diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9046 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1284,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001557632398753894,
+      "grad_norm": 1.7232097279069922,
+      "learning_rate": 9.999985033870806e-06,
+      "loss": 0.6355,
+      "step": 1
+    },
+    {
+      "epoch": 0.003115264797507788,
+      "grad_norm": 0.7663661005911214,
+      "learning_rate": 9.999940135572817e-06,
+      "loss": 0.2544,
+      "step": 2
+    },
+    {
+      "epoch": 0.004672897196261682,
+      "grad_norm": 0.8839734168626316,
+      "learning_rate": 9.999865305374812e-06,
+      "loss": 0.3767,
+      "step": 3
+    },
+    {
+      "epoch": 0.006230529595015576,
+      "grad_norm": 0.8042784289646617,
+      "learning_rate": 9.999760543724761e-06,
+      "loss": 0.184,
+      "step": 4
+    },
+    {
+      "epoch": 0.00778816199376947,
+      "grad_norm": 0.6765931399234855,
+      "learning_rate": 9.999625851249816e-06,
+      "loss": 0.2941,
+      "step": 5
+    },
+    {
+      "epoch": 0.009345794392523364,
+      "grad_norm": 0.7193614710776666,
+      "learning_rate": 9.999461228756304e-06,
+      "loss": 0.2705,
+      "step": 6
+    },
+    {
+      "epoch": 0.010903426791277258,
+      "grad_norm": 0.7344615984114217,
+      "learning_rate": 9.99926667722973e-06,
+      "loss": 0.3259,
+      "step": 7
+    },
+    {
+      "epoch": 0.012461059190031152,
+      "grad_norm": 0.73017739071341,
+      "learning_rate": 9.999042197834769e-06,
+      "loss": 0.2852,
+      "step": 8
+    },
+    {
+      "epoch": 0.014018691588785047,
+      "grad_norm": 0.592507619306345,
+      "learning_rate": 9.998787791915254e-06,
+      "loss": 0.2649,
+      "step": 9
+    },
+    {
+      "epoch": 0.01557632398753894,
+      "grad_norm": 0.9399500241178238,
+      "learning_rate": 9.998503460994176e-06,
+      "loss": 0.2521,
+      "step": 10
+    },
+    {
+      "epoch": 0.017133956386292833,
+      "grad_norm": 0.7765731965837874,
+      "learning_rate": 9.998189206773665e-06,
+      "loss": 0.3556,
+      "step": 11
+    },
+    {
+      "epoch": 0.018691588785046728,
+      "grad_norm": 0.6693444621163488,
+      "learning_rate": 9.997845031134992e-06,
+      "loss": 0.2239,
+      "step": 12
+    },
+    {
+      "epoch": 0.020249221183800622,
+      "grad_norm": 1.4936804808243616,
+      "learning_rate": 9.997470936138547e-06,
+      "loss": 0.8968,
+      "step": 13
+    },
+    {
+      "epoch": 0.021806853582554516,
+      "grad_norm": 0.7687864170878725,
+      "learning_rate": 9.997066924023832e-06,
+      "loss": 0.294,
+      "step": 14
+    },
+    {
+      "epoch": 0.02336448598130841,
+      "grad_norm": 1.0982871522503537,
+      "learning_rate": 9.996632997209444e-06,
+      "loss": 0.3007,
+      "step": 15
+    },
+    {
+      "epoch": 0.024922118380062305,
+      "grad_norm": 0.9371896720117513,
+      "learning_rate": 9.996169158293066e-06,
+      "loss": 0.2438,
+      "step": 16
+    },
+    {
+      "epoch": 0.0264797507788162,
+      "grad_norm": 0.7798940457815309,
+      "learning_rate": 9.995675410051448e-06,
+      "loss": 0.2212,
+      "step": 17
+    },
+    {
+      "epoch": 0.028037383177570093,
+      "grad_norm": 0.7355604168049059,
+      "learning_rate": 9.99515175544039e-06,
+      "loss": 0.2793,
+      "step": 18
+    },
+    {
+      "epoch": 0.029595015576323987,
+      "grad_norm": 0.8219475835131775,
+      "learning_rate": 9.994598197594723e-06,
+      "loss": 0.2868,
+      "step": 19
+    },
+    {
+      "epoch": 0.03115264797507788,
+      "grad_norm": 0.8165917677127098,
+      "learning_rate": 9.994014739828298e-06,
+      "loss": 0.2703,
+      "step": 20
+    },
+    {
+      "epoch": 0.03271028037383177,
+      "grad_norm": 0.733767269375867,
+      "learning_rate": 9.993401385633951e-06,
+      "loss": 0.2059,
+      "step": 21
+    },
+    {
+      "epoch": 0.03426791277258567,
+      "grad_norm": 1.2694949029078932,
+      "learning_rate": 9.992758138683502e-06,
+      "loss": 0.9151,
+      "step": 22
+    },
+    {
+      "epoch": 0.03582554517133956,
+      "grad_norm": 0.7777986747788719,
+      "learning_rate": 9.992085002827719e-06,
+      "loss": 0.2599,
+      "step": 23
+    },
+    {
+      "epoch": 0.037383177570093455,
+      "grad_norm": 0.610655735988334,
+      "learning_rate": 9.991381982096293e-06,
+      "loss": 0.2347,
+      "step": 24
+    },
+    {
+      "epoch": 0.03894080996884735,
+      "grad_norm": 0.8562287986067203,
+      "learning_rate": 9.990649080697825e-06,
+      "loss": 0.2359,
+      "step": 25
+    },
+    {
+      "epoch": 0.040498442367601244,
+      "grad_norm": 0.6673172477309072,
+      "learning_rate": 9.989886303019793e-06,
+      "loss": 0.2271,
+      "step": 26
+    },
+    {
+      "epoch": 0.04205607476635514,
+      "grad_norm": 0.8629676432514921,
+      "learning_rate": 9.989093653628532e-06,
+      "loss": 0.2304,
+      "step": 27
+    },
+    {
+      "epoch": 0.04361370716510903,
+      "grad_norm": 0.6898071747588274,
+      "learning_rate": 9.988271137269197e-06,
+      "loss": 0.3192,
+      "step": 28
+    },
+    {
+      "epoch": 0.045171339563862926,
+      "grad_norm": 0.5010462374255644,
+      "learning_rate": 9.987418758865743e-06,
+      "loss": 0.2406,
+      "step": 29
+    },
+    {
+      "epoch": 0.04672897196261682,
+      "grad_norm": 0.8259320779398918,
+      "learning_rate": 9.98653652352089e-06,
+      "loss": 0.3266,
+      "step": 30
+    },
+    {
+      "epoch": 0.048286604361370715,
+      "grad_norm": 0.6076733459531145,
+      "learning_rate": 9.9856244365161e-06,
+      "loss": 0.2997,
+      "step": 31
+    },
+    {
+      "epoch": 0.04984423676012461,
+      "grad_norm": 0.7298955355144913,
+      "learning_rate": 9.984682503311536e-06,
+      "loss": 0.2967,
+      "step": 32
+    },
+    {
+      "epoch": 0.0514018691588785,
+      "grad_norm": 0.7346554751645953,
+      "learning_rate": 9.983710729546038e-06,
+      "loss": 0.3105,
+      "step": 33
+    },
+    {
+      "epoch": 0.0529595015576324,
+      "grad_norm": 0.668130630437843,
+      "learning_rate": 9.98270912103708e-06,
+      "loss": 0.3535,
+      "step": 34
+    },
+    {
+      "epoch": 0.05451713395638629,
+      "grad_norm": 0.6379634068346577,
+      "learning_rate": 9.981677683780743e-06,
+      "loss": 0.3709,
+      "step": 35
+    },
+    {
+      "epoch": 0.056074766355140186,
+      "grad_norm": 0.798860986544702,
+      "learning_rate": 9.98061642395168e-06,
+      "loss": 0.2243,
+      "step": 36
+    },
+    {
+      "epoch": 0.05763239875389408,
+      "grad_norm": 0.6593735789582216,
+      "learning_rate": 9.979525347903067e-06,
+      "loss": 0.2541,
+      "step": 37
+    },
+    {
+      "epoch": 0.059190031152647975,
+      "grad_norm": 0.7224987244711324,
+      "learning_rate": 9.978404462166582e-06,
+      "loss": 0.2736,
+      "step": 38
+    },
+    {
+      "epoch": 0.06074766355140187,
+      "grad_norm": 0.6440291059790796,
+      "learning_rate": 9.977253773452349e-06,
+      "loss": 0.2764,
+      "step": 39
+    },
+    {
+      "epoch": 0.06230529595015576,
+      "grad_norm": 0.6222848807943162,
+      "learning_rate": 9.976073288648913e-06,
+      "loss": 0.2698,
+      "step": 40
+    },
+    {
+      "epoch": 0.06386292834890965,
+      "grad_norm": 0.7394026438854199,
+      "learning_rate": 9.97486301482319e-06,
+      "loss": 0.2522,
+      "step": 41
+    },
+    {
+      "epoch": 0.06542056074766354,
+      "grad_norm": 0.7520093536738988,
+      "learning_rate": 9.973622959220426e-06,
+      "loss": 0.2137,
+      "step": 42
+    },
+    {
+      "epoch": 0.06697819314641744,
+      "grad_norm": 0.744961609793494,
+      "learning_rate": 9.97235312926415e-06,
+      "loss": 0.2328,
+      "step": 43
+    },
+    {
+      "epoch": 0.06853582554517133,
+      "grad_norm": 0.7656967225499874,
+      "learning_rate": 9.971053532556143e-06,
+      "loss": 0.2669,
+      "step": 44
+    },
+    {
+      "epoch": 0.07009345794392523,
+      "grad_norm": 0.6382236962718162,
+      "learning_rate": 9.969724176876373e-06,
+      "loss": 0.2192,
+      "step": 45
+    },
+    {
+      "epoch": 0.07165109034267912,
+      "grad_norm": 0.6472374848660734,
+      "learning_rate": 9.968365070182967e-06,
+      "loss": 0.3144,
+      "step": 46
+    },
+    {
+      "epoch": 0.07320872274143302,
+      "grad_norm": 0.6391395623819942,
+      "learning_rate": 9.966976220612151e-06,
+      "loss": 0.2048,
+      "step": 47
+    },
+    {
+      "epoch": 0.07476635514018691,
+      "grad_norm": 0.7757553970772143,
+      "learning_rate": 9.965557636478203e-06,
+      "loss": 0.2579,
+      "step": 48
+    },
+    {
+      "epoch": 0.0763239875389408,
+      "grad_norm": 0.8133958395260938,
+      "learning_rate": 9.964109326273411e-06,
+      "loss": 0.3226,
+      "step": 49
+    },
+    {
+      "epoch": 0.0778816199376947,
+      "grad_norm": 0.4637065889612604,
+      "learning_rate": 9.962631298668015e-06,
+      "loss": 0.185,
+      "step": 50
+    },
+    {
+      "epoch": 0.0794392523364486,
+      "grad_norm": 0.6273381176066575,
+      "learning_rate": 9.961123562510153e-06,
+      "loss": 0.3509,
+      "step": 51
+    },
+    {
+      "epoch": 0.08099688473520249,
+      "grad_norm": 0.6419234115207709,
+      "learning_rate": 9.959586126825818e-06,
+      "loss": 0.2806,
+      "step": 52
+    },
+    {
+      "epoch": 0.08255451713395638,
+      "grad_norm": 0.7160133507575562,
+      "learning_rate": 9.95801900081879e-06,
+      "loss": 0.1838,
+      "step": 53
+    },
+    {
+      "epoch": 0.08411214953271028,
+      "grad_norm": 0.7848165329694581,
+      "learning_rate": 9.956422193870597e-06,
+      "loss": 0.255,
+      "step": 54
+    },
+    {
+      "epoch": 0.08566978193146417,
+      "grad_norm": 0.6040970248107416,
+      "learning_rate": 9.954795715540447e-06,
+      "loss": 0.2142,
+      "step": 55
+    },
+    {
+      "epoch": 0.08722741433021806,
+      "grad_norm": 0.6115559611935754,
+      "learning_rate": 9.953139575565169e-06,
+      "loss": 0.2668,
+      "step": 56
+    },
+    {
+      "epoch": 0.08878504672897196,
+      "grad_norm": 0.6298402847496053,
+      "learning_rate": 9.95145378385917e-06,
+      "loss": 0.2881,
+      "step": 57
+    },
+    {
+      "epoch": 0.09034267912772585,
+      "grad_norm": 0.5535053280930566,
+      "learning_rate": 9.949738350514358e-06,
+      "loss": 0.2202,
+      "step": 58
+    },
+    {
+      "epoch": 0.09190031152647975,
+      "grad_norm": 0.6352050541969153,
+      "learning_rate": 9.947993285800093e-06,
+      "loss": 0.3398,
+      "step": 59
+    },
+    {
+      "epoch": 0.09345794392523364,
+      "grad_norm": 0.826557642193822,
+      "learning_rate": 9.94621860016312e-06,
+      "loss": 0.3364,
+      "step": 60
+    },
+    {
+      "epoch": 0.09501557632398754,
+      "grad_norm": 0.7298038729692221,
+      "learning_rate": 9.944414304227508e-06,
+      "loss": 0.2634,
+      "step": 61
+    },
+    {
+      "epoch": 0.09657320872274143,
+      "grad_norm": 1.1507218720730263,
+      "learning_rate": 9.94258040879459e-06,
+      "loss": 0.8284,
+      "step": 62
+    },
+    {
+      "epoch": 0.09813084112149532,
+      "grad_norm": 0.6885391578463873,
+      "learning_rate": 9.94071692484289e-06,
+      "loss": 0.2336,
+      "step": 63
+    },
+    {
+      "epoch": 0.09968847352024922,
+      "grad_norm": 0.7334194656077329,
+      "learning_rate": 9.938823863528065e-06,
+      "loss": 0.2652,
+      "step": 64
+    },
+    {
+      "epoch": 0.10124610591900311,
+      "grad_norm": 1.0514064266169973,
+      "learning_rate": 9.936901236182836e-06,
+      "loss": 0.835,
+      "step": 65
+    },
+    {
+      "epoch": 0.102803738317757,
+      "grad_norm": 0.6334750887159478,
+      "learning_rate": 9.934949054316917e-06,
+      "loss": 0.2812,
+      "step": 66
+    },
+    {
+      "epoch": 0.1043613707165109,
+      "grad_norm": 0.4857684609448803,
+      "learning_rate": 9.932967329616953e-06,
+      "loss": 0.202,
+      "step": 67
+    },
+    {
+      "epoch": 0.1059190031152648,
+      "grad_norm": 0.690411540980916,
+      "learning_rate": 9.930956073946442e-06,
+      "loss": 0.2828,
+      "step": 68
+    },
+    {
+      "epoch": 0.10747663551401869,
+      "grad_norm": 0.6862713871911896,
+      "learning_rate": 9.928915299345669e-06,
+      "loss": 0.3246,
+      "step": 69
+    },
+    {
+      "epoch": 0.10903426791277258,
+      "grad_norm": 0.6617081291471549,
+      "learning_rate": 9.926845018031632e-06,
+      "loss": 0.2621,
+      "step": 70
+    },
+    {
+      "epoch": 0.11059190031152648,
+      "grad_norm": 0.5745643134324515,
+      "learning_rate": 9.924745242397968e-06,
+      "loss": 0.2224,
+      "step": 71
+    },
+    {
+      "epoch": 0.11214953271028037,
+      "grad_norm": 0.7300598195282793,
+      "learning_rate": 9.922615985014887e-06,
+      "loss": 0.2297,
+      "step": 72
+    },
+    {
+      "epoch": 0.11370716510903427,
+      "grad_norm": 0.6648229017872759,
+      "learning_rate": 9.920457258629081e-06,
+      "loss": 0.2461,
+      "step": 73
+    },
+    {
+      "epoch": 0.11526479750778816,
+      "grad_norm": 0.6868885744716628,
+      "learning_rate": 9.918269076163664e-06,
+      "loss": 0.2728,
+      "step": 74
+    },
+    {
+      "epoch": 0.11682242990654206,
+      "grad_norm": 0.6982895714335172,
+      "learning_rate": 9.916051450718085e-06,
+      "loss": 0.257,
+      "step": 75
+    },
+    {
+      "epoch": 0.11838006230529595,
+      "grad_norm": 0.5677598559231926,
+      "learning_rate": 9.91380439556805e-06,
+      "loss": 0.2614,
+      "step": 76
+    },
+    {
+      "epoch": 0.11993769470404984,
+      "grad_norm": 0.5071293538430349,
+      "learning_rate": 9.911527924165445e-06,
+      "loss": 0.2793,
+      "step": 77
+    },
+    {
+      "epoch": 0.12149532710280374,
+      "grad_norm": 0.7281461877976642,
+      "learning_rate": 9.909222050138259e-06,
+      "loss": 0.2861,
+      "step": 78
+    },
+    {
+      "epoch": 0.12305295950155763,
+      "grad_norm": 0.6491525810286177,
+      "learning_rate": 9.906886787290492e-06,
+      "loss": 0.2799,
+      "step": 79
+    },
+    {
+      "epoch": 0.12461059190031153,
+      "grad_norm": 0.9626139969649712,
+      "learning_rate": 9.904522149602087e-06,
+      "loss": 0.2251,
+      "step": 80
+    },
+    {
+      "epoch": 0.1261682242990654,
+      "grad_norm": 0.705551656345552,
+      "learning_rate": 9.902128151228827e-06,
+      "loss": 0.2899,
+      "step": 81
+    },
+    {
+      "epoch": 0.1277258566978193,
+      "grad_norm": 0.5796094318767273,
+      "learning_rate": 9.899704806502272e-06,
+      "loss": 0.3023,
+      "step": 82
+    },
+    {
+      "epoch": 0.1292834890965732,
+      "grad_norm": 0.5778885783153778,
+      "learning_rate": 9.89725212992966e-06,
+      "loss": 0.2596,
+      "step": 83
+    },
+    {
+      "epoch": 0.1308411214953271,
+      "grad_norm": 0.6407326481544323,
+      "learning_rate": 9.894770136193814e-06,
+      "loss": 0.3182,
+      "step": 84
+    },
+    {
+      "epoch": 0.13239875389408098,
+      "grad_norm": 0.8761396102635524,
+      "learning_rate": 9.892258840153075e-06,
+      "loss": 0.3261,
+      "step": 85
+    },
+    {
+      "epoch": 0.13395638629283488,
+      "grad_norm": 0.8248387465515045,
+      "learning_rate": 9.889718256841195e-06,
+      "loss": 0.3105,
+      "step": 86
+    },
+    {
+      "epoch": 0.13551401869158877,
+      "grad_norm": 2.168823178408131,
+      "learning_rate": 9.88714840146725e-06,
+      "loss": 0.7869,
+      "step": 87
+    },
+    {
+      "epoch": 0.13707165109034267,
+      "grad_norm": 0.6163273279750748,
+      "learning_rate": 9.884549289415556e-06,
+      "loss": 0.2249,
+      "step": 88
+    },
+    {
+      "epoch": 0.13862928348909656,
+      "grad_norm": 1.5331833228458356,
+      "learning_rate": 9.881920936245577e-06,
+      "loss": 0.7778,
+      "step": 89
+    },
+    {
+      "epoch": 0.14018691588785046,
+      "grad_norm": 1.5428867178521768,
+      "learning_rate": 9.879263357691815e-06,
+      "loss": 0.8951,
+      "step": 90
+    },
+    {
+      "epoch": 0.14174454828660435,
+      "grad_norm": 0.6177160477898084,
+      "learning_rate": 9.876576569663739e-06,
+      "loss": 0.224,
+      "step": 91
+    },
+    {
+      "epoch": 0.14330218068535824,
+      "grad_norm": 0.7981644114495157,
+      "learning_rate": 9.873860588245675e-06,
+      "loss": 0.2892,
+      "step": 92
+    },
+    {
+      "epoch": 0.14485981308411214,
+      "grad_norm": 0.600801453450908,
+      "learning_rate": 9.871115429696716e-06,
+      "loss": 0.2782,
+      "step": 93
+    },
+    {
+      "epoch": 0.14641744548286603,
+      "grad_norm": 0.6547888169420532,
+      "learning_rate": 9.868341110450618e-06,
+      "loss": 0.3347,
+      "step": 94
+    },
+    {
+      "epoch": 0.14797507788161993,
+      "grad_norm": 0.5823853177715245,
+      "learning_rate": 9.865537647115713e-06,
+      "loss": 0.2778,
+      "step": 95
+    },
+    {
+      "epoch": 0.14953271028037382,
+      "grad_norm": 0.9116578300075924,
+      "learning_rate": 9.862705056474795e-06,
+      "loss": 0.3503,
+      "step": 96
+    },
+    {
+      "epoch": 0.15109034267912771,
+      "grad_norm": 0.5426782446240797,
+      "learning_rate": 9.859843355485033e-06,
+      "loss": 0.2386,
+      "step": 97
+    },
+    {
+      "epoch": 0.1526479750778816,
+      "grad_norm": 0.5543482391953797,
+      "learning_rate": 9.856952561277862e-06,
+      "loss": 0.2344,
+      "step": 98
+    },
+    {
+      "epoch": 0.1542056074766355,
+      "grad_norm": 0.7660668413572433,
+      "learning_rate": 9.854032691158881e-06,
+      "loss": 0.2692,
+      "step": 99
+    },
+    {
+      "epoch": 0.1557632398753894,
+      "grad_norm": 0.7745787606896338,
+      "learning_rate": 9.851083762607752e-06,
+      "loss": 0.2374,
+      "step": 100
+    },
+    {
+      "epoch": 0.1573208722741433,
+      "grad_norm": 0.8283533162770704,
+      "learning_rate": 9.848105793278092e-06,
+      "loss": 0.2699,
+      "step": 101
+    },
+    {
+      "epoch": 0.1588785046728972,
+      "grad_norm": 0.5323847631668214,
+      "learning_rate": 9.84509880099737e-06,
+      "loss": 0.2177,
+      "step": 102
+    },
+    {
+      "epoch": 0.16043613707165108,
+      "grad_norm": 0.7022062850923283,
+      "learning_rate": 9.842062803766804e-06,
+      "loss": 0.2258,
+      "step": 103
+    },
+    {
+      "epoch": 0.16199376947040497,
+      "grad_norm": 1.9739756098890826,
+      "learning_rate": 9.838997819761241e-06,
+      "loss": 0.9263,
+      "step": 104
+    },
+    {
+      "epoch": 0.16355140186915887,
+      "grad_norm": 0.7651110334163559,
+      "learning_rate": 9.835903867329061e-06,
+      "loss": 0.2654,
+      "step": 105
+    },
+    {
+      "epoch": 0.16510903426791276,
+      "grad_norm": 0.6634416278579489,
+      "learning_rate": 9.832780964992059e-06,
+      "loss": 0.2959,
+      "step": 106
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.62947915393871,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.2608,
+      "step": 107
+    },
+    {
+      "epoch": 0.16822429906542055,
+      "grad_norm": 1.55385845539561,
+      "learning_rate": 9.826448385557208e-06,
+      "loss": 0.842,
+      "step": 108
+    },
+    {
+      "epoch": 0.16978193146417445,
+      "grad_norm": 0.6067362647236724,
+      "learning_rate": 9.823238746369038e-06,
+      "loss": 0.3388,
+      "step": 109
+    },
+    {
+      "epoch": 0.17133956386292834,
+      "grad_norm": 0.7095716481506402,
+      "learning_rate": 9.820000233095179e-06,
+      "loss": 0.2953,
+      "step": 110
+    },
+    {
+      "epoch": 0.17289719626168223,
+      "grad_norm": 1.4374378842104667,
+      "learning_rate": 9.81673286512284e-06,
+      "loss": 0.7775,
+      "step": 111
+    },
+    {
+      "epoch": 0.17445482866043613,
+      "grad_norm": 0.6761770106689733,
+      "learning_rate": 9.813436662011958e-06,
+      "loss": 0.2929,
+      "step": 112
+    },
+    {
+      "epoch": 0.17601246105919002,
+      "grad_norm": 0.7954842040051252,
+      "learning_rate": 9.810111643495095e-06,
+      "loss": 0.2734,
+      "step": 113
+    },
+    {
+      "epoch": 0.17757009345794392,
+      "grad_norm": 1.271851983971207,
+      "learning_rate": 9.806757829477313e-06,
+      "loss": 0.7987,
+      "step": 114
+    },
+    {
+      "epoch": 0.1791277258566978,
+      "grad_norm": 0.526209190777812,
+      "learning_rate": 9.803375240036057e-06,
+      "loss": 0.3054,
+      "step": 115
+    },
+    {
+      "epoch": 0.1806853582554517,
+      "grad_norm": 0.6440505513615228,
+      "learning_rate": 9.799963895421036e-06,
+      "loss": 0.2416,
+      "step": 116
+    },
+    {
+      "epoch": 0.1822429906542056,
+      "grad_norm": 0.9367999131179932,
+      "learning_rate": 9.7965238160541e-06,
+      "loss": 0.3409,
+      "step": 117
+    },
+    {
+      "epoch": 0.1838006230529595,
+      "grad_norm": 0.8415567444030029,
+      "learning_rate": 9.793055022529116e-06,
+      "loss": 0.2847,
+      "step": 118
+    },
+    {
+      "epoch": 0.1853582554517134,
+      "grad_norm": 0.5786248320276434,
+      "learning_rate": 9.789557535611853e-06,
+      "loss": 0.2644,
+      "step": 119
+    },
+    {
+      "epoch": 0.18691588785046728,
+      "grad_norm": 0.687463433264138,
+      "learning_rate": 9.786031376239842e-06,
+      "loss": 0.2569,
+      "step": 120
+    },
+    {
+      "epoch": 0.18847352024922118,
+      "grad_norm": 0.45347828103414345,
+      "learning_rate": 9.78247656552227e-06,
+      "loss": 0.2415,
+      "step": 121
+    },
+    {
+      "epoch": 0.19003115264797507,
+      "grad_norm": 0.5650726061282849,
+      "learning_rate": 9.778893124739836e-06,
+      "loss": 0.2547,
+      "step": 122
+    },
+    {
+      "epoch": 0.19158878504672897,
+      "grad_norm": 0.5103395837306397,
+      "learning_rate": 9.775281075344639e-06,
+      "loss": 0.3065,
+      "step": 123
+    },
+    {
+      "epoch": 0.19314641744548286,
+      "grad_norm": 0.5183636563691485,
+      "learning_rate": 9.771640438960037e-06,
+      "loss": 0.2922,
+      "step": 124
+    },
+    {
+      "epoch": 0.19470404984423675,
+      "grad_norm": 0.5026042818767394,
+      "learning_rate": 9.76797123738052e-06,
+      "loss": 0.2588,
+      "step": 125
+    },
+    {
+      "epoch": 0.19626168224299065,
+      "grad_norm": 0.6794540572358105,
+      "learning_rate": 9.76427349257159e-06,
+      "loss": 0.3041,
+      "step": 126
+    },
+    {
+      "epoch": 0.19781931464174454,
+      "grad_norm": 0.6283325050944261,
+      "learning_rate": 9.76054722666962e-06,
+      "loss": 0.2991,
+      "step": 127
+    },
+    {
+      "epoch": 0.19937694704049844,
+      "grad_norm": 0.7184815540245492,
+      "learning_rate": 9.756792461981713e-06,
+      "loss": 0.2599,
+      "step": 128
+    },
+    {
+      "epoch": 0.20093457943925233,
+      "grad_norm": 0.701839843872524,
+      "learning_rate": 9.753009220985593e-06,
+      "loss": 0.2711,
+      "step": 129
+    },
+    {
+      "epoch": 0.20249221183800623,
+      "grad_norm": 0.6505643462960924,
+      "learning_rate": 9.749197526329446e-06,
+      "loss": 0.2947,
+      "step": 130
+    },
+    {
+      "epoch": 0.20404984423676012,
+      "grad_norm": 0.6104336914925119,
+      "learning_rate": 9.745357400831799e-06,
+      "loss": 0.2331,
+      "step": 131
+    },
+    {
+      "epoch": 0.205607476635514,
+      "grad_norm": 0.908932761888463,
+      "learning_rate": 9.741488867481377e-06,
+      "loss": 0.2377,
+      "step": 132
+    },
+    {
+      "epoch": 0.2071651090342679,
+      "grad_norm": 0.5980236644974856,
+      "learning_rate": 9.737591949436969e-06,
+      "loss": 0.1601,
+      "step": 133
+    },
+    {
+      "epoch": 0.2087227414330218,
+      "grad_norm": 0.5937884904455855,
+      "learning_rate": 9.733666670027288e-06,
+      "loss": 0.2542,
+      "step": 134
+    },
+    {
+      "epoch": 0.2102803738317757,
+      "grad_norm": 0.7666646448496817,
+      "learning_rate": 9.729713052750827e-06,
+      "loss": 0.2465,
+      "step": 135
+    },
+    {
+      "epoch": 0.2118380062305296,
+      "grad_norm": 0.5386370084617528,
+      "learning_rate": 9.725731121275725e-06,
+      "loss": 0.3172,
+      "step": 136
+    },
+    {
+      "epoch": 0.21339563862928349,
+      "grad_norm": 0.5517332858518383,
+      "learning_rate": 9.721720899439621e-06,
+      "loss": 0.2864,
+      "step": 137
+    },
+    {
+      "epoch": 0.21495327102803738,
+      "grad_norm": 0.60004714901346,
+      "learning_rate": 9.71768241124952e-06,
+      "loss": 0.3101,
+      "step": 138
+    },
+    {
+      "epoch": 0.21651090342679127,
+      "grad_norm": 0.7525697267510544,
+      "learning_rate": 9.71361568088163e-06,
+      "loss": 0.2235,
+      "step": 139
+    },
+    {
+      "epoch": 0.21806853582554517,
+      "grad_norm": 0.5929490381364206,
+      "learning_rate": 9.709520732681238e-06,
+      "loss": 0.231,
+      "step": 140
+    },
+    {
+      "epoch": 0.21962616822429906,
+      "grad_norm": 0.7027523851155623,
+      "learning_rate": 9.705397591162556e-06,
+      "loss": 0.2379,
+      "step": 141
+    },
+    {
+      "epoch": 0.22118380062305296,
+      "grad_norm": 0.6504000514836087,
+      "learning_rate": 9.701246281008568e-06,
+      "loss": 0.2462,
+      "step": 142
+    },
+    {
+      "epoch": 0.22274143302180685,
+      "grad_norm": 0.5770728592720746,
+      "learning_rate": 9.697066827070894e-06,
+      "loss": 0.2859,
+      "step": 143
+    },
+    {
+      "epoch": 0.22429906542056074,
+      "grad_norm": 0.5798579381876525,
+      "learning_rate": 9.692859254369631e-06,
+      "loss": 0.2663,
+      "step": 144
+    },
+    {
+      "epoch": 0.22585669781931464,
+      "grad_norm": 0.5033709259076921,
+      "learning_rate": 9.68862358809321e-06,
+      "loss": 0.1913,
+      "step": 145
+    },
+    {
+      "epoch": 0.22741433021806853,
+      "grad_norm": 0.5493196423085126,
+      "learning_rate": 9.684359853598245e-06,
+      "loss": 0.2197,
+      "step": 146
+    },
+    {
+      "epoch": 0.22897196261682243,
+      "grad_norm": 0.5652921319820055,
+      "learning_rate": 9.680068076409373e-06,
+      "loss": 0.2601,
+      "step": 147
+    },
+    {
+      "epoch": 0.23052959501557632,
+      "grad_norm": 0.6783593909741826,
+      "learning_rate": 9.675748282219114e-06,
+      "loss": 0.2922,
+      "step": 148
+    },
+    {
+      "epoch": 0.23208722741433022,
+      "grad_norm": 0.7088765017570264,
+      "learning_rate": 9.671400496887704e-06,
+      "loss": 0.2997,
+      "step": 149
+    },
+    {
+      "epoch": 0.2336448598130841,
+      "grad_norm": 0.6064540912244738,
+      "learning_rate": 9.667024746442953e-06,
+      "loss": 0.3283,
+      "step": 150
+    },
+    {
+      "epoch": 0.235202492211838,
+      "grad_norm": 0.6900794118386824,
+      "learning_rate": 9.662621057080077e-06,
+      "loss": 0.2447,
+      "step": 151
+    },
+    {
+      "epoch": 0.2367601246105919,
+      "grad_norm": 0.8898873494327856,
+      "learning_rate": 9.65818945516155e-06,
+      "loss": 0.2857,
+      "step": 152
+    },
+    {
+      "epoch": 0.2383177570093458,
+      "grad_norm": 0.7596388046727889,
+      "learning_rate": 9.653729967216944e-06,
+      "loss": 0.2765,
+      "step": 153
+    },
+    {
+      "epoch": 0.2398753894080997,
+      "grad_norm": 0.5897245550654492,
+      "learning_rate": 9.64924261994277e-06,
+      "loss": 0.2629,
+      "step": 154
+    },
+    {
+      "epoch": 0.24143302180685358,
+      "grad_norm": 0.6261944121589039,
+      "learning_rate": 9.644727440202308e-06,
+      "loss": 0.3554,
+      "step": 155
+    },
+    {
+      "epoch": 0.24299065420560748,
+      "grad_norm": 0.7471727372694826,
+      "learning_rate": 9.640184455025472e-06,
+      "loss": 0.3595,
+      "step": 156
+    },
+    {
+      "epoch": 0.24454828660436137,
+      "grad_norm": 0.6539521944383881,
+      "learning_rate": 9.635613691608619e-06,
+      "loss": 0.2351,
+      "step": 157
+    },
+    {
+      "epoch": 0.24610591900311526,
+      "grad_norm": 0.7372613573398835,
+      "learning_rate": 9.631015177314402e-06,
+      "loss": 0.3061,
+      "step": 158
+    },
+    {
+      "epoch": 0.24766355140186916,
+      "grad_norm": 0.802899194753761,
+      "learning_rate": 9.62638893967161e-06,
+      "loss": 0.3606,
+      "step": 159
+    },
+    {
+      "epoch": 0.24922118380062305,
+      "grad_norm": 0.754935748177354,
+      "learning_rate": 9.621735006374984e-06,
+      "loss": 0.223,
+      "step": 160
+    },
+    {
+      "epoch": 0.2507788161993769,
+      "grad_norm": 0.4173684214144698,
