 9733
+    },
+    {
+      "epoch": 2.696398891966759,
+      "grad_norm": 0.7553840279579163,
+      "learning_rate": 2.915532958088976e-06,
+      "loss": 0.5387,
+      "step": 9734
+    },
+    {
+      "epoch": 2.6966759002770084,
+      "grad_norm": 0.7403333783149719,
+      "learning_rate": 2.915173736076172e-06,
+      "loss": 0.5533,
+      "step": 9735
+    },
+    {
+      "epoch": 2.6969529085872574,
+      "grad_norm": 0.7409501671791077,
+      "learning_rate": 2.91481450524817e-06,
+      "loss": 0.6253,
+      "step": 9736
+    },
+    {
+      "epoch": 2.697229916897507,
+      "grad_norm": 0.7569580078125,
+      "learning_rate": 2.9144552656125963e-06,
+      "loss": 0.5649,
+      "step": 9737
+    },
+    {
+      "epoch": 2.6975069252077564,
+      "grad_norm": 0.7319048047065735,
+      "learning_rate": 2.9140960171770795e-06,
+      "loss": 0.6189,
+      "step": 9738
+    },
+    {
+      "epoch": 2.6977839335180054,
+      "grad_norm": 0.7332355976104736,
+      "learning_rate": 2.9137367599492463e-06,
+      "loss": 0.54,
+      "step": 9739
+    },
+    {
+      "epoch": 2.698060941828255,
+      "grad_norm": 0.7282864451408386,
+      "learning_rate": 2.9133774939367248e-06,
+      "loss": 0.5977,
+      "step": 9740
+    },
+    {
+      "epoch": 2.6983379501385043,
+      "grad_norm": 0.7558872699737549,
+      "learning_rate": 2.9130182191471445e-06,
+      "loss": 0.6258,
+      "step": 9741
+    },
+    {
+      "epoch": 2.6986149584487533,
+      "grad_norm": 0.7340047359466553,
+      "learning_rate": 2.9126589355881317e-06,
+      "loss": 0.6142,
+      "step": 9742
+    },
+    {
+      "epoch": 2.698891966759003,
+      "grad_norm": 0.7753874659538269,
+      "learning_rate": 2.9122996432673165e-06,
+      "loss": 0.5904,
+      "step": 9743
+    },
+    {
+      "epoch": 2.6991689750692522,
+      "grad_norm": 0.6965166926383972,
+      "learning_rate": 2.911940342192327e-06,
+      "loss": 0.5466,
+      "step": 9744
+    },
+    {
+      "epoch": 2.6994459833795013,
+      "grad_norm": 0.7482168674468994,
+      "learning_rate": 2.9115810323707917e-06,
+      "loss": 0.5935,
+      "step": 9745
+    },
+    {
+      "epoch": 2.6997229916897507,
+      "grad_norm": 0.6827434301376343,
+      "learning_rate": 2.91122171381034e-06,
+      "loss": 0.5701,
+      "step": 9746
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.6957672834396362,
+      "learning_rate": 2.9108623865186015e-06,
+      "loss": 0.5836,
+      "step": 9747
+    },
+    {
+      "epoch": 2.700277008310249,
+      "grad_norm": 0.7049537897109985,
+      "learning_rate": 2.910503050503205e-06,
+      "loss": 0.5234,
+      "step": 9748
+    },
+    {
+      "epoch": 2.7005540166204987,
+      "grad_norm": 0.7790834307670593,
+      "learning_rate": 2.9101437057717808e-06,
+      "loss": 0.6205,
+      "step": 9749
+    },
+    {
+      "epoch": 2.700831024930748,
+      "grad_norm": 0.7092089653015137,
+      "learning_rate": 2.9097843523319574e-06,
+      "loss": 0.5931,
+      "step": 9750
+    },
+    {
+      "epoch": 2.701108033240997,
+      "grad_norm": 0.8164337277412415,
+      "learning_rate": 2.9094249901913666e-06,
+      "loss": 0.6217,
+      "step": 9751
+    },
+    {
+      "epoch": 2.7013850415512466,
+      "grad_norm": 0.7725958228111267,
+      "learning_rate": 2.9090656193576374e-06,
+      "loss": 0.5835,
+      "step": 9752
+    },
+    {
+      "epoch": 2.701662049861496,
+      "grad_norm": 0.7309401035308838,
+      "learning_rate": 2.9087062398384005e-06,
+      "loss": 0.5764,
+      "step": 9753
+    },
+    {
+      "epoch": 2.701939058171745,
+      "grad_norm": 0.7685689330101013,
+      "learning_rate": 2.9083468516412865e-06,
+      "loss": 0.5686,
+      "step": 9754
+    },
+    {
+      "epoch": 2.7022160664819945,
+      "grad_norm": 0.7559168338775635,
+      "learning_rate": 2.9079874547739255e-06,
+      "loss": 0.6157,
+      "step": 9755
+    },
+    {
+      "epoch": 2.702493074792244,
+      "grad_norm": 0.7864164113998413,
+      "learning_rate": 2.90762804924395e-06,
+      "loss": 0.6118,
+      "step": 9756
+    },
+    {
+      "epoch": 2.702770083102493,
+      "grad_norm": 0.7336371541023254,
+      "learning_rate": 2.9072686350589894e-06,
+      "loss": 0.5844,
+      "step": 9757
+    },
+    {
+      "epoch": 2.7030470914127425,
+      "grad_norm": 0.7496275901794434,
+      "learning_rate": 2.9069092122266758e-06,
+      "loss": 0.6102,
+      "step": 9758
+    },
+    {
+      "epoch": 2.703324099722992,
+      "grad_norm": 0.7452666163444519,
+      "learning_rate": 2.90654978075464e-06,
+      "loss": 0.6431,
+      "step": 9759
+    },
+    {
+      "epoch": 2.703601108033241,
+      "grad_norm": 0.777228593826294,
+      "learning_rate": 2.9061903406505153e-06,
+      "loss": 0.5739,
+      "step": 9760
+    },
+    {
+      "epoch": 2.7038781163434904,
+      "grad_norm": 0.7218241095542908,
+      "learning_rate": 2.9058308919219315e-06,
+      "loss": 0.5931,
+      "step": 9761
+    },
+    {
+      "epoch": 2.70415512465374,
+      "grad_norm": 0.7450538873672485,
+      "learning_rate": 2.905471434576521e-06,
+      "loss": 0.6052,
+      "step": 9762
+    },
+    {
+      "epoch": 2.704432132963989,
+      "grad_norm": 0.7024423480033875,
+      "learning_rate": 2.9051119686219176e-06,
+      "loss": 0.5941,
+      "step": 9763
+    },
+    {
+      "epoch": 2.7047091412742383,
+      "grad_norm": 0.736329972743988,
+      "learning_rate": 2.904752494065752e-06,
+      "loss": 0.5991,
+      "step": 9764
+    },
+    {
+      "epoch": 2.704986149584488,
+      "grad_norm": 0.6893607378005981,
+      "learning_rate": 2.904393010915657e-06,
+      "loss": 0.5577,
+      "step": 9765
+    },
+    {
+      "epoch": 2.705263157894737,
+      "grad_norm": 0.7273480892181396,
+      "learning_rate": 2.9040335191792663e-06,
+      "loss": 0.5537,
+      "step": 9766
+    },
+    {
+      "epoch": 2.7055401662049863,
+      "grad_norm": 0.6868672370910645,
+      "learning_rate": 2.9036740188642123e-06,
+      "loss": 0.5484,
+      "step": 9767
+    },
+    {
+      "epoch": 2.7058171745152357,
+      "grad_norm": 0.7351400852203369,
+      "learning_rate": 2.9033145099781268e-06,
+      "loss": 0.5935,
+      "step": 9768
+    },
+    {
+      "epoch": 2.7060941828254848,
+      "grad_norm": 0.7292349934577942,
+      "learning_rate": 2.9029549925286453e-06,
+      "loss": 0.5882,
+      "step": 9769
+    },
+    {
+      "epoch": 2.706371191135734,
+      "grad_norm": 0.7196292877197266,
+      "learning_rate": 2.9025954665234e-06,
+      "loss": 0.6021,
+      "step": 9770
+    },
+    {
+      "epoch": 2.7066481994459832,
+      "grad_norm": 0.751592755317688,
+      "learning_rate": 2.9022359319700254e-06,
+      "loss": 0.5931,
+      "step": 9771
+    },
+    {
+      "epoch": 2.7069252077562327,
+      "grad_norm": 0.7287519574165344,
+      "learning_rate": 2.9018763888761535e-06,
+      "loss": 0.5459,
+      "step": 9772
+    },
+    {
+      "epoch": 2.707202216066482,
+      "grad_norm": 0.7140438556671143,
+      "learning_rate": 2.901516837249421e-06,
+      "loss": 0.5956,
+      "step": 9773
+    },
+    {
+      "epoch": 2.707479224376731,
+      "grad_norm": 0.7360804080963135,
+      "learning_rate": 2.90115727709746e-06,
+      "loss": 0.6057,
+      "step": 9774
+    },
+    {
+      "epoch": 2.7077562326869806,
+      "grad_norm": 0.7283020615577698,
+      "learning_rate": 2.9007977084279058e-06,
+      "loss": 0.5403,
+      "step": 9775
+    },
+    {
+      "epoch": 2.7080332409972296,
+      "grad_norm": 0.7723913192749023,
+      "learning_rate": 2.9004381312483926e-06,
+      "loss": 0.5868,
+      "step": 9776
+    },
+    {
+      "epoch": 2.708310249307479,
+      "grad_norm": 0.7189229726791382,
+      "learning_rate": 2.9000785455665553e-06,
+      "loss": 0.6085,
+      "step": 9777
+    },
+    {
+      "epoch": 2.7085872576177286,
+      "grad_norm": 0.7812227010726929,
+      "learning_rate": 2.899718951390029e-06,
+      "loss": 0.5868,
+      "step": 9778
+    },
+    {
+      "epoch": 2.7088642659279776,
+      "grad_norm": 0.7589960098266602,
+      "learning_rate": 2.899359348726448e-06,
+      "loss": 0.6533,
+      "step": 9779
+    },
+    {
+      "epoch": 2.709141274238227,
+      "grad_norm": 0.7362611889839172,
+      "learning_rate": 2.8989997375834485e-06,
+      "loss": 0.576,
+      "step": 9780
+    },
+    {
+      "epoch": 2.7094182825484765,
+      "grad_norm": 0.7336854934692383,
+      "learning_rate": 2.898640117968665e-06,
+      "loss": 0.5834,
+      "step": 9781
+    },
+    {
+      "epoch": 2.7096952908587255,
+      "grad_norm": 0.7211445569992065,
+      "learning_rate": 2.898280489889735e-06,
+      "loss": 0.5809,
+      "step": 9782
+    },
+    {
+      "epoch": 2.709972299168975,
+      "grad_norm": 0.7566702365875244,
+      "learning_rate": 2.8979208533542923e-06,
+      "loss": 0.6097,
+      "step": 9783
+    },
+    {
+      "epoch": 2.7102493074792244,
+      "grad_norm": 0.7137244343757629,
+      "learning_rate": 2.8975612083699737e-06,
+      "loss": 0.6096,
+      "step": 9784
+    },
+    {
+      "epoch": 2.7105263157894735,
+      "grad_norm": 0.7393102049827576,
+      "learning_rate": 2.8972015549444162e-06,
+      "loss": 0.6004,
+      "step": 9785
+    },
+    {
+      "epoch": 2.710803324099723,
+      "grad_norm": 0.7391594648361206,
+      "learning_rate": 2.896841893085255e-06,
+      "loss": 0.5695,
+      "step": 9786
+    },
+    {
+      "epoch": 2.7110803324099724,
+      "grad_norm": 0.780170738697052,
+      "learning_rate": 2.896482222800126e-06,
+      "loss": 0.5785,
+      "step": 9787
+    },
+    {
+      "epoch": 2.7113573407202214,
+      "grad_norm": 0.7280192375183105,
+      "learning_rate": 2.8961225440966678e-06,
+      "loss": 0.5825,
+      "step": 9788
+    },
+    {
+      "epoch": 2.711634349030471,
+      "grad_norm": 0.7467705011367798,
+      "learning_rate": 2.8957628569825165e-06,
+      "loss": 0.6411,
+      "step": 9789
+    },
+    {
+      "epoch": 2.7119113573407203,
+      "grad_norm": 0.7193437814712524,
+      "learning_rate": 2.895403161465309e-06,
+      "loss": 0.5487,
+      "step": 9790
+    },
+    {
+      "epoch": 2.7121883656509693,
+      "grad_norm": 0.7068139910697937,
+      "learning_rate": 2.895043457552682e-06,
+      "loss": 0.5979,
+      "step": 9791
+    },
+    {
+      "epoch": 2.712465373961219,
+      "grad_norm": 0.7164737582206726,
+      "learning_rate": 2.8946837452522746e-06,
+      "loss": 0.5583,
+      "step": 9792
+    },
+    {
+      "epoch": 2.7127423822714682,
+      "grad_norm": 0.838596761226654,
+      "learning_rate": 2.8943240245717225e-06,
+      "loss": 0.6158,
+      "step": 9793
+    },
+    {
+      "epoch": 2.7130193905817173,
+      "grad_norm": 0.7566062808036804,
+      "learning_rate": 2.893964295518664e-06,
+      "loss": 0.5801,
+      "step": 9794
+    },
+    {
+      "epoch": 2.7132963988919667,
+      "grad_norm": 0.7423161864280701,
+      "learning_rate": 2.893604558100739e-06,
+      "loss": 0.6034,
+      "step": 9795
+    },
+    {
+      "epoch": 2.713573407202216,
+      "grad_norm": 0.7267448306083679,
+      "learning_rate": 2.8932448123255825e-06,
+      "loss": 0.6231,
+      "step": 9796
+    },
+    {
+      "epoch": 2.713850415512465,
+      "grad_norm": 0.7339398860931396,
+      "learning_rate": 2.892885058200835e-06,
+      "loss": 0.5997,
+      "step": 9797
+    },
+    {
+      "epoch": 2.7141274238227147,
+      "grad_norm": 0.7441668510437012,
+      "learning_rate": 2.8925252957341347e-06,
+      "loss": 0.6246,
+      "step": 9798
+    },
+    {
+      "epoch": 2.714404432132964,
+      "grad_norm": 0.7524739503860474,
+      "learning_rate": 2.8921655249331198e-06,
+      "loss": 0.6254,
+      "step": 9799
+    },
+    {
+      "epoch": 2.714681440443213,
+      "grad_norm": 0.7392253875732422,
+      "learning_rate": 2.8918057458054288e-06,
+      "loss": 0.5989,
+      "step": 9800
+    },
+    {
+      "epoch": 2.7149584487534626,
+      "grad_norm": 0.746084988117218,
+      "learning_rate": 2.891445958358702e-06,
+      "loss": 0.6185,
+      "step": 9801
+    },
+    {
+      "epoch": 2.715235457063712,
+      "grad_norm": 0.7321019768714905,
+      "learning_rate": 2.8910861626005774e-06,
+      "loss": 0.5733,
+      "step": 9802
+    },
+    {
+      "epoch": 2.715512465373961,
+      "grad_norm": 0.7405261397361755,
+      "learning_rate": 2.8907263585386954e-06,
+      "loss": 0.5753,
+      "step": 9803
+    },
+    {
+      "epoch": 2.7157894736842105,
+      "grad_norm": 0.7382570505142212,
+      "learning_rate": 2.890366546180694e-06,
+      "loss": 0.6047,
+      "step": 9804
+    },
+    {
+      "epoch": 2.71606648199446,
+      "grad_norm": 0.7403123378753662,
+      "learning_rate": 2.8900067255342152e-06,
+      "loss": 0.5944,
+      "step": 9805
+    },
+    {
+      "epoch": 2.716343490304709,
+      "grad_norm": 0.7846508622169495,
+      "learning_rate": 2.889646896606897e-06,
+      "loss": 0.6135,
+      "step": 9806
+    },
+    {
+      "epoch": 2.7166204986149585,
+      "grad_norm": 0.7530471682548523,
+      "learning_rate": 2.88928705940638e-06,
+      "loss": 0.5899,
+      "step": 9807
+    },
+    {
+      "epoch": 2.716897506925208,
+      "grad_norm": 0.6979814767837524,
+      "learning_rate": 2.8889272139403047e-06,
+      "loss": 0.6406,
+      "step": 9808
+    },
+    {
+      "epoch": 2.717174515235457,
+      "grad_norm": 0.7776252627372742,
+      "learning_rate": 2.888567360216311e-06,
+      "loss": 0.6207,
+      "step": 9809
+    },
+    {
+      "epoch": 2.7174515235457064,
+      "grad_norm": 0.7415891289710999,
+      "learning_rate": 2.8882074982420405e-06,
+      "loss": 0.5763,
+      "step": 9810
+    },
+    {
+      "epoch": 2.717728531855956,
+      "grad_norm": 0.7440627217292786,
+      "learning_rate": 2.8878476280251333e-06,
+      "loss": 0.5803,
+      "step": 9811
+    },
+    {
+      "epoch": 2.718005540166205,
+      "grad_norm": 0.744896411895752,
+      "learning_rate": 2.8874877495732306e-06,
+      "loss": 0.5769,
+      "step": 9812
+    },
+    {
+      "epoch": 2.7182825484764543,
+      "grad_norm": 0.7108137607574463,
+      "learning_rate": 2.887127862893973e-06,
+      "loss": 0.5794,
+      "step": 9813
+    },
+    {
+      "epoch": 2.718559556786704,
+      "grad_norm": 0.7342739701271057,
+      "learning_rate": 2.8867679679950032e-06,
+      "loss": 0.5807,
+      "step": 9814
+    },
+    {
+      "epoch": 2.718836565096953,
+      "grad_norm": 0.7428385615348816,
+      "learning_rate": 2.886408064883961e-06,
+      "loss": 0.5883,
+      "step": 9815
+    },
+    {
+      "epoch": 2.7191135734072023,
+      "grad_norm": 0.7807708978652954,
+      "learning_rate": 2.886048153568489e-06,
+      "loss": 0.6403,
+      "step": 9816
+    },
+    {
+      "epoch": 2.7193905817174517,
+      "grad_norm": 0.7568454146385193,
+      "learning_rate": 2.885688234056229e-06,
+      "loss": 0.657,
+      "step": 9817
+    },
+    {
+      "epoch": 2.7196675900277008,
+      "grad_norm": 0.7030062079429626,
+      "learning_rate": 2.885328306354823e-06,
+      "loss": 0.6298,
+      "step": 9818
+    },
+    {
+      "epoch": 2.71994459833795,
+      "grad_norm": 0.7581620216369629,
+      "learning_rate": 2.8849683704719123e-06,
+      "loss": 0.5754,
+      "step": 9819
+    },
+    {
+      "epoch": 2.7202216066481997,
+      "grad_norm": 0.727524995803833,
+      "learning_rate": 2.8846084264151403e-06,
+      "loss": 0.6101,
+      "step": 9820
+    },
+    {
+      "epoch": 2.7204986149584487,
+      "grad_norm": 0.7349806427955627,
+      "learning_rate": 2.8842484741921496e-06,
+      "loss": 0.6251,
+      "step": 9821
+    },
+    {
+      "epoch": 2.720775623268698,
+      "grad_norm": 0.754910945892334,
+      "learning_rate": 2.883888513810582e-06,
+      "loss": 0.5552,
+      "step": 9822
+    },
+    {
+      "epoch": 2.7210526315789476,
+      "grad_norm": 0.7700474858283997,
+      "learning_rate": 2.883528545278081e-06,
+      "loss": 0.615,
+      "step": 9823
+    },
+    {
+      "epoch": 2.7213296398891966,
+      "grad_norm": 0.7219333648681641,
+      "learning_rate": 2.8831685686022897e-06,
+      "loss": 0.598,
+      "step": 9824
+    },
+    {
+      "epoch": 2.721606648199446,
+      "grad_norm": 0.7458710074424744,
+      "learning_rate": 2.8828085837908514e-06,
+      "loss": 0.6061,
+      "step": 9825
+    },
+    {
+      "epoch": 2.7218836565096955,
+      "grad_norm": 0.7252250909805298,
+      "learning_rate": 2.8824485908514087e-06,
+      "loss": 0.5542,
+      "step": 9826
+    },
+    {
+      "epoch": 2.7221606648199446,
+      "grad_norm": 0.7319718599319458,
+      "learning_rate": 2.882088589791607e-06,
+      "loss": 0.5981,
+      "step": 9827
+    },
+    {
+      "epoch": 2.722437673130194,
+      "grad_norm": 0.7807872295379639,
+      "learning_rate": 2.8817285806190877e-06,
+      "loss": 0.6203,
+      "step": 9828
+    },
+    {
+      "epoch": 2.7227146814404435,
+      "grad_norm": 0.7514304518699646,
+      "learning_rate": 2.881368563341496e-06,
+      "loss": 0.5926,
+      "step": 9829
+    },
+    {
+      "epoch": 2.7229916897506925,
+      "grad_norm": 0.7369164824485779,
+      "learning_rate": 2.881008537966476e-06,
+      "loss": 0.601,
+      "step": 9830
+    },
+    {
+      "epoch": 2.723268698060942,
+      "grad_norm": 0.7181528210639954,
+      "learning_rate": 2.880648504501671e-06,
+      "loss": 0.5683,
+      "step": 9831
+    },
+    {
+      "epoch": 2.723545706371191,
+      "grad_norm": 0.730129599571228,
+      "learning_rate": 2.880288462954727e-06,
+      "loss": 0.6088,
+      "step": 9832
+    },
+    {
+      "epoch": 2.7238227146814404,
+      "grad_norm": 0.7116170525550842,
+      "learning_rate": 2.8799284133332878e-06,
+      "loss": 0.5718,
+      "step": 9833
+    },
+    {
+      "epoch": 2.72409972299169,
+      "grad_norm": 0.7043094038963318,
+      "learning_rate": 2.879568355644998e-06,
+      "loss": 0.5554,
+      "step": 9834
+    },
+    {
+      "epoch": 2.724376731301939,
+      "grad_norm": 0.7326838374137878,
+      "learning_rate": 2.8792082898975028e-06,
+      "loss": 0.5893,
+      "step": 9835
+    },
+    {
+      "epoch": 2.7246537396121884,
+      "grad_norm": 0.7812146544456482,
+      "learning_rate": 2.878848216098446e-06,
+      "loss": 0.6242,
+      "step": 9836
+    },
+    {
+      "epoch": 2.724930747922438,
+      "grad_norm": 0.7676199674606323,
+      "learning_rate": 2.8784881342554753e-06,
+      "loss": 0.6502,
+      "step": 9837
+    },
+    {
+      "epoch": 2.725207756232687,
+      "grad_norm": 0.7518264055252075,
+      "learning_rate": 2.8781280443762348e-06,
+      "loss": 0.5986,
+      "step": 9838
+    },
+    {
+      "epoch": 2.7254847645429363,
+      "grad_norm": 0.7407569885253906,
+      "learning_rate": 2.8777679464683704e-06,
+      "loss": 0.6142,
+      "step": 9839
+    },
+    {
+      "epoch": 2.7257617728531853,
+      "grad_norm": 0.7235451340675354,
+      "learning_rate": 2.8774078405395276e-06,
+      "loss": 0.6065,
+      "step": 9840
+    },
+    {
+      "epoch": 2.726038781163435,
+      "grad_norm": 0.714812159538269,
+      "learning_rate": 2.8770477265973525e-06,
+      "loss": 0.6125,
+      "step": 9841
+    },
+    {
+      "epoch": 2.7263157894736842,
+      "grad_norm": 0.8110000491142273,
+      "learning_rate": 2.876687604649491e-06,
+      "loss": 0.6032,
+      "step": 9842
+    },
+    {
+      "epoch": 2.7265927977839333,
+      "grad_norm": 0.7523626089096069,
+      "learning_rate": 2.8763274747035907e-06,
+      "loss": 0.6239,
+      "step": 9843
+    },
+    {
+      "epoch": 2.7268698060941827,
+      "grad_norm": 0.7214419841766357,
+      "learning_rate": 2.8759673367672964e-06,
+      "loss": 0.6413,
+      "step": 9844
+    },
+    {
+      "epoch": 2.727146814404432,
+      "grad_norm": 0.7556858658790588,
+      "learning_rate": 2.8756071908482553e-06,
+      "loss": 0.6243,
+      "step": 9845
+    },
+    {
+      "epoch": 2.727423822714681,
+      "grad_norm": 0.7446470856666565,
+      "learning_rate": 2.8752470369541152e-06,
+      "loss": 0.5993,
+      "step": 9846
+    },
+    {
+      "epoch": 2.7277008310249307,
+      "grad_norm": 0.7211043834686279,
+      "learning_rate": 2.8748868750925207e-06,
+      "loss": 0.594,
+      "step": 9847
+    },
+    {
+      "epoch": 2.72797783933518,
+      "grad_norm": 0.7198621034622192,
+      "learning_rate": 2.8745267052711217e-06,
+      "loss": 0.6002,
+      "step": 9848
+    },
+    {
+      "epoch": 2.728254847645429,
+      "grad_norm": 0.7595424652099609,
+      "learning_rate": 2.874166527497564e-06,
+      "loss": 0.5839,
+      "step": 9849
+    },
+    {
+      "epoch": 2.7285318559556786,
+      "grad_norm": 0.7390759587287903,
+      "learning_rate": 2.8738063417794947e-06,
+      "loss": 0.5863,
+      "step": 9850
+    },
+    {
+      "epoch": 2.728808864265928,
+      "grad_norm": 0.746052622795105,
+      "learning_rate": 2.873446148124563e-06,
+      "loss": 0.5581,
+      "step": 9851
+    },
+    {
+      "epoch": 2.729085872576177,
+      "grad_norm": 0.732109785079956,
+      "learning_rate": 2.873085946540416e-06,
+      "loss": 0.6028,
+      "step": 9852
+    },
+    {
+      "epoch": 2.7293628808864265,
+      "grad_norm": 0.7366575002670288,
+      "learning_rate": 2.872725737034701e-06,
+      "loss": 0.6197,
+      "step": 9853
+    },
+    {
+      "epoch": 2.729639889196676,
+      "grad_norm": 0.824176549911499,
+      "learning_rate": 2.872365519615067e-06,
+      "loss": 0.6023,
+      "step": 9854
+    },
+    {
+      "epoch": 2.729916897506925,
+      "grad_norm": 0.767004132270813,
+      "learning_rate": 2.8720052942891624e-06,
+      "loss": 0.6114,
+      "step": 9855
+    },
+    {
+      "epoch": 2.7301939058171745,
+      "grad_norm": 0.738639235496521,
+      "learning_rate": 2.871645061064635e-06,
+      "loss": 0.5904,
+      "step": 9856
+    },
+    {
+      "epoch": 2.730470914127424,
+      "grad_norm": 0.7889578342437744,
+      "learning_rate": 2.8712848199491343e-06,
+      "loss": 0.5975,
+      "step": 9857
+    },
+    {
+      "epoch": 2.730747922437673,
+      "grad_norm": 0.7595368027687073,
+      "learning_rate": 2.870924570950308e-06,
+      "loss": 0.5689,
+      "step": 9858
+    },
+    {
+      "epoch": 2.7310249307479224,
+      "grad_norm": 0.7748566269874573,
+      "learning_rate": 2.8705643140758065e-06,
+      "loss": 0.6055,
+      "step": 9859
+    },
+    {
+      "epoch": 2.731301939058172,
+      "grad_norm": 0.7329906821250916,
+      "learning_rate": 2.8702040493332776e-06,
+      "loss": 0.5957,
+      "step": 9860
+    },
+    {
+      "epoch": 2.731578947368421,
+      "grad_norm": 0.7348831295967102,
+      "learning_rate": 2.8698437767303724e-06,
+      "loss": 0.6031,
+      "step": 9861
+    },
+    {
+      "epoch": 2.7318559556786703,
+      "grad_norm": 0.7103914618492126,
+      "learning_rate": 2.8694834962747386e-06,
+      "loss": 0.5777,
+      "step": 9862
+    },
+    {
+      "epoch": 2.73213296398892,
+      "grad_norm": 0.8002786040306091,
+      "learning_rate": 2.869123207974027e-06,
+      "loss": 0.597,
+      "step": 9863
+    },
+    {
+      "epoch": 2.732409972299169,
+      "grad_norm": 0.7433820366859436,
+      "learning_rate": 2.868762911835886e-06,
+      "loss": 0.5898,
+      "step": 9864
+    },
+    {
+      "epoch": 2.7326869806094183,
+      "grad_norm": 0.7274856567382812,
+      "learning_rate": 2.8684026078679684e-06,
+      "loss": 0.5686,
+      "step": 9865
+    },
+    {
+      "epoch": 2.7329639889196677,
+      "grad_norm": 0.7254496216773987,
+      "learning_rate": 2.8680422960779215e-06,
+      "loss": 0.5935,
+      "step": 9866
+    },
+    {
+      "epoch": 2.7332409972299168,
+      "grad_norm": 0.7380785346031189,
+      "learning_rate": 2.8676819764733965e-06,
+      "loss": 0.5996,
+      "step": 9867
+    },
+    {
+      "epoch": 2.733518005540166,
+      "grad_norm": 0.7902967929840088,
+      "learning_rate": 2.8673216490620453e-06,
+      "loss": 0.6329,
+      "step": 9868
+    },
+    {
+      "epoch": 2.7337950138504157,
+      "grad_norm": 0.7636850476264954,
+      "learning_rate": 2.8669613138515163e-06,
+      "loss": 0.5976,
+      "step": 9869
+    },
+    {
+      "epoch": 2.7340720221606647,
+      "grad_norm": 0.6949307918548584,
+      "learning_rate": 2.8666009708494624e-06,
+      "loss": 0.5478,
+      "step": 9870
+    },
+    {
+      "epoch": 2.734349030470914,
+      "grad_norm": 0.7565574049949646,
+      "learning_rate": 2.8662406200635336e-06,
