diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9077 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.998067259373792,
+  "eval_steps": 500,
+  "global_step": 1292,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0015461925009663702,
+      "grad_norm": 0.305040568113327,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.9801,
+      "step": 1
+    },
+    {
+      "epoch": 0.0030923850019327404,
+      "grad_norm": 0.3316026031970978,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.0582,
+      "step": 2
+    },
+    {
+      "epoch": 0.004638577502899111,
+      "grad_norm": 0.321511834859848,
+      "learning_rate": 1.5e-06,
+      "loss": 1.0877,
+      "step": 3
+    },
+    {
+      "epoch": 0.006184770003865481,
+      "grad_norm": 0.32884451746940613,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.1082,
+      "step": 4
+    },
+    {
+      "epoch": 0.007730962504831852,
+      "grad_norm": 0.3244039714336395,
+      "learning_rate": 2.5e-06,
+      "loss": 1.1352,
+      "step": 5
+    },
+    {
+      "epoch": 0.009277155005798222,
+      "grad_norm": 0.319381445646286,
+      "learning_rate": 3e-06,
+      "loss": 1.0808,
+      "step": 6
+    },
+    {
+      "epoch": 0.010823347506764593,
+      "grad_norm": 0.3265005946159363,
+      "learning_rate": 3.5000000000000004e-06,
+      "loss": 1.0554,
+      "step": 7
+    },
+    {
+      "epoch": 0.012369540007730962,
+      "grad_norm": 0.3248363137245178,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.0834,
+      "step": 8
+    },
+    {
+      "epoch": 0.013915732508697333,
+      "grad_norm": 0.3369300365447998,
+      "learning_rate": 4.5e-06,
+      "loss": 1.1304,
+      "step": 9
+    },
+    {
+      "epoch": 0.015461925009663703,
+      "grad_norm": 0.3302851915359497,
+      "learning_rate": 5e-06,
+      "loss": 1.0683,
+      "step": 10
+    },
+    {
+      "epoch": 0.017008117510630073,
+      "grad_norm": 0.3330378830432892,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.1028,
+      "step": 11
+    },
+    {
+      "epoch": 0.018554310011596443,
+      "grad_norm": 0.332900732755661,
+      "learning_rate": 6e-06,
+      "loss": 1.1557,
+      "step": 12
+    },
+    {
+      "epoch": 0.020100502512562814,
+      "grad_norm": 0.36040395498275757,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.1344,
+      "step": 13
+    },
+    {
+      "epoch": 0.021646695013529185,
+      "grad_norm": 0.37165355682373047,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 1.1522,
+      "step": 14
+    },
+    {
+      "epoch": 0.023192887514495556,
+      "grad_norm": 0.35057052969932556,
+      "learning_rate": 7.5e-06,
+      "loss": 1.1397,
+      "step": 15
+    },
+    {
+      "epoch": 0.024739080015461924,
+      "grad_norm": 0.3664647042751312,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.1833,
+      "step": 16
+    },
+    {
+      "epoch": 0.026285272516428294,
+      "grad_norm": 0.38355353474617004,
+      "learning_rate": 8.500000000000002e-06,
+      "loss": 1.1925,
+      "step": 17
+    },
+    {
+      "epoch": 0.027831465017394665,
+      "grad_norm": 0.3568861186504364,
+      "learning_rate": 9e-06,
+      "loss": 1.1113,
+      "step": 18
+    },
+    {
+      "epoch": 0.029377657518361036,
+      "grad_norm": 0.3547118604183197,
+      "learning_rate": 9.5e-06,
+      "loss": 1.1403,
+      "step": 19
+    },
+    {
+      "epoch": 0.030923850019327407,
+      "grad_norm": 0.3746045231819153,
+      "learning_rate": 1e-05,
+      "loss": 1.1356,
+      "step": 20
+    },
+    {
+      "epoch": 0.03247004252029378,
+      "grad_norm": 0.38165998458862305,
+      "learning_rate": 1.05e-05,
+      "loss": 1.1239,
+      "step": 21
+    },
+    {
+      "epoch": 0.034016235021260145,
+      "grad_norm": 0.4228748679161072,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 1.1263,
+      "step": 22
+    },
+    {
+      "epoch": 0.03556242752222652,
+      "grad_norm": 0.36566364765167236,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 1.1362,
+      "step": 23
+    },
+    {
+      "epoch": 0.03710862002319289,
+      "grad_norm": 0.37338364124298096,
+      "learning_rate": 1.2e-05,
+      "loss": 1.1816,
+      "step": 24
+    },
+    {
+      "epoch": 0.038654812524159254,
+      "grad_norm": 0.3412342667579651,
+      "learning_rate": 1.25e-05,
+      "loss": 1.1341,
+      "step": 25
+    },
+    {
+      "epoch": 0.04020100502512563,
+      "grad_norm": 0.38790470361709595,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.2407,
+      "step": 26
+    },
+    {
+      "epoch": 0.041747197526091996,
+      "grad_norm": 0.38183456659317017,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 1.1206,
+      "step": 27
+    },
+    {
+      "epoch": 0.04329339002705837,
+      "grad_norm": 0.4037930965423584,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.1111,
+      "step": 28
+    },
+    {
+      "epoch": 0.04483958252802474,
+      "grad_norm": 0.3747173547744751,
+      "learning_rate": 1.45e-05,
+      "loss": 1.14,
+      "step": 29
+    },
+    {
+      "epoch": 0.04638577502899111,
+      "grad_norm": 0.37349933385849,
+      "learning_rate": 1.5e-05,
+      "loss": 1.1629,
+      "step": 30
+    },
+    {
+      "epoch": 0.04793196752995748,
+      "grad_norm": 0.3699789345264435,
+      "learning_rate": 1.55e-05,
+      "loss": 1.1417,
+      "step": 31
+    },
+    {
+      "epoch": 0.04947816003092385,
+      "grad_norm": 0.37245669960975647,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.1563,
+      "step": 32
+    },
+    {
+      "epoch": 0.05102435253189022,
+      "grad_norm": 0.3655848503112793,
+      "learning_rate": 1.65e-05,
+      "loss": 1.1556,
+      "step": 33
+    },
+    {
+      "epoch": 0.05257054503285659,
+      "grad_norm": 0.3637336194515228,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 1.1589,
+      "step": 34
+    },
+    {
+      "epoch": 0.05411673753382296,
+      "grad_norm": 0.35633373260498047,
+      "learning_rate": 1.75e-05,
+      "loss": 1.0969,
+      "step": 35
+    },
+    {
+      "epoch": 0.05566293003478933,
+      "grad_norm": 0.35284438729286194,
+      "learning_rate": 1.8e-05,
+      "loss": 1.0921,
+      "step": 36
+    },
+    {
+      "epoch": 0.057209122535755705,
+      "grad_norm": 0.3530278205871582,
+      "learning_rate": 1.85e-05,
+      "loss": 1.0982,
+      "step": 37
+    },
+    {
+      "epoch": 0.05875531503672207,
+      "grad_norm": 0.3533940315246582,
+      "learning_rate": 1.9e-05,
+      "loss": 1.1329,
+      "step": 38
+    },
+    {
+      "epoch": 0.06030150753768844,
+      "grad_norm": 0.35752803087234497,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 1.1617,
+      "step": 39
+    },
+    {
+      "epoch": 0.061847700038654814,
+      "grad_norm": 0.34077584743499756,
+      "learning_rate": 2e-05,
+      "loss": 1.0956,
+      "step": 40
+    },
+    {
+      "epoch": 0.06339389253962119,
+      "grad_norm": 0.36975380778312683,
+      "learning_rate": 2.05e-05,
+      "loss": 1.0757,
+      "step": 41
+    },
+    {
+      "epoch": 0.06494008504058756,
+      "grad_norm": 0.36623820662498474,
+      "learning_rate": 2.1e-05,
+      "loss": 1.1465,
+      "step": 42
+    },
+    {
+      "epoch": 0.06648627754155392,
+      "grad_norm": 0.3988489508628845,
+      "learning_rate": 2.15e-05,
+      "loss": 1.1701,
+      "step": 43
+    },
+    {
+      "epoch": 0.06803247004252029,
+      "grad_norm": 0.37758833169937134,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.0647,
+      "step": 44
+    },
+    {
+      "epoch": 0.06957866254348666,
+      "grad_norm": 0.4047185182571411,
+      "learning_rate": 2.25e-05,
+      "loss": 1.1057,
+      "step": 45
+    },
+    {
+      "epoch": 0.07112485504445304,
+      "grad_norm": 0.39636239409446716,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 1.1294,
+      "step": 46
+    },
+    {
+      "epoch": 0.0726710475454194,
+      "grad_norm": 0.4347302317619324,
+      "learning_rate": 2.35e-05,
+      "loss": 1.152,
+      "step": 47
+    },
+    {
+      "epoch": 0.07421724004638577,
+      "grad_norm": 0.4483806788921356,
+      "learning_rate": 2.4e-05,
+      "loss": 1.1676,
+      "step": 48
+    },
+    {
+      "epoch": 0.07576343254735214,
+      "grad_norm": 0.5394858121871948,
+      "learning_rate": 2.45e-05,
+      "loss": 1.2228,
+      "step": 49
+    },
+    {
+      "epoch": 0.07730962504831851,
+      "grad_norm": 0.7282954454421997,
+      "learning_rate": 2.5e-05,
+      "loss": 1.2397,
+      "step": 50
+    },
+    {
+      "epoch": 0.07885581754928489,
+      "grad_norm": 0.3573731482028961,
+      "learning_rate": 2.5500000000000003e-05,
+      "loss": 0.9119,
+      "step": 51
+    },
+    {
+      "epoch": 0.08040201005025126,
+      "grad_norm": 0.36452800035476685,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.983,
+      "step": 52
+    },
+    {
+      "epoch": 0.08194820255121762,
+      "grad_norm": 0.3743703067302704,
+      "learning_rate": 2.6500000000000004e-05,
+      "loss": 0.9428,
+      "step": 53
+    },
+    {
+      "epoch": 0.08349439505218399,
+      "grad_norm": 0.3590448498725891,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.9967,
+      "step": 54
+    },
+    {
+      "epoch": 0.08504058755315037,
+      "grad_norm": 0.35900965332984924,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 0.9687,
+      "step": 55
+    },
+    {
+      "epoch": 0.08658678005411674,
+      "grad_norm": 0.3654124140739441,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.9342,
+      "step": 56
+    },
+    {
+      "epoch": 0.08813297255508311,
+      "grad_norm": 0.3421430289745331,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 0.9207,
+      "step": 57
+    },
+    {
+      "epoch": 0.08967916505604948,
+      "grad_norm": 0.3365703523159027,
+      "learning_rate": 2.9e-05,
+      "loss": 0.9487,
+      "step": 58
+    },
+    {
+      "epoch": 0.09122535755701584,
+      "grad_norm": 0.32828226685523987,
+      "learning_rate": 2.95e-05,
+      "loss": 0.9236,
+      "step": 59
+    },
+    {
+      "epoch": 0.09277155005798222,
+      "grad_norm": 0.31865429878234863,
+      "learning_rate": 3e-05,
+      "loss": 0.9071,
+      "step": 60
+    },
+    {
+      "epoch": 0.09431774255894859,
+      "grad_norm": 0.3250292241573334,
+      "learning_rate": 3.05e-05,
+      "loss": 0.9609,
+      "step": 61
+    },
+    {
+      "epoch": 0.09586393505991496,
+      "grad_norm": 0.3177807033061981,
+      "learning_rate": 3.1e-05,
+      "loss": 0.9727,
+      "step": 62
+    },
+    {
+      "epoch": 0.09741012756088133,
+      "grad_norm": 0.3273405134677887,
+      "learning_rate": 3.15e-05,
+      "loss": 1.0691,
+      "step": 63
+    },
+    {
+      "epoch": 0.0989563200618477,
+      "grad_norm": 0.30870744585990906,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.9726,
+      "step": 64
+    },
+    {
+      "epoch": 0.10050251256281408,
+      "grad_norm": 0.3052951395511627,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 0.8804,
+      "step": 65
+    },
+    {
+      "epoch": 0.10204870506378044,
+      "grad_norm": 0.2952975928783417,
+      "learning_rate": 3.3e-05,
+      "loss": 0.9563,
+      "step": 66
+    },
+    {
+      "epoch": 0.10359489756474681,
+      "grad_norm": 0.29250308871269226,
+      "learning_rate": 3.35e-05,
+      "loss": 0.9601,
+      "step": 67
+    },
+    {
+      "epoch": 0.10514109006571318,
+      "grad_norm": 0.2929418683052063,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.8986,
+      "step": 68
+    },
+    {
+      "epoch": 0.10668728256667954,
+      "grad_norm": 0.28594040870666504,
+      "learning_rate": 3.45e-05,
+      "loss": 0.8887,
+      "step": 69
+    },
+    {
+      "epoch": 0.10823347506764593,
+      "grad_norm": 0.2964857816696167,
+      "learning_rate": 3.5e-05,
+      "loss": 0.9407,
+      "step": 70
+    },
+    {
+      "epoch": 0.1097796675686123,
+      "grad_norm": 0.31084346771240234,
+      "learning_rate": 3.55e-05,
+      "loss": 0.9907,
+      "step": 71
+    },
+    {
+      "epoch": 0.11132586006957866,
+      "grad_norm": 0.31780946254730225,
+      "learning_rate": 3.6e-05,
+      "loss": 0.9753,
+      "step": 72
+    },
+    {
+      "epoch": 0.11287205257054503,
+      "grad_norm": 0.31742170453071594,
+      "learning_rate": 3.65e-05,
+      "loss": 0.9411,
+      "step": 73
+    },
+    {
+      "epoch": 0.11441824507151141,
+      "grad_norm": 0.3252294361591339,
+      "learning_rate": 3.7e-05,
+      "loss": 0.9973,
+      "step": 74
+    },
+    {
+      "epoch": 0.11596443757247778,
+      "grad_norm": 0.3107585608959198,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.9186,
+      "step": 75
+    },
+    {
+      "epoch": 0.11751063007344414,
+      "grad_norm": 0.32054993510246277,
+      "learning_rate": 3.8e-05,
+      "loss": 0.9264,
+      "step": 76
+    },
+    {
+      "epoch": 0.11905682257441051,
+      "grad_norm": 0.31898629665374756,
+      "learning_rate": 3.85e-05,
+      "loss": 0.9958,
+      "step": 77
+    },
+    {
+      "epoch": 0.12060301507537688,
+      "grad_norm": 0.3400716185569763,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.9793,
+      "step": 78
+    },
+    {
+      "epoch": 0.12214920757634326,
+      "grad_norm": 0.347918838262558,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 0.914,
+      "step": 79
+    },
+    {
+      "epoch": 0.12369540007730963,
+      "grad_norm": 0.33906570076942444,
+      "learning_rate": 4e-05,
+      "loss": 0.93,
+      "step": 80
+    },
+    {
+      "epoch": 0.125241592578276,
+      "grad_norm": 0.31842949986457825,
+      "learning_rate": 4.05e-05,
+      "loss": 0.8957,
+      "step": 81
+    },
+    {
+      "epoch": 0.12678778507924238,
+      "grad_norm": 0.3301747739315033,
+      "learning_rate": 4.1e-05,
+      "loss": 0.9606,
+      "step": 82
+    },
+    {
+      "epoch": 0.12833397758020873,
+      "grad_norm": 0.33865636587142944,
+      "learning_rate": 4.15e-05,
+      "loss": 0.9846,
+      "step": 83
+    },
+    {
+      "epoch": 0.1298801700811751,
+      "grad_norm": 0.31991609930992126,
+      "learning_rate": 4.2e-05,
+      "loss": 0.9574,
+      "step": 84
+    },
+    {
+      "epoch": 0.13142636258214146,
+      "grad_norm": 0.3442661762237549,
+      "learning_rate": 4.25e-05,
+      "loss": 0.9398,
+      "step": 85
+    },
+    {
+      "epoch": 0.13297255508310785,
+      "grad_norm": 0.33919069170951843,
+      "learning_rate": 4.3e-05,
+      "loss": 0.9736,
+      "step": 86
+    },
+    {
+      "epoch": 0.13451874758407423,
+      "grad_norm": 0.3447560966014862,
+      "learning_rate": 4.35e-05,
+      "loss": 0.9589,
+      "step": 87
+    },
+    {
+      "epoch": 0.13606494008504058,
+      "grad_norm": 0.3504810333251953,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.0156,
+      "step": 88
+    },
+    {
+      "epoch": 0.13761113258600696,
+      "grad_norm": 0.35089215636253357,
+      "learning_rate": 4.4500000000000004e-05,
+      "loss": 0.9732,
+      "step": 89
+    },
+    {
+      "epoch": 0.13915732508697332,
+      "grad_norm": 0.346599280834198,
+      "learning_rate": 4.5e-05,
+      "loss": 0.9705,
+      "step": 90
+    },
+    {
+      "epoch": 0.1407035175879397,
+      "grad_norm": 0.3983485698699951,
+      "learning_rate": 4.55e-05,
+      "loss": 1.0236,
+      "step": 91
+    },
+    {
+      "epoch": 0.14224971008890608,
+      "grad_norm": 0.38158902525901794,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.0174,
+      "step": 92
+    },
+    {
+      "epoch": 0.14379590258987243,
+      "grad_norm": 0.3771804869174957,
+      "learning_rate": 4.6500000000000005e-05,
+      "loss": 1.012,
+      "step": 93
+    },
+    {
+      "epoch": 0.1453420950908388,
+      "grad_norm": 0.38228315114974976,
+      "learning_rate": 4.7e-05,
+      "loss": 0.9786,
+      "step": 94
+    },
+    {
+      "epoch": 0.14688828759180517,
+      "grad_norm": 0.3989880681037903,
+      "learning_rate": 4.75e-05,
+      "loss": 0.98,
+      "step": 95
+    },
+    {
+      "epoch": 0.14843448009277155,
+      "grad_norm": 0.41765791177749634,
+      "learning_rate": 4.8e-05,
+      "loss": 1.0502,
+      "step": 96
+    },
+    {
+      "epoch": 0.14998067259373793,
+      "grad_norm": 0.4242067039012909,
+      "learning_rate": 4.85e-05,
+      "loss": 1.0911,
+      "step": 97
+    },
+    {
+      "epoch": 0.15152686509470428,
+      "grad_norm": 0.4490616023540497,
+      "learning_rate": 4.9e-05,
+      "loss": 1.0791,
+      "step": 98
+    },
+    {
+      "epoch": 0.15307305759567066,
+      "grad_norm": 0.4694664478302002,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 1.0374,
+      "step": 99
+    },
+    {
+      "epoch": 0.15461925009663702,
+      "grad_norm": 0.7335464954376221,
+      "learning_rate": 5e-05,
+      "loss": 1.093,
+      "step": 100
+    },
+    {
+      "epoch": 0.1561654425976034,
+      "grad_norm": 0.3335883915424347,
+      "learning_rate": 4.995805369127517e-05,
+      "loss": 0.8461,
+      "step": 101
+    },
+    {
+      "epoch": 0.15771163509856978,
+      "grad_norm": 0.3368653357028961,
+      "learning_rate": 4.9916107382550336e-05,
+      "loss": 0.8619,
+      "step": 102
+    },
+    {
+      "epoch": 0.15925782759953613,
+      "grad_norm": 0.34530943632125854,
+      "learning_rate": 4.9874161073825505e-05,
+      "loss": 0.8399,
+      "step": 103
+    },
+    {
+      "epoch": 0.16080402010050251,
+      "grad_norm": 0.3303908407688141,
+      "learning_rate": 4.983221476510067e-05,
+      "loss": 0.891,
+      "step": 104
+    },
+    {
+      "epoch": 0.1623502126014689,
+      "grad_norm": 0.3104991614818573,
+      "learning_rate": 4.9790268456375845e-05,
+      "loss": 0.8851,
+      "step": 105
+    },
+    {
+      "epoch": 0.16389640510243525,
+      "grad_norm": 0.3138999342918396,
+      "learning_rate": 4.974832214765101e-05,
+      "loss": 0.8668,
+      "step": 106
+    },
+    {
+      "epoch": 0.16544259760340163,
+      "grad_norm": 0.3183957040309906,
+      "learning_rate": 4.970637583892618e-05,
+      "loss": 0.8494,
+      "step": 107
+    },
+    {
+      "epoch": 0.16698879010436798,
+      "grad_norm": 0.337296724319458,
+      "learning_rate": 4.966442953020135e-05,
+      "loss": 0.872,
+      "step": 108
+    },
+    {
+      "epoch": 0.16853498260533437,
+      "grad_norm": 0.29680609703063965,
+      "learning_rate": 4.962248322147651e-05,
+      "loss": 0.8473,
+      "step": 109
+    },
+    {
+      "epoch": 0.17008117510630075,
+      "grad_norm": 0.31526410579681396,
+      "learning_rate": 4.958053691275168e-05,
+      "loss": 0.9042,
+      "step": 110
+    },
+    {
+      "epoch": 0.1716273676072671,
+      "grad_norm": 0.3531287610530853,
+      "learning_rate": 4.9538590604026845e-05,
+      "loss": 0.9444,
+      "step": 111
+    },
+    {
+      "epoch": 0.17317356010823348,
+      "grad_norm": 0.324305921792984,
+      "learning_rate": 4.9496644295302015e-05,
+      "loss": 0.8867,
+      "step": 112
+    },
+    {
+      "epoch": 0.17471975260919984,
+      "grad_norm": 0.32134464383125305,
+      "learning_rate": 4.945469798657718e-05,
+      "loss": 0.8575,
+      "step": 113
+    },
+    {
+      "epoch": 0.17626594511016622,
+      "grad_norm": 0.32409024238586426,
+      "learning_rate": 4.9412751677852355e-05,
+      "loss": 0.8937,
+      "step": 114
+    },
+    {
+      "epoch": 0.1778121376111326,
+      "grad_norm": 0.3491647243499756,
+      "learning_rate": 4.937080536912752e-05,
+      "loss": 0.8902,
+      "step": 115
+    },
+    {
+      "epoch": 0.17935833011209895,
+      "grad_norm": 0.3334656357765198,
+      "learning_rate": 4.932885906040269e-05,
+      "loss": 0.8788,
+      "step": 116
+    },
+    {
+      "epoch": 0.18090452261306533,
+      "grad_norm": 0.3333323299884796,
+      "learning_rate": 4.928691275167786e-05,
+      "loss": 0.907,
+      "step": 117
+    },
+    {
+      "epoch": 0.1824507151140317,
+      "grad_norm": 0.3563230335712433,
+      "learning_rate": 4.924496644295302e-05,
+      "loss": 0.9043,
+      "step": 118
+    },
+    {
+      "epoch": 0.18399690761499807,
+      "grad_norm": 0.35443171858787537,
+      "learning_rate": 4.920302013422819e-05,
+      "loss": 0.8966,
+      "step": 119
+    },
+    {
+      "epoch": 0.18554310011596445,
+      "grad_norm": 0.36155420541763306,
+      "learning_rate": 4.9161073825503354e-05,
+      "loss": 0.903,
+      "step": 120
+    },
+    {
+      "epoch": 0.1870892926169308,
+      "grad_norm": 0.3701721131801605,
+      "learning_rate": 4.9119127516778524e-05,
+      "loss": 0.8786,
+      "step": 121
+    },
+    {
+      "epoch": 0.18863548511789718,
+      "grad_norm": 0.3697488307952881,
+      "learning_rate": 4.9077181208053694e-05,
+      "loss": 0.9279,
+      "step": 122
+    },
+    {
+      "epoch": 0.19018167761886354,
+      "grad_norm": 0.35044190287590027,
+      "learning_rate": 4.9035234899328864e-05,
+      "loss": 0.8577,
+      "step": 123
+    },
+    {
+      "epoch": 0.19172787011982992,
+      "grad_norm": 0.3510192632675171,
+      "learning_rate": 4.8993288590604034e-05,
+      "loss": 0.8831,
+      "step": 124
+    },
+    {
+      "epoch": 0.1932740626207963,
+      "grad_norm": 0.3509821593761444,
+      "learning_rate": 4.89513422818792e-05,
+      "loss": 0.8507,
+      "step": 125
+    },
+    {
+      "epoch": 0.19482025512176265,
+      "grad_norm": 0.3761618137359619,
+      "learning_rate": 4.890939597315437e-05,
+      "loss": 0.9151,
+      "step": 126
+    },
+    {
+      "epoch": 0.19636644762272903,
+      "grad_norm": 0.3758131265640259,
+      "learning_rate": 4.886744966442953e-05,
+      "loss": 0.9067,
+      "step": 127
+    },
+    {
+      "epoch": 0.1979126401236954,
+      "grad_norm": 0.39383259415626526,
+      "learning_rate": 4.88255033557047e-05,
+      "loss": 0.9002,
+      "step": 128
+    },
+    {
+      "epoch": 0.19945883262466177,
+      "grad_norm": 0.37084588408470154,
+      "learning_rate": 4.878355704697986e-05,
+      "loss": 0.8459,
+      "step": 129
+    },
+    {
+      "epoch": 0.20100502512562815,
+      "grad_norm": 0.389466792345047,
+      "learning_rate": 4.874161073825503e-05,
+      "loss": 0.9218,
+      "step": 130
+    },
+    {
+      "epoch": 0.2025512176265945,
+      "grad_norm": 0.38749396800994873,
+      "learning_rate": 4.86996644295302e-05,
+      "loss": 0.892,
+      "step": 131
+    },
+    {
+      "epoch": 0.20409741012756089,
+      "grad_norm": 0.36285966634750366,
+      "learning_rate": 4.865771812080537e-05,
+      "loss": 0.8458,
+      "step": 132
+    },
+    {
+      "epoch": 0.20564360262852724,
+      "grad_norm": 0.38966134190559387,
+      "learning_rate": 4.861577181208054e-05,
+      "loss": 0.9189,
+      "step": 133
+    },
+    {
+      "epoch": 0.20718979512949362,
+      "grad_norm": 0.3792301118373871,
+      "learning_rate": 4.8573825503355706e-05,
+      "loss": 0.891,
+      "step": 134
+    },
+    {
+      "epoch": 0.20873598763046,
+      "grad_norm": 0.39519718289375305,
+      "learning_rate": 4.8531879194630876e-05,
+      "loss": 0.8946,
+      "step": 135
+    },
+    {
+      "epoch": 0.21028218013142636,
+      "grad_norm": 0.399034708738327,
+      "learning_rate": 4.848993288590604e-05,
+      "loss": 0.8692,
+      "step": 136
+    },
+    {
+      "epoch": 0.21182837263239274,
+      "grad_norm": 0.451168954372406,
+      "learning_rate": 4.844798657718121e-05,
+      "loss": 0.9005,
+      "step": 137
+    },
+    {
+      "epoch": 0.2133745651333591,
+      "grad_norm": 0.41019031405448914,
+      "learning_rate": 4.840604026845638e-05,
+      "loss": 0.9766,
+      "step": 138
+    },
+    {
+      "epoch": 0.21492075763432547,
+      "grad_norm": 0.4109886586666107,
+      "learning_rate": 4.836409395973154e-05,
+      "loss": 0.9078,
+      "step": 139
+    },
+    {
+      "epoch": 0.21646695013529185,
+      "grad_norm": 0.4546094834804535,
+      "learning_rate": 4.832214765100672e-05,
+      "loss": 0.9621,
+      "step": 140
+    },
+    {
+      "epoch": 0.2180131426362582,
+      "grad_norm": 0.44215071201324463,
+      "learning_rate": 4.828020134228188e-05,
+      "loss": 0.9965,
+      "step": 141
+    },
+    {
+      "epoch": 0.2195593351372246,
+      "grad_norm": 0.4271162748336792,
+      "learning_rate": 4.823825503355705e-05,
+      "loss": 0.9277,
+      "step": 142
+    },
+    {
+      "epoch": 0.22110552763819097,
+      "grad_norm": 0.43438541889190674,
+      "learning_rate": 4.8196308724832215e-05,
+      "loss": 0.9727,
+      "step": 143
+    },
+    {
+      "epoch": 0.22265172013915732,
+      "grad_norm": 0.45643556118011475,
+      "learning_rate": 4.8154362416107385e-05,
+      "loss": 0.9284,
+      "step": 144
+    },
+    {
+      "epoch": 0.2241979126401237,
+      "grad_norm": 0.5079519748687744,
+      "learning_rate": 4.8112416107382555e-05,
+      "loss": 1.004,
+      "step": 145
+    },
+    {
+      "epoch": 0.22574410514109006,
+      "grad_norm": 0.4708511531352997,
+      "learning_rate": 4.807046979865772e-05,
+      "loss": 1.001,
+      "step": 146
+    },
+    {
+      "epoch": 0.22729029764205644,
+      "grad_norm": 0.510336697101593,
+      "learning_rate": 4.802852348993289e-05,
+      "loss": 0.9808,
+      "step": 147
+    },
+    {
+      "epoch": 0.22883649014302282,
+      "grad_norm": 0.5595572590827942,
+      "learning_rate": 4.798657718120805e-05,
+      "loss": 0.9983,
+      "step": 148
+    },
+    {
+      "epoch": 0.23038268264398917,
+      "grad_norm": 0.5763404965400696,
+      "learning_rate": 4.794463087248323e-05,
+      "loss": 0.9841,
+      "step": 149
+    },
+    {
+      "epoch": 0.23192887514495555,
+      "grad_norm": 0.9527286887168884,
+      "learning_rate": 4.790268456375839e-05,
+      "loss": 0.9827,
+      "step": 150
+    },
+    {
+      "epoch": 0.2334750676459219,
+      "grad_norm": 0.35568949580192566,
+      "learning_rate": 4.786073825503356e-05,
+      "loss": 0.8084,
+      "step": 151
+    },
+    {
+      "epoch": 0.2350212601468883,
+      "grad_norm": 0.3866717517375946,
+      "learning_rate": 4.7818791946308725e-05,
+      "loss": 0.8538,
+      "step": 152
+    },
+    {
+      "epoch": 0.23656745264785467,
+      "grad_norm": 0.3884953260421753,
+      "learning_rate": 4.7776845637583895e-05,
+      "loss": 0.853,
+      "step": 153
+    },
+    {
+      "epoch": 0.23811364514882102,
+      "grad_norm": 0.40036800503730774,
+      "learning_rate": 4.7734899328859064e-05,
+      "loss": 0.8604,
+      "step": 154
+    },
+    {
+      "epoch": 0.2396598376497874,
+      "grad_norm": 0.4019627273082733,
+      "learning_rate": 4.769295302013423e-05,
+      "loss": 0.8409,
+      "step": 155
+    },
+    {
+      "epoch": 0.24120603015075376,
+      "grad_norm": 0.3811728358268738,
+      "learning_rate": 4.76510067114094e-05,
+      "loss": 0.8602,
+      "step": 156
+    },
+    {
+      "epoch": 0.24275222265172014,
+      "grad_norm": 0.39744654297828674,
+      "learning_rate": 4.760906040268457e-05,
+      "loss": 0.8011,
+      "step": 157
+    },
+    {
+      "epoch": 0.24429841515268652,
+      "grad_norm": 0.38136810064315796,
+      "learning_rate": 4.756711409395974e-05,
+      "loss": 0.9004,
+      "step": 158
+    },
+    {
+      "epoch": 0.24584460765365287,
+      "grad_norm": 0.3670859932899475,
+      "learning_rate": 4.75251677852349e-05,
+      "loss": 0.8506,
+      "step": 159
+    },
+    {
+      "epoch": 0.24739080015461926,
+      "grad_norm": 0.4202471375465393,
+      "learning_rate": 4.748322147651007e-05,
+      "loss": 0.7832,
+      "step": 160
+    },
+    {
+      "epoch": 0.2489369926555856,
+      "grad_norm": 0.40722745656967163,
+      "learning_rate": 4.744127516778524e-05,
+      "loss": 0.8813,
+      "step": 161
+    },
+    {
+      "epoch": 0.250483185156552,
+      "grad_norm": 0.4037550389766693,
+      "learning_rate": 4.7399328859060404e-05,
+      "loss": 0.8328,
+      "step": 162
+    },
+    {
+      "epoch": 0.25202937765751837,
+      "grad_norm": 0.4262521266937256,
+      "learning_rate": 4.7357382550335574e-05,
+      "loss": 0.9148,
+      "step": 163
+    },
+    {
+      "epoch": 0.25357557015848475,
+      "grad_norm": 0.40966907143592834,
+      "learning_rate": 4.731543624161074e-05,
+      "loss": 0.8689,
+      "step": 164
+    },
+    {
+      "epoch": 0.2551217626594511,
+      "grad_norm": 0.41350632905960083,
+      "learning_rate": 4.727348993288591e-05,
+      "loss": 0.8253,
+      "step": 165
+    },
+    {
+      "epoch": 0.25666795516041746,
+      "grad_norm": 0.3822355568408966,
+      "learning_rate": 4.723154362416108e-05,
+      "loss": 0.8176,
+      "step": 166
+    },
