diff --git "a/checkpoint-1563/trainer_state.json" "b/checkpoint-1563/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1563/trainer_state.json"
@@ -0,0 +1,10975 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1563,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0019203072491598655,
+      "grad_norm": 3.050241640622783,
+      "learning_rate": 0.0,
+      "loss": 1.5054,
+      "step": 1
+    },
+    {
+      "epoch": 0.003840614498319731,
+      "grad_norm": 2.974162230197891,
+      "learning_rate": 3.831417624521073e-08,
+      "loss": 1.7399,
+      "step": 2
+    },
+    {
+      "epoch": 0.005760921747479597,
+      "grad_norm": 3.006494012208738,
+      "learning_rate": 7.662835249042146e-08,
+      "loss": 1.8793,
+      "step": 3
+    },
+    {
+      "epoch": 0.007681228996639462,
+      "grad_norm": 2.910641370990376,
+      "learning_rate": 1.1494252873563219e-07,
+      "loss": 1.7187,
+      "step": 4
+    },
+    {
+      "epoch": 0.009601536245799328,
+      "grad_norm": 2.6996013829444134,
+      "learning_rate": 1.5325670498084292e-07,
+      "loss": 1.6475,
+      "step": 5
+    },
+    {
+      "epoch": 0.011521843494959194,
+      "grad_norm": 3.0173297525530525,
+      "learning_rate": 1.9157088122605365e-07,
+      "loss": 1.6964,
+      "step": 6
+    },
+    {
+      "epoch": 0.01344215074411906,
+      "grad_norm": 3.0082364904112566,
+      "learning_rate": 2.2988505747126437e-07,
+      "loss": 1.6851,
+      "step": 7
+    },
+    {
+      "epoch": 0.015362457993278924,
+      "grad_norm": 2.880521601062656,
+      "learning_rate": 2.681992337164751e-07,
+      "loss": 1.557,
+      "step": 8
+    },
+    {
+      "epoch": 0.01728276524243879,
+      "grad_norm": 2.8379079611890896,
+      "learning_rate": 3.0651340996168583e-07,
+      "loss": 1.6569,
+      "step": 9
+    },
+    {
+      "epoch": 0.019203072491598656,
+      "grad_norm": 2.902207545955143,
+      "learning_rate": 3.4482758620689656e-07,
+      "loss": 1.637,
+      "step": 10
+    },
+    {
+      "epoch": 0.02112337974075852,
+      "grad_norm": 2.8146383036248883,
+      "learning_rate": 3.831417624521073e-07,
+      "loss": 1.7074,
+      "step": 11
+    },
+    {
+      "epoch": 0.023043686989918388,
+      "grad_norm": 2.7702439023075422,
+      "learning_rate": 4.2145593869731807e-07,
+      "loss": 1.6621,
+      "step": 12
+    },
+    {
+      "epoch": 0.024963994239078253,
+      "grad_norm": 2.948897328700205,
+      "learning_rate": 4.5977011494252875e-07,
+      "loss": 1.7141,
+      "step": 13
+    },
+    {
+      "epoch": 0.02688430148823812,
+      "grad_norm": 2.6155066202719666,
+      "learning_rate": 4.980842911877395e-07,
+      "loss": 1.6133,
+      "step": 14
+    },
+    {
+      "epoch": 0.028804608737397985,
+      "grad_norm": 2.7754744384918335,
+      "learning_rate": 5.363984674329502e-07,
+      "loss": 1.6922,
+      "step": 15
+    },
+    {
+      "epoch": 0.030724915986557848,
+      "grad_norm": 3.005540548334177,
+      "learning_rate": 5.747126436781609e-07,
+      "loss": 1.5753,
+      "step": 16
+    },
+    {
+      "epoch": 0.03264522323571772,
+      "grad_norm": 2.6675051377737486,
+      "learning_rate": 6.130268199233717e-07,
+      "loss": 1.6697,
+      "step": 17
+    },
+    {
+      "epoch": 0.03456553048487758,
+      "grad_norm": 2.7911452731594326,
+      "learning_rate": 6.513409961685824e-07,
+      "loss": 1.6166,
+      "step": 18
+    },
+    {
+      "epoch": 0.03648583773403745,
+      "grad_norm": 2.754640389635153,
+      "learning_rate": 6.896551724137931e-07,
+      "loss": 1.4773,
+      "step": 19
+    },
+    {
+      "epoch": 0.03840614498319731,
+      "grad_norm": 2.58056430792743,
+      "learning_rate": 7.27969348659004e-07,
+      "loss": 1.5533,
+      "step": 20
+    },
+    {
+      "epoch": 0.040326452232357174,
+      "grad_norm": 2.5780862888672047,
+      "learning_rate": 7.662835249042146e-07,
+      "loss": 1.5408,
+      "step": 21
+    },
+    {