+      "loss": 0.5925,
+      "step": 9871
+    },
+    {
+      "epoch": 2.7346260387811636,
+      "grad_norm": 0.7426764369010925,
+      "learning_rate": 2.865880261501381e-06,
+      "loss": 0.5982,
+      "step": 9872
+    },
+    {
+      "epoch": 2.7349030470914126,
+      "grad_norm": 0.7588077783584595,
+      "learning_rate": 2.8655198951706553e-06,
+      "loss": 0.5879,
+      "step": 9873
+    },
+    {
+      "epoch": 2.735180055401662,
+      "grad_norm": 0.8109205365180969,
+      "learning_rate": 2.8651595210790097e-06,
+      "loss": 0.6319,
+      "step": 9874
+    },
+    {
+      "epoch": 2.7354570637119116,
+      "grad_norm": 0.7463670372962952,
+      "learning_rate": 2.8647991392340945e-06,
+      "loss": 0.5441,
+      "step": 9875
+    },
+    {
+      "epoch": 2.7357340720221606,
+      "grad_norm": 0.7290921807289124,
+      "learning_rate": 2.864438749643563e-06,
+      "loss": 0.557,
+      "step": 9876
+    },
+    {
+      "epoch": 2.73601108033241,
+      "grad_norm": 0.7342871427536011,
+      "learning_rate": 2.8640783523150646e-06,
+      "loss": 0.5994,
+      "step": 9877
+    },
+    {
+      "epoch": 2.7362880886426595,
+      "grad_norm": 0.7484715580940247,
+      "learning_rate": 2.8637179472562543e-06,
+      "loss": 0.6144,
+      "step": 9878
+    },
+    {
+      "epoch": 2.7365650969529085,
+      "grad_norm": 0.7376574873924255,
+      "learning_rate": 2.863357534474782e-06,
+      "loss": 0.6049,
+      "step": 9879
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.7331152558326721,
+      "learning_rate": 2.862997113978302e-06,
+      "loss": 0.5784,
+      "step": 9880
+    },
+    {
+      "epoch": 2.7371191135734074,
+      "grad_norm": 0.7068518996238708,
+      "learning_rate": 2.8626366857744663e-06,
+      "loss": 0.6158,
+      "step": 9881
+    },
+    {
+      "epoch": 2.7373961218836564,
+      "grad_norm": 0.7011471390724182,
+      "learning_rate": 2.862276249870927e-06,
+      "loss": 0.5773,
+      "step": 9882
+    },
+    {
+      "epoch": 2.737673130193906,
+      "grad_norm": 0.7379846572875977,
+      "learning_rate": 2.8619158062753387e-06,
+      "loss": 0.5378,
+      "step": 9883
+    },
+    {
+      "epoch": 2.7379501385041554,
+      "grad_norm": 0.7986716032028198,
+      "learning_rate": 2.8615553549953534e-06,
+      "loss": 0.5816,
+      "step": 9884
+    },
+    {
+      "epoch": 2.7382271468144044,
+      "grad_norm": 0.7555115222930908,
+      "learning_rate": 2.861194896038625e-06,
+      "loss": 0.5616,
+      "step": 9885
+    },
+    {
+      "epoch": 2.738504155124654,
+      "grad_norm": 0.7393010854721069,
+      "learning_rate": 2.8608344294128054e-06,
+      "loss": 0.6302,
+      "step": 9886
+    },
+    {
+      "epoch": 2.7387811634349033,
+      "grad_norm": 0.7402096390724182,
+      "learning_rate": 2.8604739551255502e-06,
+      "loss": 0.5949,
+      "step": 9887
+    },
+    {
+      "epoch": 2.7390581717451523,
+      "grad_norm": 0.7098001837730408,
+      "learning_rate": 2.8601134731845126e-06,
+      "loss": 0.5708,
+      "step": 9888
+    },
+    {
+      "epoch": 2.7393351800554018,
+      "grad_norm": 0.7866057753562927,
+      "learning_rate": 2.8597529835973458e-06,
+      "loss": 0.5667,
+      "step": 9889
+    },
+    {
+      "epoch": 2.7396121883656512,
+      "grad_norm": 0.7264719605445862,
+      "learning_rate": 2.859392486371705e-06,
+      "loss": 0.5916,
+      "step": 9890
+    },
+    {
+      "epoch": 2.7398891966759003,
+      "grad_norm": 0.7185991406440735,
+      "learning_rate": 2.8590319815152433e-06,
+      "loss": 0.5991,
+      "step": 9891
+    },
+    {
+      "epoch": 2.7401662049861497,
+      "grad_norm": 0.7286472320556641,
+      "learning_rate": 2.8586714690356156e-06,
+      "loss": 0.6273,
+      "step": 9892
+    },
+    {
+      "epoch": 2.7404432132963987,
+      "grad_norm": 0.7181364297866821,
+      "learning_rate": 2.858310948940477e-06,
+      "loss": 0.6077,
+      "step": 9893
+    },
+    {
+      "epoch": 2.740720221606648,
+      "grad_norm": 0.739396870136261,
+      "learning_rate": 2.8579504212374827e-06,
+      "loss": 0.5982,
+      "step": 9894
+    },
+    {
+      "epoch": 2.7409972299168976,
+      "grad_norm": 0.7353379130363464,
+      "learning_rate": 2.8575898859342854e-06,
+      "loss": 0.5989,
+      "step": 9895
+    },
+    {
+      "epoch": 2.7412742382271467,
+      "grad_norm": 0.7130840420722961,
+      "learning_rate": 2.8572293430385423e-06,
+      "loss": 0.5807,
+      "step": 9896
+    },
+    {
+      "epoch": 2.741551246537396,
+      "grad_norm": 0.746138334274292,
+      "learning_rate": 2.8568687925579084e-06,
+      "loss": 0.6147,
+      "step": 9897
+    },
+    {
+      "epoch": 2.7418282548476456,
+      "grad_norm": 0.773597240447998,
+      "learning_rate": 2.8565082345000384e-06,
+      "loss": 0.6073,
+      "step": 9898
+    },
+    {
+      "epoch": 2.7421052631578946,
+      "grad_norm": 0.7569901943206787,
+      "learning_rate": 2.8561476688725877e-06,
+      "loss": 0.6017,
+      "step": 9899
+    },
+    {
+      "epoch": 2.742382271468144,
+      "grad_norm": 0.7603424787521362,
+      "learning_rate": 2.8557870956832135e-06,
+      "loss": 0.6067,
+      "step": 9900
+    },
+    {
+      "epoch": 2.742659279778393,
+      "grad_norm": 0.7068322896957397,
+      "learning_rate": 2.85542651493957e-06,
+      "loss": 0.6157,
+      "step": 9901
+    },
+    {
+      "epoch": 2.7429362880886425,
+      "grad_norm": 0.7092251777648926,
+      "learning_rate": 2.855065926649314e-06,
+      "loss": 0.5645,
+      "step": 9902
+    },
+    {
+      "epoch": 2.743213296398892,
+      "grad_norm": 0.7123865485191345,
+      "learning_rate": 2.8547053308201016e-06,
+      "loss": 0.5271,
+      "step": 9903
+    },
+    {
+      "epoch": 2.743490304709141,
+      "grad_norm": 0.7555539608001709,
+      "learning_rate": 2.854344727459589e-06,
+      "loss": 0.6047,
+      "step": 9904
+    },
+    {
+      "epoch": 2.7437673130193905,
+      "grad_norm": 0.7480126619338989,
+      "learning_rate": 2.853984116575433e-06,
+      "loss": 0.5918,
+      "step": 9905
+    },
+    {
+      "epoch": 2.74404432132964,
+      "grad_norm": 0.7401037216186523,
+      "learning_rate": 2.853623498175291e-06,
+      "loss": 0.5512,
+      "step": 9906
+    },
+    {
+      "epoch": 2.744321329639889,
+      "grad_norm": 0.7168396711349487,
+      "learning_rate": 2.853262872266818e-06,
+      "loss": 0.5981,
+      "step": 9907
+    },
+    {
+      "epoch": 2.7445983379501384,
+      "grad_norm": 0.7359672784805298,
+      "learning_rate": 2.8529022388576722e-06,
+      "loss": 0.6046,
+      "step": 9908
+    },
+    {
+      "epoch": 2.744875346260388,
+      "grad_norm": 0.7549652457237244,
+      "learning_rate": 2.8525415979555114e-06,
+      "loss": 0.56,
+      "step": 9909
+    },
+    {
+      "epoch": 2.745152354570637,
+      "grad_norm": 0.7408303022384644,
+      "learning_rate": 2.852180949567992e-06,
+      "loss": 0.6197,
+      "step": 9910
+    },
+    {
+      "epoch": 2.7454293628808863,
+      "grad_norm": 0.7500876784324646,
+      "learning_rate": 2.8518202937027716e-06,
+      "loss": 0.5886,
+      "step": 9911
+    },
+    {
+      "epoch": 2.745706371191136,
+      "grad_norm": 0.7531888484954834,
+      "learning_rate": 2.8514596303675073e-06,
+      "loss": 0.6038,
+      "step": 9912
+    },
+    {
+      "epoch": 2.745983379501385,
+      "grad_norm": 0.7636232376098633,
+      "learning_rate": 2.8510989595698586e-06,
+      "loss": 0.593,
+      "step": 9913
+    },
+    {
+      "epoch": 2.7462603878116343,
+      "grad_norm": 0.8050282001495361,
+      "learning_rate": 2.850738281317482e-06,
+      "loss": 0.614,
+      "step": 9914
+    },
+    {
+      "epoch": 2.7465373961218837,
+      "grad_norm": 0.7462037205696106,
+      "learning_rate": 2.850377595618035e-06,
+      "loss": 0.5723,
+      "step": 9915
+    },
+    {
+      "epoch": 2.7468144044321328,
+      "grad_norm": 0.7207900285720825,
+      "learning_rate": 2.850016902479178e-06,
+      "loss": 0.6055,
+      "step": 9916
+    },
+    {
+      "epoch": 2.747091412742382,
+      "grad_norm": 0.7226089835166931,
+      "learning_rate": 2.8496562019085685e-06,
+      "loss": 0.6262,
+      "step": 9917
+    },
+    {
+      "epoch": 2.7473684210526317,
+      "grad_norm": 0.7438740730285645,
+      "learning_rate": 2.849295493913864e-06,
+      "loss": 0.6241,
+      "step": 9918
+    },
+    {
+      "epoch": 2.7476454293628807,
+      "grad_norm": 0.7763872146606445,
+      "learning_rate": 2.8489347785027248e-06,
+      "loss": 0.6049,
+      "step": 9919
+    },
+    {
+      "epoch": 2.74792243767313,
+      "grad_norm": 0.7234786152839661,
+      "learning_rate": 2.8485740556828082e-06,
+      "loss": 0.582,
+      "step": 9920
+    },
+    {
+      "epoch": 2.7481994459833796,
+      "grad_norm": 0.7248966097831726,
+      "learning_rate": 2.848213325461775e-06,
+      "loss": 0.5763,
+      "step": 9921
+    },
+    {
+      "epoch": 2.7484764542936286,
+      "grad_norm": 0.7010312080383301,
+      "learning_rate": 2.847852587847284e-06,
+      "loss": 0.5621,
+      "step": 9922
+    },
+    {
+      "epoch": 2.748753462603878,
+      "grad_norm": 0.6976631879806519,
+      "learning_rate": 2.8474918428469937e-06,
+      "loss": 0.6227,
+      "step": 9923
+    },
+    {
+      "epoch": 2.7490304709141276,
+      "grad_norm": 0.7105882167816162,
+      "learning_rate": 2.8471310904685644e-06,
+      "loss": 0.597,
+      "step": 9924
+    },
+    {
+      "epoch": 2.7493074792243766,
+      "grad_norm": 0.7569697499275208,
+      "learning_rate": 2.846770330719656e-06,
+      "loss": 0.6418,
+      "step": 9925
+    },
+    {
+      "epoch": 2.749584487534626,
+      "grad_norm": 0.716084361076355,
+      "learning_rate": 2.846409563607927e-06,
+      "loss": 0.6109,
+      "step": 9926
+    },
+    {
+      "epoch": 2.7498614958448755,
+      "grad_norm": 0.7867611646652222,
+      "learning_rate": 2.846048789141039e-06,
+      "loss": 0.5647,
+      "step": 9927
+    },
+    {
+      "epoch": 2.7501385041551245,
+      "grad_norm": 0.7521023154258728,
+      "learning_rate": 2.8456880073266514e-06,
+      "loss": 0.5844,
+      "step": 9928
+    },
+    {
+      "epoch": 2.750415512465374,
+      "grad_norm": 0.7604575157165527,
+      "learning_rate": 2.845327218172424e-06,
+      "loss": 0.6146,
+      "step": 9929
+    },
+    {
+      "epoch": 2.7506925207756234,
+      "grad_norm": 0.7507315278053284,
+      "learning_rate": 2.8449664216860183e-06,
+      "loss": 0.6499,
+      "step": 9930
+    },
+    {
+      "epoch": 2.7509695290858724,
+      "grad_norm": 0.7635160088539124,
+      "learning_rate": 2.8446056178750945e-06,
+      "loss": 0.6021,
+      "step": 9931
+    },
+    {
+      "epoch": 2.751246537396122,
+      "grad_norm": 0.7149273157119751,
+      "learning_rate": 2.8442448067473142e-06,
+      "loss": 0.5803,
+      "step": 9932
+    },
+    {
+      "epoch": 2.7515235457063714,
+      "grad_norm": 0.7017562389373779,
+      "learning_rate": 2.8438839883103363e-06,
+      "loss": 0.5707,
+      "step": 9933
+    },
+    {
+      "epoch": 2.7518005540166204,
+      "grad_norm": 0.760026216506958,
+      "learning_rate": 2.8435231625718242e-06,
+      "loss": 0.5867,
+      "step": 9934
+    },
+    {
+      "epoch": 2.75207756232687,
+      "grad_norm": 0.7673696875572205,
+      "learning_rate": 2.8431623295394374e-06,
+      "loss": 0.58,
+      "step": 9935
+    },
+    {
+      "epoch": 2.7523545706371193,
+      "grad_norm": 0.8185964822769165,
+      "learning_rate": 2.842801489220838e-06,
+      "loss": 0.5845,
+      "step": 9936
+    },
+    {
+      "epoch": 2.7526315789473683,
+      "grad_norm": 0.718859851360321,
+      "learning_rate": 2.8424406416236878e-06,
+      "loss": 0.5931,
+      "step": 9937
+    },
+    {
+      "epoch": 2.752908587257618,
+      "grad_norm": 0.743505597114563,
+      "learning_rate": 2.8420797867556483e-06,
+      "loss": 0.5825,
+      "step": 9938
+    },
+    {
+      "epoch": 2.7531855955678672,
+      "grad_norm": 0.7636753916740417,
+      "learning_rate": 2.8417189246243813e-06,
+      "loss": 0.6018,
+      "step": 9939
+    },
+    {
+      "epoch": 2.7534626038781163,
+      "grad_norm": 0.7697381973266602,
+      "learning_rate": 2.8413580552375487e-06,
+      "loss": 0.6298,
+      "step": 9940
+    },
+    {
+      "epoch": 2.7537396121883657,
+      "grad_norm": 0.7020732760429382,
+      "learning_rate": 2.8409971786028127e-06,
+      "loss": 0.5888,
+      "step": 9941
+    },
+    {
+      "epoch": 2.754016620498615,
+      "grad_norm": 0.7665590643882751,
+      "learning_rate": 2.8406362947278365e-06,
+      "loss": 0.6058,
+      "step": 9942
+    },
+    {
+      "epoch": 2.754293628808864,
+      "grad_norm": 0.7251317501068115,
+      "learning_rate": 2.8402754036202813e-06,
+      "loss": 0.5865,
+      "step": 9943
+    },
+    {
+      "epoch": 2.7545706371191137,
+      "grad_norm": 0.736558735370636,
+      "learning_rate": 2.839914505287811e-06,
+      "loss": 0.5718,
+      "step": 9944
+    },
+    {
+      "epoch": 2.754847645429363,
+      "grad_norm": 0.7454231381416321,
+      "learning_rate": 2.839553599738087e-06,
+      "loss": 0.5996,
+      "step": 9945
+    },
+    {
+      "epoch": 2.755124653739612,
+      "grad_norm": 0.7132341861724854,
+      "learning_rate": 2.8391926869787726e-06,
+      "loss": 0.5469,
+      "step": 9946
+    },
+    {
+      "epoch": 2.7554016620498616,
+      "grad_norm": 0.7635960578918457,
+      "learning_rate": 2.838831767017532e-06,
+      "loss": 0.6274,
+      "step": 9947
+    },
+    {
+      "epoch": 2.755678670360111,
+      "grad_norm": 0.7447330355644226,
+      "learning_rate": 2.8384708398620273e-06,
+      "loss": 0.6008,
+      "step": 9948
+    },
+    {
+      "epoch": 2.75595567867036,
+      "grad_norm": 0.7645748257637024,
+      "learning_rate": 2.838109905519922e-06,
+      "loss": 0.5936,
+      "step": 9949
+    },
+    {
+      "epoch": 2.7562326869806095,
+      "grad_norm": 0.758327305316925,
+      "learning_rate": 2.8377489639988808e-06,
+      "loss": 0.5703,
+      "step": 9950
+    },
+    {
+      "epoch": 2.756509695290859,
+      "grad_norm": 0.7296598553657532,
+      "learning_rate": 2.837388015306566e-06,
+      "loss": 0.6457,
+      "step": 9951
+    },
+    {
+      "epoch": 2.756786703601108,
+      "grad_norm": 0.7532568573951721,
+      "learning_rate": 2.8370270594506432e-06,
+      "loss": 0.5712,
+      "step": 9952
+    },
+    {
+      "epoch": 2.7570637119113575,
+      "grad_norm": 0.7405272722244263,
+      "learning_rate": 2.836666096438774e-06,
+      "loss": 0.5734,
+      "step": 9953
+    },
+    {
+      "epoch": 2.757340720221607,
+      "grad_norm": 0.7674778699874878,
+      "learning_rate": 2.8363051262786247e-06,
+      "loss": 0.5953,
+      "step": 9954
+    },
+    {
+      "epoch": 2.757617728531856,
+      "grad_norm": 0.754295825958252,
+      "learning_rate": 2.835944148977858e-06,
+      "loss": 0.5753,
+      "step": 9955
+    },
+    {
+      "epoch": 2.7578947368421054,
+      "grad_norm": 0.8103058338165283,
+      "learning_rate": 2.835583164544139e-06,
+      "loss": 0.5539,
+      "step": 9956
+    },
+    {
+      "epoch": 2.7581717451523544,
+      "grad_norm": 0.7378125786781311,
+      "learning_rate": 2.835222172985133e-06,
+      "loss": 0.5761,
+      "step": 9957
+    },
+    {
+      "epoch": 2.758448753462604,
+      "grad_norm": 0.7553445100784302,
+      "learning_rate": 2.8348611743085044e-06,
+      "loss": 0.6038,
+      "step": 9958
+    },
+    {
+      "epoch": 2.7587257617728533,
+      "grad_norm": 0.7586629986763,
+      "learning_rate": 2.8345001685219166e-06,
+      "loss": 0.5723,
+      "step": 9959
+    },
+    {
+      "epoch": 2.7590027700831024,
+      "grad_norm": 0.7592271566390991,
+      "learning_rate": 2.8341391556330372e-06,
+      "loss": 0.5991,
+      "step": 9960
+    },
+    {
+      "epoch": 2.759279778393352,
+      "grad_norm": 0.7449199557304382,
+      "learning_rate": 2.8337781356495297e-06,
+      "loss": 0.6285,
+      "step": 9961
+    },
+    {
+      "epoch": 2.7595567867036013,
+      "grad_norm": 0.8248133659362793,
+      "learning_rate": 2.8334171085790606e-06,
+      "loss": 0.5723,
+      "step": 9962
+    },
+    {
+      "epoch": 2.7598337950138503,
+      "grad_norm": 0.754346489906311,
+      "learning_rate": 2.833056074429294e-06,
+      "loss": 0.6417,
+      "step": 9963
+    },
+    {
+      "epoch": 2.7601108033240997,
+      "grad_norm": 0.7287446856498718,
+      "learning_rate": 2.8326950332078967e-06,
+      "loss": 0.6107,
+      "step": 9964
+    },
+    {
+      "epoch": 2.7603878116343488,
+      "grad_norm": 0.7717081308364868,
+      "learning_rate": 2.8323339849225345e-06,
+      "loss": 0.6411,
+      "step": 9965
+    },
+    {
+      "epoch": 2.7606648199445982,
+      "grad_norm": 0.7579557299613953,
+      "learning_rate": 2.831972929580873e-06,
+      "loss": 0.6529,
+      "step": 9966
+    },
+    {
+      "epoch": 2.7609418282548477,
+      "grad_norm": 0.7521576285362244,
+      "learning_rate": 2.831611867190579e-06,
+      "loss": 0.5969,
+      "step": 9967
+    },
+    {
+      "epoch": 2.7612188365650967,
+      "grad_norm": 0.7382582426071167,
+      "learning_rate": 2.831250797759317e-06,
+      "loss": 0.5763,
+      "step": 9968
+    },
+    {
+      "epoch": 2.761495844875346,
+      "grad_norm": 0.7770794630050659,
+      "learning_rate": 2.8308897212947558e-06,
+      "loss": 0.6108,
+      "step": 9969
+    },
+    {
+      "epoch": 2.7617728531855956,
+      "grad_norm": 0.7501707673072815,
+      "learning_rate": 2.83052863780456e-06,
+      "loss": 0.6017,
+      "step": 9970
+    },
+    {
+      "epoch": 2.7620498614958446,
+      "grad_norm": 0.7477511763572693,
+      "learning_rate": 2.8301675472963973e-06,
+      "loss": 0.5902,
+      "step": 9971
+    },
+    {
+      "epoch": 2.762326869806094,
+      "grad_norm": 0.7257628440856934,
+      "learning_rate": 2.829806449777934e-06,
+      "loss": 0.6117,
+      "step": 9972
+    },
+    {
+      "epoch": 2.7626038781163436,
+      "grad_norm": 0.7329751253128052,
+      "learning_rate": 2.8294453452568395e-06,
+      "loss": 0.6202,
+      "step": 9973
+    },
+    {
+      "epoch": 2.7628808864265926,
+      "grad_norm": 0.7268105149269104,
+      "learning_rate": 2.829084233740777e-06,
+      "loss": 0.5372,
+      "step": 9974
+    },
+    {
+      "epoch": 2.763157894736842,
+      "grad_norm": 0.7518357038497925,
+      "learning_rate": 2.8287231152374164e-06,
+      "loss": 0.592,
+      "step": 9975
+    },
+    {
+      "epoch": 2.7634349030470915,
+      "grad_norm": 0.7339359521865845,
+      "learning_rate": 2.8283619897544247e-06,
+      "loss": 0.6167,
+      "step": 9976
+    },
+    {
+      "epoch": 2.7637119113573405,
+      "grad_norm": 0.7902821898460388,
+      "learning_rate": 2.828000857299469e-06,
+      "loss": 0.5682,
+      "step": 9977
+    },
+    {
+      "epoch": 2.76398891966759,
+      "grad_norm": 0.7443130016326904,
+      "learning_rate": 2.827639717880218e-06,
+      "loss": 0.6265,
+      "step": 9978
+    },
+    {
+      "epoch": 2.7642659279778394,
+      "grad_norm": 0.7677735686302185,
+      "learning_rate": 2.827278571504339e-06,
+      "loss": 0.5631,
+      "step": 9979
+    },
+    {
+      "epoch": 2.7645429362880884,
+      "grad_norm": 0.742179811000824,
+      "learning_rate": 2.8269174181795e-06,
+      "loss": 0.5674,
+      "step": 9980
+    },
+    {
+      "epoch": 2.764819944598338,
+      "grad_norm": 0.744736909866333,
+      "learning_rate": 2.8265562579133694e-06,
+      "loss": 0.6166,
+      "step": 9981
+    },
+    {
+      "epoch": 2.7650969529085874,
+      "grad_norm": 0.7325068712234497,
+      "learning_rate": 2.8261950907136154e-06,
+      "loss": 0.5825,
+      "step": 9982
+    },
+    {
+      "epoch": 2.7653739612188364,
+      "grad_norm": 0.7419037818908691,
+      "learning_rate": 2.825833916587907e-06,
+      "loss": 0.567,
+      "step": 9983
+    },
+    {
+      "epoch": 2.765650969529086,
+      "grad_norm": 0.7596971988677979,
+      "learning_rate": 2.8254727355439122e-06,
+      "loss": 0.6059,
+      "step": 9984
+    },
+    {
+      "epoch": 2.7659279778393353,
+      "grad_norm": 0.7159470319747925,
+      "learning_rate": 2.8251115475893e-06,
+      "loss": 0.5671,
+      "step": 9985
+    },
+    {
+      "epoch": 2.7662049861495843,
+      "grad_norm": 0.7521191239356995,
+      "learning_rate": 2.82475035273174e-06,
+      "loss": 0.5968,
+      "step": 9986
+    },
+    {
+      "epoch": 2.766481994459834,
+      "grad_norm": 0.7235638499259949,
+      "learning_rate": 2.8243891509788997e-06,
+      "loss": 0.6249,
+      "step": 9987
+    },
+    {
+      "epoch": 2.7667590027700832,
+      "grad_norm": 0.7232115268707275,
+      "learning_rate": 2.8240279423384497e-06,
+      "loss": 0.52,
+      "step": 9988
+    },
+    {
+      "epoch": 2.7670360110803323,
+      "grad_norm": 0.7460992336273193,
+      "learning_rate": 2.823666726818059e-06,
+      "loss": 0.6231,
+      "step": 9989
+    },
+    {
+      "epoch": 2.7673130193905817,
+      "grad_norm": 0.7331829071044922,
+      "learning_rate": 2.8233055044253975e-06,
+      "loss": 0.5745,
+      "step": 9990
+    },
+    {
+      "epoch": 2.767590027700831,
+      "grad_norm": 0.7375158667564392,
+      "learning_rate": 2.8229442751681334e-06,
+      "loss": 0.5944,
+      "step": 9991
+    },
+    {
+      "epoch": 2.76786703601108,
+      "grad_norm": 0.7081493139266968,
+      "learning_rate": 2.822583039053939e-06,
+      "loss": 0.6253,
+      "step": 9992
+    },
+    {
+      "epoch": 2.7681440443213297,
+      "grad_norm": 0.7295372486114502,
+      "learning_rate": 2.8222217960904823e-06,
+      "loss": 0.5776,
+      "step": 9993
+    },
+    {
+      "epoch": 2.768421052631579,
+      "grad_norm": 0.7875064015388489,
+      "learning_rate": 2.821860546285434e-06,
+      "loss": 0.5834,
+      "step": 9994
+    },
+    {
+      "epoch": 2.768698060941828,
+      "grad_norm": 0.7445015907287598,
+      "learning_rate": 2.8214992896464643e-06,
+      "loss": 0.6145,
+      "step": 9995
+    },
+    {
+      "epoch": 2.7689750692520776,
+      "grad_norm": 0.7500941157341003,
+      "learning_rate": 2.821138026181244e-06,
+      "loss": 0.5726,
+      "step": 9996
+    },
+    {
+      "epoch": 2.769252077562327,
+      "grad_norm": 0.7612580060958862,
+      "learning_rate": 2.820776755897443e-06,
+      "loss": 0.6014,
+      "step": 9997
+    },
+    {
+      "epoch": 2.769529085872576,
+      "grad_norm": 0.7436586022377014,
+      "learning_rate": 2.820415478802733e-06,
+      "loss": 0.5925,
+      "step": 9998
+    },
+    {
+      "epoch": 2.7698060941828255,
+      "grad_norm": 0.7628955245018005,
+      "learning_rate": 2.820054194904783e-06,
+      "loss": 0.6223,
+      "step": 9999
+    },
+    {
+      "epoch": 2.770083102493075,
+      "grad_norm": 0.7731196880340576,
+      "learning_rate": 2.8196929042112652e-06,
+      "loss": 0.5961,
+      "step": 10000
+    },
+    {
+      "epoch": 2.770360110803324,
+      "grad_norm": 0.6865490078926086,
+      "learning_rate": 2.8193316067298514e-06,
+      "loss": 0.6111,
+      "step": 10001
+    },
+    {
+      "epoch": 2.7706371191135735,
+      "grad_norm": 0.7498258352279663,
+      "learning_rate": 2.818970302468212e-06,
+      "loss": 0.6077,
+      "step": 10002
+    },
+    {
+      "epoch": 2.770914127423823,
+      "grad_norm": 0.7517375349998474,
+      "learning_rate": 2.818608991434018e-06,
+      "loss": 0.5781,
+      "step": 10003
+    },
+    {
+      "epoch": 2.771191135734072,
+      "grad_norm": 0.7093813419342041,
+      "learning_rate": 2.8182476736349414e-06,
+      "loss": 0.5823,
+      "step": 10004
+    },
+    {
+      "epoch": 2.7714681440443214,
+      "grad_norm": 0.7525933384895325,
+      "learning_rate": 2.817886349078655e-06,
+      "loss": 0.5924,
+      "step": 10005
+    },
+    {
+      "epoch": 2.771745152354571,
+      "grad_norm": 0.7898561954498291,
+      "learning_rate": 2.8175250177728288e-06,
+      "loss": 0.5583,
+      "step": 10006
+    },
+    {
+      "epoch": 2.77202216066482,
+      "grad_norm": 0.698114275932312,
+      "learning_rate": 2.817163679725135e-06,
+      "loss": 0.5231,
+      "step": 10007
+    },
+    {
+      "epoch": 2.7722991689750693,
+      "grad_norm": 0.7242182493209839,
+      "learning_rate": 2.816802334943248e-06,
+      "loss": 0.5472,
+      "step": 10008
+    },
+    {
+      "epoch": 2.772576177285319,
+      "grad_norm": 0.7450344562530518,