+    {
+      "epoch": 0.25821414766138384,
+      "grad_norm": 0.4168022572994232,
+      "learning_rate": 4.718959731543625e-05,
+      "loss": 0.8821,
+      "step": 167
+    },
+    {
+      "epoch": 0.2597603401623502,
+      "grad_norm": 0.4205927848815918,
+      "learning_rate": 4.714765100671141e-05,
+      "loss": 0.9162,
+      "step": 168
+    },
+    {
+      "epoch": 0.2613065326633166,
+      "grad_norm": 0.41793638467788696,
+      "learning_rate": 4.710570469798658e-05,
+      "loss": 0.9045,
+      "step": 169
+    },
+    {
+      "epoch": 0.26285272516428293,
+      "grad_norm": 0.41607388854026794,
+      "learning_rate": 4.706375838926175e-05,
+      "loss": 0.8474,
+      "step": 170
+    },
+    {
+      "epoch": 0.2643989176652493,
+      "grad_norm": 0.3936135172843933,
+      "learning_rate": 4.702181208053691e-05,
+      "loss": 0.8859,
+      "step": 171
+    },
+    {
+      "epoch": 0.2659451101662157,
+      "grad_norm": 0.39022210240364075,
+      "learning_rate": 4.697986577181208e-05,
+      "loss": 0.8455,
+      "step": 172
+    },
+    {
+      "epoch": 0.2674913026671821,
+      "grad_norm": 0.40289798378944397,
+      "learning_rate": 4.6937919463087246e-05,
+      "loss": 0.9244,
+      "step": 173
+    },
+    {
+      "epoch": 0.26903749516814845,
+      "grad_norm": 0.4135661721229553,
+      "learning_rate": 4.6895973154362416e-05,
+      "loss": 0.8427,
+      "step": 174
+    },
+    {
+      "epoch": 0.2705836876691148,
+      "grad_norm": 0.4445662796497345,
+      "learning_rate": 4.6854026845637586e-05,
+      "loss": 0.8042,
+      "step": 175
+    },
+    {
+      "epoch": 0.27212988017008116,
+      "grad_norm": 0.42175382375717163,
+      "learning_rate": 4.6812080536912756e-05,
+      "loss": 0.8447,
+      "step": 176
+    },
+    {
+      "epoch": 0.27367607267104754,
+      "grad_norm": 0.42236700654029846,
+      "learning_rate": 4.6770134228187926e-05,
+      "loss": 0.8964,
+      "step": 177
+    },
+    {
+      "epoch": 0.2752222651720139,
+      "grad_norm": 0.44100165367126465,
+      "learning_rate": 4.672818791946309e-05,
+      "loss": 0.7989,
+      "step": 178
+    },
+    {
+      "epoch": 0.2767684576729803,
+      "grad_norm": 0.43265581130981445,
+      "learning_rate": 4.668624161073826e-05,
+      "loss": 0.8795,
+      "step": 179
+    },
+    {
+      "epoch": 0.27831465017394663,
+      "grad_norm": 0.43812161684036255,
+      "learning_rate": 4.664429530201342e-05,
+      "loss": 0.8984,
+      "step": 180
+    },
+    {
+      "epoch": 0.279860842674913,
+      "grad_norm": 0.4735831916332245,
+      "learning_rate": 4.660234899328859e-05,
+      "loss": 0.8868,
+      "step": 181
+    },
+    {
+      "epoch": 0.2814070351758794,
+      "grad_norm": 0.4555661678314209,
+      "learning_rate": 4.6560402684563755e-05,
+      "loss": 0.9057,
+      "step": 182
+    },
+    {
+      "epoch": 0.2829532276768458,
+      "grad_norm": 0.44536617398262024,
+      "learning_rate": 4.6518456375838925e-05,
+      "loss": 0.8956,
+      "step": 183
+    },
+    {
+      "epoch": 0.28449942017781216,
+      "grad_norm": 0.4430801272392273,
+      "learning_rate": 4.6476510067114095e-05,
+      "loss": 0.9171,
+      "step": 184
+    },
+    {
+      "epoch": 0.2860456126787785,
+      "grad_norm": 0.4534691572189331,
+      "learning_rate": 4.6434563758389265e-05,
+      "loss": 0.7871,
+      "step": 185
+    },
+    {
+      "epoch": 0.28759180517974486,
+      "grad_norm": 0.46288105845451355,
+      "learning_rate": 4.6392617449664435e-05,
+      "loss": 0.8819,
+      "step": 186
+    },
+    {
+      "epoch": 0.28913799768071125,
+      "grad_norm": 0.4744343161582947,
+      "learning_rate": 4.63506711409396e-05,
+      "loss": 0.9415,
+      "step": 187
+    },
+    {
+      "epoch": 0.2906841901816776,
+      "grad_norm": 0.46916016936302185,
+      "learning_rate": 4.630872483221477e-05,
+      "loss": 0.9436,
+      "step": 188
+    },
+    {
+      "epoch": 0.292230382682644,
+      "grad_norm": 0.4746655821800232,
+      "learning_rate": 4.626677852348993e-05,
+      "loss": 0.9039,
+      "step": 189
+    },
+    {
+      "epoch": 0.29377657518361033,
+      "grad_norm": 0.48333853483200073,
+      "learning_rate": 4.62248322147651e-05,
+      "loss": 0.9466,
+      "step": 190
+    },
+    {
+      "epoch": 0.2953227676845767,
+      "grad_norm": 0.4735361933708191,
+      "learning_rate": 4.618288590604027e-05,
+      "loss": 0.9088,
+      "step": 191
+    },
+    {
+      "epoch": 0.2968689601855431,
+      "grad_norm": 0.484523206949234,
+      "learning_rate": 4.6140939597315434e-05,
+      "loss": 0.909,
+      "step": 192
+    },
+    {
+      "epoch": 0.2984151526865095,
+      "grad_norm": 0.5111984610557556,
+      "learning_rate": 4.609899328859061e-05,
+      "loss": 0.9451,
+      "step": 193
+    },
+    {
+      "epoch": 0.29996134518747586,
+      "grad_norm": 0.5157277584075928,
+      "learning_rate": 4.6057046979865774e-05,
+      "loss": 0.9768,
+      "step": 194
+    },
+    {
+      "epoch": 0.3015075376884422,
+      "grad_norm": 0.5363951325416565,
+      "learning_rate": 4.6015100671140944e-05,
+      "loss": 0.9411,
+      "step": 195
+    },
+    {
+      "epoch": 0.30305373018940857,
+      "grad_norm": 0.5306119918823242,
+      "learning_rate": 4.597315436241611e-05,
+      "loss": 0.9238,
+      "step": 196
+    },
+    {
+      "epoch": 0.30459992269037495,
+      "grad_norm": 0.5576231479644775,
+      "learning_rate": 4.593120805369128e-05,
+      "loss": 0.9866,
+      "step": 197
+    },
+    {
+      "epoch": 0.30614611519134133,
+      "grad_norm": 0.6054885983467102,
+      "learning_rate": 4.588926174496645e-05,
+      "loss": 0.9987,
+      "step": 198
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.6553575396537781,
+      "learning_rate": 4.584731543624161e-05,
+      "loss": 1.0413,
+      "step": 199
+    },
+    {
+      "epoch": 0.30923850019327404,
+      "grad_norm": 1.023345708847046,
+      "learning_rate": 4.580536912751678e-05,
+      "loss": 1.0355,
+      "step": 200
+    },
+    {
+      "epoch": 0.3107846926942404,
+      "grad_norm": 0.4243873059749603,
+      "learning_rate": 4.576342281879195e-05,
+      "loss": 0.7642,
+      "step": 201
+    },
+    {
+      "epoch": 0.3123308851952068,
+      "grad_norm": 0.4183518886566162,
+      "learning_rate": 4.572147651006712e-05,
+      "loss": 0.8115,
+      "step": 202
+    },
+    {
+      "epoch": 0.3138770776961732,
+      "grad_norm": 0.43979495763778687,
+      "learning_rate": 4.5679530201342284e-05,
+      "loss": 0.7705,
+      "step": 203
+    },
+    {
+      "epoch": 0.31542327019713956,
+      "grad_norm": 0.44140326976776123,
+      "learning_rate": 4.5637583892617453e-05,
+      "loss": 0.7784,
+      "step": 204
+    },
+    {
+      "epoch": 0.31696946269810594,
+      "grad_norm": 0.4253062903881073,
+      "learning_rate": 4.559563758389262e-05,
+      "loss": 0.8006,
+      "step": 205
+    },
+    {
+      "epoch": 0.31851565519907227,
+      "grad_norm": 0.41898655891418457,
+      "learning_rate": 4.5553691275167787e-05,
+      "loss": 0.8304,
+      "step": 206
+    },
+    {
+      "epoch": 0.32006184770003865,
+      "grad_norm": 0.3986593186855316,
+      "learning_rate": 4.5511744966442957e-05,
+      "loss": 0.7815,
+      "step": 207
+    },
+    {
+      "epoch": 0.32160804020100503,
+      "grad_norm": 0.42180249094963074,
+      "learning_rate": 4.546979865771812e-05,
+      "loss": 0.859,
+      "step": 208
+    },
+    {
+      "epoch": 0.3231542327019714,
+      "grad_norm": 0.429800808429718,
+      "learning_rate": 4.542785234899329e-05,
+      "loss": 0.8188,
+      "step": 209
+    },
+    {
+      "epoch": 0.3247004252029378,
+      "grad_norm": 0.4582015573978424,
+      "learning_rate": 4.538590604026846e-05,
+      "loss": 0.8069,
+      "step": 210
+    },
+    {
+      "epoch": 0.3262466177039041,
+      "grad_norm": 0.44153645634651184,
+      "learning_rate": 4.534395973154363e-05,
+      "loss": 0.8107,
+      "step": 211
+    },
+    {
+      "epoch": 0.3277928102048705,
+      "grad_norm": 0.4466538429260254,
+      "learning_rate": 4.530201342281879e-05,
+      "loss": 0.8076,
+      "step": 212
+    },
+    {
+      "epoch": 0.3293390027058369,
+      "grad_norm": 0.4263439178466797,
+      "learning_rate": 4.526006711409396e-05,
+      "loss": 0.8532,
+      "step": 213
+    },
+    {
+      "epoch": 0.33088519520680326,
+      "grad_norm": 0.469927042722702,
+      "learning_rate": 4.521812080536913e-05,
+      "loss": 0.8264,
+      "step": 214
+    },
+    {
+      "epoch": 0.33243138770776964,
+      "grad_norm": 0.45597830414772034,
+      "learning_rate": 4.5176174496644296e-05,
+      "loss": 0.8125,
+      "step": 215
+    },
+    {
+      "epoch": 0.33397758020873597,
+      "grad_norm": 0.43289533257484436,
+      "learning_rate": 4.5134228187919466e-05,
+      "loss": 0.8223,
+      "step": 216
+    },
+    {
+      "epoch": 0.33552377270970235,
+      "grad_norm": 0.4476693272590637,
+      "learning_rate": 4.509228187919463e-05,
+      "loss": 0.8206,
+      "step": 217
+    },
+    {
+      "epoch": 0.33706996521066873,
+      "grad_norm": 0.43755078315734863,
+      "learning_rate": 4.50503355704698e-05,
+      "loss": 0.8084,
+      "step": 218
+    },
+    {
+      "epoch": 0.3386161577116351,
+      "grad_norm": 0.4320957064628601,
+      "learning_rate": 4.500838926174497e-05,
+      "loss": 0.8127,
+      "step": 219
+    },
+    {
+      "epoch": 0.3401623502126015,
+      "grad_norm": 0.45105233788490295,
+      "learning_rate": 4.496644295302014e-05,
+      "loss": 0.8208,
+      "step": 220
+    },
+    {
+      "epoch": 0.3417085427135678,
+      "grad_norm": 0.42372819781303406,
+      "learning_rate": 4.49244966442953e-05,
+      "loss": 0.8685,
+      "step": 221
+    },
+    {
+      "epoch": 0.3432547352145342,
+      "grad_norm": 0.4444120526313782,
+      "learning_rate": 4.488255033557047e-05,
+      "loss": 0.8501,
+      "step": 222
+    },
+    {
+      "epoch": 0.3448009277155006,
+      "grad_norm": 0.4355757236480713,
+      "learning_rate": 4.484060402684564e-05,
+      "loss": 0.8507,
+      "step": 223
+    },
+    {
+      "epoch": 0.34634712021646696,
+      "grad_norm": 0.45914098620414734,
+      "learning_rate": 4.4798657718120805e-05,
+      "loss": 0.9056,
+      "step": 224
+    },
+    {
+      "epoch": 0.34789331271743335,
+      "grad_norm": 0.4455774128437042,
+      "learning_rate": 4.4756711409395975e-05,
+      "loss": 0.8716,
+      "step": 225
+    },
+    {
+      "epoch": 0.34943950521839967,
+      "grad_norm": 0.4548904299736023,
+      "learning_rate": 4.471476510067114e-05,
+      "loss": 0.7938,
+      "step": 226
+    },
+    {
+      "epoch": 0.35098569771936605,
+      "grad_norm": 0.44709107279777527,
+      "learning_rate": 4.467281879194631e-05,
+      "loss": 0.8248,
+      "step": 227
+    },
+    {
+      "epoch": 0.35253189022033243,
+      "grad_norm": 0.4717913866043091,
+      "learning_rate": 4.463087248322148e-05,
+      "loss": 0.864,
+      "step": 228
+    },
+    {
+      "epoch": 0.3540780827212988,
+      "grad_norm": 0.48902633786201477,
+      "learning_rate": 4.458892617449665e-05,
+      "loss": 0.7545,
+      "step": 229
+    },
+    {
+      "epoch": 0.3556242752222652,
+      "grad_norm": 0.4466339945793152,
+      "learning_rate": 4.454697986577182e-05,
+      "loss": 0.8869,
+      "step": 230
+    },
+    {
+      "epoch": 0.3571704677232315,
+      "grad_norm": 0.4503350257873535,
+      "learning_rate": 4.450503355704698e-05,
+      "loss": 0.8719,
+      "step": 231
+    },
+    {
+      "epoch": 0.3587166602241979,
+      "grad_norm": 0.4927978217601776,
+      "learning_rate": 4.446308724832215e-05,
+      "loss": 0.9214,
+      "step": 232
+    },
+    {
+      "epoch": 0.3602628527251643,
+      "grad_norm": 0.4751432240009308,
+      "learning_rate": 4.4421140939597314e-05,
+      "loss": 0.8735,
+      "step": 233
+    },
+    {
+      "epoch": 0.36180904522613067,
+      "grad_norm": 0.4933432638645172,
+      "learning_rate": 4.4379194630872484e-05,
+      "loss": 0.8353,
+      "step": 234
+    },
+    {
+      "epoch": 0.36335523772709705,
+      "grad_norm": 0.5034083127975464,
+      "learning_rate": 4.4337248322147654e-05,
+      "loss": 0.9168,
+      "step": 235
+    },
+    {
+      "epoch": 0.3649014302280634,
+      "grad_norm": 0.5039856433868408,
+      "learning_rate": 4.4295302013422824e-05,
+      "loss": 0.9183,
+      "step": 236
+    },
+    {
+      "epoch": 0.36644762272902975,
+      "grad_norm": 0.47999468445777893,
+      "learning_rate": 4.4253355704697994e-05,
+      "loss": 0.9402,
+      "step": 237
+    },
+    {
+      "epoch": 0.36799381522999614,
+      "grad_norm": 0.5011245608329773,
+      "learning_rate": 4.421140939597316e-05,
+      "loss": 0.9587,
+      "step": 238
+    },
+    {
+      "epoch": 0.3695400077309625,
+      "grad_norm": 0.49425458908081055,
+      "learning_rate": 4.416946308724833e-05,
+      "loss": 0.8474,
+      "step": 239
+    },
+    {
+      "epoch": 0.3710862002319289,
+      "grad_norm": 0.514880359172821,
+      "learning_rate": 4.412751677852349e-05,
+      "loss": 0.9016,
+      "step": 240
+    },
+    {
+      "epoch": 0.3726323927328952,
+      "grad_norm": 0.5200977325439453,
+      "learning_rate": 4.408557046979866e-05,
+      "loss": 0.9552,
+      "step": 241
+    },
+    {
+      "epoch": 0.3741785852338616,
+      "grad_norm": 0.5420643091201782,
+      "learning_rate": 4.4043624161073823e-05,
+      "loss": 0.9023,
+      "step": 242
+    },
+    {
+      "epoch": 0.375724777734828,
+      "grad_norm": 0.5034189820289612,
+      "learning_rate": 4.4001677852348993e-05,
+      "loss": 0.9255,
+      "step": 243
+    },
+    {
+      "epoch": 0.37727097023579437,
+      "grad_norm": 0.529698371887207,
+      "learning_rate": 4.395973154362416e-05,
+      "loss": 0.9337,
+      "step": 244
+    },
+    {
+      "epoch": 0.37881716273676075,
+      "grad_norm": 0.5248023867607117,
+      "learning_rate": 4.391778523489933e-05,
+      "loss": 0.9626,
+      "step": 245
+    },
+    {
+      "epoch": 0.3803633552377271,
+      "grad_norm": 0.5809698700904846,
+      "learning_rate": 4.38758389261745e-05,
+      "loss": 0.9598,
+      "step": 246
+    },
+    {
+      "epoch": 0.38190954773869346,
+      "grad_norm": 0.5663429498672485,
+      "learning_rate": 4.3833892617449666e-05,
+      "loss": 0.979,
+      "step": 247
+    },
+    {
+      "epoch": 0.38345574023965984,
+      "grad_norm": 0.6147710680961609,
+      "learning_rate": 4.3791946308724836e-05,
+      "loss": 0.9747,
+      "step": 248
+    },
+    {
+      "epoch": 0.3850019327406262,
+      "grad_norm": 0.7502771615982056,
+      "learning_rate": 4.375e-05,
+      "loss": 0.9344,
+      "step": 249
+    },
+    {
+      "epoch": 0.3865481252415926,
+      "grad_norm": 1.0911471843719482,
+      "learning_rate": 4.370805369127517e-05,
+      "loss": 1.0075,
+      "step": 250
+    },
+    {
+      "epoch": 0.3880943177425589,
+      "grad_norm": 0.4232114255428314,
+      "learning_rate": 4.366610738255034e-05,
+      "loss": 0.7643,
+      "step": 251
+    },
+    {
+      "epoch": 0.3896405102435253,
+      "grad_norm": 0.423627644777298,
+      "learning_rate": 4.36241610738255e-05,
+      "loss": 0.7551,
+      "step": 252
+    },
+    {
+      "epoch": 0.3911867027444917,
+      "grad_norm": 0.4198389947414398,
+      "learning_rate": 4.358221476510067e-05,
+      "loss": 0.7888,
+      "step": 253
+    },
+    {
+      "epoch": 0.39273289524545807,
+      "grad_norm": 0.43714022636413574,
+      "learning_rate": 4.354026845637584e-05,
+      "loss": 0.7958,
+      "step": 254
+    },
+    {
+      "epoch": 0.39427908774642445,
+      "grad_norm": 0.4278389513492584,
+      "learning_rate": 4.349832214765101e-05,
+      "loss": 0.7959,
+      "step": 255
+    },
+    {
+      "epoch": 0.3958252802473908,
+      "grad_norm": 0.45499542355537415,
+      "learning_rate": 4.3456375838926176e-05,
+      "loss": 0.7461,
+      "step": 256
+    },
+    {
+      "epoch": 0.39737147274835716,
+      "grad_norm": 0.46723300218582153,
+      "learning_rate": 4.3414429530201346e-05,
+      "loss": 0.8312,
+      "step": 257
+    },
+    {
+      "epoch": 0.39891766524932354,
+      "grad_norm": 0.44024282693862915,
+      "learning_rate": 4.337248322147651e-05,
+      "loss": 0.8517,
+      "step": 258
+    },
+    {
+      "epoch": 0.4004638577502899,
+      "grad_norm": 0.42471960186958313,
+      "learning_rate": 4.333053691275168e-05,
+      "loss": 0.8312,
+      "step": 259
+    },
+    {
+      "epoch": 0.4020100502512563,
+      "grad_norm": 0.47083520889282227,
+      "learning_rate": 4.328859060402685e-05,
+      "loss": 0.8618,
+      "step": 260
+    },
+    {
+      "epoch": 0.4035562427522226,
+      "grad_norm": 0.45681217312812805,
+      "learning_rate": 4.324664429530201e-05,
+      "loss": 0.8986,
+      "step": 261
+    },
+    {
+      "epoch": 0.405102435253189,
+      "grad_norm": 0.42533236742019653,
+      "learning_rate": 4.320469798657718e-05,
+      "loss": 0.824,
+      "step": 262
+    },
+    {
+      "epoch": 0.4066486277541554,
+      "grad_norm": 0.4372871220111847,
+      "learning_rate": 4.316275167785235e-05,
+      "loss": 0.8077,
+      "step": 263
+    },
+    {
+      "epoch": 0.40819482025512177,
+      "grad_norm": 0.45690760016441345,
+      "learning_rate": 4.312080536912752e-05,
+      "loss": 0.8735,
+      "step": 264
+    },
+    {
+      "epoch": 0.40974101275608815,
+      "grad_norm": 0.46814388036727905,
+      "learning_rate": 4.3078859060402685e-05,
+      "loss": 0.8161,
+      "step": 265
+    },
+    {
+      "epoch": 0.4112872052570545,
+      "grad_norm": 0.4442387521266937,
+      "learning_rate": 4.3036912751677855e-05,
+      "loss": 0.8352,
+      "step": 266
+    },
+    {
+      "epoch": 0.41283339775802086,
+      "grad_norm": 0.4814305305480957,
+      "learning_rate": 4.2994966442953025e-05,
+      "loss": 0.7508,
+      "step": 267
+    },
+    {
+      "epoch": 0.41437959025898724,
+      "grad_norm": 0.4563637673854828,
+      "learning_rate": 4.295302013422819e-05,
+      "loss": 0.8639,
+      "step": 268
+    },
+    {
+      "epoch": 0.4159257827599536,
+      "grad_norm": 0.4439164698123932,
+      "learning_rate": 4.291107382550336e-05,
+      "loss": 0.8466,
+      "step": 269
+    },
+    {
+      "epoch": 0.41747197526092,
+      "grad_norm": 0.45698022842407227,
+      "learning_rate": 4.286912751677852e-05,
+      "loss": 0.8784,
+      "step": 270
+    },
+    {
+      "epoch": 0.41901816776188633,
+      "grad_norm": 0.46446678042411804,
+      "learning_rate": 4.28271812080537e-05,
+      "loss": 0.7856,
+      "step": 271
+    },
+    {
+      "epoch": 0.4205643602628527,
+      "grad_norm": 0.4599681496620178,
+      "learning_rate": 4.278523489932886e-05,
+      "loss": 0.8241,
+      "step": 272
+    },
+    {
+      "epoch": 0.4221105527638191,
+      "grad_norm": 0.4912761449813843,
+      "learning_rate": 4.274328859060403e-05,
+      "loss": 0.8697,
+      "step": 273
+    },
+    {
+      "epoch": 0.4236567452647855,
+      "grad_norm": 0.4875286817550659,
+      "learning_rate": 4.27013422818792e-05,
+      "loss": 0.7999,
+      "step": 274
+    },
+    {
+      "epoch": 0.42520293776575185,
+      "grad_norm": 0.4821873903274536,
+      "learning_rate": 4.2659395973154364e-05,
+      "loss": 0.7944,
+      "step": 275
+    },
+    {
+      "epoch": 0.4267491302667182,
+      "grad_norm": 0.4670597016811371,
+      "learning_rate": 4.2617449664429534e-05,
+      "loss": 0.9423,
+      "step": 276
+    },
+    {
+      "epoch": 0.42829532276768456,
+      "grad_norm": 0.5004227757453918,
+      "learning_rate": 4.25755033557047e-05,
+      "loss": 0.8073,
+      "step": 277
+    },
+    {
+      "epoch": 0.42984151526865094,
+      "grad_norm": 0.510180652141571,
+      "learning_rate": 4.253355704697987e-05,
+      "loss": 0.8658,
+      "step": 278
+    },
+    {
+      "epoch": 0.4313877077696173,
+      "grad_norm": 0.4866536855697632,
+      "learning_rate": 4.249161073825503e-05,
+      "loss": 0.7532,
+      "step": 279
+    },
+    {
+      "epoch": 0.4329339002705837,
+      "grad_norm": 0.5010313987731934,
+      "learning_rate": 4.244966442953021e-05,
+      "loss": 0.9368,
+      "step": 280
+    },
+    {
+      "epoch": 0.43448009277155003,
+      "grad_norm": 0.5455735325813293,
+      "learning_rate": 4.240771812080537e-05,
+      "loss": 0.9058,
+      "step": 281
+    },
+    {
+      "epoch": 0.4360262852725164,
+      "grad_norm": 0.47950518131256104,
+      "learning_rate": 4.236577181208054e-05,
+      "loss": 0.8709,
+      "step": 282
+    },
+    {
+      "epoch": 0.4375724777734828,
+      "grad_norm": 0.5143994092941284,
+      "learning_rate": 4.232382550335571e-05,
+      "loss": 0.8528,
+      "step": 283
+    },
+    {
+      "epoch": 0.4391186702744492,
+      "grad_norm": 0.5421144366264343,
+      "learning_rate": 4.228187919463087e-05,
+      "loss": 0.8989,
+      "step": 284
+    },
+    {
+      "epoch": 0.44066486277541556,
+      "grad_norm": 0.5018422603607178,
+      "learning_rate": 4.223993288590604e-05,
+      "loss": 0.8903,
+      "step": 285
+    },
+    {
+      "epoch": 0.44221105527638194,
+      "grad_norm": 0.4975266754627228,
+      "learning_rate": 4.2197986577181206e-05,
+      "loss": 0.9568,
+      "step": 286
+    },
+    {
+      "epoch": 0.44375724777734826,
+      "grad_norm": 0.4982987642288208,
+      "learning_rate": 4.2156040268456376e-05,
+      "loss": 0.946,
+      "step": 287
+    },
+    {
+      "epoch": 0.44530344027831464,
+      "grad_norm": 0.5241461992263794,
+      "learning_rate": 4.2114093959731546e-05,
+      "loss": 0.9359,
+      "step": 288
+    },
+    {
+      "epoch": 0.446849632779281,
+      "grad_norm": 0.4898720681667328,
+      "learning_rate": 4.2072147651006716e-05,
+      "loss": 0.9798,
+      "step": 289
+    },
+    {
+      "epoch": 0.4483958252802474,
+      "grad_norm": 0.5274215936660767,
+      "learning_rate": 4.2030201342281886e-05,
+      "loss": 0.8603,
+      "step": 290
+    },
+    {
+      "epoch": 0.4499420177812138,
+      "grad_norm": 0.5538789629936218,
+      "learning_rate": 4.198825503355705e-05,
+      "loss": 0.8873,
+      "step": 291
+    },
+    {
+      "epoch": 0.4514882102821801,
+      "grad_norm": 0.5374253392219543,
+      "learning_rate": 4.194630872483222e-05,
+      "loss": 1.0174,
+      "step": 292
+    },
+    {
+      "epoch": 0.4530344027831465,
+      "grad_norm": 0.5286028385162354,
+      "learning_rate": 4.190436241610738e-05,
+      "loss": 0.9891,
+      "step": 293
+    },
+    {
+      "epoch": 0.4545805952841129,
+      "grad_norm": 0.5447548627853394,
+      "learning_rate": 4.186241610738255e-05,
+      "loss": 0.9587,
+      "step": 294
+    },
+    {
+      "epoch": 0.45612678778507926,
+      "grad_norm": 0.5515534281730652,
+      "learning_rate": 4.1820469798657716e-05,
+      "loss": 0.9663,
+      "step": 295
+    },
+    {
+      "epoch": 0.45767298028604564,
+      "grad_norm": 0.6086519956588745,
+      "learning_rate": 4.1778523489932886e-05,
+      "loss": 1.0357,
+      "step": 296
+    },
+    {
+      "epoch": 0.45921917278701196,
+      "grad_norm": 0.6332388520240784,
+      "learning_rate": 4.1736577181208055e-05,
+      "loss": 0.9999,
+      "step": 297
+    },
+    {
+      "epoch": 0.46076536528797835,
+      "grad_norm": 0.6928420066833496,
+      "learning_rate": 4.1694630872483225e-05,
+      "loss": 1.0243,
+      "step": 298
+    },
+    {
+      "epoch": 0.4623115577889447,
+      "grad_norm": 0.7226160168647766,
+      "learning_rate": 4.1652684563758395e-05,
+      "loss": 0.9209,
+      "step": 299
+    },
+    {
+      "epoch": 0.4638577502899111,
+      "grad_norm": 1.0083510875701904,
+      "learning_rate": 4.161073825503356e-05,
+      "loss": 0.9493,
+      "step": 300
+    },
+    {
+      "epoch": 0.4654039427908775,
+      "grad_norm": 0.4649132192134857,
+      "learning_rate": 4.156879194630873e-05,
+      "loss": 0.7743,
+      "step": 301
+    },
+    {
+      "epoch": 0.4669501352918438,
+      "grad_norm": 0.43430233001708984,
+      "learning_rate": 4.152684563758389e-05,
+      "loss": 0.8197,
+      "step": 302
+    },
+    {
+      "epoch": 0.4684963277928102,
+      "grad_norm": 0.44487112760543823,
+      "learning_rate": 4.148489932885906e-05,
+      "loss": 0.7632,
+      "step": 303
+    },
+    {
+      "epoch": 0.4700425202937766,
+      "grad_norm": 0.4654642343521118,
+      "learning_rate": 4.144295302013423e-05,
+      "loss": 0.7637,
+      "step": 304
+    },
+    {
+      "epoch": 0.47158871279474296,
+      "grad_norm": 0.48062166571617126,
+      "learning_rate": 4.1401006711409395e-05,
+      "loss": 0.7771,
+      "step": 305
+    },
+    {
+      "epoch": 0.47313490529570934,
+      "grad_norm": 0.4307418167591095,
+      "learning_rate": 4.135906040268457e-05,
+      "loss": 0.8336,
+      "step": 306
+    },
+    {
+      "epoch": 0.47468109779667567,
+      "grad_norm": 0.4675520658493042,
+      "learning_rate": 4.1317114093959735e-05,
+      "loss": 0.7969,
+      "step": 307
+    },
+    {
+      "epoch": 0.47622729029764205,
+      "grad_norm": 0.48613134026527405,
+      "learning_rate": 4.1275167785234905e-05,
+      "loss": 0.8821,
+      "step": 308
+    },
+    {
+      "epoch": 0.47777348279860843,
+      "grad_norm": 0.46094274520874023,
+      "learning_rate": 4.123322147651007e-05,
+      "loss": 0.8103,
+      "step": 309
+    },
+    {
+      "epoch": 0.4793196752995748,
+      "grad_norm": 0.47018101811408997,
+      "learning_rate": 4.119127516778524e-05,
+      "loss": 0.8515,
+      "step": 310
+    },
+    {
+      "epoch": 0.4808658678005412,
+      "grad_norm": 0.43754515051841736,
+      "learning_rate": 4.11493288590604e-05,
+      "loss": 0.8218,
+      "step": 311
+    },
+    {
+      "epoch": 0.4824120603015075,
+      "grad_norm": 0.49217140674591064,
+      "learning_rate": 4.110738255033557e-05,
+      "loss": 0.7866,
+      "step": 312
+    },
+    {
+      "epoch": 0.4839582528024739,
+      "grad_norm": 0.47612541913986206,
+      "learning_rate": 4.106543624161074e-05,
+      "loss": 0.8088,
+      "step": 313
+    },
+    {
+      "epoch": 0.4855044453034403,
+      "grad_norm": 0.475328654050827,
+      "learning_rate": 4.1023489932885904e-05,
+      "loss": 0.8165,
+      "step": 314
+    },
+    {
+      "epoch": 0.48705063780440666,
+      "grad_norm": 0.4439482092857361,
+      "learning_rate": 4.098154362416108e-05,
+      "loss": 0.7833,
+      "step": 315
+    },
+    {
+      "epoch": 0.48859683030537304,
+      "grad_norm": 0.5019133687019348,
+      "learning_rate": 4.0939597315436244e-05,
+      "loss": 0.7942,
+      "step": 316
+    },
+    {
+      "epoch": 0.49014302280633937,
+      "grad_norm": 0.451492577791214,
+      "learning_rate": 4.0897651006711414e-05,
+      "loss": 0.8039,
+      "step": 317
+    },
+    {
+      "epoch": 0.49168921530730575,
+      "grad_norm": 0.46894291043281555,
+      "learning_rate": 4.085570469798658e-05,
+      "loss": 0.843,
+      "step": 318
+    },
+    {
+      "epoch": 0.49323540780827213,
+      "grad_norm": 0.49084803462028503,
+      "learning_rate": 4.081375838926175e-05,
+      "loss": 0.8121,
+      "step": 319
+    },
+    {
+      "epoch": 0.4947816003092385,
+      "grad_norm": 0.46862879395484924,
+      "learning_rate": 4.077181208053692e-05,
+      "loss": 0.901,
+      "step": 320
+    },
+    {
+      "epoch": 0.4963277928102049,
+      "grad_norm": 0.4881606698036194,
+      "learning_rate": 4.072986577181208e-05,
+      "loss": 0.9001,
+      "step": 321
+    },
+    {
+      "epoch": 0.4978739853111712,
+      "grad_norm": 0.5040379166603088,
+      "learning_rate": 4.068791946308725e-05,
+      "loss": 0.801,
+      "step": 322
+    },
+    {
+      "epoch": 0.4994201778121376,
+      "grad_norm": 0.46498745679855347,
+      "learning_rate": 4.064597315436241e-05,
+      "loss": 0.8473,
+      "step": 323
+    },
+    {
+      "epoch": 0.500966370313104,
+      "grad_norm": 0.4876708388328552,
+      "learning_rate": 4.060402684563759e-05,
+      "loss": 0.8418,
+      "step": 324
+    },
+    {
+      "epoch": 0.5025125628140703,
+      "grad_norm": 0.4962271451950073,
+      "learning_rate": 4.056208053691275e-05,
+      "loss": 0.7611,
+      "step": 325
+    },
+    {
+      "epoch": 0.5040587553150367,
+      "grad_norm": 0.48774364590644836,
+      "learning_rate": 4.052013422818792e-05,
+      "loss": 0.8825,
+      "step": 326
+    },
+    {
+      "epoch": 0.5056049478160031,
+      "grad_norm": 0.5011964440345764,