+      "epoch": 0.04224675948151704,
+      "grad_norm": 2.2939155489081293,
+      "learning_rate": 8.045977011494253e-07,
+      "loss": 1.4477,
+      "step": 22
+    },
+    {
+      "epoch": 0.044167066730676906,
+      "grad_norm": 2.326783109749172,
+      "learning_rate": 8.429118773946361e-07,
+      "loss": 1.6446,
+      "step": 23
+    },
+    {
+      "epoch": 0.046087373979836775,
+      "grad_norm": 1.9784500676559427,
+      "learning_rate": 8.812260536398468e-07,
+      "loss": 1.4741,
+      "step": 24
+    },
+    {
+      "epoch": 0.04800768122899664,
+      "grad_norm": 2.158116045848602,
+      "learning_rate": 9.195402298850575e-07,
+      "loss": 1.8359,
+      "step": 25
+    },
+    {
+      "epoch": 0.04992798847815651,
+      "grad_norm": 2.035070253571019,
+      "learning_rate": 9.578544061302683e-07,
+      "loss": 1.6079,
+      "step": 26
+    },
+    {
+      "epoch": 0.05184829572731637,
+      "grad_norm": 1.9828772550714657,
+      "learning_rate": 9.96168582375479e-07,
+      "loss": 1.6275,
+      "step": 27
+    },
+    {
+      "epoch": 0.05376860297647624,
+      "grad_norm": 2.063763559262756,
+      "learning_rate": 1.0344827586206898e-06,
+      "loss": 1.6298,
+      "step": 28
+    },
+    {
+      "epoch": 0.0556889102256361,
+      "grad_norm": 1.9339129722646953,
+      "learning_rate": 1.0727969348659004e-06,
+      "loss": 1.5816,
+      "step": 29
+    },
+    {
+      "epoch": 0.05760921747479597,
+      "grad_norm": 1.916925980149754,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 1.6064,
+      "step": 30
+    },
+    {
+      "epoch": 0.05952952472395583,
+      "grad_norm": 1.6450444187274702,
+      "learning_rate": 1.1494252873563219e-06,
+      "loss": 1.5159,
+      "step": 31
+    },
+    {
+      "epoch": 0.061449831973115696,
+      "grad_norm": 1.6132237873302806,
+      "learning_rate": 1.1877394636015327e-06,
+      "loss": 1.5367,
+      "step": 32
+    },
+    {
+      "epoch": 0.06337013922227556,
+      "grad_norm": 1.818936433606168,
+      "learning_rate": 1.2260536398467433e-06,
+      "loss": 1.4623,
+      "step": 33
+    },
+    {
+      "epoch": 0.06529044647143543,
+      "grad_norm": 1.7241994573893415,
+      "learning_rate": 1.2643678160919542e-06,
+      "loss": 1.6151,
+      "step": 34
+    },
+    {
+      "epoch": 0.06721075372059529,
+      "grad_norm": 1.5649464939481088,
+      "learning_rate": 1.3026819923371648e-06,
+      "loss": 1.6588,
+      "step": 35
+    },
+    {
+      "epoch": 0.06913106096975516,
+      "grad_norm": 1.5636025180091495,
+      "learning_rate": 1.3409961685823756e-06,
+      "loss": 1.3946,
+      "step": 36
+    },
+    {
+      "epoch": 0.07105136821891503,
+      "grad_norm": 1.5461645880192942,
+      "learning_rate": 1.3793103448275862e-06,
+      "loss": 1.6768,
+      "step": 37
+    },
+    {
+      "epoch": 0.0729716754680749,
+      "grad_norm": 1.4344137662856442,
+      "learning_rate": 1.417624521072797e-06,
+      "loss": 1.4397,
+      "step": 38
+    },
+    {
+      "epoch": 0.07489198271723475,
+      "grad_norm": 1.3960130764357097,
+      "learning_rate": 1.455938697318008e-06,
+      "loss": 1.6793,
+      "step": 39
+    },
+    {
+      "epoch": 0.07681228996639462,
+      "grad_norm": 1.2396850770578887,
+      "learning_rate": 1.4942528735632185e-06,
+      "loss": 1.5595,
+      "step": 40
+    },
+    {
+      "epoch": 0.07873259721555449,
+      "grad_norm": 1.2226941074613882,
+      "learning_rate": 1.5325670498084292e-06,
+      "loss": 1.4513,
+      "step": 41
+    },
+    {
+      "epoch": 0.08065290446471435,
+      "grad_norm": 1.2420772331499856,
+      "learning_rate": 1.57088122605364e-06,
+      "loss": 1.5815,
+      "step": 42
+    },
+    {
+      "epoch": 0.08257321171387422,
+      "grad_norm": 1.1487654644464287,
+      "learning_rate": 1.6091954022988506e-06,
+      "loss": 1.5939,
+      "step": 43
+    },
+    {
+      "epoch": 0.08449351896303409,
+      "grad_norm": 1.2124391317795524,