+      "learning_rate": 2.8164409834348366e-06,
+      "loss": 0.6198,
+      "step": 10009
+    },
+    {
+      "epoch": 2.772853185595568,
+      "grad_norm": 0.7549633383750916,
+      "learning_rate": 2.816079625207576e-06,
+      "loss": 0.5971,
+      "step": 10010
+    },
+    {
+      "epoch": 2.7731301939058173,
+      "grad_norm": 0.7310160398483276,
+      "learning_rate": 2.815718260269138e-06,
+      "loss": 0.5808,
+      "step": 10011
+    },
+    {
+      "epoch": 2.7734072022160667,
+      "grad_norm": 0.6927647590637207,
+      "learning_rate": 2.815356888627195e-06,
+      "loss": 0.5657,
+      "step": 10012
+    },
+    {
+      "epoch": 2.7736842105263158,
+      "grad_norm": 0.7121282815933228,
+      "learning_rate": 2.814995510289419e-06,
+      "loss": 0.5943,
+      "step": 10013
+    },
+    {
+      "epoch": 2.773961218836565,
+      "grad_norm": 0.7226163148880005,
+      "learning_rate": 2.8146341252634846e-06,
+      "loss": 0.6026,
+      "step": 10014
+    },
+    {
+      "epoch": 2.7742382271468147,
+      "grad_norm": 0.746878981590271,
+      "learning_rate": 2.8142727335570647e-06,
+      "loss": 0.5921,
+      "step": 10015
+    },
+    {
+      "epoch": 2.7745152354570637,
+      "grad_norm": 0.7342948913574219,
+      "learning_rate": 2.813911335177831e-06,
+      "loss": 0.5626,
+      "step": 10016
+    },
+    {
+      "epoch": 2.774792243767313,
+      "grad_norm": 0.7332433462142944,
+      "learning_rate": 2.813549930133459e-06,
+      "loss": 0.6213,
+      "step": 10017
+    },
+    {
+      "epoch": 2.775069252077562,
+      "grad_norm": 0.7966634035110474,
+      "learning_rate": 2.8131885184316214e-06,
+      "loss": 0.6069,
+      "step": 10018
+    },
+    {
+      "epoch": 2.7753462603878116,
+      "grad_norm": 0.7166701555252075,
+      "learning_rate": 2.8128271000799904e-06,
+      "loss": 0.5755,
+      "step": 10019
+    },
+    {
+      "epoch": 2.775623268698061,
+      "grad_norm": 0.7254489660263062,
+      "learning_rate": 2.8124656750862422e-06,
+      "loss": 0.5584,
+      "step": 10020
+    },
+    {
+      "epoch": 2.77590027700831,
+      "grad_norm": 0.73366779088974,
+      "learning_rate": 2.8121042434580504e-06,
+      "loss": 0.6032,
+      "step": 10021
+    },
+    {
+      "epoch": 2.7761772853185596,
+      "grad_norm": 0.8047178983688354,
+      "learning_rate": 2.811742805203087e-06,
+      "loss": 0.6413,
+      "step": 10022
+    },
+    {
+      "epoch": 2.776454293628809,
+      "grad_norm": 0.7692306041717529,
+      "learning_rate": 2.8113813603290287e-06,
+      "loss": 0.5748,
+      "step": 10023
+    },
+    {
+      "epoch": 2.776731301939058,
+      "grad_norm": 0.7090165019035339,
+      "learning_rate": 2.8110199088435487e-06,
+      "loss": 0.5303,
+      "step": 10024
+    },
+    {
+      "epoch": 2.7770083102493075,
+      "grad_norm": 0.7563058137893677,
+      "learning_rate": 2.8106584507543217e-06,
+      "loss": 0.5874,
+      "step": 10025
+    },
+    {
+      "epoch": 2.7772853185595565,
+      "grad_norm": 0.7437519431114197,
+      "learning_rate": 2.810296986069022e-06,
+      "loss": 0.5511,
+      "step": 10026
+    },
+    {
+      "epoch": 2.777562326869806,
+      "grad_norm": 0.750501811504364,
+      "learning_rate": 2.809935514795326e-06,
+      "loss": 0.5746,
+      "step": 10027
+    },
+    {
+      "epoch": 2.7778393351800554,
+      "grad_norm": 0.7103888392448425,
+      "learning_rate": 2.8095740369409062e-06,
+      "loss": 0.5923,
+      "step": 10028
+    },
+    {
+      "epoch": 2.7781163434903045,
+      "grad_norm": 0.7357938289642334,
+      "learning_rate": 2.80921255251344e-06,
+      "loss": 0.6062,
+      "step": 10029
+    },
+    {
+      "epoch": 2.778393351800554,
+      "grad_norm": 0.7615418434143066,
+      "learning_rate": 2.8088510615206014e-06,
+      "loss": 0.5656,
+      "step": 10030
+    },
+    {
+      "epoch": 2.7786703601108034,
+      "grad_norm": 0.7450594902038574,
+      "learning_rate": 2.8084895639700664e-06,
+      "loss": 0.595,
+      "step": 10031
+    },
+    {
+      "epoch": 2.7789473684210524,
+      "grad_norm": 0.7462698817253113,
+      "learning_rate": 2.8081280598695095e-06,
+      "loss": 0.5784,
+      "step": 10032
+    },
+    {
+      "epoch": 2.779224376731302,
+      "grad_norm": 0.7743009328842163,
+      "learning_rate": 2.8077665492266077e-06,
+      "loss": 0.6279,
+      "step": 10033
+    },
+    {
+      "epoch": 2.7795013850415513,
+      "grad_norm": 0.6979047060012817,
+      "learning_rate": 2.807405032049035e-06,
+      "loss": 0.5675,
+      "step": 10034
+    },
+    {
+      "epoch": 2.7797783933518003,
+      "grad_norm": 0.7872177958488464,
+      "learning_rate": 2.807043508344469e-06,
+      "loss": 0.5971,
+      "step": 10035
+    },
+    {
+      "epoch": 2.78005540166205,
+      "grad_norm": 0.7440598607063293,
+      "learning_rate": 2.8066819781205855e-06,
+      "loss": 0.5577,
+      "step": 10036
+    },
+    {
+      "epoch": 2.7803324099722992,
+      "grad_norm": 0.7891249060630798,
+      "learning_rate": 2.806320441385061e-06,
+      "loss": 0.6543,
+      "step": 10037
+    },
+    {
+      "epoch": 2.7806094182825483,
+      "grad_norm": 0.7209693789482117,
+      "learning_rate": 2.8059588981455703e-06,
+      "loss": 0.567,
+      "step": 10038
+    },
+    {
+      "epoch": 2.7808864265927977,
+      "grad_norm": 0.7257687449455261,
+      "learning_rate": 2.8055973484097907e-06,
+      "loss": 0.5811,
+      "step": 10039
+    },
+    {
+      "epoch": 2.781163434903047,
+      "grad_norm": 0.730828046798706,
+      "learning_rate": 2.8052357921854e-06,
+      "loss": 0.5483,
+      "step": 10040
+    },
+    {
+      "epoch": 2.781440443213296,
+      "grad_norm": 0.7517199516296387,
+      "learning_rate": 2.8048742294800733e-06,
+      "loss": 0.6337,
+      "step": 10041
+    },
+    {
+      "epoch": 2.7817174515235457,
+      "grad_norm": 0.704637348651886,
+      "learning_rate": 2.8045126603014884e-06,
+      "loss": 0.587,
+      "step": 10042
+    },
+    {
+      "epoch": 2.781994459833795,
+      "grad_norm": 0.7503658533096313,
+      "learning_rate": 2.8041510846573215e-06,
+      "loss": 0.6262,
+      "step": 10043
+    },
+    {
+      "epoch": 2.782271468144044,
+      "grad_norm": 0.7343692183494568,
+      "learning_rate": 2.8037895025552513e-06,
+      "loss": 0.5447,
+      "step": 10044
+    },
+    {
+      "epoch": 2.7825484764542936,
+      "grad_norm": 0.7240074872970581,
+      "learning_rate": 2.803427914002953e-06,
+      "loss": 0.6007,
+      "step": 10045
+    },
+    {
+      "epoch": 2.782825484764543,
+      "grad_norm": 0.7470293045043945,
+      "learning_rate": 2.8030663190081063e-06,
+      "loss": 0.5813,
+      "step": 10046
+    },
+    {
+      "epoch": 2.783102493074792,
+      "grad_norm": 0.7610907554626465,
+      "learning_rate": 2.802704717578387e-06,
+      "loss": 0.5667,
+      "step": 10047
+    },
+    {
+      "epoch": 2.7833795013850415,
+      "grad_norm": 0.7097828388214111,
+      "learning_rate": 2.8023431097214737e-06,
+      "loss": 0.5882,
+      "step": 10048
+    },
+    {
+      "epoch": 2.783656509695291,
+      "grad_norm": 0.7703149914741516,
+      "learning_rate": 2.8019814954450446e-06,
+      "loss": 0.6014,
+      "step": 10049
+    },
+    {
+      "epoch": 2.78393351800554,
+      "grad_norm": 0.7168884873390198,
+      "learning_rate": 2.8016198747567764e-06,
+      "loss": 0.589,
+      "step": 10050
+    },
+    {
+      "epoch": 2.7842105263157895,
+      "grad_norm": 0.7583420276641846,
+      "learning_rate": 2.8012582476643487e-06,
+      "loss": 0.599,
+      "step": 10051
+    },
+    {
+      "epoch": 2.784487534626039,
+      "grad_norm": 0.7327255010604858,
+      "learning_rate": 2.8008966141754383e-06,
+      "loss": 0.6251,
+      "step": 10052
+    },
+    {
+      "epoch": 2.784764542936288,
+      "grad_norm": 0.7265594005584717,
+      "learning_rate": 2.800534974297725e-06,
+      "loss": 0.5959,
+      "step": 10053
+    },
+    {
+      "epoch": 2.7850415512465374,
+      "grad_norm": 0.7101613879203796,
+      "learning_rate": 2.800173328038886e-06,
+      "loss": 0.594,
+      "step": 10054
+    },
+    {
+      "epoch": 2.785318559556787,
+      "grad_norm": 0.7457520365715027,
+      "learning_rate": 2.799811675406601e-06,
+      "loss": 0.5728,
+      "step": 10055
+    },
+    {
+      "epoch": 2.785595567867036,
+      "grad_norm": 0.7473955750465393,
+      "learning_rate": 2.7994500164085494e-06,
+      "loss": 0.5967,
+      "step": 10056
+    },
+    {
+      "epoch": 2.7858725761772853,
+      "grad_norm": 0.7834071516990662,
+      "learning_rate": 2.7990883510524087e-06,
+      "loss": 0.6262,
+      "step": 10057
+    },
+    {
+      "epoch": 2.786149584487535,
+      "grad_norm": 0.7517677545547485,
+      "learning_rate": 2.798726679345858e-06,
+      "loss": 0.5976,
+      "step": 10058
+    },
+    {
+      "epoch": 2.786426592797784,
+      "grad_norm": 0.717934787273407,
+      "learning_rate": 2.7983650012965773e-06,
+      "loss": 0.5945,
+      "step": 10059
+    },
+    {
+      "epoch": 2.7867036011080333,
+      "grad_norm": 0.7563323974609375,
+      "learning_rate": 2.7980033169122455e-06,
+      "loss": 0.5763,
+      "step": 10060
+    },
+    {
+      "epoch": 2.7869806094182827,
+      "grad_norm": 0.7239896059036255,
+      "learning_rate": 2.797641626200543e-06,
+      "loss": 0.5837,
+      "step": 10061
+    },
+    {
+      "epoch": 2.7872576177285318,
+      "grad_norm": 0.7858687043190002,
+      "learning_rate": 2.797279929169148e-06,
+      "loss": 0.6311,
+      "step": 10062
+    },
+    {
+      "epoch": 2.787534626038781,
+      "grad_norm": 0.7397139668464661,
+      "learning_rate": 2.796918225825741e-06,
+      "loss": 0.5855,
+      "step": 10063
+    },
+    {
+      "epoch": 2.7878116343490307,
+      "grad_norm": 0.7133927941322327,
+      "learning_rate": 2.7965565161780017e-06,
+      "loss": 0.6059,
+      "step": 10064
+    },
+    {
+      "epoch": 2.7880886426592797,
+      "grad_norm": 0.7488803267478943,
+      "learning_rate": 2.796194800233611e-06,
+      "loss": 0.5584,
+      "step": 10065
+    },
+    {
+      "epoch": 2.788365650969529,
+      "grad_norm": 0.7250664234161377,
+      "learning_rate": 2.7958330780002472e-06,
+      "loss": 0.5363,
+      "step": 10066
+    },
+    {
+      "epoch": 2.7886426592797786,
+      "grad_norm": 0.7320556640625,
+      "learning_rate": 2.7954713494855924e-06,
+      "loss": 0.5771,
+      "step": 10067
+    },
+    {
+      "epoch": 2.7889196675900276,
+      "grad_norm": 0.7644149661064148,
+      "learning_rate": 2.7951096146973263e-06,
+      "loss": 0.5977,
+      "step": 10068
+    },
+    {
+      "epoch": 2.789196675900277,
+      "grad_norm": 0.6871392726898193,
+      "learning_rate": 2.7947478736431293e-06,
+      "loss": 0.5614,
+      "step": 10069
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.731040358543396,
+      "learning_rate": 2.7943861263306822e-06,
+      "loss": 0.5968,
+      "step": 10070
+    },
+    {
+      "epoch": 2.7897506925207756,
+      "grad_norm": 0.7525355815887451,
+      "learning_rate": 2.7940243727676653e-06,
+      "loss": 0.5914,
+      "step": 10071
+    },
+    {
+      "epoch": 2.790027700831025,
+      "grad_norm": 0.7682839632034302,
+      "learning_rate": 2.793662612961762e-06,
+      "loss": 0.6167,
+      "step": 10072
+    },
+    {
+      "epoch": 2.7903047091412745,
+      "grad_norm": 0.750568151473999,
+      "learning_rate": 2.793300846920649e-06,
+      "loss": 0.5705,
+      "step": 10073
+    },
+    {
+      "epoch": 2.7905817174515235,
+      "grad_norm": 0.771504282951355,
+      "learning_rate": 2.7929390746520117e-06,
+      "loss": 0.5881,
+      "step": 10074
+    },
+    {
+      "epoch": 2.790858725761773,
+      "grad_norm": 0.753092885017395,
+      "learning_rate": 2.7925772961635294e-06,
+      "loss": 0.6019,
+      "step": 10075
+    },
+    {
+      "epoch": 2.7911357340720224,
+      "grad_norm": 0.7207103371620178,
+      "learning_rate": 2.7922155114628834e-06,
+      "loss": 0.5513,
+      "step": 10076
+    },
+    {
+      "epoch": 2.7914127423822714,
+      "grad_norm": 0.80680251121521,
+      "learning_rate": 2.791853720557756e-06,
+      "loss": 0.6031,
+      "step": 10077
+    },
+    {
+      "epoch": 2.791689750692521,
+      "grad_norm": 0.7253208160400391,
+      "learning_rate": 2.79149192345583e-06,
+      "loss": 0.5972,
+      "step": 10078
+    },
+    {
+      "epoch": 2.7919667590027704,
+      "grad_norm": 0.738048791885376,
+      "learning_rate": 2.7911301201647846e-06,
+      "loss": 0.5545,
+      "step": 10079
+    },
+    {
+      "epoch": 2.7922437673130194,
+      "grad_norm": 0.798290491104126,
+      "learning_rate": 2.7907683106923034e-06,
+      "loss": 0.5873,
+      "step": 10080
+    },
+    {
+      "epoch": 2.792520775623269,
+      "grad_norm": 0.753075122833252,
+      "learning_rate": 2.7904064950460696e-06,
+      "loss": 0.5438,
+      "step": 10081
+    },
+    {
+      "epoch": 2.792797783933518,
+      "grad_norm": 0.7435222268104553,
+      "learning_rate": 2.7900446732337634e-06,
+      "loss": 0.5823,
+      "step": 10082
+    },
+    {
+      "epoch": 2.7930747922437673,
+      "grad_norm": 0.7120912671089172,
+      "learning_rate": 2.789682845263068e-06,
+      "loss": 0.5491,
+      "step": 10083
+    },
+    {
+      "epoch": 2.7933518005540168,
+      "grad_norm": 0.7340899109840393,
+      "learning_rate": 2.7893210111416665e-06,
+      "loss": 0.6199,
+      "step": 10084
+    },
+    {
+      "epoch": 2.793628808864266,
+      "grad_norm": 0.720433235168457,
+      "learning_rate": 2.7889591708772405e-06,
+      "loss": 0.5926,
+      "step": 10085
+    },
+    {
+      "epoch": 2.7939058171745152,
+      "grad_norm": 0.7632058262825012,
+      "learning_rate": 2.788597324477474e-06,
+      "loss": 0.5928,
+      "step": 10086
+    },
+    {
+      "epoch": 2.7941828254847643,
+      "grad_norm": 0.7659016847610474,
+      "learning_rate": 2.7882354719500485e-06,
+      "loss": 0.6098,
+      "step": 10087
+    },
+    {
+      "epoch": 2.7944598337950137,
+      "grad_norm": 0.7464661002159119,
+      "learning_rate": 2.787873613302649e-06,
+      "loss": 0.615,
+      "step": 10088
+    },
+    {
+      "epoch": 2.794736842105263,
+      "grad_norm": 0.7274541854858398,
+      "learning_rate": 2.787511748542957e-06,
+      "loss": 0.5968,
+      "step": 10089
+    },
+    {
+      "epoch": 2.795013850415512,
+      "grad_norm": 0.7382057905197144,
+      "learning_rate": 2.787149877678656e-06,
+      "loss": 0.607,
+      "step": 10090
+    },
+    {
+      "epoch": 2.7952908587257617,
+      "grad_norm": 0.7254559397697449,
+      "learning_rate": 2.7867880007174295e-06,
+      "loss": 0.5698,
+      "step": 10091
+    },
+    {
+      "epoch": 2.795567867036011,
+      "grad_norm": 0.7147872447967529,
+      "learning_rate": 2.7864261176669625e-06,
+      "loss": 0.5256,
+      "step": 10092
+    },
+    {
+      "epoch": 2.79584487534626,
+      "grad_norm": 0.7616999745368958,
+      "learning_rate": 2.7860642285349366e-06,
+      "loss": 0.6358,
+      "step": 10093
+    },
+    {
+      "epoch": 2.7961218836565096,
+      "grad_norm": 0.7379427552223206,
+      "learning_rate": 2.7857023333290377e-06,
+      "loss": 0.6253,
+      "step": 10094
+    },
+    {
+      "epoch": 2.796398891966759,
+      "grad_norm": 0.74202960729599,
+      "learning_rate": 2.7853404320569477e-06,
+      "loss": 0.5957,
+      "step": 10095
+    },
+    {
+      "epoch": 2.796675900277008,
+      "grad_norm": 0.7566871643066406,
+      "learning_rate": 2.7849785247263515e-06,
+      "loss": 0.5535,
+      "step": 10096
+    },
+    {
+      "epoch": 2.7969529085872575,
+      "grad_norm": 0.7378758788108826,
+      "learning_rate": 2.7846166113449345e-06,
+      "loss": 0.5844,
+      "step": 10097
+    },
+    {
+      "epoch": 2.797229916897507,
+      "grad_norm": 0.696129322052002,
+      "learning_rate": 2.7842546919203795e-06,
+      "loss": 0.5659,
+      "step": 10098
+    },
+    {
+      "epoch": 2.797506925207756,
+      "grad_norm": 0.7553945779800415,
+      "learning_rate": 2.783892766460371e-06,
+      "loss": 0.6302,
+      "step": 10099
+    },
+    {
+      "epoch": 2.7977839335180055,
+      "grad_norm": 0.7305325865745544,
+      "learning_rate": 2.783530834972594e-06,
+      "loss": 0.5851,
+      "step": 10100
+    },
+    {
+      "epoch": 2.798060941828255,
+      "grad_norm": 0.754202127456665,
+      "learning_rate": 2.783168897464734e-06,
+      "loss": 0.6049,
+      "step": 10101
+    },
+    {
+      "epoch": 2.798337950138504,
+      "grad_norm": 0.7164472341537476,
+      "learning_rate": 2.782806953944475e-06,
+      "loss": 0.591,
+      "step": 10102
+    },
+    {
+      "epoch": 2.7986149584487534,
+      "grad_norm": 0.7178779244422913,
+      "learning_rate": 2.782445004419503e-06,
+      "loss": 0.5708,
+      "step": 10103
+    },
+    {
+      "epoch": 2.798891966759003,
+      "grad_norm": 0.727104663848877,
+      "learning_rate": 2.7820830488975004e-06,
+      "loss": 0.5752,
+      "step": 10104
+    },
+    {
+      "epoch": 2.799168975069252,
+      "grad_norm": 0.7736073136329651,
+      "learning_rate": 2.7817210873861555e-06,
+      "loss": 0.5803,
+      "step": 10105
+    },
+    {
+      "epoch": 2.7994459833795013,
+      "grad_norm": 0.7395794987678528,
+      "learning_rate": 2.7813591198931524e-06,
+      "loss": 0.5339,
+      "step": 10106
+    },
+    {
+      "epoch": 2.799722991689751,
+      "grad_norm": 0.7755144834518433,
+      "learning_rate": 2.7809971464261763e-06,
+      "loss": 0.6031,
+      "step": 10107
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7793182134628296,
+      "learning_rate": 2.780635166992913e-06,
+      "loss": 0.5834,
+      "step": 10108
+    },
+    {
+      "epoch": 2.8002770083102493,
+      "grad_norm": 0.7401881814002991,
+      "learning_rate": 2.7802731816010494e-06,
+      "loss": 0.5644,
+      "step": 10109
+    },
+    {
+      "epoch": 2.8005540166204987,
+      "grad_norm": 0.7597428560256958,
+      "learning_rate": 2.7799111902582697e-06,
+      "loss": 0.6007,
+      "step": 10110
+    },
+    {
+      "epoch": 2.8008310249307478,
+      "grad_norm": 0.7522040605545044,
+      "learning_rate": 2.7795491929722612e-06,
+      "loss": 0.5958,
+      "step": 10111
+    },
+    {
+      "epoch": 2.801108033240997,
+      "grad_norm": 0.7653687596321106,
+      "learning_rate": 2.779187189750709e-06,
+      "loss": 0.6162,
+      "step": 10112
+    },
+    {
+      "epoch": 2.8013850415512467,
+      "grad_norm": 0.7406167387962341,
+      "learning_rate": 2.7788251806013e-06,
+      "loss": 0.5886,
+      "step": 10113
+    },
+    {
+      "epoch": 2.8016620498614957,
+      "grad_norm": 0.7772331833839417,
+      "learning_rate": 2.77846316553172e-06,
+      "loss": 0.5792,
+      "step": 10114
+    },
+    {
+      "epoch": 2.801939058171745,
+      "grad_norm": 0.7543328404426575,
+      "learning_rate": 2.7781011445496562e-06,
+      "loss": 0.6009,
+      "step": 10115
+    },
+    {
+      "epoch": 2.8022160664819946,
+      "grad_norm": 0.7309496998786926,
+      "learning_rate": 2.777739117662795e-06,
+      "loss": 0.5857,
+      "step": 10116
+    },
+    {
+      "epoch": 2.8024930747922436,
+      "grad_norm": 0.7182099223136902,
+      "learning_rate": 2.7773770848788227e-06,
+      "loss": 0.6056,
+      "step": 10117
+    },
+    {
+      "epoch": 2.802770083102493,
+      "grad_norm": 0.8034688830375671,
+      "learning_rate": 2.7770150462054262e-06,
+      "loss": 0.5207,
+      "step": 10118
+    },
+    {
+      "epoch": 2.8030470914127426,
+      "grad_norm": 0.7384365200996399,
+      "learning_rate": 2.7766530016502936e-06,
+      "loss": 0.5801,
+      "step": 10119
+    },
+    {
+      "epoch": 2.8033240997229916,
+      "grad_norm": 0.7842321991920471,
+      "learning_rate": 2.776290951221111e-06,
+      "loss": 0.6051,
+      "step": 10120
+    },
+    {
+      "epoch": 2.803601108033241,
+      "grad_norm": 0.7223963737487793,
+      "learning_rate": 2.775928894925566e-06,
+      "loss": 0.6145,
+      "step": 10121
+    },
+    {
+      "epoch": 2.8038781163434905,
+      "grad_norm": 0.7191910147666931,
+      "learning_rate": 2.775566832771346e-06,
+      "loss": 0.5643,
+      "step": 10122
+    },
+    {
+      "epoch": 2.8041551246537395,
+      "grad_norm": 0.6896896958351135,
+      "learning_rate": 2.7752047647661384e-06,
+      "loss": 0.5521,
+      "step": 10123
+    },
+    {
+      "epoch": 2.804432132963989,
+      "grad_norm": 0.7741313576698303,
+      "learning_rate": 2.7748426909176308e-06,
+      "loss": 0.5841,
+      "step": 10124
+    },
+    {
+      "epoch": 2.8047091412742384,
+      "grad_norm": 0.7306196689605713,
+      "learning_rate": 2.7744806112335104e-06,
+      "loss": 0.5889,
+      "step": 10125
+    },
+    {
+      "epoch": 2.8049861495844874,
+      "grad_norm": 0.7337019443511963,
+      "learning_rate": 2.774118525721467e-06,
+      "loss": 0.559,
+      "step": 10126
+    },
+    {
+      "epoch": 2.805263157894737,
+      "grad_norm": 0.7309577465057373,
+      "learning_rate": 2.773756434389186e-06,
+      "loss": 0.5988,
+      "step": 10127
+    },
+    {
+      "epoch": 2.8055401662049864,
+      "grad_norm": 0.7765200734138489,
+      "learning_rate": 2.7733943372443572e-06,
+      "loss": 0.6217,
+      "step": 10128
+    },
+    {
+      "epoch": 2.8058171745152354,
+      "grad_norm": 0.7576752305030823,
+      "learning_rate": 2.773032234294669e-06,
+      "loss": 0.5931,
+      "step": 10129
+    },
+    {
+      "epoch": 2.806094182825485,
+      "grad_norm": 0.7484496235847473,
+      "learning_rate": 2.772670125547809e-06,
+      "loss": 0.5675,
+      "step": 10130
+    },
+    {
+      "epoch": 2.8063711911357343,
+      "grad_norm": 0.7473483085632324,
+      "learning_rate": 2.7723080110114655e-06,
+      "loss": 0.6268,
+      "step": 10131
+    },
+    {
+      "epoch": 2.8066481994459833,
+      "grad_norm": 0.7621675729751587,
+      "learning_rate": 2.7719458906933277e-06,
+      "loss": 0.5431,
+      "step": 10132
+    },
+    {
+      "epoch": 2.8069252077562328,
+      "grad_norm": 0.7543573379516602,
+      "learning_rate": 2.771583764601084e-06,
+      "loss": 0.5843,
+      "step": 10133
+    },
+    {
+      "epoch": 2.8072022160664822,
+      "grad_norm": 0.7247941493988037,
+      "learning_rate": 2.771221632742423e-06,
+      "loss": 0.6032,
+      "step": 10134
+    },
+    {
+      "epoch": 2.8074792243767313,
+      "grad_norm": 0.7635888457298279,
+      "learning_rate": 2.7708594951250358e-06,
+      "loss": 0.5738,
+      "step": 10135
+    },
+    {
+      "epoch": 2.8077562326869807,
+      "grad_norm": 0.7076152563095093,
+      "learning_rate": 2.770497351756608e-06,
+      "loss": 0.5828,
+      "step": 10136
+    },
+    {
+      "epoch": 2.80803324099723,
+      "grad_norm": 0.7250659465789795,
+      "learning_rate": 2.7701352026448313e-06,
+      "loss": 0.6096,
+      "step": 10137
+    },
+    {
+      "epoch": 2.808310249307479,
+      "grad_norm": 0.7596719264984131,
+      "learning_rate": 2.769773047797395e-06,
+      "loss": 0.6277,
+      "step": 10138
+    },
+    {
+      "epoch": 2.8085872576177286,
+      "grad_norm": 0.7516382336616516,
+      "learning_rate": 2.7694108872219873e-06,
+      "loss": 0.5905,
+      "step": 10139
+    },
+    {
+      "epoch": 2.808864265927978,
+      "grad_norm": 0.7437866926193237,
+      "learning_rate": 2.7690487209262985e-06,
+      "loss": 0.5777,
+      "step": 10140
+    },
+    {
+      "epoch": 2.809141274238227,
+      "grad_norm": 0.7432383298873901,
+      "learning_rate": 2.7686865489180183e-06,
+      "loss": 0.5942,
+      "step": 10141
+    },
+    {
+      "epoch": 2.8094182825484766,
+      "grad_norm": 0.7514076828956604,
+      "learning_rate": 2.768324371204837e-06,
+      "loss": 0.6108,
+      "step": 10142
+    },
+    {
+      "epoch": 2.8096952908587256,
+      "grad_norm": 0.7362116575241089,
+      "learning_rate": 2.7679621877944434e-06,
+      "loss": 0.5703,
+      "step": 10143
+    },
+    {
+      "epoch": 2.809972299168975,
+      "grad_norm": 0.7133225202560425,
+      "learning_rate": 2.7675999986945282e-06,
+      "loss": 0.5773,
+      "step": 10144
+    },
+    {
+      "epoch": 2.8102493074792245,
+      "grad_norm": 0.7191652059555054,
+      "learning_rate": 2.767237803912783e-06,
+      "loss": 0.595,
+      "step": 10145
+    },
+    {
+      "epoch": 2.8105263157894735,
+      "grad_norm": 0.7200401425361633,
+      "learning_rate": 2.766875603456895e-06,