+      "learning_rate": 4.047818791946309e-05,
+      "loss": 0.7659,
+      "step": 327
+    },
+    {
+      "epoch": 0.5071511403169695,
+      "grad_norm": 0.5307598114013672,
+      "learning_rate": 4.0436241610738256e-05,
+      "loss": 0.8629,
+      "step": 328
+    },
+    {
+      "epoch": 0.5086973328179358,
+      "grad_norm": 0.49158573150634766,
+      "learning_rate": 4.0394295302013426e-05,
+      "loss": 0.9071,
+      "step": 329
+    },
+    {
+      "epoch": 0.5102435253189022,
+      "grad_norm": 0.5126063227653503,
+      "learning_rate": 4.035234899328859e-05,
+      "loss": 0.7998,
+      "step": 330
+    },
+    {
+      "epoch": 0.5117897178198686,
+      "grad_norm": 0.5247028470039368,
+      "learning_rate": 4.031040268456376e-05,
+      "loss": 0.8443,
+      "step": 331
+    },
+    {
+      "epoch": 0.5133359103208349,
+      "grad_norm": 0.5511295199394226,
+      "learning_rate": 4.026845637583892e-05,
+      "loss": 0.8946,
+      "step": 332
+    },
+    {
+      "epoch": 0.5148821028218014,
+      "grad_norm": 0.5266992449760437,
+      "learning_rate": 4.02265100671141e-05,
+      "loss": 0.8172,
+      "step": 333
+    },
+    {
+      "epoch": 0.5164282953227677,
+      "grad_norm": 0.5326645374298096,
+      "learning_rate": 4.018456375838926e-05,
+      "loss": 0.8407,
+      "step": 334
+    },
+    {
+      "epoch": 0.517974487823734,
+      "grad_norm": 0.5065395832061768,
+      "learning_rate": 4.014261744966443e-05,
+      "loss": 0.8751,
+      "step": 335
+    },
+    {
+      "epoch": 0.5195206803247004,
+      "grad_norm": 0.53061443567276,
+      "learning_rate": 4.01006711409396e-05,
+      "loss": 0.9204,
+      "step": 336
+    },
+    {
+      "epoch": 0.5210668728256668,
+      "grad_norm": 0.5223432779312134,
+      "learning_rate": 4.0058724832214765e-05,
+      "loss": 0.8933,
+      "step": 337
+    },
+    {
+      "epoch": 0.5226130653266332,
+      "grad_norm": 0.5290891528129578,
+      "learning_rate": 4.0016778523489935e-05,
+      "loss": 0.977,
+      "step": 338
+    },
+    {
+      "epoch": 0.5241592578275995,
+      "grad_norm": 0.5419663786888123,
+      "learning_rate": 3.99748322147651e-05,
+      "loss": 0.9521,
+      "step": 339
+    },
+    {
+      "epoch": 0.5257054503285659,
+      "grad_norm": 0.5404860973358154,
+      "learning_rate": 3.993288590604027e-05,
+      "loss": 0.8907,
+      "step": 340
+    },
+    {
+      "epoch": 0.5272516428295323,
+      "grad_norm": 0.5783061385154724,
+      "learning_rate": 3.989093959731544e-05,
+      "loss": 0.9162,
+      "step": 341
+    },
+    {
+      "epoch": 0.5287978353304986,
+      "grad_norm": 0.5513641834259033,
+      "learning_rate": 3.984899328859061e-05,
+      "loss": 0.8912,
+      "step": 342
+    },
+    {
+      "epoch": 0.5303440278314651,
+      "grad_norm": 0.5670628547668457,
+      "learning_rate": 3.980704697986578e-05,
+      "loss": 0.9502,
+      "step": 343
+    },
+    {
+      "epoch": 0.5318902203324314,
+      "grad_norm": 0.5960517525672913,
+      "learning_rate": 3.976510067114094e-05,
+      "loss": 0.9976,
+      "step": 344
+    },
+    {
+      "epoch": 0.5334364128333977,
+      "grad_norm": 0.6074213981628418,
+      "learning_rate": 3.972315436241611e-05,
+      "loss": 0.9861,
+      "step": 345
+    },
+    {
+      "epoch": 0.5349826053343641,
+      "grad_norm": 0.6100422143936157,
+      "learning_rate": 3.9681208053691275e-05,
+      "loss": 0.9483,
+      "step": 346
+    },
+    {
+      "epoch": 0.5365287978353305,
+      "grad_norm": 0.6308846473693848,
+      "learning_rate": 3.9639261744966445e-05,
+      "loss": 0.9455,
+      "step": 347
+    },
+    {
+      "epoch": 0.5380749903362969,
+      "grad_norm": 0.6239264607429504,
+      "learning_rate": 3.959731543624161e-05,
+      "loss": 0.9551,
+      "step": 348
+    },
+    {
+      "epoch": 0.5396211828372632,
+      "grad_norm": 0.6669812798500061,
+      "learning_rate": 3.955536912751678e-05,
+      "loss": 0.9251,
+      "step": 349
+    },
+    {
+      "epoch": 0.5411673753382296,
+      "grad_norm": 0.8175333738327026,
+      "learning_rate": 3.951342281879195e-05,
+      "loss": 0.9795,
+      "step": 350
+    },
+    {
+      "epoch": 0.542713567839196,
+      "grad_norm": 0.44942405819892883,
+      "learning_rate": 3.947147651006712e-05,
+      "loss": 0.7184,
+      "step": 351
+    },
+    {
+      "epoch": 0.5442597603401623,
+      "grad_norm": 0.46283286809921265,
+      "learning_rate": 3.942953020134229e-05,
+      "loss": 0.7251,
+      "step": 352
+    },
+    {
+      "epoch": 0.5458059528411288,
+      "grad_norm": 0.47923749685287476,
+      "learning_rate": 3.938758389261745e-05,
+      "loss": 0.766,
+      "step": 353
+    },
+    {
+      "epoch": 0.5473521453420951,
+      "grad_norm": 0.4575538635253906,
+      "learning_rate": 3.934563758389262e-05,
+      "loss": 0.7539,
+      "step": 354
+    },
+    {
+      "epoch": 0.5488983378430614,
+      "grad_norm": 0.48839524388313293,
+      "learning_rate": 3.9303691275167784e-05,
+      "loss": 0.7633,
+      "step": 355
+    },
+    {
+      "epoch": 0.5504445303440278,
+      "grad_norm": 0.45980286598205566,
+      "learning_rate": 3.9261744966442954e-05,
+      "loss": 0.774,
+      "step": 356
+    },
+    {
+      "epoch": 0.5519907228449942,
+      "grad_norm": 0.4767687916755676,
+      "learning_rate": 3.9219798657718124e-05,
+      "loss": 0.7588,
+      "step": 357
+    },
+    {
+      "epoch": 0.5535369153459606,
+      "grad_norm": 0.4907895028591156,
+      "learning_rate": 3.917785234899329e-05,
+      "loss": 0.7305,
+      "step": 358
+    },
+    {
+      "epoch": 0.5550831078469269,
+      "grad_norm": 0.46510565280914307,
+      "learning_rate": 3.9135906040268464e-05,
+      "loss": 0.8727,
+      "step": 359
+    },
+    {
+      "epoch": 0.5566293003478933,
+      "grad_norm": 0.4917638599872589,
+      "learning_rate": 3.909395973154363e-05,
+      "loss": 0.8013,
+      "step": 360
+    },
+    {
+      "epoch": 0.5581754928488597,
+      "grad_norm": 0.49160364270210266,
+      "learning_rate": 3.90520134228188e-05,
+      "loss": 0.8871,
+      "step": 361
+    },
+    {
+      "epoch": 0.559721685349826,
+      "grad_norm": 0.5044489502906799,
+      "learning_rate": 3.901006711409396e-05,
+      "loss": 0.7945,
+      "step": 362
+    },
+    {
+      "epoch": 0.5612678778507925,
+      "grad_norm": 0.49622687697410583,
+      "learning_rate": 3.896812080536913e-05,
+      "loss": 0.7616,
+      "step": 363
+    },
+    {
+      "epoch": 0.5628140703517588,
+      "grad_norm": 0.5114305019378662,
+      "learning_rate": 3.89261744966443e-05,
+      "loss": 0.7593,
+      "step": 364
+    },
+    {
+      "epoch": 0.5643602628527251,
+      "grad_norm": 0.499931663274765,
+      "learning_rate": 3.888422818791946e-05,
+      "loss": 0.7935,
+      "step": 365
+    },
+    {
+      "epoch": 0.5659064553536916,
+      "grad_norm": 0.511566698551178,
+      "learning_rate": 3.884228187919463e-05,
+      "loss": 0.7665,
+      "step": 366
+    },
+    {
+      "epoch": 0.5674526478546579,
+      "grad_norm": 0.49260082840919495,
+      "learning_rate": 3.8800335570469796e-05,
+      "loss": 0.8207,
+      "step": 367
+    },
+    {
+      "epoch": 0.5689988403556243,
+      "grad_norm": 0.48072633147239685,
+      "learning_rate": 3.875838926174497e-05,
+      "loss": 0.7739,
+      "step": 368
+    },
+    {
+      "epoch": 0.5705450328565906,
+      "grad_norm": 0.4969918727874756,
+      "learning_rate": 3.8716442953020136e-05,
+      "loss": 0.8155,
+      "step": 369
+    },
+    {
+      "epoch": 0.572091225357557,
+      "grad_norm": 0.49780312180519104,
+      "learning_rate": 3.8674496644295306e-05,
+      "loss": 0.8397,
+      "step": 370
+    },
+    {
+      "epoch": 0.5736374178585234,
+      "grad_norm": 0.5126276612281799,
+      "learning_rate": 3.863255033557047e-05,
+      "loss": 0.8101,
+      "step": 371
+    },
+    {
+      "epoch": 0.5751836103594897,
+      "grad_norm": 0.48839470744132996,
+      "learning_rate": 3.859060402684564e-05,
+      "loss": 0.7629,
+      "step": 372
+    },
+    {
+      "epoch": 0.5767298028604562,
+      "grad_norm": 0.4904743432998657,
+      "learning_rate": 3.854865771812081e-05,
+      "loss": 0.8126,
+      "step": 373
+    },
+    {
+      "epoch": 0.5782759953614225,
+      "grad_norm": 0.48648014664649963,
+      "learning_rate": 3.850671140939597e-05,
+      "loss": 0.845,
+      "step": 374
+    },
+    {
+      "epoch": 0.5798221878623888,
+      "grad_norm": 0.537968635559082,
+      "learning_rate": 3.846476510067114e-05,
+      "loss": 0.863,
+      "step": 375
+    },
+    {
+      "epoch": 0.5813683803633553,
+      "grad_norm": 0.5034117102622986,
+      "learning_rate": 3.8422818791946305e-05,
+      "loss": 0.8333,
+      "step": 376
+    },
+    {
+      "epoch": 0.5829145728643216,
+      "grad_norm": 0.529244601726532,
+      "learning_rate": 3.838087248322148e-05,
+      "loss": 0.7756,
+      "step": 377
+    },
+    {
+      "epoch": 0.584460765365288,
+      "grad_norm": 0.5105958580970764,
+      "learning_rate": 3.8338926174496645e-05,
+      "loss": 0.8372,
+      "step": 378
+    },
+    {
+      "epoch": 0.5860069578662543,
+      "grad_norm": 0.5209059715270996,
+      "learning_rate": 3.8296979865771815e-05,
+      "loss": 0.8347,
+      "step": 379
+    },
+    {
+      "epoch": 0.5875531503672207,
+      "grad_norm": 0.5022969841957092,
+      "learning_rate": 3.8255033557046985e-05,
+      "loss": 0.8657,
+      "step": 380
+    },
+    {
+      "epoch": 0.5890993428681871,
+      "grad_norm": 0.5265902876853943,
+      "learning_rate": 3.821308724832215e-05,
+      "loss": 0.8333,
+      "step": 381
+    },
+    {
+      "epoch": 0.5906455353691534,
+      "grad_norm": 0.5082066655158997,
+      "learning_rate": 3.817114093959732e-05,
+      "loss": 0.8834,
+      "step": 382
+    },
+    {
+      "epoch": 0.5921917278701199,
+      "grad_norm": 0.5264511704444885,
+      "learning_rate": 3.812919463087248e-05,
+      "loss": 0.908,
+      "step": 383
+    },
+    {
+      "epoch": 0.5937379203710862,
+      "grad_norm": 0.5576880574226379,
+      "learning_rate": 3.808724832214765e-05,
+      "loss": 0.9174,
+      "step": 384
+    },
+    {
+      "epoch": 0.5952841128720525,
+      "grad_norm": 0.5170852541923523,
+      "learning_rate": 3.804530201342282e-05,
+      "loss": 0.8983,
+      "step": 385
+    },
+    {
+      "epoch": 0.596830305373019,
+      "grad_norm": 0.5212023854255676,
+      "learning_rate": 3.800335570469799e-05,
+      "loss": 0.8684,
+      "step": 386
+    },
+    {
+      "epoch": 0.5983764978739853,
+      "grad_norm": 0.5252931714057922,
+      "learning_rate": 3.7961409395973154e-05,
+      "loss": 0.9016,
+      "step": 387
+    },
+    {
+      "epoch": 0.5999226903749517,
+      "grad_norm": 0.5745819807052612,
+      "learning_rate": 3.7919463087248324e-05,
+      "loss": 0.8521,
+      "step": 388
+    },
+    {
+      "epoch": 0.601468882875918,
+      "grad_norm": 0.5259801149368286,
+      "learning_rate": 3.7877516778523494e-05,
+      "loss": 0.9101,
+      "step": 389
+    },
+    {
+      "epoch": 0.6030150753768844,
+      "grad_norm": 0.5453861355781555,
+      "learning_rate": 3.783557046979866e-05,
+      "loss": 0.8816,
+      "step": 390
+    },
+    {
+      "epoch": 0.6045612678778508,
+      "grad_norm": 0.5567570328712463,
+      "learning_rate": 3.779362416107383e-05,
+      "loss": 0.9126,
+      "step": 391
+    },
+    {
+      "epoch": 0.6061074603788171,
+      "grad_norm": 0.5654810070991516,
+      "learning_rate": 3.775167785234899e-05,
+      "loss": 0.8147,
+      "step": 392
+    },
+    {
+      "epoch": 0.6076536528797836,
+      "grad_norm": 0.5371072888374329,
+      "learning_rate": 3.770973154362416e-05,
+      "loss": 0.936,
+      "step": 393
+    },
+    {
+      "epoch": 0.6091998453807499,
+      "grad_norm": 0.6840121746063232,
+      "learning_rate": 3.766778523489933e-05,
+      "loss": 1.0121,
+      "step": 394
+    },
+    {
+      "epoch": 0.6107460378817162,
+      "grad_norm": 0.6007458567619324,
+      "learning_rate": 3.76258389261745e-05,
+      "loss": 0.895,
+      "step": 395
+    },
+    {
+      "epoch": 0.6122922303826827,
+      "grad_norm": 0.6644576191902161,
+      "learning_rate": 3.758389261744967e-05,
+      "loss": 0.9978,
+      "step": 396
+    },
+    {
+      "epoch": 0.613838422883649,
+      "grad_norm": 0.6520909667015076,
+      "learning_rate": 3.7541946308724834e-05,
+      "loss": 0.9551,
+      "step": 397
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.66536945104599,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.9551,
+      "step": 398
+    },
+    {
+      "epoch": 0.6169308078855817,
+      "grad_norm": 0.7479529976844788,
+      "learning_rate": 3.745805369127517e-05,
+      "loss": 1.022,
+      "step": 399
+    },
+    {
+      "epoch": 0.6184770003865481,
+      "grad_norm": 1.1065740585327148,
+      "learning_rate": 3.741610738255034e-05,
+      "loss": 0.9431,
+      "step": 400
+    },
+    {
+      "epoch": 0.6200231928875145,
+      "grad_norm": 0.4878085255622864,
+      "learning_rate": 3.7374161073825507e-05,
+      "loss": 0.728,
+      "step": 401
+    },
+    {
+      "epoch": 0.6215693853884808,
+      "grad_norm": 0.4753767251968384,
+      "learning_rate": 3.733221476510067e-05,
+      "loss": 0.7575,
+      "step": 402
+    },
+    {
+      "epoch": 0.6231155778894473,
+      "grad_norm": 0.4956800043582916,
+      "learning_rate": 3.7290268456375846e-05,
+      "loss": 0.7292,
+      "step": 403
+    },
+    {
+      "epoch": 0.6246617703904136,
+      "grad_norm": 0.48405370116233826,
+      "learning_rate": 3.724832214765101e-05,
+      "loss": 0.7879,
+      "step": 404
+    },
+    {
+      "epoch": 0.62620796289138,
+      "grad_norm": 0.46832674741744995,
+      "learning_rate": 3.720637583892618e-05,
+      "loss": 0.7819,
+      "step": 405
+    },
+    {
+      "epoch": 0.6277541553923464,
+      "grad_norm": 0.4691636264324188,
+      "learning_rate": 3.716442953020134e-05,
+      "loss": 0.7984,
+      "step": 406
+    },
+    {
+      "epoch": 0.6293003478933127,
+      "grad_norm": 0.5165214538574219,
+      "learning_rate": 3.712248322147651e-05,
+      "loss": 0.7296,
+      "step": 407
+    },
+    {
+      "epoch": 0.6308465403942791,
+      "grad_norm": 0.5001758933067322,
+      "learning_rate": 3.7080536912751676e-05,
+      "loss": 0.808,
+      "step": 408
+    },
+    {
+      "epoch": 0.6323927328952454,
+      "grad_norm": 0.4783785045146942,
+      "learning_rate": 3.7038590604026846e-05,
+      "loss": 0.7856,
+      "step": 409
+    },
+    {
+      "epoch": 0.6339389253962119,
+      "grad_norm": 0.5511770248413086,
+      "learning_rate": 3.6996644295302016e-05,
+      "loss": 0.711,
+      "step": 410
+    },
+    {
+      "epoch": 0.6354851178971782,
+      "grad_norm": 0.5434517860412598,
+      "learning_rate": 3.695469798657718e-05,
+      "loss": 0.8498,
+      "step": 411
+    },
+    {
+      "epoch": 0.6370313103981445,
+      "grad_norm": 0.5018695592880249,
+      "learning_rate": 3.6912751677852356e-05,
+      "loss": 0.7898,
+      "step": 412
+    },
+    {
+      "epoch": 0.638577502899111,
+      "grad_norm": 0.5000431537628174,
+      "learning_rate": 3.687080536912752e-05,
+      "loss": 0.7803,
+      "step": 413
+    },
+    {
+      "epoch": 0.6401236954000773,
+      "grad_norm": 0.5340814590454102,
+      "learning_rate": 3.682885906040269e-05,
+      "loss": 0.7671,
+      "step": 414
+    },
+    {
+      "epoch": 0.6416698879010437,
+      "grad_norm": 0.5046684741973877,
+      "learning_rate": 3.678691275167785e-05,
+      "loss": 0.7876,
+      "step": 415
+    },
+    {
+      "epoch": 0.6432160804020101,
+      "grad_norm": 0.5036927461624146,
+      "learning_rate": 3.674496644295302e-05,
+      "loss": 0.8726,
+      "step": 416
+    },
+    {
+      "epoch": 0.6447622729029764,
+      "grad_norm": 0.5092120170593262,
+      "learning_rate": 3.670302013422819e-05,
+      "loss": 0.806,
+      "step": 417
+    },
+    {
+      "epoch": 0.6463084654039428,
+      "grad_norm": 0.5002840757369995,
+      "learning_rate": 3.6661073825503355e-05,
+      "loss": 0.8814,
+      "step": 418
+    },
+    {
+      "epoch": 0.6478546579049091,
+      "grad_norm": 0.5107703804969788,
+      "learning_rate": 3.6619127516778525e-05,
+      "loss": 0.7984,
+      "step": 419
+    },
+    {
+      "epoch": 0.6494008504058756,
+      "grad_norm": 0.5407206416130066,
+      "learning_rate": 3.6577181208053695e-05,
+      "loss": 0.7979,
+      "step": 420
+    },
+    {
+      "epoch": 0.6509470429068419,
+      "grad_norm": 0.5402006506919861,
+      "learning_rate": 3.6535234899328865e-05,
+      "loss": 0.7547,
+      "step": 421
+    },
+    {
+      "epoch": 0.6524932354078082,
+      "grad_norm": 0.5221443176269531,
+      "learning_rate": 3.649328859060403e-05,
+      "loss": 0.8426,
+      "step": 422
+    },
+    {
+      "epoch": 0.6540394279087747,
+      "grad_norm": 0.5319470167160034,
+      "learning_rate": 3.64513422818792e-05,
+      "loss": 0.7866,
+      "step": 423
+    },
+    {
+      "epoch": 0.655585620409741,
+      "grad_norm": 0.5060355067253113,
+      "learning_rate": 3.640939597315436e-05,
+      "loss": 0.8102,
+      "step": 424
+    },
+    {
+      "epoch": 0.6571318129107074,
+      "grad_norm": 0.5067129731178284,
+      "learning_rate": 3.636744966442953e-05,
+      "loss": 0.8316,
+      "step": 425
+    },
+    {
+      "epoch": 0.6586780054116738,
+      "grad_norm": 0.5815935730934143,
+      "learning_rate": 3.63255033557047e-05,
+      "loss": 0.7207,
+      "step": 426
+    },
+    {
+      "epoch": 0.6602241979126401,
+      "grad_norm": 0.518981397151947,
+      "learning_rate": 3.6283557046979864e-05,
+      "loss": 0.7909,
+      "step": 427
+    },
+    {
+      "epoch": 0.6617703904136065,
+      "grad_norm": 0.5797117948532104,
+      "learning_rate": 3.6241610738255034e-05,
+      "loss": 0.8346,
+      "step": 428
+    },
+    {
+      "epoch": 0.6633165829145728,
+      "grad_norm": 0.565744936466217,
+      "learning_rate": 3.6199664429530204e-05,
+      "loss": 0.8366,
+      "step": 429
+    },
+    {
+      "epoch": 0.6648627754155393,
+      "grad_norm": 0.5538972020149231,
+      "learning_rate": 3.6157718120805374e-05,
+      "loss": 0.8346,
+      "step": 430
+    },
+    {
+      "epoch": 0.6664089679165056,
+      "grad_norm": 0.561184823513031,
+      "learning_rate": 3.611577181208054e-05,
+      "loss": 0.8389,
+      "step": 431
+    },
+    {
+      "epoch": 0.6679551604174719,
+      "grad_norm": 0.5446305871009827,
+      "learning_rate": 3.607382550335571e-05,
+      "loss": 0.7981,
+      "step": 432
+    },
+    {
+      "epoch": 0.6695013529184384,
+      "grad_norm": 0.6099398136138916,
+      "learning_rate": 3.603187919463088e-05,
+      "loss": 0.8988,
+      "step": 433
+    },
+    {
+      "epoch": 0.6710475454194047,
+      "grad_norm": 0.572850227355957,
+      "learning_rate": 3.598993288590604e-05,
+      "loss": 0.8981,
+      "step": 434
+    },
+    {
+      "epoch": 0.6725937379203711,
+      "grad_norm": 0.5837070941925049,
+      "learning_rate": 3.594798657718121e-05,
+      "loss": 0.9049,
+      "step": 435
+    },
+    {
+      "epoch": 0.6741399304213375,
+      "grad_norm": 0.5394341945648193,
+      "learning_rate": 3.5906040268456373e-05,
+      "loss": 0.8744,
+      "step": 436
+    },
+    {
+      "epoch": 0.6756861229223038,
+      "grad_norm": 0.5494775772094727,
+      "learning_rate": 3.5864093959731543e-05,
+      "loss": 0.9612,
+      "step": 437
+    },
+    {
+      "epoch": 0.6772323154232702,
+      "grad_norm": 0.5773348808288574,
+      "learning_rate": 3.582214765100671e-05,
+      "loss": 0.9005,
+      "step": 438
+    },
+    {
+      "epoch": 0.6787785079242366,
+      "grad_norm": 0.5757828950881958,
+      "learning_rate": 3.578020134228188e-05,
+      "loss": 0.9121,
+      "step": 439
+    },
+    {
+      "epoch": 0.680324700425203,
+      "grad_norm": 0.6010167598724365,
+      "learning_rate": 3.5738255033557046e-05,
+      "loss": 0.8575,
+      "step": 440
+    },
+    {
+      "epoch": 0.6818708929261693,
+      "grad_norm": 0.621435284614563,
+      "learning_rate": 3.5696308724832216e-05,
+      "loss": 0.9356,
+      "step": 441
+    },
+    {
+      "epoch": 0.6834170854271356,
+      "grad_norm": 0.5952706933021545,
+      "learning_rate": 3.5654362416107386e-05,
+      "loss": 0.9653,
+      "step": 442
+    },
+    {
+      "epoch": 0.6849632779281021,
+      "grad_norm": 0.59757000207901,
+      "learning_rate": 3.561241610738255e-05,
+      "loss": 0.9147,
+      "step": 443
+    },
+    {
+      "epoch": 0.6865094704290684,
+      "grad_norm": 0.6352919936180115,
+      "learning_rate": 3.557046979865772e-05,
+      "loss": 0.8931,
+      "step": 444
+    },
+    {
+      "epoch": 0.6880556629300348,
+      "grad_norm": 0.5814509987831116,
+      "learning_rate": 3.552852348993288e-05,
+      "loss": 0.9158,
+      "step": 445
+    },
+    {
+      "epoch": 0.6896018554310012,
+      "grad_norm": 0.7028055787086487,
+      "learning_rate": 3.548657718120805e-05,
+      "loss": 0.8891,
+      "step": 446
+    },
+    {
+      "epoch": 0.6911480479319675,
+      "grad_norm": 0.659831166267395,
+      "learning_rate": 3.544463087248322e-05,
+      "loss": 0.9629,
+      "step": 447
+    },
+    {
+      "epoch": 0.6926942404329339,
+      "grad_norm": 0.7297990918159485,
+      "learning_rate": 3.540268456375839e-05,
+      "loss": 0.9856,
+      "step": 448
+    },
+    {
+      "epoch": 0.6942404329339003,
+      "grad_norm": 0.8682158589363098,
+      "learning_rate": 3.536073825503356e-05,
+      "loss": 1.0006,
+      "step": 449
+    },
+    {
+      "epoch": 0.6957866254348667,
+      "grad_norm": 1.2786856889724731,
+      "learning_rate": 3.5318791946308726e-05,
+      "loss": 0.9672,
+      "step": 450
+    },
+    {
+      "epoch": 0.697332817935833,
+      "grad_norm": 0.5570630431175232,
+      "learning_rate": 3.5276845637583896e-05,
+      "loss": 0.6907,
+      "step": 451
+    },
+    {
+      "epoch": 0.6988790104367993,
+      "grad_norm": 0.4869362413883209,
+      "learning_rate": 3.523489932885906e-05,
+      "loss": 0.7276,
+      "step": 452
+    },
+    {
+      "epoch": 0.7004252029377658,
+      "grad_norm": 0.4926219582557678,
+      "learning_rate": 3.519295302013423e-05,
+      "loss": 0.7988,
+      "step": 453
+    },
+    {
+      "epoch": 0.7019713954387321,
+      "grad_norm": 0.49141383171081543,
+      "learning_rate": 3.51510067114094e-05,
+      "loss": 0.7625,
+      "step": 454
+    },
+    {
+      "epoch": 0.7035175879396985,
+      "grad_norm": 0.48099425435066223,
+      "learning_rate": 3.510906040268457e-05,
+      "loss": 0.7932,
+      "step": 455
+    },
+    {
+      "epoch": 0.7050637804406649,
+      "grad_norm": 0.48948779702186584,
+      "learning_rate": 3.506711409395974e-05,
+      "loss": 0.725,
+      "step": 456
+    },
+    {
+      "epoch": 0.7066099729416312,
+      "grad_norm": 0.5123636722564697,
+      "learning_rate": 3.50251677852349e-05,
+      "loss": 0.7704,
+      "step": 457
+    },
+    {
+      "epoch": 0.7081561654425976,
+      "grad_norm": 0.49425482749938965,
+      "learning_rate": 3.498322147651007e-05,
+      "loss": 0.8,
+      "step": 458
+    },
+    {
+      "epoch": 0.709702357943564,
+      "grad_norm": 0.5385423898696899,
+      "learning_rate": 3.4941275167785235e-05,
+      "loss": 0.8093,
+      "step": 459
+    },
+    {
+      "epoch": 0.7112485504445304,
+      "grad_norm": 0.5466816425323486,
+      "learning_rate": 3.4899328859060405e-05,
+      "loss": 0.7991,
+      "step": 460
+    },
+    {
+      "epoch": 0.7127947429454967,
+      "grad_norm": 0.5069779753684998,
+      "learning_rate": 3.485738255033557e-05,
+      "loss": 0.7989,
+      "step": 461
+    },
+    {
+      "epoch": 0.714340935446463,
+      "grad_norm": 0.5083027482032776,
+      "learning_rate": 3.481543624161074e-05,
+      "loss": 0.7891,
+      "step": 462
+    },
+    {
+      "epoch": 0.7158871279474295,
+      "grad_norm": 0.5162233710289001,
+      "learning_rate": 3.477348993288591e-05,
+      "loss": 0.7208,
+      "step": 463
+    },
+    {
+      "epoch": 0.7174333204483958,
+      "grad_norm": 0.5323002338409424,
+      "learning_rate": 3.473154362416108e-05,
+      "loss": 0.7471,
+      "step": 464
+    },
+    {
+      "epoch": 0.7189795129493622,
+      "grad_norm": 0.5144820809364319,
+      "learning_rate": 3.468959731543625e-05,
+      "loss": 0.8335,
+      "step": 465
+    },
+    {
+      "epoch": 0.7205257054503286,
+      "grad_norm": 0.5344799160957336,
+      "learning_rate": 3.464765100671141e-05,
+      "loss": 0.7575,
+      "step": 466
+    },
+    {
+      "epoch": 0.7220718979512949,
+      "grad_norm": 0.5163140296936035,
+      "learning_rate": 3.460570469798658e-05,
+      "loss": 0.7471,
+      "step": 467
+    },
+    {
+      "epoch": 0.7236180904522613,
+      "grad_norm": 0.537284791469574,
+      "learning_rate": 3.4563758389261744e-05,
+      "loss": 0.8013,
+      "step": 468
+    },
+    {
+      "epoch": 0.7251642829532277,
+      "grad_norm": 0.5194945335388184,
+      "learning_rate": 3.4521812080536914e-05,
+      "loss": 0.7855,
+      "step": 469
+    },
+    {
+      "epoch": 0.7267104754541941,
+      "grad_norm": 0.5371651649475098,
+      "learning_rate": 3.4479865771812084e-05,
+      "loss": 0.7345,
+      "step": 470
+    },
+    {
+      "epoch": 0.7282566679551604,
+      "grad_norm": 0.5294081568717957,
+      "learning_rate": 3.443791946308725e-05,
+      "loss": 0.854,
+      "step": 471
+    },
+    {
+      "epoch": 0.7298028604561267,
+      "grad_norm": 0.5522701144218445,
+      "learning_rate": 3.439597315436242e-05,
+      "loss": 0.8338,
+      "step": 472
+    },
+    {
+      "epoch": 0.7313490529570932,
+      "grad_norm": 0.5342947244644165,
+      "learning_rate": 3.435402684563759e-05,
+      "loss": 0.8429,
+      "step": 473
+    },
+    {
+      "epoch": 0.7328952454580595,
+      "grad_norm": 0.5579586029052734,
+      "learning_rate": 3.431208053691276e-05,
+      "loss": 0.77,
+      "step": 474
+    },
+    {
+      "epoch": 0.734441437959026,
+      "grad_norm": 0.5210466384887695,
+      "learning_rate": 3.427013422818792e-05,
+      "loss": 0.8538,
+      "step": 475
+    },
+    {
+      "epoch": 0.7359876304599923,
+      "grad_norm": 0.5218535661697388,
+      "learning_rate": 3.422818791946309e-05,
+      "loss": 0.8771,
+      "step": 476
+    },
+    {
+      "epoch": 0.7375338229609586,
+      "grad_norm": 0.5211417078971863,
+      "learning_rate": 3.418624161073825e-05,
+      "loss": 0.827,
+      "step": 477
+    },
+    {
+      "epoch": 0.739080015461925,
+      "grad_norm": 0.5523113012313843,
+      "learning_rate": 3.414429530201342e-05,
+      "loss": 0.8367,
+      "step": 478
+    },
+    {
+      "epoch": 0.7406262079628914,
+      "grad_norm": 0.5218014717102051,
+      "learning_rate": 3.410234899328859e-05,
+      "loss": 0.8319,
+      "step": 479
+    },
+    {
+      "epoch": 0.7421724004638578,
+      "grad_norm": 0.5514734387397766,
+      "learning_rate": 3.4060402684563756e-05,
+      "loss": 0.8703,
+      "step": 480
+    },
+    {
+      "epoch": 0.7437185929648241,
+      "grad_norm": 0.5442553162574768,
+      "learning_rate": 3.4018456375838926e-05,
+      "loss": 0.8094,
+      "step": 481
+    },
+    {
+      "epoch": 0.7452647854657904,
+      "grad_norm": 0.5472526550292969,
+      "learning_rate": 3.3976510067114096e-05,
+      "loss": 0.8419,
+      "step": 482
+    },
+    {
+      "epoch": 0.7468109779667569,
+      "grad_norm": 0.5330756306648254,
+      "learning_rate": 3.3934563758389266e-05,
+      "loss": 0.8497,
+      "step": 483
+    },
+    {
+      "epoch": 0.7483571704677232,
+      "grad_norm": 0.5805166363716125,
+      "learning_rate": 3.389261744966443e-05,
+      "loss": 0.8505,
+      "step": 484
+    },
+    {
+      "epoch": 0.7499033629686896,
+      "grad_norm": 0.5910758376121521,
+      "learning_rate": 3.38506711409396e-05,
+      "loss": 0.8154,
+      "step": 485
+    },
+    {
+      "epoch": 0.751449555469656,
+      "grad_norm": 0.579165518283844,
+      "learning_rate": 3.380872483221477e-05,
+      "loss": 0.8398,
+      "step": 486
+    },
+    {
+      "epoch": 0.7529957479706223,
+      "grad_norm": 0.5517799258232117,
+      "learning_rate": 3.376677852348993e-05,
+      "loss": 0.8426,
+      "step": 487
+    },
+    {
+      "epoch": 0.7545419404715887,
+      "grad_norm": 0.588138997554779,
+      "learning_rate": 3.37248322147651e-05,
+      "loss": 0.8658,