+      "learning_rate": 1.6475095785440615e-06,
+      "loss": 1.3123,
+      "step": 44
+    },
+    {
+      "epoch": 0.08641382621219396,
+      "grad_norm": 1.2527034268005153,
+      "learning_rate": 1.6858237547892723e-06,
+      "loss": 1.6962,
+      "step": 45
+    },
+    {
+      "epoch": 0.08833413346135381,
+      "grad_norm": 1.22540150664902,
+      "learning_rate": 1.724137931034483e-06,
+      "loss": 1.5573,
+      "step": 46
+    },
+    {
+      "epoch": 0.09025444071051368,
+      "grad_norm": 1.0822330995590186,
+      "learning_rate": 1.7624521072796935e-06,
+      "loss": 1.6174,
+      "step": 47
+    },
+    {
+      "epoch": 0.09217474795967355,
+      "grad_norm": 1.3725369113494932,
+      "learning_rate": 1.8007662835249044e-06,
+      "loss": 1.5724,
+      "step": 48
+    },
+    {
+      "epoch": 0.09409505520883342,
+      "grad_norm": 1.1246522439222844,
+      "learning_rate": 1.839080459770115e-06,
+      "loss": 1.6577,
+      "step": 49
+    },
+    {
+      "epoch": 0.09601536245799328,
+      "grad_norm": 1.153435955007641,
+      "learning_rate": 1.8773946360153258e-06,
+      "loss": 1.6045,
+      "step": 50
+    },
+    {
+      "epoch": 0.09793566970715314,
+      "grad_norm": 1.1142741754051628,
+      "learning_rate": 1.9157088122605367e-06,
+      "loss": 1.6562,
+      "step": 51
+    },
+    {
+      "epoch": 0.09985597695631301,
+      "grad_norm": 1.0608003189926583,
+      "learning_rate": 1.9540229885057475e-06,
+      "loss": 1.4951,
+      "step": 52
+    },
+    {
+      "epoch": 0.10177628420547287,
+      "grad_norm": 1.0049278590265711,
+      "learning_rate": 1.992337164750958e-06,
+      "loss": 1.5432,
+      "step": 53
+    },
+    {
+      "epoch": 0.10369659145463274,
+      "grad_norm": 0.895715467779371,
+      "learning_rate": 2.0306513409961687e-06,
+      "loss": 1.4494,
+      "step": 54
+    },
+    {
+      "epoch": 0.10561689870379261,
+      "grad_norm": 0.948460307989256,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 1.6377,
+      "step": 55
+    },
+    {
+      "epoch": 0.10753720595295248,
+      "grad_norm": 0.8960302855343928,
+      "learning_rate": 2.1072796934865904e-06,
+      "loss": 1.5476,
+      "step": 56
+    },
+    {
+      "epoch": 0.10945751320211233,
+      "grad_norm": 0.8555220673502346,
+      "learning_rate": 2.145593869731801e-06,
+      "loss": 1.2937,
+      "step": 57
+    },
+    {
+      "epoch": 0.1113778204512722,
+      "grad_norm": 0.9297918052847216,
+      "learning_rate": 2.1839080459770117e-06,
+      "loss": 1.5018,
+      "step": 58
+    },
+    {
+      "epoch": 0.11329812770043207,
+      "grad_norm": 0.8615767845690797,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.4411,
+      "step": 59
+    },
+    {
+      "epoch": 0.11521843494959194,
+      "grad_norm": 0.9800340953282465,
+      "learning_rate": 2.260536398467433e-06,
+      "loss": 1.6592,
+      "step": 60
+    },
+    {
+      "epoch": 0.1171387421987518,
+      "grad_norm": 0.8319044483038768,
+      "learning_rate": 2.2988505747126437e-06,
+      "loss": 1.4999,
+      "step": 61
+    },
+    {
+      "epoch": 0.11905904944791167,
+      "grad_norm": 0.8113904065691636,
+      "learning_rate": 2.3371647509578546e-06,
+      "loss": 1.4058,
+      "step": 62
+    },
+    {
+      "epoch": 0.12097935669707154,
+      "grad_norm": 0.7584474578599453,
+      "learning_rate": 2.3754789272030654e-06,
+      "loss": 1.1995,
+      "step": 63
+    },
+    {
+      "epoch": 0.12289966394623139,
+      "grad_norm": 0.8112593744821786,
+      "learning_rate": 2.4137931034482762e-06,
+      "loss": 1.622,
+      "step": 64
+    },
+    {
+      "epoch": 0.12481997119539126,
+      "grad_norm": 0.8465973433211744,
+      "learning_rate": 2.4521072796934867e-06,
+      "loss": 1.5474,
+      "step": 65
+    },
+    {
+      "epoch": 0.12674027844455113,
+      "grad_norm": 0.829645722256521,
+      "learning_rate": 2.4904214559386975e-06,
+      "loss": 1.485,