+      "loss": 0.5711,
+      "step": 10146
+    },
+    {
+      "epoch": 2.810803324099723,
+      "grad_norm": 0.7562271356582642,
+      "learning_rate": 2.766513397334557e-06,
+      "loss": 0.6237,
+      "step": 10147
+    },
+    {
+      "epoch": 2.8110803324099725,
+      "grad_norm": 0.735927402973175,
+      "learning_rate": 2.7661511855534595e-06,
+      "loss": 0.5886,
+      "step": 10148
+    },
+    {
+      "epoch": 2.8113573407202215,
+      "grad_norm": 0.7801263928413391,
+      "learning_rate": 2.765788968121292e-06,
+      "loss": 0.5874,
+      "step": 10149
+    },
+    {
+      "epoch": 2.811634349030471,
+      "grad_norm": 0.6956904530525208,
+      "learning_rate": 2.7654267450457463e-06,
+      "loss": 0.5136,
+      "step": 10150
+    },
+    {
+      "epoch": 2.81191135734072,
+      "grad_norm": 0.7640595436096191,
+      "learning_rate": 2.765064516334513e-06,
+      "loss": 0.6038,
+      "step": 10151
+    },
+    {
+      "epoch": 2.8121883656509694,
+      "grad_norm": 0.7327202558517456,
+      "learning_rate": 2.7647022819952833e-06,
+      "loss": 0.5634,
+      "step": 10152
+    },
+    {
+      "epoch": 2.812465373961219,
+      "grad_norm": 0.7656522989273071,
+      "learning_rate": 2.7643400420357475e-06,
+      "loss": 0.6169,
+      "step": 10153
+    },
+    {
+      "epoch": 2.812742382271468,
+      "grad_norm": 0.7299674153327942,
+      "learning_rate": 2.763977796463599e-06,
+      "loss": 0.6227,
+      "step": 10154
+    },
+    {
+      "epoch": 2.8130193905817173,
+      "grad_norm": 0.7286268472671509,
+      "learning_rate": 2.763615545286526e-06,
+      "loss": 0.6531,
+      "step": 10155
+    },
+    {
+      "epoch": 2.813296398891967,
+      "grad_norm": 0.7793086171150208,
+      "learning_rate": 2.7632532885122227e-06,
+      "loss": 0.5584,
+      "step": 10156
+    },
+    {
+      "epoch": 2.813573407202216,
+      "grad_norm": 0.7282178997993469,
+      "learning_rate": 2.76289102614838e-06,
+      "loss": 0.6217,
+      "step": 10157
+    },
+    {
+      "epoch": 2.8138504155124653,
+      "grad_norm": 0.7371731400489807,
+      "learning_rate": 2.7625287582026888e-06,
+      "loss": 0.6266,
+      "step": 10158
+    },
+    {
+      "epoch": 2.8141274238227147,
+      "grad_norm": 0.7733290195465088,
+      "learning_rate": 2.7621664846828415e-06,
+      "loss": 0.6445,
+      "step": 10159
+    },
+    {
+      "epoch": 2.8144044321329638,
+      "grad_norm": 0.7712571024894714,
+      "learning_rate": 2.761804205596531e-06,
+      "loss": 0.6249,
+      "step": 10160
+    },
+    {
+      "epoch": 2.8146814404432132,
+      "grad_norm": 0.7427979111671448,
+      "learning_rate": 2.761441920951448e-06,
+      "loss": 0.6307,
+      "step": 10161
+    },
+    {
+      "epoch": 2.8149584487534627,
+      "grad_norm": 0.7213217616081238,
+      "learning_rate": 2.7610796307552854e-06,
+      "loss": 0.5625,
+      "step": 10162
+    },
+    {
+      "epoch": 2.8152354570637117,
+      "grad_norm": 0.7439704537391663,
+      "learning_rate": 2.7607173350157357e-06,
+      "loss": 0.5922,
+      "step": 10163
+    },
+    {
+      "epoch": 2.815512465373961,
+      "grad_norm": 0.7900652289390564,
+      "learning_rate": 2.7603550337404916e-06,
+      "loss": 0.6083,
+      "step": 10164
+    },
+    {
+      "epoch": 2.8157894736842106,
+      "grad_norm": 0.7682667970657349,
+      "learning_rate": 2.7599927269372453e-06,
+      "loss": 0.5536,
+      "step": 10165
+    },
+    {
+      "epoch": 2.8160664819944596,
+      "grad_norm": 0.7357151508331299,
+      "learning_rate": 2.7596304146136883e-06,
+      "loss": 0.6209,
+      "step": 10166
+    },
+    {
+      "epoch": 2.816343490304709,
+      "grad_norm": 0.7456122636795044,
+      "learning_rate": 2.759268096777516e-06,
+      "loss": 0.5957,
+      "step": 10167
+    },
+    {
+      "epoch": 2.8166204986149586,
+      "grad_norm": 0.7296216487884521,
+      "learning_rate": 2.758905773436419e-06,
+      "loss": 0.6071,
+      "step": 10168
+    },
+    {
+      "epoch": 2.8168975069252076,
+      "grad_norm": 0.7860580682754517,
+      "learning_rate": 2.758543444598091e-06,
+      "loss": 0.6382,
+      "step": 10169
+    },
+    {
+      "epoch": 2.817174515235457,
+      "grad_norm": 0.7770671844482422,
+      "learning_rate": 2.758181110270227e-06,
+      "loss": 0.6017,
+      "step": 10170
+    },
+    {
+      "epoch": 2.8174515235457065,
+      "grad_norm": 0.6933687329292297,
+      "learning_rate": 2.7578187704605176e-06,
+      "loss": 0.5789,
+      "step": 10171
+    },
+    {
+      "epoch": 2.8177285318559555,
+      "grad_norm": 0.722878098487854,
+      "learning_rate": 2.757456425176657e-06,
+      "loss": 0.6016,
+      "step": 10172
+    },
+    {
+      "epoch": 2.818005540166205,
+      "grad_norm": 0.7481485605239868,
+      "learning_rate": 2.757094074426339e-06,
+      "loss": 0.6036,
+      "step": 10173
+    },
+    {
+      "epoch": 2.8182825484764544,
+      "grad_norm": 0.7612590193748474,
+      "learning_rate": 2.7567317182172583e-06,
+      "loss": 0.5457,
+      "step": 10174
+    },
+    {
+      "epoch": 2.8185595567867034,
+      "grad_norm": 0.7551214694976807,
+      "learning_rate": 2.756369356557106e-06,
+      "loss": 0.6061,
+      "step": 10175
+    },
+    {
+      "epoch": 2.818836565096953,
+      "grad_norm": 0.7312377691268921,
+      "learning_rate": 2.7560069894535783e-06,
+      "loss": 0.5956,
+      "step": 10176
+    },
+    {
+      "epoch": 2.8191135734072024,
+      "grad_norm": 0.7991149425506592,
+      "learning_rate": 2.755644616914369e-06,
+      "loss": 0.5719,
+      "step": 10177
+    },
+    {
+      "epoch": 2.8193905817174514,
+      "grad_norm": 0.741660475730896,
+      "learning_rate": 2.7552822389471716e-06,
+      "loss": 0.5659,
+      "step": 10178
+    },
+    {
+      "epoch": 2.819667590027701,
+      "grad_norm": 0.7375673055648804,
+      "learning_rate": 2.7549198555596794e-06,
+      "loss": 0.5978,
+      "step": 10179
+    },
+    {
+      "epoch": 2.8199445983379503,
+      "grad_norm": 0.7576779723167419,
+      "learning_rate": 2.7545574667595894e-06,
+      "loss": 0.5936,
+      "step": 10180
+    },
+    {
+      "epoch": 2.8202216066481993,
+      "grad_norm": 0.7339960932731628,
+      "learning_rate": 2.7541950725545924e-06,
+      "loss": 0.6356,
+      "step": 10181
+    },
+    {
+      "epoch": 2.820498614958449,
+      "grad_norm": 0.722320556640625,
+      "learning_rate": 2.7538326729523858e-06,
+      "loss": 0.57,
+      "step": 10182
+    },
+    {
+      "epoch": 2.8207756232686982,
+      "grad_norm": 0.7453545928001404,
+      "learning_rate": 2.7534702679606635e-06,
+      "loss": 0.6029,
+      "step": 10183
+    },
+    {
+      "epoch": 2.8210526315789473,
+      "grad_norm": 0.7736693620681763,
+      "learning_rate": 2.7531078575871196e-06,
+      "loss": 0.5681,
+      "step": 10184
+    },
+    {
+      "epoch": 2.8213296398891967,
+      "grad_norm": 0.734969437122345,
+      "learning_rate": 2.7527454418394494e-06,
+      "loss": 0.5816,
+      "step": 10185
+    },
+    {
+      "epoch": 2.821606648199446,
+      "grad_norm": 0.7231078743934631,
+      "learning_rate": 2.7523830207253494e-06,
+      "loss": 0.5659,
+      "step": 10186
+    },
+    {
+      "epoch": 2.821883656509695,
+      "grad_norm": 0.7357196807861328,
+      "learning_rate": 2.752020594252511e-06,
+      "loss": 0.6238,
+      "step": 10187
+    },
+    {
+      "epoch": 2.8221606648199447,
+      "grad_norm": 0.7271388173103333,
+      "learning_rate": 2.7516581624286337e-06,
+      "loss": 0.5628,
+      "step": 10188
+    },
+    {
+      "epoch": 2.822437673130194,
+      "grad_norm": 0.7925208806991577,
+      "learning_rate": 2.7512957252614108e-06,
+      "loss": 0.5855,
+      "step": 10189
+    },
+    {
+      "epoch": 2.822714681440443,
+      "grad_norm": 0.7478787899017334,
+      "learning_rate": 2.7509332827585367e-06,
+      "loss": 0.5793,
+      "step": 10190
+    },
+    {
+      "epoch": 2.8229916897506926,
+      "grad_norm": 0.7537767887115479,
+      "learning_rate": 2.7505708349277093e-06,
+      "loss": 0.6236,
+      "step": 10191
+    },
+    {
+      "epoch": 2.823268698060942,
+      "grad_norm": 0.775567889213562,
+      "learning_rate": 2.7502083817766227e-06,
+      "loss": 0.6143,
+      "step": 10192
+    },
+    {
+      "epoch": 2.823545706371191,
+      "grad_norm": 0.744401752948761,
+      "learning_rate": 2.7498459233129738e-06,
+      "loss": 0.6268,
+      "step": 10193
+    },
+    {
+      "epoch": 2.8238227146814405,
+      "grad_norm": 0.762998640537262,
+      "learning_rate": 2.749483459544457e-06,
+      "loss": 0.5823,
+      "step": 10194
+    },
+    {
+      "epoch": 2.82409972299169,
+      "grad_norm": 0.7275715470314026,
+      "learning_rate": 2.7491209904787695e-06,
+      "loss": 0.57,
+      "step": 10195
+    },
+    {
+      "epoch": 2.824376731301939,
+      "grad_norm": 0.6960054039955139,
+      "learning_rate": 2.7487585161236076e-06,
+      "loss": 0.5769,
+      "step": 10196
+    },
+    {
+      "epoch": 2.8246537396121885,
+      "grad_norm": 0.7539833188056946,
+      "learning_rate": 2.748396036486667e-06,
+      "loss": 0.5861,
+      "step": 10197
+    },
+    {
+      "epoch": 2.824930747922438,
+      "grad_norm": 0.7374334931373596,
+      "learning_rate": 2.748033551575644e-06,
+      "loss": 0.584,
+      "step": 10198
+    },
+    {
+      "epoch": 2.825207756232687,
+      "grad_norm": 0.7197996973991394,
+      "learning_rate": 2.747671061398236e-06,
+      "loss": 0.6054,
+      "step": 10199
+    },
+    {
+      "epoch": 2.8254847645429364,
+      "grad_norm": 0.7191854119300842,
+      "learning_rate": 2.747308565962138e-06,
+      "loss": 0.5984,
+      "step": 10200
+    },
+    {
+      "epoch": 2.825761772853186,
+      "grad_norm": 0.7698421478271484,
+      "learning_rate": 2.7469460652750477e-06,
+      "loss": 0.5802,
+      "step": 10201
+    },
+    {
+      "epoch": 2.826038781163435,
+      "grad_norm": 0.7502759099006653,
+      "learning_rate": 2.7465835593446622e-06,
+      "loss": 0.6245,
+      "step": 10202
+    },
+    {
+      "epoch": 2.8263157894736843,
+      "grad_norm": 0.7236973643302917,
+      "learning_rate": 2.7462210481786776e-06,
+      "loss": 0.5862,
+      "step": 10203
+    },
+    {
+      "epoch": 2.8265927977839334,
+      "grad_norm": 0.676241397857666,
+      "learning_rate": 2.745858531784791e-06,
+      "loss": 0.556,
+      "step": 10204
+    },
+    {
+      "epoch": 2.826869806094183,
+      "grad_norm": 0.7392049431800842,
+      "learning_rate": 2.7454960101707012e-06,
+      "loss": 0.5983,
+      "step": 10205
+    },
+    {
+      "epoch": 2.8271468144044323,
+      "grad_norm": 0.7158267498016357,
+      "learning_rate": 2.7451334833441035e-06,
+      "loss": 0.5992,
+      "step": 10206
+    },
+    {
+      "epoch": 2.8274238227146813,
+      "grad_norm": 0.7335191369056702,
+      "learning_rate": 2.744770951312696e-06,
+      "loss": 0.6643,
+      "step": 10207
+    },
+    {
+      "epoch": 2.8277008310249307,
+      "grad_norm": 0.707643985748291,
+      "learning_rate": 2.7444084140841766e-06,
+      "loss": 0.5584,
+      "step": 10208
+    },
+    {
+      "epoch": 2.82797783933518,
+      "grad_norm": 0.7364636063575745,
+      "learning_rate": 2.744045871666241e-06,
+      "loss": 0.5815,
+      "step": 10209
+    },
+    {
+      "epoch": 2.8282548476454292,
+      "grad_norm": 0.7579964995384216,
+      "learning_rate": 2.7436833240665892e-06,
+      "loss": 0.5676,
+      "step": 10210
+    },
+    {
+      "epoch": 2.8285318559556787,
+      "grad_norm": 0.773270845413208,
+      "learning_rate": 2.743320771292919e-06,
+      "loss": 0.6214,
+      "step": 10211
+    },
+    {
+      "epoch": 2.8288088642659277,
+      "grad_norm": 0.7789325714111328,
+      "learning_rate": 2.7429582133529266e-06,
+      "loss": 0.5636,
+      "step": 10212
+    },
+    {
+      "epoch": 2.829085872576177,
+      "grad_norm": 0.719110369682312,
+      "learning_rate": 2.74259565025431e-06,
+      "loss": 0.5672,
+      "step": 10213
+    },
+    {
+      "epoch": 2.8293628808864266,
+      "grad_norm": 0.7776400446891785,
+      "learning_rate": 2.7422330820047693e-06,
+      "loss": 0.5938,
+      "step": 10214
+    },
+    {
+      "epoch": 2.8296398891966756,
+      "grad_norm": 0.7782211899757385,
+      "learning_rate": 2.7418705086120018e-06,
+      "loss": 0.6078,
+      "step": 10215
+    },
+    {
+      "epoch": 2.829916897506925,
+      "grad_norm": 0.7914127111434937,
+      "learning_rate": 2.7415079300837056e-06,
+      "loss": 0.6053,
+      "step": 10216
+    },
+    {
+      "epoch": 2.8301939058171746,
+      "grad_norm": 0.7330632209777832,
+      "learning_rate": 2.7411453464275783e-06,
+      "loss": 0.5704,
+      "step": 10217
+    },
+    {
+      "epoch": 2.8304709141274236,
+      "grad_norm": 0.7698874473571777,
+      "learning_rate": 2.74078275765132e-06,
+      "loss": 0.5881,
+      "step": 10218
+    },
+    {
+      "epoch": 2.830747922437673,
+      "grad_norm": 0.7458370923995972,
+      "learning_rate": 2.74042016376263e-06,
+      "loss": 0.6182,
+      "step": 10219
+    },
+    {
+      "epoch": 2.8310249307479225,
+      "grad_norm": 0.7117974758148193,
+      "learning_rate": 2.7400575647692046e-06,
+      "loss": 0.5814,
+      "step": 10220
+    },
+    {
+      "epoch": 2.8313019390581715,
+      "grad_norm": 0.7336068153381348,
+      "learning_rate": 2.739694960678745e-06,
+      "loss": 0.5504,
+      "step": 10221
+    },
+    {
+      "epoch": 2.831578947368421,
+      "grad_norm": 0.7284829020500183,
+      "learning_rate": 2.7393323514989485e-06,
+      "loss": 0.5956,
+      "step": 10222
+    },
+    {
+      "epoch": 2.8318559556786704,
+      "grad_norm": 0.7642632722854614,
+      "learning_rate": 2.738969737237515e-06,
+      "loss": 0.6281,
+      "step": 10223
+    },
+    {
+      "epoch": 2.8321329639889194,
+      "grad_norm": 0.754257321357727,
+      "learning_rate": 2.7386071179021445e-06,
+      "loss": 0.569,
+      "step": 10224
+    },
+    {
+      "epoch": 2.832409972299169,
+      "grad_norm": 0.7193533182144165,
+      "learning_rate": 2.7382444935005353e-06,
+      "loss": 0.5377,
+      "step": 10225
+    },
+    {
+      "epoch": 2.8326869806094184,
+      "grad_norm": 0.784979522228241,
+      "learning_rate": 2.7378818640403865e-06,
+      "loss": 0.6296,
+      "step": 10226
+    },
+    {
+      "epoch": 2.8329639889196674,
+      "grad_norm": 0.75473552942276,
+      "learning_rate": 2.7375192295293984e-06,
+      "loss": 0.6027,
+      "step": 10227
+    },
+    {
+      "epoch": 2.833240997229917,
+      "grad_norm": 0.7351518869400024,
+      "learning_rate": 2.737156589975271e-06,
+      "loss": 0.5944,
+      "step": 10228
+    },
+    {
+      "epoch": 2.8335180055401663,
+      "grad_norm": 0.7493429780006409,
+      "learning_rate": 2.7367939453857034e-06,
+      "loss": 0.5478,
+      "step": 10229
+    },
+    {
+      "epoch": 2.8337950138504153,
+      "grad_norm": 0.735837996006012,
+      "learning_rate": 2.736431295768395e-06,
+      "loss": 0.5491,
+      "step": 10230
+    },
+    {
+      "epoch": 2.834072022160665,
+      "grad_norm": 0.7217791676521301,
+      "learning_rate": 2.736068641131047e-06,
+      "loss": 0.5784,
+      "step": 10231
+    },
+    {
+      "epoch": 2.8343490304709142,
+      "grad_norm": 0.730108380317688,
+      "learning_rate": 2.735705981481359e-06,
+      "loss": 0.5939,
+      "step": 10232
+    },
+    {
+      "epoch": 2.8346260387811633,
+      "grad_norm": 0.7806057929992676,
+      "learning_rate": 2.735343316827031e-06,
+      "loss": 0.5877,
+      "step": 10233
+    },
+    {
+      "epoch": 2.8349030470914127,
+      "grad_norm": 0.7238426208496094,
+      "learning_rate": 2.7349806471757634e-06,
+      "loss": 0.6003,
+      "step": 10234
+    },
+    {
+      "epoch": 2.835180055401662,
+      "grad_norm": 0.7456830143928528,
+      "learning_rate": 2.7346179725352562e-06,
+      "loss": 0.5845,
+      "step": 10235
+    },
+    {
+      "epoch": 2.835457063711911,
+      "grad_norm": 0.7526271939277649,
+      "learning_rate": 2.7342552929132106e-06,
+      "loss": 0.5942,
+      "step": 10236
+    },
+    {
+      "epoch": 2.8357340720221607,
+      "grad_norm": 0.7471960186958313,
+      "learning_rate": 2.7338926083173266e-06,
+      "loss": 0.6783,
+      "step": 10237
+    },
+    {
+      "epoch": 2.83601108033241,
+      "grad_norm": 0.7458435297012329,
+      "learning_rate": 2.733529918755306e-06,
+      "loss": 0.6002,
+      "step": 10238
+    },
+    {
+      "epoch": 2.836288088642659,
+      "grad_norm": 0.7718545794487,
+      "learning_rate": 2.733167224234848e-06,
+      "loss": 0.5943,
+      "step": 10239
+    },
+    {
+      "epoch": 2.8365650969529086,
+      "grad_norm": 0.7507277131080627,
+      "learning_rate": 2.7328045247636555e-06,
+      "loss": 0.615,
+      "step": 10240
+    },
+    {
+      "epoch": 2.836842105263158,
+      "grad_norm": 0.7238956689834595,
+      "learning_rate": 2.732441820349427e-06,
+      "loss": 0.602,
+      "step": 10241
+    },
+    {
+      "epoch": 2.837119113573407,
+      "grad_norm": 0.6814334392547607,
+      "learning_rate": 2.7320791109998655e-06,
+      "loss": 0.5912,
+      "step": 10242
+    },
+    {
+      "epoch": 2.8373961218836565,
+      "grad_norm": 0.7484681010246277,
+      "learning_rate": 2.7317163967226722e-06,
+      "loss": 0.5704,
+      "step": 10243
+    },
+    {
+      "epoch": 2.837673130193906,
+      "grad_norm": 0.710078775882721,
+      "learning_rate": 2.731353677525548e-06,
+      "loss": 0.5733,
+      "step": 10244
+    },
+    {
+      "epoch": 2.837950138504155,
+      "grad_norm": 0.7513867616653442,
+      "learning_rate": 2.730990953416194e-06,
+      "loss": 0.5329,
+      "step": 10245
+    },
+    {
+      "epoch": 2.8382271468144045,
+      "grad_norm": 0.7298266887664795,
+      "learning_rate": 2.730628224402313e-06,
+      "loss": 0.563,
+      "step": 10246
+    },
+    {
+      "epoch": 2.838504155124654,
+      "grad_norm": 0.7499540448188782,
+      "learning_rate": 2.7302654904916055e-06,
+      "loss": 0.5934,
+      "step": 10247
+    },
+    {
+      "epoch": 2.838781163434903,
+      "grad_norm": 0.7679840922355652,
+      "learning_rate": 2.7299027516917725e-06,
+      "loss": 0.5884,
+      "step": 10248
+    },
+    {
+      "epoch": 2.8390581717451524,
+      "grad_norm": 0.7211563587188721,
+      "learning_rate": 2.729540008010518e-06,
+      "loss": 0.5442,
+      "step": 10249
+    },
+    {
+      "epoch": 2.839335180055402,
+      "grad_norm": 0.7048432230949402,
+      "learning_rate": 2.729177259455543e-06,
+      "loss": 0.5749,
+      "step": 10250
+    },
+    {
+      "epoch": 2.839612188365651,
+      "grad_norm": 0.7010490298271179,
+      "learning_rate": 2.7288145060345495e-06,
+      "loss": 0.5917,
+      "step": 10251
+    },
+    {
+      "epoch": 2.8398891966759003,
+      "grad_norm": 0.7572813034057617,
+      "learning_rate": 2.7284517477552393e-06,
+      "loss": 0.5683,
+      "step": 10252
+    },
+    {
+      "epoch": 2.84016620498615,
+      "grad_norm": 0.7326565980911255,
+      "learning_rate": 2.728088984625316e-06,
+      "loss": 0.6065,
+      "step": 10253
+    },
+    {
+      "epoch": 2.840443213296399,
+      "grad_norm": 0.7696539163589478,
+      "learning_rate": 2.72772621665248e-06,
+      "loss": 0.6366,
+      "step": 10254
+    },
+    {
+      "epoch": 2.8407202216066483,
+      "grad_norm": 0.711655855178833,
+      "learning_rate": 2.727363443844435e-06,
+      "loss": 0.6076,
+      "step": 10255
+    },
+    {
+      "epoch": 2.8409972299168977,
+      "grad_norm": 0.7610486149787903,
+      "learning_rate": 2.7270006662088842e-06,
+      "loss": 0.5638,
+      "step": 10256
+    },
+    {
+      "epoch": 2.8412742382271468,
+      "grad_norm": 0.7857367992401123,
+      "learning_rate": 2.7266378837535296e-06,
+      "loss": 0.601,
+      "step": 10257
+    },
+    {
+      "epoch": 2.841551246537396,
+      "grad_norm": 0.7206484079360962,
+      "learning_rate": 2.7262750964860727e-06,
+      "loss": 0.5781,
+      "step": 10258
+    },
+    {
+      "epoch": 2.8418282548476457,
+      "grad_norm": 0.7355260848999023,
+      "learning_rate": 2.7259123044142188e-06,
+      "loss": 0.5942,
+      "step": 10259
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.7465602159500122,
+      "learning_rate": 2.7255495075456693e-06,
+      "loss": 0.6028,
+      "step": 10260
+    },
+    {
+      "epoch": 2.842382271468144,
+      "grad_norm": 0.7275111079216003,
+      "learning_rate": 2.7251867058881277e-06,
+      "loss": 0.6069,
+      "step": 10261
+    },
+    {
+      "epoch": 2.8426592797783936,
+      "grad_norm": 0.7320525050163269,
+      "learning_rate": 2.724823899449298e-06,
+      "loss": 0.6253,
+      "step": 10262
+    },
+    {
+      "epoch": 2.8429362880886426,
+      "grad_norm": 0.7653960585594177,
+      "learning_rate": 2.7244610882368815e-06,
+      "loss": 0.5776,
+      "step": 10263
+    },
+    {
+      "epoch": 2.843213296398892,
+      "grad_norm": 0.7047906517982483,
+      "learning_rate": 2.724098272258584e-06,
+      "loss": 0.5906,
+      "step": 10264
+    },
+    {
+      "epoch": 2.8434903047091415,
+      "grad_norm": 0.7334323525428772,
+      "learning_rate": 2.7237354515221075e-06,
+      "loss": 0.5913,
+      "step": 10265
+    },
+    {
+      "epoch": 2.8437673130193906,
+      "grad_norm": 0.7125048041343689,
+      "learning_rate": 2.723372626035156e-06,
+      "loss": 0.5951,
+      "step": 10266
+    },
+    {
+      "epoch": 2.84404432132964,
+      "grad_norm": 0.7758623957633972,
+      "learning_rate": 2.723009795805433e-06,
+      "loss": 0.6011,
+      "step": 10267
+    },
+    {
+      "epoch": 2.844321329639889,
+      "grad_norm": 0.6995490193367004,
+      "learning_rate": 2.722646960840643e-06,
+      "loss": 0.5852,
+      "step": 10268
+    },
+    {
+      "epoch": 2.8445983379501385,
+      "grad_norm": 0.6949390172958374,
+      "learning_rate": 2.7222841211484896e-06,
+      "loss": 0.5442,
+      "step": 10269
+    },
+    {
+      "epoch": 2.844875346260388,
+      "grad_norm": 0.7149237990379333,
+      "learning_rate": 2.721921276736676e-06,
+      "loss": 0.6074,
+      "step": 10270
+    },
+    {
+      "epoch": 2.845152354570637,
+      "grad_norm": 0.6994629502296448,
+      "learning_rate": 2.721558427612907e-06,
+      "loss": 0.6086,
+      "step": 10271
+    },
+    {
+      "epoch": 2.8454293628808864,
+      "grad_norm": 0.7139860987663269,
+      "learning_rate": 2.7211955737848876e-06,
+      "loss": 0.5656,
+      "step": 10272
+    },
+    {
+      "epoch": 2.845706371191136,
+      "grad_norm": 0.7294262647628784,
+      "learning_rate": 2.7208327152603204e-06,
+      "loss": 0.6139,
+      "step": 10273
+    },
+    {
+      "epoch": 2.845983379501385,
+      "grad_norm": 0.7386606335639954,
+      "learning_rate": 2.7204698520469103e-06,
+      "loss": 0.5715,
+      "step": 10274
+    },
+    {
+      "epoch": 2.8462603878116344,
+      "grad_norm": 0.8146941065788269,
+      "learning_rate": 2.720106984152364e-06,
+      "loss": 0.6356,
+      "step": 10275
+    },
+    {
+      "epoch": 2.8465373961218834,
+      "grad_norm": 0.7665870189666748,
+      "learning_rate": 2.7197441115843824e-06,
+      "loss": 0.6195,
+      "step": 10276
+    },
+    {
+      "epoch": 2.846814404432133,
+      "grad_norm": 0.7315971255302429,
+      "learning_rate": 2.719381234350673e-06,
+      "loss": 0.5145,
+      "step": 10277
+    },
+    {
+      "epoch": 2.8470914127423823,
+      "grad_norm": 0.7454190850257874,
+      "learning_rate": 2.7190183524589405e-06,
+      "loss": 0.6395,
+      "step": 10278
+    },
+    {
+      "epoch": 2.8473684210526313,
+      "grad_norm": 0.7426220178604126,
+      "learning_rate": 2.7186554659168886e-06,
+      "loss": 0.5366,
+      "step": 10279
+    },
+    {
+      "epoch": 2.847645429362881,
+      "grad_norm": 0.7703450918197632,
+      "learning_rate": 2.718292574732222e-06,
+      "loss": 0.5854,
+      "step": 10280
+    },
+    {
+      "epoch": 2.8479224376731302,
+      "grad_norm": 0.7275317907333374,
+      "learning_rate": 2.717929678912647e-06,
+      "loss": 0.5968,
+      "step": 10281
+    },
+    {
+      "epoch": 2.8481994459833793,
+      "grad_norm": 0.72344970703125,
+      "learning_rate": 2.717566778465869e-06,
+      "loss": 0.6176,
+      "step": 10282
+    },
+    {
+      "epoch": 2.8484764542936287,
+      "grad_norm": 0.7199895977973938,
+      "learning_rate": 2.717203873399592e-06,
+      "loss": 0.5797,