+      "step": 488
+    },
+    {
+      "epoch": 0.7560881329725551,
+      "grad_norm": 0.5962609648704529,
+      "learning_rate": 3.3682885906040266e-05,
+      "loss": 0.9157,
+      "step": 489
+    },
+    {
+      "epoch": 0.7576343254735215,
+      "grad_norm": 0.5821203589439392,
+      "learning_rate": 3.3640939597315436e-05,
+      "loss": 0.9158,
+      "step": 490
+    },
+    {
+      "epoch": 0.7591805179744878,
+      "grad_norm": 0.5653342008590698,
+      "learning_rate": 3.3598993288590605e-05,
+      "loss": 0.924,
+      "step": 491
+    },
+    {
+      "epoch": 0.7607267104754541,
+      "grad_norm": 0.6114529967308044,
+      "learning_rate": 3.3557046979865775e-05,
+      "loss": 0.9249,
+      "step": 492
+    },
+    {
+      "epoch": 0.7622729029764206,
+      "grad_norm": 0.6291983723640442,
+      "learning_rate": 3.3515100671140945e-05,
+      "loss": 0.8665,
+      "step": 493
+    },
+    {
+      "epoch": 0.7638190954773869,
+      "grad_norm": 0.6541888117790222,
+      "learning_rate": 3.347315436241611e-05,
+      "loss": 0.9678,
+      "step": 494
+    },
+    {
+      "epoch": 0.7653652879783533,
+      "grad_norm": 0.6252798438072205,
+      "learning_rate": 3.343120805369128e-05,
+      "loss": 0.8691,
+      "step": 495
+    },
+    {
+      "epoch": 0.7669114804793197,
+      "grad_norm": 0.6714550852775574,
+      "learning_rate": 3.338926174496644e-05,
+      "loss": 0.9581,
+      "step": 496
+    },
+    {
+      "epoch": 0.768457672980286,
+      "grad_norm": 0.7336750626564026,
+      "learning_rate": 3.334731543624161e-05,
+      "loss": 0.9024,
+      "step": 497
+    },
+    {
+      "epoch": 0.7700038654812524,
+      "grad_norm": 0.7459555864334106,
+      "learning_rate": 3.3305369127516775e-05,
+      "loss": 0.9153,
+      "step": 498
+    },
+    {
+      "epoch": 0.7715500579822188,
+      "grad_norm": 0.7960460186004639,
+      "learning_rate": 3.326342281879195e-05,
+      "loss": 0.9776,
+      "step": 499
+    },
+    {
+      "epoch": 0.7730962504831852,
+      "grad_norm": 1.188610315322876,
+      "learning_rate": 3.3221476510067115e-05,
+      "loss": 0.9963,
+      "step": 500
+    },
+    {
+      "epoch": 0.7746424429841515,
+      "grad_norm": 0.5373630523681641,
+      "learning_rate": 3.3179530201342285e-05,
+      "loss": 0.7497,
+      "step": 501
+    },
+    {
+      "epoch": 0.7761886354851179,
+      "grad_norm": 0.5297138094902039,
+      "learning_rate": 3.3137583892617455e-05,
+      "loss": 0.6842,
+      "step": 502
+    },
+    {
+      "epoch": 0.7777348279860843,
+      "grad_norm": 0.5116965770721436,
+      "learning_rate": 3.309563758389262e-05,
+      "loss": 0.7081,
+      "step": 503
+    },
+    {
+      "epoch": 0.7792810204870506,
+      "grad_norm": 0.523707389831543,
+      "learning_rate": 3.305369127516779e-05,
+      "loss": 0.7354,
+      "step": 504
+    },
+    {
+      "epoch": 0.780827212988017,
+      "grad_norm": 0.5127305388450623,
+      "learning_rate": 3.301174496644295e-05,
+      "loss": 0.7478,
+      "step": 505
+    },
+    {
+      "epoch": 0.7823734054889834,
+      "grad_norm": 0.5249508619308472,
+      "learning_rate": 3.296979865771812e-05,
+      "loss": 0.7198,
+      "step": 506
+    },
+    {
+      "epoch": 0.7839195979899497,
+      "grad_norm": 0.4998358190059662,
+      "learning_rate": 3.292785234899329e-05,
+      "loss": 0.7726,
+      "step": 507
+    },
+    {
+      "epoch": 0.7854657904909161,
+      "grad_norm": 0.48964107036590576,
+      "learning_rate": 3.288590604026846e-05,
+      "loss": 0.7643,
+      "step": 508
+    },
+    {
+      "epoch": 0.7870119829918825,
+      "grad_norm": 0.4985421299934387,
+      "learning_rate": 3.284395973154363e-05,
+      "loss": 0.7807,
+      "step": 509
+    },
+    {
+      "epoch": 0.7885581754928489,
+      "grad_norm": 0.5199535489082336,
+      "learning_rate": 3.2802013422818794e-05,
+      "loss": 0.7899,
+      "step": 510
+    },
+    {
+      "epoch": 0.7901043679938152,
+      "grad_norm": 0.5251364707946777,
+      "learning_rate": 3.2760067114093964e-05,
+      "loss": 0.7619,
+      "step": 511
+    },
+    {
+      "epoch": 0.7916505604947816,
+      "grad_norm": 0.5302333831787109,
+      "learning_rate": 3.271812080536913e-05,
+      "loss": 0.7561,
+      "step": 512
+    },
+    {
+      "epoch": 0.793196752995748,
+      "grad_norm": 0.544916033744812,
+      "learning_rate": 3.26761744966443e-05,
+      "loss": 0.7497,
+      "step": 513
+    },
+    {
+      "epoch": 0.7947429454967143,
+      "grad_norm": 0.5568458437919617,
+      "learning_rate": 3.263422818791946e-05,
+      "loss": 0.7872,
+      "step": 514
+    },
+    {
+      "epoch": 0.7962891379976808,
+      "grad_norm": 0.5136657357215881,
+      "learning_rate": 3.259228187919463e-05,
+      "loss": 0.8583,
+      "step": 515
+    },
+    {
+      "epoch": 0.7978353304986471,
+      "grad_norm": 0.5261164903640747,
+      "learning_rate": 3.25503355704698e-05,
+      "loss": 0.7806,
+      "step": 516
+    },
+    {
+      "epoch": 0.7993815229996134,
+      "grad_norm": 0.5282236933708191,
+      "learning_rate": 3.250838926174497e-05,
+      "loss": 0.7391,
+      "step": 517
+    },
+    {
+      "epoch": 0.8009277155005798,
+      "grad_norm": 0.5518252849578857,
+      "learning_rate": 3.246644295302014e-05,
+      "loss": 0.7981,
+      "step": 518
+    },
+    {
+      "epoch": 0.8024739080015462,
+      "grad_norm": 0.5146390199661255,
+      "learning_rate": 3.24244966442953e-05,
+      "loss": 0.7831,
+      "step": 519
+    },
+    {
+      "epoch": 0.8040201005025126,
+      "grad_norm": 0.5058282017707825,
+      "learning_rate": 3.238255033557047e-05,
+      "loss": 0.807,
+      "step": 520
+    },
+    {
+      "epoch": 0.8055662930034789,
+      "grad_norm": 0.5313502550125122,
+      "learning_rate": 3.2340604026845636e-05,
+      "loss": 0.8176,
+      "step": 521
+    },
+    {
+      "epoch": 0.8071124855044453,
+      "grad_norm": 0.5471131801605225,
+      "learning_rate": 3.2298657718120806e-05,
+      "loss": 0.81,
+      "step": 522
+    },
+    {
+      "epoch": 0.8086586780054117,
+      "grad_norm": 0.5153350830078125,
+      "learning_rate": 3.2256711409395976e-05,
+      "loss": 0.8729,
+      "step": 523
+    },
+    {
+      "epoch": 0.810204870506378,
+      "grad_norm": 0.5148741006851196,
+      "learning_rate": 3.221476510067114e-05,
+      "loss": 0.7776,
+      "step": 524
+    },
+    {
+      "epoch": 0.8117510630073445,
+      "grad_norm": 0.5151890516281128,
+      "learning_rate": 3.217281879194631e-05,
+      "loss": 0.8355,
+      "step": 525
+    },
+    {
+      "epoch": 0.8132972555083108,
+      "grad_norm": 0.5539215207099915,
+      "learning_rate": 3.213087248322148e-05,
+      "loss": 0.8074,
+      "step": 526
+    },
+    {
+      "epoch": 0.8148434480092771,
+      "grad_norm": 0.5416978597640991,
+      "learning_rate": 3.208892617449665e-05,
+      "loss": 0.8077,
+      "step": 527
+    },
+    {
+      "epoch": 0.8163896405102435,
+      "grad_norm": 0.5648258924484253,
+      "learning_rate": 3.204697986577181e-05,
+      "loss": 0.7928,
+      "step": 528
+    },
+    {
+      "epoch": 0.8179358330112099,
+      "grad_norm": 0.549947202205658,
+      "learning_rate": 3.200503355704698e-05,
+      "loss": 0.8371,
+      "step": 529
+    },
+    {
+      "epoch": 0.8194820255121763,
+      "grad_norm": 0.5815473198890686,
+      "learning_rate": 3.196308724832215e-05,
+      "loss": 0.8236,
+      "step": 530
+    },
+    {
+      "epoch": 0.8210282180131426,
+      "grad_norm": 0.5924739241600037,
+      "learning_rate": 3.1921140939597315e-05,
+      "loss": 0.7953,
+      "step": 531
+    },
+    {
+      "epoch": 0.822574410514109,
+      "grad_norm": 0.5691911578178406,
+      "learning_rate": 3.1879194630872485e-05,
+      "loss": 0.8312,
+      "step": 532
+    },
+    {
+      "epoch": 0.8241206030150754,
+      "grad_norm": 0.5568677186965942,
+      "learning_rate": 3.183724832214765e-05,
+      "loss": 0.839,
+      "step": 533
+    },
+    {
+      "epoch": 0.8256667955160417,
+      "grad_norm": 0.584894061088562,
+      "learning_rate": 3.1795302013422825e-05,
+      "loss": 0.9014,
+      "step": 534
+    },
+    {
+      "epoch": 0.8272129880170082,
+      "grad_norm": 0.5695962309837341,
+      "learning_rate": 3.175335570469799e-05,
+      "loss": 0.8177,
+      "step": 535
+    },
+    {
+      "epoch": 0.8287591805179745,
+      "grad_norm": 0.5662206411361694,
+      "learning_rate": 3.171140939597316e-05,
+      "loss": 0.8666,
+      "step": 536
+    },
+    {
+      "epoch": 0.8303053730189408,
+      "grad_norm": 0.6044044494628906,
+      "learning_rate": 3.166946308724832e-05,
+      "loss": 0.8757,
+      "step": 537
+    },
+    {
+      "epoch": 0.8318515655199072,
+      "grad_norm": 0.5584191083908081,
+      "learning_rate": 3.162751677852349e-05,
+      "loss": 0.8701,
+      "step": 538
+    },
+    {
+      "epoch": 0.8333977580208736,
+      "grad_norm": 0.6086747646331787,
+      "learning_rate": 3.158557046979866e-05,
+      "loss": 0.9369,
+      "step": 539
+    },
+    {
+      "epoch": 0.83494395052184,
+      "grad_norm": 0.5706461071968079,
+      "learning_rate": 3.1543624161073825e-05,
+      "loss": 0.9749,
+      "step": 540
+    },
+    {
+      "epoch": 0.8364901430228063,
+      "grad_norm": 0.5996837615966797,
+      "learning_rate": 3.1501677852348995e-05,
+      "loss": 0.881,
+      "step": 541
+    },
+    {
+      "epoch": 0.8380363355237727,
+      "grad_norm": 0.645283043384552,
+      "learning_rate": 3.145973154362416e-05,
+      "loss": 0.9201,
+      "step": 542
+    },
+    {
+      "epoch": 0.8395825280247391,
+      "grad_norm": 0.6247162222862244,
+      "learning_rate": 3.1417785234899334e-05,
+      "loss": 0.9159,
+      "step": 543
+    },
+    {
+      "epoch": 0.8411287205257054,
+      "grad_norm": 0.6042696833610535,
+      "learning_rate": 3.13758389261745e-05,
+      "loss": 0.9362,
+      "step": 544
+    },
+    {
+      "epoch": 0.8426749130266719,
+      "grad_norm": 0.6593182682991028,
+      "learning_rate": 3.133389261744967e-05,
+      "loss": 0.9722,
+      "step": 545
+    },
+    {
+      "epoch": 0.8442211055276382,
+      "grad_norm": 0.665708601474762,
+      "learning_rate": 3.129194630872484e-05,
+      "loss": 0.9028,
+      "step": 546
+    },
+    {
+      "epoch": 0.8457672980286045,
+      "grad_norm": 0.6491990089416504,
+      "learning_rate": 3.125e-05,
+      "loss": 0.9169,
+      "step": 547
+    },
+    {
+      "epoch": 0.847313490529571,
+      "grad_norm": 0.7431784868240356,
+      "learning_rate": 3.120805369127517e-05,
+      "loss": 0.9927,
+      "step": 548
+    },
+    {
+      "epoch": 0.8488596830305373,
+      "grad_norm": 0.8240005970001221,
+      "learning_rate": 3.1166107382550334e-05,
+      "loss": 1.0051,
+      "step": 549
+    },
+    {
+      "epoch": 0.8504058755315037,
+      "grad_norm": 1.1392122507095337,
+      "learning_rate": 3.1124161073825504e-05,
+      "loss": 0.9428,
+      "step": 550
+    },
+    {
+      "epoch": 0.85195206803247,
+      "grad_norm": 0.48860234022140503,
+      "learning_rate": 3.108221476510067e-05,
+      "loss": 0.7069,
+      "step": 551
+    },
+    {
+      "epoch": 0.8534982605334364,
+      "grad_norm": 0.4950990378856659,
+      "learning_rate": 3.1040268456375844e-05,
+      "loss": 0.7463,
+      "step": 552
+    },
+    {
+      "epoch": 0.8550444530344028,
+      "grad_norm": 0.48706600069999695,
+      "learning_rate": 3.099832214765101e-05,
+      "loss": 0.786,
+      "step": 553
+    },
+    {
+      "epoch": 0.8565906455353691,
+      "grad_norm": 0.5027382373809814,
+      "learning_rate": 3.095637583892618e-05,
+      "loss": 0.7882,
+      "step": 554
+    },
+    {
+      "epoch": 0.8581368380363356,
+      "grad_norm": 0.5104815363883972,
+      "learning_rate": 3.091442953020135e-05,
+      "loss": 0.7301,
+      "step": 555
+    },
+    {
+      "epoch": 0.8596830305373019,
+      "grad_norm": 0.49334749579429626,
+      "learning_rate": 3.087248322147651e-05,
+      "loss": 0.786,
+      "step": 556
+    },
+    {
+      "epoch": 0.8612292230382682,
+      "grad_norm": 0.5204259157180786,
+      "learning_rate": 3.083053691275168e-05,
+      "loss": 0.7809,
+      "step": 557
+    },
+    {
+      "epoch": 0.8627754155392346,
+      "grad_norm": 0.5358408093452454,
+      "learning_rate": 3.078859060402684e-05,
+      "loss": 0.7887,
+      "step": 558
+    },
+    {
+      "epoch": 0.864321608040201,
+      "grad_norm": 0.515990674495697,
+      "learning_rate": 3.074664429530201e-05,
+      "loss": 0.8452,
+      "step": 559
+    },
+    {
+      "epoch": 0.8658678005411674,
+      "grad_norm": 0.5604992508888245,
+      "learning_rate": 3.070469798657718e-05,
+      "loss": 0.7178,
+      "step": 560
+    },
+    {
+      "epoch": 0.8674139930421337,
+      "grad_norm": 0.5511056184768677,
+      "learning_rate": 3.066275167785235e-05,
+      "loss": 0.7577,
+      "step": 561
+    },
+    {
+      "epoch": 0.8689601855431001,
+      "grad_norm": 0.5570728778839111,
+      "learning_rate": 3.062080536912752e-05,
+      "loss": 0.7799,
+      "step": 562
+    },
+    {
+      "epoch": 0.8705063780440665,
+      "grad_norm": 0.535253643989563,
+      "learning_rate": 3.0578859060402686e-05,
+      "loss": 0.85,
+      "step": 563
+    },
+    {
+      "epoch": 0.8720525705450328,
+      "grad_norm": 0.5327886343002319,
+      "learning_rate": 3.0536912751677856e-05,
+      "loss": 0.8057,
+      "step": 564
+    },
+    {
+      "epoch": 0.8735987630459993,
+      "grad_norm": 0.5440667867660522,
+      "learning_rate": 3.0494966442953022e-05,
+      "loss": 0.7407,
+      "step": 565
+    },
+    {
+      "epoch": 0.8751449555469656,
+      "grad_norm": 0.5552430152893066,
+      "learning_rate": 3.045302013422819e-05,
+      "loss": 0.8022,
+      "step": 566
+    },
+    {
+      "epoch": 0.876691148047932,
+      "grad_norm": 0.523145854473114,
+      "learning_rate": 3.0411073825503356e-05,
+      "loss": 0.7694,
+      "step": 567
+    },
+    {
+      "epoch": 0.8782373405488983,
+      "grad_norm": 0.5461344122886658,
+      "learning_rate": 3.0369127516778522e-05,
+      "loss": 0.7849,
+      "step": 568
+    },
+    {
+      "epoch": 0.8797835330498647,
+      "grad_norm": 0.548468828201294,
+      "learning_rate": 3.0327181208053695e-05,
+      "loss": 0.8127,
+      "step": 569
+    },
+    {
+      "epoch": 0.8813297255508311,
+      "grad_norm": 0.5320367813110352,
+      "learning_rate": 3.0285234899328862e-05,
+      "loss": 0.8243,
+      "step": 570
+    },
+    {
+      "epoch": 0.8828759180517974,
+      "grad_norm": 0.5147125124931335,
+      "learning_rate": 3.024328859060403e-05,
+      "loss": 0.8784,
+      "step": 571
+    },
+    {
+      "epoch": 0.8844221105527639,
+      "grad_norm": 0.5353872776031494,
+      "learning_rate": 3.02013422818792e-05,
+      "loss": 0.7841,
+      "step": 572
+    },
+    {
+      "epoch": 0.8859683030537302,
+      "grad_norm": 0.5554389953613281,
+      "learning_rate": 3.0159395973154365e-05,
+      "loss": 0.8601,
+      "step": 573
+    },
+    {
+      "epoch": 0.8875144955546965,
+      "grad_norm": 0.5762984752655029,
+      "learning_rate": 3.011744966442953e-05,
+      "loss": 0.7359,
+      "step": 574
+    },
+    {
+      "epoch": 0.889060688055663,
+      "grad_norm": 0.5287367701530457,
+      "learning_rate": 3.0075503355704698e-05,
+      "loss": 0.8063,
+      "step": 575
+    },
+    {
+      "epoch": 0.8906068805566293,
+      "grad_norm": 0.5730248689651489,
+      "learning_rate": 3.0033557046979865e-05,
+      "loss": 0.735,
+      "step": 576
+    },
+    {
+      "epoch": 0.8921530730575957,
+      "grad_norm": 0.557384192943573,
+      "learning_rate": 2.999161073825503e-05,
+      "loss": 0.7962,
+      "step": 577
+    },
+    {
+      "epoch": 0.893699265558562,
+      "grad_norm": 0.5956771373748779,
+      "learning_rate": 2.9949664429530205e-05,
+      "loss": 0.8867,
+      "step": 578
+    },
+    {
+      "epoch": 0.8952454580595284,
+      "grad_norm": 0.5796382427215576,
+      "learning_rate": 2.990771812080537e-05,
+      "loss": 0.858,
+      "step": 579
+    },
+    {
+      "epoch": 0.8967916505604948,
+      "grad_norm": 0.5635401010513306,
+      "learning_rate": 2.986577181208054e-05,
+      "loss": 0.7987,
+      "step": 580
+    },
+    {
+      "epoch": 0.8983378430614611,
+      "grad_norm": 0.6214528679847717,
+      "learning_rate": 2.9823825503355708e-05,
+      "loss": 0.8397,
+      "step": 581
+    },
+    {
+      "epoch": 0.8998840355624276,
+      "grad_norm": 0.5945339202880859,
+      "learning_rate": 2.9781879194630874e-05,
+      "loss": 0.8428,
+      "step": 582
+    },
+    {
+      "epoch": 0.9014302280633939,
+      "grad_norm": 0.5704767107963562,
+      "learning_rate": 2.973993288590604e-05,
+      "loss": 0.8291,
+      "step": 583
+    },
+    {
+      "epoch": 0.9029764205643602,
+      "grad_norm": 0.5803176164627075,
+      "learning_rate": 2.9697986577181207e-05,
+      "loss": 0.8792,
+      "step": 584
+    },
+    {
+      "epoch": 0.9045226130653267,
+      "grad_norm": 0.6039038300514221,
+      "learning_rate": 2.9656040268456374e-05,
+      "loss": 0.7788,
+      "step": 585
+    },
+    {
+      "epoch": 0.906068805566293,
+      "grad_norm": 0.5550165176391602,
+      "learning_rate": 2.9614093959731544e-05,
+      "loss": 0.8658,
+      "step": 586
+    },
+    {
+      "epoch": 0.9076149980672594,
+      "grad_norm": 0.6096512079238892,
+      "learning_rate": 2.9572147651006714e-05,
+      "loss": 0.9237,
+      "step": 587
+    },
+    {
+      "epoch": 0.9091611905682258,
+      "grad_norm": 0.5784430503845215,
+      "learning_rate": 2.9530201342281884e-05,
+      "loss": 0.8538,
+      "step": 588
+    },
+    {
+      "epoch": 0.9107073830691921,
+      "grad_norm": 0.5854966044425964,
+      "learning_rate": 2.948825503355705e-05,
+      "loss": 0.9009,
+      "step": 589
+    },
+    {
+      "epoch": 0.9122535755701585,
+      "grad_norm": 0.5904499292373657,
+      "learning_rate": 2.9446308724832217e-05,
+      "loss": 0.9064,
+      "step": 590
+    },
+    {
+      "epoch": 0.9137997680711248,
+      "grad_norm": 0.6125240325927734,
+      "learning_rate": 2.9404362416107384e-05,
+      "loss": 0.8539,
+      "step": 591
+    },
+    {
+      "epoch": 0.9153459605720913,
+      "grad_norm": 0.6209454536437988,
+      "learning_rate": 2.936241610738255e-05,
+      "loss": 0.8864,
+      "step": 592
+    },
+    {
+      "epoch": 0.9168921530730576,
+      "grad_norm": 0.634355902671814,
+      "learning_rate": 2.9320469798657717e-05,
+      "loss": 0.8999,
+      "step": 593
+    },
+    {
+      "epoch": 0.9184383455740239,
+      "grad_norm": 0.644378125667572,
+      "learning_rate": 2.9278523489932887e-05,
+      "loss": 0.9351,
+      "step": 594
+    },
+    {
+      "epoch": 0.9199845380749904,
+      "grad_norm": 0.638783872127533,
+      "learning_rate": 2.9236577181208053e-05,
+      "loss": 0.8522,
+      "step": 595
+    },
+    {
+      "epoch": 0.9215307305759567,
+      "grad_norm": 0.6960675716400146,
+      "learning_rate": 2.9194630872483227e-05,
+      "loss": 0.8183,
+      "step": 596
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.7445054054260254,
+      "learning_rate": 2.9152684563758393e-05,
+      "loss": 0.9749,
+      "step": 597
+    },
+    {
+      "epoch": 0.9246231155778895,
+      "grad_norm": 0.7296366095542908,
+      "learning_rate": 2.911073825503356e-05,
+      "loss": 0.9535,
+      "step": 598
+    },
+    {
+      "epoch": 0.9261693080788558,
+      "grad_norm": 0.8242074251174927,
+      "learning_rate": 2.9068791946308726e-05,
+      "loss": 0.9661,
+      "step": 599
+    },
+    {
+      "epoch": 0.9277155005798222,
+      "grad_norm": 0.983094334602356,
+      "learning_rate": 2.9026845637583893e-05,
+      "loss": 0.8454,
+      "step": 600
+    },
+    {
+      "epoch": 0.9292616930807885,
+      "grad_norm": 0.4916326403617859,
+      "learning_rate": 2.898489932885906e-05,
+      "loss": 0.7178,
+      "step": 601
+    },
+    {
+      "epoch": 0.930807885581755,
+      "grad_norm": 0.5236021876335144,
+      "learning_rate": 2.894295302013423e-05,
+      "loss": 0.7586,
+      "step": 602
+    },
+    {
+      "epoch": 0.9323540780827213,
+      "grad_norm": 0.5226188898086548,
+      "learning_rate": 2.8901006711409396e-05,
+      "loss": 0.7287,
+      "step": 603
+    },
+    {
+      "epoch": 0.9339002705836876,
+      "grad_norm": 0.5189059376716614,
+      "learning_rate": 2.885906040268457e-05,
+      "loss": 0.7176,
+      "step": 604
+    },
+    {
+      "epoch": 0.9354464630846541,
+      "grad_norm": 0.5282127857208252,
+      "learning_rate": 2.8817114093959736e-05,
+      "loss": 0.7487,
+      "step": 605
+    },
+    {
+      "epoch": 0.9369926555856204,
+      "grad_norm": 0.5156176090240479,
+      "learning_rate": 2.8775167785234902e-05,
+      "loss": 0.7929,
+      "step": 606
+    },
+    {
+      "epoch": 0.9385388480865868,
+      "grad_norm": 0.5219593644142151,
+      "learning_rate": 2.873322147651007e-05,
+      "loss": 0.7992,
+      "step": 607
+    },
+    {
+      "epoch": 0.9400850405875532,
+      "grad_norm": 0.5400338768959045,
+      "learning_rate": 2.8691275167785235e-05,
+      "loss": 0.7738,
+      "step": 608
+    },
+    {
+      "epoch": 0.9416312330885195,
+      "grad_norm": 0.5067276954650879,
+      "learning_rate": 2.8649328859060402e-05,
+      "loss": 0.8046,
+      "step": 609
+    },
+    {
+      "epoch": 0.9431774255894859,
+      "grad_norm": 0.5286040902137756,
+      "learning_rate": 2.8607382550335572e-05,
+      "loss": 0.6818,
+      "step": 610
+    },
+    {
+      "epoch": 0.9447236180904522,
+      "grad_norm": 0.5325278043746948,
+      "learning_rate": 2.856543624161074e-05,
+      "loss": 0.8112,
+      "step": 611
+    },
+    {
+      "epoch": 0.9462698105914187,
+      "grad_norm": 0.5434727668762207,
+      "learning_rate": 2.8523489932885905e-05,
+      "loss": 0.8245,
+      "step": 612
+    },
+    {
+      "epoch": 0.947816003092385,
+      "grad_norm": 0.5526994466781616,
+      "learning_rate": 2.848154362416108e-05,
+      "loss": 0.7493,
+      "step": 613
+    },
+    {
+      "epoch": 0.9493621955933513,
+      "grad_norm": 0.5655114054679871,
+      "learning_rate": 2.8439597315436245e-05,
+      "loss": 0.7514,
+      "step": 614
+    },
+    {
+      "epoch": 0.9509083880943178,
+      "grad_norm": 0.5636076331138611,
+      "learning_rate": 2.839765100671141e-05,
+      "loss": 0.7992,
+      "step": 615
+    },
+    {
+      "epoch": 0.9524545805952841,
+      "grad_norm": 0.5688204169273376,
+      "learning_rate": 2.8355704697986578e-05,
+      "loss": 0.7747,
+      "step": 616
+    },
+    {
+      "epoch": 0.9540007730962505,
+      "grad_norm": 0.5534058809280396,
+      "learning_rate": 2.8313758389261748e-05,
+      "loss": 0.7511,
+      "step": 617
+    },
+    {
+      "epoch": 0.9555469655972169,
+      "grad_norm": 0.5112160444259644,
+      "learning_rate": 2.8271812080536915e-05,
+      "loss": 0.7875,
+      "step": 618
+    },
+    {
+      "epoch": 0.9570931580981832,
+      "grad_norm": 0.553830623626709,
+      "learning_rate": 2.822986577181208e-05,
+      "loss": 0.8652,
+      "step": 619
+    },
+    {
+      "epoch": 0.9586393505991496,
+      "grad_norm": 0.5614729523658752,
+      "learning_rate": 2.8187919463087248e-05,
+      "loss": 0.8693,
+      "step": 620
+    },
+    {
+      "epoch": 0.960185543100116,
+      "grad_norm": 0.5519264340400696,
+      "learning_rate": 2.8145973154362414e-05,
+      "loss": 0.8102,
+      "step": 621
+    },
+    {
+      "epoch": 0.9617317356010824,
+      "grad_norm": 0.5544281601905823,
+      "learning_rate": 2.8104026845637588e-05,
+      "loss": 0.8263,
+      "step": 622
+    },
+    {
+      "epoch": 0.9632779281020487,
+      "grad_norm": 0.5747584700584412,
+      "learning_rate": 2.8062080536912754e-05,
+      "loss": 0.7718,
+      "step": 623
+    },
+    {
+      "epoch": 0.964824120603015,
+      "grad_norm": 0.5676540732383728,
+      "learning_rate": 2.802013422818792e-05,
+      "loss": 0.7616,
+      "step": 624
+    },
+    {
+      "epoch": 0.9663703131039815,
+      "grad_norm": 0.5307291150093079,
+      "learning_rate": 2.797818791946309e-05,
+      "loss": 0.8401,
+      "step": 625
+    },
+    {
+      "epoch": 0.9679165056049478,
+      "grad_norm": 0.5527417063713074,
+      "learning_rate": 2.7936241610738257e-05,
+      "loss": 0.8818,
+      "step": 626
+    },
+    {
+      "epoch": 0.9694626981059142,
+      "grad_norm": 0.545058012008667,
+      "learning_rate": 2.7894295302013424e-05,
+      "loss": 0.8606,
+      "step": 627
+    },
+    {
+      "epoch": 0.9710088906068806,
+      "grad_norm": 0.5928349494934082,
+      "learning_rate": 2.785234899328859e-05,
+      "loss": 0.7728,
+      "step": 628
+    },
+    {
+      "epoch": 0.9725550831078469,
+      "grad_norm": 0.5348992943763733,
+      "learning_rate": 2.7810402684563757e-05,
+      "loss": 0.8101,
+      "step": 629
+    },
+    {
+      "epoch": 0.9741012756088133,
+      "grad_norm": 0.583490788936615,
+      "learning_rate": 2.7768456375838923e-05,
+      "loss": 0.8648,
+      "step": 630
+    },
+    {
+      "epoch": 0.9756474681097796,
+      "grad_norm": 0.6235148906707764,
+      "learning_rate": 2.7726510067114097e-05,
+      "loss": 0.8425,
+      "step": 631
+    },
+    {
+      "epoch": 0.9771936606107461,
+      "grad_norm": 0.5755742788314819,
+      "learning_rate": 2.7684563758389263e-05,
+      "loss": 0.8765,
+      "step": 632
+    },
+    {
+      "epoch": 0.9787398531117124,
+      "grad_norm": 0.5636020302772522,
+      "learning_rate": 2.7642617449664433e-05,
+      "loss": 0.8114,
+      "step": 633
+    },
+    {
+      "epoch": 0.9802860456126787,
+      "grad_norm": 0.5665507316589355,
+      "learning_rate": 2.76006711409396e-05,
+      "loss": 0.9008,
+      "step": 634
+    },
+    {
+      "epoch": 0.9818322381136452,
+      "grad_norm": 0.5966489911079407,
+      "learning_rate": 2.7558724832214766e-05,
+      "loss": 0.846,
+      "step": 635
+    },
+    {
+      "epoch": 0.9833784306146115,
+      "grad_norm": 0.5960109829902649,
+      "learning_rate": 2.7516778523489933e-05,
+      "loss": 0.8962,
+      "step": 636
+    },
+    {
+      "epoch": 0.9849246231155779,
+      "grad_norm": 0.5626753568649292,
+      "learning_rate": 2.74748322147651e-05,
+      "loss": 0.9062,
+      "step": 637
+    },
+    {
+      "epoch": 0.9864708156165443,
+      "grad_norm": 0.6764492392539978,
+      "learning_rate": 2.7432885906040266e-05,
+      "loss": 0.8961,
+      "step": 638
+    },
+    {
+      "epoch": 0.9880170081175106,
+      "grad_norm": 0.5922832489013672,
+      "learning_rate": 2.7390939597315436e-05,
+      "loss": 0.8525,
+      "step": 639
+    },
+    {
+      "epoch": 0.989563200618477,
+      "grad_norm": 0.6102508902549744,
+      "learning_rate": 2.7348993288590606e-05,
+      "loss": 0.8887,
+      "step": 640
+    },
+    {
+      "epoch": 0.9911093931194433,
+      "grad_norm": 0.6205296516418457,
+      "learning_rate": 2.7307046979865776e-05,
+      "loss": 0.9007,
+      "step": 641
+    },
+    {
+      "epoch": 0.9926555856204098,
+      "grad_norm": 0.6284985542297363,
+      "learning_rate": 2.7265100671140943e-05,
+      "loss": 0.9228,
+      "step": 642
+    },
+    {
+      "epoch": 0.9942017781213761,
+      "grad_norm": 0.6276938915252686,
+      "learning_rate": 2.722315436241611e-05,
+      "loss": 0.9066,
+      "step": 643
+    },
+    {
+      "epoch": 0.9957479706223424,
+      "grad_norm": 0.6849061250686646,
+      "learning_rate": 2.7181208053691276e-05,
+      "loss": 0.9204,
+      "step": 644
+    },
+    {
+      "epoch": 0.9972941631233089,
+      "grad_norm": 0.7061152458190918,
+      "learning_rate": 2.7139261744966442e-05,
+      "loss": 0.9711,
+      "step": 645
+    },
+    {
+      "epoch": 0.9988403556242752,
+      "grad_norm": 0.7999619841575623,
+      "learning_rate": 2.709731543624161e-05,
+      "loss": 0.9722,
+      "step": 646
+    },
+    {
+      "epoch": 1.0007730962504833,
+      "grad_norm": 1.3448657989501953,
+      "learning_rate": 2.705536912751678e-05,
+      "loss": 1.2285,
+      "step": 647
+    },
+    {
+      "epoch": 1.0023192887514496,
+      "grad_norm": 0.4756757318973541,
+      "learning_rate": 2.701342281879195e-05,
+      "loss": 0.7208,
+      "step": 648
+    },
+    {
+      "epoch": 1.003865481252416,
+      "grad_norm": 0.495257169008255,
+      "learning_rate": 2.697147651006712e-05,
+      "loss": 0.7645,
+      "step": 649
+    },
+    {