+      "step": 66
+    },
+    {
+      "epoch": 0.128660585693711,
+      "grad_norm": 0.8480123859041653,
+      "learning_rate": 2.5287356321839083e-06,
+      "loss": 1.5641,
+      "step": 67
+    },
+    {
+      "epoch": 0.13058089294287087,
+      "grad_norm": 0.7945937151249043,
+      "learning_rate": 2.567049808429119e-06,
+      "loss": 1.3511,
+      "step": 68
+    },
+    {
+      "epoch": 0.13250120019203074,
+      "grad_norm": 0.8452583555772669,
+      "learning_rate": 2.6053639846743296e-06,
+      "loss": 1.4936,
+      "step": 69
+    },
+    {
+      "epoch": 0.13442150744119058,
+      "grad_norm": 0.7913167472525273,
+      "learning_rate": 2.6436781609195404e-06,
+      "loss": 1.524,
+      "step": 70
+    },
+    {
+      "epoch": 0.13634181469035045,
+      "grad_norm": 0.7632883226077274,
+      "learning_rate": 2.6819923371647512e-06,
+      "loss": 1.4957,
+      "step": 71
+    },
+    {
+      "epoch": 0.13826212193951032,
+      "grad_norm": 0.8040467176753551,
+      "learning_rate": 2.720306513409962e-06,
+      "loss": 1.5472,
+      "step": 72
+    },
+    {
+      "epoch": 0.1401824291886702,
+      "grad_norm": 0.8063495654842968,
+      "learning_rate": 2.7586206896551725e-06,
+      "loss": 1.4355,
+      "step": 73
+    },
+    {
+      "epoch": 0.14210273643783006,
+      "grad_norm": 0.8632589003342163,
+      "learning_rate": 2.796934865900383e-06,
+      "loss": 1.6251,
+      "step": 74
+    },
+    {
+      "epoch": 0.14402304368698993,
+      "grad_norm": 0.9065110727033228,
+      "learning_rate": 2.835249042145594e-06,
+      "loss": 1.3482,
+      "step": 75
+    },
+    {
+      "epoch": 0.1459433509361498,
+      "grad_norm": 0.7358621824934779,
+      "learning_rate": 2.8735632183908046e-06,
+      "loss": 1.4579,
+      "step": 76
+    },
+    {
+      "epoch": 0.14786365818530964,
+      "grad_norm": 0.7882740690499183,
+      "learning_rate": 2.911877394636016e-06,
+      "loss": 1.584,
+      "step": 77
+    },
+    {
+      "epoch": 0.1497839654344695,
+      "grad_norm": 0.8028390602285591,
+      "learning_rate": 2.9501915708812262e-06,
+      "loss": 1.5799,
+      "step": 78
+    },
+    {
+      "epoch": 0.15170427268362938,
+      "grad_norm": 0.7087212262202383,
+      "learning_rate": 2.988505747126437e-06,
+      "loss": 1.3796,
+      "step": 79
+    },
+    {
+      "epoch": 0.15362457993278925,
+      "grad_norm": 0.7477246745009772,
+      "learning_rate": 3.026819923371648e-06,
+      "loss": 1.4577,
+      "step": 80
+    },
+    {
+      "epoch": 0.15554488718194912,
+      "grad_norm": 0.7362367028641865,
+      "learning_rate": 3.0651340996168583e-06,
+      "loss": 1.3986,
+      "step": 81
+    },
+    {
+      "epoch": 0.15746519443110898,
+      "grad_norm": 0.7542749931181064,
+      "learning_rate": 3.103448275862069e-06,
+      "loss": 1.3923,
+      "step": 82
+    },
+    {
+      "epoch": 0.15938550168026885,
+      "grad_norm": 0.7591236429919996,
+      "learning_rate": 3.14176245210728e-06,
+      "loss": 1.468,
+      "step": 83
+    },
+    {
+      "epoch": 0.1613058089294287,
+      "grad_norm": 0.750073455610315,
+      "learning_rate": 3.180076628352491e-06,
+      "loss": 1.5728,
+      "step": 84
+    },
+    {
+      "epoch": 0.16322611617858857,
+      "grad_norm": 0.7463483577071213,
+      "learning_rate": 3.2183908045977012e-06,
+      "loss": 1.5751,
+      "step": 85
+    },
+    {
+      "epoch": 0.16514642342774843,
+      "grad_norm": 0.7459336348523459,
+      "learning_rate": 3.256704980842912e-06,
+      "loss": 1.5821,
+      "step": 86
+    },
+    {
+      "epoch": 0.1670667306769083,
+      "grad_norm": 0.7503745075545026,
+      "learning_rate": 3.295019157088123e-06,
+      "loss": 1.3811,
+      "step": 87
+    },
+    {
+      "epoch": 0.16898703792606817,
+      "grad_norm": 0.6781019129991772,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.4967,
+      "step": 88
+    },
+    {
+      "epoch": 0.17090734517522804,