+      "step": 10283
+    },
+    {
+      "epoch": 2.848753462603878,
+      "grad_norm": 0.725170373916626,
+      "learning_rate": 2.7168409637215213e-06,
+      "loss": 0.6077,
+      "step": 10284
+    },
+    {
+      "epoch": 2.849030470914127,
+      "grad_norm": 0.7106685638427734,
+      "learning_rate": 2.716478049439365e-06,
+      "loss": 0.5501,
+      "step": 10285
+    },
+    {
+      "epoch": 2.8493074792243767,
+      "grad_norm": 0.7496122717857361,
+      "learning_rate": 2.7161151305608258e-06,
+      "loss": 0.6609,
+      "step": 10286
+    },
+    {
+      "epoch": 2.849584487534626,
+      "grad_norm": 0.7406869530677795,
+      "learning_rate": 2.715752207093611e-06,
+      "loss": 0.5644,
+      "step": 10287
+    },
+    {
+      "epoch": 2.849861495844875,
+      "grad_norm": 0.7530012726783752,
+      "learning_rate": 2.7153892790454256e-06,
+      "loss": 0.6045,
+      "step": 10288
+    },
+    {
+      "epoch": 2.8501385041551246,
+      "grad_norm": 0.7549383044242859,
+      "learning_rate": 2.715026346423975e-06,
+      "loss": 0.6313,
+      "step": 10289
+    },
+    {
+      "epoch": 2.850415512465374,
+      "grad_norm": 0.7471486330032349,
+      "learning_rate": 2.7146634092369674e-06,
+      "loss": 0.5472,
+      "step": 10290
+    },
+    {
+      "epoch": 2.850692520775623,
+      "grad_norm": 0.7477738857269287,
+      "learning_rate": 2.7143004674921063e-06,
+      "loss": 0.583,
+      "step": 10291
+    },
+    {
+      "epoch": 2.8509695290858725,
+      "grad_norm": 0.7241801023483276,
+      "learning_rate": 2.7139375211971e-06,
+      "loss": 0.5983,
+      "step": 10292
+    },
+    {
+      "epoch": 2.851246537396122,
+      "grad_norm": 0.7376339435577393,
+      "learning_rate": 2.7135745703596528e-06,
+      "loss": 0.5867,
+      "step": 10293
+    },
+    {
+      "epoch": 2.851523545706371,
+      "grad_norm": 0.7515271306037903,
+      "learning_rate": 2.7132116149874732e-06,
+      "loss": 0.6049,
+      "step": 10294
+    },
+    {
+      "epoch": 2.8518005540166205,
+      "grad_norm": 0.7021607756614685,
+      "learning_rate": 2.7128486550882647e-06,
+      "loss": 0.5832,
+      "step": 10295
+    },
+    {
+      "epoch": 2.85207756232687,
+      "grad_norm": 0.7405130863189697,
+      "learning_rate": 2.7124856906697374e-06,
+      "loss": 0.6111,
+      "step": 10296
+    },
+    {
+      "epoch": 2.852354570637119,
+      "grad_norm": 0.7491231560707092,
+      "learning_rate": 2.7121227217395957e-06,
+      "loss": 0.647,
+      "step": 10297
+    },
+    {
+      "epoch": 2.8526315789473684,
+      "grad_norm": 0.7277181148529053,
+      "learning_rate": 2.711759748305547e-06,
+      "loss": 0.5674,
+      "step": 10298
+    },
+    {
+      "epoch": 2.852908587257618,
+      "grad_norm": 0.6759470701217651,
+      "learning_rate": 2.7113967703752974e-06,
+      "loss": 0.5795,
+      "step": 10299
+    },
+    {
+      "epoch": 2.853185595567867,
+      "grad_norm": 0.7326931953430176,
+      "learning_rate": 2.711033787956555e-06,
+      "loss": 0.6122,
+      "step": 10300
+    },
+    {
+      "epoch": 2.8534626038781163,
+      "grad_norm": 0.7393088340759277,
+      "learning_rate": 2.710670801057027e-06,
+      "loss": 0.5768,
+      "step": 10301
+    },
+    {
+      "epoch": 2.853739612188366,
+      "grad_norm": 0.7498008012771606,
+      "learning_rate": 2.7103078096844187e-06,
+      "loss": 0.5535,
+      "step": 10302
+    },
+    {
+      "epoch": 2.854016620498615,
+      "grad_norm": 0.765217125415802,
+      "learning_rate": 2.7099448138464395e-06,
+      "loss": 0.5641,
+      "step": 10303
+    },
+    {
+      "epoch": 2.8542936288088643,
+      "grad_norm": 0.7652571201324463,
+      "learning_rate": 2.7095818135507957e-06,
+      "loss": 0.5632,
+      "step": 10304
+    },
+    {
+      "epoch": 2.8545706371191137,
+      "grad_norm": 0.7151368856430054,
+      "learning_rate": 2.7092188088051945e-06,
+      "loss": 0.5924,
+      "step": 10305
+    },
+    {
+      "epoch": 2.8548476454293628,
+      "grad_norm": 0.7446417808532715,
+      "learning_rate": 2.7088557996173433e-06,
+      "loss": 0.5704,
+      "step": 10306
+    },
+    {
+      "epoch": 2.855124653739612,
+      "grad_norm": 0.74827641248703,
+      "learning_rate": 2.708492785994951e-06,
+      "loss": 0.6094,
+      "step": 10307
+    },
+    {
+      "epoch": 2.8554016620498617,
+      "grad_norm": 0.7772006392478943,
+      "learning_rate": 2.7081297679457238e-06,
+      "loss": 0.6026,
+      "step": 10308
+    },
+    {
+      "epoch": 2.8556786703601107,
+      "grad_norm": 0.6965993046760559,
+      "learning_rate": 2.7077667454773705e-06,
+      "loss": 0.5866,
+      "step": 10309
+    },
+    {
+      "epoch": 2.85595567867036,
+      "grad_norm": 0.7219796776771545,
+      "learning_rate": 2.707403718597598e-06,
+      "loss": 0.5489,
+      "step": 10310
+    },
+    {
+      "epoch": 2.8562326869806096,
+      "grad_norm": 0.712982714176178,
+      "learning_rate": 2.707040687314116e-06,
+      "loss": 0.595,
+      "step": 10311
+    },
+    {
+      "epoch": 2.8565096952908586,
+      "grad_norm": 0.7534233927726746,
+      "learning_rate": 2.70667765163463e-06,
+      "loss": 0.5422,
+      "step": 10312
+    },
+    {
+      "epoch": 2.856786703601108,
+      "grad_norm": 0.7488660216331482,
+      "learning_rate": 2.706314611566851e-06,
+      "loss": 0.5522,
+      "step": 10313
+    },
+    {
+      "epoch": 2.8570637119113576,
+      "grad_norm": 0.7368558645248413,
+      "learning_rate": 2.705951567118485e-06,
+      "loss": 0.6267,
+      "step": 10314
+    },
+    {
+      "epoch": 2.8573407202216066,
+      "grad_norm": 0.7692279815673828,
+      "learning_rate": 2.7055885182972417e-06,
+      "loss": 0.5865,
+      "step": 10315
+    },
+    {
+      "epoch": 2.857617728531856,
+      "grad_norm": 0.7746422290802002,
+      "learning_rate": 2.7052254651108295e-06,
+      "loss": 0.5904,
+      "step": 10316
+    },
+    {
+      "epoch": 2.8578947368421055,
+      "grad_norm": 0.7562121152877808,
+      "learning_rate": 2.7048624075669554e-06,
+      "loss": 0.6338,
+      "step": 10317
+    },
+    {
+      "epoch": 2.8581717451523545,
+      "grad_norm": 0.7386156320571899,
+      "learning_rate": 2.7044993456733304e-06,
+      "loss": 0.5977,
+      "step": 10318
+    },
+    {
+      "epoch": 2.858448753462604,
+      "grad_norm": 0.7650381326675415,
+      "learning_rate": 2.704136279437662e-06,
+      "loss": 0.5834,
+      "step": 10319
+    },
+    {
+      "epoch": 2.8587257617728534,
+      "grad_norm": 0.7712715864181519,
+      "learning_rate": 2.7037732088676583e-06,
+      "loss": 0.6099,
+      "step": 10320
+    },
+    {
+      "epoch": 2.8590027700831024,
+      "grad_norm": 0.7141051888465881,
+      "learning_rate": 2.703410133971029e-06,
+      "loss": 0.5966,
+      "step": 10321
+    },
+    {
+      "epoch": 2.859279778393352,
+      "grad_norm": 0.7260600328445435,
+      "learning_rate": 2.703047054755484e-06,
+      "loss": 0.5698,
+      "step": 10322
+    },
+    {
+      "epoch": 2.8595567867036014,
+      "grad_norm": 0.7182386517524719,
+      "learning_rate": 2.702683971228731e-06,
+      "loss": 0.6064,
+      "step": 10323
+    },
+    {
+      "epoch": 2.8598337950138504,
+      "grad_norm": 0.7457290291786194,
+      "learning_rate": 2.70232088339848e-06,
+      "loss": 0.6439,
+      "step": 10324
+    },
+    {
+      "epoch": 2.8601108033241,
+      "grad_norm": 0.7517281770706177,
+      "learning_rate": 2.7019577912724393e-06,
+      "loss": 0.5871,
+      "step": 10325
+    },
+    {
+      "epoch": 2.8603878116343493,
+      "grad_norm": 0.7766019701957703,
+      "learning_rate": 2.7015946948583198e-06,
+      "loss": 0.5691,
+      "step": 10326
+    },
+    {
+      "epoch": 2.8606648199445983,
+      "grad_norm": 0.77918541431427,
+      "learning_rate": 2.701231594163829e-06,
+      "loss": 0.6303,
+      "step": 10327
+    },
+    {
+      "epoch": 2.8609418282548478,
+      "grad_norm": 0.7339177131652832,
+      "learning_rate": 2.7008684891966787e-06,
+      "loss": 0.5974,
+      "step": 10328
+    },
+    {
+      "epoch": 2.861218836565097,
+      "grad_norm": 0.7705397009849548,
+      "learning_rate": 2.700505379964577e-06,
+      "loss": 0.5466,
+      "step": 10329
+    },
+    {
+      "epoch": 2.8614958448753463,
+      "grad_norm": 0.7495179772377014,
+      "learning_rate": 2.7001422664752338e-06,
+      "loss": 0.5933,
+      "step": 10330
+    },
+    {
+      "epoch": 2.8617728531855957,
+      "grad_norm": 0.7036161422729492,
+      "learning_rate": 2.699779148736359e-06,
+      "loss": 0.5784,
+      "step": 10331
+    },
+    {
+      "epoch": 2.8620498614958447,
+      "grad_norm": 0.8170588612556458,
+      "learning_rate": 2.6994160267556633e-06,
+      "loss": 0.5748,
+      "step": 10332
+    },
+    {
+      "epoch": 2.862326869806094,
+      "grad_norm": 0.7510451674461365,
+      "learning_rate": 2.6990529005408563e-06,
+      "loss": 0.6104,
+      "step": 10333
+    },
+    {
+      "epoch": 2.8626038781163436,
+      "grad_norm": 0.7337208390235901,
+      "learning_rate": 2.6986897700996467e-06,
+      "loss": 0.6053,
+      "step": 10334
+    },
+    {
+      "epoch": 2.8628808864265927,
+      "grad_norm": 0.7081796526908875,
+      "learning_rate": 2.698326635439747e-06,
+      "loss": 0.5888,
+      "step": 10335
+    },
+    {
+      "epoch": 2.863157894736842,
+      "grad_norm": 0.7299284338951111,
+      "learning_rate": 2.697963496568866e-06,
+      "loss": 0.573,
+      "step": 10336
+    },
+    {
+      "epoch": 2.863434903047091,
+      "grad_norm": 0.757738471031189,
+      "learning_rate": 2.6976003534947144e-06,
+      "loss": 0.6199,
+      "step": 10337
+    },
+    {
+      "epoch": 2.8637119113573406,
+      "grad_norm": 0.7651471495628357,
+      "learning_rate": 2.697237206225003e-06,
+      "loss": 0.5917,
+      "step": 10338
+    },
+    {
+      "epoch": 2.86398891966759,
+      "grad_norm": 0.7392405867576599,
+      "learning_rate": 2.6968740547674416e-06,
+      "loss": 0.5752,
+      "step": 10339
+    },
+    {
+      "epoch": 2.864265927977839,
+      "grad_norm": 0.761151909828186,
+      "learning_rate": 2.696510899129741e-06,
+      "loss": 0.6254,
+      "step": 10340
+    },
+    {
+      "epoch": 2.8645429362880885,
+      "grad_norm": 0.74574214220047,
+      "learning_rate": 2.696147739319613e-06,
+      "loss": 0.5659,
+      "step": 10341
+    },
+    {
+      "epoch": 2.864819944598338,
+      "grad_norm": 0.743224024772644,
+      "learning_rate": 2.695784575344767e-06,
+      "loss": 0.6149,
+      "step": 10342
+    },
+    {
+      "epoch": 2.865096952908587,
+      "grad_norm": 0.7006165981292725,
+      "learning_rate": 2.6954214072129148e-06,
+      "loss": 0.5829,
+      "step": 10343
+    },
+    {
+      "epoch": 2.8653739612188365,
+      "grad_norm": 0.7185438871383667,
+      "learning_rate": 2.6950582349317665e-06,
+      "loss": 0.6488,
+      "step": 10344
+    },
+    {
+      "epoch": 2.865650969529086,
+      "grad_norm": 0.7694297432899475,
+      "learning_rate": 2.694695058509034e-06,
+      "loss": 0.6311,
+      "step": 10345
+    },
+    {
+      "epoch": 2.865927977839335,
+      "grad_norm": 0.7526674270629883,
+      "learning_rate": 2.6943318779524285e-06,
+      "loss": 0.5898,
+      "step": 10346
+    },
+    {
+      "epoch": 2.8662049861495844,
+      "grad_norm": 0.7574858069419861,
+      "learning_rate": 2.6939686932696608e-06,
+      "loss": 0.5964,
+      "step": 10347
+    },
+    {
+      "epoch": 2.866481994459834,
+      "grad_norm": 0.690632164478302,
+      "learning_rate": 2.6936055044684427e-06,
+      "loss": 0.5699,
+      "step": 10348
+    },
+    {
+      "epoch": 2.866759002770083,
+      "grad_norm": 0.7249476909637451,
+      "learning_rate": 2.6932423115564847e-06,
+      "loss": 0.6352,
+      "step": 10349
+    },
+    {
+      "epoch": 2.8670360110803323,
+      "grad_norm": 0.6996097564697266,
+      "learning_rate": 2.6928791145414996e-06,
+      "loss": 0.5521,
+      "step": 10350
+    },
+    {
+      "epoch": 2.867313019390582,
+      "grad_norm": 0.7583314776420593,
+      "learning_rate": 2.692515913431198e-06,
+      "loss": 0.6453,
+      "step": 10351
+    },
+    {
+      "epoch": 2.867590027700831,
+      "grad_norm": 0.7682088613510132,
+      "learning_rate": 2.692152708233292e-06,
+      "loss": 0.6159,
+      "step": 10352
+    },
+    {
+      "epoch": 2.8678670360110803,
+      "grad_norm": 0.7417542934417725,
+      "learning_rate": 2.6917894989554926e-06,
+      "loss": 0.5728,
+      "step": 10353
+    },
+    {
+      "epoch": 2.8681440443213297,
+      "grad_norm": 0.7185600399971008,
+      "learning_rate": 2.6914262856055135e-06,
+      "loss": 0.561,
+      "step": 10354
+    },
+    {
+      "epoch": 2.8684210526315788,
+      "grad_norm": 0.7589703798294067,
+      "learning_rate": 2.691063068191065e-06,
+      "loss": 0.5948,
+      "step": 10355
+    },
+    {
+      "epoch": 2.868698060941828,
+      "grad_norm": 0.7474316358566284,
+      "learning_rate": 2.69069984671986e-06,
+      "loss": 0.6212,
+      "step": 10356
+    },
+    {
+      "epoch": 2.8689750692520777,
+      "grad_norm": 0.7184518575668335,
+      "learning_rate": 2.690336621199609e-06,
+      "loss": 0.5574,
+      "step": 10357
+    },
+    {
+      "epoch": 2.8692520775623267,
+      "grad_norm": 0.7420065999031067,
+      "learning_rate": 2.6899733916380265e-06,
+      "loss": 0.5569,
+      "step": 10358
+    },
+    {
+      "epoch": 2.869529085872576,
+      "grad_norm": 0.754462718963623,
+      "learning_rate": 2.689610158042824e-06,
+      "loss": 0.5758,
+      "step": 10359
+    },
+    {
+      "epoch": 2.8698060941828256,
+      "grad_norm": 0.7530493140220642,
+      "learning_rate": 2.689246920421713e-06,
+      "loss": 0.5947,
+      "step": 10360
+    },
+    {
+      "epoch": 2.8700831024930746,
+      "grad_norm": 0.7274234294891357,
+      "learning_rate": 2.6888836787824074e-06,
+      "loss": 0.5789,
+      "step": 10361
+    },
+    {
+      "epoch": 2.870360110803324,
+      "grad_norm": 0.7387732863426208,
+      "learning_rate": 2.6885204331326183e-06,
+      "loss": 0.5774,
+      "step": 10362
+    },
+    {
+      "epoch": 2.8706371191135736,
+      "grad_norm": 0.6835230588912964,
+      "learning_rate": 2.688157183480059e-06,
+      "loss": 0.5154,
+      "step": 10363
+    },
+    {
+      "epoch": 2.8709141274238226,
+      "grad_norm": 0.7362700700759888,
+      "learning_rate": 2.687793929832442e-06,
+      "loss": 0.6004,
+      "step": 10364
+    },
+    {
+      "epoch": 2.871191135734072,
+      "grad_norm": 0.7476580142974854,
+      "learning_rate": 2.687430672197481e-06,
+      "loss": 0.584,
+      "step": 10365
+    },
+    {
+      "epoch": 2.8714681440443215,
+      "grad_norm": 0.7573707699775696,
+      "learning_rate": 2.687067410582887e-06,
+      "loss": 0.6167,
+      "step": 10366
+    },
+    {
+      "epoch": 2.8717451523545705,
+      "grad_norm": 0.7462565898895264,
+      "learning_rate": 2.686704144996376e-06,
+      "loss": 0.5862,
+      "step": 10367
+    },
+    {
+      "epoch": 2.87202216066482,
+      "grad_norm": 0.743422269821167,
+      "learning_rate": 2.6863408754456577e-06,
+      "loss": 0.5973,
+      "step": 10368
+    },
+    {
+      "epoch": 2.8722991689750694,
+      "grad_norm": 0.7598124742507935,
+      "learning_rate": 2.685977601938447e-06,
+      "loss": 0.5604,
+      "step": 10369
+    },
+    {
+      "epoch": 2.8725761772853184,
+      "grad_norm": 0.767044723033905,
+      "learning_rate": 2.6856143244824573e-06,
+      "loss": 0.579,
+      "step": 10370
+    },
+    {
+      "epoch": 2.872853185595568,
+      "grad_norm": 0.764921247959137,
+      "learning_rate": 2.685251043085401e-06,
+      "loss": 0.6125,
+      "step": 10371
+    },
+    {
+      "epoch": 2.8731301939058174,
+      "grad_norm": 0.7420265674591064,
+      "learning_rate": 2.684887757754992e-06,
+      "loss": 0.6123,
+      "step": 10372
+    },
+    {
+      "epoch": 2.8734072022160664,
+      "grad_norm": 0.7900787591934204,
+      "learning_rate": 2.684524468498944e-06,
+      "loss": 0.5781,
+      "step": 10373
+    },
+    {
+      "epoch": 2.873684210526316,
+      "grad_norm": 0.7371800541877747,
+      "learning_rate": 2.684161175324971e-06,
+      "loss": 0.5815,
+      "step": 10374
+    },
+    {
+      "epoch": 2.8739612188365653,
+      "grad_norm": 0.7698126435279846,
+      "learning_rate": 2.683797878240784e-06,
+      "loss": 0.5723,
+      "step": 10375
+    },
+    {
+      "epoch": 2.8742382271468143,
+      "grad_norm": 0.729429304599762,
+      "learning_rate": 2.6834345772541005e-06,
+      "loss": 0.5586,
+      "step": 10376
+    },
+    {
+      "epoch": 2.8745152354570638,
+      "grad_norm": 0.7678404450416565,
+      "learning_rate": 2.683071272372632e-06,
+      "loss": 0.6127,
+      "step": 10377
+    },
+    {
+      "epoch": 2.8747922437673132,
+      "grad_norm": 0.7486993670463562,
+      "learning_rate": 2.6827079636040924e-06,
+      "loss": 0.5885,
+      "step": 10378
+    },
+    {
+      "epoch": 2.8750692520775623,
+      "grad_norm": 0.7203310132026672,
+      "learning_rate": 2.682344650956196e-06,
+      "loss": 0.58,
+      "step": 10379
+    },
+    {
+      "epoch": 2.8753462603878117,
+      "grad_norm": 0.7622033953666687,
+      "learning_rate": 2.6819813344366584e-06,
+      "loss": 0.568,
+      "step": 10380
+    },
+    {
+      "epoch": 2.875623268698061,
+      "grad_norm": 0.7309375405311584,
+      "learning_rate": 2.681618014053191e-06,
+      "loss": 0.5834,
+      "step": 10381
+    },
+    {
+      "epoch": 2.87590027700831,
+      "grad_norm": 0.7362342476844788,
+      "learning_rate": 2.68125468981351e-06,
+      "loss": 0.5626,
+      "step": 10382
+    },
+    {
+      "epoch": 2.8761772853185597,
+      "grad_norm": 0.7526897192001343,
+      "learning_rate": 2.680891361725329e-06,
+      "loss": 0.5786,
+      "step": 10383
+    },
+    {
+      "epoch": 2.876454293628809,
+      "grad_norm": 0.7369986772537231,
+      "learning_rate": 2.6805280297963625e-06,
+      "loss": 0.5598,
+      "step": 10384
+    },
+    {
+      "epoch": 2.876731301939058,
+      "grad_norm": 0.7252307534217834,
+      "learning_rate": 2.6801646940343245e-06,
+      "loss": 0.6002,
+      "step": 10385
+    },
+    {
+      "epoch": 2.8770083102493076,
+      "grad_norm": 0.7209944128990173,
+      "learning_rate": 2.6798013544469304e-06,
+      "loss": 0.5885,
+      "step": 10386
+    },
+    {
+      "epoch": 2.877285318559557,
+      "grad_norm": 0.7498239874839783,
+      "learning_rate": 2.679438011041894e-06,
+      "loss": 0.5409,
+      "step": 10387
+    },
+    {
+      "epoch": 2.877562326869806,
+      "grad_norm": 0.7315983772277832,
+      "learning_rate": 2.67907466382693e-06,
+      "loss": 0.5746,
+      "step": 10388
+    },
+    {
+      "epoch": 2.8778393351800555,
+      "grad_norm": 0.7259673476219177,
+      "learning_rate": 2.678711312809754e-06,
+      "loss": 0.5633,
+      "step": 10389
+    },
+    {
+      "epoch": 2.878116343490305,
+      "grad_norm": 0.7267625331878662,
+      "learning_rate": 2.678347957998081e-06,
+      "loss": 0.5913,
+      "step": 10390
+    },
+    {
+      "epoch": 2.878393351800554,
+      "grad_norm": 0.7919610738754272,
+      "learning_rate": 2.6779845993996252e-06,
+      "loss": 0.6664,
+      "step": 10391
+    },
+    {
+      "epoch": 2.8786703601108035,
+      "grad_norm": 0.7108157277107239,
+      "learning_rate": 2.6776212370221015e-06,
+      "loss": 0.5555,
+      "step": 10392
+    },
+    {
+      "epoch": 2.8789473684210525,
+      "grad_norm": 0.7125835418701172,
+      "learning_rate": 2.6772578708732254e-06,
+      "loss": 0.6006,
+      "step": 10393
+    },
+    {
+      "epoch": 2.879224376731302,
+      "grad_norm": 0.6881554126739502,
+      "learning_rate": 2.676894500960712e-06,
+      "loss": 0.5513,
+      "step": 10394
+    },
+    {
+      "epoch": 2.8795013850415514,
+      "grad_norm": 0.7503734230995178,
+      "learning_rate": 2.676531127292277e-06,
+      "loss": 0.6088,
+      "step": 10395
+    },
+    {
+      "epoch": 2.8797783933518004,
+      "grad_norm": 0.7577073574066162,
+      "learning_rate": 2.676167749875635e-06,
+      "loss": 0.5944,
+      "step": 10396
+    },
+    {
+      "epoch": 2.88005540166205,
+      "grad_norm": 0.7305821776390076,
+      "learning_rate": 2.675804368718502e-06,
+      "loss": 0.5907,
+      "step": 10397
+    },
+    {
+      "epoch": 2.8803324099722993,
+      "grad_norm": 0.7082803845405579,
+      "learning_rate": 2.6754409838285926e-06,
+      "loss": 0.5824,
+      "step": 10398
+    },
+    {
+      "epoch": 2.8806094182825484,
+      "grad_norm": 0.7570558190345764,
+      "learning_rate": 2.6750775952136244e-06,
+      "loss": 0.5769,
+      "step": 10399
+    },
+    {
+      "epoch": 2.880886426592798,
+      "grad_norm": 0.7124000191688538,
+      "learning_rate": 2.6747142028813107e-06,
+      "loss": 0.5276,
+      "step": 10400
+    },
+    {
+      "epoch": 2.881163434903047,
+      "grad_norm": 0.7314373850822449,
+      "learning_rate": 2.674350806839368e-06,
+      "loss": 0.5916,
+      "step": 10401
+    },
+    {
+      "epoch": 2.8814404432132963,
+      "grad_norm": 0.714793860912323,
+      "learning_rate": 2.6739874070955134e-06,
+      "loss": 0.5756,
+      "step": 10402
+    },
+    {
+      "epoch": 2.8817174515235457,
+      "grad_norm": 0.7305763959884644,
+      "learning_rate": 2.673624003657461e-06,
+      "loss": 0.6034,
+      "step": 10403
+    },
+    {
+      "epoch": 2.8819944598337948,
+      "grad_norm": 0.7191005349159241,
+      "learning_rate": 2.6732605965329282e-06,
+      "loss": 0.6368,
+      "step": 10404
+    },
+    {
+      "epoch": 2.8822714681440442,
+      "grad_norm": 0.7448530793190002,
+      "learning_rate": 2.6728971857296305e-06,
+      "loss": 0.5453,
+      "step": 10405
+    },
+    {
+      "epoch": 2.8825484764542937,
+      "grad_norm": 0.6936946511268616,
+      "learning_rate": 2.672533771255284e-06,
+      "loss": 0.5296,
+      "step": 10406
+    },
+    {
+      "epoch": 2.8828254847645427,
+      "grad_norm": 0.7031975388526917,
+      "learning_rate": 2.6721703531176045e-06,
+      "loss": 0.5641,
+      "step": 10407
+    },
+    {
+      "epoch": 2.883102493074792,
+      "grad_norm": 0.7554457187652588,
+      "learning_rate": 2.671806931324309e-06,
+      "loss": 0.5301,
+      "step": 10408
+    },
+    {
+      "epoch": 2.8833795013850416,
+      "grad_norm": 0.7462570071220398,
+      "learning_rate": 2.671443505883114e-06,
+      "loss": 0.5864,
+      "step": 10409
+    },
+    {
+      "epoch": 2.8836565096952906,
+      "grad_norm": 0.7643255591392517,
+      "learning_rate": 2.671080076801735e-06,
+      "loss": 0.6312,
+      "step": 10410
+    },
+    {
+      "epoch": 2.88393351800554,
+      "grad_norm": 0.7761854529380798,
+      "learning_rate": 2.670716644087889e-06,
+      "loss": 0.6093,
+      "step": 10411
+    },
+    {
+      "epoch": 2.8842105263157896,
+      "grad_norm": 0.7329185605049133,
+      "learning_rate": 2.6703532077492935e-06,
+      "loss": 0.5356,
+      "step": 10412
+    },
+    {
+      "epoch": 2.8844875346260386,
+      "grad_norm": 0.7831817269325256,
+      "learning_rate": 2.669989767793663e-06,
+      "loss": 0.5695,
+      "step": 10413
+    },
+    {
+      "epoch": 2.884764542936288,
+      "grad_norm": 0.7784004211425781,
+      "learning_rate": 2.6696263242287167e-06,
+      "loss": 0.6021,
+      "step": 10414
+    },
+    {
+      "epoch": 2.8850415512465375,
+      "grad_norm": 0.7950989007949829,
+      "learning_rate": 2.6692628770621704e-06,
+      "loss": 0.575,
+      "step": 10415
+    },
+    {
+      "epoch": 2.8853185595567865,
+      "grad_norm": 0.7388087511062622,
+      "learning_rate": 2.6688994263017405e-06,
+      "loss": 0.5906,
+      "step": 10416
+    },
+    {
+      "epoch": 2.885595567867036,
+      "grad_norm": 0.6924887299537659,
+      "learning_rate": 2.6685359719551445e-06,
+      "loss": 0.542,
+      "step": 10417
+    },
+    {
+      "epoch": 2.8858725761772854,
+      "grad_norm": 0.7233446836471558,
+      "learning_rate": 2.6681725140300995e-06,
+      "loss": 0.5887,
+      "step": 10418
+    },
+    {
+      "epoch": 2.8861495844875344,
+      "grad_norm": 0.7647252678871155,
+      "learning_rate": 2.6678090525343227e-06,
+      "loss": 0.6204,
+      "step": 10419
+    },
+    {
+      "epoch": 2.886426592797784,
+      "grad_norm": 0.7015892863273621,
+      "learning_rate": 2.6674455874755306e-06,
+      "loss": 0.5807,