+      "epoch": 1.0054116737533823,
+      "grad_norm": 0.5022267699241638,
+      "learning_rate": 2.6929530201342285e-05,
+      "loss": 0.7099,
+      "step": 650
+    },
+    {
+      "epoch": 1.0069578662543486,
+      "grad_norm": 0.5082698464393616,
+      "learning_rate": 2.6887583892617452e-05,
+      "loss": 0.7451,
+      "step": 651
+    },
+    {
+      "epoch": 1.0085040587553151,
+      "grad_norm": 0.5273095369338989,
+      "learning_rate": 2.6845637583892618e-05,
+      "loss": 0.7665,
+      "step": 652
+    },
+    {
+      "epoch": 1.0100502512562815,
+      "grad_norm": 0.5531541109085083,
+      "learning_rate": 2.6803691275167785e-05,
+      "loss": 0.7041,
+      "step": 653
+    },
+    {
+      "epoch": 1.0115964437572478,
+      "grad_norm": 0.5290402173995972,
+      "learning_rate": 2.6761744966442955e-05,
+      "loss": 0.812,
+      "step": 654
+    },
+    {
+      "epoch": 1.013142636258214,
+      "grad_norm": 0.556932270526886,
+      "learning_rate": 2.671979865771812e-05,
+      "loss": 0.7414,
+      "step": 655
+    },
+    {
+      "epoch": 1.0146888287591804,
+      "grad_norm": 0.501987874507904,
+      "learning_rate": 2.6677852348993288e-05,
+      "loss": 0.7446,
+      "step": 656
+    },
+    {
+      "epoch": 1.016235021260147,
+      "grad_norm": 0.5225424766540527,
+      "learning_rate": 2.663590604026846e-05,
+      "loss": 0.8439,
+      "step": 657
+    },
+    {
+      "epoch": 1.0177812137611133,
+      "grad_norm": 0.5627469420433044,
+      "learning_rate": 2.6593959731543628e-05,
+      "loss": 0.6966,
+      "step": 658
+    },
+    {
+      "epoch": 1.0193274062620796,
+      "grad_norm": 0.6055929660797119,
+      "learning_rate": 2.6552013422818794e-05,
+      "loss": 0.7859,
+      "step": 659
+    },
+    {
+      "epoch": 1.020873598763046,
+      "grad_norm": 0.6422880291938782,
+      "learning_rate": 2.651006711409396e-05,
+      "loss": 0.7733,
+      "step": 660
+    },
+    {
+      "epoch": 1.0224197912640123,
+      "grad_norm": 0.5887860059738159,
+      "learning_rate": 2.6468120805369128e-05,
+      "loss": 0.7041,
+      "step": 661
+    },
+    {
+      "epoch": 1.0239659837649788,
+      "grad_norm": 0.5853235125541687,
+      "learning_rate": 2.6426174496644297e-05,
+      "loss": 0.7268,
+      "step": 662
+    },
+    {
+      "epoch": 1.0255121762659452,
+      "grad_norm": 0.5295597910881042,
+      "learning_rate": 2.6384228187919464e-05,
+      "loss": 0.7482,
+      "step": 663
+    },
+    {
+      "epoch": 1.0270583687669115,
+      "grad_norm": 0.5481401085853577,
+      "learning_rate": 2.634228187919463e-05,
+      "loss": 0.7548,
+      "step": 664
+    },
+    {
+      "epoch": 1.0286045612678778,
+      "grad_norm": 0.538827121257782,
+      "learning_rate": 2.6300335570469797e-05,
+      "loss": 0.764,
+      "step": 665
+    },
+    {
+      "epoch": 1.0301507537688441,
+      "grad_norm": 0.577368974685669,
+      "learning_rate": 2.625838926174497e-05,
+      "loss": 0.6835,
+      "step": 666
+    },
+    {
+      "epoch": 1.0316969462698107,
+      "grad_norm": 0.5659049153327942,
+      "learning_rate": 2.6216442953020137e-05,
+      "loss": 0.7264,
+      "step": 667
+    },
+    {
+      "epoch": 1.033243138770777,
+      "grad_norm": 0.5179746150970459,
+      "learning_rate": 2.6174496644295304e-05,
+      "loss": 0.7911,
+      "step": 668
+    },
+    {
+      "epoch": 1.0347893312717433,
+      "grad_norm": 0.5719809532165527,
+      "learning_rate": 2.613255033557047e-05,
+      "loss": 0.7479,
+      "step": 669
+    },
+    {
+      "epoch": 1.0363355237727097,
+      "grad_norm": 0.5943763852119446,
+      "learning_rate": 2.609060402684564e-05,
+      "loss": 0.7802,
+      "step": 670
+    },
+    {
+      "epoch": 1.037881716273676,
+      "grad_norm": 0.5392094850540161,
+      "learning_rate": 2.6048657718120807e-05,
+      "loss": 0.7625,
+      "step": 671
+    },
+    {
+      "epoch": 1.0394279087746425,
+      "grad_norm": 0.5679749250411987,
+      "learning_rate": 2.6006711409395973e-05,
+      "loss": 0.7911,
+      "step": 672
+    },
+    {
+      "epoch": 1.0409741012756089,
+      "grad_norm": 0.5740141272544861,
+      "learning_rate": 2.596476510067114e-05,
+      "loss": 0.7477,
+      "step": 673
+    },
+    {
+      "epoch": 1.0425202937765752,
+      "grad_norm": 0.607397198677063,
+      "learning_rate": 2.5922818791946306e-05,
+      "loss": 0.852,
+      "step": 674
+    },
+    {
+      "epoch": 1.0440664862775415,
+      "grad_norm": 0.5706917643547058,
+      "learning_rate": 2.588087248322148e-05,
+      "loss": 0.6778,
+      "step": 675
+    },
+    {
+      "epoch": 1.0456126787785078,
+      "grad_norm": 0.5882996320724487,
+      "learning_rate": 2.5838926174496646e-05,
+      "loss": 0.7071,
+      "step": 676
+    },
+    {
+      "epoch": 1.0471588712794744,
+      "grad_norm": 0.5978296995162964,
+      "learning_rate": 2.5796979865771813e-05,
+      "loss": 0.7593,
+      "step": 677
+    },
+    {
+      "epoch": 1.0487050637804407,
+      "grad_norm": 0.6237056255340576,
+      "learning_rate": 2.5755033557046983e-05,
+      "loss": 0.8099,
+      "step": 678
+    },
+    {
+      "epoch": 1.050251256281407,
+      "grad_norm": 0.6156934499740601,
+      "learning_rate": 2.571308724832215e-05,
+      "loss": 0.8256,
+      "step": 679
+    },
+    {
+      "epoch": 1.0517974487823734,
+      "grad_norm": 0.6217848062515259,
+      "learning_rate": 2.5671140939597316e-05,
+      "loss": 0.8445,
+      "step": 680
+    },
+    {
+      "epoch": 1.0533436412833397,
+      "grad_norm": 0.6550363898277283,
+      "learning_rate": 2.5629194630872482e-05,
+      "loss": 0.7931,
+      "step": 681
+    },
+    {
+      "epoch": 1.0548898337843062,
+      "grad_norm": 0.6072224378585815,
+      "learning_rate": 2.558724832214765e-05,
+      "loss": 0.7985,
+      "step": 682
+    },
+    {
+      "epoch": 1.0564360262852726,
+      "grad_norm": 0.6478685140609741,
+      "learning_rate": 2.5545302013422822e-05,
+      "loss": 0.8055,
+      "step": 683
+    },
+    {
+      "epoch": 1.0579822187862389,
+      "grad_norm": 0.6033689975738525,
+      "learning_rate": 2.550335570469799e-05,
+      "loss": 0.9145,
+      "step": 684
+    },
+    {
+      "epoch": 1.0595284112872052,
+      "grad_norm": 0.5850486755371094,
+      "learning_rate": 2.5461409395973155e-05,
+      "loss": 0.8129,
+      "step": 685
+    },
+    {
+      "epoch": 1.0610746037881715,
+      "grad_norm": 0.6233928799629211,
+      "learning_rate": 2.5419463087248325e-05,
+      "loss": 0.9122,
+      "step": 686
+    },
+    {
+      "epoch": 1.062620796289138,
+      "grad_norm": 0.6058603525161743,
+      "learning_rate": 2.5377516778523492e-05,
+      "loss": 0.7968,
+      "step": 687
+    },
+    {
+      "epoch": 1.0641669887901044,
+      "grad_norm": 0.6053382754325867,
+      "learning_rate": 2.533557046979866e-05,
+      "loss": 0.8349,
+      "step": 688
+    },
+    {
+      "epoch": 1.0657131812910707,
+      "grad_norm": 0.6354022026062012,
+      "learning_rate": 2.5293624161073825e-05,
+      "loss": 0.8879,
+      "step": 689
+    },
+    {
+      "epoch": 1.067259373792037,
+      "grad_norm": 0.6736825108528137,
+      "learning_rate": 2.525167785234899e-05,
+      "loss": 0.8317,
+      "step": 690
+    },
+    {
+      "epoch": 1.0688055662930034,
+      "grad_norm": 0.6729496717453003,
+      "learning_rate": 2.5209731543624158e-05,
+      "loss": 0.8247,
+      "step": 691
+    },
+    {
+      "epoch": 1.07035175879397,
+      "grad_norm": 0.7000686526298523,
+      "learning_rate": 2.516778523489933e-05,
+      "loss": 0.9076,
+      "step": 692
+    },
+    {
+      "epoch": 1.0718979512949363,
+      "grad_norm": 0.7471379041671753,
+      "learning_rate": 2.5125838926174498e-05,
+      "loss": 0.861,
+      "step": 693
+    },
+    {
+      "epoch": 1.0734441437959026,
+      "grad_norm": 0.7505892515182495,
+      "learning_rate": 2.5083892617449668e-05,
+      "loss": 0.9747,
+      "step": 694
+    },
+    {
+      "epoch": 1.074990336296869,
+      "grad_norm": 0.7872920632362366,
+      "learning_rate": 2.5041946308724835e-05,
+      "loss": 0.8623,
+      "step": 695
+    },
+    {
+      "epoch": 1.0765365287978352,
+      "grad_norm": 0.97450190782547,
+      "learning_rate": 2.5e-05,
+      "loss": 0.8574,
+      "step": 696
+    },
+    {
+      "epoch": 1.0780827212988018,
+      "grad_norm": 0.6834471225738525,
+      "learning_rate": 2.4958053691275168e-05,
+      "loss": 0.7284,
+      "step": 697
+    },
+    {
+      "epoch": 1.079628913799768,
+      "grad_norm": 0.5371273756027222,
+      "learning_rate": 2.4916107382550334e-05,
+      "loss": 0.6604,
+      "step": 698
+    },
+    {
+      "epoch": 1.0811751063007344,
+      "grad_norm": 0.5436398983001709,
+      "learning_rate": 2.4874161073825504e-05,
+      "loss": 0.7106,
+      "step": 699
+    },
+    {
+      "epoch": 1.0827212988017008,
+      "grad_norm": 0.5626257061958313,
+      "learning_rate": 2.4832214765100674e-05,
+      "loss": 0.7199,
+      "step": 700
+    },
+    {
+      "epoch": 1.084267491302667,
+      "grad_norm": 0.5859701037406921,
+      "learning_rate": 2.479026845637584e-05,
+      "loss": 0.727,
+      "step": 701
+    },
+    {
+      "epoch": 1.0858136838036336,
+      "grad_norm": 0.5619367957115173,
+      "learning_rate": 2.4748322147651007e-05,
+      "loss": 0.7721,
+      "step": 702
+    },
+    {
+      "epoch": 1.0873598763046,
+      "grad_norm": 0.5503877997398376,
+      "learning_rate": 2.4706375838926177e-05,
+      "loss": 0.6971,
+      "step": 703
+    },
+    {
+      "epoch": 1.0889060688055663,
+      "grad_norm": 0.5587684512138367,
+      "learning_rate": 2.4664429530201344e-05,
+      "loss": 0.7868,
+      "step": 704
+    },
+    {
+      "epoch": 1.0904522613065326,
+      "grad_norm": 0.5771764516830444,
+      "learning_rate": 2.462248322147651e-05,
+      "loss": 0.7831,
+      "step": 705
+    },
+    {
+      "epoch": 1.091998453807499,
+      "grad_norm": 0.5702334046363831,
+      "learning_rate": 2.4580536912751677e-05,
+      "loss": 0.6751,
+      "step": 706
+    },
+    {
+      "epoch": 1.0935446463084655,
+      "grad_norm": 0.5503116250038147,
+      "learning_rate": 2.4538590604026847e-05,
+      "loss": 0.7732,
+      "step": 707
+    },
+    {
+      "epoch": 1.0950908388094318,
+      "grad_norm": 0.5595600008964539,
+      "learning_rate": 2.4496644295302017e-05,
+      "loss": 0.7168,
+      "step": 708
+    },
+    {
+      "epoch": 1.0966370313103981,
+      "grad_norm": 0.5581438541412354,
+      "learning_rate": 2.4454697986577183e-05,
+      "loss": 0.7462,
+      "step": 709
+    },
+    {
+      "epoch": 1.0981832238113645,
+      "grad_norm": 0.5454738736152649,
+      "learning_rate": 2.441275167785235e-05,
+      "loss": 0.8809,
+      "step": 710
+    },
+    {
+      "epoch": 1.0997294163123308,
+      "grad_norm": 0.5763128399848938,
+      "learning_rate": 2.4370805369127517e-05,
+      "loss": 0.6942,
+      "step": 711
+    },
+    {
+      "epoch": 1.1012756088132973,
+      "grad_norm": 0.5813524127006531,
+      "learning_rate": 2.4328859060402687e-05,
+      "loss": 0.6836,
+      "step": 712
+    },
+    {
+      "epoch": 1.1028218013142637,
+      "grad_norm": 0.5554409623146057,
+      "learning_rate": 2.4286912751677853e-05,
+      "loss": 0.7616,
+      "step": 713
+    },
+    {
+      "epoch": 1.10436799381523,
+      "grad_norm": 0.5576358437538147,
+      "learning_rate": 2.424496644295302e-05,
+      "loss": 0.7835,
+      "step": 714
+    },
+    {
+      "epoch": 1.1059141863161963,
+      "grad_norm": 0.5760726928710938,
+      "learning_rate": 2.420302013422819e-05,
+      "loss": 0.7866,
+      "step": 715
+    },
+    {
+      "epoch": 1.1074603788171626,
+      "grad_norm": 0.5910109877586365,
+      "learning_rate": 2.416107382550336e-05,
+      "loss": 0.7893,
+      "step": 716
+    },
+    {
+      "epoch": 1.1090065713181292,
+      "grad_norm": 0.6322896480560303,
+      "learning_rate": 2.4119127516778526e-05,
+      "loss": 0.7235,
+      "step": 717
+    },
+    {
+      "epoch": 1.1105527638190955,
+      "grad_norm": 0.5939295887947083,
+      "learning_rate": 2.4077181208053693e-05,
+      "loss": 0.7816,
+      "step": 718
+    },
+    {
+      "epoch": 1.1120989563200618,
+      "grad_norm": 0.5953226089477539,
+      "learning_rate": 2.403523489932886e-05,
+      "loss": 0.7444,
+      "step": 719
+    },
+    {
+      "epoch": 1.1136451488210282,
+      "grad_norm": 0.60584956407547,
+      "learning_rate": 2.3993288590604026e-05,
+      "loss": 0.7525,
+      "step": 720
+    },
+    {
+      "epoch": 1.1151913413219945,
+      "grad_norm": 0.6163296699523926,
+      "learning_rate": 2.3951342281879196e-05,
+      "loss": 0.772,
+      "step": 721
+    },
+    {
+      "epoch": 1.116737533822961,
+      "grad_norm": 0.5978072881698608,
+      "learning_rate": 2.3909395973154362e-05,
+      "loss": 0.7421,
+      "step": 722
+    },
+    {
+      "epoch": 1.1182837263239274,
+      "grad_norm": 0.5980417132377625,
+      "learning_rate": 2.3867449664429532e-05,
+      "loss": 0.7942,
+      "step": 723
+    },
+    {
+      "epoch": 1.1198299188248937,
+      "grad_norm": 0.6046664118766785,
+      "learning_rate": 2.38255033557047e-05,
+      "loss": 0.8131,
+      "step": 724
+    },
+    {
+      "epoch": 1.12137611132586,
+      "grad_norm": 0.5973670482635498,
+      "learning_rate": 2.378355704697987e-05,
+      "loss": 0.7857,
+      "step": 725
+    },
+    {
+      "epoch": 1.1229223038268263,
+      "grad_norm": 0.6115438342094421,
+      "learning_rate": 2.3741610738255035e-05,
+      "loss": 0.7942,
+      "step": 726
+    },
+    {
+      "epoch": 1.1244684963277929,
+      "grad_norm": 0.5946105718612671,
+      "learning_rate": 2.3699664429530202e-05,
+      "loss": 0.8088,
+      "step": 727
+    },
+    {
+      "epoch": 1.1260146888287592,
+      "grad_norm": 0.6095959544181824,
+      "learning_rate": 2.365771812080537e-05,
+      "loss": 0.8034,
+      "step": 728
+    },
+    {
+      "epoch": 1.1275608813297255,
+      "grad_norm": 0.6506893634796143,
+      "learning_rate": 2.361577181208054e-05,
+      "loss": 0.8469,
+      "step": 729
+    },
+    {
+      "epoch": 1.1291070738306919,
+      "grad_norm": 0.6179336309432983,
+      "learning_rate": 2.3573825503355705e-05,
+      "loss": 0.7918,
+      "step": 730
+    },
+    {
+      "epoch": 1.1306532663316582,
+      "grad_norm": 0.6253457069396973,
+      "learning_rate": 2.3531879194630875e-05,
+      "loss": 0.845,
+      "step": 731
+    },
+    {
+      "epoch": 1.1321994588326247,
+      "grad_norm": 0.6577156186103821,
+      "learning_rate": 2.348993288590604e-05,
+      "loss": 0.7475,
+      "step": 732
+    },
+    {
+      "epoch": 1.133745651333591,
+      "grad_norm": 0.6993891596794128,
+      "learning_rate": 2.3447986577181208e-05,
+      "loss": 0.7988,
+      "step": 733
+    },
+    {
+      "epoch": 1.1352918438345574,
+      "grad_norm": 0.6977733373641968,
+      "learning_rate": 2.3406040268456378e-05,
+      "loss": 0.7899,
+      "step": 734
+    },
+    {
+      "epoch": 1.1368380363355237,
+      "grad_norm": 0.6664114594459534,
+      "learning_rate": 2.3364093959731545e-05,
+      "loss": 0.8609,
+      "step": 735
+    },
+    {
+      "epoch": 1.13838422883649,
+      "grad_norm": 0.6565979719161987,
+      "learning_rate": 2.332214765100671e-05,
+      "loss": 0.8592,
+      "step": 736
+    },
+    {
+      "epoch": 1.1399304213374566,
+      "grad_norm": 0.6794628500938416,
+      "learning_rate": 2.3280201342281878e-05,
+      "loss": 0.8709,
+      "step": 737
+    },
+    {
+      "epoch": 1.141476613838423,
+      "grad_norm": 0.6833226084709167,
+      "learning_rate": 2.3238255033557048e-05,
+      "loss": 0.8121,
+      "step": 738
+    },
+    {
+      "epoch": 1.1430228063393892,
+      "grad_norm": 0.6899168491363525,
+      "learning_rate": 2.3196308724832218e-05,
+      "loss": 0.8362,
+      "step": 739
+    },
+    {
+      "epoch": 1.1445689988403556,
+      "grad_norm": 0.7028947472572327,
+      "learning_rate": 2.3154362416107384e-05,
+      "loss": 0.7995,
+      "step": 740
+    },
+    {
+      "epoch": 1.146115191341322,
+      "grad_norm": 0.7305999994277954,
+      "learning_rate": 2.311241610738255e-05,
+      "loss": 0.832,
+      "step": 741
+    },
+    {
+      "epoch": 1.1476613838422884,
+      "grad_norm": 0.7404617667198181,
+      "learning_rate": 2.3070469798657717e-05,
+      "loss": 0.9275,
+      "step": 742
+    },
+    {
+      "epoch": 1.1492075763432548,
+      "grad_norm": 0.7118339538574219,
+      "learning_rate": 2.3028523489932887e-05,
+      "loss": 0.8852,
+      "step": 743
+    },
+    {
+      "epoch": 1.150753768844221,
+      "grad_norm": 0.8329970240592957,
+      "learning_rate": 2.2986577181208054e-05,
+      "loss": 0.8482,
+      "step": 744
+    },
+    {
+      "epoch": 1.1522999613451874,
+      "grad_norm": 0.8370991945266724,
+      "learning_rate": 2.2944630872483224e-05,
+      "loss": 1.0003,
+      "step": 745
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 1.0701762437820435,
+      "learning_rate": 2.290268456375839e-05,
+      "loss": 1.0102,
+      "step": 746
+    },
+    {
+      "epoch": 1.1553923463471203,
+      "grad_norm": 0.7740350961685181,
+      "learning_rate": 2.286073825503356e-05,
+      "loss": 0.617,
+      "step": 747
+    },
+    {
+      "epoch": 1.1569385388480866,
+      "grad_norm": 0.5996577739715576,
+      "learning_rate": 2.2818791946308727e-05,
+      "loss": 0.7698,
+      "step": 748
+    },
+    {
+      "epoch": 1.158484731349053,
+      "grad_norm": 0.597896933555603,
+      "learning_rate": 2.2776845637583893e-05,
+      "loss": 0.7018,
+      "step": 749
+    },
+    {
+      "epoch": 1.1600309238500193,
+      "grad_norm": 0.5840704441070557,
+      "learning_rate": 2.273489932885906e-05,
+      "loss": 0.7634,
+      "step": 750
+    },
+    {
+      "epoch": 1.1615771163509856,
+      "grad_norm": 0.5987460613250732,
+      "learning_rate": 2.269295302013423e-05,
+      "loss": 0.7973,
+      "step": 751
+    },
+    {
+      "epoch": 1.1631233088519521,
+      "grad_norm": 0.5817953944206238,
+      "learning_rate": 2.2651006711409396e-05,
+      "loss": 0.7703,
+      "step": 752
+    },
+    {
+      "epoch": 1.1646695013529185,
+      "grad_norm": 0.5792746543884277,
+      "learning_rate": 2.2609060402684566e-05,
+      "loss": 0.7425,
+      "step": 753
+    },
+    {
+      "epoch": 1.1662156938538848,
+      "grad_norm": 0.5840611457824707,
+      "learning_rate": 2.2567114093959733e-05,
+      "loss": 0.715,
+      "step": 754
+    },
+    {
+      "epoch": 1.1677618863548511,
+      "grad_norm": 0.534000039100647,
+      "learning_rate": 2.25251677852349e-05,
+      "loss": 0.7764,
+      "step": 755
+    },
+    {
+      "epoch": 1.1693080788558174,
+      "grad_norm": 0.6020839214324951,
+      "learning_rate": 2.248322147651007e-05,
+      "loss": 0.7502,
+      "step": 756
+    },
+    {
+      "epoch": 1.170854271356784,
+      "grad_norm": 0.5764395594596863,
+      "learning_rate": 2.2441275167785236e-05,
+      "loss": 0.7298,
+      "step": 757
+    },
+    {
+      "epoch": 1.1724004638577503,
+      "grad_norm": 0.5840582251548767,
+      "learning_rate": 2.2399328859060403e-05,
+      "loss": 0.7836,
+      "step": 758
+    },
+    {
+      "epoch": 1.1739466563587166,
+      "grad_norm": 0.5759351253509521,
+      "learning_rate": 2.235738255033557e-05,
+      "loss": 0.7646,
+      "step": 759
+    },
+    {
+      "epoch": 1.175492848859683,
+      "grad_norm": 0.5777841806411743,
+      "learning_rate": 2.231543624161074e-05,
+      "loss": 0.8351,
+      "step": 760
+    },
+    {
+      "epoch": 1.1770390413606493,
+      "grad_norm": 0.6419858932495117,
+      "learning_rate": 2.227348993288591e-05,
+      "loss": 0.6833,
+      "step": 761
+    },
+    {
+      "epoch": 1.1785852338616158,
+      "grad_norm": 0.5851649045944214,
+      "learning_rate": 2.2231543624161076e-05,
+      "loss": 0.7559,
+      "step": 762
+    },
+    {
+      "epoch": 1.1801314263625822,
+      "grad_norm": 0.6243789196014404,
+      "learning_rate": 2.2189597315436242e-05,
+      "loss": 0.6587,
+      "step": 763
+    },
+    {
+      "epoch": 1.1816776188635485,
+      "grad_norm": 0.6107107996940613,
+      "learning_rate": 2.2147651006711412e-05,
+      "loss": 0.7724,
+      "step": 764
+    },
+    {
+      "epoch": 1.1832238113645148,
+      "grad_norm": 0.6164106130599976,
+      "learning_rate": 2.210570469798658e-05,
+      "loss": 0.6949,
+      "step": 765
+    },
+    {
+      "epoch": 1.1847700038654811,
+      "grad_norm": 0.6295919418334961,
+      "learning_rate": 2.2063758389261745e-05,
+      "loss": 0.7239,
+      "step": 766
+    },
+    {
+      "epoch": 1.1863161963664477,
+      "grad_norm": 0.582129955291748,
+      "learning_rate": 2.2021812080536912e-05,
+      "loss": 0.7619,
+      "step": 767
+    },
+    {
+      "epoch": 1.187862388867414,
+      "grad_norm": 0.639700174331665,
+      "learning_rate": 2.197986577181208e-05,
+      "loss": 0.7788,
+      "step": 768
+    },
+    {
+      "epoch": 1.1894085813683803,
+      "grad_norm": 0.6275160908699036,
+      "learning_rate": 2.193791946308725e-05,
+      "loss": 0.7457,
+      "step": 769
+    },
+    {
+      "epoch": 1.1909547738693467,
+      "grad_norm": 0.5969827175140381,
+      "learning_rate": 2.1895973154362418e-05,
+      "loss": 0.7981,
+      "step": 770
+    },
+    {
+      "epoch": 1.192500966370313,
+      "grad_norm": 0.5809218883514404,
+      "learning_rate": 2.1854026845637585e-05,
+      "loss": 0.8194,
+      "step": 771
+    },
+    {
+      "epoch": 1.1940471588712795,
+      "grad_norm": 0.5926761627197266,
+      "learning_rate": 2.181208053691275e-05,
+      "loss": 0.7808,
+      "step": 772
+    },
+    {
+      "epoch": 1.1955933513722459,
+      "grad_norm": 0.6315684914588928,
+      "learning_rate": 2.177013422818792e-05,
+      "loss": 0.8342,
+      "step": 773
+    },
+    {
+      "epoch": 1.1971395438732122,
+      "grad_norm": 0.5824242830276489,
+      "learning_rate": 2.1728187919463088e-05,
+      "loss": 0.8041,
+      "step": 774
+    },
+    {
+      "epoch": 1.1986857363741785,
+      "grad_norm": 0.6280264258384705,
+      "learning_rate": 2.1686241610738254e-05,
+      "loss": 0.7976,
+      "step": 775
+    },
+    {
+      "epoch": 1.2002319288751448,
+      "grad_norm": 0.6388978958129883,
+      "learning_rate": 2.1644295302013424e-05,
+      "loss": 0.7253,
+      "step": 776
+    },
+    {
+      "epoch": 1.2017781213761114,
+      "grad_norm": 0.6550062894821167,
+      "learning_rate": 2.160234899328859e-05,
+      "loss": 0.7531,
+      "step": 777
+    },
+    {
+      "epoch": 1.2033243138770777,
+      "grad_norm": 0.6625634431838989,
+      "learning_rate": 2.156040268456376e-05,
+      "loss": 0.7869,
+      "step": 778
+    },
+    {
+      "epoch": 1.204870506378044,
+      "grad_norm": 0.6421430706977844,
+      "learning_rate": 2.1518456375838927e-05,
+      "loss": 0.8198,
+      "step": 779
+    },
+    {
+      "epoch": 1.2064166988790104,
+      "grad_norm": 0.6213207840919495,
+      "learning_rate": 2.1476510067114094e-05,
+      "loss": 0.8565,
+      "step": 780
+    },
+    {
+      "epoch": 1.2079628913799767,
+      "grad_norm": 0.6527850031852722,
+      "learning_rate": 2.143456375838926e-05,
+      "loss": 0.8084,
+      "step": 781
+    },
+    {
+      "epoch": 1.2095090838809432,
+      "grad_norm": 0.6330167651176453,
+      "learning_rate": 2.139261744966443e-05,
+      "loss": 0.8062,
+      "step": 782
+    },
+    {
+      "epoch": 1.2110552763819096,
+      "grad_norm": 0.6649383902549744,
+      "learning_rate": 2.13506711409396e-05,
+      "loss": 0.7586,
+      "step": 783
+    },
+    {
+      "epoch": 1.212601468882876,
+      "grad_norm": 0.6256992220878601,
+      "learning_rate": 2.1308724832214767e-05,
+      "loss": 0.7785,
+      "step": 784
+    },
+    {
+      "epoch": 1.2141476613838422,
+      "grad_norm": 0.622163712978363,
+      "learning_rate": 2.1266778523489934e-05,
+      "loss": 0.9157,
+      "step": 785
+    },
+    {
+      "epoch": 1.2156938538848086,
+      "grad_norm": 0.6411583423614502,
+      "learning_rate": 2.1224832214765103e-05,
+      "loss": 0.8916,
+      "step": 786
+    },
+    {
+      "epoch": 1.217240046385775,
+      "grad_norm": 0.6612896919250488,
+      "learning_rate": 2.118288590604027e-05,
+      "loss": 0.8234,
+      "step": 787
+    },
+    {
+      "epoch": 1.2187862388867414,
+      "grad_norm": 0.7050024271011353,
+      "learning_rate": 2.1140939597315437e-05,
+      "loss": 0.9097,
+      "step": 788
+    },
+    {
+      "epoch": 1.2203324313877078,
+      "grad_norm": 0.7046983242034912,
+      "learning_rate": 2.1098993288590603e-05,
+      "loss": 0.8503,
+      "step": 789
+    },
+    {
+      "epoch": 1.221878623888674,
+      "grad_norm": 0.7282384634017944,
+      "learning_rate": 2.1057046979865773e-05,
+      "loss": 0.8317,
+      "step": 790
+    },
+    {
+      "epoch": 1.2234248163896404,
+      "grad_norm": 0.7514353394508362,
+      "learning_rate": 2.1015100671140943e-05,
+      "loss": 0.8562,
+      "step": 791
+    },
+    {
+      "epoch": 1.224971008890607,
+      "grad_norm": 0.6923422813415527,
+      "learning_rate": 2.097315436241611e-05,
+      "loss": 0.8234,
+      "step": 792
+    },
+    {
+      "epoch": 1.2265172013915733,
+      "grad_norm": 0.7773630023002625,
+      "learning_rate": 2.0931208053691276e-05,
+      "loss": 0.8988,
+      "step": 793
+    },
+    {
+      "epoch": 1.2280633938925396,
+      "grad_norm": 0.8075311183929443,
+      "learning_rate": 2.0889261744966443e-05,
+      "loss": 0.8421,
+      "step": 794
+    },
+    {
+      "epoch": 1.229609586393506,
+      "grad_norm": 0.8537722826004028,
+      "learning_rate": 2.0847315436241613e-05,
+      "loss": 0.835,
+      "step": 795
+    },
+    {
+      "epoch": 1.2311557788944723,
+      "grad_norm": 1.1471562385559082,
+      "learning_rate": 2.080536912751678e-05,
+      "loss": 0.9985,
+      "step": 796
+    },
+    {
+      "epoch": 1.2327019713954388,
+      "grad_norm": 0.8220142126083374,
+      "learning_rate": 2.0763422818791946e-05,
+      "loss": 0.6295,
+      "step": 797
+    },
+    {
+      "epoch": 1.2342481638964051,
+      "grad_norm": 0.6230762600898743,
+      "learning_rate": 2.0721476510067116e-05,
+      "loss": 0.6921,
+      "step": 798
+    },
+    {
+      "epoch": 1.2357943563973715,
+      "grad_norm": 0.596222996711731,
+      "learning_rate": 2.0679530201342286e-05,
+      "loss": 0.7503,
+      "step": 799
+    },
+    {
+      "epoch": 1.2373405488983378,
+      "grad_norm": 0.5620123744010925,
+      "learning_rate": 2.0637583892617452e-05,
+      "loss": 0.6849,
+      "step": 800
+    },
+    {
+      "epoch": 1.238886741399304,
+      "grad_norm": 0.5710748434066772,
+      "learning_rate": 2.059563758389262e-05,
+      "loss": 0.7431,
+      "step": 801
+    },
+    {
+      "epoch": 1.2404329339002707,
+      "grad_norm": 0.584814727306366,
+      "learning_rate": 2.0553691275167785e-05,
+      "loss": 0.721,
+      "step": 802
+    },
+    {
+      "epoch": 1.241979126401237,
+      "grad_norm": 0.576964795589447,
+      "learning_rate": 2.0511744966442952e-05,
+      "loss": 0.6897,
+      "step": 803
+    },
+    {
+      "epoch": 1.2435253189022033,
+      "grad_norm": 0.6063031554222107,
+      "learning_rate": 2.0469798657718122e-05,
+      "loss": 0.7516,
+      "step": 804
+    },
+    {
+      "epoch": 1.2450715114031696,
+      "grad_norm": 0.596976101398468,
+      "learning_rate": 2.042785234899329e-05,
+      "loss": 0.7488,
+      "step": 805
+    },
+    {
+      "epoch": 1.246617703904136,
+      "grad_norm": 0.6499916911125183,
+      "learning_rate": 2.038590604026846e-05,
+      "loss": 0.7554,
+      "step": 806
+    },
+    {
+      "epoch": 1.2481638964051025,
+      "grad_norm": 0.5964358448982239,
+      "learning_rate": 2.0343959731543625e-05,
+      "loss": 0.7417,
+      "step": 807
+    },
+    {
+      "epoch": 1.2497100889060688,
+      "grad_norm": 0.5935381650924683,
+      "learning_rate": 2.0302013422818795e-05,
+      "loss": 0.7308,
+      "step": 808
+    },
+    {
+      "epoch": 1.2512562814070352,
+      "grad_norm": 0.6293373703956604,
+      "learning_rate": 2.026006711409396e-05,
+      "loss": 0.734,
+      "step": 809
+    },
+    {
+      "epoch": 1.2528024739080015,
+      "grad_norm": 0.592042088508606,
+      "learning_rate": 2.0218120805369128e-05,
+      "loss": 0.7412,
+      "step": 810
+    },
+    {
+      "epoch": 1.2543486664089678,