+      "learning_rate": 9.617053405285077e-06,
+      "loss": 0.2319,
+      "step": 161
+    },
+    {
+      "epoch": 0.2523364485981308,
+      "grad_norm": 0.6435417325636267,
+      "learning_rate": 9.612344164428063e-06,
+      "loss": 0.3996,
+      "step": 162
+    },
+    {
+      "epoch": 0.2538940809968847,
+      "grad_norm": 0.8830545123878781,
+      "learning_rate": 9.607607311995587e-06,
+      "loss": 0.3724,
+      "step": 163
+    },
+    {
+      "epoch": 0.2554517133956386,
+      "grad_norm": 2.2533700279001554,
+      "learning_rate": 9.602842876344589e-06,
+      "loss": 0.8261,
+      "step": 164
+    },
+    {
+      "epoch": 0.2570093457943925,
+      "grad_norm": 0.6641726752237889,
+      "learning_rate": 9.59805088599713e-06,
+      "loss": 0.3339,
+      "step": 165
+    },
+    {
+      "epoch": 0.2585669781931464,
+      "grad_norm": 0.6276989828906703,
+      "learning_rate": 9.59323136964023e-06,
+      "loss": 0.219,
+      "step": 166
+    },
+    {
+      "epoch": 0.2601246105919003,
+      "grad_norm": 0.8210416456426778,
+      "learning_rate": 9.588384356125691e-06,
+      "loss": 0.282,
+      "step": 167
+    },
+    {
+      "epoch": 0.2616822429906542,
+      "grad_norm": 0.751594963703605,
+      "learning_rate": 9.583509874469924e-06,
+      "loss": 0.3373,
+      "step": 168
+    },
+    {
+      "epoch": 0.2632398753894081,
+      "grad_norm": 0.7261273311893066,
+      "learning_rate": 9.578607953853777e-06,
+      "loss": 0.28,
+      "step": 169
+    },
+    {
+      "epoch": 0.26479750778816197,
+      "grad_norm": 0.6922505407762392,
+      "learning_rate": 9.573678623622364e-06,
+      "loss": 0.3412,
+      "step": 170
+    },
+    {
+      "epoch": 0.26635514018691586,
+      "grad_norm": 0.6419868376151807,
+      "learning_rate": 9.568721913284879e-06,
+      "loss": 0.2588,
+      "step": 171
+    },
+    {
+      "epoch": 0.26791277258566976,
+      "grad_norm": 0.49324052267306684,
+      "learning_rate": 9.563737852514432e-06,
+      "loss": 0.2332,
+      "step": 172
+    },
+    {
+      "epoch": 0.26947040498442365,
+      "grad_norm": 0.5770500149070207,
+      "learning_rate": 9.558726471147862e-06,
+      "loss": 0.2937,
+      "step": 173
+    },
+    {
+      "epoch": 0.27102803738317754,
+      "grad_norm": 0.7329648250931949,
+      "learning_rate": 9.553687799185556e-06,
+      "loss": 0.3309,
+      "step": 174
+    },
+    {
+      "epoch": 0.27258566978193144,
+      "grad_norm": 0.4864393923841514,
+      "learning_rate": 9.548621866791286e-06,
+      "loss": 0.2043,
+      "step": 175
+    },
+    {
+      "epoch": 0.27414330218068533,
+      "grad_norm": 1.136074014302222,
+      "learning_rate": 9.54352870429201e-06,
+      "loss": 0.7893,
+      "step": 176
+    },
+    {
+      "epoch": 0.2757009345794392,
+      "grad_norm": 0.6443586968700755,
+      "learning_rate": 9.538408342177699e-06,
+      "loss": 0.3025,
+      "step": 177
+    },
+    {
+      "epoch": 0.2772585669781931,
+      "grad_norm": 0.6403852564269055,
+      "learning_rate": 9.533260811101152e-06,
+      "loss": 0.2629,
+      "step": 178
+    },
+    {
+      "epoch": 0.278816199376947,
+      "grad_norm": 0.6869290500301357,
+      "learning_rate": 9.528086141877817e-06,
+      "loss": 0.301,
+      "step": 179
+    },
+    {
+      "epoch": 0.2803738317757009,
+      "grad_norm": 0.9338061428659953,
+      "learning_rate": 9.522884365485599e-06,
+      "loss": 0.7925,
+      "step": 180
+    },
+    {
+      "epoch": 0.2819314641744548,
+      "grad_norm": 0.7304512334819939,
+      "learning_rate": 9.517655513064682e-06,
+      "loss": 0.44,
+      "step": 181
+    },
+    {
+      "epoch": 0.2834890965732087,
+      "grad_norm": 0.6207928517482323,
+      "learning_rate": 9.512399615917342e-06,
+      "loss": 0.228,
+      "step": 182
+    },
+    {
+      "epoch": 0.2850467289719626,
+      "grad_norm": 0.8847617104487071,
+      "learning_rate": 9.507116705507748e-06,
+      "loss": 0.4137,
+      "step": 183
+    },
+    {
+      "epoch": 0.2866043613707165,
+      "grad_norm": 0.6631070285038642,
+      "learning_rate": 9.50180681346179e-06,
+      "loss": 0.3438,
+      "step": 184
+    },
+    {
+      "epoch": 0.2881619937694704,
+      "grad_norm": 1.0647904328891042,
+      "learning_rate": 9.49646997156688e-06,
+      "loss": 0.2452,
+      "step": 185
+    },
+    {
+      "epoch": 0.2897196261682243,
+      "grad_norm": 0.4959301062995678,
+      "learning_rate": 9.491106211771765e-06,
+      "loss": 0.2915,
+      "step": 186
+    },
+    {
+      "epoch": 0.29127725856697817,
+      "grad_norm": 0.705003606824088,
+      "learning_rate": 9.485715566186333e-06,
+      "loss": 0.2945,
+      "step": 187
+    },
+    {
+      "epoch": 0.29283489096573206,
+      "grad_norm": 1.045178016891709,
+      "learning_rate": 9.480298067081422e-06,
+      "loss": 0.5356,
+      "step": 188
+    },
+    {
+      "epoch": 0.29439252336448596,
+      "grad_norm": 0.6198122538394608,
+      "learning_rate": 9.474853746888631e-06,
+      "loss": 0.2744,
+      "step": 189
+    },
+    {
+      "epoch": 0.29595015576323985,
+      "grad_norm": 0.631725313682397,
+      "learning_rate": 9.469382638200119e-06,
+      "loss": 0.3064,
+      "step": 190
+    },
+    {
+      "epoch": 0.29750778816199375,
+      "grad_norm": 0.49405112326099454,
+      "learning_rate": 9.463884773768413e-06,
+      "loss": 0.2016,
+      "step": 191
+    },
+    {
+      "epoch": 0.29906542056074764,
+      "grad_norm": 0.9255242318461628,
+      "learning_rate": 9.458360186506212e-06,
+      "loss": 0.318,
+      "step": 192
+    },
+    {
+      "epoch": 0.30062305295950154,
+      "grad_norm": 0.5172541239349757,
+      "learning_rate": 9.452808909486195e-06,
+      "loss": 0.2182,
+      "step": 193
+    },
+    {
+      "epoch": 0.30218068535825543,
+      "grad_norm": 0.6548225129226617,
+      "learning_rate": 9.447230975940808e-06,
+      "loss": 0.3501,
+      "step": 194
+    },
+    {
+      "epoch": 0.3037383177570093,
+      "grad_norm": 0.6373263585258429,
+      "learning_rate": 9.441626419262084e-06,
+      "loss": 0.2331,
+      "step": 195
+    },
+    {
+      "epoch": 0.3052959501557632,
+      "grad_norm": 0.8451103465765266,
+      "learning_rate": 9.43599527300143e-06,
+      "loss": 0.2379,
+      "step": 196
+    },
+    {
+      "epoch": 0.3068535825545171,
+      "grad_norm": 0.6257655537331657,
+      "learning_rate": 9.430337570869432e-06,
+      "loss": 0.262,
+      "step": 197
+    },
+    {
+      "epoch": 0.308411214953271,
+      "grad_norm": 0.7885047351755374,
+      "learning_rate": 9.424653346735649e-06,
+      "loss": 0.2783,
+      "step": 198
+    },
+    {
+      "epoch": 0.3099688473520249,
+      "grad_norm": 1.2716968661440529,
+      "learning_rate": 9.418942634628413e-06,
+      "loss": 0.3266,
+      "step": 199
+    },
+    {
+      "epoch": 0.3115264797507788,
+      "grad_norm": 0.709734211796566,
+      "learning_rate": 9.413205468734628e-06,
+      "loss": 0.298,
+      "step": 200
+    },
+    {
+      "epoch": 0.3130841121495327,
+      "grad_norm": 0.6966111704003821,
+      "learning_rate": 9.40744188339956e-06,
+      "loss": 0.3872,
+      "step": 201
+    },
+    {
+      "epoch": 0.3146417445482866,
+      "grad_norm": 0.6328947624746845,
+      "learning_rate": 9.401651913126634e-06,
+      "loss": 0.2855,
+      "step": 202
+    },
+    {
+      "epoch": 0.3161993769470405,
+      "grad_norm": 0.6782499058815041,
+      "learning_rate": 9.395835592577228e-06,
+      "loss": 0.2555,
+      "step": 203
+    },
+    {
+      "epoch": 0.3177570093457944,
+      "grad_norm": 0.6111146053952768,
+      "learning_rate": 9.389992956570463e-06,
+      "loss": 0.2119,
+      "step": 204
+    },
+    {
+      "epoch": 0.31931464174454827,
+      "grad_norm": 2.1014154014060646,
+      "learning_rate": 9.384124040082996e-06,
+      "loss": 1.3393,
+      "step": 205
+    },
+    {
+      "epoch": 0.32087227414330216,
+      "grad_norm": 0.6937660819701476,
+      "learning_rate": 9.378228878248814e-06,
+      "loss": 0.3003,
+      "step": 206
+    },
+    {
+      "epoch": 0.32242990654205606,
+      "grad_norm": 0.7798606308310687,
+      "learning_rate": 9.372307506359019e-06,
+      "loss": 0.2394,
+      "step": 207
+    },
+    {
+      "epoch": 0.32398753894080995,
+      "grad_norm": 0.8064220607757047,
+      "learning_rate": 9.366359959861615e-06,
+      "loss": 0.2457,
+      "step": 208
+    },
+    {
+      "epoch": 0.32554517133956384,
+      "grad_norm": 0.8663909019847741,
+      "learning_rate": 9.360386274361305e-06,
+      "loss": 0.2363,
+      "step": 209
+    },
+    {
+      "epoch": 0.32710280373831774,
+      "grad_norm": 0.7765065772183143,
+      "learning_rate": 9.354386485619264e-06,
+      "loss": 0.2519,
+      "step": 210
+    },
+    {
+      "epoch": 0.32866043613707163,
+      "grad_norm": 0.6924828379419826,
+      "learning_rate": 9.348360629552941e-06,
+      "loss": 0.2237,
+      "step": 211
+    },
+    {
+      "epoch": 0.3302180685358255,
+      "grad_norm": 0.6564177523605338,
+      "learning_rate": 9.342308742235831e-06,
+      "loss": 0.2369,
+      "step": 212
+    },
+    {
+      "epoch": 0.3317757009345794,
+      "grad_norm": 0.6166208484170821,
+      "learning_rate": 9.336230859897266e-06,
+      "loss": 0.2531,
+      "step": 213
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.6325656097098429,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.228,
+      "step": 214
+    },
+    {
+      "epoch": 0.3348909657320872,
+      "grad_norm": 0.9005060609426595,
+      "learning_rate": 9.323997255850965e-06,
+      "loss": 0.2068,
+      "step": 215
+    },
+    {
+      "epoch": 0.3364485981308411,
+      "grad_norm": 0.6030970162930338,
+      "learning_rate": 9.317841607379106e-06,
+      "loss": 0.2269,
+      "step": 216
+    },
+    {
+      "epoch": 0.338006230529595,
+      "grad_norm": 0.5757510215326483,
+      "learning_rate": 9.311660110357116e-06,
+      "loss": 0.2277,
+      "step": 217
+    },
+    {
+      "epoch": 0.3395638629283489,
+      "grad_norm": 0.48332154427815277,
+      "learning_rate": 9.30545280179022e-06,
+      "loss": 0.2088,
+      "step": 218
+    },
+    {
+      "epoch": 0.3411214953271028,
+      "grad_norm": 1.6587632436476218,
+      "learning_rate": 9.299219718838174e-06,
+      "loss": 0.8341,
+      "step": 219
+    },
+    {
+      "epoch": 0.3426791277258567,
+      "grad_norm": 0.9279443473544912,
+      "learning_rate": 9.292960898815032e-06,
+      "loss": 0.2415,
+      "step": 220
+    },
+    {
+      "epoch": 0.3442367601246106,
+      "grad_norm": 1.6924178170865158,
+      "learning_rate": 9.286676379188913e-06,
+      "loss": 0.9856,
+      "step": 221
+    },
+    {
+      "epoch": 0.34579439252336447,
+      "grad_norm": 1.8214244233005599,
+      "learning_rate": 9.280366197581792e-06,
+      "loss": 0.826,
+      "step": 222
+    },
+    {
+      "epoch": 0.34735202492211836,
+      "grad_norm": 0.783390501917877,
+      "learning_rate": 9.274030391769264e-06,
+      "loss": 0.8049,
+      "step": 223
+    },
+    {
+      "epoch": 0.34890965732087226,
+      "grad_norm": 0.5859949211179467,
+      "learning_rate": 9.267668999680328e-06,
+      "loss": 0.2939,
+      "step": 224
+    },
+    {
+      "epoch": 0.35046728971962615,
+      "grad_norm": 0.7610274820950346,
+      "learning_rate": 9.261282059397145e-06,
+      "loss": 0.2581,
+      "step": 225
+    },
+    {
+      "epoch": 0.35202492211838005,
+      "grad_norm": 0.6066786542121007,
+      "learning_rate": 9.25486960915483e-06,
+      "loss": 0.2758,
+      "step": 226
+    },
+    {
+      "epoch": 0.35358255451713394,
+      "grad_norm": 1.1073512044256302,
+      "learning_rate": 9.248431687341203e-06,
+      "loss": 0.237,
+      "step": 227
+    },
+    {
+      "epoch": 0.35514018691588783,
+      "grad_norm": 1.0141799395602682,
+      "learning_rate": 9.241968332496576e-06,
+      "loss": 0.2568,
+      "step": 228
+    },
+    {
+      "epoch": 0.35669781931464173,
+      "grad_norm": 0.5568072088382595,
+      "learning_rate": 9.235479583313504e-06,
+      "loss": 0.2334,
+      "step": 229
+    },
+    {
+      "epoch": 0.3582554517133956,
+      "grad_norm": 0.7930065901392893,
+      "learning_rate": 9.228965478636575e-06,
+      "loss": 0.2793,
+      "step": 230
+    },
+    {
+      "epoch": 0.3598130841121495,
+      "grad_norm": 0.7048596889355911,
+      "learning_rate": 9.222426057462162e-06,
+      "loss": 0.3062,
+      "step": 231
+    },
+    {
+      "epoch": 0.3613707165109034,
+      "grad_norm": 0.8914753088496622,
+      "learning_rate": 9.215861358938191e-06,
+      "loss": 0.3289,
+      "step": 232
+    },
+    {
+      "epoch": 0.3629283489096573,
+      "grad_norm": 0.8233950771824091,
+      "learning_rate": 9.209271422363915e-06,
+      "loss": 0.3021,
+      "step": 233
+    },
+    {
+      "epoch": 0.3644859813084112,
+      "grad_norm": 0.8572614177235147,
+      "learning_rate": 9.20265628718967e-06,
+      "loss": 0.3435,
+      "step": 234
+    },
+    {
+      "epoch": 0.3660436137071651,
+      "grad_norm": 0.6533204687228193,
+      "learning_rate": 9.196015993016642e-06,
+      "loss": 0.2537,
+      "step": 235
+    },
+    {
+      "epoch": 0.367601246105919,
+      "grad_norm": 0.676657852907951,
+      "learning_rate": 9.189350579596635e-06,
+      "loss": 0.273,
+      "step": 236
+    },
+    {
+      "epoch": 0.3691588785046729,
+      "grad_norm": 0.8090061594796749,
+      "learning_rate": 9.182660086831819e-06,
+      "loss": 0.3786,
+      "step": 237
+    },
+    {
+      "epoch": 0.3707165109034268,
+      "grad_norm": 0.6195251333212162,
+      "learning_rate": 9.175944554774507e-06,
+      "loss": 0.2799,
+      "step": 238
+    },
+    {
+      "epoch": 0.37227414330218067,
+      "grad_norm": 0.5501170288605118,
+      "learning_rate": 9.169204023626911e-06,
+      "loss": 0.2577,
+      "step": 239
+    },
+    {
+      "epoch": 0.37383177570093457,
+      "grad_norm": 0.6087726744520703,
+      "learning_rate": 9.162438533740891e-06,
+      "loss": 0.2578,
+      "step": 240
+    },
+    {
+      "epoch": 0.37538940809968846,
+      "grad_norm": 0.6423864822269919,
+      "learning_rate": 9.15564812561773e-06,
+      "loss": 0.2865,
+      "step": 241
+    },
+    {
+      "epoch": 0.37694704049844235,
+      "grad_norm": 0.6443057475504381,
+      "learning_rate": 9.148832839907871e-06,
+      "loss": 0.2898,
+      "step": 242
+    },
+    {
+      "epoch": 0.37850467289719625,
+      "grad_norm": 0.5688149258379418,
+      "learning_rate": 9.141992717410697e-06,
+      "loss": 0.239,
+      "step": 243
+    },
+    {
+      "epoch": 0.38006230529595014,
+      "grad_norm": 0.6629923621745606,
+      "learning_rate": 9.135127799074271e-06,
+      "loss": 0.2614,
+      "step": 244
+    },
+    {
+      "epoch": 0.38161993769470404,
+      "grad_norm": 0.6300463035343699,
+      "learning_rate": 9.128238125995093e-06,
+      "loss": 0.24,
+      "step": 245
+    },
+    {
+      "epoch": 0.38317757009345793,
+      "grad_norm": 0.643027164394109,
+      "learning_rate": 9.121323739417858e-06,
+      "loss": 0.2365,
+      "step": 246
+    },
+    {
+      "epoch": 0.3847352024922118,
+      "grad_norm": 2.8455942968073282,
+      "learning_rate": 9.114384680735211e-06,
+      "loss": 0.7814,
+      "step": 247
+    },
+    {
+      "epoch": 0.3862928348909657,
+      "grad_norm": 0.630104563255545,
+      "learning_rate": 9.107420991487488e-06,
+      "loss": 0.2234,
+      "step": 248
+    },
+    {
+      "epoch": 0.3878504672897196,
+      "grad_norm": 0.905027542232206,
+      "learning_rate": 9.100432713362477e-06,
+      "loss": 0.3142,
+      "step": 249
+    },
+    {
+      "epoch": 0.3894080996884735,
+      "grad_norm": 0.6960898261331736,
+      "learning_rate": 9.09341988819517e-06,
+      "loss": 0.3195,
+      "step": 250
+    },
+    {
+      "epoch": 0.3909657320872274,
+      "grad_norm": 0.7350311907562319,
+      "learning_rate": 9.086382557967507e-06,
+      "loss": 0.2479,
+      "step": 251
+    },
+    {
+      "epoch": 0.3925233644859813,
+      "grad_norm": 0.7212085527913289,
+      "learning_rate": 9.07932076480812e-06,
+      "loss": 0.2334,
+      "step": 252
+    },
+    {
+      "epoch": 0.3940809968847352,
+      "grad_norm": 0.6340762358232386,
+      "learning_rate": 9.072234550992099e-06,
+      "loss": 0.282,
+      "step": 253
+    },
+    {
+      "epoch": 0.3956386292834891,
+      "grad_norm": 0.5688729860657064,
+      "learning_rate": 9.065123958940716e-06,
+      "loss": 0.2928,
+      "step": 254
+    },
+    {
+      "epoch": 0.397196261682243,
+      "grad_norm": 0.8922598360101682,
+      "learning_rate": 9.057989031221187e-06,
+      "loss": 0.281,
+      "step": 255
+    },
+    {
+      "epoch": 0.3987538940809969,
+      "grad_norm": 0.7114134585553089,
+      "learning_rate": 9.050829810546416e-06,
+      "loss": 0.2635,
+      "step": 256
+    },
+    {
+      "epoch": 0.40031152647975077,
+      "grad_norm": 0.6966492509810489,
+      "learning_rate": 9.043646339774726e-06,
+      "loss": 0.3394,
+      "step": 257
+    },
+    {
+      "epoch": 0.40186915887850466,
+      "grad_norm": 0.5958512506208158,
+      "learning_rate": 9.036438661909624e-06,
+      "loss": 0.2718,
+      "step": 258
+    },
+    {
+      "epoch": 0.40342679127725856,
+      "grad_norm": 0.9116531361266647,
+      "learning_rate": 9.029206820099518e-06,
+      "loss": 0.2785,
+      "step": 259
+    },
+    {
+      "epoch": 0.40498442367601245,
+      "grad_norm": 0.5887944110529719,
+      "learning_rate": 9.021950857637486e-06,
+      "loss": 0.2668,
+      "step": 260
+    },
+    {
+      "epoch": 0.40654205607476634,
+      "grad_norm": 0.6108632031180715,
+      "learning_rate": 9.014670817960993e-06,
+      "loss": 0.1969,
+      "step": 261
+    },
+    {
+      "epoch": 0.40809968847352024,
+      "grad_norm": 0.6907915569936867,
+      "learning_rate": 9.007366744651646e-06,
+      "loss": 0.2268,
+      "step": 262
+    },
+    {
+      "epoch": 0.40965732087227413,
+      "grad_norm": 0.5793225985517833,
+      "learning_rate": 9.000038681434925e-06,
+      "loss": 0.1931,
+      "step": 263
+    },
+    {
+      "epoch": 0.411214953271028,
+      "grad_norm": 1.0784967522591073,
+      "learning_rate": 8.99268667217993e-06,
+      "loss": 0.2329,
+      "step": 264
+    },
+    {
+      "epoch": 0.4127725856697819,
+      "grad_norm": 0.602627598917929,
+      "learning_rate": 8.985310760899108e-06,
+      "loss": 0.2574,
+      "step": 265
+    },
+    {
+      "epoch": 0.4143302180685358,
+      "grad_norm": 0.6724448335681737,
+      "learning_rate": 8.977910991747993e-06,
+      "loss": 0.2781,
+      "step": 266
+    },
+    {
+      "epoch": 0.4158878504672897,
+      "grad_norm": 0.713294514234468,
+      "learning_rate": 8.970487409024949e-06,
+      "loss": 0.2284,
+      "step": 267
+    },
+    {
+      "epoch": 0.4174454828660436,
+      "grad_norm": 0.8427860833648895,
+      "learning_rate": 8.96304005717089e-06,
+      "loss": 0.2953,
+      "step": 268
+    },
+    {
+      "epoch": 0.4190031152647975,
+      "grad_norm": 0.5840913639256132,
+      "learning_rate": 8.955568980769033e-06,
+      "loss": 0.2974,
+      "step": 269
+    },
+    {
+      "epoch": 0.4205607476635514,
+      "grad_norm": 0.6131987426518342,
+      "learning_rate": 8.948074224544615e-06,
+      "loss": 0.2773,
+      "step": 270
+    },
+    {
+      "epoch": 0.4221183800623053,
+      "grad_norm": 0.5947515631497642,
+      "learning_rate": 8.94055583336463e-06,
+      "loss": 0.275,
+      "step": 271
+    },
+    {
+      "epoch": 0.4236760124610592,
+      "grad_norm": 0.5946955459161034,
+      "learning_rate": 8.933013852237564e-06,
+      "loss": 0.3307,
+      "step": 272
+    },
+    {
+      "epoch": 0.4252336448598131,
+      "grad_norm": 0.7451920245650221,
+      "learning_rate": 8.925448326313125e-06,
+      "loss": 0.286,
+      "step": 273
+    },
+    {
+      "epoch": 0.42679127725856697,
+      "grad_norm": 0.924856917547545,
+      "learning_rate": 8.917859300881965e-06,
+      "loss": 0.2879,
+      "step": 274
+    },
+    {
+      "epoch": 0.42834890965732086,
+      "grad_norm": 1.0327909293428048,
+      "learning_rate": 8.91024682137542e-06,
+      "loss": 0.2408,
+      "step": 275
+    },
+    {
+      "epoch": 0.42990654205607476,
+      "grad_norm": 0.880531058290941,
+      "learning_rate": 8.90261093336523e-06,
+      "loss": 0.3045,
+      "step": 276
+    },
+    {
+      "epoch": 0.43146417445482865,
+      "grad_norm": 0.580559025082547,
+      "learning_rate": 8.89495168256327e-06,
+      "loss": 0.2805,
+      "step": 277
+    },
+    {
+      "epoch": 0.43302180685358255,
+      "grad_norm": 0.5475610266945151,
+      "learning_rate": 8.887269114821275e-06,
+      "loss": 0.2338,
+      "step": 278
+    },
+    {
+      "epoch": 0.43457943925233644,
+      "grad_norm": 0.6043881248930308,
+      "learning_rate": 8.879563276130567e-06,
+      "loss": 0.2516,
+      "step": 279
+    },
+    {
+      "epoch": 0.43613707165109034,
+      "grad_norm": 0.6621273162261064,
+      "learning_rate": 8.871834212621773e-06,
+      "loss": 0.2485,
+      "step": 280
+    },
+    {
+      "epoch": 0.43769470404984423,
+      "grad_norm": 0.6108881149273678,
+      "learning_rate": 8.86408197056456e-06,
+      "loss": 0.2082,
+      "step": 281
+    },
+    {
+      "epoch": 0.4392523364485981,
+      "grad_norm": 0.6626887696196829,
+      "learning_rate": 8.856306596367351e-06,
+      "loss": 0.2741,
+      "step": 282
+    },
+    {
+      "epoch": 0.440809968847352,
+      "grad_norm": 0.6517640250027363,
+      "learning_rate": 8.84850813657705e-06,
+      "loss": 0.3733,
+      "step": 283
+    },
+    {
+      "epoch": 0.4423676012461059,
+      "grad_norm": 0.5979891922361272,
+      "learning_rate": 8.840686637878756e-06,
+      "loss": 0.2615,
+      "step": 284
+    },
+    {
+      "epoch": 0.4439252336448598,
+      "grad_norm": 0.5447033448285953,
+      "learning_rate": 8.832842147095495e-06,
+      "loss": 0.2204,
+      "step": 285
+    },
+    {
+      "epoch": 0.4454828660436137,
+      "grad_norm": 0.6176926760203407,
+      "learning_rate": 8.82497471118793e-06,
+      "loss": 0.259,
+      "step": 286
+    },
+    {
+      "epoch": 0.4470404984423676,
+      "grad_norm": 0.6803871655362536,
+      "learning_rate": 8.817084377254089e-06,
+      "loss": 0.3406,
+      "step": 287
+    },
+    {
+      "epoch": 0.4485981308411215,
+      "grad_norm": 1.0924030353117906,
+      "learning_rate": 8.809171192529074e-06,
+      "loss": 0.7712,
+      "step": 288
+    },
+    {
+      "epoch": 0.4501557632398754,
+      "grad_norm": 0.7261898907998104,
+      "learning_rate": 8.80123520438478e-06,
+      "loss": 0.2439,
+      "step": 289
+    },
+    {
+      "epoch": 0.4517133956386293,
+      "grad_norm": 0.5278888030187282,
+      "learning_rate": 8.79327646032962e-06,
+      "loss": 0.2838,
+      "step": 290
+    },
+    {
+      "epoch": 0.4532710280373832,
+      "grad_norm": 0.7336946619047033,
+      "learning_rate": 8.785295008008227e-06,
+      "loss": 0.2863,
+      "step": 291
+    },
+    {
+      "epoch": 0.45482866043613707,
+      "grad_norm": 0.5663680560662904,
+      "learning_rate": 8.777290895201186e-06,
+      "loss": 0.2487,
+      "step": 292
+    },
+    {
+      "epoch": 0.45638629283489096,
+      "grad_norm": 0.6674793012574268,
+      "learning_rate": 8.769264169824725e-06,
+      "loss": 0.2115,
+      "step": 293
+    },
+    {
+      "epoch": 0.45794392523364486,
+      "grad_norm": 0.629476332991769,
+      "learning_rate": 8.761214879930452e-06,
+      "loss": 0.2698,
+      "step": 294
+    },
+    {
+      "epoch": 0.45950155763239875,
+      "grad_norm": 0.5733049643694853,
+      "learning_rate": 8.753143073705048e-06,
+      "loss": 0.2755,
+      "step": 295
+    },
+    {
+      "epoch": 0.46105919003115264,
+      "grad_norm": 0.7496857759281685,
+      "learning_rate": 8.745048799469996e-06,
+      "loss": 0.2066,
+      "step": 296
+    },
+    {
+      "epoch": 0.46261682242990654,
+      "grad_norm": 0.5833814287468523,
+      "learning_rate": 8.736932105681272e-06,
+      "loss": 0.2233,
+      "step": 297
+    },
+    {
+      "epoch": 0.46417445482866043,
+      "grad_norm": 0.7333726802931715,
+      "learning_rate": 8.728793040929075e-06,
+      "loss": 0.2529,
+      "step": 298
+    },
+    {
+      "epoch": 0.4657320872274143,
+      "grad_norm": 0.6752352954357181,
+      "learning_rate": 8.720631653937522e-06,
+      "loss": 0.2487,
+      "step": 299
+    },
+    {
+      "epoch": 0.4672897196261682,
+      "grad_norm": 0.6281610479983413,
+      "learning_rate": 8.712447993564362e-06,
+      "loss": 0.2513,
+      "step": 300
+    },
+    {
+      "epoch": 0.4688473520249221,
+      "grad_norm": 1.719855082212449,
+      "learning_rate": 8.70424210880068e-06,
+      "loss": 1.2841,
+      "step": 301
+    },
+    {
+      "epoch": 0.470404984423676,
+      "grad_norm": 0.8697991461292468,
+      "learning_rate": 8.696014048770611e-06,
+      "loss": 0.1752,
+      "step": 302
+    },
+    {
+      "epoch": 0.4719626168224299,
+      "grad_norm": 0.7388848487578813,
+      "learning_rate": 8.68776386273104e-06,
+      "loss": 0.3559,
+      "step": 303
+    },
+    {
+      "epoch": 0.4735202492211838,
+      "grad_norm": 0.6740217209253545,
+      "learning_rate": 8.679491600071304e-06,
+      "loss": 0.2506,
+      "step": 304
+    },
+    {
+      "epoch": 0.4750778816199377,
+      "grad_norm": 0.8324393939489836,
+      "learning_rate": 8.671197310312905e-06,
+      "loss": 0.2951,
+      "step": 305
+    },
+    {
+      "epoch": 0.4766355140186916,
+      "grad_norm": 0.5859142473050233,
+      "learning_rate": 8.662881043109208e-06,
+      "loss": 0.2737,
+      "step": 306
+    },
+    {
+      "epoch": 0.4781931464174455,
+      "grad_norm": 0.5144143552921643,
+      "learning_rate": 8.654542848245146e-06,
+      "loss": 0.2555,
+      "step": 307
+    },
+    {
+      "epoch": 0.4797507788161994,
+      "grad_norm": 0.9870060592323339,
+      "learning_rate": 8.646182775636917e-06,
+      "loss": 0.3268,
+      "step": 308
+    },
+    {
+      "epoch": 0.48130841121495327,
+      "grad_norm": 1.22444825315241,
+      "learning_rate": 8.637800875331693e-06,
+      "loss": 0.8927,
+      "step": 309
+    },
+    {
+      "epoch": 0.48286604361370716,
+      "grad_norm": 0.6219488136095235,
+      "learning_rate": 8.629397197507315e-06,
+      "loss": 0.1925,
+      "step": 310
+    },
+    {
+      "epoch": 0.48442367601246106,
+      "grad_norm": 0.6663733783513853,
+      "learning_rate": 8.620971792471994e-06,
+      "loss": 0.2525,
+      "step": 311
+    },
+    {
+      "epoch": 0.48598130841121495,
+      "grad_norm": 0.7779309034351384,
+      "learning_rate": 8.612524710664012e-06,
+      "loss": 0.2857,
+      "step": 312
+    },
+    {
+      "epoch": 0.48753894080996885,
+      "grad_norm": 0.5815264215390993,
+      "learning_rate": 8.604056002651414e-06,
+      "loss": 0.2132,
+      "step": 313
+    },
+    {
+      "epoch": 0.48909657320872274,
+      "grad_norm": 0.639985193258881,
+      "learning_rate": 8.595565719131711e-06,
+      "loss": 0.2927,
+      "step": 314
+    },
+    {
+      "epoch": 0.49065420560747663,
+      "grad_norm": 0.5843630349664094,
+      "learning_rate": 8.587053910931576e-06,
+      "loss": 0.2073,
+      "step": 315
+    },
+    {
+      "epoch": 0.49221183800623053,
+      "grad_norm": 0.5718073343739106,
+      "learning_rate": 8.578520629006537e-06,
+      "loss": 0.2701,
+      "step": 316
+    },
+    {
+      "epoch": 0.4937694704049844,
+      "grad_norm": 0.8093573764947887,
+      "learning_rate": 8.569965924440675e-06,
+      "loss": 0.3701,
+      "step": 317
+    },
+    {
+      "epoch": 0.4953271028037383,
+      "grad_norm": 0.45875585137104996,
+      "learning_rate": 8.561389848446314e-06,
+      "loss": 0.2535,
+      "step": 318
+    },
+    {
+      "epoch": 0.4968847352024922,
+      "grad_norm": 0.6800533406672229,
+      "learning_rate": 8.55279245236372e-06,
+      "loss": 0.2542,
+      "step": 319
+    },
+    {
+      "epoch": 0.4984423676012461,
+      "grad_norm": 0.5791491500061517,
+      "learning_rate": 8.544173787660788e-06,
+      "loss": 0.2233,
+      "step": 320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6853895685598861,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.1762,
+      "step": 321
+    },
+    {
+      "epoch": 0.5015576323987538,
+      "grad_norm": 0.5400182911842227,
+      "learning_rate": 8.526872858901806e-06,
+      "loss": 0.2565,
+      "step": 322