+      "grad_norm": 0.6772160890677477,
+      "learning_rate": 3.3716475095785446e-06,
+      "loss": 1.4602,
+      "step": 89
+    },
+    {
+      "epoch": 0.1728276524243879,
+      "grad_norm": 0.7414559131468873,
+      "learning_rate": 3.409961685823755e-06,
+      "loss": 1.5384,
+      "step": 90
+    },
+    {
+      "epoch": 0.17474795967354778,
+      "grad_norm": 0.7339713894607118,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 1.3478,
+      "step": 91
+    },
+    {
+      "epoch": 0.17666826692270762,
+      "grad_norm": 0.6741291120903733,
+      "learning_rate": 3.4865900383141767e-06,
+      "loss": 1.3051,
+      "step": 92
+    },
+    {
+      "epoch": 0.1785885741718675,
+      "grad_norm": 0.8466746593898973,
+      "learning_rate": 3.524904214559387e-06,
+      "loss": 1.6414,
+      "step": 93
+    },
+    {
+      "epoch": 0.18050888142102736,
+      "grad_norm": 0.6235998792480368,
+      "learning_rate": 3.563218390804598e-06,
+      "loss": 1.3774,
+      "step": 94
+    },
+    {
+      "epoch": 0.18242918867018723,
+      "grad_norm": 0.7888875616341429,
+      "learning_rate": 3.6015325670498087e-06,
+      "loss": 1.2273,
+      "step": 95
+    },
+    {
+      "epoch": 0.1843494959193471,
+      "grad_norm": 0.7184060206606023,
+      "learning_rate": 3.6398467432950196e-06,
+      "loss": 1.4414,
+      "step": 96
+    },
+    {
+      "epoch": 0.18626980316850697,
+      "grad_norm": 0.6931282624303644,
+      "learning_rate": 3.67816091954023e-06,
+      "loss": 1.1996,
+      "step": 97
+    },
+    {
+      "epoch": 0.18819011041766684,
+      "grad_norm": 0.6898899817332669,
+      "learning_rate": 3.7164750957854412e-06,
+      "loss": 1.153,
+      "step": 98
+    },
+    {
+      "epoch": 0.19011041766682668,
+      "grad_norm": 0.6706065855351341,
+      "learning_rate": 3.7547892720306517e-06,
+      "loss": 1.3219,
+      "step": 99
+    },
+    {
+      "epoch": 0.19203072491598655,
+      "grad_norm": 0.6618017863324709,
+      "learning_rate": 3.793103448275862e-06,
+      "loss": 1.3694,
+      "step": 100
+    },
+    {
+      "epoch": 0.19395103216514642,
+      "grad_norm": 0.7084267342732283,
+      "learning_rate": 3.831417624521073e-06,
+      "loss": 1.5242,
+      "step": 101
+    },
+    {
+      "epoch": 0.1958713394143063,
+      "grad_norm": 0.726545416016091,
+      "learning_rate": 3.869731800766283e-06,
+      "loss": 1.4521,
+      "step": 102
+    },
+    {
+      "epoch": 0.19779164666346616,
+      "grad_norm": 0.7462891561612502,
+      "learning_rate": 3.908045977011495e-06,
+      "loss": 1.4687,
+      "step": 103
+    },
+    {
+      "epoch": 0.19971195391262603,
+      "grad_norm": 0.7186363274122183,
+      "learning_rate": 3.946360153256705e-06,
+      "loss": 1.4222,
+      "step": 104
+    },
+    {
+      "epoch": 0.2016322611617859,
+      "grad_norm": 0.6562522942973675,
+      "learning_rate": 3.984674329501916e-06,
+      "loss": 1.3302,
+      "step": 105
+    },
+    {
+      "epoch": 0.20355256841094574,
+      "grad_norm": 0.7593764693035522,
+      "learning_rate": 4.022988505747127e-06,
+      "loss": 1.4029,
+      "step": 106
+    },
+    {
+      "epoch": 0.2054728756601056,
+      "grad_norm": 0.674187794860422,
+      "learning_rate": 4.0613026819923375e-06,
+      "loss": 1.2851,
+      "step": 107
+    },
+    {
+      "epoch": 0.20739318290926548,
+      "grad_norm": 0.6771453444719904,
+      "learning_rate": 4.099616858237548e-06,
+      "loss": 1.3873,
+      "step": 108
+    },
+    {
+      "epoch": 0.20931349015842535,
+      "grad_norm": 0.6904184501885661,
+      "learning_rate": 4.137931034482759e-06,
+      "loss": 1.4671,
+      "step": 109
+    },
+    {
+      "epoch": 0.21123379740758522,
+      "grad_norm": 0.7386010515763627,
+      "learning_rate": 4.17624521072797e-06,
+      "loss": 1.4114,
+      "step": 110
+    },
+    {
+      "epoch": 0.21315410465674509,
+      "grad_norm": 0.6330108602988729,
+      "learning_rate": 4.214559386973181e-06,