+      "grad_norm": 0.5861983895301819,
+      "learning_rate": 2.0176174496644295e-05,
+      "loss": 0.8101,
+      "step": 811
+    },
+    {
+      "epoch": 1.2558948589099344,
+      "grad_norm": 0.6101320385932922,
+      "learning_rate": 2.013422818791946e-05,
+      "loss": 0.7394,
+      "step": 812
+    },
+    {
+      "epoch": 1.2574410514109007,
+      "grad_norm": 0.6127662062644958,
+      "learning_rate": 2.009228187919463e-05,
+      "loss": 0.7941,
+      "step": 813
+    },
+    {
+      "epoch": 1.258987243911867,
+      "grad_norm": 0.5885155200958252,
+      "learning_rate": 2.00503355704698e-05,
+      "loss": 0.7987,
+      "step": 814
+    },
+    {
+      "epoch": 1.2605334364128333,
+      "grad_norm": 0.6043751835823059,
+      "learning_rate": 2.0008389261744968e-05,
+      "loss": 0.7918,
+      "step": 815
+    },
+    {
+      "epoch": 1.2620796289137997,
+      "grad_norm": 0.6351797580718994,
+      "learning_rate": 1.9966442953020134e-05,
+      "loss": 0.7218,
+      "step": 816
+    },
+    {
+      "epoch": 1.2636258214147662,
+      "grad_norm": 0.6218336224555969,
+      "learning_rate": 1.9924496644295304e-05,
+      "loss": 0.8294,
+      "step": 817
+    },
+    {
+      "epoch": 1.2651720139157325,
+      "grad_norm": 0.6148021817207336,
+      "learning_rate": 1.988255033557047e-05,
+      "loss": 0.7767,
+      "step": 818
+    },
+    {
+      "epoch": 1.2667182064166989,
+      "grad_norm": 0.6303946375846863,
+      "learning_rate": 1.9840604026845637e-05,
+      "loss": 0.7816,
+      "step": 819
+    },
+    {
+      "epoch": 1.2682643989176652,
+      "grad_norm": 0.6076005101203918,
+      "learning_rate": 1.9798657718120804e-05,
+      "loss": 0.8023,
+      "step": 820
+    },
+    {
+      "epoch": 1.2698105914186315,
+      "grad_norm": 0.6453294157981873,
+      "learning_rate": 1.9756711409395974e-05,
+      "loss": 0.7728,
+      "step": 821
+    },
+    {
+      "epoch": 1.271356783919598,
+      "grad_norm": 0.6279742121696472,
+      "learning_rate": 1.9714765100671144e-05,
+      "loss": 0.7564,
+      "step": 822
+    },
+    {
+      "epoch": 1.2729029764205644,
+      "grad_norm": 0.6434690356254578,
+      "learning_rate": 1.967281879194631e-05,
+      "loss": 0.7989,
+      "step": 823
+    },
+    {
+      "epoch": 1.2744491689215307,
+      "grad_norm": 0.6816707253456116,
+      "learning_rate": 1.9630872483221477e-05,
+      "loss": 0.7964,
+      "step": 824
+    },
+    {
+      "epoch": 1.275995361422497,
+      "grad_norm": 0.6387498378753662,
+      "learning_rate": 1.9588926174496643e-05,
+      "loss": 0.7621,
+      "step": 825
+    },
+    {
+      "epoch": 1.2775415539234634,
+      "grad_norm": 0.6609524488449097,
+      "learning_rate": 1.9546979865771813e-05,
+      "loss": 0.8252,
+      "step": 826
+    },
+    {
+      "epoch": 1.27908774642443,
+      "grad_norm": 0.6324920654296875,
+      "learning_rate": 1.950503355704698e-05,
+      "loss": 0.7502,
+      "step": 827
+    },
+    {
+      "epoch": 1.2806339389253962,
+      "grad_norm": 0.6569236516952515,
+      "learning_rate": 1.946308724832215e-05,
+      "loss": 0.8148,
+      "step": 828
+    },
+    {
+      "epoch": 1.2821801314263626,
+      "grad_norm": 0.625497043132782,
+      "learning_rate": 1.9421140939597316e-05,
+      "loss": 0.8663,
+      "step": 829
+    },
+    {
+      "epoch": 1.2837263239273289,
+      "grad_norm": 0.6833832263946533,
+      "learning_rate": 1.9379194630872486e-05,
+      "loss": 0.692,
+      "step": 830
+    },
+    {
+      "epoch": 1.2852725164282952,
+      "grad_norm": 0.6534375548362732,
+      "learning_rate": 1.9337248322147653e-05,
+      "loss": 0.7731,
+      "step": 831
+    },
+    {
+      "epoch": 1.2868187089292618,
+      "grad_norm": 0.6668411493301392,
+      "learning_rate": 1.929530201342282e-05,
+      "loss": 0.758,
+      "step": 832
+    },
+    {
+      "epoch": 1.288364901430228,
+      "grad_norm": 0.6552398204803467,
+      "learning_rate": 1.9253355704697986e-05,
+      "loss": 0.8387,
+      "step": 833
+    },
+    {
+      "epoch": 1.2899110939311944,
+      "grad_norm": 0.6804989576339722,
+      "learning_rate": 1.9211409395973153e-05,
+      "loss": 0.823,
+      "step": 834
+    },
+    {
+      "epoch": 1.2914572864321607,
+      "grad_norm": 0.6738516092300415,
+      "learning_rate": 1.9169463087248323e-05,
+      "loss": 0.8759,
+      "step": 835
+    },
+    {
+      "epoch": 1.293003478933127,
+      "grad_norm": 0.6957226991653442,
+      "learning_rate": 1.9127516778523493e-05,
+      "loss": 0.8465,
+      "step": 836
+    },
+    {
+      "epoch": 1.2945496714340936,
+      "grad_norm": 0.7255749106407166,
+      "learning_rate": 1.908557046979866e-05,
+      "loss": 0.7872,
+      "step": 837
+    },
+    {
+      "epoch": 1.29609586393506,
+      "grad_norm": 0.7283822894096375,
+      "learning_rate": 1.9043624161073826e-05,
+      "loss": 0.8759,
+      "step": 838
+    },
+    {
+      "epoch": 1.2976420564360263,
+      "grad_norm": 0.7126689553260803,
+      "learning_rate": 1.9001677852348996e-05,
+      "loss": 0.8252,
+      "step": 839
+    },
+    {
+      "epoch": 1.2991882489369926,
+      "grad_norm": 0.7381671071052551,
+      "learning_rate": 1.8959731543624162e-05,
+      "loss": 0.9227,
+      "step": 840
+    },
+    {
+      "epoch": 1.300734441437959,
+      "grad_norm": 0.721825897693634,
+      "learning_rate": 1.891778523489933e-05,
+      "loss": 0.9147,
+      "step": 841
+    },
+    {
+      "epoch": 1.3022806339389255,
+      "grad_norm": 0.719792902469635,
+      "learning_rate": 1.8875838926174495e-05,
+      "loss": 0.9038,
+      "step": 842
+    },
+    {
+      "epoch": 1.3038268264398918,
+      "grad_norm": 0.7600436806678772,
+      "learning_rate": 1.8833892617449665e-05,
+      "loss": 0.903,
+      "step": 843
+    },
+    {
+      "epoch": 1.305373018940858,
+      "grad_norm": 0.743240475654602,
+      "learning_rate": 1.8791946308724835e-05,
+      "loss": 0.8799,
+      "step": 844
+    },
+    {
+      "epoch": 1.3069192114418244,
+      "grad_norm": 0.8364670872688293,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.9059,
+      "step": 845
+    },
+    {
+      "epoch": 1.3084654039427908,
+      "grad_norm": 0.9684616327285767,
+      "learning_rate": 1.870805369127517e-05,
+      "loss": 0.9468,
+      "step": 846
+    },
+    {
+      "epoch": 1.3100115964437573,
+      "grad_norm": 0.7947605848312378,
+      "learning_rate": 1.8666107382550335e-05,
+      "loss": 0.6681,
+      "step": 847
+    },
+    {
+      "epoch": 1.3115577889447236,
+      "grad_norm": 0.5737844109535217,
+      "learning_rate": 1.8624161073825505e-05,
+      "loss": 0.6837,
+      "step": 848
+    },
+    {
+      "epoch": 1.31310398144569,
+      "grad_norm": 0.5890055298805237,
+      "learning_rate": 1.858221476510067e-05,
+      "loss": 0.7115,
+      "step": 849
+    },
+    {
+      "epoch": 1.3146501739466563,
+      "grad_norm": 0.6102372407913208,
+      "learning_rate": 1.8540268456375838e-05,
+      "loss": 0.7131,
+      "step": 850
+    },
+    {
+      "epoch": 1.3161963664476226,
+      "grad_norm": 0.5825210809707642,
+      "learning_rate": 1.8498322147651008e-05,
+      "loss": 0.6873,
+      "step": 851
+    },
+    {
+      "epoch": 1.3177425589485892,
+      "grad_norm": 0.5857095718383789,
+      "learning_rate": 1.8456375838926178e-05,
+      "loss": 0.7435,
+      "step": 852
+    },
+    {
+      "epoch": 1.3192887514495555,
+      "grad_norm": 0.60005122423172,
+      "learning_rate": 1.8414429530201344e-05,
+      "loss": 0.6988,
+      "step": 853
+    },
+    {
+      "epoch": 1.3208349439505218,
+      "grad_norm": 0.6294332146644592,
+      "learning_rate": 1.837248322147651e-05,
+      "loss": 0.7796,
+      "step": 854
+    },
+    {
+      "epoch": 1.3223811364514881,
+      "grad_norm": 0.6006381511688232,
+      "learning_rate": 1.8330536912751678e-05,
+      "loss": 0.7677,
+      "step": 855
+    },
+    {
+      "epoch": 1.3239273289524545,
+      "grad_norm": 0.5754826664924622,
+      "learning_rate": 1.8288590604026847e-05,
+      "loss": 0.7845,
+      "step": 856
+    },
+    {
+      "epoch": 1.325473521453421,
+      "grad_norm": 0.6118499040603638,
+      "learning_rate": 1.8246644295302014e-05,
+      "loss": 0.6819,
+      "step": 857
+    },
+    {
+      "epoch": 1.3270197139543873,
+      "grad_norm": 0.5924245119094849,
+      "learning_rate": 1.820469798657718e-05,
+      "loss": 0.8092,
+      "step": 858
+    },
+    {
+      "epoch": 1.3285659064553537,
+      "grad_norm": 0.6202556490898132,
+      "learning_rate": 1.816275167785235e-05,
+      "loss": 0.7992,
+      "step": 859
+    },
+    {
+      "epoch": 1.33011209895632,
+      "grad_norm": 0.6356550455093384,
+      "learning_rate": 1.8120805369127517e-05,
+      "loss": 0.7731,
+      "step": 860
+    },
+    {
+      "epoch": 1.3316582914572863,
+      "grad_norm": 0.6205728650093079,
+      "learning_rate": 1.8078859060402687e-05,
+      "loss": 0.7951,
+      "step": 861
+    },
+    {
+      "epoch": 1.3332044839582529,
+      "grad_norm": 0.6306980848312378,
+      "learning_rate": 1.8036912751677854e-05,
+      "loss": 0.7904,
+      "step": 862
+    },
+    {
+      "epoch": 1.3347506764592192,
+      "grad_norm": 0.5935050249099731,
+      "learning_rate": 1.799496644295302e-05,
+      "loss": 0.8029,
+      "step": 863
+    },
+    {
+      "epoch": 1.3362968689601855,
+      "grad_norm": 0.6275442838668823,
+      "learning_rate": 1.7953020134228187e-05,
+      "loss": 0.7376,
+      "step": 864
+    },
+    {
+      "epoch": 1.3378430614611518,
+      "grad_norm": 0.6528346538543701,
+      "learning_rate": 1.7911073825503357e-05,
+      "loss": 0.7286,
+      "step": 865
+    },
+    {
+      "epoch": 1.3393892539621182,
+      "grad_norm": 0.6417142748832703,
+      "learning_rate": 1.7869127516778523e-05,
+      "loss": 0.7632,
+      "step": 866
+    },
+    {
+      "epoch": 1.3409354464630847,
+      "grad_norm": 0.5968343615531921,
+      "learning_rate": 1.7827181208053693e-05,
+      "loss": 0.8156,
+      "step": 867
+    },
+    {
+      "epoch": 1.342481638964051,
+      "grad_norm": 0.5820614695549011,
+      "learning_rate": 1.778523489932886e-05,
+      "loss": 0.7957,
+      "step": 868
+    },
+    {
+      "epoch": 1.3440278314650174,
+      "grad_norm": 0.576564610004425,
+      "learning_rate": 1.7743288590604026e-05,
+      "loss": 0.7799,
+      "step": 869
+    },
+    {
+      "epoch": 1.3455740239659837,
+      "grad_norm": 0.6096657514572144,
+      "learning_rate": 1.7701342281879196e-05,
+      "loss": 0.7929,
+      "step": 870
+    },
+    {
+      "epoch": 1.34712021646695,
+      "grad_norm": 0.6413806080818176,
+      "learning_rate": 1.7659395973154363e-05,
+      "loss": 0.7468,
+      "step": 871
+    },
+    {
+      "epoch": 1.3486664089679166,
+      "grad_norm": 0.6387087106704712,
+      "learning_rate": 1.761744966442953e-05,
+      "loss": 0.7342,
+      "step": 872
+    },
+    {
+      "epoch": 1.350212601468883,
+      "grad_norm": 0.6342037320137024,
+      "learning_rate": 1.75755033557047e-05,
+      "loss": 0.829,
+      "step": 873
+    },
+    {
+      "epoch": 1.3517587939698492,
+      "grad_norm": 0.6201750636100769,
+      "learning_rate": 1.753355704697987e-05,
+      "loss": 0.7289,
+      "step": 874
+    },
+    {
+      "epoch": 1.3533049864708155,
+      "grad_norm": 0.6321792006492615,
+      "learning_rate": 1.7491610738255036e-05,
+      "loss": 0.7253,
+      "step": 875
+    },
+    {
+      "epoch": 1.3548511789717819,
+      "grad_norm": 0.6561968922615051,
+      "learning_rate": 1.7449664429530202e-05,
+      "loss": 0.7157,
+      "step": 876
+    },
+    {
+      "epoch": 1.3563973714727484,
+      "grad_norm": 0.6112854480743408,
+      "learning_rate": 1.740771812080537e-05,
+      "loss": 0.7421,
+      "step": 877
+    },
+    {
+      "epoch": 1.3579435639737147,
+      "grad_norm": 0.6668578386306763,
+      "learning_rate": 1.736577181208054e-05,
+      "loss": 0.7849,
+      "step": 878
+    },
+    {
+      "epoch": 1.359489756474681,
+      "grad_norm": 0.6496410965919495,
+      "learning_rate": 1.7323825503355705e-05,
+      "loss": 0.8031,
+      "step": 879
+    },
+    {
+      "epoch": 1.3610359489756474,
+      "grad_norm": 0.6386198401451111,
+      "learning_rate": 1.7281879194630872e-05,
+      "loss": 0.8812,
+      "step": 880
+    },
+    {
+      "epoch": 1.3625821414766137,
+      "grad_norm": 0.6844744682312012,
+      "learning_rate": 1.7239932885906042e-05,
+      "loss": 0.8245,
+      "step": 881
+    },
+    {
+      "epoch": 1.3641283339775803,
+      "grad_norm": 0.6815952658653259,
+      "learning_rate": 1.719798657718121e-05,
+      "loss": 0.8751,
+      "step": 882
+    },
+    {
+      "epoch": 1.3656745264785466,
+      "grad_norm": 0.6746396422386169,
+      "learning_rate": 1.715604026845638e-05,
+      "loss": 0.812,
+      "step": 883
+    },
+    {
+      "epoch": 1.367220718979513,
+      "grad_norm": 0.6984376907348633,
+      "learning_rate": 1.7114093959731545e-05,
+      "loss": 0.8418,
+      "step": 884
+    },
+    {
+      "epoch": 1.3687669114804792,
+      "grad_norm": 0.6949034929275513,
+      "learning_rate": 1.707214765100671e-05,
+      "loss": 0.8032,
+      "step": 885
+    },
+    {
+      "epoch": 1.3703131039814456,
+      "grad_norm": 0.6689850091934204,
+      "learning_rate": 1.7030201342281878e-05,
+      "loss": 0.8363,
+      "step": 886
+    },
+    {
+      "epoch": 1.3718592964824121,
+      "grad_norm": 0.6734853386878967,
+      "learning_rate": 1.6988255033557048e-05,
+      "loss": 0.8643,
+      "step": 887
+    },
+    {
+      "epoch": 1.3734054889833784,
+      "grad_norm": 0.7207046747207642,
+      "learning_rate": 1.6946308724832215e-05,
+      "loss": 0.853,
+      "step": 888
+    },
+    {
+      "epoch": 1.3749516814843448,
+      "grad_norm": 0.7100691199302673,
+      "learning_rate": 1.6904362416107385e-05,
+      "loss": 0.8776,
+      "step": 889
+    },
+    {
+      "epoch": 1.376497873985311,
+      "grad_norm": 0.7267347574234009,
+      "learning_rate": 1.686241610738255e-05,
+      "loss": 0.817,
+      "step": 890
+    },
+    {
+      "epoch": 1.3780440664862774,
+      "grad_norm": 0.7453994154930115,
+      "learning_rate": 1.6820469798657718e-05,
+      "loss": 0.7864,
+      "step": 891
+    },
+    {
+      "epoch": 1.379590258987244,
+      "grad_norm": 0.7423779368400574,
+      "learning_rate": 1.6778523489932888e-05,
+      "loss": 0.8475,
+      "step": 892
+    },
+    {
+      "epoch": 1.3811364514882103,
+      "grad_norm": 0.7435901165008545,
+      "learning_rate": 1.6736577181208054e-05,
+      "loss": 0.8409,
+      "step": 893
+    },
+    {
+      "epoch": 1.3826826439891766,
+      "grad_norm": 0.8185101747512817,
+      "learning_rate": 1.669463087248322e-05,
+      "loss": 0.9255,
+      "step": 894
+    },
+    {
+      "epoch": 1.384228836490143,
+      "grad_norm": 0.8886857628822327,
+      "learning_rate": 1.6652684563758387e-05,
+      "loss": 0.9095,
+      "step": 895
+    },
+    {
+      "epoch": 1.3857750289911093,
+      "grad_norm": 1.2426859140396118,
+      "learning_rate": 1.6610738255033557e-05,
+      "loss": 0.9899,
+      "step": 896
+    },
+    {
+      "epoch": 1.3873212214920758,
+      "grad_norm": 0.7389244437217712,
+      "learning_rate": 1.6568791946308727e-05,
+      "loss": 0.5189,
+      "step": 897
+    },
+    {
+      "epoch": 1.3888674139930421,
+      "grad_norm": 0.5934030413627625,
+      "learning_rate": 1.6526845637583894e-05,
+      "loss": 0.6868,
+      "step": 898
+    },
+    {
+      "epoch": 1.3904136064940085,
+      "grad_norm": 0.5986191630363464,
+      "learning_rate": 1.648489932885906e-05,
+      "loss": 0.7075,
+      "step": 899
+    },
+    {
+      "epoch": 1.3919597989949748,
+      "grad_norm": 0.6108096241950989,
+      "learning_rate": 1.644295302013423e-05,
+      "loss": 0.6855,
+      "step": 900
+    },
+    {
+      "epoch": 1.3935059914959411,
+      "grad_norm": 0.5924285650253296,
+      "learning_rate": 1.6401006711409397e-05,
+      "loss": 0.7459,
+      "step": 901
+    },
+    {
+      "epoch": 1.3950521839969077,
+      "grad_norm": 0.627100944519043,
+      "learning_rate": 1.6359060402684563e-05,
+      "loss": 0.71,
+      "step": 902
+    },
+    {
+      "epoch": 1.396598376497874,
+      "grad_norm": 0.6151503920555115,
+      "learning_rate": 1.631711409395973e-05,
+      "loss": 0.6844,
+      "step": 903
+    },
+    {
+      "epoch": 1.3981445689988403,
+      "grad_norm": 0.6181209683418274,
+      "learning_rate": 1.62751677852349e-05,
+      "loss": 0.7635,
+      "step": 904
+    },
+    {
+      "epoch": 1.3996907614998066,
+      "grad_norm": 0.5866500735282898,
+      "learning_rate": 1.623322147651007e-05,
+      "loss": 0.742,
+      "step": 905
+    },
+    {
+      "epoch": 1.401236954000773,
+      "grad_norm": 0.6311929225921631,
+      "learning_rate": 1.6191275167785237e-05,
+      "loss": 0.734,
+      "step": 906
+    },
+    {
+      "epoch": 1.4027831465017395,
+      "grad_norm": 0.6104720830917358,
+      "learning_rate": 1.6149328859060403e-05,
+      "loss": 0.7187,
+      "step": 907
+    },
+    {
+      "epoch": 1.4043293390027058,
+      "grad_norm": 0.6044167280197144,
+      "learning_rate": 1.610738255033557e-05,
+      "loss": 0.717,
+      "step": 908
+    },
+    {
+      "epoch": 1.4058755315036722,
+      "grad_norm": 0.5984567999839783,
+      "learning_rate": 1.606543624161074e-05,
+      "loss": 0.8021,
+      "step": 909
+    },
+    {
+      "epoch": 1.4074217240046385,
+      "grad_norm": 0.6355494260787964,
+      "learning_rate": 1.6023489932885906e-05,
+      "loss": 0.8507,
+      "step": 910
+    },
+    {
+      "epoch": 1.4089679165056048,
+      "grad_norm": 0.6177673935890198,
+      "learning_rate": 1.5981543624161076e-05,
+      "loss": 0.7205,
+      "step": 911
+    },
+    {
+      "epoch": 1.4105141090065714,
+      "grad_norm": 0.6486150026321411,
+      "learning_rate": 1.5939597315436243e-05,
+      "loss": 0.8078,
+      "step": 912
+    },
+    {
+      "epoch": 1.4120603015075377,
+      "grad_norm": 0.6070188879966736,
+      "learning_rate": 1.5897651006711413e-05,
+      "loss": 0.7067,
+      "step": 913
+    },
+    {
+      "epoch": 1.413606494008504,
+      "grad_norm": 0.6363996267318726,
+      "learning_rate": 1.585570469798658e-05,
+      "loss": 0.7621,
+      "step": 914
+    },
+    {
+      "epoch": 1.4151526865094703,
+      "grad_norm": 0.6805376410484314,
+      "learning_rate": 1.5813758389261746e-05,
+      "loss": 0.7354,
+      "step": 915
+    },
+    {
+      "epoch": 1.4166988790104367,
+      "grad_norm": 0.6029914617538452,
+      "learning_rate": 1.5771812080536912e-05,
+      "loss": 0.7425,
+      "step": 916
+    },
+    {
+      "epoch": 1.4182450715114032,
+      "grad_norm": 0.6197159290313721,
+      "learning_rate": 1.572986577181208e-05,
+      "loss": 0.769,
+      "step": 917
+    },
+    {
+      "epoch": 1.4197912640123695,
+      "grad_norm": 0.6450138092041016,
+      "learning_rate": 1.568791946308725e-05,
+      "loss": 0.7455,
+      "step": 918
+    },
+    {
+      "epoch": 1.4213374565133359,
+      "grad_norm": 0.6024583578109741,
+      "learning_rate": 1.564597315436242e-05,
+      "loss": 0.783,
+      "step": 919
+    },
+    {
+      "epoch": 1.4228836490143022,
+      "grad_norm": 0.6291201114654541,
+      "learning_rate": 1.5604026845637585e-05,
+      "loss": 0.7888,
+      "step": 920
+    },
+    {
+      "epoch": 1.4244298415152685,
+      "grad_norm": 0.6113951802253723,
+      "learning_rate": 1.5562080536912752e-05,
+      "loss": 0.7682,
+      "step": 921
+    },
+    {
+      "epoch": 1.425976034016235,
+      "grad_norm": 0.6421045064926147,
+      "learning_rate": 1.5520134228187922e-05,
+      "loss": 0.7691,
+      "step": 922
+    },
+    {
+      "epoch": 1.4275222265172014,
+      "grad_norm": 0.6352230906486511,
+      "learning_rate": 1.547818791946309e-05,
+      "loss": 0.7902,
+      "step": 923
+    },
+    {
+      "epoch": 1.4290684190181677,
+      "grad_norm": 0.677621066570282,
+      "learning_rate": 1.5436241610738255e-05,
+      "loss": 0.765,
+      "step": 924
+    },
+    {
+      "epoch": 1.430614611519134,
+      "grad_norm": 0.6393842697143555,
+      "learning_rate": 1.539429530201342e-05,
+      "loss": 0.6943,
+      "step": 925
+    },
+    {
+      "epoch": 1.4321608040201004,
+      "grad_norm": 0.6991260051727295,
+      "learning_rate": 1.535234899328859e-05,
+      "loss": 0.7966,
+      "step": 926
+    },
+    {
+      "epoch": 1.433706996521067,
+      "grad_norm": 0.642305314540863,
+      "learning_rate": 1.531040268456376e-05,
+      "loss": 0.8245,
+      "step": 927
+    },
+    {
+      "epoch": 1.4352531890220332,
+      "grad_norm": 0.6572315692901611,
+      "learning_rate": 1.5268456375838928e-05,
+      "loss": 0.7562,
+      "step": 928
+    },
+    {
+      "epoch": 1.4367993815229996,
+      "grad_norm": 0.6849440336227417,
+      "learning_rate": 1.5226510067114095e-05,
+      "loss": 0.8522,
+      "step": 929
+    },
+    {
+      "epoch": 1.4383455740239661,
+      "grad_norm": 0.7153504490852356,
+      "learning_rate": 1.5184563758389261e-05,
+      "loss": 0.7843,
+      "step": 930
+    },
+    {
+      "epoch": 1.4398917665249322,
+      "grad_norm": 0.6780909895896912,
+      "learning_rate": 1.5142617449664431e-05,
+      "loss": 0.8062,
+      "step": 931
+    },
+    {
+      "epoch": 1.4414379590258988,
+      "grad_norm": 0.6425846815109253,
+      "learning_rate": 1.51006711409396e-05,
+      "loss": 0.7866,
+      "step": 932
+    },
+    {
+      "epoch": 1.442984151526865,
+      "grad_norm": 0.6629062294960022,
+      "learning_rate": 1.5058724832214766e-05,
+      "loss": 0.8047,
+      "step": 933
+    },
+    {
+      "epoch": 1.4445303440278314,
+      "grad_norm": 0.6615095734596252,
+      "learning_rate": 1.5016778523489932e-05,
+      "loss": 0.8117,
+      "step": 934
+    },
+    {
+      "epoch": 1.446076536528798,
+      "grad_norm": 0.718272864818573,
+      "learning_rate": 1.4974832214765102e-05,
+      "loss": 0.8416,
+      "step": 935
+    },
+    {
+      "epoch": 1.447622729029764,
+      "grad_norm": 0.7286227941513062,
+      "learning_rate": 1.493288590604027e-05,
+      "loss": 0.8347,
+      "step": 936
+    },
+    {
+      "epoch": 1.4491689215307306,
+      "grad_norm": 0.7250097990036011,
+      "learning_rate": 1.4890939597315437e-05,
+      "loss": 0.8095,
+      "step": 937
+    },
+    {
+      "epoch": 1.450715114031697,
+      "grad_norm": 0.7696204781532288,
+      "learning_rate": 1.4848993288590604e-05,
+      "loss": 0.8611,
+      "step": 938
+    },
+    {
+      "epoch": 1.4522613065326633,
+      "grad_norm": 0.7375919222831726,
+      "learning_rate": 1.4807046979865772e-05,
+      "loss": 0.8903,
+      "step": 939
+    },
+    {
+      "epoch": 1.4538074990336298,
+      "grad_norm": 0.7759246230125427,
+      "learning_rate": 1.4765100671140942e-05,
+      "loss": 0.8005,
+      "step": 940
+    },
+    {
+      "epoch": 1.455353691534596,
+      "grad_norm": 0.8237205743789673,
+      "learning_rate": 1.4723154362416108e-05,
+      "loss": 0.8204,
+      "step": 941
+    },
+    {
+      "epoch": 1.4568998840355625,
+      "grad_norm": 0.7622566819190979,
+      "learning_rate": 1.4681208053691275e-05,
+      "loss": 0.8673,
+      "step": 942
+    },
+    {
+      "epoch": 1.4584460765365288,
+      "grad_norm": 0.7782172560691833,
+      "learning_rate": 1.4639261744966443e-05,
+      "loss": 0.9434,
+      "step": 943
+    },
+    {
+      "epoch": 1.4599922690374951,
+      "grad_norm": 0.856338381767273,
+      "learning_rate": 1.4597315436241613e-05,
+      "loss": 0.9032,
+      "step": 944
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.9459933042526245,
+      "learning_rate": 1.455536912751678e-05,
+      "loss": 0.8599,
+      "step": 945
+    },
+    {
+      "epoch": 1.4630846540394278,
+      "grad_norm": 1.1424592733383179,
+      "learning_rate": 1.4513422818791946e-05,
+      "loss": 0.879,
+      "step": 946
+    },
+    {
+      "epoch": 1.4646308465403943,
+      "grad_norm": 0.9243542551994324,
+      "learning_rate": 1.4471476510067115e-05,
+      "loss": 0.6613,
+      "step": 947
+    },
+    {
+      "epoch": 1.4661770390413607,
+      "grad_norm": 0.5729179978370667,
+      "learning_rate": 1.4429530201342285e-05,
+      "loss": 0.6693,
+      "step": 948
+    },
+    {
+      "epoch": 1.467723231542327,
+      "grad_norm": 0.6040563583374023,
+      "learning_rate": 1.4387583892617451e-05,
+      "loss": 0.663,
+      "step": 949
+    },
+    {
+      "epoch": 1.4692694240432935,
+      "grad_norm": 0.6559063196182251,
+      "learning_rate": 1.4345637583892618e-05,
+      "loss": 0.691,
+      "step": 950
+    },
+    {
+      "epoch": 1.4708156165442596,
+      "grad_norm": 0.5916107296943665,
+      "learning_rate": 1.4303691275167786e-05,
+      "loss": 0.7474,
+      "step": 951
+    },
+    {
+      "epoch": 1.4723618090452262,
+      "grad_norm": 0.5876263380050659,
+      "learning_rate": 1.4261744966442953e-05,
+      "loss": 0.7287,
+      "step": 952
+    },
+    {
+      "epoch": 1.4739080015461925,
+      "grad_norm": 0.5855628848075867,
+      "learning_rate": 1.4219798657718122e-05,
+      "loss": 0.7254,
+      "step": 953
+    },
+    {
+      "epoch": 1.4754541940471588,
+      "grad_norm": 0.5839198231697083,
+      "learning_rate": 1.4177852348993289e-05,
+      "loss": 0.7462,
+      "step": 954
+    },
+    {
+      "epoch": 1.4770003865481254,
+      "grad_norm": 0.5933511257171631,
+      "learning_rate": 1.4135906040268457e-05,
+      "loss": 0.7303,
+      "step": 955
+    },
+    {
+      "epoch": 1.4785465790490915,
+      "grad_norm": 0.620617687702179,
+      "learning_rate": 1.4093959731543624e-05,
+      "loss": 0.7287,
+      "step": 956
+    },
+    {
+      "epoch": 1.480092771550058,
+      "grad_norm": 0.6549968719482422,
+      "learning_rate": 1.4052013422818794e-05,
+      "loss": 0.7426,
+      "step": 957
+    },
+    {
+      "epoch": 1.4816389640510244,
+      "grad_norm": 0.6639398336410522,
+      "learning_rate": 1.401006711409396e-05,
+      "loss": 0.7329,
+      "step": 958
+    },
+    {
+      "epoch": 1.4831851565519907,
+      "grad_norm": 0.6285966634750366,
+      "learning_rate": 1.3968120805369129e-05,
+      "loss": 0.7302,
+      "step": 959
+    },
+    {
+      "epoch": 1.4847313490529572,
+      "grad_norm": 0.6673871278762817,
+      "learning_rate": 1.3926174496644295e-05,
+      "loss": 0.7165,
+      "step": 960
+    },
+    {
+      "epoch": 1.4862775415539233,
+      "grad_norm": 0.6557399034500122,
+      "learning_rate": 1.3884228187919462e-05,
+      "loss": 0.7465,
+      "step": 961
+    },
+    {
+      "epoch": 1.4878237340548899,
+      "grad_norm": 0.615143358707428,
+      "learning_rate": 1.3842281879194632e-05,
+      "loss": 0.7531,
+      "step": 962
+    },
+    {
+      "epoch": 1.4893699265558562,
+      "grad_norm": 0.6445402503013611,
+      "learning_rate": 1.38003355704698e-05,
+      "loss": 0.7088,
+      "step": 963
+    },
+    {
+      "epoch": 1.4909161190568225,
+      "grad_norm": 0.6428129076957703,
+      "learning_rate": 1.3758389261744966e-05,
+      "loss": 0.7011,
+      "step": 964
+    },
+    {
+      "epoch": 1.492462311557789,
+      "grad_norm": 0.6472491025924683,
+      "learning_rate": 1.3716442953020133e-05,
+      "loss": 0.8012,
+      "step": 965
+    },
+    {
+      "epoch": 1.4940085040587552,
+      "grad_norm": 0.6467485427856445,
+      "learning_rate": 1.3674496644295303e-05,
+      "loss": 0.7108,
+      "step": 966
+    },
+    {
+      "epoch": 1.4955546965597217,
+      "grad_norm": 0.648501455783844,
+      "learning_rate": 1.3632550335570471e-05,
+      "loss": 0.7786,
+      "step": 967
+    },
+    {
+      "epoch": 1.497100889060688,
+      "grad_norm": 0.6614555716514587,
+      "learning_rate": 1.3590604026845638e-05,
+      "loss": 0.6745,
+      "step": 968
+    },
+    {
+      "epoch": 1.4986470815616544,
+      "grad_norm": 0.6633491516113281,
+      "learning_rate": 1.3548657718120804e-05,
+      "loss": 0.7064,
+      "step": 969
+    },
+    {
+      "epoch": 1.500193274062621,
+      "grad_norm": 0.658253014087677,
+      "learning_rate": 1.3506711409395974e-05,
+      "loss": 0.7438,
+      "step": 970
+    },
+    {
+      "epoch": 1.501739466563587,
+      "grad_norm": 0.6717640161514282,
+      "learning_rate": 1.3464765100671143e-05,
+      "loss": 0.7792,
+      "step": 971
+    },
+    {
+      "epoch": 1.5032856590645536,
+      "grad_norm": 0.6208140850067139,