+      "step": 10420
+    },
+    {
+      "epoch": 2.8867036011080334,
+      "grad_norm": 0.7895192503929138,
+      "learning_rate": 2.6670821188614422e-06,
+      "loss": 0.5693,
+      "step": 10421
+    },
+    {
+      "epoch": 2.8869806094182824,
+      "grad_norm": 0.7190335988998413,
+      "learning_rate": 2.6667186466997724e-06,
+      "loss": 0.5839,
+      "step": 10422
+    },
+    {
+      "epoch": 2.887257617728532,
+      "grad_norm": 0.7219985127449036,
+      "learning_rate": 2.666355170998241e-06,
+      "loss": 0.5896,
+      "step": 10423
+    },
+    {
+      "epoch": 2.8875346260387813,
+      "grad_norm": 0.7790842056274414,
+      "learning_rate": 2.665991691764564e-06,
+      "loss": 0.614,
+      "step": 10424
+    },
+    {
+      "epoch": 2.8878116343490303,
+      "grad_norm": 1.0126031637191772,
+      "learning_rate": 2.66562820900646e-06,
+      "loss": 0.6181,
+      "step": 10425
+    },
+    {
+      "epoch": 2.88808864265928,
+      "grad_norm": 0.7423431873321533,
+      "learning_rate": 2.6652647227316453e-06,
+      "loss": 0.627,
+      "step": 10426
+    },
+    {
+      "epoch": 2.8883656509695292,
+      "grad_norm": 0.7445407509803772,
+      "learning_rate": 2.664901232947839e-06,
+      "loss": 0.6103,
+      "step": 10427
+    },
+    {
+      "epoch": 2.8886426592797783,
+      "grad_norm": 0.7246869206428528,
+      "learning_rate": 2.6645377396627587e-06,
+      "loss": 0.5792,
+      "step": 10428
+    },
+    {
+      "epoch": 2.8889196675900277,
+      "grad_norm": 0.7190065383911133,
+      "learning_rate": 2.664174242884121e-06,
+      "loss": 0.5731,
+      "step": 10429
+    },
+    {
+      "epoch": 2.889196675900277,
+      "grad_norm": 0.7758619785308838,
+      "learning_rate": 2.6638107426196457e-06,
+      "loss": 0.5841,
+      "step": 10430
+    },
+    {
+      "epoch": 2.889473684210526,
+      "grad_norm": 0.790145754814148,
+      "learning_rate": 2.66344723887705e-06,
+      "loss": 0.6128,
+      "step": 10431
+    },
+    {
+      "epoch": 2.8897506925207757,
+      "grad_norm": 0.7663121223449707,
+      "learning_rate": 2.663083731664052e-06,
+      "loss": 0.5924,
+      "step": 10432
+    },
+    {
+      "epoch": 2.890027700831025,
+      "grad_norm": 0.7356205582618713,
+      "learning_rate": 2.6627202209883697e-06,
+      "loss": 0.6058,
+      "step": 10433
+    },
+    {
+      "epoch": 2.890304709141274,
+      "grad_norm": 0.7383139133453369,
+      "learning_rate": 2.662356706857722e-06,
+      "loss": 0.553,
+      "step": 10434
+    },
+    {
+      "epoch": 2.8905817174515236,
+      "grad_norm": 0.7479643821716309,
+      "learning_rate": 2.661993189279826e-06,
+      "loss": 0.5983,
+      "step": 10435
+    },
+    {
+      "epoch": 2.890858725761773,
+      "grad_norm": 0.7774149179458618,
+      "learning_rate": 2.6616296682624016e-06,
+      "loss": 0.6211,
+      "step": 10436
+    },
+    {
+      "epoch": 2.891135734072022,
+      "grad_norm": 0.7735824584960938,
+      "learning_rate": 2.6612661438131664e-06,
+      "loss": 0.6096,
+      "step": 10437
+    },
+    {
+      "epoch": 2.8914127423822715,
+      "grad_norm": 0.7517207860946655,
+      "learning_rate": 2.6609026159398395e-06,
+      "loss": 0.5858,
+      "step": 10438
+    },
+    {
+      "epoch": 2.891689750692521,
+      "grad_norm": 0.750087320804596,
+      "learning_rate": 2.6605390846501377e-06,
+      "loss": 0.5987,
+      "step": 10439
+    },
+    {
+      "epoch": 2.89196675900277,
+      "grad_norm": 0.7316486239433289,
+      "learning_rate": 2.6601755499517826e-06,
+      "loss": 0.5654,
+      "step": 10440
+    },
+    {
+      "epoch": 2.8922437673130195,
+      "grad_norm": 0.7494932413101196,
+      "learning_rate": 2.659812011852491e-06,
+      "loss": 0.6034,
+      "step": 10441
+    },
+    {
+      "epoch": 2.892520775623269,
+      "grad_norm": 0.7179362177848816,
+      "learning_rate": 2.659448470359983e-06,
+      "loss": 0.5445,
+      "step": 10442
+    },
+    {
+      "epoch": 2.892797783933518,
+      "grad_norm": 0.7596146464347839,
+      "learning_rate": 2.6590849254819757e-06,
+      "loss": 0.6145,
+      "step": 10443
+    },
+    {
+      "epoch": 2.8930747922437674,
+      "grad_norm": 0.7503767013549805,
+      "learning_rate": 2.6587213772261896e-06,
+      "loss": 0.5562,
+      "step": 10444
+    },
+    {
+      "epoch": 2.893351800554017,
+      "grad_norm": 0.7222331762313843,
+      "learning_rate": 2.6583578256003435e-06,
+      "loss": 0.6068,
+      "step": 10445
+    },
+    {
+      "epoch": 2.893628808864266,
+      "grad_norm": 0.735457181930542,
+      "learning_rate": 2.6579942706121566e-06,
+      "loss": 0.6146,
+      "step": 10446
+    },
+    {
+      "epoch": 2.8939058171745153,
+      "grad_norm": 0.7325124144554138,
+      "learning_rate": 2.6576307122693474e-06,
+      "loss": 0.5624,
+      "step": 10447
+    },
+    {
+      "epoch": 2.894182825484765,
+      "grad_norm": 0.7304285764694214,
+      "learning_rate": 2.6572671505796357e-06,
+      "loss": 0.5689,
+      "step": 10448
+    },
+    {
+      "epoch": 2.894459833795014,
+      "grad_norm": 0.7730352878570557,
+      "learning_rate": 2.656903585550741e-06,
+      "loss": 0.5711,
+      "step": 10449
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.7142786383628845,
+      "learning_rate": 2.656540017190382e-06,
+      "loss": 0.5513,
+      "step": 10450
+    },
+    {
+      "epoch": 2.8950138504155127,
+      "grad_norm": 0.7653011083602905,
+      "learning_rate": 2.6561764455062793e-06,
+      "loss": 0.5771,
+      "step": 10451
+    },
+    {
+      "epoch": 2.8952908587257618,
+      "grad_norm": 0.7704102993011475,
+      "learning_rate": 2.655812870506151e-06,
+      "loss": 0.6245,
+      "step": 10452
+    },
+    {
+      "epoch": 2.895567867036011,
+      "grad_norm": 0.703545093536377,
+      "learning_rate": 2.6554492921977188e-06,
+      "loss": 0.5881,
+      "step": 10453
+    },
+    {
+      "epoch": 2.8958448753462602,
+      "grad_norm": 0.7527859807014465,
+      "learning_rate": 2.6550857105887e-06,
+      "loss": 0.6229,
+      "step": 10454
+    },
+    {
+      "epoch": 2.8961218836565097,
+      "grad_norm": 0.7578487992286682,
+      "learning_rate": 2.654722125686816e-06,
+      "loss": 0.6109,
+      "step": 10455
+    },
+    {
+      "epoch": 2.896398891966759,
+      "grad_norm": 0.7411022782325745,
+      "learning_rate": 2.654358537499786e-06,
+      "loss": 0.6115,
+      "step": 10456
+    },
+    {
+      "epoch": 2.896675900277008,
+      "grad_norm": 0.712331235408783,
+      "learning_rate": 2.6539949460353297e-06,
+      "loss": 0.6168,
+      "step": 10457
+    },
+    {
+      "epoch": 2.8969529085872576,
+      "grad_norm": 0.7301256060600281,
+      "learning_rate": 2.653631351301168e-06,
+      "loss": 0.5976,
+      "step": 10458
+    },
+    {
+      "epoch": 2.897229916897507,
+      "grad_norm": 0.7509595155715942,
+      "learning_rate": 2.6532677533050207e-06,
+      "loss": 0.5968,
+      "step": 10459
+    },
+    {
+      "epoch": 2.897506925207756,
+      "grad_norm": 0.7287399172782898,
+      "learning_rate": 2.6529041520546072e-06,
+      "loss": 0.5736,
+      "step": 10460
+    },
+    {
+      "epoch": 2.8977839335180056,
+      "grad_norm": 0.7702319025993347,
+      "learning_rate": 2.6525405475576478e-06,
+      "loss": 0.5706,
+      "step": 10461
+    },
+    {
+      "epoch": 2.8980609418282546,
+      "grad_norm": 0.749108076095581,
+      "learning_rate": 2.6521769398218635e-06,
+      "loss": 0.5925,
+      "step": 10462
+    },
+    {
+      "epoch": 2.898337950138504,
+      "grad_norm": 0.757893979549408,
+      "learning_rate": 2.651813328854974e-06,
+      "loss": 0.5766,
+      "step": 10463
+    },
+    {
+      "epoch": 2.8986149584487535,
+      "grad_norm": 0.7170742750167847,
+      "learning_rate": 2.6514497146647e-06,
+      "loss": 0.5713,
+      "step": 10464
+    },
+    {
+      "epoch": 2.8988919667590025,
+      "grad_norm": 0.7580187916755676,
+      "learning_rate": 2.651086097258761e-06,
+      "loss": 0.5913,
+      "step": 10465
+    },
+    {
+      "epoch": 2.899168975069252,
+      "grad_norm": 0.7491013407707214,
+      "learning_rate": 2.6507224766448797e-06,
+      "loss": 0.5965,
+      "step": 10466
+    },
+    {
+      "epoch": 2.8994459833795014,
+      "grad_norm": 0.7685227990150452,
+      "learning_rate": 2.650358852830774e-06,
+      "loss": 0.5856,
+      "step": 10467
+    },
+    {
+      "epoch": 2.8997229916897505,
+      "grad_norm": 0.7391490340232849,
+      "learning_rate": 2.6499952258241664e-06,
+      "loss": 0.6273,
+      "step": 10468
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7236970067024231,
+      "learning_rate": 2.649631595632777e-06,
+      "loss": 0.6031,
+      "step": 10469
+    },
+    {
+      "epoch": 2.9002770083102494,
+      "grad_norm": 0.7581817507743835,
+      "learning_rate": 2.649267962264327e-06,
+      "loss": 0.6149,
+      "step": 10470
+    },
+    {
+      "epoch": 2.9005540166204984,
+      "grad_norm": 0.7591217756271362,
+      "learning_rate": 2.6489043257265362e-06,
+      "loss": 0.6211,
+      "step": 10471
+    },
+    {
+      "epoch": 2.900831024930748,
+      "grad_norm": 0.7190767526626587,
+      "learning_rate": 2.648540686027127e-06,
+      "loss": 0.618,
+      "step": 10472
+    },
+    {
+      "epoch": 2.9011080332409973,
+      "grad_norm": 0.6588174700737,
+      "learning_rate": 2.6481770431738202e-06,
+      "loss": 0.5314,
+      "step": 10473
+    },
+    {
+      "epoch": 2.9013850415512463,
+      "grad_norm": 0.7454791069030762,
+      "learning_rate": 2.647813397174335e-06,
+      "loss": 0.5785,
+      "step": 10474
+    },
+    {
+      "epoch": 2.901662049861496,
+      "grad_norm": 0.7779990434646606,
+      "learning_rate": 2.647449748036395e-06,
+      "loss": 0.5702,
+      "step": 10475
+    },
+    {
+      "epoch": 2.9019390581717452,
+      "grad_norm": 0.725070595741272,
+      "learning_rate": 2.6470860957677203e-06,
+      "loss": 0.5901,
+      "step": 10476
+    },
+    {
+      "epoch": 2.9022160664819943,
+      "grad_norm": 0.766870379447937,
+      "learning_rate": 2.6467224403760315e-06,
+      "loss": 0.5372,
+      "step": 10477
+    },
+    {
+      "epoch": 2.9024930747922437,
+      "grad_norm": 0.6898723840713501,
+      "learning_rate": 2.646358781869051e-06,
+      "loss": 0.5443,
+      "step": 10478
+    },
+    {
+      "epoch": 2.902770083102493,
+      "grad_norm": 0.7398709654808044,
+      "learning_rate": 2.6459951202544998e-06,
+      "loss": 0.5762,
+      "step": 10479
+    },
+    {
+      "epoch": 2.903047091412742,
+      "grad_norm": 0.7504476308822632,
+      "learning_rate": 2.645631455540099e-06,
+      "loss": 0.6037,
+      "step": 10480
+    },
+    {
+      "epoch": 2.9033240997229917,
+      "grad_norm": 0.7410624623298645,
+      "learning_rate": 2.645267787733571e-06,
+      "loss": 0.593,
+      "step": 10481
+    },
+    {
+      "epoch": 2.903601108033241,
+      "grad_norm": 0.694694459438324,
+      "learning_rate": 2.6449041168426372e-06,
+      "loss": 0.5549,
+      "step": 10482
+    },
+    {
+      "epoch": 2.90387811634349,
+      "grad_norm": 0.8081617951393127,
+      "learning_rate": 2.6445404428750183e-06,
+      "loss": 0.6227,
+      "step": 10483
+    },
+    {
+      "epoch": 2.9041551246537396,
+      "grad_norm": 0.7287876009941101,
+      "learning_rate": 2.6441767658384363e-06,
+      "loss": 0.576,
+      "step": 10484
+    },
+    {
+      "epoch": 2.904432132963989,
+      "grad_norm": 0.7433827519416809,
+      "learning_rate": 2.643813085740614e-06,
+      "loss": 0.6118,
+      "step": 10485
+    },
+    {
+      "epoch": 2.904709141274238,
+      "grad_norm": 0.7551594376564026,
+      "learning_rate": 2.6434494025892726e-06,
+      "loss": 0.5844,
+      "step": 10486
+    },
+    {
+      "epoch": 2.9049861495844875,
+      "grad_norm": 0.7610915303230286,
+      "learning_rate": 2.6430857163921335e-06,
+      "loss": 0.5876,
+      "step": 10487
+    },
+    {
+      "epoch": 2.905263157894737,
+      "grad_norm": 0.6847673654556274,
+      "learning_rate": 2.6427220271569206e-06,
+      "loss": 0.6123,
+      "step": 10488
+    },
+    {
+      "epoch": 2.905540166204986,
+      "grad_norm": 0.7411179542541504,
+      "learning_rate": 2.6423583348913533e-06,
+      "loss": 0.5867,
+      "step": 10489
+    },
+    {
+      "epoch": 2.9058171745152355,
+      "grad_norm": 0.7440828084945679,
+      "learning_rate": 2.6419946396031552e-06,
+      "loss": 0.5836,
+      "step": 10490
+    },
+    {
+      "epoch": 2.906094182825485,
+      "grad_norm": 0.7707085609436035,
+      "learning_rate": 2.641630941300049e-06,
+      "loss": 0.5885,
+      "step": 10491
+    },
+    {
+      "epoch": 2.906371191135734,
+      "grad_norm": 0.75754314661026,
+      "learning_rate": 2.641267239989756e-06,
+      "loss": 0.596,
+      "step": 10492
+    },
+    {
+      "epoch": 2.9066481994459834,
+      "grad_norm": 0.7997472882270813,
+      "learning_rate": 2.640903535679998e-06,
+      "loss": 0.6185,
+      "step": 10493
+    },
+    {
+      "epoch": 2.906925207756233,
+      "grad_norm": 0.754387617111206,
+      "learning_rate": 2.6405398283784984e-06,
+      "loss": 0.5758,
+      "step": 10494
+    },
+    {
+      "epoch": 2.907202216066482,
+      "grad_norm": 0.7376065254211426,
+      "learning_rate": 2.6401761180929798e-06,
+      "loss": 0.5902,
+      "step": 10495
+    },
+    {
+      "epoch": 2.9074792243767313,
+      "grad_norm": 0.7251534461975098,
+      "learning_rate": 2.639812404831164e-06,
+      "loss": 0.6071,
+      "step": 10496
+    },
+    {
+      "epoch": 2.907756232686981,
+      "grad_norm": 0.8182569742202759,
+      "learning_rate": 2.6394486886007743e-06,
+      "loss": 0.5888,
+      "step": 10497
+    },
+    {
+      "epoch": 2.90803324099723,
+      "grad_norm": 0.7495296001434326,
+      "learning_rate": 2.6390849694095322e-06,
+      "loss": 0.6519,
+      "step": 10498
+    },
+    {
+      "epoch": 2.9083102493074793,
+      "grad_norm": 0.7202491164207458,
+      "learning_rate": 2.6387212472651606e-06,
+      "loss": 0.607,
+      "step": 10499
+    },
+    {
+      "epoch": 2.9085872576177287,
+      "grad_norm": 0.7387695908546448,
+      "learning_rate": 2.638357522175383e-06,
+      "loss": 0.6281,
+      "step": 10500
+    },
+    {
+      "epoch": 2.9088642659279778,
+      "grad_norm": 0.7193158268928528,
+      "learning_rate": 2.6379937941479226e-06,
+      "loss": 0.6095,
+      "step": 10501
+    },
+    {
+      "epoch": 2.909141274238227,
+      "grad_norm": 0.7286669015884399,
+      "learning_rate": 2.6376300631905005e-06,
+      "loss": 0.5486,
+      "step": 10502
+    },
+    {
+      "epoch": 2.9094182825484767,
+      "grad_norm": 0.7364821434020996,
+      "learning_rate": 2.6372663293108415e-06,
+      "loss": 0.586,
+      "step": 10503
+    },
+    {
+      "epoch": 2.9096952908587257,
+      "grad_norm": 0.7616895437240601,
+      "learning_rate": 2.6369025925166674e-06,
+      "loss": 0.5801,
+      "step": 10504
+    },
+    {
+      "epoch": 2.909972299168975,
+      "grad_norm": 0.7166840434074402,
+      "learning_rate": 2.636538852815702e-06,
+      "loss": 0.6128,
+      "step": 10505
+    },
+    {
+      "epoch": 2.9102493074792246,
+      "grad_norm": 0.7373641729354858,
+      "learning_rate": 2.6361751102156673e-06,
+      "loss": 0.5976,
+      "step": 10506
+    },
+    {
+      "epoch": 2.9105263157894736,
+      "grad_norm": 0.7657269239425659,
+      "learning_rate": 2.6358113647242885e-06,
+      "loss": 0.5798,
+      "step": 10507
+    },
+    {
+      "epoch": 2.910803324099723,
+      "grad_norm": 0.775659441947937,
+      "learning_rate": 2.6354476163492865e-06,
+      "loss": 0.5781,
+      "step": 10508
+    },
+    {
+      "epoch": 2.9110803324099725,
+      "grad_norm": 0.7320591807365417,
+      "learning_rate": 2.635083865098386e-06,
+      "loss": 0.5592,
+      "step": 10509
+    },
+    {
+      "epoch": 2.9113573407202216,
+      "grad_norm": 0.7523388266563416,
+      "learning_rate": 2.63472011097931e-06,
+      "loss": 0.6334,
+      "step": 10510
+    },
+    {
+      "epoch": 2.911634349030471,
+      "grad_norm": 0.7414600253105164,
+      "learning_rate": 2.634356353999783e-06,
+      "loss": 0.5845,
+      "step": 10511
+    },
+    {
+      "epoch": 2.9119113573407205,
+      "grad_norm": 0.7154373526573181,
+      "learning_rate": 2.633992594167526e-06,
+      "loss": 0.5643,
+      "step": 10512
+    },
+    {
+      "epoch": 2.9121883656509695,
+      "grad_norm": 0.6937295198440552,
+      "learning_rate": 2.6336288314902645e-06,
+      "loss": 0.581,
+      "step": 10513
+    },
+    {
+      "epoch": 2.912465373961219,
+      "grad_norm": 0.7316800951957703,
+      "learning_rate": 2.6332650659757223e-06,
+      "loss": 0.5846,
+      "step": 10514
+    },
+    {
+      "epoch": 2.912742382271468,
+      "grad_norm": 0.7038574814796448,
+      "learning_rate": 2.6329012976316216e-06,
+      "loss": 0.5028,
+      "step": 10515
+    },
+    {
+      "epoch": 2.9130193905817174,
+      "grad_norm": 0.7551959156990051,
+      "learning_rate": 2.632537526465687e-06,
+      "loss": 0.5651,
+      "step": 10516
+    },
+    {
+      "epoch": 2.913296398891967,
+      "grad_norm": 0.762638509273529,
+      "learning_rate": 2.6321737524856434e-06,
+      "loss": 0.6465,
+      "step": 10517
+    },
+    {
+      "epoch": 2.913573407202216,
+      "grad_norm": 0.7476597428321838,
+      "learning_rate": 2.631809975699213e-06,
+      "loss": 0.5952,
+      "step": 10518
+    },
+    {
+      "epoch": 2.9138504155124654,
+      "grad_norm": 0.7609908580780029,
+      "learning_rate": 2.6314461961141193e-06,
+      "loss": 0.5852,
+      "step": 10519
+    },
+    {
+      "epoch": 2.914127423822715,
+      "grad_norm": 0.7875975966453552,
+      "learning_rate": 2.631082413738088e-06,
+      "loss": 0.5712,
+      "step": 10520
+    },
+    {
+      "epoch": 2.914404432132964,
+      "grad_norm": 0.7203830480575562,
+      "learning_rate": 2.6307186285788423e-06,
+      "loss": 0.5757,
+      "step": 10521
+    },
+    {
+      "epoch": 2.9146814404432133,
+      "grad_norm": 0.7585301399230957,
+      "learning_rate": 2.630354840644106e-06,
+      "loss": 0.5814,
+      "step": 10522
+    },
+    {
+      "epoch": 2.9149584487534623,
+      "grad_norm": 0.74942946434021,
+      "learning_rate": 2.6299910499416036e-06,
+      "loss": 0.5819,
+      "step": 10523
+    },
+    {
+      "epoch": 2.915235457063712,
+      "grad_norm": 0.7211716771125793,
+      "learning_rate": 2.6296272564790594e-06,
+      "loss": 0.5855,
+      "step": 10524
+    },
+    {
+      "epoch": 2.9155124653739612,
+      "grad_norm": 0.7344101071357727,
+      "learning_rate": 2.629263460264197e-06,
+      "loss": 0.6082,
+      "step": 10525
+    },
+    {
+      "epoch": 2.9157894736842103,
+      "grad_norm": 0.7525090575218201,
+      "learning_rate": 2.6288996613047424e-06,
+      "loss": 0.5673,
+      "step": 10526
+    },
+    {
+      "epoch": 2.9160664819944597,
+      "grad_norm": 0.7705866694450378,
+      "learning_rate": 2.6285358596084175e-06,
+      "loss": 0.5953,
+      "step": 10527
+    },
+    {
+      "epoch": 2.916343490304709,
+      "grad_norm": 0.7172234654426575,
+      "learning_rate": 2.628172055182948e-06,
+      "loss": 0.6043,
+      "step": 10528
+    },
+    {
+      "epoch": 2.916620498614958,
+      "grad_norm": 0.7324336171150208,
+      "learning_rate": 2.6278082480360602e-06,
+      "loss": 0.5722,
+      "step": 10529
+    },
+    {
+      "epoch": 2.9168975069252077,
+      "grad_norm": 0.7875596880912781,
+      "learning_rate": 2.627444438175475e-06,
+      "loss": 0.6104,
+      "step": 10530
+    },
+    {
+      "epoch": 2.917174515235457,
+      "grad_norm": 0.7457826733589172,
+      "learning_rate": 2.62708062560892e-06,
+      "loss": 0.6176,
+      "step": 10531
+    },
+    {
+      "epoch": 2.917451523545706,
+      "grad_norm": 0.7246255874633789,
+      "learning_rate": 2.6267168103441185e-06,
+      "loss": 0.5951,
+      "step": 10532
+    },
+    {
+      "epoch": 2.9177285318559556,
+      "grad_norm": 0.7618518471717834,
+      "learning_rate": 2.6263529923887954e-06,
+      "loss": 0.6172,
+      "step": 10533
+    },
+    {
+      "epoch": 2.918005540166205,
+      "grad_norm": 0.715311586856842,
+      "learning_rate": 2.625989171750675e-06,
+      "loss": 0.5528,
+      "step": 10534
+    },
+    {
+      "epoch": 2.918282548476454,
+      "grad_norm": 0.7357887029647827,
+      "learning_rate": 2.625625348437484e-06,
+      "loss": 0.6057,
+      "step": 10535
+    },
+    {
+      "epoch": 2.9185595567867035,
+      "grad_norm": 0.7042737007141113,
+      "learning_rate": 2.6252615224569455e-06,
+      "loss": 0.5605,
+      "step": 10536
+    },
+    {
+      "epoch": 2.918836565096953,
+      "grad_norm": 0.7708760499954224,
+      "learning_rate": 2.6248976938167855e-06,
+      "loss": 0.5948,
+      "step": 10537
+    },
+    {
+      "epoch": 2.919113573407202,
+      "grad_norm": 0.7178238034248352,
+      "learning_rate": 2.6245338625247274e-06,
+      "loss": 0.569,
+      "step": 10538
+    },
+    {
+      "epoch": 2.9193905817174515,
+      "grad_norm": 0.7370444536209106,
+      "learning_rate": 2.6241700285884985e-06,
+      "loss": 0.5457,
+      "step": 10539
+    },
+    {
+      "epoch": 2.919667590027701,
+      "grad_norm": 0.7676613926887512,
+      "learning_rate": 2.623806192015822e-06,
+      "loss": 0.5601,
+      "step": 10540
+    },
+    {
+      "epoch": 2.91994459833795,
+      "grad_norm": 0.7429258823394775,
+      "learning_rate": 2.6234423528144243e-06,
+      "loss": 0.6178,
+      "step": 10541
+    },
+    {
+      "epoch": 2.9202216066481994,
+      "grad_norm": 0.71916264295578,
+      "learning_rate": 2.62307851099203e-06,
+      "loss": 0.5349,
+      "step": 10542
+    },
+    {
+      "epoch": 2.920498614958449,
+      "grad_norm": 0.7575047612190247,
+      "learning_rate": 2.6227146665563642e-06,
+      "loss": 0.6412,
+      "step": 10543
+    },
+    {
+      "epoch": 2.920775623268698,
+      "grad_norm": 0.7574516534805298,
+      "learning_rate": 2.6223508195151532e-06,
+      "loss": 0.5718,
+      "step": 10544
+    },
+    {
+      "epoch": 2.9210526315789473,
+      "grad_norm": 0.7422577142715454,
+      "learning_rate": 2.621986969876122e-06,
+      "loss": 0.5846,
+      "step": 10545
+    },
+    {
+      "epoch": 2.921329639889197,
+      "grad_norm": 0.7510814666748047,
+      "learning_rate": 2.621623117646996e-06,
+      "loss": 0.5693,
+      "step": 10546
+    },
+    {
+      "epoch": 2.921606648199446,
+      "grad_norm": 0.7736993432044983,
+      "learning_rate": 2.6212592628354994e-06,
+      "loss": 0.6027,
+      "step": 10547
+    },
+    {
+      "epoch": 2.9218836565096953,
+      "grad_norm": 0.7615543007850647,
+      "learning_rate": 2.6208954054493602e-06,
+      "loss": 0.6346,
+      "step": 10548
+    },
+    {
+      "epoch": 2.9221606648199447,
+      "grad_norm": 0.7570958137512207,
+      "learning_rate": 2.6205315454963026e-06,
+      "loss": 0.6041,
+      "step": 10549
+    },
+    {
+      "epoch": 2.9224376731301938,
+      "grad_norm": 0.7129548192024231,
+      "learning_rate": 2.620167682984052e-06,
+      "loss": 0.5973,
+      "step": 10550
+    },
+    {
+      "epoch": 2.922714681440443,
+      "grad_norm": 0.7714894413948059,
+      "learning_rate": 2.6198038179203357e-06,
+      "loss": 0.6143,
+      "step": 10551
+    },
+    {
+      "epoch": 2.9229916897506927,
+      "grad_norm": 0.7106122374534607,
+      "learning_rate": 2.6194399503128776e-06,
+      "loss": 0.5939,
+      "step": 10552
+    },
+    {
+      "epoch": 2.9232686980609417,
+      "grad_norm": 0.7235246300697327,
+      "learning_rate": 2.6190760801694037e-06,
+      "loss": 0.61,
+      "step": 10553
+    },
+    {
+      "epoch": 2.923545706371191,
+      "grad_norm": 0.7699087262153625,
+      "learning_rate": 2.618712207497642e-06,
+      "loss": 0.608,
+      "step": 10554
+    },
+    {
+      "epoch": 2.9238227146814406,
+      "grad_norm": 0.7375195622444153,
+      "learning_rate": 2.618348332305316e-06,
+      "loss": 0.5718,
+      "step": 10555
+    },
+    {
+      "epoch": 2.9240997229916896,
+      "grad_norm": 0.7720207571983337,
+      "learning_rate": 2.617984454600153e-06,
+      "loss": 0.6399,
+      "step": 10556
+    },
+    {
+      "epoch": 2.924376731301939,
+      "grad_norm": 0.806556224822998,
+      "learning_rate": 2.617620574389878e-06,
+      "loss": 0.6269,
+      "step": 10557
+    },
+    {
+      "epoch": 2.9246537396121886,