+      "loss": 1.5344,
+      "step": 111
+    },
+    {
+      "epoch": 0.21507441190590496,
+      "grad_norm": 0.6476316086673372,
+      "learning_rate": 4.252873563218391e-06,
+      "loss": 1.4134,
+      "step": 112
+    },
+    {
+      "epoch": 0.2169947191550648,
+      "grad_norm": 0.7369608607793586,
+      "learning_rate": 4.291187739463602e-06,
+      "loss": 1.4656,
+      "step": 113
+    },
+    {
+      "epoch": 0.21891502640422467,
+      "grad_norm": 0.6991819737178624,
+      "learning_rate": 4.3295019157088125e-06,
+      "loss": 1.3801,
+      "step": 114
+    },
+    {
+      "epoch": 0.22083533365338454,
+      "grad_norm": 0.6670879445681203,
+      "learning_rate": 4.367816091954023e-06,
+      "loss": 1.4012,
+      "step": 115
+    },
+    {
+      "epoch": 0.2227556409025444,
+      "grad_norm": 0.7091321651558993,
+      "learning_rate": 4.406130268199234e-06,
+      "loss": 1.369,
+      "step": 116
+    },
+    {
+      "epoch": 0.22467594815170427,
+      "grad_norm": 0.6538069130880931,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.207,
+      "step": 117
+    },
+    {
+      "epoch": 0.22659625540086414,
+      "grad_norm": 0.7414604205506811,
+      "learning_rate": 4.482758620689656e-06,
+      "loss": 1.4466,
+      "step": 118
+    },
+    {
+      "epoch": 0.228516562650024,
+      "grad_norm": 0.6754242592483168,
+      "learning_rate": 4.521072796934866e-06,
+      "loss": 1.345,
+      "step": 119
+    },
+    {
+      "epoch": 0.23043686989918388,
+      "grad_norm": 0.7890890418316723,
+      "learning_rate": 4.5593869731800775e-06,
+      "loss": 1.4887,
+      "step": 120
+    },
+    {
+      "epoch": 0.23235717714834372,
+      "grad_norm": 0.7142672200972927,
+      "learning_rate": 4.5977011494252875e-06,
+      "loss": 1.5663,
+      "step": 121
+    },
+    {
+      "epoch": 0.2342774843975036,
+      "grad_norm": 0.6607576396465201,
+      "learning_rate": 4.636015325670498e-06,
+      "loss": 1.5097,
+      "step": 122
+    },
+    {
+      "epoch": 0.23619779164666346,
+      "grad_norm": 0.7801245141924045,
+      "learning_rate": 4.674329501915709e-06,
+      "loss": 1.5577,
+      "step": 123
+    },
+    {
+      "epoch": 0.23811809889582333,
+      "grad_norm": 0.7853237789383211,
+      "learning_rate": 4.71264367816092e-06,
+      "loss": 1.5542,
+      "step": 124
+    },
+    {
+      "epoch": 0.2400384061449832,
+      "grad_norm": 0.6731916164633106,
+      "learning_rate": 4.750957854406131e-06,
+      "loss": 1.5991,
+      "step": 125
+    },
+    {
+      "epoch": 0.24195871339414307,
+      "grad_norm": 0.7081285505123779,
+      "learning_rate": 4.789272030651342e-06,
+      "loss": 1.4219,
+      "step": 126
+    },
+    {
+      "epoch": 0.24387902064330294,
+      "grad_norm": 0.6523124091427233,
+      "learning_rate": 4.8275862068965525e-06,
+      "loss": 1.3967,
+      "step": 127
+    },
+    {
+      "epoch": 0.24579932789246278,
+      "grad_norm": 0.7010318704056578,
+      "learning_rate": 4.8659003831417625e-06,
+      "loss": 1.4079,
+      "step": 128
+    },
+    {
+      "epoch": 0.24771963514162265,
+      "grad_norm": 0.7310188880844346,
+      "learning_rate": 4.904214559386973e-06,
+      "loss": 1.5645,
+      "step": 129
+    },
+    {
+      "epoch": 0.24963994239078252,
+      "grad_norm": 0.7370196750980766,
+      "learning_rate": 4.942528735632184e-06,
+      "loss": 1.3163,
+      "step": 130
+    },
+    {
+      "epoch": 0.2515602496399424,
+      "grad_norm": 0.5988613538056734,
+      "learning_rate": 4.980842911877395e-06,
+      "loss": 1.2164,
+      "step": 131
+    },
+    {
+      "epoch": 0.25348055688910226,
+      "grad_norm": 0.6686452758995931,
+      "learning_rate": 5.019157088122606e-06,
+      "loss": 1.3216,
+      "step": 132
+    },
+    {
+      "epoch": 0.25540086413826213,
+      "grad_norm": 0.6283032311907456,
+      "learning_rate": 5.057471264367817e-06,
+      "loss": 1.3004,
+      "step": 133
+    },