+    },
+    {
+      "epoch": 0.5031152647975078,
+      "grad_norm": 0.7099762938056243,
+      "learning_rate": 8.518190698416929e-06,
+      "loss": 0.2901,
+      "step": 323
+    },
+    {
+      "epoch": 0.5046728971962616,
+      "grad_norm": 0.7351812911105055,
+      "learning_rate": 8.509487476453442e-06,
+      "loss": 0.2829,
+      "step": 324
+    },
+    {
+      "epoch": 0.5062305295950156,
+      "grad_norm": 0.6208369210464704,
+      "learning_rate": 8.500763245112763e-06,
+      "loss": 0.2916,
+      "step": 325
+    },
+    {
+      "epoch": 0.5077881619937694,
+      "grad_norm": 0.7390752110429559,
+      "learning_rate": 8.492018056622083e-06,
+      "loss": 0.3912,
+      "step": 326
+    },
+    {
+      "epoch": 0.5093457943925234,
+      "grad_norm": 0.5930641088378493,
+      "learning_rate": 8.483251963334047e-06,
+      "loss": 0.2193,
+      "step": 327
+    },
+    {
+      "epoch": 0.5109034267912772,
+      "grad_norm": 0.6344017285315496,
+      "learning_rate": 8.474465017726452e-06,
+      "loss": 0.2191,
+      "step": 328
+    },
+    {
+      "epoch": 0.5124610591900312,
+      "grad_norm": 0.7250776947740573,
+      "learning_rate": 8.465657272401921e-06,
+      "loss": 0.2435,
+      "step": 329
+    },
+    {
+      "epoch": 0.514018691588785,
+      "grad_norm": 0.6175903219629659,
+      "learning_rate": 8.456828780087598e-06,
+      "loss": 0.2005,
+      "step": 330
+    },
+    {
+      "epoch": 0.5155763239875389,
+      "grad_norm": 0.6186407482326058,
+      "learning_rate": 8.447979593634823e-06,
+      "loss": 0.2969,
+      "step": 331
+    },
+    {
+      "epoch": 0.5171339563862928,
+      "grad_norm": 0.6221498367824214,
+      "learning_rate": 8.439109766018825e-06,
+      "loss": 0.2532,
+      "step": 332
+    },
+    {
+      "epoch": 0.5186915887850467,
+      "grad_norm": 0.5991537960888588,
+      "learning_rate": 8.430219350338398e-06,
+      "loss": 0.2029,
+      "step": 333
+    },
+    {
+      "epoch": 0.5202492211838006,
+      "grad_norm": 0.5074549508649242,
+      "learning_rate": 8.421308399815586e-06,
+      "loss": 0.2452,
+      "step": 334
+    },
+    {
+      "epoch": 0.5218068535825545,
+      "grad_norm": 1.0447367430008974,
+      "learning_rate": 8.412376967795362e-06,
+      "loss": 0.7821,
+      "step": 335
+    },
+    {
+      "epoch": 0.5233644859813084,
+      "grad_norm": 0.6715190449661054,
+      "learning_rate": 8.403425107745315e-06,
+      "loss": 0.1873,
+      "step": 336
+    },
+    {
+      "epoch": 0.5249221183800623,
+      "grad_norm": 0.8046073365053287,
+      "learning_rate": 8.394452873255321e-06,
+      "loss": 0.2459,
+      "step": 337
+    },
+    {
+      "epoch": 0.5264797507788161,
+      "grad_norm": 0.6185327399453182,
+      "learning_rate": 8.385460318037228e-06,
+      "loss": 0.2347,
+      "step": 338
+    },
+    {
+      "epoch": 0.5280373831775701,
+      "grad_norm": 0.6523073370428902,
+      "learning_rate": 8.376447495924533e-06,
+      "loss": 0.2494,
+      "step": 339
+    },
+    {
+      "epoch": 0.5295950155763239,
+      "grad_norm": 0.8230020627407991,
+      "learning_rate": 8.367414460872064e-06,
+      "loss": 0.7898,
+      "step": 340
+    },
+    {
+      "epoch": 0.5311526479750779,
+      "grad_norm": 0.6739918439315772,
+      "learning_rate": 8.358361266955641e-06,
+      "loss": 0.3159,
+      "step": 341
+    },
+    {
+      "epoch": 0.5327102803738317,
+      "grad_norm": 0.7519773686193855,
+      "learning_rate": 8.34928796837178e-06,
+      "loss": 0.2593,
+      "step": 342
+    },
+    {
+      "epoch": 0.5342679127725857,
+      "grad_norm": 0.5445971280446563,
+      "learning_rate": 8.34019461943734e-06,
+      "loss": 0.2923,
+      "step": 343
+    },
+    {
+      "epoch": 0.5358255451713395,
+      "grad_norm": 0.7071155258569332,
+      "learning_rate": 8.331081274589217e-06,
+      "loss": 0.2598,
+      "step": 344
+    },
+    {
+      "epoch": 0.5373831775700935,
+      "grad_norm": 0.719502132159577,
+      "learning_rate": 8.321947988384006e-06,
+      "loss": 0.276,
+      "step": 345
+    },
+    {
+      "epoch": 0.5389408099688473,
+      "grad_norm": 0.5170866766320925,
+      "learning_rate": 8.312794815497688e-06,
+      "loss": 0.2236,
+      "step": 346
+    },
+    {
+      "epoch": 0.5404984423676013,
+      "grad_norm": 0.6654228108717598,
+      "learning_rate": 8.303621810725287e-06,
+      "loss": 0.2666,
+      "step": 347
+    },
+    {
+      "epoch": 0.5420560747663551,
+      "grad_norm": 0.6117403127181333,
+      "learning_rate": 8.294429028980555e-06,
+      "loss": 0.2767,
+      "step": 348
+    },
+    {
+      "epoch": 0.543613707165109,
+      "grad_norm": 0.8584300691210696,
+      "learning_rate": 8.285216525295636e-06,
+      "loss": 0.2521,
+      "step": 349
+    },
+    {
+      "epoch": 0.5451713395638629,
+      "grad_norm": 0.7222215782352946,
+      "learning_rate": 8.275984354820736e-06,
+      "loss": 0.2499,
+      "step": 350
+    },
+    {
+      "epoch": 0.5467289719626168,
+      "grad_norm": 0.643947565152816,
+      "learning_rate": 8.266732572823799e-06,
+      "loss": 0.2092,
+      "step": 351
+    },
+    {
+      "epoch": 0.5482866043613707,
+      "grad_norm": 0.6231517894213087,
+      "learning_rate": 8.25746123469017e-06,
+      "loss": 0.2598,
+      "step": 352
+    },
+    {
+      "epoch": 0.5498442367601246,
+      "grad_norm": 0.5983797307315736,
+      "learning_rate": 8.248170395922266e-06,
+      "loss": 0.3241,
+      "step": 353
+    },
+    {
+      "epoch": 0.5514018691588785,
+      "grad_norm": 0.6611331737862588,
+      "learning_rate": 8.238860112139246e-06,
+      "loss": 0.2292,
+      "step": 354
+    },
+    {
+      "epoch": 0.5529595015576324,
+      "grad_norm": 0.667013942050513,
+      "learning_rate": 8.229530439076674e-06,
+      "loss": 0.2911,
+      "step": 355
+    },
+    {
+      "epoch": 0.5545171339563862,
+      "grad_norm": 0.5711025724816864,
+      "learning_rate": 8.220181432586187e-06,
+      "loss": 0.3828,
+      "step": 356
+    },
+    {
+      "epoch": 0.5560747663551402,
+      "grad_norm": 0.7326983314478233,
+      "learning_rate": 8.210813148635158e-06,
+      "loss": 0.4075,
+      "step": 357
+    },
+    {
+      "epoch": 0.557632398753894,
+      "grad_norm": 0.5806269356738779,
+      "learning_rate": 8.201425643306367e-06,
+      "loss": 0.2283,
+      "step": 358
+    },
+    {
+      "epoch": 0.559190031152648,
+      "grad_norm": 0.8188821554767335,
+      "learning_rate": 8.192018972797665e-06,
+      "loss": 0.2549,
+      "step": 359
+    },
+    {
+      "epoch": 0.5607476635514018,
+      "grad_norm": 0.6809831648038962,
+      "learning_rate": 8.182593193421625e-06,
+      "loss": 0.2543,
+      "step": 360
+    },
+    {
+      "epoch": 0.5623052959501558,
+      "grad_norm": 0.6203982930897892,
+      "learning_rate": 8.173148361605224e-06,
+      "loss": 0.3205,
+      "step": 361
+    },
+    {
+      "epoch": 0.5638629283489096,
+      "grad_norm": 0.4686220806935477,
+      "learning_rate": 8.163684533889489e-06,
+      "loss": 0.2356,
+      "step": 362
+    },
+    {
+      "epoch": 0.5654205607476636,
+      "grad_norm": 0.7261908034014758,
+      "learning_rate": 8.154201766929167e-06,
+      "loss": 0.2763,
+      "step": 363
+    },
+    {
+      "epoch": 0.5669781931464174,
+      "grad_norm": 0.6301811228833201,
+      "learning_rate": 8.144700117492386e-06,
+      "loss": 0.2369,
+      "step": 364
+    },
+    {
+      "epoch": 0.5685358255451713,
+      "grad_norm": 0.7734460446532929,
+      "learning_rate": 8.135179642460308e-06,
+      "loss": 0.2722,
+      "step": 365
+    },
+    {
+      "epoch": 0.5700934579439252,
+      "grad_norm": 0.6997558366317673,
+      "learning_rate": 8.125640398826803e-06,
+      "loss": 0.2543,
+      "step": 366
+    },
+    {
+      "epoch": 0.5716510903426791,
+      "grad_norm": 0.8988136993440748,
+      "learning_rate": 8.116082443698085e-06,
+      "loss": 0.2835,
+      "step": 367
+    },
+    {
+      "epoch": 0.573208722741433,
+      "grad_norm": 0.8141909161490042,
+      "learning_rate": 8.106505834292396e-06,
+      "loss": 0.2538,
+      "step": 368
+    },
+    {
+      "epoch": 0.5747663551401869,
+      "grad_norm": 1.3911734406138567,
+      "learning_rate": 8.09691062793964e-06,
+      "loss": 0.7141,
+      "step": 369
+    },
+    {
+      "epoch": 0.5763239875389408,
+      "grad_norm": 0.6359597508417911,
+      "learning_rate": 8.087296882081062e-06,
+      "loss": 0.2794,
+      "step": 370
+    },
+    {
+      "epoch": 0.5778816199376947,
+      "grad_norm": 0.5841003451023182,
+      "learning_rate": 8.077664654268883e-06,
+      "loss": 0.2973,
+      "step": 371
+    },
+    {
+      "epoch": 0.5794392523364486,
+      "grad_norm": 0.6441301027816297,
+      "learning_rate": 8.06801400216597e-06,
+      "loss": 0.2477,
+      "step": 372
+    },
+    {
+      "epoch": 0.5809968847352025,
+      "grad_norm": 0.6922555735413835,
+      "learning_rate": 8.058344983545486e-06,
+      "loss": 0.2524,
+      "step": 373
+    },
+    {
+      "epoch": 0.5825545171339563,
+      "grad_norm": 0.6083842427437397,
+      "learning_rate": 8.048657656290545e-06,
+      "loss": 0.2837,
+      "step": 374
+    },
+    {
+      "epoch": 0.5841121495327103,
+      "grad_norm": 0.5803164555571954,
+      "learning_rate": 8.03895207839386e-06,
+      "loss": 0.3084,
+      "step": 375
+    },
+    {
+      "epoch": 0.5856697819314641,
+      "grad_norm": 0.8102916914102428,
+      "learning_rate": 8.029228307957408e-06,
+      "loss": 0.2462,
+      "step": 376
+    },
+    {
+      "epoch": 0.5872274143302181,
+      "grad_norm": 0.7061482090328073,
+      "learning_rate": 8.019486403192069e-06,
+      "loss": 0.2487,
+      "step": 377
+    },
+    {
+      "epoch": 0.5887850467289719,
+      "grad_norm": 0.5113763283575798,
+      "learning_rate": 8.009726422417286e-06,
+      "loss": 0.2931,
+      "step": 378
+    },
+    {
+      "epoch": 0.5903426791277259,
+      "grad_norm": 0.6897001371312232,
+      "learning_rate": 7.99994842406071e-06,
+      "loss": 0.2186,
+      "step": 379
+    },
+    {
+      "epoch": 0.5919003115264797,
+      "grad_norm": 0.9072221405351328,
+      "learning_rate": 7.99015246665786e-06,
+      "loss": 0.8348,
+      "step": 380
+    },
+    {
+      "epoch": 0.5934579439252337,
+      "grad_norm": 0.6531934878590409,
+      "learning_rate": 7.980338608851756e-06,
+      "loss": 0.2228,
+      "step": 381
+    },
+    {
+      "epoch": 0.5950155763239875,
+      "grad_norm": 0.7327462885118111,
+      "learning_rate": 7.970506909392588e-06,
+      "loss": 0.2433,
+      "step": 382
+    },
+    {
+      "epoch": 0.5965732087227414,
+      "grad_norm": 1.1643513095965004,
+      "learning_rate": 7.960657427137347e-06,
+      "loss": 0.3147,
+      "step": 383
+    },
+    {
+      "epoch": 0.5981308411214953,
+      "grad_norm": 0.6452843748972524,
+      "learning_rate": 7.950790221049485e-06,
+      "loss": 0.3303,
+      "step": 384
+    },
+    {
+      "epoch": 0.5996884735202492,
+      "grad_norm": 0.7274077390061954,
+      "learning_rate": 7.940905350198553e-06,
+      "loss": 0.3347,
+      "step": 385
+    },
+    {
+      "epoch": 0.6012461059190031,
+      "grad_norm": 0.6178112722471735,
+      "learning_rate": 7.931002873759852e-06,
+      "loss": 0.2276,
+      "step": 386
+    },
+    {
+      "epoch": 0.602803738317757,
+      "grad_norm": 0.7461711392201733,
+      "learning_rate": 7.921082851014079e-06,
+      "loss": 0.2634,
+      "step": 387
+    },
+    {
+      "epoch": 0.6043613707165109,
+      "grad_norm": 0.8795743223985983,
+      "learning_rate": 7.911145341346972e-06,
+      "loss": 0.2896,
+      "step": 388
+    },
+    {
+      "epoch": 0.6059190031152648,
+      "grad_norm": 0.8319649658997385,
+      "learning_rate": 7.901190404248952e-06,
+      "loss": 0.7066,
+      "step": 389
+    },
+    {
+      "epoch": 0.6074766355140186,
+      "grad_norm": 0.5267541152010679,
+      "learning_rate": 7.89121809931477e-06,
+      "loss": 0.2416,
+      "step": 390
+    },
+    {
+      "epoch": 0.6090342679127726,
+      "grad_norm": 0.46625213633043144,
+      "learning_rate": 7.881228486243144e-06,
+      "loss": 0.2482,
+      "step": 391
+    },
+    {
+      "epoch": 0.6105919003115264,
+      "grad_norm": 0.6924717742437231,
+      "learning_rate": 7.871221624836414e-06,
+      "loss": 0.3523,
+      "step": 392
+    },
+    {
+      "epoch": 0.6121495327102804,
+      "grad_norm": 0.5838406566254681,
+      "learning_rate": 7.861197575000168e-06,
+      "loss": 0.228,
+      "step": 393
+    },
+    {
+      "epoch": 0.6137071651090342,
+      "grad_norm": 0.8869116056189287,
+      "learning_rate": 7.8511563967429e-06,
+      "loss": 0.2496,
+      "step": 394
+    },
+    {
+      "epoch": 0.6152647975077882,
+      "grad_norm": 0.45593356473166835,
+      "learning_rate": 7.841098150175636e-06,
+      "loss": 0.2643,
+      "step": 395
+    },
+    {
+      "epoch": 0.616822429906542,
+      "grad_norm": 0.8435131653971922,
+      "learning_rate": 7.831022895511586e-06,
+      "loss": 0.2496,
+      "step": 396
+    },
+    {
+      "epoch": 0.618380062305296,
+      "grad_norm": 0.4987004271913618,
+      "learning_rate": 7.820930693065771e-06,
+      "loss": 0.2063,
+      "step": 397
+    },
+    {
+      "epoch": 0.6199376947040498,
+      "grad_norm": 0.7942215586211049,
+      "learning_rate": 7.810821603254677e-06,
+      "loss": 0.2181,
+      "step": 398
+    },
+    {
+      "epoch": 0.6214953271028038,
+      "grad_norm": 0.6794367411830937,
+      "learning_rate": 7.800695686595879e-06,
+      "loss": 0.3009,
+      "step": 399
+    },
+    {
+      "epoch": 0.6230529595015576,
+      "grad_norm": 0.7273130133089402,
+      "learning_rate": 7.790553003707691e-06,
+      "loss": 0.2469,
+      "step": 400
+    },
+    {
+      "epoch": 0.6246105919003115,
+      "grad_norm": 0.6902330948327355,
+      "learning_rate": 7.780393615308787e-06,
+      "loss": 0.2508,
+      "step": 401
+    },
+    {
+      "epoch": 0.6261682242990654,
+      "grad_norm": 0.7084362576430354,
+      "learning_rate": 7.770217582217863e-06,
+      "loss": 0.2551,
+      "step": 402
+    },
+    {
+      "epoch": 0.6277258566978193,
+      "grad_norm": 0.8029205384416798,
+      "learning_rate": 7.760024965353246e-06,
+      "loss": 0.2333,
+      "step": 403
+    },
+    {
+      "epoch": 0.6292834890965732,
+      "grad_norm": 0.6112029983652504,
+      "learning_rate": 7.749815825732543e-06,
+      "loss": 0.298,
+      "step": 404
+    },
+    {
+      "epoch": 0.6308411214953271,
+      "grad_norm": 0.7494581341489577,
+      "learning_rate": 7.739590224472275e-06,
+      "loss": 0.3462,
+      "step": 405
+    },
+    {
+      "epoch": 0.632398753894081,
+      "grad_norm": 0.7347669711126691,
+      "learning_rate": 7.729348222787514e-06,
+      "loss": 0.3149,
+      "step": 406
+    },
+    {
+      "epoch": 0.6339563862928349,
+      "grad_norm": 0.6796064286407987,
+      "learning_rate": 7.719089881991503e-06,
+      "loss": 0.2873,
+      "step": 407
+    },
+    {
+      "epoch": 0.6355140186915887,
+      "grad_norm": 0.7425509324765857,
+      "learning_rate": 7.708815263495307e-06,
+      "loss": 0.3278,
+      "step": 408
+    },
+    {
+      "epoch": 0.6370716510903427,
+      "grad_norm": 0.609414275478013,
+      "learning_rate": 7.698524428807431e-06,
+      "loss": 0.2708,
+      "step": 409
+    },
+    {
+      "epoch": 0.6386292834890965,
+      "grad_norm": 0.7757117977400942,
+      "learning_rate": 7.68821743953346e-06,
+      "loss": 0.2555,
+      "step": 410
+    },
+    {
+      "epoch": 0.6401869158878505,
+      "grad_norm": 0.6642687790623766,
+      "learning_rate": 7.677894357375689e-06,
+      "loss": 0.3625,
+      "step": 411
+    },
+    {
+      "epoch": 0.6417445482866043,
+      "grad_norm": 0.5791966784356082,
+      "learning_rate": 7.667555244132749e-06,
+      "loss": 0.2661,
+      "step": 412
+    },
+    {
+      "epoch": 0.6433021806853583,
+      "grad_norm": 0.5594732951892226,
+      "learning_rate": 7.65720016169924e-06,
+      "loss": 0.2995,
+      "step": 413
+    },
+    {
+      "epoch": 0.6448598130841121,
+      "grad_norm": 0.6021900759219545,
+      "learning_rate": 7.646829172065367e-06,
+      "loss": 0.3099,
+      "step": 414
+    },
+    {
+      "epoch": 0.6464174454828661,
+      "grad_norm": 0.5562483872284556,
+      "learning_rate": 7.636442337316555e-06,
+      "loss": 0.2376,
+      "step": 415
+    },
+    {
+      "epoch": 0.6479750778816199,
+      "grad_norm": 0.5829741964791303,
+      "learning_rate": 7.6260397196330895e-06,
+      "loss": 0.2774,
+      "step": 416
+    },
+    {
+      "epoch": 0.6495327102803738,
+      "grad_norm": 0.7958468559537486,
+      "learning_rate": 7.615621381289737e-06,
+      "loss": 0.2316,
+      "step": 417
+    },
+    {
+      "epoch": 0.6510903426791277,
+      "grad_norm": 0.6088648202059304,
+      "learning_rate": 7.6051873846553795e-06,
+      "loss": 0.33,
+      "step": 418
+    },
+    {
+      "epoch": 0.6526479750778816,
+      "grad_norm": 0.523142097250351,
+      "learning_rate": 7.594737792192629e-06,
+      "loss": 0.2589,
+      "step": 419
+    },
+    {
+      "epoch": 0.6542056074766355,
+      "grad_norm": 0.6233483869502079,
+      "learning_rate": 7.584272666457471e-06,
+      "loss": 0.2409,
+      "step": 420
+    },
+    {
+      "epoch": 0.6557632398753894,
+      "grad_norm": 0.8831745439168878,
+      "learning_rate": 7.573792070098873e-06,
+      "loss": 0.3156,
+      "step": 421
+    },
+    {
+      "epoch": 0.6573208722741433,
+      "grad_norm": 0.646939395915981,
+      "learning_rate": 7.5632960658584184e-06,
+      "loss": 0.1882,
+      "step": 422
+    },
+    {
+      "epoch": 0.6588785046728972,
+      "grad_norm": 0.7493677482681486,
+      "learning_rate": 7.5527847165699295e-06,
+      "loss": 0.2533,
+      "step": 423
+    },
+    {
+      "epoch": 0.660436137071651,
+      "grad_norm": 0.6895089900125264,
+      "learning_rate": 7.542258085159091e-06,
+      "loss": 0.2239,
+      "step": 424
+    },
+    {
+      "epoch": 0.661993769470405,
+      "grad_norm": 0.8218899831192643,
+      "learning_rate": 7.531716234643071e-06,
+      "loss": 0.3025,
+      "step": 425
+    },
+    {
+      "epoch": 0.6635514018691588,
+      "grad_norm": 0.8511691627825192,
+      "learning_rate": 7.5211592281301525e-06,
+      "loss": 0.2081,
+      "step": 426
+    },
+    {
+      "epoch": 0.6651090342679128,
+      "grad_norm": 0.6202374314769092,
+      "learning_rate": 7.510587128819341e-06,
+      "loss": 0.2159,
+      "step": 427
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.8637856332283039,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.6963,
+      "step": 428
+    },
+    {
+      "epoch": 0.6682242990654206,
+      "grad_norm": 0.7070722564579784,
+      "learning_rate": 7.489397905051465e-06,
+      "loss": 0.3265,
+      "step": 429
+    },
+    {
+      "epoch": 0.6697819314641744,
+      "grad_norm": 1.0912368137134154,
+      "learning_rate": 7.478780907442665e-06,
+      "loss": 0.3064,
+      "step": 430
+    },
+    {
+      "epoch": 0.6713395638629284,
+      "grad_norm": 0.4993170737898787,
+      "learning_rate": 7.468149070731742e-06,
+      "loss": 0.3532,
+      "step": 431
+    },
+    {
+      "epoch": 0.6728971962616822,
+      "grad_norm": 0.6200661171236782,
+      "learning_rate": 7.457502458565673e-06,
+      "loss": 0.2325,
+      "step": 432
+    },
+    {
+      "epoch": 0.6744548286604362,
+      "grad_norm": 0.6152606478186087,
+      "learning_rate": 7.446841134679888e-06,
+      "loss": 0.2538,
+      "step": 433
+    },
+    {
+      "epoch": 0.67601246105919,
+      "grad_norm": 0.5515844149625706,
+      "learning_rate": 7.436165162897886e-06,
+      "loss": 0.2619,
+      "step": 434
+    },
+    {
+      "epoch": 0.677570093457944,
+      "grad_norm": 0.7008609334925875,
+      "learning_rate": 7.425474607130858e-06,
+      "loss": 0.3168,
+      "step": 435
+    },
+    {
+      "epoch": 0.6791277258566978,
+      "grad_norm": 0.6379535213002501,
+      "learning_rate": 7.414769531377298e-06,
+      "loss": 0.268,
+      "step": 436
+    },
+    {
+      "epoch": 0.6806853582554517,
+      "grad_norm": 0.5317208819493666,
+      "learning_rate": 7.4040499997226245e-06,
+      "loss": 0.2193,
+      "step": 437
+    },
+    {
+      "epoch": 0.6822429906542056,
+      "grad_norm": 0.7290549365391932,
+      "learning_rate": 7.393316076338798e-06,
+      "loss": 0.3694,
+      "step": 438
+    },
+    {
+      "epoch": 0.6838006230529595,
+      "grad_norm": 0.8546888899097251,
+      "learning_rate": 7.382567825483929e-06,
+      "loss": 0.2822,
+      "step": 439
+    },
+    {
+      "epoch": 0.6853582554517134,
+      "grad_norm": 0.7390434139959143,
+      "learning_rate": 7.371805311501905e-06,
+      "loss": 0.24,
+      "step": 440
+    },
+    {
+      "epoch": 0.6869158878504673,
+      "grad_norm": 0.7021761813392882,
+      "learning_rate": 7.361028598821993e-06,
+      "loss": 0.3065,
+      "step": 441
+    },
+    {
+      "epoch": 0.6884735202492211,
+      "grad_norm": 0.5340954968447894,
+      "learning_rate": 7.350237751958466e-06,
+      "loss": 0.221,
+      "step": 442
+    },
+    {
+      "epoch": 0.6900311526479751,
+      "grad_norm": 0.5761626294301733,
+      "learning_rate": 7.339432835510203e-06,
+      "loss": 0.2345,
+      "step": 443
+    },
+    {
+      "epoch": 0.6915887850467289,
+      "grad_norm": 0.662018391928594,
+      "learning_rate": 7.328613914160319e-06,
+      "loss": 0.3171,
+      "step": 444
+    },
+    {
+      "epoch": 0.6931464174454829,
+      "grad_norm": 0.8130781056088618,
+      "learning_rate": 7.3177810526757594e-06,
+      "loss": 0.2909,
+      "step": 445
+    },
+    {
+      "epoch": 0.6947040498442367,
+      "grad_norm": 0.7219882547975953,
+      "learning_rate": 7.3069343159069296e-06,
+      "loss": 0.2481,
+      "step": 446
+    },
+    {
+      "epoch": 0.6962616822429907,
+      "grad_norm": 0.6369674341462834,
+      "learning_rate": 7.296073768787293e-06,
+      "loss": 0.3649,
+      "step": 447
+    },
+    {
+      "epoch": 0.6978193146417445,
+      "grad_norm": 0.7223244796104977,
+      "learning_rate": 7.285199476332991e-06,
+      "loss": 0.3488,
+      "step": 448
+    },
+    {
+      "epoch": 0.6993769470404985,
+      "grad_norm": 0.9091117254585579,
+      "learning_rate": 7.27431150364245e-06,
+      "loss": 0.3168,
+      "step": 449
+    },
+    {
+      "epoch": 0.7009345794392523,
+      "grad_norm": 0.7868966044967969,
+      "learning_rate": 7.263409915895992e-06,
+      "loss": 0.259,
+      "step": 450
+    },
+    {
+      "epoch": 0.7024922118380063,
+      "grad_norm": 0.6563385278402535,
+      "learning_rate": 7.252494778355444e-06,
+      "loss": 0.25,
+      "step": 451
+    },
+    {
+      "epoch": 0.7040498442367601,
+      "grad_norm": 0.641132207138942,
+      "learning_rate": 7.2415661563637506e-06,
+      "loss": 0.3307,
+      "step": 452
+    },
+    {
+      "epoch": 0.705607476635514,
+      "grad_norm": 0.7073578438725788,
+      "learning_rate": 7.23062411534458e-06,
+      "loss": 0.2261,
+      "step": 453
+    },
+    {
+      "epoch": 0.7071651090342679,
+      "grad_norm": 1.040324988179143,
+      "learning_rate": 7.2196687208019315e-06,
+      "loss": 0.2057,
+      "step": 454
+    },
+    {
+      "epoch": 0.7087227414330218,
+      "grad_norm": 0.6121914696936145,
+      "learning_rate": 7.208700038319744e-06,
+      "loss": 0.3199,
+      "step": 455
+    },
+    {
+      "epoch": 0.7102803738317757,
+      "grad_norm": 0.48700575482675645,
+      "learning_rate": 7.1977181335615085e-06,
+      "loss": 0.2259,
+      "step": 456
+    },
+    {
+      "epoch": 0.7118380062305296,
+      "grad_norm": 0.545525370035186,
+      "learning_rate": 7.186723072269863e-06,
+      "loss": 0.268,
+      "step": 457
+    },
+    {
+      "epoch": 0.7133956386292835,
+      "grad_norm": 0.846012722177333,
+      "learning_rate": 7.175714920266214e-06,
+      "loss": 0.7256,
+      "step": 458
+    },
+    {
+      "epoch": 0.7149532710280374,
+      "grad_norm": 0.5989442738821008,
+      "learning_rate": 7.164693743450329e-06,
+      "loss": 0.3005,
+      "step": 459
+    },
+    {
+      "epoch": 0.7165109034267912,
+      "grad_norm": 0.6556232944526054,
+      "learning_rate": 7.153659607799952e-06,
+      "loss": 0.2745,
+      "step": 460
+    },
+    {
+      "epoch": 0.7180685358255452,
+      "grad_norm": 0.6984028449665124,
+      "learning_rate": 7.142612579370402e-06,
+      "loss": 0.2272,
+      "step": 461
+    },
+    {
+      "epoch": 0.719626168224299,
+      "grad_norm": 0.5324547293774875,
+      "learning_rate": 7.131552724294181e-06,
+      "loss": 0.2518,
+      "step": 462
+    },
+    {
+      "epoch": 0.721183800623053,
+      "grad_norm": 0.5943008852162496,
+      "learning_rate": 7.1204801087805765e-06,
+      "loss": 0.2663,
+      "step": 463
+    },
+    {
+      "epoch": 0.7227414330218068,
+      "grad_norm": 0.5954388212811877,
+      "learning_rate": 7.109394799115268e-06,
+      "loss": 0.25,
+      "step": 464
+    },
+    {
+      "epoch": 0.7242990654205608,
+      "grad_norm": 0.5237590199785461,
+      "learning_rate": 7.098296861659925e-06,
+      "loss": 0.2451,
+      "step": 465
+    },
+    {
+      "epoch": 0.7258566978193146,
+      "grad_norm": 0.6108790141608955,
+      "learning_rate": 7.0871863628518136e-06,
+      "loss": 0.2782,
+      "step": 466
+    },
+    {
+      "epoch": 0.7274143302180686,
+      "grad_norm": 0.5784066614984076,
+      "learning_rate": 7.0760633692033975e-06,
+      "loss": 0.2588,
+      "step": 467
+    },
+    {
+      "epoch": 0.7289719626168224,
+      "grad_norm": 0.5736026805586273,
+      "learning_rate": 7.064927947301942e-06,
+      "loss": 0.3319,
+      "step": 468
+    },
+    {
+      "epoch": 0.7305295950155763,
+      "grad_norm": 0.660786534496975,
+      "learning_rate": 7.0537801638091116e-06,
+      "loss": 0.3207,
+      "step": 469
+    },
+    {
+      "epoch": 0.7320872274143302,
+      "grad_norm": 0.6164260678789174,
+      "learning_rate": 7.042620085460574e-06,
+      "loss": 0.2759,
+      "step": 470
+    },
+    {
+      "epoch": 0.7336448598130841,
+      "grad_norm": 1.0298833742062845,
+      "learning_rate": 7.0314477790656e-06,
+      "loss": 0.2769,
+      "step": 471
+    },
+    {
+      "epoch": 0.735202492211838,
+      "grad_norm": 0.7848249743313419,
+      "learning_rate": 7.020263311506659e-06,
+      "loss": 0.3963,
+      "step": 472
+    },
+    {
+      "epoch": 0.7367601246105919,
+      "grad_norm": 0.5488287365596327,
+      "learning_rate": 7.009066749739026e-06,
+      "loss": 0.2244,
+      "step": 473
+    },
+    {
+      "epoch": 0.7383177570093458,
+      "grad_norm": 0.5479634675942974,
+      "learning_rate": 6.99785816079038e-06,
+      "loss": 0.277,
+      "step": 474
+    },
+    {
+      "epoch": 0.7398753894080997,
+      "grad_norm": 0.632641842156797,
+      "learning_rate": 6.986637611760394e-06,
+      "loss": 0.2948,
+      "step": 475
+    },
+    {
+      "epoch": 0.7414330218068536,
+      "grad_norm": 0.8957280703439034,
+      "learning_rate": 6.975405169820344e-06,
+      "loss": 0.353,
+      "step": 476
+    },
+    {
+      "epoch": 0.7429906542056075,
+      "grad_norm": 0.575855558736389,
+      "learning_rate": 6.9641609022127e-06,
+      "loss": 0.2667,
+      "step": 477
+    },
+    {
+      "epoch": 0.7445482866043613,
+      "grad_norm": 0.6675031465700932,
+      "learning_rate": 6.952904876250729e-06,
+      "loss": 0.239,
+      "step": 478
+    },
+    {
+      "epoch": 0.7461059190031153,
+      "grad_norm": 0.5488652674770181,
+      "learning_rate": 6.941637159318083e-06,
+      "loss": 0.2605,
+      "step": 479
+    },
+    {
+      "epoch": 0.7476635514018691,
+      "grad_norm": 0.765674305969199,
+      "learning_rate": 6.9303578188684085e-06,
+      "loss": 0.2668,
+      "step": 480
+    },
+    {
+      "epoch": 0.7492211838006231,
+      "grad_norm": 0.6239515815918181,
+      "learning_rate": 6.919066922424931e-06,
+      "loss": 0.2883,
+      "step": 481
+    },
+    {
+      "epoch": 0.7507788161993769,
+      "grad_norm": 0.8869720139234101,
+      "learning_rate": 6.907764537580053e-06,
+      "loss": 0.2726,
+      "step": 482
+    },
+    {
+      "epoch": 0.7523364485981309,
+      "grad_norm": 0.7182796918869947,
+      "learning_rate": 6.896450731994959e-06,
+      "loss": 0.2575,
+      "step": 483
+    },
+    {
+      "epoch": 0.7538940809968847,
+      "grad_norm": 0.689714104473123,