+      "grad_norm": 0.811886727809906,
+      "learning_rate": 2.617256691682219e-06,
+      "loss": 0.6067,
+      "step": 10558
+    },
+    {
+      "epoch": 2.9249307479224376,
+      "grad_norm": 0.7293379306793213,
+      "learning_rate": 2.6168928064849004e-06,
+      "loss": 0.5955,
+      "step": 10559
+    },
+    {
+      "epoch": 2.925207756232687,
+      "grad_norm": 0.7429826259613037,
+      "learning_rate": 2.6165289188056485e-06,
+      "loss": 0.539,
+      "step": 10560
+    },
+    {
+      "epoch": 2.9254847645429365,
+      "grad_norm": 0.763191819190979,
+      "learning_rate": 2.616165028652192e-06,
+      "loss": 0.5943,
+      "step": 10561
+    },
+    {
+      "epoch": 2.9257617728531855,
+      "grad_norm": 0.7543225288391113,
+      "learning_rate": 2.615801136032253e-06,
+      "loss": 0.5751,
+      "step": 10562
+    },
+    {
+      "epoch": 2.926038781163435,
+      "grad_norm": 0.7704012989997864,
+      "learning_rate": 2.6154372409535613e-06,
+      "loss": 0.5902,
+      "step": 10563
+    },
+    {
+      "epoch": 2.9263157894736844,
+      "grad_norm": 0.8033764958381653,
+      "learning_rate": 2.615073343423842e-06,
+      "loss": 0.5975,
+      "step": 10564
+    },
+    {
+      "epoch": 2.9265927977839334,
+      "grad_norm": 0.7130028009414673,
+      "learning_rate": 2.614709443450822e-06,
+      "loss": 0.5654,
+      "step": 10565
+    },
+    {
+      "epoch": 2.926869806094183,
+      "grad_norm": 0.778659462928772,
+      "learning_rate": 2.6143455410422263e-06,
+      "loss": 0.6033,
+      "step": 10566
+    },
+    {
+      "epoch": 2.9271468144044324,
+      "grad_norm": 0.7242159247398376,
+      "learning_rate": 2.613981636205784e-06,
+      "loss": 0.5871,
+      "step": 10567
+    },
+    {
+      "epoch": 2.9274238227146814,
+      "grad_norm": 0.7225329279899597,
+      "learning_rate": 2.6136177289492205e-06,
+      "loss": 0.6044,
+      "step": 10568
+    },
+    {
+      "epoch": 2.927700831024931,
+      "grad_norm": 0.7506078481674194,
+      "learning_rate": 2.6132538192802624e-06,
+      "loss": 0.5996,
+      "step": 10569
+    },
+    {
+      "epoch": 2.9279778393351803,
+      "grad_norm": 0.7476010322570801,
+      "learning_rate": 2.612889907206636e-06,
+      "loss": 0.6208,
+      "step": 10570
+    },
+    {
+      "epoch": 2.9282548476454293,
+      "grad_norm": 0.7440158128738403,
+      "learning_rate": 2.612525992736069e-06,
+      "loss": 0.5913,
+      "step": 10571
+    },
+    {
+      "epoch": 2.9285318559556788,
+      "grad_norm": 0.7561903595924377,
+      "learning_rate": 2.6121620758762877e-06,
+      "loss": 0.5772,
+      "step": 10572
+    },
+    {
+      "epoch": 2.9288088642659282,
+      "grad_norm": 0.8481956124305725,
+      "learning_rate": 2.6117981566350187e-06,
+      "loss": 0.6188,
+      "step": 10573
+    },
+    {
+      "epoch": 2.9290858725761773,
+      "grad_norm": 0.7656933069229126,
+      "learning_rate": 2.6114342350199907e-06,
+      "loss": 0.6179,
+      "step": 10574
+    },
+    {
+      "epoch": 2.9293628808864267,
+      "grad_norm": 0.7416088581085205,
+      "learning_rate": 2.6110703110389274e-06,
+      "loss": 0.5994,
+      "step": 10575
+    },
+    {
+      "epoch": 2.929639889196676,
+      "grad_norm": 0.7594820857048035,
+      "learning_rate": 2.610706384699559e-06,
+      "loss": 0.6221,
+      "step": 10576
+    },
+    {
+      "epoch": 2.929916897506925,
+      "grad_norm": 0.7453796863555908,
+      "learning_rate": 2.6103424560096114e-06,
+      "loss": 0.5513,
+      "step": 10577
+    },
+    {
+      "epoch": 2.9301939058171746,
+      "grad_norm": 0.7599956393241882,
+      "learning_rate": 2.609978524976811e-06,
+      "loss": 0.61,
+      "step": 10578
+    },
+    {
+      "epoch": 2.9304709141274237,
+      "grad_norm": 0.7576059103012085,
+      "learning_rate": 2.6096145916088855e-06,
+      "loss": 0.5663,
+      "step": 10579
+    },
+    {
+      "epoch": 2.930747922437673,
+      "grad_norm": 0.711409330368042,
+      "learning_rate": 2.609250655913564e-06,
+      "loss": 0.5377,
+      "step": 10580
+    },
+    {
+      "epoch": 2.9310249307479226,
+      "grad_norm": 0.7231387495994568,
+      "learning_rate": 2.60888671789857e-06,
+      "loss": 0.6061,
+      "step": 10581
+    },
+    {
+      "epoch": 2.9313019390581716,
+      "grad_norm": 0.7678942084312439,
+      "learning_rate": 2.6085227775716338e-06,
+      "loss": 0.6152,
+      "step": 10582
+    },
+    {
+      "epoch": 2.931578947368421,
+      "grad_norm": 0.7268967032432556,
+      "learning_rate": 2.6081588349404825e-06,
+      "loss": 0.5442,
+      "step": 10583
+    },
+    {
+      "epoch": 2.9318559556786705,
+      "grad_norm": 0.7112714648246765,
+      "learning_rate": 2.607794890012841e-06,
+      "loss": 0.6094,
+      "step": 10584
+    },
+    {
+      "epoch": 2.9321329639889195,
+      "grad_norm": 0.7436681985855103,
+      "learning_rate": 2.6074309427964405e-06,
+      "loss": 0.5888,
+      "step": 10585
+    },
+    {
+      "epoch": 2.932409972299169,
+      "grad_norm": 0.7648245692253113,
+      "learning_rate": 2.607066993299007e-06,
+      "loss": 0.5992,
+      "step": 10586
+    },
+    {
+      "epoch": 2.932686980609418,
+      "grad_norm": 0.7910863161087036,
+      "learning_rate": 2.606703041528267e-06,
+      "loss": 0.6206,
+      "step": 10587
+    },
+    {
+      "epoch": 2.9329639889196675,
+      "grad_norm": 0.7461734414100647,
+      "learning_rate": 2.6063390874919482e-06,
+      "loss": 0.6609,
+      "step": 10588
+    },
+    {
+      "epoch": 2.933240997229917,
+      "grad_norm": 0.7565977573394775,
+      "learning_rate": 2.60597513119778e-06,
+      "loss": 0.5607,
+      "step": 10589
+    },
+    {
+      "epoch": 2.933518005540166,
+      "grad_norm": 0.7210617065429688,
+      "learning_rate": 2.605611172653489e-06,
+      "loss": 0.6282,
+      "step": 10590
+    },
+    {
+      "epoch": 2.9337950138504154,
+      "grad_norm": 0.765498697757721,
+      "learning_rate": 2.605247211866803e-06,
+      "loss": 0.5997,
+      "step": 10591
+    },
+    {
+      "epoch": 2.934072022160665,
+      "grad_norm": 0.7618787288665771,
+      "learning_rate": 2.6048832488454495e-06,
+      "loss": 0.5885,
+      "step": 10592
+    },
+    {
+      "epoch": 2.934349030470914,
+      "grad_norm": 0.7326313853263855,
+      "learning_rate": 2.6045192835971578e-06,
+      "loss": 0.5532,
+      "step": 10593
+    },
+    {
+      "epoch": 2.9346260387811633,
+      "grad_norm": 0.7318726181983948,
+      "learning_rate": 2.604155316129654e-06,
+      "loss": 0.6062,
+      "step": 10594
+    },
+    {
+      "epoch": 2.934903047091413,
+      "grad_norm": 0.729059100151062,
+      "learning_rate": 2.603791346450667e-06,
+      "loss": 0.5883,
+      "step": 10595
+    },
+    {
+      "epoch": 2.935180055401662,
+      "grad_norm": 0.7532776594161987,
+      "learning_rate": 2.603427374567925e-06,
+      "loss": 0.582,
+      "step": 10596
+    },
+    {
+      "epoch": 2.9354570637119113,
+      "grad_norm": 0.7572320103645325,
+      "learning_rate": 2.603063400489156e-06,
+      "loss": 0.5723,
+      "step": 10597
+    },
+    {
+      "epoch": 2.9357340720221607,
+      "grad_norm": 0.7214776873588562,
+      "learning_rate": 2.6026994242220866e-06,
+      "loss": 0.6043,
+      "step": 10598
+    },
+    {
+      "epoch": 2.9360110803324098,
+      "grad_norm": 0.7422539591789246,
+      "learning_rate": 2.602335445774447e-06,
+      "loss": 0.6308,
+      "step": 10599
+    },
+    {
+      "epoch": 2.936288088642659,
+      "grad_norm": 0.7235873937606812,
+      "learning_rate": 2.6019714651539647e-06,
+      "loss": 0.6344,
+      "step": 10600
+    },
+    {
+      "epoch": 2.9365650969529087,
+      "grad_norm": 0.7317372560501099,
+      "learning_rate": 2.6016074823683666e-06,
+      "loss": 0.5976,
+      "step": 10601
+    },
+    {
+      "epoch": 2.9368421052631577,
+      "grad_norm": 0.7706679105758667,
+      "learning_rate": 2.6012434974253832e-06,
+      "loss": 0.6308,
+      "step": 10602
+    },
+    {
+      "epoch": 2.937119113573407,
+      "grad_norm": 0.7357689738273621,
+      "learning_rate": 2.6008795103327416e-06,
+      "loss": 0.5425,
+      "step": 10603
+    },
+    {
+      "epoch": 2.9373961218836566,
+      "grad_norm": 0.770684540271759,
+      "learning_rate": 2.600515521098171e-06,
+      "loss": 0.6076,
+      "step": 10604
+    },
+    {
+      "epoch": 2.9376731301939056,
+      "grad_norm": 0.7773721814155579,
+      "learning_rate": 2.6001515297293982e-06,
+      "loss": 0.5507,
+      "step": 10605
+    },
+    {
+      "epoch": 2.937950138504155,
+      "grad_norm": 0.7833162546157837,
+      "learning_rate": 2.5997875362341535e-06,
+      "loss": 0.5845,
+      "step": 10606
+    },
+    {
+      "epoch": 2.9382271468144046,
+      "grad_norm": 0.7654193639755249,
+      "learning_rate": 2.599423540620164e-06,
+      "loss": 0.5849,
+      "step": 10607
+    },
+    {
+      "epoch": 2.9385041551246536,
+      "grad_norm": 0.771862268447876,
+      "learning_rate": 2.5990595428951586e-06,
+      "loss": 0.6145,
+      "step": 10608
+    },
+    {
+      "epoch": 2.938781163434903,
+      "grad_norm": 0.7339853644371033,
+      "learning_rate": 2.598695543066867e-06,
+      "loss": 0.5774,
+      "step": 10609
+    },
+    {
+      "epoch": 2.9390581717451525,
+      "grad_norm": 0.7562407851219177,
+      "learning_rate": 2.5983315411430166e-06,
+      "loss": 0.5958,
+      "step": 10610
+    },
+    {
+      "epoch": 2.9393351800554015,
+      "grad_norm": 0.7739925384521484,
+      "learning_rate": 2.5979675371313357e-06,
+      "loss": 0.5826,
+      "step": 10611
+    },
+    {
+      "epoch": 2.939612188365651,
+      "grad_norm": 0.7429248690605164,
+      "learning_rate": 2.597603531039555e-06,
+      "loss": 0.5429,
+      "step": 10612
+    },
+    {
+      "epoch": 2.9398891966759004,
+      "grad_norm": 0.7736132144927979,
+      "learning_rate": 2.5972395228754017e-06,
+      "loss": 0.5891,
+      "step": 10613
+    },
+    {
+      "epoch": 2.9401662049861494,
+      "grad_norm": 0.7905640602111816,
+      "learning_rate": 2.5968755126466044e-06,
+      "loss": 0.5947,
+      "step": 10614
+    },
+    {
+      "epoch": 2.940443213296399,
+      "grad_norm": 0.8147552609443665,
+      "learning_rate": 2.596511500360894e-06,
+      "loss": 0.6118,
+      "step": 10615
+    },
+    {
+      "epoch": 2.9407202216066484,
+      "grad_norm": 0.7183418869972229,
+      "learning_rate": 2.596147486025996e-06,
+      "loss": 0.5513,
+      "step": 10616
+    },
+    {
+      "epoch": 2.9409972299168974,
+      "grad_norm": 0.7094548344612122,
+      "learning_rate": 2.5957834696496424e-06,
+      "loss": 0.5686,
+      "step": 10617
+    },
+    {
+      "epoch": 2.941274238227147,
+      "grad_norm": 0.7191303968429565,
+      "learning_rate": 2.5954194512395614e-06,
+      "loss": 0.5691,
+      "step": 10618
+    },
+    {
+      "epoch": 2.9415512465373963,
+      "grad_norm": 0.680117666721344,
+      "learning_rate": 2.5950554308034813e-06,
+      "loss": 0.5616,
+      "step": 10619
+    },
+    {
+      "epoch": 2.9418282548476453,
+      "grad_norm": 0.779163122177124,
+      "learning_rate": 2.5946914083491314e-06,
+      "loss": 0.6328,
+      "step": 10620
+    },
+    {
+      "epoch": 2.942105263157895,
+      "grad_norm": 0.726834237575531,
+      "learning_rate": 2.594327383884242e-06,
+      "loss": 0.5863,
+      "step": 10621
+    },
+    {
+      "epoch": 2.9423822714681442,
+      "grad_norm": 0.7219435572624207,
+      "learning_rate": 2.5939633574165408e-06,
+      "loss": 0.6116,
+      "step": 10622
+    },
+    {
+      "epoch": 2.9426592797783933,
+      "grad_norm": 0.7471601366996765,
+      "learning_rate": 2.593599328953758e-06,
+      "loss": 0.597,
+      "step": 10623
+    },
+    {
+      "epoch": 2.9429362880886427,
+      "grad_norm": 0.7578615546226501,
+      "learning_rate": 2.593235298503621e-06,
+      "loss": 0.6315,
+      "step": 10624
+    },
+    {
+      "epoch": 2.943213296398892,
+      "grad_norm": 0.716760516166687,
+      "learning_rate": 2.592871266073863e-06,
+      "loss": 0.5851,
+      "step": 10625
+    },
+    {
+      "epoch": 2.943490304709141,
+      "grad_norm": 0.7250704765319824,
+      "learning_rate": 2.592507231672209e-06,
+      "loss": 0.5971,
+      "step": 10626
+    },
+    {
+      "epoch": 2.9437673130193907,
+      "grad_norm": 0.7800359725952148,
+      "learning_rate": 2.5921431953063903e-06,
+      "loss": 0.5734,
+      "step": 10627
+    },
+    {
+      "epoch": 2.94404432132964,
+      "grad_norm": 0.7490966320037842,
+      "learning_rate": 2.591779156984137e-06,
+      "loss": 0.5976,
+      "step": 10628
+    },
+    {
+      "epoch": 2.944321329639889,
+      "grad_norm": 0.7280365824699402,
+      "learning_rate": 2.591415116713177e-06,
+      "loss": 0.6236,
+      "step": 10629
+    },
+    {
+      "epoch": 2.9445983379501386,
+      "grad_norm": 0.7410351037979126,
+      "learning_rate": 2.5910510745012413e-06,
+      "loss": 0.5541,
+      "step": 10630
+    },
+    {
+      "epoch": 2.944875346260388,
+      "grad_norm": 0.7178124785423279,
+      "learning_rate": 2.590687030356058e-06,
+      "loss": 0.6049,
+      "step": 10631
+    },
+    {
+      "epoch": 2.945152354570637,
+      "grad_norm": 0.7093154191970825,
+      "learning_rate": 2.590322984285359e-06,
+      "loss": 0.5973,
+      "step": 10632
+    },
+    {
+      "epoch": 2.9454293628808865,
+      "grad_norm": 0.7441872358322144,
+      "learning_rate": 2.589958936296871e-06,
+      "loss": 0.5981,
+      "step": 10633
+    },
+    {
+      "epoch": 2.945706371191136,
+      "grad_norm": 0.7678994536399841,
+      "learning_rate": 2.5895948863983257e-06,
+      "loss": 0.6114,
+      "step": 10634
+    },
+    {
+      "epoch": 2.945983379501385,
+      "grad_norm": 0.7277671694755554,
+      "learning_rate": 2.5892308345974517e-06,
+      "loss": 0.649,
+      "step": 10635
+    },
+    {
+      "epoch": 2.9462603878116345,
+      "grad_norm": 0.7727528810501099,
+      "learning_rate": 2.5888667809019792e-06,
+      "loss": 0.5758,
+      "step": 10636
+    },
+    {
+      "epoch": 2.946537396121884,
+      "grad_norm": 0.7813266515731812,
+      "learning_rate": 2.5885027253196384e-06,
+      "loss": 0.6389,
+      "step": 10637
+    },
+    {
+      "epoch": 2.946814404432133,
+      "grad_norm": 0.7353212237358093,
+      "learning_rate": 2.5881386678581587e-06,
+      "loss": 0.6088,
+      "step": 10638
+    },
+    {
+      "epoch": 2.9470914127423824,
+      "grad_norm": 0.7160367369651794,
+      "learning_rate": 2.587774608525269e-06,
+      "loss": 0.5732,
+      "step": 10639
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.7289338111877441,
+      "learning_rate": 2.5874105473287015e-06,
+      "loss": 0.6094,
+      "step": 10640
+    },
+    {
+      "epoch": 2.947645429362881,
+      "grad_norm": 0.7848336696624756,
+      "learning_rate": 2.5870464842761844e-06,
+      "loss": 0.5968,
+      "step": 10641
+    },
+    {
+      "epoch": 2.9479224376731303,
+      "grad_norm": 0.7372105717658997,
+      "learning_rate": 2.5866824193754482e-06,
+      "loss": 0.5872,
+      "step": 10642
+    },
+    {
+      "epoch": 2.9481994459833794,
+      "grad_norm": 0.7537047863006592,
+      "learning_rate": 2.586318352634222e-06,
+      "loss": 0.592,
+      "step": 10643
+    },
+    {
+      "epoch": 2.948476454293629,
+      "grad_norm": 0.7874873280525208,
+      "learning_rate": 2.585954284060238e-06,
+      "loss": 0.5799,
+      "step": 10644
+    },
+    {
+      "epoch": 2.9487534626038783,
+      "grad_norm": 0.6731414198875427,
+      "learning_rate": 2.5855902136612247e-06,
+      "loss": 0.5007,
+      "step": 10645
+    },
+    {
+      "epoch": 2.9490304709141273,
+      "grad_norm": 0.747714638710022,
+      "learning_rate": 2.585226141444912e-06,
+      "loss": 0.5941,
+      "step": 10646
+    },
+    {
+      "epoch": 2.9493074792243767,
+      "grad_norm": 0.7554249167442322,
+      "learning_rate": 2.584862067419032e-06,
+      "loss": 0.5363,
+      "step": 10647
+    },
+    {
+      "epoch": 2.9495844875346258,
+      "grad_norm": 0.7585732340812683,
+      "learning_rate": 2.584497991591313e-06,
+      "loss": 0.6131,
+      "step": 10648
+    },
+    {
+      "epoch": 2.9498614958448752,
+      "grad_norm": 0.7329699397087097,
+      "learning_rate": 2.5841339139694856e-06,
+      "loss": 0.5882,
+      "step": 10649
+    },
+    {
+      "epoch": 2.9501385041551247,
+      "grad_norm": 0.7495070695877075,
+      "learning_rate": 2.58376983456128e-06,
+      "loss": 0.582,
+      "step": 10650
+    },
+    {
+      "epoch": 2.9504155124653737,
+      "grad_norm": 0.7455748319625854,
+      "learning_rate": 2.5834057533744272e-06,
+      "loss": 0.578,
+      "step": 10651
+    },
+    {
+      "epoch": 2.950692520775623,
+      "grad_norm": 0.7285454869270325,
+      "learning_rate": 2.583041670416657e-06,
+      "loss": 0.625,
+      "step": 10652
+    },
+    {
+      "epoch": 2.9509695290858726,
+      "grad_norm": 0.7877806425094604,
+      "learning_rate": 2.5826775856957005e-06,
+      "loss": 0.6188,
+      "step": 10653
+    },
+    {
+      "epoch": 2.9512465373961216,
+      "grad_norm": 0.7521032094955444,
+      "learning_rate": 2.5823134992192877e-06,
+      "loss": 0.619,
+      "step": 10654
+    },
+    {
+      "epoch": 2.951523545706371,
+      "grad_norm": 0.7143591046333313,
+      "learning_rate": 2.5819494109951486e-06,
+      "loss": 0.5876,
+      "step": 10655
+    },
+    {
+      "epoch": 2.9518005540166206,
+      "grad_norm": 0.7743513584136963,
+      "learning_rate": 2.5815853210310154e-06,
+      "loss": 0.6256,
+      "step": 10656
+    },
+    {
+      "epoch": 2.9520775623268696,
+      "grad_norm": 0.7508295178413391,
+      "learning_rate": 2.5812212293346163e-06,
+      "loss": 0.5781,
+      "step": 10657
+    },
+    {
+      "epoch": 2.952354570637119,
+      "grad_norm": 0.7436931133270264,
+      "learning_rate": 2.5808571359136836e-06,
+      "loss": 0.6053,
+      "step": 10658
+    },
+    {
+      "epoch": 2.9526315789473685,
+      "grad_norm": 0.7597281336784363,
+      "learning_rate": 2.5804930407759477e-06,
+      "loss": 0.586,
+      "step": 10659
+    },
+    {
+      "epoch": 2.9529085872576175,
+      "grad_norm": 0.7129910588264465,
+      "learning_rate": 2.580128943929139e-06,
+      "loss": 0.5702,
+      "step": 10660
+    },
+    {
+      "epoch": 2.953185595567867,
+      "grad_norm": 0.7608397006988525,
+      "learning_rate": 2.579764845380987e-06,
+      "loss": 0.6034,
+      "step": 10661
+    },
+    {
+      "epoch": 2.9534626038781164,
+      "grad_norm": 0.7523941993713379,
+      "learning_rate": 2.5794007451392246e-06,
+      "loss": 0.5824,
+      "step": 10662
+    },
+    {
+      "epoch": 2.9537396121883654,
+      "grad_norm": 0.7492373585700989,
+      "learning_rate": 2.5790366432115815e-06,
+      "loss": 0.6168,
+      "step": 10663
+    },
+    {
+      "epoch": 2.954016620498615,
+      "grad_norm": 0.7219147682189941,
+      "learning_rate": 2.5786725396057886e-06,
+      "loss": 0.571,
+      "step": 10664
+    },
+    {
+      "epoch": 2.9542936288088644,
+      "grad_norm": 0.7466295957565308,
+      "learning_rate": 2.578308434329576e-06,
+      "loss": 0.5993,
+      "step": 10665
+    },
+    {
+      "epoch": 2.9545706371191134,
+      "grad_norm": 0.7330300807952881,
+      "learning_rate": 2.5779443273906773e-06,
+      "loss": 0.5982,
+      "step": 10666
+    },
+    {
+      "epoch": 2.954847645429363,
+      "grad_norm": 0.7423272132873535,
+      "learning_rate": 2.5775802187968197e-06,
+      "loss": 0.6273,
+      "step": 10667
+    },
+    {
+      "epoch": 2.9551246537396123,
+      "grad_norm": 0.7298309206962585,
+      "learning_rate": 2.5772161085557367e-06,
+      "loss": 0.5768,
+      "step": 10668
+    },
+    {
+      "epoch": 2.9554016620498613,
+      "grad_norm": 0.7723406553268433,
+      "learning_rate": 2.5768519966751587e-06,
+      "loss": 0.605,
+      "step": 10669
+    },
+    {
+      "epoch": 2.955678670360111,
+      "grad_norm": 0.7510731220245361,
+      "learning_rate": 2.5764878831628154e-06,
+      "loss": 0.596,
+      "step": 10670
+    },
+    {
+      "epoch": 2.9559556786703602,
+      "grad_norm": 0.7381873726844788,
+      "learning_rate": 2.57612376802644e-06,
+      "loss": 0.6044,
+      "step": 10671
+    },
+    {
+      "epoch": 2.9562326869806093,
+      "grad_norm": 0.7327526807785034,
+      "learning_rate": 2.5757596512737626e-06,
+      "loss": 0.6019,
+      "step": 10672
+    },
+    {
+      "epoch": 2.9565096952908587,
+      "grad_norm": 0.7478799223899841,
+      "learning_rate": 2.5753955329125148e-06,
+      "loss": 0.6089,
+      "step": 10673
+    },
+    {
+      "epoch": 2.956786703601108,
+      "grad_norm": 0.7106507420539856,
+      "learning_rate": 2.5750314129504255e-06,
+      "loss": 0.5661,
+      "step": 10674
+    },
+    {
+      "epoch": 2.957063711911357,
+      "grad_norm": 0.8187487721443176,
+      "learning_rate": 2.574667291395229e-06,
+      "loss": 0.6081,
+      "step": 10675
+    },
+    {
+      "epoch": 2.9573407202216067,
+      "grad_norm": 0.7382457852363586,
+      "learning_rate": 2.5743031682546555e-06,
+      "loss": 0.5384,
+      "step": 10676
+    },
+    {
+      "epoch": 2.957617728531856,
+      "grad_norm": 0.7540406584739685,
+      "learning_rate": 2.573939043536436e-06,
+      "loss": 0.6568,
+      "step": 10677
+    },
+    {
+      "epoch": 2.957894736842105,
+      "grad_norm": 0.7558385729789734,
+      "learning_rate": 2.5735749172483003e-06,
+      "loss": 0.5759,
+      "step": 10678
+    },
+    {
+      "epoch": 2.9581717451523546,
+      "grad_norm": 0.7608779072761536,
+      "learning_rate": 2.5732107893979834e-06,
+      "loss": 0.5739,
+      "step": 10679
+    },
+    {
+      "epoch": 2.958448753462604,
+      "grad_norm": 0.7513423562049866,
+      "learning_rate": 2.5728466599932127e-06,
+      "loss": 0.5456,
+      "step": 10680
+    },
+    {
+      "epoch": 2.958725761772853,
+      "grad_norm": 0.7584231495857239,
+      "learning_rate": 2.5724825290417223e-06,
+      "loss": 0.586,
+      "step": 10681
+    },
+    {
+      "epoch": 2.9590027700831025,
+      "grad_norm": 0.7205784916877747,
+      "learning_rate": 2.5721183965512424e-06,
+      "loss": 0.5438,
+      "step": 10682
+    },
+    {
+      "epoch": 2.959279778393352,
+      "grad_norm": 0.7798346281051636,
+      "learning_rate": 2.5717542625295044e-06,
+      "loss": 0.5974,
+      "step": 10683
+    },
+    {
+      "epoch": 2.959556786703601,
+      "grad_norm": 0.7180570960044861,
+      "learning_rate": 2.5713901269842405e-06,
+      "loss": 0.5726,
+      "step": 10684
+    },
+    {
+      "epoch": 2.9598337950138505,
+      "grad_norm": 0.7144060134887695,
+      "learning_rate": 2.5710259899231827e-06,
+      "loss": 0.5293,
+      "step": 10685
+    },
+    {
+      "epoch": 2.9601108033241,
+      "grad_norm": 0.7333686947822571,
+      "learning_rate": 2.570661851354061e-06,
+      "loss": 0.5776,
+      "step": 10686
+    },
+    {
+      "epoch": 2.960387811634349,
+      "grad_norm": 0.7347108721733093,
+      "learning_rate": 2.5702977112846073e-06,
+      "loss": 0.5931,
+      "step": 10687
+    },
+    {
+      "epoch": 2.9606648199445984,
+      "grad_norm": 0.7476058602333069,
+      "learning_rate": 2.5699335697225545e-06,
+      "loss": 0.6298,
+      "step": 10688
+    },
+    {
+      "epoch": 2.960941828254848,
+      "grad_norm": 0.760506808757782,
+      "learning_rate": 2.5695694266756326e-06,
+      "loss": 0.6123,
+      "step": 10689
+    },
+    {
+      "epoch": 2.961218836565097,
+      "grad_norm": 0.7429872155189514,
+      "learning_rate": 2.569205282151575e-06,
+      "loss": 0.5671,
+      "step": 10690
+    },
+    {
+      "epoch": 2.9614958448753463,
+      "grad_norm": 0.7676592469215393,
+      "learning_rate": 2.5688411361581123e-06,
+      "loss": 0.5654,
+      "step": 10691
+    },
+    {
+      "epoch": 2.961772853185596,
+      "grad_norm": 0.7633059024810791,
+      "learning_rate": 2.5684769887029764e-06,
+      "loss": 0.5604,
+      "step": 10692
+    },
+    {
+      "epoch": 2.962049861495845,
+      "grad_norm": 0.719962477684021,
+      "learning_rate": 2.5681128397938987e-06,
+      "loss": 0.6238,
+      "step": 10693
+    },
+    {
+      "epoch": 2.9623268698060943,
+      "grad_norm": 0.7325805425643921,
+      "learning_rate": 2.5677486894386118e-06,
+      "loss": 0.5728,
+      "step": 10694
+    },
+    {
+      "epoch": 2.9626038781163437,