+    {
+      "epoch": 0.257321171387422,
+      "grad_norm": 0.7131274515928692,
+      "learning_rate": 5.095785440613027e-06,
+      "loss": 1.3574,
+      "step": 134
+    },
+    {
+      "epoch": 0.25924147863658187,
+      "grad_norm": 0.7562023685666645,
+      "learning_rate": 5.134099616858238e-06,
+      "loss": 1.4755,
+      "step": 135
+    },
+    {
+      "epoch": 0.26116178588574174,
+      "grad_norm": 0.7260126586378673,
+      "learning_rate": 5.172413793103449e-06,
+      "loss": 1.3512,
+      "step": 136
+    },
+    {
+      "epoch": 0.2630820931349016,
+      "grad_norm": 0.7027786093003081,
+      "learning_rate": 5.210727969348659e-06,
+      "loss": 1.3988,
+      "step": 137
+    },
+    {
+      "epoch": 0.2650024003840615,
+      "grad_norm": 0.6536542941035637,
+      "learning_rate": 5.24904214559387e-06,
+      "loss": 1.3403,
+      "step": 138
+    },
+    {
+      "epoch": 0.2669227076332213,
+      "grad_norm": 0.6887374540995002,
+      "learning_rate": 5.287356321839081e-06,
+      "loss": 1.4009,
+      "step": 139
+    },
+    {
+      "epoch": 0.26884301488238116,
+      "grad_norm": 0.6904650924816887,
+      "learning_rate": 5.3256704980842925e-06,
+      "loss": 1.3674,
+      "step": 140
+    },
+    {
+      "epoch": 0.27076332213154103,
+      "grad_norm": 0.7396766783545521,
+      "learning_rate": 5.3639846743295025e-06,
+      "loss": 1.2224,
+      "step": 141
+    },
+    {
+      "epoch": 0.2726836293807009,
+      "grad_norm": 0.6618698222762107,
+      "learning_rate": 5.402298850574713e-06,
+      "loss": 1.3261,
+      "step": 142
+    },
+    {
+      "epoch": 0.27460393662986077,
+      "grad_norm": 0.7346869512327489,
+      "learning_rate": 5.440613026819924e-06,
+      "loss": 1.5539,
+      "step": 143
+    },
+    {
+      "epoch": 0.27652424387902064,
+      "grad_norm": 0.6542272405748857,
+      "learning_rate": 5.478927203065134e-06,
+      "loss": 1.2301,
+      "step": 144
+    },
+    {
+      "epoch": 0.2784445511281805,
+      "grad_norm": 0.652844033827654,
+      "learning_rate": 5.517241379310345e-06,
+      "loss": 1.3203,
+      "step": 145
+    },
+    {
+      "epoch": 0.2803648583773404,
+      "grad_norm": 0.7093255205591261,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 1.4194,
+      "step": 146
+    },
+    {
+      "epoch": 0.28228516562650025,
+      "grad_norm": 0.7713263587807493,
+      "learning_rate": 5.593869731800766e-06,
+      "loss": 1.3986,
+      "step": 147
+    },
+    {
+      "epoch": 0.2842054728756601,
+      "grad_norm": 0.7273045777114594,
+      "learning_rate": 5.6321839080459775e-06,
+      "loss": 1.5226,
+      "step": 148
+    },
+    {
+      "epoch": 0.28612578012482,
+      "grad_norm": 0.6416135588874926,
+      "learning_rate": 5.670498084291188e-06,
+      "loss": 1.4148,
+      "step": 149
+    },
+    {
+      "epoch": 0.28804608737397985,
+      "grad_norm": 0.6767941047038493,
+      "learning_rate": 5.708812260536399e-06,
+      "loss": 1.4741,
+      "step": 150
+    },
+    {
+      "epoch": 0.2899663946231397,
+      "grad_norm": 0.7125976753134923,
+      "learning_rate": 5.747126436781609e-06,
+      "loss": 1.4689,
+      "step": 151
+    },
+    {
+      "epoch": 0.2918867018722996,
+      "grad_norm": 0.6848434615690435,
+      "learning_rate": 5.78544061302682e-06,
+      "loss": 1.3208,
+      "step": 152
+    },
+    {
+      "epoch": 0.2938070091214594,
+      "grad_norm": 0.7637154008633908,
+      "learning_rate": 5.823754789272032e-06,
+      "loss": 1.5237,
+      "step": 153
+    },
+    {
+      "epoch": 0.2957273163706193,
+      "grad_norm": 0.6483331857341429,
+      "learning_rate": 5.862068965517242e-06,
+      "loss": 1.3073,
+      "step": 154
+    },
+    {
+      "epoch": 0.29764762361977914,
+      "grad_norm": 0.684818235289912,
+      "learning_rate": 5.9003831417624525e-06,
+      "loss": 1.4076,
+      "step": 155
+    },
+    {
+      "epoch": 0.299567930868939,
+      "grad_norm": 0.6683114123494819,