+      "learning_rate": 6.8851255733992006e-06,
+      "loss": 0.2548,
+      "step": 484
+    },
+    {
+      "epoch": 0.7554517133956387,
+      "grad_norm": 0.8752253075858156,
+      "learning_rate": 6.873789129590287e-06,
+      "loss": 0.2598,
+      "step": 485
+    },
+    {
+      "epoch": 0.7570093457943925,
+      "grad_norm": 0.6547980788626615,
+      "learning_rate": 6.862441468433298e-06,
+      "loss": 0.274,
+      "step": 486
+    },
+    {
+      "epoch": 0.7585669781931464,
+      "grad_norm": 0.6955009265885427,
+      "learning_rate": 6.851082657860453e-06,
+      "loss": 0.286,
+      "step": 487
+    },
+    {
+      "epoch": 0.7601246105919003,
+      "grad_norm": 0.6057981135550708,
+      "learning_rate": 6.839712765870725e-06,
+      "loss": 0.3072,
+      "step": 488
+    },
+    {
+      "epoch": 0.7616822429906542,
+      "grad_norm": 0.5562050274960125,
+      "learning_rate": 6.828331860529422e-06,
+      "loss": 0.2765,
+      "step": 489
+    },
+    {
+      "epoch": 0.7632398753894081,
+      "grad_norm": 0.9242326126038012,
+      "learning_rate": 6.816940009967787e-06,
+      "loss": 0.8322,
+      "step": 490
+    },
+    {
+      "epoch": 0.764797507788162,
+      "grad_norm": 0.7207873437414208,
+      "learning_rate": 6.805537282382581e-06,
+      "loss": 0.2175,
+      "step": 491
+    },
+    {
+      "epoch": 0.7663551401869159,
+      "grad_norm": 0.5928431496932391,
+      "learning_rate": 6.79412374603568e-06,
+      "loss": 0.24,
+      "step": 492
+    },
+    {
+      "epoch": 0.7679127725856698,
+      "grad_norm": 0.8086943486132299,
+      "learning_rate": 6.782699469253671e-06,
+      "loss": 0.3252,
+      "step": 493
+    },
+    {
+      "epoch": 0.7694704049844237,
+      "grad_norm": 0.6500702055304157,
+      "learning_rate": 6.771264520427432e-06,
+      "loss": 0.2831,
+      "step": 494
+    },
+    {
+      "epoch": 0.7710280373831776,
+      "grad_norm": 0.5699110226071109,
+      "learning_rate": 6.759818968011731e-06,
+      "loss": 0.2604,
+      "step": 495
+    },
+    {
+      "epoch": 0.7725856697819314,
+      "grad_norm": 0.5580360203832775,
+      "learning_rate": 6.748362880524819e-06,
+      "loss": 0.2684,
+      "step": 496
+    },
+    {
+      "epoch": 0.7741433021806854,
+      "grad_norm": 0.7886971673525824,
+      "learning_rate": 6.736896326548006e-06,
+      "loss": 0.2123,
+      "step": 497
+    },
+    {
+      "epoch": 0.7757009345794392,
+      "grad_norm": 0.6054794222872896,
+      "learning_rate": 6.7254193747252645e-06,
+      "loss": 0.3127,
+      "step": 498
+    },
+    {
+      "epoch": 0.7772585669781932,
+      "grad_norm": 0.7462264966697667,
+      "learning_rate": 6.713932093762811e-06,
+      "loss": 0.3051,
+      "step": 499
+    },
+    {
+      "epoch": 0.778816199376947,
+      "grad_norm": 0.6730107352048917,
+      "learning_rate": 6.702434552428702e-06,
+      "loss": 0.3007,
+      "step": 500
+    },
+    {
+      "epoch": 0.778816199376947,
+      "eval_loss": 0.366039514541626,
+      "eval_runtime": 2.8278,
+      "eval_samples_per_second": 9.194,
+      "eval_steps_per_second": 2.475,
+      "step": 500
+    },
+    {
+      "epoch": 0.780373831775701,
+      "grad_norm": 0.6861562261314369,
+      "learning_rate": 6.690926819552408e-06,
+      "loss": 0.287,
+      "step": 501
+    },
+    {
+      "epoch": 0.7819314641744548,
+      "grad_norm": 0.9308842751027873,
+      "learning_rate": 6.679408964024419e-06,
+      "loss": 0.8811,
+      "step": 502
+    },
+    {
+      "epoch": 0.7834890965732088,
+      "grad_norm": 0.6729340583401545,
+      "learning_rate": 6.667881054795818e-06,
+      "loss": 0.2304,
+      "step": 503
+    },
+    {
+      "epoch": 0.7850467289719626,
+      "grad_norm": 0.7848693380567189,
+      "learning_rate": 6.65634316087788e-06,
+      "loss": 0.2965,
+      "step": 504
+    },
+    {
+      "epoch": 0.7866043613707165,
+      "grad_norm": 0.7035209610164758,
+      "learning_rate": 6.6447953513416474e-06,
+      "loss": 0.2589,
+      "step": 505
+    },
+    {
+      "epoch": 0.7881619937694704,
+      "grad_norm": 0.5912497530045528,
+      "learning_rate": 6.633237695317523e-06,
+      "loss": 0.2566,
+      "step": 506
+    },
+    {
+      "epoch": 0.7897196261682243,
+      "grad_norm": 0.7247030582803601,
+      "learning_rate": 6.621670261994857e-06,
+      "loss": 0.2726,
+      "step": 507
+    },
+    {
+      "epoch": 0.7912772585669782,
+      "grad_norm": 0.6656199289111854,
+      "learning_rate": 6.610093120621532e-06,
+      "loss": 0.2999,
+      "step": 508
+    },
+    {
+      "epoch": 0.7928348909657321,
+      "grad_norm": 0.8158049430019846,
+      "learning_rate": 6.598506340503541e-06,
+      "loss": 0.2453,
+      "step": 509
+    },
+    {
+      "epoch": 0.794392523364486,
+      "grad_norm": 0.6984808487227331,
+      "learning_rate": 6.586909991004587e-06,
+      "loss": 0.3149,
+      "step": 510
+    },
+    {
+      "epoch": 0.7959501557632399,
+      "grad_norm": 0.6126995511808185,
+      "learning_rate": 6.575304141545653e-06,
+      "loss": 0.2666,
+      "step": 511
+    },
+    {
+      "epoch": 0.7975077881619937,
+      "grad_norm": 0.5505067863178127,
+      "learning_rate": 6.5636888616046e-06,
+      "loss": 0.2998,
+      "step": 512
+    },
+    {
+      "epoch": 0.7990654205607477,
+      "grad_norm": 0.7685023837309144,
+      "learning_rate": 6.552064220715737e-06,
+      "loss": 0.1876,
+      "step": 513
+    },
+    {
+      "epoch": 0.8006230529595015,
+      "grad_norm": 0.6186176349687995,
+      "learning_rate": 6.5404302884694145e-06,
+      "loss": 0.2823,
+      "step": 514
+    },
+    {
+      "epoch": 0.8021806853582555,
+      "grad_norm": 0.6245560858038731,
+      "learning_rate": 6.528787134511608e-06,
+      "loss": 0.2063,
+      "step": 515
+    },
+    {
+      "epoch": 0.8037383177570093,
+      "grad_norm": 0.6027333943405707,
+      "learning_rate": 6.5171348285434965e-06,
+      "loss": 0.3079,
+      "step": 516
+    },
+    {
+      "epoch": 0.8052959501557633,
+      "grad_norm": 0.7334879584524152,
+      "learning_rate": 6.505473440321044e-06,
+      "loss": 0.2906,
+      "step": 517
+    },
+    {
+      "epoch": 0.8068535825545171,
+      "grad_norm": 0.5611150628723894,
+      "learning_rate": 6.493803039654589e-06,
+      "loss": 0.2437,
+      "step": 518
+    },
+    {
+      "epoch": 0.8084112149532711,
+      "grad_norm": 0.5709098533110826,
+      "learning_rate": 6.48212369640842e-06,
+      "loss": 0.2695,
+      "step": 519
+    },
+    {
+      "epoch": 0.8099688473520249,
+      "grad_norm": 0.7469745364074795,
+      "learning_rate": 6.4704354805003626e-06,
+      "loss": 0.2828,
+      "step": 520
+    },
+    {
+      "epoch": 0.8115264797507789,
+      "grad_norm": 0.7580041865120294,
+      "learning_rate": 6.458738461901354e-06,
+      "loss": 0.2456,
+      "step": 521
+    },
+    {
+      "epoch": 0.8130841121495327,
+      "grad_norm": 0.7356278600835281,
+      "learning_rate": 6.447032710635035e-06,
+      "loss": 0.2325,
+      "step": 522
+    },
+    {
+      "epoch": 0.8146417445482866,
+      "grad_norm": 0.5694798902062997,
+      "learning_rate": 6.435318296777316e-06,
+      "loss": 0.2763,
+      "step": 523
+    },
+    {
+      "epoch": 0.8161993769470405,
+      "grad_norm": 0.6227419240061058,
+      "learning_rate": 6.423595290455971e-06,
+      "loss": 0.2871,
+      "step": 524
+    },
+    {
+      "epoch": 0.8177570093457944,
+      "grad_norm": 0.7005734890264759,
+      "learning_rate": 6.41186376185021e-06,
+      "loss": 0.3003,
+      "step": 525
+    },
+    {
+      "epoch": 0.8193146417445483,
+      "grad_norm": 0.8909957470129115,
+      "learning_rate": 6.400123781190265e-06,
+      "loss": 0.3328,
+      "step": 526
+    },
+    {
+      "epoch": 0.8208722741433022,
+      "grad_norm": 0.5534613081412874,
+      "learning_rate": 6.388375418756959e-06,
+      "loss": 0.2816,
+      "step": 527
+    },
+    {
+      "epoch": 0.822429906542056,
+      "grad_norm": 0.7546735553270361,
+      "learning_rate": 6.3766187448813e-06,
+      "loss": 0.249,
+      "step": 528
+    },
+    {
+      "epoch": 0.82398753894081,
+      "grad_norm": 1.08724358600162,
+      "learning_rate": 6.3648538299440444e-06,
+      "loss": 0.2978,
+      "step": 529
+    },
+    {
+      "epoch": 0.8255451713395638,
+      "grad_norm": 0.5510668935985298,
+      "learning_rate": 6.35308074437529e-06,
+      "loss": 0.2586,
+      "step": 530
+    },
+    {
+      "epoch": 0.8271028037383178,
+      "grad_norm": 0.9295747471096518,
+      "learning_rate": 6.341299558654042e-06,
+      "loss": 0.4423,
+      "step": 531
+    },
+    {
+      "epoch": 0.8286604361370716,
+      "grad_norm": 0.6078601432381958,
+      "learning_rate": 6.329510343307801e-06,
+      "loss": 0.3089,
+      "step": 532
+    },
+    {
+      "epoch": 0.8302180685358256,
+      "grad_norm": 0.7009464939112006,
+      "learning_rate": 6.3177131689121325e-06,
+      "loss": 0.276,
+      "step": 533
+    },
+    {
+      "epoch": 0.8317757009345794,
+      "grad_norm": 0.6055023924718476,
+      "learning_rate": 6.305908106090255e-06,
+      "loss": 0.289,
+      "step": 534
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 1.0146910918408667,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.2117,
+      "step": 535
+    },
+    {
+      "epoch": 0.8348909657320872,
+      "grad_norm": 0.5684609592585866,
+      "learning_rate": 6.282274597896421e-06,
+      "loss": 0.268,
+      "step": 536
+    },
+    {
+      "epoch": 0.8364485981308412,
+      "grad_norm": 0.6324240721524141,
+      "learning_rate": 6.2704462940053165e-06,
+      "loss": 0.2348,
+      "step": 537
+    },
+    {
+      "epoch": 0.838006230529595,
+      "grad_norm": 0.582281483203043,
+      "learning_rate": 6.2586103846488654e-06,
+      "loss": 0.2975,
+      "step": 538
+    },
+    {
+      "epoch": 0.839563862928349,
+      "grad_norm": 0.9776397911217686,
+      "learning_rate": 6.246766940682165e-06,
+      "loss": 0.7799,
+      "step": 539
+    },
+    {
+      "epoch": 0.8411214953271028,
+      "grad_norm": 0.5174311636719064,
+      "learning_rate": 6.234916033005421e-06,
+      "loss": 0.1973,
+      "step": 540
+    },
+    {
+      "epoch": 0.8426791277258567,
+      "grad_norm": 0.6331348424871293,
+      "learning_rate": 6.22305773256352e-06,
+      "loss": 0.2749,
+      "step": 541
+    },
+    {
+      "epoch": 0.8442367601246106,
+      "grad_norm": 0.5229024799327089,
+      "learning_rate": 6.211192110345603e-06,
+      "loss": 0.2811,
+      "step": 542
+    },
+    {
+      "epoch": 0.8457943925233645,
+      "grad_norm": 0.5575336291274628,
+      "learning_rate": 6.199319237384645e-06,
+      "loss": 0.2534,
+      "step": 543
+    },
+    {
+      "epoch": 0.8473520249221184,
+      "grad_norm": 0.7289277957152529,
+      "learning_rate": 6.187439184757025e-06,
+      "loss": 0.7677,
+      "step": 544
+    },
+    {
+      "epoch": 0.8489096573208723,
+      "grad_norm": 0.540862000313681,
+      "learning_rate": 6.1755520235821055e-06,
+      "loss": 0.3294,
+      "step": 545
+    },
+    {
+      "epoch": 0.8504672897196262,
+      "grad_norm": 0.7035928510596402,
+      "learning_rate": 6.163657825021802e-06,
+      "loss": 0.3147,
+      "step": 546
+    },
+    {
+      "epoch": 0.8520249221183801,
+      "grad_norm": 0.5438019081147566,
+      "learning_rate": 6.1517566602801596e-06,
+      "loss": 0.2003,
+      "step": 547
+    },
+    {
+      "epoch": 0.8535825545171339,
+      "grad_norm": 0.8025432437697821,
+      "learning_rate": 6.139848600602926e-06,
+      "loss": 0.2756,
+      "step": 548
+    },
+    {
+      "epoch": 0.8551401869158879,
+      "grad_norm": 0.654724311621637,
+      "learning_rate": 6.127933717277123e-06,
+      "loss": 0.2934,
+      "step": 549
+    },
+    {
+      "epoch": 0.8566978193146417,
+      "grad_norm": 1.328494349119985,
+      "learning_rate": 6.116012081630629e-06,
+      "loss": 0.2731,
+      "step": 550
+    },
+    {
+      "epoch": 0.8582554517133957,
+      "grad_norm": 0.5435085655801555,
+      "learning_rate": 6.104083765031734e-06,
+      "loss": 0.1934,
+      "step": 551
+    },
+    {
+      "epoch": 0.8598130841121495,
+      "grad_norm": 0.7875228878328122,
+      "learning_rate": 6.0921488388887315e-06,
+      "loss": 0.7651,
+      "step": 552
+    },
+    {
+      "epoch": 0.8613707165109035,
+      "grad_norm": 0.9138141754922854,
+      "learning_rate": 6.080207374649482e-06,
+      "loss": 0.6927,
+      "step": 553
+    },
+    {
+      "epoch": 0.8629283489096573,
+      "grad_norm": 0.658537276564411,
+      "learning_rate": 6.068259443800981e-06,
+      "loss": 0.3088,
+      "step": 554
+    },
+    {
+      "epoch": 0.8644859813084113,
+      "grad_norm": 0.8424646536545584,
+      "learning_rate": 6.0563051178689395e-06,
+      "loss": 0.6504,
+      "step": 555
+    },
+    {
+      "epoch": 0.8660436137071651,
+      "grad_norm": 0.5487578856405264,
+      "learning_rate": 6.0443444684173524e-06,
+      "loss": 0.2504,
+      "step": 556
+    },
+    {
+      "epoch": 0.867601246105919,
+      "grad_norm": 0.577742497246078,
+      "learning_rate": 6.032377567048071e-06,
+      "loss": 0.2724,
+      "step": 557
+    },
+    {
+      "epoch": 0.8691588785046729,
+      "grad_norm": 0.5720861179090082,
+      "learning_rate": 6.0204044854003705e-06,
+      "loss": 0.2494,
+      "step": 558
+    },
+    {
+      "epoch": 0.8707165109034268,
+      "grad_norm": 0.614579567677496,
+      "learning_rate": 6.008425295150526e-06,
+      "loss": 0.2431,
+      "step": 559
+    },
+    {
+      "epoch": 0.8722741433021807,
+      "grad_norm": 0.7053311054530548,
+      "learning_rate": 5.996440068011383e-06,
+      "loss": 0.3007,
+      "step": 560
+    },
+    {
+      "epoch": 0.8738317757009346,
+      "grad_norm": 0.6676390464189279,
+      "learning_rate": 5.9844488757319205e-06,
+      "loss": 0.2309,
+      "step": 561
+    },
+    {
+      "epoch": 0.8753894080996885,
+      "grad_norm": 0.7749136512043995,
+      "learning_rate": 5.972451790096837e-06,
+      "loss": 0.3327,
+      "step": 562
+    },
+    {
+      "epoch": 0.8769470404984424,
+      "grad_norm": 0.6384638289071073,
+      "learning_rate": 5.960448882926101e-06,
+      "loss": 0.3447,
+      "step": 563
+    },
+    {
+      "epoch": 0.8785046728971962,
+      "grad_norm": 0.5992691951009588,
+      "learning_rate": 5.948440226074539e-06,
+      "loss": 0.2181,
+      "step": 564
+    },
+    {
+      "epoch": 0.8800623052959502,
+      "grad_norm": 0.7553595588283479,
+      "learning_rate": 5.936425891431394e-06,
+      "loss": 0.2307,
+      "step": 565
+    },
+    {
+      "epoch": 0.881619937694704,
+      "grad_norm": 0.7852167594459105,
+      "learning_rate": 5.924405950919902e-06,
+      "loss": 0.3119,
+      "step": 566
+    },
+    {
+      "epoch": 0.883177570093458,
+      "grad_norm": 0.5241767306732831,
+      "learning_rate": 5.91238047649685e-06,
+      "loss": 0.2293,
+      "step": 567
+    },
+    {
+      "epoch": 0.8847352024922118,
+      "grad_norm": 0.7471757161407457,
+      "learning_rate": 5.900349540152167e-06,
+      "loss": 0.3251,
+      "step": 568
+    },
+    {
+      "epoch": 0.8862928348909658,
+      "grad_norm": 0.7843562558921093,
+      "learning_rate": 5.888313213908468e-06,
+      "loss": 0.2868,
+      "step": 569
+    },
+    {
+      "epoch": 0.8878504672897196,
+      "grad_norm": 0.7503757117692468,
+      "learning_rate": 5.876271569820638e-06,
+      "loss": 0.2555,
+      "step": 570
+    },
+    {
+      "epoch": 0.8894080996884736,
+      "grad_norm": 0.7275473381141497,
+      "learning_rate": 5.864224679975399e-06,
+      "loss": 0.2945,
+      "step": 571
+    },
+    {
+      "epoch": 0.8909657320872274,
+      "grad_norm": 0.7108035276261796,
+      "learning_rate": 5.852172616490875e-06,
+      "loss": 0.2826,
+      "step": 572
+    },
+    {
+      "epoch": 0.8925233644859814,
+      "grad_norm": 0.7292991034746364,
+      "learning_rate": 5.84011545151616e-06,
+      "loss": 0.2052,
+      "step": 573
+    },
+    {
+      "epoch": 0.8940809968847352,
+      "grad_norm": 0.7701552350107903,
+      "learning_rate": 5.828053257230893e-06,
+      "loss": 0.272,
+      "step": 574
+    },
+    {
+      "epoch": 0.8956386292834891,
+      "grad_norm": 0.5385532832886137,
+      "learning_rate": 5.815986105844813e-06,
+      "loss": 0.2859,
+      "step": 575
+    },
+    {
+      "epoch": 0.897196261682243,
+      "grad_norm": 0.4784130573714912,
+      "learning_rate": 5.803914069597342e-06,
+      "loss": 0.2385,
+      "step": 576
+    },
+    {
+      "epoch": 0.8987538940809969,
+      "grad_norm": 0.6661872649899948,
+      "learning_rate": 5.791837220757139e-06,
+      "loss": 0.2601,
+      "step": 577
+    },
+    {
+      "epoch": 0.9003115264797508,
+      "grad_norm": 0.6423784935357807,
+      "learning_rate": 5.779755631621679e-06,
+      "loss": 0.2861,
+      "step": 578
+    },
+    {
+      "epoch": 0.9018691588785047,
+      "grad_norm": 0.5564393840451491,
+      "learning_rate": 5.767669374516807e-06,
+      "loss": 0.2247,
+      "step": 579
+    },
+    {
+      "epoch": 0.9034267912772586,
+      "grad_norm": 1.0831258133460262,
+      "learning_rate": 5.755578521796321e-06,
+      "loss": 0.7525,
+      "step": 580
+    },
+    {
+      "epoch": 0.9049844236760125,
+      "grad_norm": 0.7513349003919912,
+      "learning_rate": 5.743483145841525e-06,
+      "loss": 0.2417,
+      "step": 581
+    },
+    {
+      "epoch": 0.9065420560747663,
+      "grad_norm": 0.6795049942946425,
+      "learning_rate": 5.731383319060805e-06,
+      "loss": 0.3177,
+      "step": 582
+    },
+    {
+      "epoch": 0.9080996884735203,
+      "grad_norm": 0.7390882203257174,
+      "learning_rate": 5.719279113889184e-06,
+      "loss": 0.3581,
+      "step": 583
+    },
+    {
+      "epoch": 0.9096573208722741,
+      "grad_norm": 0.582238701325137,
+      "learning_rate": 5.707170602787908e-06,
+      "loss": 0.2755,
+      "step": 584
+    },
+    {
+      "epoch": 0.9112149532710281,
+      "grad_norm": 0.5600730304265147,
+      "learning_rate": 5.695057858243989e-06,
+      "loss": 0.2745,
+      "step": 585
+    },
+    {
+      "epoch": 0.9127725856697819,
+      "grad_norm": 0.638261445819756,
+      "learning_rate": 5.682940952769788e-06,
+      "loss": 0.177,
+      "step": 586
+    },
+    {
+      "epoch": 0.9143302180685359,
+      "grad_norm": 0.6191376433403027,
+      "learning_rate": 5.670819958902576e-06,
+      "loss": 0.2447,
+      "step": 587
+    },
+    {
+      "epoch": 0.9158878504672897,
+      "grad_norm": 0.8780779667116613,
+      "learning_rate": 5.658694949204094e-06,
+      "loss": 0.2438,
+      "step": 588
+    },
+    {
+      "epoch": 0.9174454828660437,
+      "grad_norm": 0.7198384217051569,
+      "learning_rate": 5.646565996260129e-06,
+      "loss": 0.2408,
+      "step": 589
+    },
+    {
+      "epoch": 0.9190031152647975,
+      "grad_norm": 1.0167707545321327,
+      "learning_rate": 5.634433172680072e-06,
+      "loss": 0.7316,
+      "step": 590
+    },
+    {
+      "epoch": 0.9205607476635514,
+      "grad_norm": 0.6546640817603999,
+      "learning_rate": 5.622296551096481e-06,
+      "loss": 0.3253,
+      "step": 591
+    },
+    {
+      "epoch": 0.9221183800623053,
+      "grad_norm": 0.8382914937621136,
+      "learning_rate": 5.61015620416466e-06,
+      "loss": 0.2981,
+      "step": 592
+    },
+    {
+      "epoch": 0.9236760124610592,
+      "grad_norm": 0.6120676244562511,
+      "learning_rate": 5.598012204562204e-06,
+      "loss": 0.2647,
+      "step": 593
+    },
+    {
+      "epoch": 0.9252336448598131,
+      "grad_norm": 0.700304920368459,
+      "learning_rate": 5.5858646249885855e-06,
+      "loss": 0.2249,
+      "step": 594
+    },
+    {
+      "epoch": 0.926791277258567,
+      "grad_norm": 0.6840607666164885,
+      "learning_rate": 5.573713538164698e-06,
+      "loss": 0.2795,
+      "step": 595
+    },
+    {
+      "epoch": 0.9283489096573209,
+      "grad_norm": 0.8513439406827232,
+      "learning_rate": 5.561559016832438e-06,
+      "loss": 0.2931,
+      "step": 596
+    },
+    {
+      "epoch": 0.9299065420560748,
+      "grad_norm": 0.7770854828276103,
+      "learning_rate": 5.549401133754259e-06,
+      "loss": 0.2819,
+      "step": 597
+    },
+    {
+      "epoch": 0.9314641744548287,
+      "grad_norm": 0.5874428152071298,
+      "learning_rate": 5.5372399617127415e-06,
+      "loss": 0.4483,
+      "step": 598
+    },
+    {
+      "epoch": 0.9330218068535826,
+      "grad_norm": 0.6488291930752155,
+      "learning_rate": 5.525075573510154e-06,
+      "loss": 0.3873,
+      "step": 599
+    },
+    {
+      "epoch": 0.9345794392523364,
+      "grad_norm": 0.5286049009581415,
+      "learning_rate": 5.512908041968018e-06,
+      "loss": 0.3113,
+      "step": 600
+    },
+    {
+      "epoch": 0.9361370716510904,
+      "grad_norm": 0.5929114341675884,
+      "learning_rate": 5.500737439926674e-06,
+      "loss": 0.334,
+      "step": 601
+    },
+    {
+      "epoch": 0.9376947040498442,
+      "grad_norm": 0.847988678139219,
+      "learning_rate": 5.488563840244843e-06,
+      "loss": 0.3026,
+      "step": 602
+    },
+    {
+      "epoch": 0.9392523364485982,
+      "grad_norm": 0.5635536494534855,
+      "learning_rate": 5.476387315799189e-06,
+      "loss": 0.2146,
+      "step": 603
+    },
+    {
+      "epoch": 0.940809968847352,
+      "grad_norm": 0.5570339586064699,
+      "learning_rate": 5.464207939483891e-06,
+      "loss": 0.2407,
+      "step": 604
+    },
+    {
+      "epoch": 0.942367601246106,
+      "grad_norm": 0.5996622051462993,
+      "learning_rate": 5.452025784210193e-06,
+      "loss": 0.2301,
+      "step": 605
+    },
+    {
+      "epoch": 0.9439252336448598,
+      "grad_norm": 0.6134433053870486,
+      "learning_rate": 5.439840922905982e-06,
+      "loss": 0.2881,
+      "step": 606
+    },
+    {
+      "epoch": 0.9454828660436138,
+      "grad_norm": 0.6534399078221188,
+      "learning_rate": 5.42765342851534e-06,
+      "loss": 0.2991,
+      "step": 607
+    },
+    {
+      "epoch": 0.9470404984423676,
+      "grad_norm": 0.690575160461211,
+      "learning_rate": 5.415463373998112e-06,
+      "loss": 0.3353,
+      "step": 608
+    },
+    {
+      "epoch": 0.9485981308411215,
+      "grad_norm": 0.8921403858615625,
+      "learning_rate": 5.403270832329473e-06,
+      "loss": 0.2008,
+      "step": 609
+    },
+    {
+      "epoch": 0.9501557632398754,
+      "grad_norm": 0.7025072704858318,
+      "learning_rate": 5.391075876499483e-06,
+      "loss": 0.2621,
+      "step": 610
+    },
+    {
+      "epoch": 0.9517133956386293,
+      "grad_norm": 0.6757376481036776,
+      "learning_rate": 5.3788785795126554e-06,
+      "loss": 0.2469,
+      "step": 611
+    },
+    {
+      "epoch": 0.9532710280373832,
+      "grad_norm": 0.6875817981934039,
+      "learning_rate": 5.36667901438752e-06,
+      "loss": 0.236,
+      "step": 612
+    },
+    {
+      "epoch": 0.9548286604361371,
+      "grad_norm": 0.7529907066389188,
+      "learning_rate": 5.354477254156184e-06,
+      "loss": 0.2755,
+      "step": 613
+    },
+    {
+      "epoch": 0.956386292834891,
+      "grad_norm": 0.5896914001022201,
+      "learning_rate": 5.342273371863895e-06,
+      "loss": 0.2634,
+      "step": 614
+    },
+    {
+      "epoch": 0.9579439252336449,
+      "grad_norm": 0.7027739875415374,
+      "learning_rate": 5.330067440568605e-06,
+      "loss": 0.2829,
+      "step": 615
+    },
+    {
+      "epoch": 0.9595015576323987,
+      "grad_norm": 0.7667507882778263,
+      "learning_rate": 5.317859533340532e-06,
+      "loss": 0.3506,
+      "step": 616
+    },
+    {
+      "epoch": 0.9610591900311527,
+      "grad_norm": 0.8456849495117152,
+      "learning_rate": 5.30564972326172e-06,
+      "loss": 0.3054,
+      "step": 617
+    },
+    {
+      "epoch": 0.9626168224299065,
+      "grad_norm": 0.51248054747711,
+      "learning_rate": 5.293438083425611e-06,
+      "loss": 0.2301,
+      "step": 618
+    },
+    {
+      "epoch": 0.9641744548286605,
+      "grad_norm": 0.7467447701014356,
+      "learning_rate": 5.281224686936594e-06,
+      "loss": 0.3769,
+      "step": 619
+    },
+    {
+      "epoch": 0.9657320872274143,
+      "grad_norm": 0.6414852824849853,
+      "learning_rate": 5.26900960690958e-06,
+      "loss": 0.2779,
+      "step": 620
+    },
+    {
+      "epoch": 0.9672897196261683,
+      "grad_norm": 0.6295561918474766,
+      "learning_rate": 5.256792916469552e-06,
+      "loss": 0.2586,
+      "step": 621
+    },
+    {
+      "epoch": 0.9688473520249221,
+      "grad_norm": 0.6049915483456579,
+      "learning_rate": 5.244574688751138e-06,
+      "loss": 0.2195,
+      "step": 622
+    },
+    {
+      "epoch": 0.9704049844236761,
+      "grad_norm": 1.0443726591172307,
+      "learning_rate": 5.23235499689817e-06,
+      "loss": 0.7974,
+      "step": 623
+    },
+    {
+      "epoch": 0.9719626168224299,
+      "grad_norm": 0.4947295180592083,
+      "learning_rate": 5.220133914063239e-06,
+      "loss": 0.2173,
+      "step": 624
+    },
+    {
+      "epoch": 0.9735202492211839,
+      "grad_norm": 0.774384784856486,
+      "learning_rate": 5.20791151340727e-06,
+      "loss": 0.2196,
+      "step": 625
+    },
+    {
+      "epoch": 0.9750778816199377,
+      "grad_norm": 0.6848166382044595,
+      "learning_rate": 5.195687868099073e-06,
+      "loss": 0.2784,
+      "step": 626
+    },
+    {
+      "epoch": 0.9766355140186916,
+      "grad_norm": 1.5114886266488685,
+      "learning_rate": 5.1834630513149086e-06,
+      "loss": 0.8531,
+      "step": 627
+    },
+    {
+      "epoch": 0.9781931464174455,
+      "grad_norm": 1.0107267590162416,
+      "learning_rate": 5.171237136238054e-06,
+      "loss": 0.2692,
+      "step": 628
+    },
+    {
+      "epoch": 0.9797507788161994,
+      "grad_norm": 0.7292686375997546,
+      "learning_rate": 5.159010196058356e-06,
+      "loss": 0.2305,
+      "step": 629
+    },
+    {
+      "epoch": 0.9813084112149533,
+      "grad_norm": 0.565137268345777,
+      "learning_rate": 5.1467823039718046e-06,
+      "loss": 0.3076,
+      "step": 630
+    },
+    {
+      "epoch": 0.9828660436137072,
+      "grad_norm": 1.0043606696953649,
+      "learning_rate": 5.134553533180082e-06,
+      "loss": 0.7515,
+      "step": 631
+    },
+    {
+      "epoch": 0.9844236760124611,
+      "grad_norm": 0.6233222851344706,
+      "learning_rate": 5.122323956890136e-06,
+      "loss": 0.3019,
+      "step": 632
+    },
+    {
+      "epoch": 0.985981308411215,
+      "grad_norm": 0.730841019679576,
+      "learning_rate": 5.110093648313732e-06,
+      "loss": 0.3483,
+      "step": 633
+    },
+    {
+      "epoch": 0.9875389408099688,
+      "grad_norm": 0.6632089778706831,
+      "learning_rate": 5.097862680667024e-06,
+      "loss": 0.3608,
+      "step": 634
+    },
+    {
+      "epoch": 0.9890965732087228,
+      "grad_norm": 0.6691349603071851,
+      "learning_rate": 5.085631127170106e-06,
+      "loss": 0.2524,
+      "step": 635
+    },
+    {
+      "epoch": 0.9906542056074766,
+      "grad_norm": 0.73951871814352,
+      "learning_rate": 5.073399061046584e-06,
+      "loss": 0.3134,
+      "step": 636
+    },
+    {
+      "epoch": 0.9922118380062306,
+      "grad_norm": 0.6079224831882191,
+      "learning_rate": 5.061166555523129e-06,
+      "loss": 0.308,
+      "step": 637
+    },
+    {
+      "epoch": 0.9937694704049844,
+      "grad_norm": 0.6166501321049228,
+      "learning_rate": 5.048933683829046e-06,
+      "loss": 0.3073,
+      "step": 638
+    },
+    {
+      "epoch": 0.9953271028037384,
+      "grad_norm": 1.003680827370098,
+      "learning_rate": 5.0367005191958275e-06,
+      "loss": 0.3229,
+      "step": 639
+    },
+    {
+      "epoch": 0.9968847352024922,
+      "grad_norm": 0.7517097037555416,
+      "learning_rate": 5.024467134856725e-06,
+      "loss": 0.2744,
+      "step": 640
+    },
+    {
+      "epoch": 0.9984423676012462,
+      "grad_norm": 0.5399606823979543,
+      "learning_rate": 5.012233604046303e-06,
+      "loss": 0.247,
+      "step": 641
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.602288219366613,
+      "learning_rate": 5e-06,
+      "loss": 0.2401,
+      "step": 642
+    },
+    {
+      "epoch": 1.0015576323987538,
+      "grad_norm": 0.5865025697495287,
+      "learning_rate": 4.987766395953699e-06,
+      "loss": 0.1907,
+      "step": 643
+    },
+    {
+      "epoch": 1.0031152647975077,
+      "grad_norm": 0.6271078194072086,
+      "learning_rate": 4.975532865143277e-06,
+      "loss": 0.1837,