+      "learning_rate": 1.3422818791946309e-05,
+      "loss": 0.7973,
+      "step": 972
+    },
+    {
+      "epoch": 1.50483185156552,
+      "grad_norm": 0.6302463412284851,
+      "learning_rate": 1.3380872483221477e-05,
+      "loss": 0.785,
+      "step": 973
+    },
+    {
+      "epoch": 1.5063780440664862,
+      "grad_norm": 0.6479527950286865,
+      "learning_rate": 1.3338926174496644e-05,
+      "loss": 0.8266,
+      "step": 974
+    },
+    {
+      "epoch": 1.5079242365674528,
+      "grad_norm": 0.6340357661247253,
+      "learning_rate": 1.3296979865771814e-05,
+      "loss": 0.7134,
+      "step": 975
+    },
+    {
+      "epoch": 1.5094704290684189,
+      "grad_norm": 0.6478318572044373,
+      "learning_rate": 1.325503355704698e-05,
+      "loss": 0.7638,
+      "step": 976
+    },
+    {
+      "epoch": 1.5110166215693854,
+      "grad_norm": 0.6769170761108398,
+      "learning_rate": 1.3213087248322149e-05,
+      "loss": 0.7754,
+      "step": 977
+    },
+    {
+      "epoch": 1.5125628140703518,
+      "grad_norm": 0.6835642457008362,
+      "learning_rate": 1.3171140939597315e-05,
+      "loss": 0.7856,
+      "step": 978
+    },
+    {
+      "epoch": 1.514109006571318,
+      "grad_norm": 0.700343668460846,
+      "learning_rate": 1.3129194630872485e-05,
+      "loss": 0.7428,
+      "step": 979
+    },
+    {
+      "epoch": 1.5156551990722846,
+      "grad_norm": 0.6960814595222473,
+      "learning_rate": 1.3087248322147652e-05,
+      "loss": 0.7649,
+      "step": 980
+    },
+    {
+      "epoch": 1.5172013915732507,
+      "grad_norm": 0.6891623735427856,
+      "learning_rate": 1.304530201342282e-05,
+      "loss": 0.8444,
+      "step": 981
+    },
+    {
+      "epoch": 1.5187475840742173,
+      "grad_norm": 0.6881600618362427,
+      "learning_rate": 1.3003355704697987e-05,
+      "loss": 0.7956,
+      "step": 982
+    },
+    {
+      "epoch": 1.5202937765751836,
+      "grad_norm": 0.6770979762077332,
+      "learning_rate": 1.2961409395973153e-05,
+      "loss": 0.8173,
+      "step": 983
+    },
+    {
+      "epoch": 1.52183996907615,
+      "grad_norm": 0.6920093894004822,
+      "learning_rate": 1.2919463087248323e-05,
+      "loss": 0.8522,
+      "step": 984
+    },
+    {
+      "epoch": 1.5233861615771165,
+      "grad_norm": 0.6803298592567444,
+      "learning_rate": 1.2877516778523491e-05,
+      "loss": 0.8345,
+      "step": 985
+    },
+    {
+      "epoch": 1.5249323540780826,
+      "grad_norm": 0.6997124552726746,
+      "learning_rate": 1.2835570469798658e-05,
+      "loss": 0.8141,
+      "step": 986
+    },
+    {
+      "epoch": 1.5264785465790491,
+      "grad_norm": 0.7112342715263367,
+      "learning_rate": 1.2793624161073825e-05,
+      "loss": 0.8196,
+      "step": 987
+    },
+    {
+      "epoch": 1.5280247390800155,
+      "grad_norm": 0.7443994879722595,
+      "learning_rate": 1.2751677852348994e-05,
+      "loss": 0.8575,
+      "step": 988
+    },
+    {
+      "epoch": 1.5295709315809818,
+      "grad_norm": 0.6785229444503784,
+      "learning_rate": 1.2709731543624163e-05,
+      "loss": 0.8843,
+      "step": 989
+    },
+    {
+      "epoch": 1.5311171240819483,
+      "grad_norm": 0.7394726276397705,
+      "learning_rate": 1.266778523489933e-05,
+      "loss": 0.9496,
+      "step": 990
+    },
+    {
+      "epoch": 1.5326633165829144,
+      "grad_norm": 0.738186776638031,
+      "learning_rate": 1.2625838926174496e-05,
+      "loss": 0.9172,
+      "step": 991
+    },
+    {
+      "epoch": 1.534209509083881,
+      "grad_norm": 0.7867004871368408,
+      "learning_rate": 1.2583892617449666e-05,
+      "loss": 0.8743,
+      "step": 992
+    },
+    {
+      "epoch": 1.5357557015848473,
+      "grad_norm": 0.8075705170631409,
+      "learning_rate": 1.2541946308724834e-05,
+      "loss": 0.9021,
+      "step": 993
+    },
+    {
+      "epoch": 1.5373018940858136,
+      "grad_norm": 0.8144130110740662,
+      "learning_rate": 1.25e-05,
+      "loss": 0.8258,
+      "step": 994
+    },
+    {
+      "epoch": 1.5388480865867802,
+      "grad_norm": 0.8460942506790161,
+      "learning_rate": 1.2458053691275167e-05,
+      "loss": 0.8051,
+      "step": 995
+    },
+    {
+      "epoch": 1.5403942790877463,
+      "grad_norm": 1.0733917951583862,
+      "learning_rate": 1.2416107382550337e-05,
+      "loss": 0.9469,
+      "step": 996
+    },
+    {
+      "epoch": 1.5419404715887128,
+      "grad_norm": 0.6984366178512573,
+      "learning_rate": 1.2374161073825504e-05,
+      "loss": 0.6271,
+      "step": 997
+    },
+    {
+      "epoch": 1.5434866640896792,
+      "grad_norm": 0.5974945425987244,
+      "learning_rate": 1.2332214765100672e-05,
+      "loss": 0.6767,
+      "step": 998
+    },
+    {
+      "epoch": 1.5450328565906455,
+      "grad_norm": 0.5917332172393799,
+      "learning_rate": 1.2290268456375838e-05,
+      "loss": 0.7027,
+      "step": 999
+    },
+    {
+      "epoch": 1.546579049091612,
+      "grad_norm": 0.5879162549972534,
+      "learning_rate": 1.2248322147651008e-05,
+      "loss": 0.6444,
+      "step": 1000
+    },
+    {
+      "epoch": 1.5481252415925781,
+      "grad_norm": 0.5611408948898315,
+      "learning_rate": 1.2206375838926175e-05,
+      "loss": 0.6743,
+      "step": 1001
+    },
+    {
+      "epoch": 1.5496714340935447,
+      "grad_norm": 0.6579827666282654,
+      "learning_rate": 1.2164429530201343e-05,
+      "loss": 0.6863,
+      "step": 1002
+    },
+    {
+      "epoch": 1.551217626594511,
+      "grad_norm": 0.6543861031532288,
+      "learning_rate": 1.212248322147651e-05,
+      "loss": 0.6911,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5527638190954773,
+      "grad_norm": 0.6316918134689331,
+      "learning_rate": 1.208053691275168e-05,
+      "loss": 0.7126,
+      "step": 1004
+    },
+    {
+      "epoch": 1.5543100115964439,
+      "grad_norm": 0.6070011258125305,
+      "learning_rate": 1.2038590604026846e-05,
+      "loss": 0.5802,
+      "step": 1005
+    },
+    {
+      "epoch": 1.55585620409741,
+      "grad_norm": 0.6225163340568542,
+      "learning_rate": 1.1996644295302013e-05,
+      "loss": 0.7565,
+      "step": 1006
+    },
+    {
+      "epoch": 1.5574023965983765,
+      "grad_norm": 0.6303391456604004,
+      "learning_rate": 1.1954697986577181e-05,
+      "loss": 0.7537,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5589485890993429,
+      "grad_norm": 0.6572604179382324,
+      "learning_rate": 1.191275167785235e-05,
+      "loss": 0.7332,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5604947816003092,
+      "grad_norm": 0.6323292255401611,
+      "learning_rate": 1.1870805369127518e-05,
+      "loss": 0.7537,
+      "step": 1009
+    },
+    {
+      "epoch": 1.5620409741012757,
+      "grad_norm": 0.621051549911499,
+      "learning_rate": 1.1828859060402684e-05,
+      "loss": 0.7371,
+      "step": 1010
+    },
+    {
+      "epoch": 1.5635871666022418,
+      "grad_norm": 0.6632164120674133,
+      "learning_rate": 1.1786912751677852e-05,
+      "loss": 0.7169,
+      "step": 1011
+    },
+    {
+      "epoch": 1.5651333591032084,
+      "grad_norm": 0.6632513999938965,
+      "learning_rate": 1.174496644295302e-05,
+      "loss": 0.726,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5666795516041747,
+      "grad_norm": 0.6338618397712708,
+      "learning_rate": 1.1703020134228189e-05,
+      "loss": 0.7774,
+      "step": 1013
+    },
+    {
+      "epoch": 1.568225744105141,
+      "grad_norm": 0.6426889896392822,
+      "learning_rate": 1.1661073825503356e-05,
+      "loss": 0.712,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5697719366061076,
+      "grad_norm": 0.6418123841285706,
+      "learning_rate": 1.1619127516778524e-05,
+      "loss": 0.692,
+      "step": 1015
+    },
+    {
+      "epoch": 1.5713181291070737,
+      "grad_norm": 0.6338704228401184,
+      "learning_rate": 1.1577181208053692e-05,
+      "loss": 0.7498,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5728643216080402,
+      "grad_norm": 0.6129948496818542,
+      "learning_rate": 1.1535234899328859e-05,
+      "loss": 0.8721,
+      "step": 1017
+    },
+    {
+      "epoch": 1.5744105141090066,
+      "grad_norm": 0.6413542032241821,
+      "learning_rate": 1.1493288590604027e-05,
+      "loss": 0.8381,
+      "step": 1018
+    },
+    {
+      "epoch": 1.575956706609973,
+      "grad_norm": 0.6431681513786316,
+      "learning_rate": 1.1451342281879195e-05,
+      "loss": 0.7365,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5775028991109394,
+      "grad_norm": 0.6414735317230225,
+      "learning_rate": 1.1409395973154363e-05,
+      "loss": 0.8006,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5790490916119055,
+      "grad_norm": 0.6431716680526733,
+      "learning_rate": 1.136744966442953e-05,
+      "loss": 0.7759,
+      "step": 1021
+    },
+    {
+      "epoch": 1.580595284112872,
+      "grad_norm": 0.6646420359611511,
+      "learning_rate": 1.1325503355704698e-05,
+      "loss": 0.7038,
+      "step": 1022
+    },
+    {
+      "epoch": 1.5821414766138384,
+      "grad_norm": 0.7008923292160034,
+      "learning_rate": 1.1283557046979866e-05,
+      "loss": 0.719,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5836876691148047,
+      "grad_norm": 0.6393096446990967,
+      "learning_rate": 1.1241610738255035e-05,
+      "loss": 0.7841,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5852338616157713,
+      "grad_norm": 0.6879417300224304,
+      "learning_rate": 1.1199664429530201e-05,
+      "loss": 0.8218,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5867800541167374,
+      "grad_norm": 0.6500439047813416,
+      "learning_rate": 1.115771812080537e-05,
+      "loss": 0.8026,
+      "step": 1026
+    },
+    {
+      "epoch": 1.588326246617704,
+      "grad_norm": 0.6967812180519104,
+      "learning_rate": 1.1115771812080538e-05,
+      "loss": 0.7494,
+      "step": 1027
+    },
+    {
+      "epoch": 1.5898724391186703,
+      "grad_norm": 0.6527352929115295,
+      "learning_rate": 1.1073825503355706e-05,
+      "loss": 0.8472,
+      "step": 1028
+    },
+    {
+      "epoch": 1.5914186316196366,
+      "grad_norm": 0.7033571004867554,
+      "learning_rate": 1.1031879194630873e-05,
+      "loss": 0.7553,
+      "step": 1029
+    },
+    {
+      "epoch": 1.5929648241206031,
+      "grad_norm": 0.6790737509727478,
+      "learning_rate": 1.098993288590604e-05,
+      "loss": 0.8045,
+      "step": 1030
+    },
+    {
+      "epoch": 1.5945110166215692,
+      "grad_norm": 0.7032893300056458,
+      "learning_rate": 1.0947986577181209e-05,
+      "loss": 0.8292,
+      "step": 1031
+    },
+    {
+      "epoch": 1.5960572091225358,
+      "grad_norm": 0.6649371981620789,
+      "learning_rate": 1.0906040268456376e-05,
+      "loss": 0.8032,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5976034016235021,
+      "grad_norm": 0.7009012699127197,
+      "learning_rate": 1.0864093959731544e-05,
+      "loss": 0.9107,
+      "step": 1033
+    },
+    {
+      "epoch": 1.5991495941244684,
+      "grad_norm": 0.7047061324119568,
+      "learning_rate": 1.0822147651006712e-05,
+      "loss": 0.904,
+      "step": 1034
+    },
+    {
+      "epoch": 1.600695786625435,
+      "grad_norm": 0.7520581483840942,
+      "learning_rate": 1.078020134228188e-05,
+      "loss": 0.8972,
+      "step": 1035
+    },
+    {
+      "epoch": 1.602241979126401,
+      "grad_norm": 0.710966944694519,
+      "learning_rate": 1.0738255033557047e-05,
+      "loss": 0.8133,
+      "step": 1036
+    },
+    {
+      "epoch": 1.6037881716273676,
+      "grad_norm": 0.7045226693153381,
+      "learning_rate": 1.0696308724832215e-05,
+      "loss": 0.8474,
+      "step": 1037
+    },
+    {
+      "epoch": 1.605334364128334,
+      "grad_norm": 0.7222912311553955,
+      "learning_rate": 1.0654362416107383e-05,
+      "loss": 0.8577,
+      "step": 1038
+    },
+    {
+      "epoch": 1.6068805566293003,
+      "grad_norm": 0.7768924236297607,
+      "learning_rate": 1.0612416107382552e-05,
+      "loss": 0.8906,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6084267491302668,
+      "grad_norm": 0.7447935342788696,
+      "learning_rate": 1.0570469798657718e-05,
+      "loss": 0.8571,
+      "step": 1040
+    },
+    {
+      "epoch": 1.609972941631233,
+      "grad_norm": 0.7860161662101746,
+      "learning_rate": 1.0528523489932887e-05,
+      "loss": 0.868,
+      "step": 1041
+    },
+    {
+      "epoch": 1.6115191341321995,
+      "grad_norm": 0.772621750831604,
+      "learning_rate": 1.0486577181208055e-05,
+      "loss": 0.9064,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6130653266331658,
+      "grad_norm": 0.7927963733673096,
+      "learning_rate": 1.0444630872483221e-05,
+      "loss": 0.8628,
+      "step": 1043
+    },
+    {
+      "epoch": 1.6146115191341321,
+      "grad_norm": 0.8811314105987549,
+      "learning_rate": 1.040268456375839e-05,
+      "loss": 0.9203,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6161577116350987,
+      "grad_norm": 1.0155308246612549,
+      "learning_rate": 1.0360738255033558e-05,
+      "loss": 0.9134,
+      "step": 1045
+    },
+    {
+      "epoch": 1.6177039041360648,
+      "grad_norm": 1.3558322191238403,
+      "learning_rate": 1.0318791946308726e-05,
+      "loss": 1.007,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6192500966370313,
+      "grad_norm": 0.771702766418457,
+      "learning_rate": 1.0276845637583893e-05,
+      "loss": 0.5492,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6207962891379977,
+      "grad_norm": 0.5778831243515015,
+      "learning_rate": 1.0234899328859061e-05,
+      "loss": 0.5969,
+      "step": 1048
+    },
+    {
+      "epoch": 1.622342481638964,
+      "grad_norm": 0.5768334865570068,
+      "learning_rate": 1.019295302013423e-05,
+      "loss": 0.7561,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6238886741399305,
+      "grad_norm": 0.6801527142524719,
+      "learning_rate": 1.0151006711409397e-05,
+      "loss": 0.6649,
+      "step": 1050
+    },
+    {
+      "epoch": 1.6254348666408966,
+      "grad_norm": 0.6493854522705078,
+      "learning_rate": 1.0109060402684564e-05,
+      "loss": 0.6788,
+      "step": 1051
+    },
+    {
+      "epoch": 1.6269810591418632,
+      "grad_norm": 0.6070857644081116,
+      "learning_rate": 1.006711409395973e-05,
+      "loss": 0.6509,
+      "step": 1052
+    },
+    {
+      "epoch": 1.6285272516428295,
+      "grad_norm": 0.6425489783287048,
+      "learning_rate": 1.00251677852349e-05,
+      "loss": 0.7303,
+      "step": 1053
+    },
+    {
+      "epoch": 1.6300734441437958,
+      "grad_norm": 0.6311664581298828,
+      "learning_rate": 9.983221476510067e-06,
+      "loss": 0.6492,
+      "step": 1054
+    },
+    {
+      "epoch": 1.6316196366447624,
+      "grad_norm": 0.5983323454856873,
+      "learning_rate": 9.941275167785235e-06,
+      "loss": 0.7778,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6331658291457285,
+      "grad_norm": 0.5775598287582397,
+      "learning_rate": 9.899328859060402e-06,
+      "loss": 0.7594,
+      "step": 1056
+    },
+    {
+      "epoch": 1.634712021646695,
+      "grad_norm": 0.6420764327049255,
+      "learning_rate": 9.857382550335572e-06,
+      "loss": 0.7812,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6362582141476614,
+      "grad_norm": 0.6438013911247253,
+      "learning_rate": 9.815436241610738e-06,
+      "loss": 0.7245,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6378044066486277,
+      "grad_norm": 0.6310963034629822,
+      "learning_rate": 9.773489932885907e-06,
+      "loss": 0.7321,
+      "step": 1059
+    },
+    {
+      "epoch": 1.6393505991495942,
+      "grad_norm": 0.6203198432922363,
+      "learning_rate": 9.731543624161075e-06,
+      "loss": 0.6914,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6408967916505603,
+      "grad_norm": 0.638454258441925,
+      "learning_rate": 9.689597315436243e-06,
+      "loss": 0.7478,
+      "step": 1061
+    },
+    {
+      "epoch": 1.642442984151527,
+      "grad_norm": 0.6612226963043213,
+      "learning_rate": 9.64765100671141e-06,
+      "loss": 0.7209,
+      "step": 1062
+    },
+    {
+      "epoch": 1.6439891766524932,
+      "grad_norm": 0.6566651463508606,
+      "learning_rate": 9.605704697986576e-06,
+      "loss": 0.8045,
+      "step": 1063
+    },
+    {
+      "epoch": 1.6455353691534595,
+      "grad_norm": 0.6543892621994019,
+      "learning_rate": 9.563758389261746e-06,
+      "loss": 0.7727,
+      "step": 1064
+    },
+    {
+      "epoch": 1.647081561654426,
+      "grad_norm": 0.6707619428634644,
+      "learning_rate": 9.521812080536913e-06,
+      "loss": 0.7271,
+      "step": 1065
+    },
+    {
+      "epoch": 1.6486277541553922,
+      "grad_norm": 0.6582595705986023,
+      "learning_rate": 9.479865771812081e-06,
+      "loss": 0.7332,
+      "step": 1066
+    },
+    {
+      "epoch": 1.6501739466563587,
+      "grad_norm": 0.6757826209068298,
+      "learning_rate": 9.437919463087248e-06,
+      "loss": 0.7663,
+      "step": 1067
+    },
+    {
+      "epoch": 1.651720139157325,
+      "grad_norm": 0.698391318321228,
+      "learning_rate": 9.395973154362418e-06,
+      "loss": 0.7277,
+      "step": 1068
+    },
+    {
+      "epoch": 1.6532663316582914,
+      "grad_norm": 0.6885313987731934,
+      "learning_rate": 9.354026845637584e-06,
+      "loss": 0.7959,
+      "step": 1069
+    },
+    {
+      "epoch": 1.654812524159258,
+      "grad_norm": 0.664113461971283,
+      "learning_rate": 9.312080536912752e-06,
+      "loss": 0.7571,
+      "step": 1070
+    },
+    {
+      "epoch": 1.656358716660224,
+      "grad_norm": 0.6730982065200806,
+      "learning_rate": 9.270134228187919e-06,
+      "loss": 0.709,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6579049091611906,
+      "grad_norm": 0.684267520904541,
+      "learning_rate": 9.228187919463089e-06,
+      "loss": 0.6932,
+      "step": 1072
+    },
+    {
+      "epoch": 1.659451101662157,
+      "grad_norm": 0.631272554397583,
+      "learning_rate": 9.186241610738255e-06,
+      "loss": 0.8569,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6609972941631233,
+      "grad_norm": 0.7173067927360535,
+      "learning_rate": 9.144295302013424e-06,
+      "loss": 0.777,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6625434866640898,
+      "grad_norm": 0.6663726568222046,
+      "learning_rate": 9.10234899328859e-06,
+      "loss": 0.7523,
+      "step": 1075
+    },
+    {
+      "epoch": 1.664089679165056,
+      "grad_norm": 0.689304530620575,
+      "learning_rate": 9.060402684563759e-06,
+      "loss": 0.7922,
+      "step": 1076
+    },
+    {
+      "epoch": 1.6656358716660224,
+      "grad_norm": 0.6620500683784485,
+      "learning_rate": 9.018456375838927e-06,
+      "loss": 0.8587,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6671820641669888,
+      "grad_norm": 0.6905922889709473,
+      "learning_rate": 8.976510067114093e-06,
+      "loss": 0.7897,
+      "step": 1078
+    },
+    {
+      "epoch": 1.668728256667955,
+      "grad_norm": 0.6763781309127808,
+      "learning_rate": 8.934563758389262e-06,
+      "loss": 0.7886,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6702744491689216,
+      "grad_norm": 0.684766948223114,
+      "learning_rate": 8.89261744966443e-06,
+      "loss": 0.8122,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6718206416698878,
+      "grad_norm": 0.6934885382652283,
+      "learning_rate": 8.850671140939598e-06,
+      "loss": 0.7894,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6733668341708543,
+      "grad_norm": 0.7051049470901489,
+      "learning_rate": 8.808724832214765e-06,
+      "loss": 0.8047,
+      "step": 1082
+    },
+    {
+      "epoch": 1.6749130266718206,
+      "grad_norm": 0.7332410216331482,
+      "learning_rate": 8.766778523489935e-06,
+      "loss": 0.7371,
+      "step": 1083
+    },
+    {
+      "epoch": 1.676459219172787,
+      "grad_norm": 0.7231751084327698,
+      "learning_rate": 8.724832214765101e-06,
+      "loss": 0.7804,
+      "step": 1084
+    },
+    {
+      "epoch": 1.6780054116737535,
+      "grad_norm": 0.7421995997428894,
+      "learning_rate": 8.68288590604027e-06,
+      "loss": 0.8364,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6795516041747196,
+      "grad_norm": 0.7182170152664185,
+      "learning_rate": 8.640939597315436e-06,
+      "loss": 0.848,
+      "step": 1086
+    },
+    {
+      "epoch": 1.6810977966756862,
+      "grad_norm": 0.7546189427375793,
+      "learning_rate": 8.598993288590604e-06,
+      "loss": 0.8336,
+      "step": 1087
+    },
+    {
+      "epoch": 1.6826439891766525,
+      "grad_norm": 0.7409399747848511,
+      "learning_rate": 8.557046979865773e-06,
+      "loss": 0.8151,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6841901816776188,
+      "grad_norm": 0.7315151691436768,
+      "learning_rate": 8.515100671140939e-06,
+      "loss": 0.8126,
+      "step": 1089
+    },
+    {
+      "epoch": 1.6857363741785854,
+      "grad_norm": 0.7949945330619812,
+      "learning_rate": 8.473154362416107e-06,
+      "loss": 0.8533,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6872825666795515,
+      "grad_norm": 0.7690507769584656,
+      "learning_rate": 8.431208053691276e-06,
+      "loss": 0.846,
+      "step": 1091
+    },
+    {
+      "epoch": 1.688828759180518,
+      "grad_norm": 0.7926616072654724,
+      "learning_rate": 8.389261744966444e-06,
+      "loss": 0.8402,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6903749516814843,
+      "grad_norm": 0.8056994676589966,
+      "learning_rate": 8.34731543624161e-06,
+      "loss": 0.8191,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6919211441824507,
+      "grad_norm": 0.8308284282684326,
+      "learning_rate": 8.305369127516779e-06,
+      "loss": 0.9042,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6934673366834172,
+      "grad_norm": 0.9019404053688049,
+      "learning_rate": 8.263422818791947e-06,
+      "loss": 0.8962,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6950135291843833,
+      "grad_norm": 1.144425630569458,
+      "learning_rate": 8.221476510067115e-06,
+      "loss": 1.0569,
+      "step": 1096
+    },
+    {
+      "epoch": 1.6965597216853499,
+      "grad_norm": 0.7823913097381592,
+      "learning_rate": 8.179530201342282e-06,
+      "loss": 0.5956,
+      "step": 1097
+    },
+    {
+      "epoch": 1.6981059141863162,
+      "grad_norm": 0.5778486132621765,
+      "learning_rate": 8.13758389261745e-06,
+      "loss": 0.6246,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6996521066872825,
+      "grad_norm": 0.6111435890197754,
+      "learning_rate": 8.095637583892618e-06,
+      "loss": 0.6681,
+      "step": 1099
+    },
+    {
+      "epoch": 1.701198299188249,
+      "grad_norm": 0.6174157857894897,
+      "learning_rate": 8.053691275167785e-06,
+      "loss": 0.6635,
+      "step": 1100
+    },
+    {
+      "epoch": 1.7027444916892152,
+      "grad_norm": 0.660089910030365,
+      "learning_rate": 8.011744966442953e-06,
+      "loss": 0.6994,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7042906841901817,
+      "grad_norm": 0.6478524208068848,
+      "learning_rate": 7.969798657718121e-06,
+      "loss": 0.6157,
+      "step": 1102
+    },
+    {
+      "epoch": 1.705836876691148,
+      "grad_norm": 0.6385321617126465,
+      "learning_rate": 7.92785234899329e-06,
+      "loss": 0.6967,
+      "step": 1103
+    },
+    {
+      "epoch": 1.7073830691921144,
+      "grad_norm": 0.6223974227905273,
+      "learning_rate": 7.885906040268456e-06,
+      "loss": 0.7572,
+      "step": 1104
+    },
+    {
+      "epoch": 1.708929261693081,
+      "grad_norm": 0.6408361792564392,
+      "learning_rate": 7.843959731543624e-06,
+      "loss": 0.8266,
+      "step": 1105
+    },
+    {
+      "epoch": 1.710475454194047,
+      "grad_norm": 0.6401566863059998,
+      "learning_rate": 7.802013422818793e-06,
+      "loss": 0.7073,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7120216466950136,
+      "grad_norm": 0.6204596757888794,
+      "learning_rate": 7.760067114093961e-06,
+      "loss": 0.6831,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7135678391959799,
+      "grad_norm": 0.6418159604072571,
+      "learning_rate": 7.718120805369127e-06,
+      "loss": 0.6925,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7151140316969462,
+      "grad_norm": 0.6294611692428589,
+      "learning_rate": 7.676174496644296e-06,
+      "loss": 0.7594,
+      "step": 1109
+    },
+    {
+      "epoch": 1.7166602241979128,
+      "grad_norm": 0.6831735968589783,
+      "learning_rate": 7.634228187919464e-06,
+      "loss": 0.6813,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7182064166988789,
+      "grad_norm": 0.6496366262435913,
+      "learning_rate": 7.5922818791946305e-06,
+      "loss": 0.7343,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7197526091998454,
+      "grad_norm": 0.6550260782241821,
+      "learning_rate": 7.5503355704698e-06,
+      "loss": 0.7389,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7212988017008117,
+      "grad_norm": 0.6319887638092041,
+      "learning_rate": 7.508389261744966e-06,
+      "loss": 0.7011,
+      "step": 1113
+    },
+    {
+      "epoch": 1.722844994201778,
+      "grad_norm": 0.6578625440597534,
+      "learning_rate": 7.466442953020135e-06,
+      "loss": 0.7221,
+      "step": 1114
+    },
+    {
+      "epoch": 1.7243911867027446,
+      "grad_norm": 0.6553155779838562,
+      "learning_rate": 7.424496644295302e-06,
+      "loss": 0.7053,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7259373792037107,
+      "grad_norm": 0.6502689719200134,
+      "learning_rate": 7.382550335570471e-06,
+      "loss": 0.8073,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7274835717046773,
+      "grad_norm": 0.6422849893569946,
+      "learning_rate": 7.3406040268456375e-06,
+      "loss": 0.6863,
+      "step": 1117
+    },
+    {
+      "epoch": 1.7290297642056436,
+      "grad_norm": 0.6866220831871033,
+      "learning_rate": 7.298657718120807e-06,
+      "loss": 0.7959,
+      "step": 1118
+    },
+    {
+      "epoch": 1.73057595670661,
+      "grad_norm": 0.645169198513031,
+      "learning_rate": 7.256711409395973e-06,
+      "loss": 0.7729,
+      "step": 1119
+    },
+    {
+      "epoch": 1.7321221492075765,
+      "grad_norm": 0.654244601726532,
+      "learning_rate": 7.214765100671142e-06,
+      "loss": 0.7707,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7336683417085426,
+      "grad_norm": 0.6695640087127686,
+      "learning_rate": 7.172818791946309e-06,
+      "loss": 0.714,
+      "step": 1121
+    },
+    {
+      "epoch": 1.735214534209509,
+      "grad_norm": 0.6387913227081299,
+      "learning_rate": 7.130872483221476e-06,
+      "loss": 0.7654,
+      "step": 1122
+    },
+    {
+      "epoch": 1.7367607267104754,
+      "grad_norm": 0.664368748664856,
+      "learning_rate": 7.0889261744966445e-06,
+      "loss": 0.7525,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7383069192114418,
+      "grad_norm": 0.7358962893486023,
+      "learning_rate": 7.046979865771812e-06,
+      "loss": 0.747,
+      "step": 1124
+    },
+    {
+      "epoch": 1.7398531117124083,
+      "grad_norm": 0.6654751896858215,
+      "learning_rate": 7.00503355704698e-06,
+      "loss": 0.7669,
+      "step": 1125
+    },
+    {
+      "epoch": 1.7413993042133744,
+      "grad_norm": 0.6763463020324707,
+      "learning_rate": 6.963087248322148e-06,
+      "loss": 0.8077,
+      "step": 1126
+    },
+    {
+      "epoch": 1.742945496714341,
+      "grad_norm": 0.6764401197433472,
+      "learning_rate": 6.921140939597316e-06,
+      "loss": 0.7438,
+      "step": 1127
+    },
+    {
+      "epoch": 1.7444916892153073,
+      "grad_norm": 0.7182475924491882,
+      "learning_rate": 6.879194630872483e-06,
+      "loss": 0.728,
+      "step": 1128
+    },
+    {
+      "epoch": 1.7460378817162736,
+      "grad_norm": 0.6785516142845154,
+      "learning_rate": 6.8372483221476515e-06,
+      "loss": 0.7939,
+      "step": 1129
+    },
+    {
+      "epoch": 1.7475840742172402,
+      "grad_norm": 0.7003461122512817,
+      "learning_rate": 6.795302013422819e-06,
+      "loss": 0.7729,
+      "step": 1130
+    },
+    {
+      "epoch": 1.7491302667182063,
+      "grad_norm": 0.6882821917533875,
+      "learning_rate": 6.753355704697987e-06,
+      "loss": 0.9204,
+      "step": 1131
+    },
+    {
+      "epoch": 1.7506764592191728,
+      "grad_norm": 0.7342216372489929,
+      "learning_rate": 6.7114093959731546e-06,
+      "loss": 0.7385,
+      "step": 1132
+    },
+    {
+      "epoch": 1.7522226517201391,