+      "grad_norm": 0.7231216430664062,
+      "learning_rate": 2.5673845376448473e-06,
+      "loss": 0.5906,
+      "step": 10695
+    },
+    {
+      "epoch": 2.9628808864265928,
+      "grad_norm": 0.7547472715377808,
+      "learning_rate": 2.567020384420337e-06,
+      "loss": 0.6111,
+      "step": 10696
+    },
+    {
+      "epoch": 2.963157894736842,
+      "grad_norm": 0.7770267128944397,
+      "learning_rate": 2.566656229772812e-06,
+      "loss": 0.5794,
+      "step": 10697
+    },
+    {
+      "epoch": 2.9634349030470917,
+      "grad_norm": 0.7602913975715637,
+      "learning_rate": 2.566292073710006e-06,
+      "loss": 0.6388,
+      "step": 10698
+    },
+    {
+      "epoch": 2.9637119113573407,
+      "grad_norm": 0.752708375453949,
+      "learning_rate": 2.56592791623965e-06,
+      "loss": 0.5852,
+      "step": 10699
+    },
+    {
+      "epoch": 2.96398891966759,
+      "grad_norm": 0.7117756009101868,
+      "learning_rate": 2.565563757369475e-06,
+      "loss": 0.5784,
+      "step": 10700
+    },
+    {
+      "epoch": 2.9642659279778396,
+      "grad_norm": 0.7588861584663391,
+      "learning_rate": 2.565199597107216e-06,
+      "loss": 0.5937,
+      "step": 10701
+    },
+    {
+      "epoch": 2.9645429362880886,
+      "grad_norm": 0.7347100377082825,
+      "learning_rate": 2.5648354354606007e-06,
+      "loss": 0.6163,
+      "step": 10702
+    },
+    {
+      "epoch": 2.964819944598338,
+      "grad_norm": 0.751549482345581,
+      "learning_rate": 2.5644712724373648e-06,
+      "loss": 0.6389,
+      "step": 10703
+    },
+    {
+      "epoch": 2.965096952908587,
+      "grad_norm": 0.7782813906669617,
+      "learning_rate": 2.564107108045239e-06,
+      "loss": 0.6091,
+      "step": 10704
+    },
+    {
+      "epoch": 2.9653739612188366,
+      "grad_norm": 0.6835278868675232,
+      "learning_rate": 2.5637429422919556e-06,
+      "loss": 0.5734,
+      "step": 10705
+    },
+    {
+      "epoch": 2.965650969529086,
+      "grad_norm": 0.7344762086868286,
+      "learning_rate": 2.563378775185246e-06,
+      "loss": 0.5731,
+      "step": 10706
+    },
+    {
+      "epoch": 2.965927977839335,
+      "grad_norm": 0.7300355434417725,
+      "learning_rate": 2.563014606732844e-06,
+      "loss": 0.6025,
+      "step": 10707
+    },
+    {
+      "epoch": 2.9662049861495845,
+      "grad_norm": 0.7507689595222473,
+      "learning_rate": 2.562650436942481e-06,
+      "loss": 0.5756,
+      "step": 10708
+    },
+    {
+      "epoch": 2.966481994459834,
+      "grad_norm": 0.7238181233406067,
+      "learning_rate": 2.5622862658218893e-06,
+      "loss": 0.5627,
+      "step": 10709
+    },
+    {
+      "epoch": 2.966759002770083,
+      "grad_norm": 0.7482140064239502,
+      "learning_rate": 2.5619220933788007e-06,
+      "loss": 0.5917,
+      "step": 10710
+    },
+    {
+      "epoch": 2.9670360110803324,
+      "grad_norm": 0.7118654847145081,
+      "learning_rate": 2.5615579196209477e-06,
+      "loss": 0.595,
+      "step": 10711
+    },
+    {
+      "epoch": 2.9673130193905815,
+      "grad_norm": 0.7792865037918091,
+      "learning_rate": 2.5611937445560635e-06,
+      "loss": 0.5944,
+      "step": 10712
+    },
+    {
+      "epoch": 2.967590027700831,
+      "grad_norm": 0.7508601546287537,
+      "learning_rate": 2.5608295681918794e-06,
+      "loss": 0.6245,
+      "step": 10713
+    },
+    {
+      "epoch": 2.9678670360110804,
+      "grad_norm": 0.7758817672729492,
+      "learning_rate": 2.5604653905361286e-06,
+      "loss": 0.6158,
+      "step": 10714
+    },
+    {
+      "epoch": 2.9681440443213294,
+      "grad_norm": 0.7338311076164246,
+      "learning_rate": 2.5601012115965425e-06,
+      "loss": 0.557,
+      "step": 10715
+    },
+    {
+      "epoch": 2.968421052631579,
+      "grad_norm": 0.7640522122383118,
+      "learning_rate": 2.5597370313808552e-06,
+      "loss": 0.6413,
+      "step": 10716
+    },
+    {
+      "epoch": 2.9686980609418283,
+      "grad_norm": 0.7294602394104004,
+      "learning_rate": 2.5593728498967975e-06,
+      "loss": 0.5962,
+      "step": 10717
+    },
+    {
+      "epoch": 2.9689750692520773,
+      "grad_norm": 0.744451642036438,
+      "learning_rate": 2.5590086671521023e-06,
+      "loss": 0.5961,
+      "step": 10718
+    },
+    {
+      "epoch": 2.969252077562327,
+      "grad_norm": 0.7545126676559448,
+      "learning_rate": 2.5586444831545027e-06,
+      "loss": 0.5904,
+      "step": 10719
+    },
+    {
+      "epoch": 2.9695290858725762,
+      "grad_norm": 0.7299168705940247,
+      "learning_rate": 2.5582802979117316e-06,
+      "loss": 0.5959,
+      "step": 10720
+    },
+    {
+      "epoch": 2.9698060941828253,
+      "grad_norm": 0.7411100268363953,
+      "learning_rate": 2.55791611143152e-06,
+      "loss": 0.6012,
+      "step": 10721
+    },
+    {
+      "epoch": 2.9700831024930747,
+      "grad_norm": 0.7515658736228943,
+      "learning_rate": 2.557551923721602e-06,
+      "loss": 0.6036,
+      "step": 10722
+    },
+    {
+      "epoch": 2.970360110803324,
+      "grad_norm": 0.7271835207939148,
+      "learning_rate": 2.5571877347897096e-06,
+      "loss": 0.55,
+      "step": 10723
+    },
+    {
+      "epoch": 2.970637119113573,
+      "grad_norm": 0.7838777303695679,
+      "learning_rate": 2.5568235446435757e-06,
+      "loss": 0.5862,
+      "step": 10724
+    },
+    {
+      "epoch": 2.9709141274238227,
+      "grad_norm": 0.7445371150970459,
+      "learning_rate": 2.556459353290932e-06,
+      "loss": 0.6078,
+      "step": 10725
+    },
+    {
+      "epoch": 2.971191135734072,
+      "grad_norm": 0.7580215930938721,
+      "learning_rate": 2.556095160739513e-06,
+      "loss": 0.6034,
+      "step": 10726
+    },
+    {
+      "epoch": 2.971468144044321,
+      "grad_norm": 0.7328695058822632,
+      "learning_rate": 2.55573096699705e-06,
+      "loss": 0.5894,
+      "step": 10727
+    },
+    {
+      "epoch": 2.9717451523545706,
+      "grad_norm": 0.7478681802749634,
+      "learning_rate": 2.555366772071276e-06,
+      "loss": 0.5915,
+      "step": 10728
+    },
+    {
+      "epoch": 2.97202216066482,
+      "grad_norm": 0.7482941150665283,
+      "learning_rate": 2.555002575969924e-06,
+      "loss": 0.6263,
+      "step": 10729
+    },
+    {
+      "epoch": 2.972299168975069,
+      "grad_norm": 0.7541418075561523,
+      "learning_rate": 2.5546383787007276e-06,
+      "loss": 0.6189,
+      "step": 10730
+    },
+    {
+      "epoch": 2.9725761772853185,
+      "grad_norm": 0.7668107748031616,
+      "learning_rate": 2.5542741802714188e-06,
+      "loss": 0.6036,
+      "step": 10731
+    },
+    {
+      "epoch": 2.972853185595568,
+      "grad_norm": 0.770712673664093,
+      "learning_rate": 2.55390998068973e-06,
+      "loss": 0.5466,
+      "step": 10732
+    },
+    {
+      "epoch": 2.973130193905817,
+      "grad_norm": 0.801234245300293,
+      "learning_rate": 2.5535457799633957e-06,
+      "loss": 0.5953,
+      "step": 10733
+    },
+    {
+      "epoch": 2.9734072022160665,
+      "grad_norm": 0.708030104637146,
+      "learning_rate": 2.5531815781001462e-06,
+      "loss": 0.567,
+      "step": 10734
+    },
+    {
+      "epoch": 2.973684210526316,
+      "grad_norm": 0.7175045013427734,
+      "learning_rate": 2.552817375107717e-06,
+      "loss": 0.5635,
+      "step": 10735
+    },
+    {
+      "epoch": 2.973961218836565,
+      "grad_norm": 0.7897862195968628,
+      "learning_rate": 2.55245317099384e-06,
+      "loss": 0.619,
+      "step": 10736
+    },
+    {
+      "epoch": 2.9742382271468144,
+      "grad_norm": 0.8118829727172852,
+      "learning_rate": 2.5520889657662482e-06,
+      "loss": 0.5977,
+      "step": 10737
+    },
+    {
+      "epoch": 2.974515235457064,
+      "grad_norm": 0.7560293078422546,
+      "learning_rate": 2.5517247594326744e-06,
+      "loss": 0.6059,
+      "step": 10738
+    },
+    {
+      "epoch": 2.974792243767313,
+      "grad_norm": 0.7857494354248047,
+      "learning_rate": 2.5513605520008516e-06,
+      "loss": 0.6039,
+      "step": 10739
+    },
+    {
+      "epoch": 2.9750692520775623,
+      "grad_norm": 0.7377294301986694,
+      "learning_rate": 2.550996343478514e-06,
+      "loss": 0.5649,
+      "step": 10740
+    },
+    {
+      "epoch": 2.975346260387812,
+      "grad_norm": 0.8070437908172607,
+      "learning_rate": 2.5506321338733924e-06,
+      "loss": 0.5845,
+      "step": 10741
+    },
+    {
+      "epoch": 2.975623268698061,
+      "grad_norm": 0.757997989654541,
+      "learning_rate": 2.5502679231932232e-06,
+      "loss": 0.5986,
+      "step": 10742
+    },
+    {
+      "epoch": 2.9759002770083103,
+      "grad_norm": 0.7478947639465332,
+      "learning_rate": 2.5499037114457358e-06,
+      "loss": 0.6166,
+      "step": 10743
+    },
+    {
+      "epoch": 2.9761772853185597,
+      "grad_norm": 0.7756154537200928,
+      "learning_rate": 2.549539498638666e-06,
+      "loss": 0.5843,
+      "step": 10744
+    },
+    {
+      "epoch": 2.9764542936288088,
+      "grad_norm": 0.8048810362815857,
+      "learning_rate": 2.549175284779746e-06,
+      "loss": 0.6279,
+      "step": 10745
+    },
+    {
+      "epoch": 2.976731301939058,
+      "grad_norm": 0.7790731191635132,
+      "learning_rate": 2.54881106987671e-06,
+      "loss": 0.5502,
+      "step": 10746
+    },
+    {
+      "epoch": 2.9770083102493077,
+      "grad_norm": 0.7683466076850891,
+      "learning_rate": 2.548446853937289e-06,
+      "loss": 0.5858,
+      "step": 10747
+    },
+    {
+      "epoch": 2.9772853185595567,
+      "grad_norm": 0.7681669592857361,
+      "learning_rate": 2.5480826369692178e-06,
+      "loss": 0.5647,
+      "step": 10748
+    },
+    {
+      "epoch": 2.977562326869806,
+      "grad_norm": 0.7260222434997559,
+      "learning_rate": 2.54771841898023e-06,
+      "loss": 0.6206,
+      "step": 10749
+    },
+    {
+      "epoch": 2.9778393351800556,
+      "grad_norm": 0.7287525534629822,
+      "learning_rate": 2.5473541999780583e-06,
+      "loss": 0.5813,
+      "step": 10750
+    },
+    {
+      "epoch": 2.9781163434903046,
+      "grad_norm": 0.7364113926887512,
+      "learning_rate": 2.5469899799704356e-06,
+      "loss": 0.5709,
+      "step": 10751
+    },
+    {
+      "epoch": 2.978393351800554,
+      "grad_norm": 0.7547378540039062,
+      "learning_rate": 2.546625758965096e-06,
+      "loss": 0.6324,
+      "step": 10752
+    },
+    {
+      "epoch": 2.9786703601108035,
+      "grad_norm": 0.7311744689941406,
+      "learning_rate": 2.5462615369697726e-06,
+      "loss": 0.5771,
+      "step": 10753
+    },
+    {
+      "epoch": 2.9789473684210526,
+      "grad_norm": 0.753720760345459,
+      "learning_rate": 2.5458973139921974e-06,
+      "loss": 0.5625,
+      "step": 10754
+    },
+    {
+      "epoch": 2.979224376731302,
+      "grad_norm": 0.7415842413902283,
+      "learning_rate": 2.5455330900401066e-06,
+      "loss": 0.5854,
+      "step": 10755
+    },
+    {
+      "epoch": 2.9795013850415515,
+      "grad_norm": 0.7530409693717957,
+      "learning_rate": 2.545168865121231e-06,
+      "loss": 0.5897,
+      "step": 10756
+    },
+    {
+      "epoch": 2.9797783933518005,
+      "grad_norm": 0.7591843008995056,
+      "learning_rate": 2.5448046392433057e-06,
+      "loss": 0.5495,
+      "step": 10757
+    },
+    {
+      "epoch": 2.98005540166205,
+      "grad_norm": 0.7462211847305298,
+      "learning_rate": 2.544440412414063e-06,
+      "loss": 0.5712,
+      "step": 10758
+    },
+    {
+      "epoch": 2.9803324099722994,
+      "grad_norm": 0.716187596321106,
+      "learning_rate": 2.544076184641237e-06,
+      "loss": 0.593,
+      "step": 10759
+    },
+    {
+      "epoch": 2.9806094182825484,
+      "grad_norm": 0.7935730814933777,
+      "learning_rate": 2.5437119559325606e-06,
+      "loss": 0.602,
+      "step": 10760
+    },
+    {
+      "epoch": 2.980886426592798,
+      "grad_norm": 0.7291024923324585,
+      "learning_rate": 2.5433477262957686e-06,
+      "loss": 0.5809,
+      "step": 10761
+    },
+    {
+      "epoch": 2.9811634349030474,
+      "grad_norm": 0.7451568841934204,
+      "learning_rate": 2.5429834957385937e-06,
+      "loss": 0.5973,
+      "step": 10762
+    },
+    {
+      "epoch": 2.9814404432132964,
+      "grad_norm": 0.8070505261421204,
+      "learning_rate": 2.5426192642687687e-06,
+      "loss": 0.6404,
+      "step": 10763
+    },
+    {
+      "epoch": 2.981717451523546,
+      "grad_norm": 0.7522998452186584,
+      "learning_rate": 2.5422550318940283e-06,
+      "loss": 0.58,
+      "step": 10764
+    },
+    {
+      "epoch": 2.981994459833795,
+      "grad_norm": 0.7420567274093628,
+      "learning_rate": 2.541890798622106e-06,
+      "loss": 0.5826,
+      "step": 10765
+    },
+    {
+      "epoch": 2.9822714681440443,
+      "grad_norm": 0.7697232961654663,
+      "learning_rate": 2.5415265644607338e-06,
+      "loss": 0.6362,
+      "step": 10766
+    },
+    {
+      "epoch": 2.9825484764542938,
+      "grad_norm": 0.7693185210227966,
+      "learning_rate": 2.5411623294176473e-06,
+      "loss": 0.6172,
+      "step": 10767
+    },
+    {
+      "epoch": 2.982825484764543,
+      "grad_norm": 0.6835063695907593,
+      "learning_rate": 2.54079809350058e-06,
+      "loss": 0.552,
+      "step": 10768
+    },
+    {
+      "epoch": 2.9831024930747922,
+      "grad_norm": 0.77583247423172,
+      "learning_rate": 2.5404338567172635e-06,
+      "loss": 0.6076,
+      "step": 10769
+    },
+    {
+      "epoch": 2.9833795013850417,
+      "grad_norm": 0.7879008054733276,
+      "learning_rate": 2.5400696190754347e-06,
+      "loss": 0.5839,
+      "step": 10770
+    },
+    {
+      "epoch": 2.9836565096952907,
+      "grad_norm": 0.7527600526809692,
+      "learning_rate": 2.5397053805828248e-06,
+      "loss": 0.5529,
+      "step": 10771
+    },
+    {
+      "epoch": 2.98393351800554,
+      "grad_norm": 0.7148284912109375,
+      "learning_rate": 2.5393411412471685e-06,
+      "loss": 0.6178,
+      "step": 10772
+    },
+    {
+      "epoch": 2.984210526315789,
+      "grad_norm": 0.8008612394332886,
+      "learning_rate": 2.538976901076199e-06,
+      "loss": 0.564,
+      "step": 10773
+    },
+    {
+      "epoch": 2.9844875346260387,
+      "grad_norm": 0.7103780508041382,
+      "learning_rate": 2.538612660077651e-06,
+      "loss": 0.5598,
+      "step": 10774
+    },
+    {
+      "epoch": 2.984764542936288,
+      "grad_norm": 0.7940399050712585,
+      "learning_rate": 2.5382484182592567e-06,
+      "loss": 0.6056,
+      "step": 10775
+    },
+    {
+      "epoch": 2.985041551246537,
+      "grad_norm": 0.7633231282234192,
+      "learning_rate": 2.537884175628751e-06,
+      "loss": 0.5574,
+      "step": 10776
+    },
+    {
+      "epoch": 2.9853185595567866,
+      "grad_norm": 0.7409139275550842,
+      "learning_rate": 2.537519932193868e-06,
+      "loss": 0.5227,
+      "step": 10777
+    },
+    {
+      "epoch": 2.985595567867036,
+      "grad_norm": 0.7334606051445007,
+      "learning_rate": 2.537155687962341e-06,
+      "loss": 0.6251,
+      "step": 10778
+    },
+    {
+      "epoch": 2.985872576177285,
+      "grad_norm": 0.7315794825553894,
+      "learning_rate": 2.536791442941903e-06,
+      "loss": 0.5971,
+      "step": 10779
+    },
+    {
+      "epoch": 2.9861495844875345,
+      "grad_norm": 0.735944390296936,
+      "learning_rate": 2.5364271971402893e-06,
+      "loss": 0.6195,
+      "step": 10780
+    },
+    {
+      "epoch": 2.986426592797784,
+      "grad_norm": 0.7769847512245178,
+      "learning_rate": 2.5360629505652334e-06,
+      "loss": 0.6028,
+      "step": 10781
+    },
+    {
+      "epoch": 2.986703601108033,
+      "grad_norm": 0.7395898103713989,
+      "learning_rate": 2.5356987032244686e-06,
+      "loss": 0.5689,
+      "step": 10782
+    },
+    {
+      "epoch": 2.9869806094182825,
+      "grad_norm": 0.7860537767410278,
+      "learning_rate": 2.535334455125729e-06,
+      "loss": 0.6261,
+      "step": 10783
+    },
+    {
+      "epoch": 2.987257617728532,
+      "grad_norm": 0.7746903300285339,
+      "learning_rate": 2.534970206276749e-06,
+      "loss": 0.5677,
+      "step": 10784
+    },
+    {
+      "epoch": 2.987534626038781,
+      "grad_norm": 0.7295857071876526,
+      "learning_rate": 2.5346059566852626e-06,
+      "loss": 0.5774,
+      "step": 10785
+    },
+    {
+      "epoch": 2.9878116343490304,
+      "grad_norm": 0.7481796741485596,
+      "learning_rate": 2.5342417063590023e-06,
+      "loss": 0.5934,
+      "step": 10786
+    },
+    {
+      "epoch": 2.98808864265928,
+      "grad_norm": 0.7683750987052917,
+      "learning_rate": 2.5338774553057044e-06,
+      "loss": 0.6032,
+      "step": 10787
+    },
+    {
+      "epoch": 2.988365650969529,
+      "grad_norm": 0.7441554665565491,
+      "learning_rate": 2.5335132035331006e-06,
+      "loss": 0.6215,
+      "step": 10788
+    },
+    {
+      "epoch": 2.9886426592797783,
+      "grad_norm": 0.729632556438446,
+      "learning_rate": 2.5331489510489265e-06,
+      "loss": 0.5929,
+      "step": 10789
+    },
+    {
+      "epoch": 2.988919667590028,
+      "grad_norm": 0.7185140252113342,
+      "learning_rate": 2.5327846978609154e-06,
+      "loss": 0.5782,
+      "step": 10790
+    },
+    {
+      "epoch": 2.989196675900277,
+      "grad_norm": 0.7277169227600098,
+      "learning_rate": 2.532420443976802e-06,
+      "loss": 0.5951,
+      "step": 10791
+    },
+    {
+      "epoch": 2.9894736842105263,
+      "grad_norm": 0.7362697720527649,
+      "learning_rate": 2.532056189404318e-06,
+      "loss": 0.5914,
+      "step": 10792
+    },
+    {
+      "epoch": 2.9897506925207757,
+      "grad_norm": 0.7093611359596252,
+      "learning_rate": 2.5316919341512013e-06,
+      "loss": 0.5739,
+      "step": 10793
+    },
+    {
+      "epoch": 2.9900277008310248,
+      "grad_norm": 0.7320445775985718,
+      "learning_rate": 2.531327678225183e-06,
+      "loss": 0.5517,
+      "step": 10794
+    },
+    {
+      "epoch": 2.990304709141274,
+      "grad_norm": 0.7347859740257263,
+      "learning_rate": 2.530963421633998e-06,
+      "loss": 0.564,
+      "step": 10795
+    },
+    {
+      "epoch": 2.9905817174515237,
+      "grad_norm": 0.7222632765769958,
+      "learning_rate": 2.5305991643853804e-06,
+      "loss": 0.599,
+      "step": 10796
+    },
+    {
+      "epoch": 2.9908587257617727,
+      "grad_norm": 0.7426832318305969,
+      "learning_rate": 2.5302349064870644e-06,
+      "loss": 0.5932,
+      "step": 10797
+    },
+    {
+      "epoch": 2.991135734072022,
+      "grad_norm": 0.7762872576713562,
+      "learning_rate": 2.5298706479467843e-06,
+      "loss": 0.5427,
+      "step": 10798
+    },
+    {
+      "epoch": 2.9914127423822716,
+      "grad_norm": 0.7226398587226868,
+      "learning_rate": 2.5295063887722742e-06,
+      "loss": 0.5753,
+      "step": 10799
+    },
+    {
+      "epoch": 2.9916897506925206,
+      "grad_norm": 0.7660290002822876,
+      "learning_rate": 2.529142128971268e-06,
+      "loss": 0.5759,
+      "step": 10800
+    },
+    {
+      "epoch": 2.99196675900277,
+      "grad_norm": 0.7273073792457581,
+      "learning_rate": 2.5287778685514998e-06,
+      "loss": 0.5764,
+      "step": 10801
+    },
+    {
+      "epoch": 2.9922437673130196,
+      "grad_norm": 0.7794416546821594,
+      "learning_rate": 2.5284136075207043e-06,
+      "loss": 0.6293,
+      "step": 10802
+    },
+    {
+      "epoch": 2.9925207756232686,
+      "grad_norm": 0.7536942362785339,
+      "learning_rate": 2.528049345886615e-06,
+      "loss": 0.5782,
+      "step": 10803
+    },
+    {
+      "epoch": 2.992797783933518,
+      "grad_norm": 0.7645159363746643,
+      "learning_rate": 2.5276850836569665e-06,
+      "loss": 0.6137,
+      "step": 10804
+    },
+    {
+      "epoch": 2.9930747922437675,
+      "grad_norm": 0.7383338212966919,
+      "learning_rate": 2.5273208208394927e-06,
+      "loss": 0.6231,
+      "step": 10805
+    },
+    {
+      "epoch": 2.9933518005540165,
+      "grad_norm": 0.7565742135047913,
+      "learning_rate": 2.5269565574419297e-06,
+      "loss": 0.5862,
+      "step": 10806
+    },
+    {
+      "epoch": 2.993628808864266,
+      "grad_norm": 0.772453248500824,
+      "learning_rate": 2.5265922934720082e-06,
+      "loss": 0.5827,
+      "step": 10807
+    },
+    {
+      "epoch": 2.9939058171745154,
+      "grad_norm": 0.751247227191925,
+      "learning_rate": 2.5262280289374652e-06,
+      "loss": 0.5273,
+      "step": 10808
+    },
+    {
+      "epoch": 2.9941828254847644,
+      "grad_norm": 0.7555593848228455,
+      "learning_rate": 2.525863763846034e-06,
+      "loss": 0.5786,
+      "step": 10809
+    },
+    {
+      "epoch": 2.994459833795014,
+      "grad_norm": 0.753018856048584,
+      "learning_rate": 2.5254994982054494e-06,
+      "loss": 0.5976,
+      "step": 10810
+    },
+    {
+      "epoch": 2.9947368421052634,
+      "grad_norm": 0.7415595650672913,
+      "learning_rate": 2.525135232023445e-06,
+      "loss": 0.5904,
+      "step": 10811
+    },
+    {
+      "epoch": 2.9950138504155124,
+      "grad_norm": 0.7019715905189514,
+      "learning_rate": 2.524770965307756e-06,
+      "loss": 0.5546,
+      "step": 10812
+    },
+    {
+      "epoch": 2.995290858725762,
+      "grad_norm": 0.7442302703857422,
+      "learning_rate": 2.5244066980661154e-06,
+      "loss": 0.5846,
+      "step": 10813
+    },
+    {
+      "epoch": 2.9955678670360113,
+      "grad_norm": 0.7354565262794495,
+      "learning_rate": 2.524042430306258e-06,
+      "loss": 0.6103,
+      "step": 10814
+    },
+    {
+      "epoch": 2.9958448753462603,
+      "grad_norm": 0.755585253238678,
+      "learning_rate": 2.5236781620359196e-06,
+      "loss": 0.6171,
+      "step": 10815
+    },
+    {
+      "epoch": 2.9961218836565098,
+      "grad_norm": 0.7500039935112,
+      "learning_rate": 2.5233138932628326e-06,
+      "loss": 0.5986,
+      "step": 10816
+    },
+    {
+      "epoch": 2.9963988919667592,
+      "grad_norm": 0.7086367011070251,
+      "learning_rate": 2.522949623994732e-06,
+      "loss": 0.6021,
+      "step": 10817
+    },
+    {
+      "epoch": 2.9966759002770083,
+      "grad_norm": 0.7707514762878418,
+      "learning_rate": 2.522585354239353e-06,
+      "loss": 0.6043,
+      "step": 10818
+    },
+    {
+      "epoch": 2.9969529085872577,
+      "grad_norm": 0.7526091933250427,
+      "learning_rate": 2.5222210840044287e-06,
+      "loss": 0.5745,
+      "step": 10819
+    },
+    {
+      "epoch": 2.997229916897507,
+      "grad_norm": 0.7560510635375977,
+      "learning_rate": 2.521856813297694e-06,
+      "loss": 0.6333,
+      "step": 10820
+    },
+    {
+      "epoch": 2.997506925207756,
+      "grad_norm": 0.7667151093482971,
+      "learning_rate": 2.521492542126883e-06,
+      "loss": 0.6159,
+      "step": 10821
+    },
+    {
+      "epoch": 2.9977839335180057,
+      "grad_norm": 0.7421969771385193,
+      "learning_rate": 2.5211282704997314e-06,
+      "loss": 0.6622,
+      "step": 10822
+    },
+    {
+      "epoch": 2.998060941828255,
+      "grad_norm": 0.7187398672103882,
+      "learning_rate": 2.5207639984239723e-06,
+      "loss": 0.5686,
+      "step": 10823
+    },
+    {
+      "epoch": 2.998337950138504,
+      "grad_norm": 0.7653238773345947,
+      "learning_rate": 2.5203997259073398e-06,
+      "loss": 0.5997,
+      "step": 10824
+    },
+    {
+      "epoch": 2.9986149584487536,
+      "grad_norm": 0.6978490352630615,
+      "learning_rate": 2.5200354529575693e-06,
+      "loss": 0.6116,
+      "step": 10825
+    },
+    {
+      "epoch": 2.998891966759003,
+      "grad_norm": 0.749488115310669,
+      "learning_rate": 2.5196711795823957e-06,
+      "loss": 0.5505,
+      "step": 10826
+    },
+    {
+      "epoch": 2.999168975069252,
+      "grad_norm": 0.7787477374076843,
+      "learning_rate": 2.519306905789552e-06,
+      "loss": 0.6134,
+      "step": 10827
+    },
+    {
+      "epoch": 2.9994459833795015,
+      "grad_norm": 0.7261177897453308,
+      "learning_rate": 2.518942631586774e-06,
+      "loss": 0.5708,
+      "step": 10828
+    },
+    {
+      "epoch": 2.9997229916897505,
+      "grad_norm": 0.7337374091148376,
+      "learning_rate": 2.5185783569817944e-06,
+      "loss": 0.5427,
+      "step": 10829
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7540554404258728,
+      "learning_rate": 2.5182140819823503e-06,
+      "loss": 0.5834,
+      "step": 10830
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 21660,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 3610,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.426074873341149e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}