+      "step": 644
+    },
+    {
+      "epoch": 1.0046728971962617,
+      "grad_norm": 0.5433097442594456,
+      "learning_rate": 4.963299480804173e-06,
+      "loss": 0.2122,
+      "step": 645
+    },
+    {
+      "epoch": 1.0062305295950156,
+      "grad_norm": 0.6939024837652152,
+      "learning_rate": 4.951066316170956e-06,
+      "loss": 0.2464,
+      "step": 646
+    },
+    {
+      "epoch": 1.0077881619937694,
+      "grad_norm": 0.6120591075726564,
+      "learning_rate": 4.938833444476873e-06,
+      "loss": 0.2943,
+      "step": 647
+    },
+    {
+      "epoch": 1.0093457943925233,
+      "grad_norm": 0.5279986678775014,
+      "learning_rate": 4.926600938953418e-06,
+      "loss": 0.2139,
+      "step": 648
+    },
+    {
+      "epoch": 1.0109034267912773,
+      "grad_norm": 0.5465124083851354,
+      "learning_rate": 4.9143688728298946e-06,
+      "loss": 0.2946,
+      "step": 649
+    },
+    {
+      "epoch": 1.0124610591900312,
+      "grad_norm": 0.598599797464655,
+      "learning_rate": 4.9021373193329775e-06,
+      "loss": 0.2467,
+      "step": 650
+    },
+    {
+      "epoch": 1.014018691588785,
+      "grad_norm": 0.5911973059402644,
+      "learning_rate": 4.889906351686269e-06,
+      "loss": 0.2439,
+      "step": 651
+    },
+    {
+      "epoch": 1.0155763239875388,
+      "grad_norm": 0.5007183880780492,
+      "learning_rate": 4.8776760431098665e-06,
+      "loss": 0.2448,
+      "step": 652
+    },
+    {
+      "epoch": 1.017133956386293,
+      "grad_norm": 0.6198547585957133,
+      "learning_rate": 4.865446466819918e-06,
+      "loss": 0.2107,
+      "step": 653
+    },
+    {
+      "epoch": 1.0186915887850467,
+      "grad_norm": 0.6141060329229784,
+      "learning_rate": 4.853217696028197e-06,
+      "loss": 0.2124,
+      "step": 654
+    },
+    {
+      "epoch": 1.0202492211838006,
+      "grad_norm": 0.5500970083546799,
+      "learning_rate": 4.840989803941645e-06,
+      "loss": 0.2413,
+      "step": 655
+    },
+    {
+      "epoch": 1.0218068535825544,
+      "grad_norm": 0.5785491153029504,
+      "learning_rate": 4.828762863761948e-06,
+      "loss": 0.322,
+      "step": 656
+    },
+    {
+      "epoch": 1.0233644859813085,
+      "grad_norm": 0.5158799094788992,
+      "learning_rate": 4.816536948685091e-06,
+      "loss": 0.2458,
+      "step": 657
+    },
+    {
+      "epoch": 1.0249221183800623,
+      "grad_norm": 0.7110301489590548,
+      "learning_rate": 4.804312131900929e-06,
+      "loss": 0.1491,
+      "step": 658
+    },
+    {
+      "epoch": 1.0264797507788161,
+      "grad_norm": 0.49142641785128816,
+      "learning_rate": 4.792088486592731e-06,
+      "loss": 0.2403,
+      "step": 659
+    },
+    {
+      "epoch": 1.02803738317757,
+      "grad_norm": 0.6365446048004576,
+      "learning_rate": 4.779866085936762e-06,
+      "loss": 0.2723,
+      "step": 660
+    },
+    {
+      "epoch": 1.029595015576324,
+      "grad_norm": 0.5622051291501104,
+      "learning_rate": 4.767645003101831e-06,
+      "loss": 0.1508,
+      "step": 661
+    },
+    {
+      "epoch": 1.0311526479750779,
+      "grad_norm": 0.7978535815313595,
+      "learning_rate": 4.755425311248863e-06,
+      "loss": 0.7126,
+      "step": 662
+    },
+    {
+      "epoch": 1.0327102803738317,
+      "grad_norm": 0.49688606974260013,
+      "learning_rate": 4.7432070835304494e-06,
+      "loss": 0.1821,
+      "step": 663
+    },
+    {
+      "epoch": 1.0342679127725856,
+      "grad_norm": 0.9843736648935775,
+      "learning_rate": 4.730990393090422e-06,
+      "loss": 0.6915,
+      "step": 664
+    },
+    {
+      "epoch": 1.0358255451713396,
+      "grad_norm": 0.5197585726631249,
+      "learning_rate": 4.718775313063406e-06,
+      "loss": 0.1985,
+      "step": 665
+    },
+    {
+      "epoch": 1.0373831775700935,
+      "grad_norm": 0.7127374001801975,
+      "learning_rate": 4.70656191657439e-06,
+      "loss": 0.2209,
+      "step": 666
+    },
+    {
+      "epoch": 1.0389408099688473,
+      "grad_norm": 0.5805515887865611,
+      "learning_rate": 4.6943502767382815e-06,
+      "loss": 0.1781,
+      "step": 667
+    },
+    {
+      "epoch": 1.0404984423676011,
+      "grad_norm": 0.5812301249853751,
+      "learning_rate": 4.6821404666594715e-06,
+      "loss": 0.2362,
+      "step": 668
+    },
+    {
+      "epoch": 1.0420560747663552,
+      "grad_norm": 0.5137436961715455,
+      "learning_rate": 4.669932559431396e-06,
+      "loss": 0.2381,
+      "step": 669
+    },
+    {
+      "epoch": 1.043613707165109,
+      "grad_norm": 0.49254347681935357,
+      "learning_rate": 4.657726628136105e-06,
+      "loss": 0.2335,
+      "step": 670
+    },
+    {
+      "epoch": 1.0451713395638629,
+      "grad_norm": 0.807702133830089,
+      "learning_rate": 4.645522745843817e-06,
+      "loss": 0.1636,
+      "step": 671
+    },
+    {
+      "epoch": 1.0467289719626167,
+      "grad_norm": 0.6598785992415666,
+      "learning_rate": 4.6333209856124814e-06,
+      "loss": 0.2039,
+      "step": 672
+    },
+    {
+      "epoch": 1.0482866043613708,
+      "grad_norm": 0.5735476926058175,
+      "learning_rate": 4.621121420487345e-06,
+      "loss": 0.2344,
+      "step": 673
+    },
+    {
+      "epoch": 1.0498442367601246,
+      "grad_norm": 0.5406292279451508,
+      "learning_rate": 4.608924123500519e-06,
+      "loss": 0.1802,
+      "step": 674
+    },
+    {
+      "epoch": 1.0514018691588785,
+      "grad_norm": 0.5983081571916266,
+      "learning_rate": 4.596729167670529e-06,
+      "loss": 0.1676,
+      "step": 675
+    },
+    {
+      "epoch": 1.0529595015576323,
+      "grad_norm": 0.5810265835484077,
+      "learning_rate": 4.58453662600189e-06,
+      "loss": 0.2822,
+      "step": 676
+    },
+    {
+      "epoch": 1.0545171339563864,
+      "grad_norm": 0.6855576677633736,
+      "learning_rate": 4.572346571484661e-06,
+      "loss": 0.1977,
+      "step": 677
+    },
+    {
+      "epoch": 1.0560747663551402,
+      "grad_norm": 0.5335540200996262,
+      "learning_rate": 4.5601590770940195e-06,
+      "loss": 0.2152,
+      "step": 678
+    },
+    {
+      "epoch": 1.057632398753894,
+      "grad_norm": 0.6453393789551976,
+      "learning_rate": 4.547974215789808e-06,
+      "loss": 0.1886,
+      "step": 679
+    },
+    {
+      "epoch": 1.0591900311526479,
+      "grad_norm": 0.5915758123605618,
+      "learning_rate": 4.535792060516112e-06,
+      "loss": 0.2209,
+      "step": 680
+    },
+    {
+      "epoch": 1.060747663551402,
+      "grad_norm": 0.6302498485071517,
+      "learning_rate": 4.523612684200811e-06,
+      "loss": 0.239,
+      "step": 681
+    },
+    {
+      "epoch": 1.0623052959501558,
+      "grad_norm": 0.7064325822119844,
+      "learning_rate": 4.511436159755159e-06,
+      "loss": 0.1988,
+      "step": 682
+    },
+    {
+      "epoch": 1.0638629283489096,
+      "grad_norm": 0.5229568055974967,
+      "learning_rate": 4.499262560073328e-06,
+      "loss": 0.2219,
+      "step": 683
+    },
+    {
+      "epoch": 1.0654205607476634,
+      "grad_norm": 0.848853210404809,
+      "learning_rate": 4.487091958031984e-06,
+      "loss": 0.1813,
+      "step": 684
+    },
+    {
+      "epoch": 1.0669781931464175,
+      "grad_norm": 0.62716757784117,
+      "learning_rate": 4.474924426489847e-06,
+      "loss": 0.2433,
+      "step": 685
+    },
+    {
+      "epoch": 1.0685358255451713,
+      "grad_norm": 0.5961449452514619,
+      "learning_rate": 4.46276003828726e-06,
+      "loss": 0.2576,
+      "step": 686
+    },
+    {
+      "epoch": 1.0700934579439252,
+      "grad_norm": 0.5336164214991632,
+      "learning_rate": 4.450598866245743e-06,
+      "loss": 0.2094,
+      "step": 687
+    },
+    {
+      "epoch": 1.071651090342679,
+      "grad_norm": 0.5350296348658355,
+      "learning_rate": 4.438440983167564e-06,
+      "loss": 0.2883,
+      "step": 688
+    },
+    {
+      "epoch": 1.073208722741433,
+      "grad_norm": 0.53810152034016,
+      "learning_rate": 4.426286461835303e-06,
+      "loss": 0.1968,
+      "step": 689
+    },
+    {
+      "epoch": 1.074766355140187,
+      "grad_norm": 0.4674678946148812,
+      "learning_rate": 4.414135375011416e-06,
+      "loss": 0.1468,
+      "step": 690
+    },
+    {
+      "epoch": 1.0763239875389408,
+      "grad_norm": 0.5672398628672646,
+      "learning_rate": 4.401987795437797e-06,
+      "loss": 0.1796,
+      "step": 691
+    },
+    {
+      "epoch": 1.0778816199376946,
+      "grad_norm": 0.6409062495534954,
+      "learning_rate": 4.3898437958353435e-06,
+      "loss": 0.1536,
+      "step": 692
+    },
+    {
+      "epoch": 1.0794392523364487,
+      "grad_norm": 0.7979975112099554,
+      "learning_rate": 4.377703448903519e-06,
+      "loss": 0.7814,
+      "step": 693
+    },
+    {
+      "epoch": 1.0809968847352025,
+      "grad_norm": 0.6128214407671699,
+      "learning_rate": 4.3655668273199305e-06,
+      "loss": 0.249,
+      "step": 694
+    },
+    {
+      "epoch": 1.0825545171339563,
+      "grad_norm": 0.6671186706045426,
+      "learning_rate": 4.353434003739872e-06,
+      "loss": 0.162,
+      "step": 695
+    },
+    {
+      "epoch": 1.0841121495327102,
+      "grad_norm": 0.5657453591665842,
+      "learning_rate": 4.341305050795907e-06,
+      "loss": 0.2355,
+      "step": 696
+    },
+    {
+      "epoch": 1.0856697819314642,
+      "grad_norm": 0.5876907165277706,
+      "learning_rate": 4.329180041097425e-06,
+      "loss": 0.2898,
+      "step": 697
+    },
+    {
+      "epoch": 1.087227414330218,
+      "grad_norm": 0.5418335332942893,
+      "learning_rate": 4.3170590472302125e-06,
+      "loss": 0.2049,
+      "step": 698
+    },
+    {
+      "epoch": 1.088785046728972,
+      "grad_norm": 0.6194906159202879,
+      "learning_rate": 4.304942141756012e-06,
+      "loss": 0.2716,
+      "step": 699
+    },
+    {
+      "epoch": 1.0903426791277258,
+      "grad_norm": 0.6321743236263448,
+      "learning_rate": 4.292829397212094e-06,
+      "loss": 0.1858,
+      "step": 700
+    },
+    {
+      "epoch": 1.0919003115264798,
+      "grad_norm": 0.5884478735032194,
+      "learning_rate": 4.280720886110815e-06,
+      "loss": 0.2181,
+      "step": 701
+    },
+    {
+      "epoch": 1.0934579439252337,
+      "grad_norm": 0.512888102374138,
+      "learning_rate": 4.268616680939197e-06,
+      "loss": 0.2495,
+      "step": 702
+    },
+    {
+      "epoch": 1.0950155763239875,
+      "grad_norm": 0.6806148269647997,
+      "learning_rate": 4.256516854158476e-06,
+      "loss": 0.1811,
+      "step": 703
+    },
+    {
+      "epoch": 1.0965732087227413,
+      "grad_norm": 0.590183243846604,
+      "learning_rate": 4.244421478203681e-06,
+      "loss": 0.1759,
+      "step": 704
+    },
+    {
+      "epoch": 1.0981308411214954,
+      "grad_norm": 0.7141876070652246,
+      "learning_rate": 4.232330625483194e-06,
+      "loss": 0.2155,
+      "step": 705
+    },
+    {
+      "epoch": 1.0996884735202492,
+      "grad_norm": 0.5606859986060387,
+      "learning_rate": 4.220244368378324e-06,
+      "loss": 0.1748,
+      "step": 706
+    },
+    {
+      "epoch": 1.101246105919003,
+      "grad_norm": 0.5374363163472021,
+      "learning_rate": 4.208162779242862e-06,
+      "loss": 0.15,
+      "step": 707
+    },
+    {
+      "epoch": 1.102803738317757,
+      "grad_norm": 0.6337046437481759,
+      "learning_rate": 4.19608593040266e-06,
+      "loss": 0.2094,
+      "step": 708
+    },
+    {
+      "epoch": 1.104361370716511,
+      "grad_norm": 0.6441679702023525,
+      "learning_rate": 4.184013894155187e-06,
+      "loss": 0.1347,
+      "step": 709
+    },
+    {
+      "epoch": 1.1059190031152648,
+      "grad_norm": 0.5105399247363097,
+      "learning_rate": 4.171946742769109e-06,
+      "loss": 0.2492,
+      "step": 710
+    },
+    {
+      "epoch": 1.1074766355140186,
+      "grad_norm": 0.835104618126293,
+      "learning_rate": 4.1598845484838405e-06,
+      "loss": 0.6552,
+      "step": 711
+    },
+    {
+      "epoch": 1.1090342679127725,
+      "grad_norm": 0.5091044357624278,
+      "learning_rate": 4.147827383509127e-06,
+      "loss": 0.2459,
+      "step": 712
+    },
+    {
+      "epoch": 1.1105919003115265,
+      "grad_norm": 0.4173053058204635,
+      "learning_rate": 4.135775320024601e-06,
+      "loss": 0.1834,
+      "step": 713
+    },
+    {
+      "epoch": 1.1121495327102804,
+      "grad_norm": 0.516073956820414,
+      "learning_rate": 4.123728430179363e-06,
+      "loss": 0.2096,
+      "step": 714
+    },
+    {
+      "epoch": 1.1137071651090342,
+      "grad_norm": 0.5229385439265322,
+      "learning_rate": 4.111686786091534e-06,
+      "loss": 0.1619,
+      "step": 715
+    },
+    {
+      "epoch": 1.115264797507788,
+      "grad_norm": 0.697243292089673,
+      "learning_rate": 4.099650459847835e-06,
+      "loss": 0.2181,
+      "step": 716
+    },
+    {
+      "epoch": 1.1168224299065421,
+      "grad_norm": 0.7935991971149668,
+      "learning_rate": 4.087619523503149e-06,
+      "loss": 0.7976,
+      "step": 717
+    },
+    {
+      "epoch": 1.118380062305296,
+      "grad_norm": 0.7468568843345416,
+      "learning_rate": 4.0755940490801e-06,
+      "loss": 0.1844,
+      "step": 718
+    },
+    {
+      "epoch": 1.1199376947040498,
+      "grad_norm": 0.655225509433663,
+      "learning_rate": 4.0635741085686065e-06,
+      "loss": 0.2016,
+      "step": 719
+    },
+    {
+      "epoch": 1.1214953271028036,
+      "grad_norm": 0.6590392753571896,
+      "learning_rate": 4.051559773925462e-06,
+      "loss": 0.2175,
+      "step": 720
+    },
+    {
+      "epoch": 1.1230529595015577,
+      "grad_norm": 0.5673140068671504,
+      "learning_rate": 4.039551117073899e-06,
+      "loss": 0.1377,
+      "step": 721
+    },
+    {
+      "epoch": 1.1246105919003115,
+      "grad_norm": 0.5308315587462934,
+      "learning_rate": 4.027548209903165e-06,
+      "loss": 0.1628,
+      "step": 722
+    },
+    {
+      "epoch": 1.1261682242990654,
+      "grad_norm": 0.5915750734678845,
+      "learning_rate": 4.01555112426808e-06,
+      "loss": 0.1005,
+      "step": 723
+    },
+    {
+      "epoch": 1.1277258566978192,
+      "grad_norm": 0.6605084387776982,
+      "learning_rate": 4.00355993198862e-06,
+      "loss": 0.1835,
+      "step": 724
+    },
+    {
+      "epoch": 1.1292834890965733,
+      "grad_norm": 0.5150684098328038,
+      "learning_rate": 3.991574704849474e-06,
+      "loss": 0.2452,
+      "step": 725
+    },
+    {
+      "epoch": 1.1308411214953271,
+      "grad_norm": 0.5762133251543371,
+      "learning_rate": 3.97959551459963e-06,
+      "loss": 0.7117,
+      "step": 726
+    },
+    {
+      "epoch": 1.132398753894081,
+      "grad_norm": 0.5992236027329398,
+      "learning_rate": 3.967622432951931e-06,
+      "loss": 0.2487,
+      "step": 727
+    },
+    {
+      "epoch": 1.1339563862928348,
+      "grad_norm": 0.6328864728376592,
+      "learning_rate": 3.955655531582649e-06,
+      "loss": 0.237,
+      "step": 728
+    },
+    {
+      "epoch": 1.1355140186915889,
+      "grad_norm": 0.5851714493036804,
+      "learning_rate": 3.943694882131061e-06,
+      "loss": 0.2293,
+      "step": 729
+    },
+    {
+      "epoch": 1.1370716510903427,
+      "grad_norm": 0.6659466040289207,
+      "learning_rate": 3.931740556199021e-06,
+      "loss": 0.7144,
+      "step": 730
+    },
+    {
+      "epoch": 1.1386292834890965,
+      "grad_norm": 0.8181017630361279,
+      "learning_rate": 3.91979262535052e-06,
+      "loss": 0.4389,
+      "step": 731
+    },
+    {
+      "epoch": 1.1401869158878504,
+      "grad_norm": 0.6310519685249386,
+      "learning_rate": 3.907851161111269e-06,
+      "loss": 0.1969,
+      "step": 732
+    },
+    {
+      "epoch": 1.1417445482866044,
+      "grad_norm": 0.5414429317292473,
+      "learning_rate": 3.895916234968267e-06,
+      "loss": 0.1969,
+      "step": 733
+    },
+    {
+      "epoch": 1.1433021806853583,
+      "grad_norm": 0.5694107952685924,
+      "learning_rate": 3.883987918369373e-06,
+      "loss": 0.2409,
+      "step": 734
+    },
+    {
+      "epoch": 1.144859813084112,
+      "grad_norm": 0.5303093457895504,
+      "learning_rate": 3.8720662827228774e-06,
+      "loss": 0.1599,
+      "step": 735
+    },
+    {
+      "epoch": 1.146417445482866,
+      "grad_norm": 0.5798145908998529,
+      "learning_rate": 3.860151399397077e-06,
+      "loss": 0.2864,
+      "step": 736
+    },
+    {
+      "epoch": 1.14797507788162,
+      "grad_norm": 0.7216318366333795,
+      "learning_rate": 3.848243339719841e-06,
+      "loss": 0.1274,
+      "step": 737
+    },
+    {
+      "epoch": 1.1495327102803738,
+      "grad_norm": 0.5363385976748041,
+      "learning_rate": 3.836342174978199e-06,
+      "loss": 0.1679,
+      "step": 738
+    },
+    {
+      "epoch": 1.1510903426791277,
+      "grad_norm": 0.5904714591263907,
+      "learning_rate": 3.824447976417897e-06,
+      "loss": 0.1803,
+      "step": 739
+    },
+    {
+      "epoch": 1.1526479750778815,
+      "grad_norm": 0.7957820087907058,
+      "learning_rate": 3.8125608152429777e-06,
+      "loss": 0.2525,
+      "step": 740
+    },
+    {
+      "epoch": 1.1542056074766356,
+      "grad_norm": 0.6852233460884398,
+      "learning_rate": 3.8006807626153565e-06,
+      "loss": 0.6679,
+      "step": 741
+    },
+    {
+      "epoch": 1.1557632398753894,
+      "grad_norm": 0.7058769394932962,
+      "learning_rate": 3.7888078896543984e-06,
+      "loss": 0.1596,
+      "step": 742
+    },
+    {
+      "epoch": 1.1573208722741433,
+      "grad_norm": 0.5829336602902025,
+      "learning_rate": 3.776942267436482e-06,
+      "loss": 0.1256,
+      "step": 743
+    },
+    {
+      "epoch": 1.158878504672897,
+      "grad_norm": 0.4936789586579176,
+      "learning_rate": 3.7650839669945804e-06,
+      "loss": 0.2213,
+      "step": 744
+    },
+    {
+      "epoch": 1.1604361370716512,
+      "grad_norm": 0.6225558777529572,
+      "learning_rate": 3.7532330593178356e-06,
+      "loss": 0.1667,
+      "step": 745
+    },
+    {
+      "epoch": 1.161993769470405,
+      "grad_norm": 0.5614342757375034,
+      "learning_rate": 3.741389615351136e-06,
+      "loss": 0.1611,
+      "step": 746
+    },
+    {
+      "epoch": 1.1635514018691588,
+      "grad_norm": 0.5754099281842366,
+      "learning_rate": 3.729553705994685e-06,
+      "loss": 0.2156,
+      "step": 747
+    },
+    {
+      "epoch": 1.1651090342679127,
+      "grad_norm": 0.5573995209639809,
+      "learning_rate": 3.7177254021035824e-06,
+      "loss": 0.1777,
+      "step": 748
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 0.6243355246035951,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.2092,
+      "step": 749
+    },
+    {
+      "epoch": 1.1682242990654206,
+      "grad_norm": 0.7072395196568583,
+      "learning_rate": 3.694091893909746e-06,
+      "loss": 0.1916,
+      "step": 750
+    },
+    {
+      "epoch": 1.1697819314641744,
+      "grad_norm": 0.5876756814093886,
+      "learning_rate": 3.6822868310878683e-06,
+      "loss": 0.1514,
+      "step": 751
+    },
+    {
+      "epoch": 1.1713395638629283,
+      "grad_norm": 0.51963445472172,
+      "learning_rate": 3.670489656692202e-06,
+      "loss": 0.2446,
+      "step": 752
+    },
+    {
+      "epoch": 1.1728971962616823,
+      "grad_norm": 0.6338437719274572,
+      "learning_rate": 3.658700441345959e-06,
+      "loss": 0.2041,
+      "step": 753
+    },
+    {
+      "epoch": 1.1744548286604362,
+      "grad_norm": 0.7029067603645407,
+      "learning_rate": 3.646919255624711e-06,
+      "loss": 0.1852,
+      "step": 754
+    },
+    {
+      "epoch": 1.17601246105919,
+      "grad_norm": 2.8451831521816353,
+      "learning_rate": 3.6351461700559564e-06,
+      "loss": 0.6003,
+      "step": 755
+    },
+    {
+      "epoch": 1.1775700934579438,
+      "grad_norm": 0.5800825083133982,
+      "learning_rate": 3.623381255118702e-06,
+      "loss": 0.1983,
+      "step": 756
+    },
+    {
+      "epoch": 1.179127725856698,
+      "grad_norm": 0.7108249246289187,
+      "learning_rate": 3.6116245812430404e-06,
+      "loss": 0.1904,
+      "step": 757
+    },
+    {
+      "epoch": 1.1806853582554517,
+      "grad_norm": 0.5537550002098082,
+      "learning_rate": 3.5998762188097364e-06,
+      "loss": 0.2697,
+      "step": 758
+    },
+    {
+      "epoch": 1.1822429906542056,
+      "grad_norm": 0.5354252171867416,
+      "learning_rate": 3.588136238149791e-06,
+      "loss": 0.2256,
+      "step": 759
+    },
+    {
+      "epoch": 1.1838006230529594,
+      "grad_norm": 0.5746205607851972,
+      "learning_rate": 3.5764047095440313e-06,
+      "loss": 0.2216,
+      "step": 760
+    },
+    {
+      "epoch": 1.1853582554517135,
+      "grad_norm": 0.6474421569647941,
+      "learning_rate": 3.5646817032226855e-06,
+      "loss": 0.18,
+      "step": 761
+    },
+    {
+      "epoch": 1.1869158878504673,
+      "grad_norm": 0.6499250531970322,
+      "learning_rate": 3.552967289364967e-06,
+      "loss": 0.1564,
+      "step": 762
+    },
+    {
+      "epoch": 1.1884735202492211,
+      "grad_norm": 0.6915421275577466,
+      "learning_rate": 3.541261538098647e-06,
+      "loss": 0.1937,
+      "step": 763
+    },
+    {
+      "epoch": 1.190031152647975,
+      "grad_norm": 0.7484332653622916,
+      "learning_rate": 3.529564519499641e-06,
+      "loss": 0.8181,
+      "step": 764
+    },
+    {
+      "epoch": 1.191588785046729,
+      "grad_norm": 0.5487998303234818,
+      "learning_rate": 3.517876303591581e-06,
+      "loss": 0.2187,
+      "step": 765
+    },
+    {
+      "epoch": 1.1931464174454829,
+      "grad_norm": 0.5468867601412329,
+      "learning_rate": 3.506196960345413e-06,
+      "loss": 0.2741,
+      "step": 766
+    },
+    {
+      "epoch": 1.1947040498442367,
+      "grad_norm": 0.5333599222192844,
+      "learning_rate": 3.494526559678958e-06,
+      "loss": 0.196,
+      "step": 767
+    },
+    {
+      "epoch": 1.1962616822429906,
+      "grad_norm": 0.6163139809793089,
+      "learning_rate": 3.4828651714565056e-06,
+      "loss": 0.1689,
+      "step": 768
+    },
+    {
+      "epoch": 1.1978193146417446,
+      "grad_norm": 0.6118788241013663,
+      "learning_rate": 3.4712128654883915e-06,
+      "loss": 0.6684,
+      "step": 769
+    },
+    {
+      "epoch": 1.1993769470404985,
+      "grad_norm": 0.7588288371409719,
+      "learning_rate": 3.459569711530586e-06,
+      "loss": 0.1844,
+      "step": 770
+    },
+    {
+      "epoch": 1.2009345794392523,
+      "grad_norm": 0.512022978994329,
+      "learning_rate": 3.447935779284265e-06,
+      "loss": 0.2842,
+      "step": 771
+    },
+    {
+      "epoch": 1.2024922118380061,
+      "grad_norm": 0.5591196557510058,
+      "learning_rate": 3.436311138395402e-06,
+      "loss": 0.1664,
+      "step": 772
+    },
+    {
+      "epoch": 1.2040498442367602,
+      "grad_norm": 0.8115349813699498,
+      "learning_rate": 3.424695858454347e-06,
+      "loss": 0.307,
+      "step": 773
+    },
+    {
+      "epoch": 1.205607476635514,
+      "grad_norm": 0.5218970414061956,
+      "learning_rate": 3.4130900089954142e-06,
+      "loss": 0.2348,
+      "step": 774
+    },
+    {
+      "epoch": 1.2071651090342679,
+      "grad_norm": 0.6595592104442388,
+      "learning_rate": 3.4014936594964608e-06,
+      "loss": 0.2381,
+      "step": 775
+    },
+    {
+      "epoch": 1.2087227414330217,
+      "grad_norm": 0.5724303041429597,
+      "learning_rate": 3.3899068793784717e-06,
+      "loss": 0.2087,
+      "step": 776
+    },
+    {
+      "epoch": 1.2102803738317758,
+      "grad_norm": 0.5855841777284334,
+      "learning_rate": 3.378329738005144e-06,
+      "loss": 0.2215,
+      "step": 777
+    },
+    {
+      "epoch": 1.2118380062305296,
+      "grad_norm": 0.5794484038490832,
+      "learning_rate": 3.3667623046824783e-06,
+      "loss": 0.1275,
+      "step": 778
+    },
+    {
+      "epoch": 1.2133956386292835,
+      "grad_norm": 0.8252092993882152,
+      "learning_rate": 3.3552046486583547e-06,
+      "loss": 0.2076,
+      "step": 779
+    },
+    {
+      "epoch": 1.2149532710280373,
+      "grad_norm": 0.5054517090774381,
+      "learning_rate": 3.3436568391221215e-06,
+      "loss": 0.2269,
+      "step": 780
+    },
+    {
+      "epoch": 1.2165109034267914,
+      "grad_norm": 0.6361245982706095,
+      "learning_rate": 3.3321189452041814e-06,
+      "loss": 0.1765,
+      "step": 781
+    },
+    {
+      "epoch": 1.2180685358255452,
+      "grad_norm": 0.6359846065805195,
+      "learning_rate": 3.3205910359755823e-06,
+      "loss": 0.213,
+      "step": 782
+    },
+    {
+      "epoch": 1.219626168224299,
+      "grad_norm": 0.575827824012367,
+      "learning_rate": 3.309073180447593e-06,
+      "loss": 0.2128,
+      "step": 783
+    },
+    {
+      "epoch": 1.2211838006230529,
+      "grad_norm": 0.594472470089166,
+      "learning_rate": 3.2975654475713005e-06,
+      "loss": 0.1867,
+      "step": 784
+    },
+    {
+      "epoch": 1.222741433021807,
+      "grad_norm": 0.6841555235014364,
+      "learning_rate": 3.286067906237188e-06,
+      "loss": 0.1615,
+      "step": 785
+    },
+    {
+      "epoch": 1.2242990654205608,
+      "grad_norm": 0.5190015178190842,
+      "learning_rate": 3.274580625274737e-06,
+      "loss": 0.2139,
+      "step": 786
+    },
+    {
+      "epoch": 1.2258566978193146,
+      "grad_norm": 0.675548730024154,
+      "learning_rate": 3.263103673451996e-06,
+      "loss": 0.2535,
+      "step": 787
+    },
+    {
+      "epoch": 1.2274143302180685,
+      "grad_norm": 0.7466550657712839,
+      "learning_rate": 3.2516371194751838e-06,
+      "loss": 0.2706,
+      "step": 788
+    },
+    {
+      "epoch": 1.2289719626168225,
+      "grad_norm": 0.6666523188287753,
+      "learning_rate": 3.24018103198827e-06,
+      "loss": 0.1727,
+      "step": 789
+    },
+    {
+      "epoch": 1.2305295950155763,
+      "grad_norm": 0.6337742576262074,
+      "learning_rate": 3.22873547957257e-06,
+      "loss": 0.1979,
+      "step": 790
+    },
+    {
+      "epoch": 1.2320872274143302,
+      "grad_norm": 0.5592619596660579,
+      "learning_rate": 3.217300530746331e-06,
+      "loss": 0.2195,
+      "step": 791
+    },
+    {
+      "epoch": 1.233644859813084,
+      "grad_norm": 0.6098769962820866,
+      "learning_rate": 3.2058762539643214e-06,
+      "loss": 0.231,
+      "step": 792
+    },
+    {
+      "epoch": 1.235202492211838,
+      "grad_norm": 0.5644158815873026,
+      "learning_rate": 3.1944627176174204e-06,
+      "loss": 0.2373,
+      "step": 793
+    },
+    {
+      "epoch": 1.236760124610592,
+      "grad_norm": 0.9387436364990357,
+      "learning_rate": 3.1830599900322135e-06,
+      "loss": 0.8066,
+      "step": 794
+    },
+    {
+      "epoch": 1.2383177570093458,
+      "grad_norm": 0.5202721107932754,
+      "learning_rate": 3.1716681394705783e-06,
+      "loss": 0.2198,
+      "step": 795
+    },
+    {
+      "epoch": 1.2398753894080996,
+      "grad_norm": 0.5496594883977267,
+      "learning_rate": 3.1602872341292772e-06,
+      "loss": 0.1961,
+      "step": 796
+    },
+    {
+      "epoch": 1.2414330218068537,
+      "grad_norm": 0.6098209049046375,
+      "learning_rate": 3.148917342139548e-06,
+      "loss": 0.2209,
+      "step": 797
+    },
+    {
+      "epoch": 1.2429906542056075,
+      "grad_norm": 0.6346192193491992,
+      "learning_rate": 3.1375585315667047e-06,
+      "loss": 0.1802,
+      "step": 798
+    },
+    {
+      "epoch": 1.2445482866043613,
+      "grad_norm": 0.765060316147878,
+      "learning_rate": 3.1262108704097137e-06,
+      "loss": 0.7635,
+      "step": 799
+    },
+    {
+      "epoch": 1.2461059190031152,
+      "grad_norm": 0.5914294283180993,
+      "learning_rate": 3.1148744266008024e-06,
+      "loss": 0.1632,
+      "step": 800
+    },
+    {
+      "epoch": 1.2476635514018692,
+      "grad_norm": 0.6570214224554458,
+      "learning_rate": 3.1035492680050416e-06,
+      "loss": 0.2406,
+      "step": 801
+    },
+    {
+      "epoch": 1.249221183800623,
+      "grad_norm": 0.5452895252692876,
+      "learning_rate": 3.0922354624199487e-06,
+      "loss": 0.2458,
+      "step": 802
+    },
+    {
+      "epoch": 1.250778816199377,
+      "grad_norm": 0.8021393101552716,
+      "learning_rate": 3.080933077575071e-06,
+      "loss": 0.2422,
+      "step": 803
+    },
+    {
+      "epoch": 1.2523364485981308,
+      "grad_norm": 0.5980931498074656,
+      "learning_rate": 3.0696421811315923e-06,
+      "loss": 0.3071,
+      "step": 804
+    },
+    {
+      "epoch": 1.2538940809968846,
+      "grad_norm": 0.7151706000107999,
+      "learning_rate": 3.0583628406819167e-06,
+      "loss": 0.1693,
+      "step": 805
+    },
+    {
+      "epoch": 1.2554517133956387,