+      "grad_norm": 0.7073139548301697,
+      "learning_rate": 6.669463087248322e-06,
+      "loss": 0.9268,
+      "step": 1133
+    },
+    {
+      "epoch": 1.7537688442211055,
+      "grad_norm": 0.7231860756874084,
+      "learning_rate": 6.62751677852349e-06,
+      "loss": 0.8596,
+      "step": 1134
+    },
+    {
+      "epoch": 1.755315036722072,
+      "grad_norm": 0.7145645618438721,
+      "learning_rate": 6.585570469798658e-06,
+      "loss": 0.8466,
+      "step": 1135
+    },
+    {
+      "epoch": 1.7568612292230381,
+      "grad_norm": 0.7454160451889038,
+      "learning_rate": 6.543624161073826e-06,
+      "loss": 0.7629,
+      "step": 1136
+    },
+    {
+      "epoch": 1.7584074217240047,
+      "grad_norm": 0.7323503494262695,
+      "learning_rate": 6.501677852348993e-06,
+      "loss": 0.8666,
+      "step": 1137
+    },
+    {
+      "epoch": 1.759953614224971,
+      "grad_norm": 0.7377263307571411,
+      "learning_rate": 6.4597315436241616e-06,
+      "loss": 0.8515,
+      "step": 1138
+    },
+    {
+      "epoch": 1.7614998067259373,
+      "grad_norm": 0.7750667929649353,
+      "learning_rate": 6.417785234899329e-06,
+      "loss": 0.8033,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7630459992269039,
+      "grad_norm": 0.7829360365867615,
+      "learning_rate": 6.375838926174497e-06,
+      "loss": 0.8903,
+      "step": 1140
+    },
+    {
+      "epoch": 1.76459219172787,
+      "grad_norm": 0.7346689701080322,
+      "learning_rate": 6.333892617449665e-06,
+      "loss": 0.9772,
+      "step": 1141
+    },
+    {
+      "epoch": 1.7661383842288365,
+      "grad_norm": 0.7309790253639221,
+      "learning_rate": 6.291946308724833e-06,
+      "loss": 0.8317,
+      "step": 1142
+    },
+    {
+      "epoch": 1.7676845767298028,
+      "grad_norm": 0.8332825899124146,
+      "learning_rate": 6.25e-06,
+      "loss": 0.7692,
+      "step": 1143
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.8708672523498535,
+      "learning_rate": 6.2080536912751686e-06,
+      "loss": 0.946,
+      "step": 1144
+    },
+    {
+      "epoch": 1.7707769617317357,
+      "grad_norm": 0.9887493848800659,
+      "learning_rate": 6.166107382550336e-06,
+      "loss": 0.9102,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7723231542327018,
+      "grad_norm": 1.3964698314666748,
+      "learning_rate": 6.124161073825504e-06,
+      "loss": 0.8942,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7738693467336684,
+      "grad_norm": 0.9236196875572205,
+      "learning_rate": 6.082214765100672e-06,
+      "loss": 0.6239,
+      "step": 1147
+    },
+    {
+      "epoch": 1.7754155392346347,
+      "grad_norm": 0.5753183960914612,
+      "learning_rate": 6.04026845637584e-06,
+      "loss": 0.6674,
+      "step": 1148
+    },
+    {
+      "epoch": 1.776961731735601,
+      "grad_norm": 0.6132087111473083,
+      "learning_rate": 5.9983221476510064e-06,
+      "loss": 0.7186,
+      "step": 1149
+    },
+    {
+      "epoch": 1.7785079242365676,
+      "grad_norm": 0.586824893951416,
+      "learning_rate": 5.956375838926175e-06,
+      "loss": 0.7077,
+      "step": 1150
+    },
+    {
+      "epoch": 1.7800541167375337,
+      "grad_norm": 0.6105204224586487,
+      "learning_rate": 5.914429530201342e-06,
+      "loss": 0.7448,
+      "step": 1151
+    },
+    {
+      "epoch": 1.7816003092385002,
+      "grad_norm": 0.6133350133895874,
+      "learning_rate": 5.87248322147651e-06,
+      "loss": 0.7761,
+      "step": 1152
+    },
+    {
+      "epoch": 1.7831465017394665,
+      "grad_norm": 0.6162737011909485,
+      "learning_rate": 5.830536912751678e-06,
+      "loss": 0.7204,
+      "step": 1153
+    },
+    {
+      "epoch": 1.7846926942404329,
+      "grad_norm": 0.6450937390327454,
+      "learning_rate": 5.788590604026846e-06,
+      "loss": 0.7163,
+      "step": 1154
+    },
+    {
+      "epoch": 1.7862388867413994,
+      "grad_norm": 0.6387403011322021,
+      "learning_rate": 5.7466442953020134e-06,
+      "loss": 0.7454,
+      "step": 1155
+    },
+    {
+      "epoch": 1.7877850792423655,
+      "grad_norm": 0.6444848775863647,
+      "learning_rate": 5.704697986577182e-06,
+      "loss": 0.7048,
+      "step": 1156
+    },
+    {
+      "epoch": 1.789331271743332,
+      "grad_norm": 0.6471322774887085,
+      "learning_rate": 5.662751677852349e-06,
+      "loss": 0.75,
+      "step": 1157
+    },
+    {
+      "epoch": 1.7908774642442984,
+      "grad_norm": 0.6798081398010254,
+      "learning_rate": 5.620805369127517e-06,
+      "loss": 0.7096,
+      "step": 1158
+    },
+    {
+      "epoch": 1.7924236567452647,
+      "grad_norm": 0.6646496653556824,
+      "learning_rate": 5.578859060402685e-06,
+      "loss": 0.8017,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7939698492462313,
+      "grad_norm": 0.6455252170562744,
+      "learning_rate": 5.536912751677853e-06,
+      "loss": 0.7323,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7955160417471974,
+      "grad_norm": 0.6841678023338318,
+      "learning_rate": 5.49496644295302e-06,
+      "loss": 0.7373,
+      "step": 1161
+    },
+    {
+      "epoch": 1.797062234248164,
+      "grad_norm": 0.6608906388282776,
+      "learning_rate": 5.453020134228188e-06,
+      "loss": 0.7039,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7986084267491302,
+      "grad_norm": 0.6611918807029724,
+      "learning_rate": 5.411073825503356e-06,
+      "loss": 0.768,
+      "step": 1163
+    },
+    {
+      "epoch": 1.8001546192500966,
+      "grad_norm": 0.6479055881500244,
+      "learning_rate": 5.3691275167785235e-06,
+      "loss": 0.7672,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8017008117510631,
+      "grad_norm": 0.6645311117172241,
+      "learning_rate": 5.327181208053692e-06,
+      "loss": 0.8146,
+      "step": 1165
+    },
+    {
+      "epoch": 1.8032470042520292,
+      "grad_norm": 0.6700178980827332,
+      "learning_rate": 5.285234899328859e-06,
+      "loss": 0.7019,
+      "step": 1166
+    },
+    {
+      "epoch": 1.8047931967529958,
+      "grad_norm": 0.6892951130867004,
+      "learning_rate": 5.243288590604027e-06,
+      "loss": 0.7982,
+      "step": 1167
+    },
+    {
+      "epoch": 1.806339389253962,
+      "grad_norm": 0.6578270196914673,
+      "learning_rate": 5.201342281879195e-06,
+      "loss": 0.7844,
+      "step": 1168
+    },
+    {
+      "epoch": 1.8078855817549284,
+      "grad_norm": 0.6844699382781982,
+      "learning_rate": 5.159395973154363e-06,
+      "loss": 0.73,
+      "step": 1169
+    },
+    {
+      "epoch": 1.809431774255895,
+      "grad_norm": 0.67879718542099,
+      "learning_rate": 5.1174496644295305e-06,
+      "loss": 0.8256,
+      "step": 1170
+    },
+    {
+      "epoch": 1.810977966756861,
+      "grad_norm": 0.6748877167701721,
+      "learning_rate": 5.075503355704699e-06,
+      "loss": 0.7733,
+      "step": 1171
+    },
+    {
+      "epoch": 1.8125241592578276,
+      "grad_norm": 0.6248130798339844,
+      "learning_rate": 5.033557046979865e-06,
+      "loss": 0.7489,
+      "step": 1172
+    },
+    {
+      "epoch": 1.814070351758794,
+      "grad_norm": 0.6643280386924744,
+      "learning_rate": 4.9916107382550336e-06,
+      "loss": 0.726,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8156165442597603,
+      "grad_norm": 0.6891630291938782,
+      "learning_rate": 4.949664429530201e-06,
+      "loss": 0.7442,
+      "step": 1174
+    },
+    {
+      "epoch": 1.8171627367607268,
+      "grad_norm": 0.6768823862075806,
+      "learning_rate": 4.907718120805369e-06,
+      "loss": 0.7376,
+      "step": 1175
+    },
+    {
+      "epoch": 1.818708929261693,
+      "grad_norm": 0.669154167175293,
+      "learning_rate": 4.8657718120805375e-06,
+      "loss": 0.7775,
+      "step": 1176
+    },
+    {
+      "epoch": 1.8202551217626595,
+      "grad_norm": 0.7224996089935303,
+      "learning_rate": 4.823825503355705e-06,
+      "loss": 0.7535,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8218013142636258,
+      "grad_norm": 0.6954566240310669,
+      "learning_rate": 4.781879194630873e-06,
+      "loss": 0.77,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8233475067645921,
+      "grad_norm": 0.6938985586166382,
+      "learning_rate": 4.7399328859060405e-06,
+      "loss": 0.7963,
+      "step": 1179
+    },
+    {
+      "epoch": 1.8248936992655587,
+      "grad_norm": 0.6518607139587402,
+      "learning_rate": 4.697986577181209e-06,
+      "loss": 0.7652,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8264398917665248,
+      "grad_norm": 0.7072224617004395,
+      "learning_rate": 4.656040268456376e-06,
+      "loss": 0.7875,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8279860842674913,
+      "grad_norm": 0.689007580280304,
+      "learning_rate": 4.6140939597315445e-06,
+      "loss": 0.8659,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8295322767684576,
+      "grad_norm": 0.7117170095443726,
+      "learning_rate": 4.572147651006712e-06,
+      "loss": 0.7761,
+      "step": 1183
+    },
+    {
+      "epoch": 1.831078469269424,
+      "grad_norm": 0.6940242052078247,
+      "learning_rate": 4.530201342281879e-06,
+      "loss": 0.8305,
+      "step": 1184
+    },
+    {
+      "epoch": 1.8326246617703905,
+      "grad_norm": 0.7297004461288452,
+      "learning_rate": 4.488255033557047e-06,
+      "loss": 0.8451,
+      "step": 1185
+    },
+    {
+      "epoch": 1.8341708542713566,
+      "grad_norm": 0.7414118647575378,
+      "learning_rate": 4.446308724832215e-06,
+      "loss": 0.7932,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8357170467723232,
+      "grad_norm": 0.7365830540657043,
+      "learning_rate": 4.404362416107382e-06,
+      "loss": 0.8153,
+      "step": 1187
+    },
+    {
+      "epoch": 1.8372632392732895,
+      "grad_norm": 0.7378683686256409,
+      "learning_rate": 4.362416107382551e-06,
+      "loss": 0.8795,
+      "step": 1188
+    },
+    {
+      "epoch": 1.8388094317742558,
+      "grad_norm": 0.7360830307006836,
+      "learning_rate": 4.320469798657718e-06,
+      "loss": 0.863,
+      "step": 1189
+    },
+    {
+      "epoch": 1.8403556242752224,
+      "grad_norm": 0.7365934252738953,
+      "learning_rate": 4.278523489932886e-06,
+      "loss": 0.8836,
+      "step": 1190
+    },
+    {
+      "epoch": 1.8419018167761885,
+      "grad_norm": 0.7860958576202393,
+      "learning_rate": 4.236577181208054e-06,
+      "loss": 0.8834,
+      "step": 1191
+    },
+    {
+      "epoch": 1.843448009277155,
+      "grad_norm": 0.8424365520477295,
+      "learning_rate": 4.194630872483222e-06,
+      "loss": 0.8303,
+      "step": 1192
+    },
+    {
+      "epoch": 1.8449942017781213,
+      "grad_norm": 0.8641267418861389,
+      "learning_rate": 4.152684563758389e-06,
+      "loss": 0.7928,
+      "step": 1193
+    },
+    {
+      "epoch": 1.8465403942790877,
+      "grad_norm": 0.8995117545127869,
+      "learning_rate": 4.110738255033558e-06,
+      "loss": 0.8474,
+      "step": 1194
+    },
+    {
+      "epoch": 1.8480865867800542,
+      "grad_norm": 0.9355427026748657,
+      "learning_rate": 4.068791946308725e-06,
+      "loss": 0.8516,
+      "step": 1195
+    },
+    {
+      "epoch": 1.8496327792810203,
+      "grad_norm": 1.3238755464553833,
+      "learning_rate": 4.026845637583892e-06,
+      "loss": 1.0038,
+      "step": 1196
+    },
+    {
+      "epoch": 1.8511789717819869,
+      "grad_norm": 0.8634078502655029,
+      "learning_rate": 3.984899328859061e-06,
+      "loss": 0.6067,
+      "step": 1197
+    },
+    {
+      "epoch": 1.8527251642829532,
+      "grad_norm": 0.5876993536949158,
+      "learning_rate": 3.942953020134228e-06,
+      "loss": 0.6623,
+      "step": 1198
+    },
+    {
+      "epoch": 1.8542713567839195,
+      "grad_norm": 0.6208466291427612,
+      "learning_rate": 3.901006711409396e-06,
+      "loss": 0.6686,
+      "step": 1199
+    },
+    {
+      "epoch": 1.855817549284886,
+      "grad_norm": 0.5899839401245117,
+      "learning_rate": 3.859060402684564e-06,
+      "loss": 0.7108,
+      "step": 1200
+    },
+    {
+      "epoch": 1.8573637417858522,
+      "grad_norm": 0.6139258742332458,
+      "learning_rate": 3.817114093959732e-06,
+      "loss": 0.6853,
+      "step": 1201
+    },
+    {
+      "epoch": 1.8589099342868187,
+      "grad_norm": 0.6186558604240417,
+      "learning_rate": 3.7751677852349e-06,
+      "loss": 0.6619,
+      "step": 1202
+    },
+    {
+      "epoch": 1.860456126787785,
+      "grad_norm": 0.6083664298057556,
+      "learning_rate": 3.7332214765100677e-06,
+      "loss": 0.7119,
+      "step": 1203
+    },
+    {
+      "epoch": 1.8620023192887514,
+      "grad_norm": 0.6227363348007202,
+      "learning_rate": 3.6912751677852355e-06,
+      "loss": 0.7251,
+      "step": 1204
+    },
+    {
+      "epoch": 1.863548511789718,
+      "grad_norm": 0.6329985857009888,
+      "learning_rate": 3.6493288590604033e-06,
+      "loss": 0.7796,
+      "step": 1205
+    },
+    {
+      "epoch": 1.865094704290684,
+      "grad_norm": 0.644888162612915,
+      "learning_rate": 3.607382550335571e-06,
+      "loss": 0.7747,
+      "step": 1206
+    },
+    {
+      "epoch": 1.8666408967916506,
+      "grad_norm": 0.6022448539733887,
+      "learning_rate": 3.565436241610738e-06,
+      "loss": 0.7027,
+      "step": 1207
+    },
+    {
+      "epoch": 1.868187089292617,
+      "grad_norm": 0.6584210991859436,
+      "learning_rate": 3.523489932885906e-06,
+      "loss": 0.6816,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8697332817935832,
+      "grad_norm": 0.6227755546569824,
+      "learning_rate": 3.481543624161074e-06,
+      "loss": 0.7245,
+      "step": 1209
+    },
+    {
+      "epoch": 1.8712794742945498,
+      "grad_norm": 0.6115254759788513,
+      "learning_rate": 3.4395973154362416e-06,
+      "loss": 0.7511,
+      "step": 1210
+    },
+    {
+      "epoch": 1.8728256667955159,
+      "grad_norm": 0.638425886631012,
+      "learning_rate": 3.3976510067114095e-06,
+      "loss": 0.742,
+      "step": 1211
+    },
+    {
+      "epoch": 1.8743718592964824,
+      "grad_norm": 0.654451310634613,
+      "learning_rate": 3.3557046979865773e-06,
+      "loss": 0.7323,
+      "step": 1212
+    },
+    {
+      "epoch": 1.8759180517974487,
+      "grad_norm": 0.6516563892364502,
+      "learning_rate": 3.313758389261745e-06,
+      "loss": 0.6654,
+      "step": 1213
+    },
+    {
+      "epoch": 1.877464244298415,
+      "grad_norm": 0.6713635325431824,
+      "learning_rate": 3.271812080536913e-06,
+      "loss": 0.7639,
+      "step": 1214
+    },
+    {
+      "epoch": 1.8790104367993816,
+      "grad_norm": 0.6944072246551514,
+      "learning_rate": 3.2298657718120808e-06,
+      "loss": 0.7274,
+      "step": 1215
+    },
+    {
+      "epoch": 1.880556629300348,
+      "grad_norm": 0.6565424799919128,
+      "learning_rate": 3.1879194630872486e-06,
+      "loss": 0.7757,
+      "step": 1216
+    },
+    {
+      "epoch": 1.8821028218013143,
+      "grad_norm": 0.6998873949050903,
+      "learning_rate": 3.1459731543624164e-06,
+      "loss": 0.8059,
+      "step": 1217
+    },
+    {
+      "epoch": 1.8836490143022806,
+      "grad_norm": 0.6804831624031067,
+      "learning_rate": 3.1040268456375843e-06,
+      "loss": 0.7777,
+      "step": 1218
+    },
+    {
+      "epoch": 1.885195206803247,
+      "grad_norm": 0.6547567248344421,
+      "learning_rate": 3.062080536912752e-06,
+      "loss": 0.746,
+      "step": 1219
+    },
+    {
+      "epoch": 1.8867413993042135,
+      "grad_norm": 0.6822516322135925,
+      "learning_rate": 3.02013422818792e-06,
+      "loss": 0.7518,
+      "step": 1220
+    },
+    {
+      "epoch": 1.8882875918051798,
+      "grad_norm": 0.6750561594963074,
+      "learning_rate": 2.9781879194630873e-06,
+      "loss": 0.7261,
+      "step": 1221
+    },
+    {
+      "epoch": 1.8898337843061461,
+      "grad_norm": 0.6937829256057739,
+      "learning_rate": 2.936241610738255e-06,
+      "loss": 0.7716,
+      "step": 1222
+    },
+    {
+      "epoch": 1.8913799768071125,
+      "grad_norm": 0.6683257818222046,
+      "learning_rate": 2.894295302013423e-06,
+      "loss": 0.7657,
+      "step": 1223
+    },
+    {
+      "epoch": 1.8929261693080788,
+      "grad_norm": 0.6856616735458374,
+      "learning_rate": 2.852348993288591e-06,
+      "loss": 0.7833,
+      "step": 1224
+    },
+    {
+      "epoch": 1.8944723618090453,
+      "grad_norm": 0.719022274017334,
+      "learning_rate": 2.8104026845637587e-06,
+      "loss": 0.7462,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8960185543100117,
+      "grad_norm": 0.6999730467796326,
+      "learning_rate": 2.7684563758389265e-06,
+      "loss": 0.7777,
+      "step": 1226
+    },
+    {
+      "epoch": 1.897564746810978,
+      "grad_norm": 0.710363507270813,
+      "learning_rate": 2.726510067114094e-06,
+      "loss": 0.7624,
+      "step": 1227
+    },
+    {
+      "epoch": 1.8991109393119443,
+      "grad_norm": 0.6629136204719543,
+      "learning_rate": 2.6845637583892617e-06,
+      "loss": 0.7578,
+      "step": 1228
+    },
+    {
+      "epoch": 1.9006571318129106,
+      "grad_norm": 0.6639029383659363,
+      "learning_rate": 2.6426174496644296e-06,
+      "loss": 0.8438,
+      "step": 1229
+    },
+    {
+      "epoch": 1.9022033243138772,
+      "grad_norm": 0.6924868226051331,
+      "learning_rate": 2.6006711409395974e-06,
+      "loss": 0.7844,
+      "step": 1230
+    },
+    {
+      "epoch": 1.9037495168148435,
+      "grad_norm": 0.7059327960014343,
+      "learning_rate": 2.5587248322147652e-06,
+      "loss": 0.7654,
+      "step": 1231
+    },
+    {
+      "epoch": 1.9052957093158098,
+      "grad_norm": 0.6732707619667053,
+      "learning_rate": 2.5167785234899326e-06,
+      "loss": 0.7158,
+      "step": 1232
+    },
+    {
+      "epoch": 1.9068419018167762,
+      "grad_norm": 0.712011456489563,
+      "learning_rate": 2.4748322147651005e-06,
+      "loss": 0.8323,
+      "step": 1233
+    },
+    {
+      "epoch": 1.9083880943177425,
+      "grad_norm": 0.7283968329429626,
+      "learning_rate": 2.4328859060402687e-06,
+      "loss": 0.86,
+      "step": 1234
+    },
+    {
+      "epoch": 1.909934286818709,
+      "grad_norm": 0.736113429069519,
+      "learning_rate": 2.3909395973154366e-06,
+      "loss": 0.8779,
+      "step": 1235
+    },
+    {
+      "epoch": 1.9114804793196754,
+      "grad_norm": 0.7328662276268005,
+      "learning_rate": 2.3489932885906044e-06,
+      "loss": 0.8673,
+      "step": 1236
+    },
+    {
+      "epoch": 1.9130266718206417,
+      "grad_norm": 0.7270045280456543,
+      "learning_rate": 2.3070469798657722e-06,
+      "loss": 0.8348,
+      "step": 1237
+    },
+    {
+      "epoch": 1.914572864321608,
+      "grad_norm": 0.7681441903114319,
+      "learning_rate": 2.2651006711409396e-06,
+      "loss": 0.818,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9161190568225743,
+      "grad_norm": 0.7744415998458862,
+      "learning_rate": 2.2231543624161075e-06,
+      "loss": 0.8879,
+      "step": 1239
+    },
+    {
+      "epoch": 1.9176652493235409,
+      "grad_norm": 0.7644962072372437,
+      "learning_rate": 2.1812080536912753e-06,
+      "loss": 0.8545,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9192114418245072,
+      "grad_norm": 0.7949373722076416,
+      "learning_rate": 2.139261744966443e-06,
+      "loss": 0.8603,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9207576343254735,
+      "grad_norm": 0.8180006146430969,
+      "learning_rate": 2.097315436241611e-06,
+      "loss": 0.8146,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9223038268264399,
+      "grad_norm": 0.8280307650566101,
+      "learning_rate": 2.055369127516779e-06,
+      "loss": 0.9628,
+      "step": 1243
+    },
+    {
+      "epoch": 1.9238500193274062,
+      "grad_norm": 0.8613501787185669,
+      "learning_rate": 2.013422818791946e-06,
+      "loss": 0.8326,
+      "step": 1244
+    },
+    {
+      "epoch": 1.9253962118283727,
+      "grad_norm": 0.9388262629508972,
+      "learning_rate": 1.971476510067114e-06,
+      "loss": 0.9428,
+      "step": 1245
+    },
+    {
+      "epoch": 1.926942404329339,
+      "grad_norm": 1.25608491897583,
+      "learning_rate": 1.929530201342282e-06,
+      "loss": 0.8971,
+      "step": 1246
+    },
+    {
+      "epoch": 1.9284885968303054,
+      "grad_norm": 0.8791813254356384,
+      "learning_rate": 1.88758389261745e-06,
+      "loss": 0.6034,
+      "step": 1247
+    },
+    {
+      "epoch": 1.9300347893312717,
+      "grad_norm": 0.5444540977478027,
+      "learning_rate": 1.8456375838926177e-06,
+      "loss": 0.6993,
+      "step": 1248
+    },
+    {
+      "epoch": 1.931580981832238,
+      "grad_norm": 0.6073552966117859,
+      "learning_rate": 1.8036912751677856e-06,
+      "loss": 0.7178,
+      "step": 1249
+    },
+    {
+      "epoch": 1.9331271743332046,
+      "grad_norm": 0.6113264560699463,
+      "learning_rate": 1.761744966442953e-06,
+      "loss": 0.6966,
+      "step": 1250
+    },
+    {
+      "epoch": 1.934673366834171,
+      "grad_norm": 0.5979108810424805,
+      "learning_rate": 1.7197986577181208e-06,
+      "loss": 0.7133,
+      "step": 1251
+    },
+    {
+      "epoch": 1.9362195593351372,
+      "grad_norm": 0.6315338611602783,
+      "learning_rate": 1.6778523489932886e-06,
+      "loss": 0.7721,
+      "step": 1252
+    },
+    {
+      "epoch": 1.9377657518361036,
+      "grad_norm": 0.6291902661323547,
+      "learning_rate": 1.6359060402684565e-06,
+      "loss": 0.7006,
+      "step": 1253
+    },
+    {
+      "epoch": 1.9393119443370699,
+      "grad_norm": 0.6238860487937927,
+      "learning_rate": 1.5939597315436243e-06,
+      "loss": 0.7787,
+      "step": 1254
+    },
+    {
+      "epoch": 1.9408581368380364,
+      "grad_norm": 0.6401642560958862,
+      "learning_rate": 1.5520134228187921e-06,
+      "loss": 0.7603,
+      "step": 1255
+    },
+    {
+      "epoch": 1.9424043293390028,
+      "grad_norm": 0.6174636483192444,
+      "learning_rate": 1.51006711409396e-06,
+      "loss": 0.7259,
+      "step": 1256
+    },
+    {
+      "epoch": 1.943950521839969,
+      "grad_norm": 0.6495472192764282,
+      "learning_rate": 1.4681208053691276e-06,
+      "loss": 0.6654,
+      "step": 1257
+    },
+    {
+      "epoch": 1.9454967143409354,
+      "grad_norm": 0.667812168598175,
+      "learning_rate": 1.4261744966442954e-06,
+      "loss": 0.7147,
+      "step": 1258
+    },
+    {
+      "epoch": 1.9470429068419017,
+      "grad_norm": 0.6613562703132629,
+      "learning_rate": 1.3842281879194633e-06,
+      "loss": 0.7437,
+      "step": 1259
+    },
+    {
+      "epoch": 1.9485890993428683,
+      "grad_norm": 0.7061516046524048,
+      "learning_rate": 1.3422818791946309e-06,
+      "loss": 0.6831,
+      "step": 1260
+    },
+    {
+      "epoch": 1.9501352918438346,
+      "grad_norm": 0.6739703416824341,
+      "learning_rate": 1.3003355704697987e-06,
+      "loss": 0.747,
+      "step": 1261
+    },
+    {
+      "epoch": 1.951681484344801,
+      "grad_norm": 0.6422568559646606,
+      "learning_rate": 1.2583892617449663e-06,
+      "loss": 0.7569,
+      "step": 1262
+    },
+    {
+      "epoch": 1.9532276768457673,
+      "grad_norm": 0.6601455211639404,
+      "learning_rate": 1.2164429530201344e-06,
+      "loss": 0.7461,
+      "step": 1263
+    },
+    {
+      "epoch": 1.9547738693467336,
+      "grad_norm": 0.6846182942390442,
+      "learning_rate": 1.1744966442953022e-06,
+      "loss": 0.7185,
+      "step": 1264
+    },
+    {
+      "epoch": 1.9563200618477001,
+      "grad_norm": 0.6746420860290527,
+      "learning_rate": 1.1325503355704698e-06,
+      "loss": 0.6806,
+      "step": 1265
+    },
+    {
+      "epoch": 1.9578662543486665,
+      "grad_norm": 0.6510487198829651,
+      "learning_rate": 1.0906040268456377e-06,
+      "loss": 0.7693,
+      "step": 1266
+    },
+    {
+      "epoch": 1.9594124468496328,
+      "grad_norm": 0.6752596497535706,
+      "learning_rate": 1.0486577181208055e-06,
+      "loss": 0.899,
+      "step": 1267
+    },
+    {
+      "epoch": 1.960958639350599,
+      "grad_norm": 0.698874294757843,
+      "learning_rate": 1.006711409395973e-06,
+      "loss": 0.7521,
+      "step": 1268
+    },
+    {
+      "epoch": 1.9625048318515654,
+      "grad_norm": 0.6726242899894714,
+      "learning_rate": 9.64765100671141e-07,
+      "loss": 0.7782,
+      "step": 1269
+    },
+    {
+      "epoch": 1.964051024352532,
+      "grad_norm": 0.6857203245162964,
+      "learning_rate": 9.228187919463089e-07,
+      "loss": 0.7344,
+      "step": 1270
+    },
+    {
+      "epoch": 1.9655972168534983,
+      "grad_norm": 0.7023079991340637,
+      "learning_rate": 8.808724832214765e-07,
+      "loss": 0.7533,
+      "step": 1271
+    },
+    {
+      "epoch": 1.9671434093544646,
+      "grad_norm": 0.6352121233940125,
+      "learning_rate": 8.389261744966443e-07,
+      "loss": 0.7554,
+      "step": 1272
+    },
+    {
+      "epoch": 1.968689601855431,
+      "grad_norm": 0.6913352608680725,
+      "learning_rate": 7.969798657718122e-07,
+      "loss": 0.7468,
+      "step": 1273
+    },
+    {
+      "epoch": 1.9702357943563973,
+      "grad_norm": 0.6950103640556335,
+      "learning_rate": 7.5503355704698e-07,
+      "loss": 0.8532,
+      "step": 1274
+    },
+    {
+      "epoch": 1.9717819868573638,
+      "grad_norm": 0.6684133410453796,
+      "learning_rate": 7.130872483221477e-07,
+      "loss": 0.8302,
+      "step": 1275
+    },
+    {
+      "epoch": 1.9733281793583302,
+      "grad_norm": 0.6860172152519226,
+      "learning_rate": 6.711409395973154e-07,
+      "loss": 0.7342,
+      "step": 1276
+    },
+    {
+      "epoch": 1.9748743718592965,
+      "grad_norm": 0.6672773957252502,
+      "learning_rate": 6.291946308724832e-07,
+      "loss": 0.7901,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9764205643602628,
+      "grad_norm": 0.7022169828414917,
+      "learning_rate": 5.872483221476511e-07,
+      "loss": 0.7796,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9779667568612291,
+      "grad_norm": 0.6778275370597839,
+      "learning_rate": 5.453020134228188e-07,
+      "loss": 0.8389,
+      "step": 1279
+    },
+    {
+      "epoch": 1.9795129493621957,
+      "grad_norm": 0.721585750579834,
+      "learning_rate": 5.033557046979866e-07,
+      "loss": 0.848,
+      "step": 1280
+    },
+    {
+      "epoch": 1.981059141863162,
+      "grad_norm": 0.6909308433532715,
+      "learning_rate": 4.6140939597315444e-07,
+      "loss": 0.8214,
+      "step": 1281
+    },
+    {
+      "epoch": 1.9826053343641283,
+      "grad_norm": 0.7320754528045654,
+      "learning_rate": 4.1946308724832216e-07,
+      "loss": 0.8179,
+      "step": 1282
+    },
+    {
+      "epoch": 1.9841515268650947,
+      "grad_norm": 0.7341198325157166,
+      "learning_rate": 3.7751677852349e-07,
+      "loss": 0.8212,
+      "step": 1283
+    },
+    {
+      "epoch": 1.985697719366061,
+      "grad_norm": 0.7166808843612671,
+      "learning_rate": 3.355704697986577e-07,
+      "loss": 0.8472,
+      "step": 1284
+    },
+    {
+      "epoch": 1.9872439118670275,
+      "grad_norm": 0.7207593321800232,
+      "learning_rate": 2.9362416107382555e-07,
+      "loss": 0.8408,
+      "step": 1285
+    },
+    {
+      "epoch": 1.9887901043679939,
+      "grad_norm": 0.7933880090713501,
+      "learning_rate": 2.516778523489933e-07,
+      "loss": 0.8781,
+      "step": 1286
+    },
+    {
+      "epoch": 1.9903362968689602,
+      "grad_norm": 0.7923394441604614,
+      "learning_rate": 2.0973154362416108e-07,
+      "loss": 0.9313,
+      "step": 1287
+    },
+    {
+      "epoch": 1.9918824893699265,
+      "grad_norm": 0.7550710439682007,
+      "learning_rate": 1.6778523489932886e-07,
+      "loss": 0.8893,
+      "step": 1288
+    },
+    {
+      "epoch": 1.9934286818708928,
+      "grad_norm": 0.8748595118522644,
+      "learning_rate": 1.2583892617449664e-07,
+      "loss": 0.9071,
+      "step": 1289
+    },
+    {
+      "epoch": 1.9949748743718594,
+      "grad_norm": 0.8392748236656189,
+      "learning_rate": 8.389261744966443e-08,
+      "loss": 0.9107,
+      "step": 1290
+    },
+    {
+      "epoch": 1.9965210668728257,
+      "grad_norm": 0.8647322058677673,
+      "learning_rate": 4.1946308724832215e-08,
+      "loss": 0.8547,
+      "step": 1291
+    },
+    {
+      "epoch": 1.998067259373792,
+      "grad_norm": 1.0320862531661987,
+      "learning_rate": 0.0,
+      "loss": 0.8732,
+      "step": 1292
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1292,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.697955930285742e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}