+      "grad_norm": 0.7557667483815892,
+      "learning_rate": 3.0470951237492724e-06,
+      "loss": 0.6472,
+      "step": 806
+    },
+    {
+      "epoch": 1.2570093457943925,
+      "grad_norm": 0.5383018776710425,
+      "learning_rate": 3.035839097787301e-06,
+      "loss": 0.2179,
+      "step": 807
+    },
+    {
+      "epoch": 1.2585669781931463,
+      "grad_norm": 0.5591749963571674,
+      "learning_rate": 3.024594830179658e-06,
+      "loss": 0.2537,
+      "step": 808
+    },
+    {
+      "epoch": 1.2601246105919004,
+      "grad_norm": 0.5434411022961728,
+      "learning_rate": 3.0133623882396067e-06,
+      "loss": 0.1933,
+      "step": 809
+    },
+    {
+      "epoch": 1.2616822429906542,
+      "grad_norm": 0.6201199414157127,
+      "learning_rate": 3.0021418392096215e-06,
+      "loss": 0.2193,
+      "step": 810
+    },
+    {
+      "epoch": 1.263239875389408,
+      "grad_norm": 0.669872737304252,
+      "learning_rate": 2.9909332502609744e-06,
+      "loss": 0.1446,
+      "step": 811
+    },
+    {
+      "epoch": 1.264797507788162,
+      "grad_norm": 0.5828212724381946,
+      "learning_rate": 2.979736688493343e-06,
+      "loss": 0.253,
+      "step": 812
+    },
+    {
+      "epoch": 1.2663551401869158,
+      "grad_norm": 0.5644246073635343,
+      "learning_rate": 2.968552220934402e-06,
+      "loss": 0.2241,
+      "step": 813
+    },
+    {
+      "epoch": 1.2679127725856698,
+      "grad_norm": 0.5607183966563147,
+      "learning_rate": 2.9573799145394266e-06,
+      "loss": 0.2745,
+      "step": 814
+    },
+    {
+      "epoch": 1.2694704049844237,
+      "grad_norm": 0.474278573645314,
+      "learning_rate": 2.9462198361908893e-06,
+      "loss": 0.1337,
+      "step": 815
+    },
+    {
+      "epoch": 1.2710280373831775,
+      "grad_norm": 1.3740104676242695,
+      "learning_rate": 2.9350720526980592e-06,
+      "loss": 0.7275,
+      "step": 816
+    },
+    {
+      "epoch": 1.2725856697819315,
+      "grad_norm": 0.711919465609596,
+      "learning_rate": 2.9239366307966033e-06,
+      "loss": 0.7703,
+      "step": 817
+    },
+    {
+      "epoch": 1.2741433021806854,
+      "grad_norm": 0.6630095165519386,
+      "learning_rate": 2.912813637148187e-06,
+      "loss": 0.1362,
+      "step": 818
+    },
+    {
+      "epoch": 1.2757009345794392,
+      "grad_norm": 0.602978479954925,
+      "learning_rate": 2.9017031383400772e-06,
+      "loss": 0.1538,
+      "step": 819
+    },
+    {
+      "epoch": 1.277258566978193,
+      "grad_norm": 0.7257094178132808,
+      "learning_rate": 2.890605200884733e-06,
+      "loss": 0.1533,
+      "step": 820
+    },
+    {
+      "epoch": 1.278816199376947,
+      "grad_norm": 0.5041273431174441,
+      "learning_rate": 2.879519891219424e-06,
+      "loss": 0.195,
+      "step": 821
+    },
+    {
+      "epoch": 1.280373831775701,
+      "grad_norm": 0.5315316010733497,
+      "learning_rate": 2.868447275705821e-06,
+      "loss": 0.1634,
+      "step": 822
+    },
+    {
+      "epoch": 1.2819314641744548,
+      "grad_norm": 0.6131393193633405,
+      "learning_rate": 2.8573874206296005e-06,
+      "loss": 0.2044,
+      "step": 823
+    },
+    {
+      "epoch": 1.2834890965732086,
+      "grad_norm": 0.5970183436040595,
+      "learning_rate": 2.846340392200051e-06,
+      "loss": 0.2539,
+      "step": 824
+    },
+    {
+      "epoch": 1.2850467289719627,
+      "grad_norm": 0.5065607828613443,
+      "learning_rate": 2.8353062565496715e-06,
+      "loss": 0.166,
+      "step": 825
+    },
+    {
+      "epoch": 1.2866043613707165,
+      "grad_norm": 0.6837585450327539,
+      "learning_rate": 2.824285079733788e-06,
+      "loss": 0.2406,
+      "step": 826
+    },
+    {
+      "epoch": 1.2881619937694704,
+      "grad_norm": 0.5599124367538377,
+      "learning_rate": 2.8132769277301374e-06,
+      "loss": 0.157,
+      "step": 827
+    },
+    {
+      "epoch": 1.2897196261682242,
+      "grad_norm": 0.49511182258884123,
+      "learning_rate": 2.8022818664384945e-06,
+      "loss": 0.1824,
+      "step": 828
+    },
+    {
+      "epoch": 1.291277258566978,
+      "grad_norm": 0.5896227768783965,
+      "learning_rate": 2.791299961680255e-06,
+      "loss": 0.258,
+      "step": 829
+    },
+    {
+      "epoch": 1.2928348909657321,
+      "grad_norm": 0.5730839746888057,
+      "learning_rate": 2.7803312791980697e-06,
+      "loss": 0.2379,
+      "step": 830
+    },
+    {
+      "epoch": 1.294392523364486,
+      "grad_norm": 0.6736027436106719,
+      "learning_rate": 2.769375884655421e-06,
+      "loss": 0.2515,
+      "step": 831
+    },
+    {
+      "epoch": 1.2959501557632398,
+      "grad_norm": 0.6242264709025079,
+      "learning_rate": 2.758433843636252e-06,
+      "loss": 0.1923,
+      "step": 832
+    },
+    {
+      "epoch": 1.2975077881619939,
+      "grad_norm": 0.6286501649370759,
+      "learning_rate": 2.7475052216445588e-06,
+      "loss": 0.1842,
+      "step": 833
+    },
+    {
+      "epoch": 1.2990654205607477,
+      "grad_norm": 0.5914028874575505,
+      "learning_rate": 2.7365900841040104e-06,
+      "loss": 0.1473,
+      "step": 834
+    },
+    {
+      "epoch": 1.3006230529595015,
+      "grad_norm": 0.6185390608342917,
+      "learning_rate": 2.7256884963575536e-06,
+      "loss": 0.2347,
+      "step": 835
+    },
+    {
+      "epoch": 1.3021806853582554,
+      "grad_norm": 0.5490349592608532,
+      "learning_rate": 2.714800523667011e-06,
+      "loss": 0.1875,
+      "step": 836
+    },
+    {
+      "epoch": 1.3037383177570092,
+      "grad_norm": 0.619509184599072,
+      "learning_rate": 2.703926231212708e-06,
+      "loss": 0.2472,
+      "step": 837
+    },
+    {
+      "epoch": 1.3052959501557633,
+      "grad_norm": 0.7158852617243849,
+      "learning_rate": 2.6930656840930713e-06,
+      "loss": 0.2468,
+      "step": 838
+    },
+    {
+      "epoch": 1.3068535825545171,
+      "grad_norm": 0.723996693315863,
+      "learning_rate": 2.6822189473242422e-06,
+      "loss": 0.2276,
+      "step": 839
+    },
+    {
+      "epoch": 1.308411214953271,
+      "grad_norm": 0.5632679246702721,
+      "learning_rate": 2.671386085839682e-06,
+      "loss": 0.2215,
+      "step": 840
+    },
+    {
+      "epoch": 1.309968847352025,
+      "grad_norm": 0.5985281299233187,
+      "learning_rate": 2.6605671644897967e-06,
+      "loss": 0.3572,
+      "step": 841
+    },
+    {
+      "epoch": 1.3115264797507789,
+      "grad_norm": 0.686365038362172,
+      "learning_rate": 2.6497622480415346e-06,
+      "loss": 0.3266,
+      "step": 842
+    },
+    {
+      "epoch": 1.3130841121495327,
+      "grad_norm": 0.6582167035193466,
+      "learning_rate": 2.6389714011780078e-06,
+      "loss": 0.2669,
+      "step": 843
+    },
+    {
+      "epoch": 1.3146417445482865,
+      "grad_norm": 0.5788593198598405,
+      "learning_rate": 2.628194688498096e-06,
+      "loss": 0.2572,
+      "step": 844
+    },
+    {
+      "epoch": 1.3161993769470404,
+      "grad_norm": 0.5457859405003289,
+      "learning_rate": 2.6174321745160707e-06,
+      "loss": 0.2016,
+      "step": 845
+    },
+    {
+      "epoch": 1.3177570093457944,
+      "grad_norm": 0.5093352344419979,
+      "learning_rate": 2.606683923661204e-06,
+      "loss": 0.2152,
+      "step": 846
+    },
+    {
+      "epoch": 1.3193146417445483,
+      "grad_norm": 0.6332315738079737,
+      "learning_rate": 2.5959500002773763e-06,
+      "loss": 0.1449,
+      "step": 847
+    },
+    {
+      "epoch": 1.320872274143302,
+      "grad_norm": 0.4685817262067038,
+      "learning_rate": 2.585230468622705e-06,
+      "loss": 0.2152,
+      "step": 848
+    },
+    {
+      "epoch": 1.3224299065420562,
+      "grad_norm": 0.49981674536826676,
+      "learning_rate": 2.574525392869143e-06,
+      "loss": 0.1783,
+      "step": 849
+    },
+    {
+      "epoch": 1.32398753894081,
+      "grad_norm": 1.0724140185603979,
+      "learning_rate": 2.563834837102115e-06,
+      "loss": 0.6521,
+      "step": 850
+    },
+    {
+      "epoch": 1.3255451713395638,
+      "grad_norm": 0.5582924411201017,
+      "learning_rate": 2.5531588653201134e-06,
+      "loss": 0.1696,
+      "step": 851
+    },
+    {
+      "epoch": 1.3271028037383177,
+      "grad_norm": 0.5064036048508646,
+      "learning_rate": 2.542497541434329e-06,
+      "loss": 0.1616,
+      "step": 852
+    },
+    {
+      "epoch": 1.3286604361370715,
+      "grad_norm": 0.7424483742643145,
+      "learning_rate": 2.531850929268258e-06,
+      "loss": 0.3,
+      "step": 853
+    },
+    {
+      "epoch": 1.3302180685358256,
+      "grad_norm": 0.6230465177616844,
+      "learning_rate": 2.5212190925573366e-06,
+      "loss": 0.638,
+      "step": 854
+    },
+    {
+      "epoch": 1.3317757009345794,
+      "grad_norm": 0.5918692120070366,
+      "learning_rate": 2.510602094948535e-06,
+      "loss": 0.1742,
+      "step": 855
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.6010294004442239,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.2177,
+      "step": 856
+    },
+    {
+      "epoch": 1.3348909657320873,
+      "grad_norm": 0.8792131655034126,
+      "learning_rate": 2.4894128711806603e-06,
+      "loss": 0.1769,
+      "step": 857
+    },
+    {
+      "epoch": 1.3364485981308412,
+      "grad_norm": 0.5439058673773991,
+      "learning_rate": 2.4788407718698487e-06,
+      "loss": 0.1423,
+      "step": 858
+    },
+    {
+      "epoch": 1.338006230529595,
+      "grad_norm": 0.6698324814866782,
+      "learning_rate": 2.468283765356931e-06,
+      "loss": 0.2016,
+      "step": 859
+    },
+    {
+      "epoch": 1.3395638629283488,
+      "grad_norm": 0.7209231410413558,
+      "learning_rate": 2.4577419148409123e-06,
+      "loss": 0.2526,
+      "step": 860
+    },
+    {
+      "epoch": 1.3411214953271027,
+      "grad_norm": 0.5554513132189719,
+      "learning_rate": 2.447215283430072e-06,
+      "loss": 0.2054,
+      "step": 861
+    },
+    {
+      "epoch": 1.3426791277258567,
+      "grad_norm": 0.6204181777681318,
+      "learning_rate": 2.436703934141583e-06,
+      "loss": 0.2055,
+      "step": 862
+    },
+    {
+      "epoch": 1.3442367601246106,
+      "grad_norm": 0.6174847596208328,
+      "learning_rate": 2.4262079299011287e-06,
+      "loss": 0.2214,
+      "step": 863
+    },
+    {
+      "epoch": 1.3457943925233644,
+      "grad_norm": 0.596347448095447,
+      "learning_rate": 2.4157273335425296e-06,
+      "loss": 0.2396,
+      "step": 864
+    },
+    {
+      "epoch": 1.3473520249221185,
+      "grad_norm": 0.6945784756475979,
+      "learning_rate": 2.4052622078073714e-06,
+      "loss": 0.6505,
+      "step": 865
+    },
+    {
+      "epoch": 1.3489096573208723,
+      "grad_norm": 0.6484352021895371,
+      "learning_rate": 2.394812615344622e-06,
+      "loss": 0.2738,
+      "step": 866
+    },
+    {
+      "epoch": 1.3504672897196262,
+      "grad_norm": 0.5351329498045785,
+      "learning_rate": 2.384378618710265e-06,
+      "loss": 0.1667,
+      "step": 867
+    },
+    {
+      "epoch": 1.35202492211838,
+      "grad_norm": 0.5884847134807437,
+      "learning_rate": 2.3739602803669126e-06,
+      "loss": 0.2267,
+      "step": 868
+    },
+    {
+      "epoch": 1.3535825545171338,
+      "grad_norm": 0.7327148616338638,
+      "learning_rate": 2.363557662683446e-06,
+      "loss": 0.7327,
+      "step": 869
+    },
+    {
+      "epoch": 1.355140186915888,
+      "grad_norm": 0.712258388986809,
+      "learning_rate": 2.3531708279346347e-06,
+      "loss": 0.2281,
+      "step": 870
+    },
+    {
+      "epoch": 1.3566978193146417,
+      "grad_norm": 0.598978404959667,
+      "learning_rate": 2.3427998383007605e-06,
+      "loss": 0.2433,
+      "step": 871
+    },
+    {
+      "epoch": 1.3582554517133956,
+      "grad_norm": 0.6295698566133894,
+      "learning_rate": 2.3324447558672543e-06,
+      "loss": 0.2569,
+      "step": 872
+    },
+    {
+      "epoch": 1.3598130841121496,
+      "grad_norm": 0.5986998801181967,
+      "learning_rate": 2.3221056426243112e-06,
+      "loss": 0.1962,
+      "step": 873
+    },
+    {
+      "epoch": 1.3613707165109035,
+      "grad_norm": 0.5674067782697227,
+      "learning_rate": 2.3117825604665405e-06,
+      "loss": 0.2158,
+      "step": 874
+    },
+    {
+      "epoch": 1.3629283489096573,
+      "grad_norm": 0.6884839665928056,
+      "learning_rate": 2.3014755711925695e-06,
+      "loss": 0.2329,
+      "step": 875
+    },
+    {
+      "epoch": 1.3644859813084111,
+      "grad_norm": 0.6885485571502845,
+      "learning_rate": 2.291184736504695e-06,
+      "loss": 0.1955,
+      "step": 876
+    },
+    {
+      "epoch": 1.366043613707165,
+      "grad_norm": 0.5473669303308986,
+      "learning_rate": 2.2809101180084963e-06,
+      "loss": 0.2171,
+      "step": 877
+    },
+    {
+      "epoch": 1.367601246105919,
+      "grad_norm": 0.5860469422390239,
+      "learning_rate": 2.2706517772124875e-06,
+      "loss": 0.3142,
+      "step": 878
+    },
+    {
+      "epoch": 1.3691588785046729,
+      "grad_norm": 0.4589130720603424,
+      "learning_rate": 2.2604097755277244e-06,
+      "loss": 0.1892,
+      "step": 879
+    },
+    {
+      "epoch": 1.3707165109034267,
+      "grad_norm": 0.56817784386682,
+      "learning_rate": 2.250184174267459e-06,
+      "loss": 0.2104,
+      "step": 880
+    },
+    {
+      "epoch": 1.3722741433021808,
+      "grad_norm": 2.9170440438070244,
+      "learning_rate": 2.239975034646756e-06,
+      "loss": 0.4075,
+      "step": 881
+    },
+    {
+      "epoch": 1.3738317757009346,
+      "grad_norm": 0.5936408877749505,
+      "learning_rate": 2.2297824177821374e-06,
+      "loss": 0.1725,
+      "step": 882
+    },
+    {
+      "epoch": 1.3753894080996885,
+      "grad_norm": 0.5570669743525561,
+      "learning_rate": 2.2196063846912142e-06,
+      "loss": 0.1089,
+      "step": 883
+    },
+    {
+      "epoch": 1.3769470404984423,
+      "grad_norm": 0.7043574775467037,
+      "learning_rate": 2.209446996292312e-06,
+      "loss": 0.2412,
+      "step": 884
+    },
+    {
+      "epoch": 1.3785046728971961,
+      "grad_norm": 0.6356878135461869,
+      "learning_rate": 2.1993043134041213e-06,
+      "loss": 0.2314,
+      "step": 885
+    },
+    {
+      "epoch": 1.3800623052959502,
+      "grad_norm": 0.6088173529635026,
+      "learning_rate": 2.1891783967453235e-06,
+      "loss": 0.1986,
+      "step": 886
+    },
+    {
+      "epoch": 1.381619937694704,
+      "grad_norm": 0.6505134414780895,
+      "learning_rate": 2.1790693069342293e-06,
+      "loss": 0.2621,
+      "step": 887
+    },
+    {
+      "epoch": 1.3831775700934579,
+      "grad_norm": 0.5229082442758375,
+      "learning_rate": 2.168977104488415e-06,
+      "loss": 0.1857,
+      "step": 888
+    },
+    {
+      "epoch": 1.384735202492212,
+      "grad_norm": 0.6219752894683416,
+      "learning_rate": 2.158901849824363e-06,
+      "loss": 0.2393,
+      "step": 889
+    },
+    {
+      "epoch": 1.3862928348909658,
+      "grad_norm": 0.6712717762459158,
+      "learning_rate": 2.1488436032571e-06,
+      "loss": 0.1975,
+      "step": 890
+    },
+    {
+      "epoch": 1.3878504672897196,
+      "grad_norm": 0.5882444001428154,
+      "learning_rate": 2.138802424999833e-06,
+      "loss": 0.1938,
+      "step": 891
+    },
+    {
+      "epoch": 1.3894080996884735,
+      "grad_norm": 0.7887785551105917,
+      "learning_rate": 2.128778375163588e-06,
+      "loss": 0.2438,
+      "step": 892
+    },
+    {
+      "epoch": 1.3909657320872273,
+      "grad_norm": 0.6726478463164812,
+      "learning_rate": 2.1187715137568564e-06,
+      "loss": 0.1609,
+      "step": 893
+    },
+    {
+      "epoch": 1.3925233644859814,
+      "grad_norm": 1.1772556659919378,
+      "learning_rate": 2.1087819006852327e-06,
+      "loss": 0.1874,
+      "step": 894
+    },
+    {
+      "epoch": 1.3940809968847352,
+      "grad_norm": 0.6141735195326121,
+      "learning_rate": 2.098809595751049e-06,
+      "loss": 0.1771,
+      "step": 895
+    },
+    {
+      "epoch": 1.395638629283489,
+      "grad_norm": 0.574616048911299,
+      "learning_rate": 2.08885465865303e-06,
+      "loss": 0.1973,
+      "step": 896
+    },
+    {
+      "epoch": 1.397196261682243,
+      "grad_norm": 0.7291041551854346,
+      "learning_rate": 2.078917148985921e-06,
+      "loss": 0.7534,
+      "step": 897
+    },
+    {
+      "epoch": 1.398753894080997,
+      "grad_norm": 0.5907252824302475,
+      "learning_rate": 2.06899712624015e-06,
+      "loss": 0.2802,
+      "step": 898
+    },
+    {
+      "epoch": 1.4003115264797508,
+      "grad_norm": 0.5328199626427012,
+      "learning_rate": 2.0590946498014493e-06,
+      "loss": 0.2303,
+      "step": 899
+    },
+    {
+      "epoch": 1.4018691588785046,
+      "grad_norm": 0.6150130691322249,
+      "learning_rate": 2.049209778950518e-06,
+      "loss": 0.2424,
+      "step": 900
+    },
+    {
+      "epoch": 1.4034267912772584,
+      "grad_norm": 0.6127344165171512,
+      "learning_rate": 2.0393425728626526e-06,
+      "loss": 0.1674,
+      "step": 901
+    },
+    {
+      "epoch": 1.4049844236760125,
+      "grad_norm": 0.6455010665412071,
+      "learning_rate": 2.029493090607413e-06,
+      "loss": 0.201,
+      "step": 902
+    },
+    {
+      "epoch": 1.4065420560747663,
+      "grad_norm": 0.6680633935331085,
+      "learning_rate": 2.0196613911482445e-06,
+      "loss": 0.1652,
+      "step": 903
+    },
+    {
+      "epoch": 1.4080996884735202,
+      "grad_norm": 0.5230495421575583,
+      "learning_rate": 2.0098475333421426e-06,
+      "loss": 0.1869,
+      "step": 904
+    },
+    {
+      "epoch": 1.4096573208722742,
+      "grad_norm": 0.6484127593933029,
+      "learning_rate": 2.0000515759392904e-06,
+      "loss": 0.2167,
+      "step": 905
+    },
+    {
+      "epoch": 1.411214953271028,
+      "grad_norm": 0.6594541288417016,
+      "learning_rate": 1.9902735775827146e-06,
+      "loss": 0.3135,
+      "step": 906
+    },
+    {
+      "epoch": 1.412772585669782,
+      "grad_norm": 0.6546159891601314,
+      "learning_rate": 1.980513596807932e-06,
+      "loss": 0.1546,
+      "step": 907
+    },
+    {
+      "epoch": 1.4143302180685358,
+      "grad_norm": 0.6853904172880749,
+      "learning_rate": 1.9707716920425923e-06,
+      "loss": 0.2063,
+      "step": 908
+    },
+    {
+      "epoch": 1.4158878504672896,
+      "grad_norm": 0.6131385223592029,
+      "learning_rate": 1.9610479216061393e-06,
+      "loss": 0.2792,
+      "step": 909
+    },
+    {
+      "epoch": 1.4174454828660437,
+      "grad_norm": 0.47832488985489935,
+      "learning_rate": 1.951342343709456e-06,
+      "loss": 0.2819,
+      "step": 910
+    },
+    {
+      "epoch": 1.4190031152647975,
+      "grad_norm": 0.7260584028544175,
+      "learning_rate": 1.941655016454515e-06,
+      "loss": 0.1664,
+      "step": 911
+    },
+    {
+      "epoch": 1.4205607476635513,
+      "grad_norm": 0.531050945261339,
+      "learning_rate": 1.9319859978340312e-06,
+      "loss": 0.218,
+      "step": 912
+    },
+    {
+      "epoch": 1.4221183800623054,
+      "grad_norm": 0.7077461606190044,
+      "learning_rate": 1.9223353457311178e-06,
+      "loss": 0.2129,
+      "step": 913
+    },
+    {
+      "epoch": 1.4236760124610592,
+      "grad_norm": 0.6914074130890262,
+      "learning_rate": 1.912703117918938e-06,
+      "loss": 0.1846,
+      "step": 914
+    },
+    {
+      "epoch": 1.425233644859813,
+      "grad_norm": 0.7199368018705896,
+      "learning_rate": 1.9030893720603604e-06,
+      "loss": 0.2253,
+      "step": 915
+    },
+    {
+      "epoch": 1.426791277258567,
+      "grad_norm": 0.5583301531155793,
+      "learning_rate": 1.8934941657076057e-06,
+      "loss": 0.2059,
+      "step": 916
+    },
+    {
+      "epoch": 1.4283489096573208,
+      "grad_norm": 0.5646047767722963,
+      "learning_rate": 1.8839175563019145e-06,
+      "loss": 0.2514,
+      "step": 917
+    },
+    {
+      "epoch": 1.4299065420560748,
+      "grad_norm": 0.6817330899431177,
+      "learning_rate": 1.8743596011731984e-06,
+      "loss": 0.2018,
+      "step": 918
+    },
+    {
+      "epoch": 1.4314641744548287,
+      "grad_norm": 0.6478720709719429,
+      "learning_rate": 1.8648203575396912e-06,
+      "loss": 0.1172,
+      "step": 919
+    },
+    {
+      "epoch": 1.4330218068535825,
+      "grad_norm": 0.6481753359929265,
+      "learning_rate": 1.8552998825076163e-06,
+      "loss": 0.2396,
+      "step": 920
+    },
+    {
+      "epoch": 1.4345794392523366,
+      "grad_norm": 0.6132125740119907,
+      "learning_rate": 1.845798233070833e-06,
+      "loss": 0.2065,
+      "step": 921
+    },
+    {
+      "epoch": 1.4361370716510904,
+      "grad_norm": 0.5448210616281429,
+      "learning_rate": 1.8363154661105131e-06,
+      "loss": 0.2022,
+      "step": 922
+    },
+    {
+      "epoch": 1.4376947040498442,
+      "grad_norm": 0.5931140592473454,
+      "learning_rate": 1.826851638394777e-06,
+      "loss": 0.2241,
+      "step": 923
+    },
+    {
+      "epoch": 1.439252336448598,
+      "grad_norm": 0.5732273344982183,
+      "learning_rate": 1.8174068065783768e-06,
+      "loss": 0.238,
+      "step": 924
+    },
+    {
+      "epoch": 1.440809968847352,
+      "grad_norm": 0.6031290862741796,
+      "learning_rate": 1.8079810272023373e-06,
+      "loss": 0.2065,
+      "step": 925
+    },
+    {
+      "epoch": 1.442367601246106,
+      "grad_norm": 0.7027323074692515,
+      "learning_rate": 1.7985743566936338e-06,
+      "loss": 0.1353,
+      "step": 926
+    },
+    {
+      "epoch": 1.4439252336448598,
+      "grad_norm": 0.679661751499936,
+      "learning_rate": 1.7891868513648436e-06,
+      "loss": 0.1636,
+      "step": 927
+    },
+    {
+      "epoch": 1.4454828660436136,
+      "grad_norm": 0.6123098696997167,
+      "learning_rate": 1.7798185674138163e-06,
+      "loss": 0.7362,
+      "step": 928
+    },
+    {
+      "epoch": 1.4470404984423677,
+      "grad_norm": 0.6533459270958951,
+      "learning_rate": 1.7704695609233275e-06,
+      "loss": 0.2479,
+      "step": 929
+    },
+    {
+      "epoch": 1.4485981308411215,
+      "grad_norm": 0.6018511814849575,
+      "learning_rate": 1.7611398878607544e-06,
+      "loss": 0.2135,
+      "step": 930
+    },
+    {
+      "epoch": 1.4501557632398754,
+      "grad_norm": 0.6980529317039801,
+      "learning_rate": 1.7518296040777355e-06,
+      "loss": 0.1639,
+      "step": 931
+    },
+    {
+      "epoch": 1.4517133956386292,
+      "grad_norm": 0.6062016861868827,
+      "learning_rate": 1.742538765309832e-06,
+      "loss": 0.2538,
+      "step": 932
+    },
+    {
+      "epoch": 1.453271028037383,
+      "grad_norm": 0.6539624285807091,
+      "learning_rate": 1.7332674271762018e-06,
+      "loss": 0.1811,
+      "step": 933
+    },
+    {
+      "epoch": 1.4548286604361371,
+      "grad_norm": 0.7122687298587204,
+      "learning_rate": 1.724015645179264e-06,
+      "loss": 0.1891,
+      "step": 934
+    },
+    {
+      "epoch": 1.456386292834891,
+      "grad_norm": 0.7014287750312252,
+      "learning_rate": 1.7147834747043651e-06,
+      "loss": 0.1866,
+      "step": 935
+    },
+    {
+      "epoch": 1.4579439252336448,
+      "grad_norm": 0.6319275863012598,
+      "learning_rate": 1.7055709710194452e-06,
+      "loss": 0.1921,
+      "step": 936
+    },
+    {
+      "epoch": 1.4595015576323989,
+      "grad_norm": 0.6620029301577026,
+      "learning_rate": 1.6963781892747128e-06,
+      "loss": 0.1889,
+      "step": 937
+    },
+    {
+      "epoch": 1.4610591900311527,
+      "grad_norm": 0.6613697258053355,
+      "learning_rate": 1.6872051845023124e-06,
+      "loss": 0.2277,
+      "step": 938
+    },
+    {
+      "epoch": 1.4626168224299065,
+      "grad_norm": 0.6709827699093153,
+      "learning_rate": 1.6780520116159954e-06,
+      "loss": 0.2345,
+      "step": 939
+    },
+    {
+      "epoch": 1.4641744548286604,
+      "grad_norm": 0.6308994122097046,
+      "learning_rate": 1.6689187254107852e-06,
+      "loss": 0.2384,
+      "step": 940
+    },
+    {
+      "epoch": 1.4657320872274142,
+      "grad_norm": 0.49997735693518447,
+      "learning_rate": 1.6598053805626607e-06,
+      "loss": 0.2885,
+      "step": 941
+    },
+    {
+      "epoch": 1.4672897196261683,
+      "grad_norm": 0.5769855416317906,
+      "learning_rate": 1.6507120316282222e-06,
+      "loss": 0.2262,
+      "step": 942
+    },
+    {
+      "epoch": 1.4688473520249221,
+      "grad_norm": 0.7766043611879518,
+      "learning_rate": 1.6416387330443594e-06,
+      "loss": 0.2012,
+      "step": 943
+    },
+    {
+      "epoch": 1.470404984423676,
+      "grad_norm": 0.6306011227357949,
+      "learning_rate": 1.6325855391279399e-06,
+      "loss": 0.1672,
+      "step": 944
+    },
+    {
+      "epoch": 1.47196261682243,
+      "grad_norm": 0.6252880509484167,
+      "learning_rate": 1.623552504075467e-06,
+      "loss": 0.21,
+      "step": 945
+    },
+    {
+      "epoch": 1.4735202492211839,
+      "grad_norm": 0.5733165288371812,
+      "learning_rate": 1.6145396819627734e-06,
+      "loss": 0.203,
+      "step": 946
+    },
+    {
+      "epoch": 1.4750778816199377,
+      "grad_norm": 0.7571650504143702,
+      "learning_rate": 1.6055471267446804e-06,
+      "loss": 0.2244,
+      "step": 947
+    },
+    {
+      "epoch": 1.4766355140186915,
+      "grad_norm": 0.7478203882298783,
+      "learning_rate": 1.5965748922546876e-06,
+      "loss": 0.1799,
+      "step": 948
+    },
+    {
+      "epoch": 1.4781931464174454,
+      "grad_norm": 0.6868312970173284,
+      "learning_rate": 1.5876230322046393e-06,
+      "loss": 0.6746,
+      "step": 949
+    },
+    {
+      "epoch": 1.4797507788161994,
+      "grad_norm": 0.6037632401087822,
+      "learning_rate": 1.578691600184416e-06,
+      "loss": 0.2057,
+      "step": 950
+    },
+    {
+      "epoch": 1.4813084112149533,
+      "grad_norm": 0.6046826777481786,
+      "learning_rate": 1.569780649661603e-06,
+      "loss": 0.2926,
+      "step": 951
+    },
+    {
+      "epoch": 1.482866043613707,
+      "grad_norm": 0.6848611609536812,
+      "learning_rate": 1.5608902339811765e-06,
+      "loss": 0.2328,
+      "step": 952
+    },
+    {
+      "epoch": 1.4844236760124612,
+      "grad_norm": 0.5462582273781476,
+      "learning_rate": 1.5520204063651784e-06,
+      "loss": 0.2703,
+      "step": 953
+    },
+    {
+      "epoch": 1.485981308411215,
+      "grad_norm": 0.5459862014182448,
+      "learning_rate": 1.5431712199124033e-06,
+      "loss": 0.2268,
+      "step": 954
+    },
+    {
+      "epoch": 1.4875389408099688,
+      "grad_norm": 0.5909228089011523,
+      "learning_rate": 1.5343427275980804e-06,
+      "loss": 0.1918,
+      "step": 955
+    },
+    {
+      "epoch": 1.4890965732087227,
+      "grad_norm": 0.6922392218824316,
+      "learning_rate": 1.5255349822735494e-06,
+      "loss": 0.2065,
+      "step": 956
+    },
+    {
+      "epoch": 1.4906542056074765,
+      "grad_norm": 0.6347770175659823,
+      "learning_rate": 1.5167480366659538e-06,
+      "loss": 0.1683,
+      "step": 957
+    },
+    {
+      "epoch": 1.4922118380062306,
+      "grad_norm": 0.5516070303093964,
+      "learning_rate": 1.507981943377918e-06,
+      "loss": 0.2173,
+      "step": 958
+    },
+    {
+      "epoch": 1.4937694704049844,
+      "grad_norm": 0.6071049706243482,
+      "learning_rate": 1.4992367548872383e-06,
+      "loss": 0.2458,
+      "step": 959
+    },
+    {
+      "epoch": 1.4953271028037383,
+      "grad_norm": 0.6720642183082266,
+      "learning_rate": 1.490512523546559e-06,
+      "loss": 0.1915,
+      "step": 960
+    },
+    {
+      "epoch": 1.4968847352024923,
+      "grad_norm": 0.5147991401453758,
+      "learning_rate": 1.481809301583072e-06,
+      "loss": 0.2315,
+      "step": 961
+    },
+    {
+      "epoch": 1.4984423676012462,
+      "grad_norm": 0.5193622138782028,
+      "learning_rate": 1.4731271410981961e-06,
+      "loss": 0.2994,
+      "step": 962
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6483302843412619,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.1486,
+      "step": 963
+    },
+    {
+      "epoch": 1.5015576323987538,
+      "grad_norm": 0.6211856895290482,
+      "learning_rate": 1.4558262123392125e-06,
+      "loss": 0.1979,
+      "step": 964
+    },
+    {
+      "epoch": 1.5031152647975077,
+      "grad_norm": 0.6267158569891346,
+      "learning_rate": 1.4472075476362819e-06,
+      "loss": 0.226,
+      "step": 965
+    },
+    {
+      "epoch": 1.5046728971962615,
+      "grad_norm": 0.6878081812868871,
+      "learning_rate": 1.4386101515536865e-06,
+      "loss": 0.215,
+      "step": 966
+    },
+    {
+      "epoch": 1.5062305295950156,
+      "grad_norm": 0.6013152247422209,
+      "learning_rate": 1.4300340755593256e-06,
+      "loss": 0.1853,
+      "step": 967
+    },
+    {
+      "epoch": 1.5077881619937694,
+      "grad_norm": 0.5961333640534994,
+      "learning_rate": 1.4214793709934644e-06,
+      "loss": 0.2134,
+      "step": 968
+    },
+    {
+      "epoch": 1.5093457943925235,
+      "grad_norm": 0.6714185068729337,
+      "learning_rate": 1.4129460890684255e-06,
+      "loss": 0.2285,
+      "step": 969
+    },
+    {
+      "epoch": 1.5109034267912773,
+      "grad_norm": 0.5287976544223729,
+      "learning_rate": 1.4044342808682904e-06,
+      "loss": 0.226,
+      "step": 970
+    },
+    {
+      "epoch": 1.5124610591900312,
+      "grad_norm": 0.7244263964282488,
+      "learning_rate": 1.3959439973485872e-06,
+      "loss": 0.1919,
+      "step": 971
+    },
+    {
+      "epoch": 1.514018691588785,
+      "grad_norm": 0.7411473501185608,
+      "learning_rate": 1.38747528933599e-06,
+      "loss": 0.1573,
+      "step": 972
+    },
+    {
+      "epoch": 1.5155763239875388,
+      "grad_norm": 0.7535346088046625,
+      "learning_rate": 1.3790282075280064e-06,
+      "loss": 0.1939,
+      "step": 973
+    },
+    {
+      "epoch": 1.5171339563862927,
+      "grad_norm": 0.7535256340431424,
+      "learning_rate": 1.3706028024926855e-06,
+      "loss": 0.1736,
+      "step": 974
+    },
+    {
+      "epoch": 1.5186915887850467,
+      "grad_norm": 0.5709920085521653,
+      "learning_rate": 1.3621991246683093e-06,
+      "loss": 0.222,
+      "step": 975
+    },
+    {
+      "epoch": 1.5202492211838006,
+      "grad_norm": 0.7297081050093409,
+      "learning_rate": 1.3538172243630848e-06,
+      "loss": 0.1788,
+      "step": 976
+    },
+    {
+      "epoch": 1.5218068535825546,
+      "grad_norm": 0.6180657113624616,
+      "learning_rate": 1.3454571517548554e-06,
+      "loss": 0.177,
+      "step": 977
+    },
+    {
+      "epoch": 1.5233644859813085,
+      "grad_norm": 0.591364040153287,
+      "learning_rate": 1.3371189568907915e-06,
+      "loss": 0.2128,
+      "step": 978
+    },
+    {
+      "epoch": 1.5249221183800623,
+      "grad_norm": 0.5022261849806481,
+      "learning_rate": 1.3288026896870964e-06,
+      "loss": 0.2267,
+      "step": 979
+    },
+    {
+      "epoch": 1.5264797507788161,
+      "grad_norm": 0.6203532747393247,
+      "learning_rate": 1.3205083999286973e-06,
+      "loss": 0.1955,
+      "step": 980
+    },
+    {
+      "epoch": 1.52803738317757,
+      "grad_norm": 0.6729761565050358,
+      "learning_rate": 1.3122361372689624e-06,
+      "loss": 0.1657,
+      "step": 981
+    },
+    {
+      "epoch": 1.5295950155763238,
+      "grad_norm": 0.6536059366129146,
+      "learning_rate": 1.3039859512293885e-06,
+      "loss": 0.2198,
+      "step": 982
+    },
+    {
+      "epoch": 1.5311526479750779,
+      "grad_norm": 0.695539877357984,
+      "learning_rate": 1.2957578911993212e-06,
+      "loss": 0.2023,
+      "step": 983
+    },
+    {
+      "epoch": 1.5327102803738317,
+      "grad_norm": 0.4424609062997814,
+      "learning_rate": 1.28755200643564e-06,
+      "loss": 0.2238,
+      "step": 984
+    },
+    {
+      "epoch": 1.5342679127725858,
+      "grad_norm": 0.5843664425742057,
+      "learning_rate": 1.27936834606248e-06,
+      "loss": 0.2197,
+      "step": 985
+    },
+    {
+      "epoch": 1.5358255451713396,
+      "grad_norm": 0.6148367721669108,
+      "learning_rate": 1.2712069590709265e-06,
+      "loss": 0.2004,
+      "step": 986
+    },
+    {
+      "epoch": 1.5373831775700935,
+      "grad_norm": 0.562472309572901,
+      "learning_rate": 1.2630678943187292e-06,
+      "loss": 0.1835,
+      "step": 987
+    },
+    {
+      "epoch": 1.5389408099688473,
+      "grad_norm": 0.7987485306921659,
+      "learning_rate": 1.2549512005300068e-06,
+      "loss": 0.1995,
+      "step": 988
+    },
+    {
+      "epoch": 1.5404984423676011,
+      "grad_norm": 0.6656097374020112,
+      "learning_rate": 1.246856926294953e-06,
+      "loss": 0.1874,
+      "step": 989
+    },
+    {
+      "epoch": 1.542056074766355,
+      "grad_norm": 0.7982996720144556,
+      "learning_rate": 1.23878512006955e-06,
+      "loss": 0.351,
+      "step": 990
+    },
+    {
+      "epoch": 1.543613707165109,
+      "grad_norm": 0.7237322332450993,
+      "learning_rate": 1.2307358301752753e-06,
+      "loss": 0.6736,
+      "step": 991
+    },
+    {
+      "epoch": 1.5451713395638629,
+      "grad_norm": 0.6606058936315011,
+      "learning_rate": 1.222709104798816e-06,
+      "loss": 0.2446,
+      "step": 992
+    },
+    {
+      "epoch": 1.546728971962617,
+      "grad_norm": 0.7491652827965601,
+      "learning_rate": 1.214704991991773e-06,
+      "loss": 0.1912,
+      "step": 993
+    },
+    {
+      "epoch": 1.5482866043613708,
+      "grad_norm": 0.5316071822548012,
+      "learning_rate": 1.2067235396703819e-06,
+      "loss": 0.2006,
+      "step": 994
+    },
+    {
+      "epoch": 1.5498442367601246,
+      "grad_norm": 0.4870862919772457,
+      "learning_rate": 1.198764795615221e-06,
+      "loss": 0.1721,
+      "step": 995
+    },
+    {
+      "epoch": 1.5514018691588785,
+      "grad_norm": 0.45786723638909316,
+      "learning_rate": 1.190828807470929e-06,
+      "loss": 0.2847,
+      "step": 996
+    },
+    {
+      "epoch": 1.5529595015576323,
+      "grad_norm": 0.5220478168315765,
+      "learning_rate": 1.182915622745912e-06,
+      "loss": 0.1942,
+      "step": 997
+    },
+    {
+      "epoch": 1.5545171339563861,
+      "grad_norm": 1.0355443091689227,
+      "learning_rate": 1.17502528881207e-06,
+      "loss": 0.6648,
+      "step": 998
+    },
+    {
+      "epoch": 1.5560747663551402,
+      "grad_norm": 0.6645917404798355,
+      "learning_rate": 1.1671578529045075e-06,
+      "loss": 0.1686,
+      "step": 999
+    },
+    {
+      "epoch": 1.557632398753894,
+      "grad_norm": 0.5772980090930082,
+      "learning_rate": 1.1593133621212454e-06,
+      "loss": 0.1601,
+      "step": 1000
+    },
+    {
+      "epoch": 1.557632398753894,
+      "eval_loss": 0.36974087357521057,
+      "eval_runtime": 2.7835,
+      "eval_samples_per_second": 9.341,
+      "eval_steps_per_second": 2.515,
+      "step": 1000
+    },
+    {
+      "epoch": 1.559190031152648,
+      "grad_norm": 0.6634169549696792,
+      "learning_rate": 1.151491863422951e-06,
+      "loss": 0.178,
+      "step": 1001
+    },
+    {
+      "epoch": 1.560747663551402,
+      "grad_norm": 0.6293088334490837,
+      "learning_rate": 1.1436934036326485e-06,
+      "loss": 0.2054,
+      "step": 1002
+    },
+    {
+      "epoch": 1.5623052959501558,
+      "grad_norm": 0.6633410484590091,
+      "learning_rate": 1.135918029435441e-06,
+      "loss": 0.2654,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5638629283489096,
+      "grad_norm": 0.6622964143754947,
+      "learning_rate": 1.1281657873782287e-06,
+      "loss": 0.2793,
+      "step": 1004
+    },
+    {
+      "epoch": 1.5654205607476634,
+      "grad_norm": 0.6119616717415093,
+      "learning_rate": 1.1204367238694357e-06,
+      "loss": 0.2711,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5669781931464173,
+      "grad_norm": 0.47949709495438925,
+      "learning_rate": 1.1127308851787244e-06,
+      "loss": 0.2021,
+      "step": 1006
+    },
+    {
+      "epoch": 1.5685358255451713,
+      "grad_norm": 0.6934647914219084,
+      "learning_rate": 1.1050483174367304e-06,
+      "loss": 0.203,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5700934579439252,
+      "grad_norm": 0.5481726220781126,
+      "learning_rate": 1.0973890666347703e-06,
+      "loss": 0.2368,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5716510903426792,
+      "grad_norm": 0.6509857110936647,
+      "learning_rate": 1.0897531786245819e-06,
+      "loss": 0.2125,
+      "step": 1009
+    },
+    {
+      "epoch": 1.573208722741433,
+      "grad_norm": 0.5617926695948265,
+      "learning_rate": 1.0821406991180367e-06,
+      "loss": 0.2398,
+      "step": 1010
+    },
+    {
+      "epoch": 1.574766355140187,
+      "grad_norm": 0.6111199895380683,
+      "learning_rate": 1.0745516736868766e-06,
+      "loss": 0.2485,
+      "step": 1011
+    },
+    {
+      "epoch": 1.5763239875389408,
+      "grad_norm": 0.7124396857915882,
+      "learning_rate": 1.0669861477624376e-06,
+      "loss": 0.2517,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5778816199376946,
+      "grad_norm": 0.5990035454998449,
+      "learning_rate": 1.0594441666353722e-06,
+      "loss": 0.151,
+      "step": 1013
+    },
+    {
+      "epoch": 1.5794392523364484,
+      "grad_norm": 0.6520260923864126,
+      "learning_rate": 1.0519257754553868e-06,
+      "loss": 0.1888,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5809968847352025,
+      "grad_norm": 0.7169934380455352,
+      "learning_rate": 1.0444310192309675e-06,
+      "loss": 0.2318,
+      "step": 1015
+    },
+    {
+      "epoch": 1.5825545171339563,
+      "grad_norm": 0.5993914903355098,
+      "learning_rate": 1.0369599428291116e-06,
+      "loss": 0.2124,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5841121495327104,
+      "grad_norm": 0.49513044936794753,
+      "learning_rate": 1.0295125909750537e-06,
+      "loss": 0.2402,
+      "step": 1017
+    },
+    {
+      "epoch": 1.5856697819314642,
+      "grad_norm": 0.7226758243950167,
+      "learning_rate": 1.0220890082520074e-06,
+      "loss": 0.2398,
+      "step": 1018
+    },
+    {
+      "epoch": 1.587227414330218,
+      "grad_norm": 0.6288423009834784,
+      "learning_rate": 1.0146892391008929e-06,
+      "loss": 0.1948,
+      "step": 1019
+    },
+    {
+      "epoch": 1.588785046728972,
+      "grad_norm": 0.596604250578145,
+      "learning_rate": 1.0073133278200702e-06,
+      "loss": 0.232,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5903426791277258,
+      "grad_norm": 0.6135147035860976,
+      "learning_rate": 9.999613185650748e-07,
+      "loss": 0.2614,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5919003115264796,
+      "grad_norm": 0.5404711008426897,
+      "learning_rate": 9.926332553483547e-07,
+      "loss": 0.1885,
+      "step": 1022
+    },
+    {
+      "epoch": 1.5934579439252337,
+      "grad_norm": 0.7031398847787177,
+      "learning_rate": 9.853291820390087e-07,
+      "loss": 0.2325,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5950155763239875,
+      "grad_norm": 0.6348294111355727,
+      "learning_rate": 9.780491423625154e-07,
+      "loss": 0.2385,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5965732087227416,
+      "grad_norm": 0.5521768442130486,
+      "learning_rate": 9.70793179900484e-07,
+      "loss": 0.1988,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5981308411214954,
+      "grad_norm": 0.613980141544231,
+      "learning_rate": 9.63561338090378e-07,
+      "loss": 0.1027,
+      "step": 1026
+    },
+    {
+      "epoch": 1.5996884735202492,
+      "grad_norm": 0.7720687456198224,
+      "learning_rate": 9.563536602252749e-07,
+      "loss": 0.2905,
+      "step": 1027
+    },
+    {
+      "epoch": 1.601246105919003,
+      "grad_norm": 0.6319851908391722,
+      "learning_rate": 9.49170189453586e-07,
+      "loss": 0.2617,
+      "step": 1028
+    },
+    {
+      "epoch": 1.602803738317757,
+      "grad_norm": 0.6736383488893641,
+      "learning_rate": 9.420109687788148e-07,
+      "loss": 0.2641,
+      "step": 1029
+    },
+    {
+      "epoch": 1.6043613707165107,
+      "grad_norm": 0.6656334468054872,
+      "learning_rate": 9.348760410592855e-07,
+      "loss": 0.2171,
+      "step": 1030
+    },
+    {
+      "epoch": 1.6059190031152648,
+      "grad_norm": 0.6957076685070199,
+      "learning_rate": 9.277654490079035e-07,
+      "loss": 0.236,
+      "step": 1031
+    },
+    {
+      "epoch": 1.6074766355140186,
+      "grad_norm": 0.6288117317451123,
+      "learning_rate": 9.206792351918809e-07,
+      "loss": 0.1562,
+      "step": 1032
+    },
+    {
+      "epoch": 1.6090342679127727,
+      "grad_norm": 0.6616920441316214,
+      "learning_rate": 9.136174420324962e-07,
+      "loss": 0.218,
+      "step": 1033
+    },
+    {
+      "epoch": 1.6105919003115265,
+      "grad_norm": 0.6342551825259773,
+      "learning_rate": 9.065801118048312e-07,
+      "loss": 0.2252,
+      "step": 1034
+    },
+    {
+      "epoch": 1.6121495327102804,
+      "grad_norm": 0.5589154301426593,
+      "learning_rate": 8.995672866375237e-07,
+      "loss": 0.1551,
+      "step": 1035
+    },
+    {
+      "epoch": 1.6137071651090342,
+      "grad_norm": 0.4635682927354973,
+      "learning_rate": 8.925790085125152e-07,
+      "loss": 0.1833,
+      "step": 1036
+    },
+    {
+      "epoch": 1.615264797507788,
+      "grad_norm": 0.5543844541586774,
+      "learning_rate": 8.856153192647909e-07,
+      "loss": 0.2067,
+      "step": 1037
+    },
+    {
+      "epoch": 1.616822429906542,
+      "grad_norm": 0.6110967305529611,
+      "learning_rate": 8.786762605821419e-07,
+      "loss": 0.2331,
+      "step": 1038
+    },
+    {
+      "epoch": 1.618380062305296,
+      "grad_norm": 0.6930362377044883,
+      "learning_rate": 8.717618740049083e-07,
+      "loss": 0.2451,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6199376947040498,
+      "grad_norm": 0.8137005439911986,
+      "learning_rate": 8.648722009257315e-07,
+      "loss": 0.1626,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6214953271028039,
+      "grad_norm": 0.7004708747608873,
+      "learning_rate": 8.580072825893049e-07,
+      "loss": 0.1831,
+      "step": 1041
+    },
+    {
+      "epoch": 1.6230529595015577,
+      "grad_norm": 0.607273962654228,
+      "learning_rate": 8.511671600921306e-07,
+      "loss": 0.161,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6246105919003115,
+      "grad_norm": 0.5132546282344246,
+      "learning_rate": 8.443518743822726e-07,
+      "loss": 0.2764,
+      "step": 1043
+    },
+    {
+      "epoch": 1.6261682242990654,
+      "grad_norm": 0.6893277741210059,
+      "learning_rate": 8.375614662591097e-07,
+      "loss": 0.1793,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6277258566978192,
+      "grad_norm": 0.46872986327742755,
+      "learning_rate": 8.307959763730899e-07,
+      "loss": 0.1761,
+      "step": 1045
+    },
+    {
+      "epoch": 1.629283489096573,
+      "grad_norm": 0.5290839127332908,
+      "learning_rate": 8.240554452254929e-07,
+      "loss": 0.2329,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6308411214953271,
+      "grad_norm": 0.6109039006881162,
+      "learning_rate": 8.173399131681831e-07,
+      "loss": 0.2121,
+      "step": 1047
+    },
+    {
+      "epoch": 1.632398753894081,
+      "grad_norm": 0.6091367601010639,
+      "learning_rate": 8.106494204033677e-07,
+      "loss": 0.1873,
+      "step": 1048
+    },
+    {
+      "epoch": 1.633956386292835,
+      "grad_norm": 0.5867126273845332,
+      "learning_rate": 8.039840069833594e-07,
+      "loss": 0.7795,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6355140186915889,
+      "grad_norm": 0.5685700476106482,
+      "learning_rate": 7.973437128103306e-07,
+      "loss": 0.2459,
+      "step": 1050
+    },
+    {
+      "epoch": 1.6370716510903427,
+      "grad_norm": 0.5135498312661578,
+      "learning_rate": 7.907285776360863e-07,
+      "loss": 0.255,
+      "step": 1051
+    },
+    {
+      "epoch": 1.6386292834890965,
+      "grad_norm": 0.6667857649972504,
+      "learning_rate": 7.841386410618096e-07,
+      "loss": 0.2476,
+      "step": 1052
+    },
+    {
+      "epoch": 1.6401869158878504,
+      "grad_norm": 0.5960262952580946,
+      "learning_rate": 7.775739425378398e-07,
+      "loss": 0.2462,
+      "step": 1053
+    },
+    {
+      "epoch": 1.6417445482866042,
+      "grad_norm": 0.8148918825910239,
+      "learning_rate": 7.710345213634247e-07,
+      "loss": 0.2342,
+      "step": 1054
+    },
+    {
+      "epoch": 1.6433021806853583,
+      "grad_norm": 0.6117412041332031,
+      "learning_rate": 7.645204166864967e-07,
+      "loss": 0.1889,
+      "step": 1055
+    },
+    {
+      "epoch": 1.644859813084112,
+      "grad_norm": 0.6194608199292435,
+      "learning_rate": 7.580316675034255e-07,
+      "loss": 0.1807,
+      "step": 1056
+    },
+    {
+      "epoch": 1.6464174454828662,
+      "grad_norm": 0.9323712553721532,
+      "learning_rate": 7.515683126587975e-07,
+      "loss": 0.1879,
+      "step": 1057
+    },
+    {
+      "epoch": 1.64797507788162,
+      "grad_norm": 0.5493635025344051,
+      "learning_rate": 7.451303908451707e-07,
+      "loss": 0.2106,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6495327102803738,
+      "grad_norm": 0.7158277912716375,
+      "learning_rate": 7.387179406028549e-07,
+      "loss": 0.2354,
+      "step": 1059
+    },
+    {
+      "epoch": 1.6510903426791277,
+      "grad_norm": 0.5571552725370883,
+      "learning_rate": 7.323310003196749e-07,
+      "loss": 0.2145,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6526479750778815,
+      "grad_norm": 0.7538580273599934,
+      "learning_rate": 7.259696082307372e-07,
+      "loss": 0.2482,
+      "step": 1061
+    },
+    {
+      "epoch": 1.6542056074766354,
+      "grad_norm": 0.5327326528906995,
+      "learning_rate": 7.196338024182098e-07,
+      "loss": 0.2354,
+      "step": 1062
+    },
+    {
+      "epoch": 1.6557632398753894,
+      "grad_norm": 0.557287992992955,
+      "learning_rate": 7.133236208110878e-07,
+      "loss": 0.1693,
+      "step": 1063
+    },
+    {
+      "epoch": 1.6573208722741433,
+      "grad_norm": 0.5104215870170742,
+      "learning_rate": 7.070391011849698e-07,
+      "loss": 0.1717,
+      "step": 1064
+    },
+    {
+      "epoch": 1.6588785046728973,
+      "grad_norm": 0.5917806139945608,
+      "learning_rate": 7.007802811618258e-07,
+      "loss": 0.2085,
+      "step": 1065
+    },
+    {
+      "epoch": 1.6604361370716512,
+      "grad_norm": 0.6306946462467933,
+      "learning_rate": 6.945471982097818e-07,
+      "loss": 0.2145,
+      "step": 1066
+    },
+    {
+      "epoch": 1.661993769470405,
+      "grad_norm": 0.7196660102146798,
+      "learning_rate": 6.883398896428867e-07,
+      "loss": 0.2321,
+      "step": 1067
+    },
+    {
+      "epoch": 1.6635514018691588,
+      "grad_norm": 0.532187998071398,
+      "learning_rate": 6.821583926208947e-07,
+      "loss": 0.2366,
+      "step": 1068
+    },
+    {
+      "epoch": 1.6651090342679127,
+      "grad_norm": 0.5827926596867945,
+      "learning_rate": 6.760027441490369e-07,
+      "loss": 0.285,
+      "step": 1069
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.7057955872693081,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.6118,
+      "step": 1070
+    },
+    {
+      "epoch": 1.6682242990654206,
+      "grad_norm": 0.7121732240730015,
+      "learning_rate": 6.637691401027352e-07,
+      "loss": 0.8544,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6697819314641744,
+      "grad_norm": 0.49184886655898746,
+      "learning_rate": 6.576912577641697e-07,
+      "loss": 0.2325,
+      "step": 1072
+    },
+    {
+      "epoch": 1.6713395638629285,
+      "grad_norm": 0.6760569677497837,
+      "learning_rate": 6.516393704470608e-07,
+      "loss": 0.8096,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6728971962616823,
+      "grad_norm": 0.7478627429100672,
+      "learning_rate": 6.456135143807369e-07,
+      "loss": 0.1594,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6744548286604362,
+      "grad_norm": 0.5379337402997096,
+      "learning_rate": 6.396137256386975e-07,
+      "loss": 0.1994,
+      "step": 1075
+    },
+    {
+      "epoch": 1.67601246105919,
+      "grad_norm": 0.5032927663415819,
+      "learning_rate": 6.336400401383857e-07,
+      "loss": 0.2132,
+      "step": 1076
+    },
+    {
+      "epoch": 1.6775700934579438,
+      "grad_norm": 0.47582414487022096,
+      "learning_rate": 6.276924936409829e-07,
+      "loss": 0.2478,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6791277258566977,
+      "grad_norm": 0.6434383172973982,
+      "learning_rate": 6.217711217511857e-07,
+      "loss": 0.2143,
+      "step": 1078
+    },
+    {
+      "epoch": 1.6806853582554517,
+      "grad_norm": 0.6604513508530203,
+      "learning_rate": 6.158759599170045e-07,
+      "loss": 0.19,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6822429906542056,
+      "grad_norm": 0.5696994691557421,
+      "learning_rate": 6.100070434295379e-07,
+      "loss": 0.2234,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6838006230529596,
+      "grad_norm": 0.668728026470969,
+      "learning_rate": 6.041644074227738e-07,
+      "loss": 0.1627,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6853582554517135,
+      "grad_norm": 0.5959783592005928,
+      "learning_rate": 5.983480868733666e-07,
+      "loss": 0.2239,
+      "step": 1082
+    },
+    {
+      "epoch": 1.6869158878504673,
+      "grad_norm": 0.7134998370022319,
+      "learning_rate": 5.925581166004407e-07,
+      "loss": 0.1782,
+      "step": 1083
+    },
+    {
+      "epoch": 1.6884735202492211,
+      "grad_norm": 0.5805335392567044,
+      "learning_rate": 5.867945312653733e-07,
+      "loss": 0.1725,
+      "step": 1084
+    },
+    {
+      "epoch": 1.690031152647975,
+      "grad_norm": 0.938179646578548,
+      "learning_rate": 5.810573653715884e-07,
+      "loss": 0.6935,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6915887850467288,
+      "grad_norm": 0.6255817452393061,
+      "learning_rate": 5.753466532643531e-07,
+      "loss": 0.2057,
+      "step": 1086
+    },
+    {
+      "epoch": 1.6931464174454829,
+      "grad_norm": 0.6164273521801443,
+      "learning_rate": 5.696624291305692e-07,
+      "loss": 0.2354,
+      "step": 1087
+    },
+    {
+      "epoch": 1.6947040498442367,
+      "grad_norm": 0.6024358987476546,
+      "learning_rate": 5.64004726998571e-07,
+      "loss": 0.1628,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6962616822429908,
+      "grad_norm": 0.583734062156969,
+      "learning_rate": 5.583735807379165e-07,
+      "loss": 0.258,
+      "step": 1089
+    },
+    {
+      "epoch": 1.6978193146417446,
+      "grad_norm": 0.5381732861463082,
+      "learning_rate": 5.527690240591927e-07,
+      "loss": 0.259,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6993769470404985,
+      "grad_norm": 0.6432738953861825,
+      "learning_rate": 5.471910905138062e-07,
+      "loss": 0.2233,
+      "step": 1091
+    },
+    {
+      "epoch": 1.7009345794392523,
+      "grad_norm": 0.7405741920054726,
+      "learning_rate": 5.416398134937878e-07,
+      "loss": 0.2732,
+      "step": 1092
+    },
+    {
+      "epoch": 1.7024922118380061,
+      "grad_norm": 0.5681851173547261,
+      "learning_rate": 5.361152262315883e-07,
+      "loss": 0.1976,
+      "step": 1093
+    },
+    {
+      "epoch": 1.70404984423676,
+      "grad_norm": 0.5473658208153401,
+      "learning_rate": 5.306173617998823e-07,
+      "loss": 0.2139,
+      "step": 1094
+    },
+    {
+      "epoch": 1.705607476635514,
+      "grad_norm": 0.5255770004268006,
+      "learning_rate": 5.251462531113705e-07,
+      "loss": 0.1707,
+      "step": 1095
+    },
+    {
+      "epoch": 1.7071651090342679,
+      "grad_norm": 0.5374196236910275,
+      "learning_rate": 5.197019329185787e-07,
+      "loss": 0.2281,
+      "step": 1096
+    },
+    {
+      "epoch": 1.708722741433022,
+      "grad_norm": 0.5260272455503919,
+      "learning_rate": 5.142844338136693e-07,
+      "loss": 0.2471,
+      "step": 1097
+    },
+    {
+      "epoch": 1.7102803738317758,
+      "grad_norm": 0.5646648215018458,
+      "learning_rate": 5.088937882282358e-07,
+      "loss": 0.1951,
+      "step": 1098
+    },
+    {
+      "epoch": 1.7118380062305296,
+      "grad_norm": 0.5458034326374251,
+      "learning_rate": 5.035300284331213e-07,
+      "loss": 0.2432,
+      "step": 1099
+    },
+    {
+      "epoch": 1.7133956386292835,
+      "grad_norm": 0.6037542349280097,
+      "learning_rate": 4.981931865382111e-07,
+      "loss": 0.2618,
+      "step": 1100
+    },
+    {
+      "epoch": 1.7149532710280373,
+      "grad_norm": 0.5948181143891571,
+      "learning_rate": 4.92883294492254e-07,
+      "loss": 0.1635,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7165109034267911,
+      "grad_norm": 0.5776080511589572,
+      "learning_rate": 4.876003840826588e-07,
+      "loss": 0.2451,
+      "step": 1102
+    },
+    {
+      "epoch": 1.7180685358255452,
+      "grad_norm": 0.5026789603534472,
+      "learning_rate": 4.823444869353177e-07,
+      "loss": 0.1562,
+      "step": 1103
+    },
+    {
+      "epoch": 1.719626168224299,
+      "grad_norm": 0.5367686440727194,
+      "learning_rate": 4.771156345144018e-07,
+      "loss": 0.2041,
+      "step": 1104
+    },
+    {
+      "epoch": 1.721183800623053,
+      "grad_norm": 0.658875304880407,
+      "learning_rate": 4.7191385812218547e-07,
+      "loss": 0.7385,
+      "step": 1105
+    },
+    {
+      "epoch": 1.722741433021807,
+      "grad_norm": 0.6189314088368518,
+      "learning_rate": 4.6673918889884985e-07,
+      "loss": 0.269,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7242990654205608,
+      "grad_norm": 0.7000827321303451,
+      "learning_rate": 4.615916578223029e-07,
+      "loss": 0.219,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7258566978193146,
+      "grad_norm": 0.6218605150183268,
+      "learning_rate": 4.5647129570799196e-07,
+      "loss": 0.2765,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7274143302180685,
+      "grad_norm": 0.7349959501432963,
+      "learning_rate": 4.513781332087153e-07,
+      "loss": 0.2457,
+      "step": 1109
+    },
+    {
+      "epoch": 1.7289719626168223,
+      "grad_norm": 0.7008089188891375,
+      "learning_rate": 4.463122008144449e-07,
+      "loss": 0.2083,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7305295950155763,
+      "grad_norm": 0.5607384923036217,
+      "learning_rate": 4.4127352885214026e-07,
+      "loss": 0.2166,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7320872274143302,
+      "grad_norm": 0.645885067353582,
+      "learning_rate": 4.362621474855688e-07,
+      "loss": 0.2312,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7336448598130842,
+      "grad_norm": 1.0208563939368382,
+      "learning_rate": 4.3127808671512114e-07,
+      "loss": 0.8117,
+      "step": 1113
+    },
+    {
+      "epoch": 1.735202492211838,
+      "grad_norm": 0.6252228702704483,
+      "learning_rate": 4.263213763776369e-07,
+      "loss": 0.2111,
+      "step": 1114
+    },
+    {
+      "epoch": 1.736760124610592,
+      "grad_norm": 0.5295697965398305,
+      "learning_rate": 4.213920461462234e-07,
+      "loss": 0.1693,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7383177570093458,
+      "grad_norm": 0.5321708104559851,
+      "learning_rate": 4.1649012553007795e-07,
+      "loss": 0.1598,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7398753894080996,
+      "grad_norm": 0.6297092340917954,
+      "learning_rate": 4.116156438743102e-07,
+      "loss": 0.6563,
+      "step": 1117
+    },
+    {
+      "epoch": 1.7414330218068534,
+      "grad_norm": 0.6502646751207796,
+      "learning_rate": 4.067686303597701e-07,
+      "loss": 0.828,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7429906542056075,
+      "grad_norm": 0.6299291575701418,
+      "learning_rate": 4.0194911400287084e-07,
+      "loss": 0.1618,
+      "step": 1119
+    },
+    {
+      "epoch": 1.7445482866043613,
+      "grad_norm": 0.6763449045114679,
+      "learning_rate": 3.971571236554117e-07,
+      "loss": 0.1755,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7461059190031154,
+      "grad_norm": 0.6097387148482865,
+      "learning_rate": 3.923926880044132e-07,
+      "loss": 0.2144,
+      "step": 1121
+    },
+    {
+      "epoch": 1.7476635514018692,
+      "grad_norm": 0.6555185956249271,
+      "learning_rate": 3.876558355719373e-07,
+      "loss": 0.7322,
+      "step": 1122
+    },
+    {
+      "epoch": 1.749221183800623,
+      "grad_norm": 0.529186428154948,
+      "learning_rate": 3.8294659471492466e-07,
+      "loss": 0.2299,
+      "step": 1123
+    },
+    {
+      "epoch": 1.750778816199377,
+      "grad_norm": 0.5210926587424783,
+      "learning_rate": 3.7826499362501657e-07,
+      "loss": 0.7365,
+      "step": 1124
+    },
+    {
+      "epoch": 1.7523364485981308,
+      "grad_norm": 0.5705242040491778,
+      "learning_rate": 3.7361106032839266e-07,
+      "loss": 0.1769,
+      "step": 1125
+    },
+    {
+      "epoch": 1.7538940809968846,
+      "grad_norm": 0.6221568867702569,
+      "learning_rate": 3.6898482268559745e-07,
+      "loss": 0.2641,
+      "step": 1126
+    },
+    {
+      "epoch": 1.7554517133956387,
+      "grad_norm": 0.5761942177039651,
+      "learning_rate": 3.643863083913829e-07,
+      "loss": 0.1756,
+      "step": 1127
+    },
+    {
+      "epoch": 1.7570093457943925,
+      "grad_norm": 0.5598040118881634,
+      "learning_rate": 3.5981554497452886e-07,
+      "loss": 0.1968,
+      "step": 1128
+    },
+    {
+      "epoch": 1.7585669781931466,
+      "grad_norm": 0.6115168676953736,
+      "learning_rate": 3.552725597976925e-07,
+      "loss": 0.1656,
+      "step": 1129
+    },
+    {
+      "epoch": 1.7601246105919004,
+      "grad_norm": 0.5618155080059914,
+      "learning_rate": 3.507573800572328e-07,
+      "loss": 0.2284,
+      "step": 1130
+    },
+    {
+      "epoch": 1.7616822429906542,
+      "grad_norm": 0.5746401068264577,
+      "learning_rate": 3.462700327830559e-07,
+      "loss": 0.2942,
+      "step": 1131
+    },
+    {
+      "epoch": 1.763239875389408,
+      "grad_norm": 0.5812252313415237,
+      "learning_rate": 3.418105448384507e-07,
+      "loss": 0.2478,
+      "step": 1132
+    },
+    {
+      "epoch": 1.764797507788162,
+      "grad_norm": 0.5205912713872596,
+      "learning_rate": 3.3737894291992426e-07,
+      "loss": 0.1793,
+      "step": 1133
+    },
+    {
+      "epoch": 1.7663551401869158,
+      "grad_norm": 0.59827648712384,
+      "learning_rate": 3.32975253557048e-07,
+      "loss": 0.1643,
+      "step": 1134
+    },
+    {
+      "epoch": 1.7679127725856698,
+      "grad_norm": 0.5476307417528222,
+      "learning_rate": 3.2859950311229625e-07,
+      "loss": 0.2551,
+      "step": 1135
+    },
+    {
+      "epoch": 1.7694704049844237,
+      "grad_norm": 0.5912924193469346,
+      "learning_rate": 3.242517177808874e-07,
+      "loss": 0.2958,
+      "step": 1136
+    },
+    {
+      "epoch": 1.7710280373831777,
+      "grad_norm": 0.6149100468721329,
+      "learning_rate": 3.199319235906273e-07,
+      "loss": 0.2391,
+      "step": 1137
+    },
+    {
+      "epoch": 1.7725856697819315,
+      "grad_norm": 0.5135930360411095,
+      "learning_rate": 3.1564014640175613e-07,
+      "loss": 0.1651,
+      "step": 1138
+    },
+    {
+      "epoch": 1.7741433021806854,
+      "grad_norm": 0.545428012439618,
+      "learning_rate": 3.1137641190678967e-07,
+      "loss": 0.1866,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7757009345794392,
+      "grad_norm": 0.7534351814752701,
+      "learning_rate": 3.0714074563037043e-07,
+      "loss": 0.1752,
+      "step": 1140
+    },
+    {
+      "epoch": 1.777258566978193,
+      "grad_norm": 0.6368630911991421,
+      "learning_rate": 3.029331729291074e-07,
+      "loss": 0.2049,
+      "step": 1141
+    },
+    {
+      "epoch": 1.778816199376947,
+      "grad_norm": 0.594879170975228,
+      "learning_rate": 2.9875371899143246e-07,
+      "loss": 0.6515,
+      "step": 1142
+    },
+    {
+      "epoch": 1.780373831775701,
+      "grad_norm": 0.5107491983811178,
+      "learning_rate": 2.9460240883744496e-07,
+      "loss": 0.1768,
+      "step": 1143
+    },
+    {
+      "epoch": 1.7819314641744548,
+      "grad_norm": 0.5417424283009235,
+      "learning_rate": 2.9047926731876177e-07,
+      "loss": 0.2087,
+      "step": 1144
+    },
+    {
+      "epoch": 1.7834890965732089,
+      "grad_norm": 0.6798250278910997,
+      "learning_rate": 2.8638431911837126e-07,
+      "loss": 0.28,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7850467289719627,
+      "grad_norm": 1.0039697477055525,
+      "learning_rate": 2.8231758875048087e-07,
+      "loss": 0.2178,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7866043613707165,
+      "grad_norm": 0.6262879884675681,
+      "learning_rate": 2.7827910056037857e-07,
+      "loss": 0.1775,
+      "step": 1147
+    },
+    {
+      "epoch": 1.7881619937694704,
+      "grad_norm": 0.6215897703245564,
+      "learning_rate": 2.742688787242764e-07,
+      "loss": 0.1726,
+      "step": 1148
+    },
+    {
+      "epoch": 1.7897196261682242,
+      "grad_norm": 0.7119013996583601,
+      "learning_rate": 2.702869472491748e-07,
+      "loss": 0.6972,
+      "step": 1149
+    },
+    {
+      "epoch": 1.791277258566978,
+      "grad_norm": 0.5300668663016045,
+      "learning_rate": 2.6633332997271277e-07,
+      "loss": 0.1797,
+      "step": 1150
+    },
+    {
+      "epoch": 1.7928348909657321,
+      "grad_norm": 0.6010057607669876,
+      "learning_rate": 2.624080505630311e-07,
+      "loss": 0.2206,
+      "step": 1151
+    },
+    {
+      "epoch": 1.794392523364486,
+      "grad_norm": 0.5566227596606641,
+      "learning_rate": 2.585111325186235e-07,
+      "loss": 0.2991,
+      "step": 1152
+    },
+    {
+      "epoch": 1.79595015576324,
+      "grad_norm": 0.7373120599532028,
+      "learning_rate": 2.5464259916820276e-07,
+      "loss": 0.2155,
+      "step": 1153
+    },
+    {
+      "epoch": 1.7975077881619939,
+      "grad_norm": 0.5444732937470245,
+      "learning_rate": 2.508024736705561e-07,
+      "loss": 0.2394,
+      "step": 1154
+    },
+    {
+      "epoch": 1.7990654205607477,
+      "grad_norm": 0.6187155054450917,
+      "learning_rate": 2.469907790144088e-07,
+      "loss": 0.1835,
+      "step": 1155
+    },
+    {
+      "epoch": 1.8006230529595015,
+      "grad_norm": 0.45905420748321907,
+      "learning_rate": 2.4320753801828853e-07,
+      "loss": 0.1604,
+      "step": 1156
+    },
+    {
+      "epoch": 1.8021806853582554,
+      "grad_norm": 0.43879992123704614,
+      "learning_rate": 2.394527733303831e-07,
+      "loss": 0.1723,
+      "step": 1157
+    },
+    {
+      "epoch": 1.8037383177570092,
+      "grad_norm": 0.5693210977954104,
+      "learning_rate": 2.3572650742840985e-07,
+      "loss": 0.1148,
+      "step": 1158
+    },
+    {
+      "epoch": 1.8052959501557633,
+      "grad_norm": 0.5824148680871707,
+      "learning_rate": 2.3202876261948004e-07,
+      "loss": 0.2304,
+      "step": 1159
+    },
+    {
+      "epoch": 1.8068535825545171,
+      "grad_norm": 0.692163156120945,
+      "learning_rate": 2.2835956103996525e-07,
+      "loss": 0.1877,
+      "step": 1160
+    },
+    {
+      "epoch": 1.8084112149532712,
+      "grad_norm": 0.627351603757966,
+      "learning_rate": 2.247189246553616e-07,
+      "loss": 0.2133,
+      "step": 1161
+    },
+    {
+      "epoch": 1.809968847352025,
+      "grad_norm": 0.6447153050268107,
+      "learning_rate": 2.2110687526016418e-07,
+      "loss": 0.1753,
+      "step": 1162
+    },
+    {
+      "epoch": 1.8115264797507789,
+      "grad_norm": 0.6498742455536043,
+      "learning_rate": 2.1752343447773105e-07,
+      "loss": 0.2801,
+      "step": 1163
+    },
+    {
+      "epoch": 1.8130841121495327,
+      "grad_norm": 0.5556829005135521,
+      "learning_rate": 2.1396862376015904e-07,
+      "loss": 0.2613,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8146417445482865,
+      "grad_norm": 0.5643429928213758,
+      "learning_rate": 2.1044246438814918e-07,
+      "loss": 0.2529,
+      "step": 1165
+    },
+    {
+      "epoch": 1.8161993769470404,
+      "grad_norm": 0.747941354496335,
+      "learning_rate": 2.0694497747088428e-07,
+      "loss": 0.2321,
+      "step": 1166
+    },
+    {
+      "epoch": 1.8177570093457944,
+      "grad_norm": 0.6102806034003578,
+      "learning_rate": 2.034761839459015e-07,
+      "loss": 0.218,
+      "step": 1167
+    },
+    {
+      "epoch": 1.8193146417445483,
+      "grad_norm": 0.7249346044063573,
+      "learning_rate": 2.0003610457896506e-07,
+      "loss": 0.2317,
+      "step": 1168
+    },
+    {
+      "epoch": 1.8208722741433023,
+      "grad_norm": 0.5858879975789372,
+      "learning_rate": 1.966247599639448e-07,
+      "loss": 0.1886,
+      "step": 1169
+    },
+    {
+      "epoch": 1.8224299065420562,
+      "grad_norm": 0.458859760014875,
+      "learning_rate": 1.9324217052268835e-07,
+      "loss": 0.2037,
+      "step": 1170
+    },
+    {
+      "epoch": 1.82398753894081,
+      "grad_norm": 0.6009222197974706,
+      "learning_rate": 1.8988835650490645e-07,
+      "loss": 0.1845,
+      "step": 1171
+    },
+    {
+      "epoch": 1.8255451713395638,
+      "grad_norm": 0.7980290452291322,
+      "learning_rate": 1.865633379880427e-07,
+      "loss": 0.2341,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8271028037383177,
+      "grad_norm": 0.5432836689095761,
+      "learning_rate": 1.832671348771614e-07,
+      "loss": 0.2282,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8286604361370715,
+      "grad_norm": 0.628743562125201,
+      "learning_rate": 1.799997669048209e-07,
+      "loss": 0.261,
+      "step": 1174
+    },
+    {
+      "epoch": 1.8302180685358256,
+      "grad_norm": 0.5832048925707848,
+      "learning_rate": 1.7676125363096431e-07,
+      "loss": 0.2157,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8317757009345794,
+      "grad_norm": 0.6281683541540782,
+      "learning_rate": 1.7355161444279346e-07,
+      "loss": 0.2799,
+      "step": 1176
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 0.6990945601104308,
+      "learning_rate": 1.7037086855465902e-07,
+      "loss": 0.2248,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8348909657320873,
+      "grad_norm": 0.48096314961131165,
+      "learning_rate": 1.6721903500794112e-07,
+      "loss": 0.2014,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8364485981308412,
+      "grad_norm": 0.6720925885316696,
+      "learning_rate": 1.6409613267094006e-07,
+      "loss": 0.1839,
+      "step": 1179
+    },
+    {
+      "epoch": 1.838006230529595,
+      "grad_norm": 0.5023236174498488,
+      "learning_rate": 1.6100218023876013e-07,
+      "loss": 0.1729,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8395638629283488,
+      "grad_norm": 0.6472364425599902,
+      "learning_rate": 1.5793719623319713e-07,
+      "loss": 0.2147,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8411214953271027,
+      "grad_norm": 0.5750750734641096,
+      "learning_rate": 1.5490119900263e-07,
+      "loss": 0.1734,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8426791277258567,
+      "grad_norm": 0.640400103725089,
+      "learning_rate": 1.5189420672190924e-07,
+      "loss": 0.2149,
+      "step": 1183
+    },
+    {
+      "epoch": 1.8442367601246106,
+      "grad_norm": 0.6159833044318871,
+      "learning_rate": 1.489162373922498e-07,
+      "loss": 0.2426,
+      "step": 1184
+    },
+    {
+      "epoch": 1.8457943925233646,
+      "grad_norm": 0.5891969287939168,
+      "learning_rate": 1.4596730884112008e-07,
+      "loss": 0.2164,
+      "step": 1185
+    },
+    {
+      "epoch": 1.8473520249221185,
+      "grad_norm": 0.6060235472721057,
+      "learning_rate": 1.4304743872213868e-07,
+      "loss": 0.2567,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8489096573208723,
+      "grad_norm": 0.6985533236531171,
+      "learning_rate": 1.4015664451496713e-07,
+      "loss": 0.1688,
+      "step": 1187
+    },
+    {
+      "epoch": 1.8504672897196262,
+      "grad_norm": 0.5859326752612632,
+      "learning_rate": 1.372949435252058e-07,
+      "loss": 0.2311,
+      "step": 1188
+    },
+    {
+      "epoch": 1.85202492211838,
+      "grad_norm": 0.7294577853076925,
+      "learning_rate": 1.3446235288428867e-07,
+      "loss": 0.2023,
+      "step": 1189
+    },
+    {
+      "epoch": 1.8535825545171338,
+      "grad_norm": 0.6888020410981194,
+      "learning_rate": 1.31658889549382e-07,
+      "loss": 0.2295,
+      "step": 1190
+    },
+    {
+      "epoch": 1.855140186915888,
+      "grad_norm": 0.6480785185138133,
+      "learning_rate": 1.288845703032854e-07,
+      "loss": 0.2359,
+      "step": 1191
+    },
+    {
+      "epoch": 1.8566978193146417,
+      "grad_norm": 0.7015546454899397,
+      "learning_rate": 1.2613941175432577e-07,
+      "loss": 0.2293,
+      "step": 1192
+    },
+    {
+      "epoch": 1.8582554517133958,
+      "grad_norm": 0.5363209042476926,
+      "learning_rate": 1.2342343033626248e-07,
+      "loss": 0.18,
+      "step": 1193
+    },
+    {
+      "epoch": 1.8598130841121496,
+      "grad_norm": 0.6459401688040521,
+      "learning_rate": 1.2073664230818571e-07,
+      "loss": 0.1598,
+      "step": 1194
+    },
+    {
+      "epoch": 1.8613707165109035,
+      "grad_norm": 0.638563320329426,
+      "learning_rate": 1.180790637544249e-07,
+      "loss": 0.2476,
+      "step": 1195
+    },
+    {
+      "epoch": 1.8629283489096573,
+      "grad_norm": 0.6002853300206833,
+      "learning_rate": 1.1545071058444324e-07,
+      "loss": 0.2199,
+      "step": 1196
+    },
+    {
+      "epoch": 1.8644859813084111,
+      "grad_norm": 0.5764195151127395,
+      "learning_rate": 1.1285159853275107e-07,
+      "loss": 0.1227,
+      "step": 1197
+    },
+    {
+      "epoch": 1.866043613707165,
+      "grad_norm": 0.6123103816053441,
+      "learning_rate": 1.1028174315880657e-07,
+      "loss": 0.815,
+      "step": 1198
+    },
+    {
+      "epoch": 1.867601246105919,
+      "grad_norm": 0.8123905890394445,
+      "learning_rate": 1.0774115984692523e-07,
+      "loss": 0.7364,
+      "step": 1199
+    },
+    {
+      "epoch": 1.8691588785046729,
+      "grad_norm": 0.5718174425382163,
+      "learning_rate": 1.0522986380618606e-07,
+      "loss": 0.2227,
+      "step": 1200
+    },
+    {
+      "epoch": 1.870716510903427,
+      "grad_norm": 0.5280345967170433,
+      "learning_rate": 1.0274787007034215e-07,
+      "loss": 0.232,
+      "step": 1201
+    },
+    {
+      "epoch": 1.8722741433021808,
+      "grad_norm": 0.6572165953398976,
+      "learning_rate": 1.0029519349772754e-07,
+      "loss": 0.2738,
+      "step": 1202
+    },
+    {
+      "epoch": 1.8738317757009346,
+      "grad_norm": 0.5244811712323645,
+      "learning_rate": 9.787184877117328e-08,
+      "loss": 0.2594,
+      "step": 1203
+    },
+    {
+      "epoch": 1.8753894080996885,
+      "grad_norm": 0.6676225172095591,
+      "learning_rate": 9.547785039791535e-08,
+      "loss": 0.1547,
+      "step": 1204
+    },
+    {
+      "epoch": 1.8769470404984423,
+      "grad_norm": 0.5476881051496193,
+      "learning_rate": 9.311321270950801e-08,
+      "loss": 0.2128,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8785046728971961,
+      "grad_norm": 0.5345787062303978,
+      "learning_rate": 9.077794986174226e-08,
+      "loss": 0.229,
+      "step": 1206
+    },
+    {
+      "epoch": 1.8800623052959502,
+      "grad_norm": 0.6008670505158255,
+      "learning_rate": 8.84720758345553e-08,
+      "loss": 0.2073,
+      "step": 1207
+    },
+    {
+      "epoch": 1.881619937694704,
+      "grad_norm": 0.6417543491732723,
+      "learning_rate": 8.619560443195174e-08,
+      "loss": 0.2276,
+      "step": 1208
+    },
+    {
+      "epoch": 1.883177570093458,
+      "grad_norm": 0.5552051709339759,
+      "learning_rate": 8.394854928191587e-08,
+      "loss": 0.2693,
+      "step": 1209
+    },
+    {
+      "epoch": 1.884735202492212,
+      "grad_norm": 0.6694794375263459,
+      "learning_rate": 8.173092383633563e-08,
+      "loss": 0.1767,
+      "step": 1210
+    },
+    {
+      "epoch": 1.8862928348909658,
+      "grad_norm": 0.5551294915003306,
+      "learning_rate": 7.954274137091877e-08,
+      "loss": 0.147,
+      "step": 1211
+    },
+    {
+      "epoch": 1.8878504672897196,
+      "grad_norm": 0.534343522912461,
+      "learning_rate": 7.738401498511406e-08,
+      "loss": 0.1891,
+      "step": 1212
+    },
+    {
+      "epoch": 1.8894080996884735,
+      "grad_norm": 0.6889315198186039,
+      "learning_rate": 7.525475760203239e-08,
+      "loss": 0.2832,
+      "step": 1213
+    },
+    {
+      "epoch": 1.8909657320872273,
+      "grad_norm": 0.6872767665873238,
+      "learning_rate": 7.315498196836968e-08,
+      "loss": 0.1472,
+      "step": 1214
+    },
+    {
+      "epoch": 1.8925233644859814,
+      "grad_norm": 0.6612078267914279,
+      "learning_rate": 7.108470065433193e-08,
+      "loss": 0.2704,
+      "step": 1215
+    },
+    {
+      "epoch": 1.8940809968847352,
+      "grad_norm": 0.560144578359172,
+      "learning_rate": 6.904392605355803e-08,
+      "loss": 0.2943,
+      "step": 1216
+    },
+    {
+      "epoch": 1.8956386292834893,
+      "grad_norm": 0.590357610353342,
+      "learning_rate": 6.703267038304706e-08,
+      "loss": 0.229,
+      "step": 1217
+    },
+    {
+      "epoch": 1.897196261682243,
+      "grad_norm": 0.496234632076301,
+      "learning_rate": 6.505094568308223e-08,
+      "loss": 0.2807,
+      "step": 1218
+    },
+    {
+      "epoch": 1.898753894080997,
+      "grad_norm": 0.6744039849106092,
+      "learning_rate": 6.309876381716429e-08,
+      "loss": 0.1744,
+      "step": 1219
+    },
+    {
+      "epoch": 1.9003115264797508,
+      "grad_norm": 0.6282540133425786,
+      "learning_rate": 6.117613647193543e-08,
+      "loss": 0.1896,
+      "step": 1220
+    },
+    {
+      "epoch": 1.9018691588785046,
+      "grad_norm": 0.545913837904785,
+      "learning_rate": 5.928307515711107e-08,
+      "loss": 0.1798,
+      "step": 1221
+    },
+    {
+      "epoch": 1.9034267912772584,
+      "grad_norm": 0.5750165821891594,
+      "learning_rate": 5.741959120541096e-08,
+      "loss": 0.2328,
+      "step": 1222
+    },
+    {
+      "epoch": 1.9049844236760125,
+      "grad_norm": 0.6138340557718204,
+      "learning_rate": 5.558569577249207e-08,
+      "loss": 0.1179,
+      "step": 1223
+    },
+    {
+      "epoch": 1.9065420560747663,
+      "grad_norm": 0.5250776885173793,
+      "learning_rate": 5.378139983688135e-08,
+      "loss": 0.178,
+      "step": 1224
+    },
+    {
+      "epoch": 1.9080996884735204,
+      "grad_norm": 0.5353000767702344,
+      "learning_rate": 5.2006714199908106e-08,
+      "loss": 0.1881,
+      "step": 1225
+    },
+    {
+      "epoch": 1.9096573208722742,
+      "grad_norm": 0.6835275834635214,
+      "learning_rate": 5.026164948564283e-08,
+      "loss": 0.2075,
+      "step": 1226
+    },
+    {
+      "epoch": 1.911214953271028,
+      "grad_norm": 0.48910598318494225,
+      "learning_rate": 4.854621614083122e-08,
+      "loss": 0.1864,
+      "step": 1227
+    },
+    {
+      "epoch": 1.912772585669782,
+      "grad_norm": 0.5349299444820935,
+      "learning_rate": 4.686042443483196e-08,
+      "loss": 0.297,
+      "step": 1228
+    },
+    {
+      "epoch": 1.9143302180685358,
+      "grad_norm": 0.42493561251864087,
+      "learning_rate": 4.520428445955571e-08,
+      "loss": 0.1656,
+      "step": 1229
+    },
+    {
+      "epoch": 1.9158878504672896,
+      "grad_norm": 0.5786034487132414,
+      "learning_rate": 4.357780612940343e-08,
+      "loss": 0.2243,
+      "step": 1230
+    },
+    {
+      "epoch": 1.9174454828660437,
+      "grad_norm": 0.6628360349913834,
+      "learning_rate": 4.1980999181210344e-08,
+      "loss": 0.237,
+      "step": 1231
+    },
+    {
+      "epoch": 1.9190031152647975,
+      "grad_norm": 0.5726232559997532,
+      "learning_rate": 4.041387317418377e-08,
+      "loss": 0.19,
+      "step": 1232
+    },
+    {
+      "epoch": 1.9205607476635516,
+      "grad_norm": 0.5957569462038291,
+      "learning_rate": 3.88764374898476e-08,
+      "loss": 0.1922,
+      "step": 1233
+    },
+    {
+      "epoch": 1.9221183800623054,
+      "grad_norm": 0.5981083547680797,
+      "learning_rate": 3.736870133198622e-08,
+      "loss": 0.1861,
+      "step": 1234
+    },
+    {
+      "epoch": 1.9236760124610592,
+      "grad_norm": 0.5039034286835927,
+      "learning_rate": 3.589067372658961e-08,
+      "loss": 0.1587,
+      "step": 1235
+    },
+    {
+      "epoch": 1.925233644859813,
+      "grad_norm": 0.5946390323111295,
+      "learning_rate": 3.444236352179831e-08,
+      "loss": 0.1658,
+      "step": 1236
+    },
+    {
+      "epoch": 1.926791277258567,
+      "grad_norm": 0.595459693543413,
+      "learning_rate": 3.302377938785128e-08,
+      "loss": 0.1509,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9283489096573208,
+      "grad_norm": 0.6496361440227659,
+      "learning_rate": 3.163492981703431e-08,
+      "loss": 0.693,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9299065420560748,
+      "grad_norm": 0.6473442377574977,
+      "learning_rate": 3.027582312362776e-08,
+      "loss": 0.2082,
+      "step": 1239
+    },
+    {
+      "epoch": 1.9314641744548287,
+      "grad_norm": 0.6541508019021366,
+      "learning_rate": 2.8946467443858873e-08,
+      "loss": 0.1911,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9330218068535827,
+      "grad_norm": 0.5343926005691645,
+      "learning_rate": 2.764687073585015e-08,
+      "loss": 0.1724,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9345794392523366,
+      "grad_norm": 0.6477434345058528,
+      "learning_rate": 2.6377040779574924e-08,
+      "loss": 0.2109,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9361370716510904,
+      "grad_norm": 0.4788618278691479,
+      "learning_rate": 2.5136985176809625e-08,
+      "loss": 0.2257,
+      "step": 1243
+    },
+    {
+      "epoch": 1.9376947040498442,
+      "grad_norm": 0.7161454071348646,
+      "learning_rate": 2.3926711351086594e-08,
+      "loss": 0.2201,
+      "step": 1244
+    },
+    {
+      "epoch": 1.939252336448598,
+      "grad_norm": 0.6162622985277744,
+      "learning_rate": 2.2746226547651352e-08,
+      "loss": 0.2353,
+      "step": 1245
+    },
+    {
+      "epoch": 1.940809968847352,
+      "grad_norm": 0.6572303875322422,
+      "learning_rate": 2.1595537833419276e-08,
+      "loss": 0.1238,
+      "step": 1246
+    },
+    {
+      "epoch": 1.942367601246106,
+      "grad_norm": 0.4905959486856584,
+      "learning_rate": 2.047465209693289e-08,
+      "loss": 0.1913,
+      "step": 1247
+    },
+    {
+      "epoch": 1.9439252336448598,
+      "grad_norm": 0.6262698157072888,
+      "learning_rate": 1.9383576048320752e-08,
+      "loss": 0.2135,
+      "step": 1248
+    },
+    {
+      "epoch": 1.9454828660436139,
+      "grad_norm": 0.5977256427085367,
+      "learning_rate": 1.8322316219256396e-08,
+      "loss": 0.3103,
+      "step": 1249
+    },
+    {
+      "epoch": 1.9470404984423677,
+      "grad_norm": 0.5597696638121235,
+      "learning_rate": 1.7290878962920587e-08,
+      "loss": 0.2003,
+      "step": 1250
+    },
+    {
+      "epoch": 1.9485981308411215,
+      "grad_norm": 0.7414334515835818,
+      "learning_rate": 1.6289270453963e-08,
+      "loss": 0.133,
+      "step": 1251
+    },
+    {
+      "epoch": 1.9501557632398754,
+      "grad_norm": 0.6577156096873663,
+      "learning_rate": 1.531749668846394e-08,
+      "loss": 0.1938,
+      "step": 1252
+    },
+    {
+      "epoch": 1.9517133956386292,
+      "grad_norm": 0.44233229164345883,
+      "learning_rate": 1.4375563483901011e-08,
+      "loss": 0.2196,
+      "step": 1253
+    },
+    {
+      "epoch": 1.953271028037383,
+      "grad_norm": 0.5494650237391789,
+      "learning_rate": 1.3463476479110837e-08,
+      "loss": 0.239,
+      "step": 1254
+    },
+    {
+      "epoch": 1.9548286604361371,
+      "grad_norm": 0.4597714878638595,
+      "learning_rate": 1.2581241134258515e-08,
+      "loss": 0.1569,
+      "step": 1255
+    },
+    {
+      "epoch": 1.956386292834891,
+      "grad_norm": 0.5344018379047227,
+      "learning_rate": 1.1728862730803759e-08,
+      "loss": 0.2506,
+      "step": 1256
+    },
+    {
+      "epoch": 1.957943925233645,
+      "grad_norm": 0.6170756968114266,
+      "learning_rate": 1.0906346371468147e-08,
+      "loss": 0.1756,
+      "step": 1257
+    },
+    {
+      "epoch": 1.9595015576323989,
+      "grad_norm": 0.7298638061579327,
+      "learning_rate": 1.011369698020681e-08,
+      "loss": 1.2315,
+      "step": 1258
+    },
+    {
+      "epoch": 1.9610591900311527,
+      "grad_norm": 0.7726214050156128,
+      "learning_rate": 9.350919302176242e-09,
+      "loss": 0.2132,
+      "step": 1259
+    },
+    {
+      "epoch": 1.9626168224299065,
+      "grad_norm": 0.7662101654548159,
+      "learning_rate": 8.618017903708198e-09,
+      "loss": 0.1979,
+      "step": 1260
+    },
+    {
+      "epoch": 1.9641744548286604,
+      "grad_norm": 0.7759433714405526,
+      "learning_rate": 7.914997172282502e-09,
+      "loss": 0.7825,
+      "step": 1261
+    },
+    {
+      "epoch": 1.9657320872274142,
+      "grad_norm": 0.5561454562523979,
+      "learning_rate": 7.241861316497623e-09,
+      "loss": 0.1787,
+      "step": 1262
+    },
+    {
+      "epoch": 1.9672897196261683,
+      "grad_norm": 0.6088898793952118,
+      "learning_rate": 6.598614366049027e-09,
+      "loss": 0.1525,
+      "step": 1263
+    },
+    {
+      "epoch": 1.9688473520249221,
+      "grad_norm": 0.6388680112674057,
+      "learning_rate": 5.985260171704199e-09,
+      "loss": 0.2538,
+      "step": 1264
+    },
+    {
+      "epoch": 1.9704049844236762,
+      "grad_norm": 1.2537539847150645,
+      "learning_rate": 5.401802405277656e-09,
+      "loss": 0.208,
+      "step": 1265
+    },
+    {
+      "epoch": 1.97196261682243,
+      "grad_norm": 0.5173311007016309,
+      "learning_rate": 4.848244559610971e-09,
+      "loss": 0.1927,
+      "step": 1266
+    },
+    {
+      "epoch": 1.9735202492211839,
+      "grad_norm": 0.7196153604341586,
+      "learning_rate": 4.324589948552227e-09,
+      "loss": 0.1292,
+      "step": 1267
+    },
+    {
+      "epoch": 1.9750778816199377,
+      "grad_norm": 0.8257531526111838,
+      "learning_rate": 3.830841706934374e-09,
+      "loss": 0.2268,
+      "step": 1268
+    },
+    {
+      "epoch": 1.9766355140186915,
+      "grad_norm": 0.6450834847941774,
+      "learning_rate": 3.367002790556906e-09,
+      "loss": 0.8063,
+      "step": 1269
+    },
+    {
+      "epoch": 1.9781931464174454,
+      "grad_norm": 0.5873649988075496,
+      "learning_rate": 2.9330759761692086e-09,
+      "loss": 0.1333,
+      "step": 1270
+    },
+    {
+      "epoch": 1.9797507788161994,
+      "grad_norm": 0.5338211522410232,
+      "learning_rate": 2.5290638614533516e-09,
+      "loss": 0.1651,
+      "step": 1271
+    },
+    {
+      "epoch": 1.9813084112149533,
+      "grad_norm": 0.5778154838677994,
+      "learning_rate": 2.154968865007989e-09,
+      "loss": 0.1375,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9828660436137073,
+      "grad_norm": 0.6159025357834766,
+      "learning_rate": 1.8107932263350392e-09,
+      "loss": 0.2551,
+      "step": 1273
+    },
+    {
+      "epoch": 1.9844236760124612,
+      "grad_norm": 0.5276126425838753,
+      "learning_rate": 1.4965390058258035e-09,
+      "loss": 0.2857,
+      "step": 1274
+    },
+    {
+      "epoch": 1.985981308411215,
+      "grad_norm": 0.578702087814479,
+      "learning_rate": 1.2122080847470907e-09,
+      "loss": 0.1644,
+      "step": 1275
+    },
+    {
+      "epoch": 1.9875389408099688,
+      "grad_norm": 0.6156843300297827,
+      "learning_rate": 9.578021652323356e-10,
+      "loss": 0.203,
+      "step": 1276
+    },
+    {
+      "epoch": 1.9890965732087227,
+      "grad_norm": 0.5234412510622443,
+      "learning_rate": 7.3332277027105e-10,
+      "loss": 0.2086,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9906542056074765,
+      "grad_norm": 0.6889557509874759,
+      "learning_rate": 5.38771243697167e-10,
+      "loss": 0.2412,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9922118380062306,
+      "grad_norm": 0.5571525262541802,
+      "learning_rate": 3.7414875018515483e-10,
+      "loss": 0.314,
+      "step": 1279
+    },
+    {
+      "epoch": 1.9937694704049844,
+      "grad_norm": 0.48476554367462193,
+      "learning_rate": 2.3945627523891403e-10,
+      "loss": 0.1878,
+      "step": 1280
+    },
+    {
+      "epoch": 1.9953271028037385,
+      "grad_norm": 0.5759369634737167,
+      "learning_rate": 1.3469462518844733e-10,
+      "loss": 0.181,
+      "step": 1281
+    },
+    {
+      "epoch": 1.9968847352024923,
+      "grad_norm": 0.6640918787171276,
+      "learning_rate": 5.986442718486363e-11,
+      "loss": 0.1886,
+      "step": 1282
+    },
+    {
+      "epoch": 1.9984423676012462,
+      "grad_norm": 0.6150458129744069,
+      "learning_rate": 1.496612919482665e-11,
+      "loss": 0.2397,
+      "step": 1283
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6727485887005468,
+      "learning_rate": 0.0,
+      "loss": 0.2332,
+      "step": 1284
+    },
+    {
+      "epoch": 2.0,
+      "step": 1284,
+      "total_flos": 17255618273280.0,
+      "train_loss": 0.27858535300162723,
+      "train_runtime": 1476.77,
+      "train_samples_per_second": 3.477,
+      "train_steps_per_second": 0.869
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1284,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 70000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 17255618273280.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}