+      "learning_rate": 5.938697318007663e-06,
+      "loss": 1.3518,
+      "step": 156
+    },
+    {
+      "epoch": 0.3014882381180989,
+      "grad_norm": 0.8301668940114854,
+      "learning_rate": 5.977011494252874e-06,
+      "loss": 1.5087,
+      "step": 157
+    },
+    {
+      "epoch": 0.30340854536725875,
+      "grad_norm": 0.7306107739349151,
+      "learning_rate": 6.015325670498084e-06,
+      "loss": 1.4187,
+      "step": 158
+    },
+    {
+      "epoch": 0.3053288526164186,
+      "grad_norm": 0.692887187718929,
+      "learning_rate": 6.053639846743296e-06,
+      "loss": 1.5365,
+      "step": 159
+    },
+    {
+      "epoch": 0.3072491598655785,
+      "grad_norm": 0.6502853635559999,
+      "learning_rate": 6.091954022988507e-06,
+      "loss": 1.3497,
+      "step": 160
+    },
+    {
+      "epoch": 0.30916946711473836,
+      "grad_norm": 0.7309674020853324,
+      "learning_rate": 6.130268199233717e-06,
+      "loss": 1.4965,
+      "step": 161
+    },
+    {
+      "epoch": 0.31108977436389823,
+      "grad_norm": 0.7124865800378807,
+      "learning_rate": 6.1685823754789275e-06,
+      "loss": 1.4449,
+      "step": 162
+    },
+    {
+      "epoch": 0.3130100816130581,
+      "grad_norm": 0.692769353628125,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 1.2823,
+      "step": 163
+    },
+    {
+      "epoch": 0.31493038886221797,
+      "grad_norm": 0.8120061941639997,
+      "learning_rate": 6.24521072796935e-06,
+      "loss": 1.5322,
+      "step": 164
+    },
+    {
+      "epoch": 0.31685069611137784,
+      "grad_norm": 0.6506906315887734,
+      "learning_rate": 6.28352490421456e-06,
+      "loss": 1.4138,
+      "step": 165
+    },
+    {
+      "epoch": 0.3187710033605377,
+      "grad_norm": 0.7021274588006857,
+      "learning_rate": 6.321839080459771e-06,
+      "loss": 1.5534,
+      "step": 166
+    },
+    {
+      "epoch": 0.3206913106096976,
+      "grad_norm": 0.6965648094070332,
+      "learning_rate": 6.360153256704982e-06,
+      "loss": 1.3055,
+      "step": 167
+    },
+    {
+      "epoch": 0.3226116178588574,
+      "grad_norm": 0.7498732720147595,
+      "learning_rate": 6.398467432950192e-06,
+      "loss": 1.2775,
+      "step": 168
+    },
+    {
+      "epoch": 0.32453192510801726,
+      "grad_norm": 0.7097564261194121,
+      "learning_rate": 6.4367816091954025e-06,
+      "loss": 1.3029,
+      "step": 169
+    },
+    {
+      "epoch": 0.32645223235717713,
+      "grad_norm": 0.6382917432482148,
+      "learning_rate": 6.475095785440614e-06,
+      "loss": 1.3569,
+      "step": 170
+    },
+    {
+      "epoch": 0.328372539606337,
+      "grad_norm": 0.6745359700702276,
+      "learning_rate": 6.513409961685824e-06,
+      "loss": 1.3339,
+      "step": 171
+    },
+    {
+      "epoch": 0.33029284685549687,
+      "grad_norm": 0.8056549461489378,
+      "learning_rate": 6.551724137931035e-06,
+      "loss": 1.4649,
+      "step": 172
+    },
+    {
+      "epoch": 0.33221315410465674,
+      "grad_norm": 0.7305964712888819,
+      "learning_rate": 6.590038314176246e-06,
+      "loss": 1.4451,
+      "step": 173
+    },
+    {
+      "epoch": 0.3341334613538166,
+      "grad_norm": 0.6201165387321146,
+      "learning_rate": 6.628352490421457e-06,
+      "loss": 1.3352,
+      "step": 174
+    },
+    {
+      "epoch": 0.3360537686029765,
+      "grad_norm": 0.8265036713279577,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.4794,
+      "step": 175
+    },
+    {
+      "epoch": 0.33797407585213635,
+      "grad_norm": 0.6893681404195715,
+      "learning_rate": 6.7049808429118775e-06,
+      "loss": 1.4419,
+      "step": 176
+    },
+    {
+      "epoch": 0.3398943831012962,
+      "grad_norm": 0.7509743815259918,
+      "learning_rate": 6.743295019157089e-06,
+      "loss": 1.3666,
+      "step": 177
+    },
+    {
+      "epoch": 0.3418146903504561,
+      "grad_norm": 0.7122960885573658,
+      "learning_rate": 6.781609195402299e-06,
+      "loss": 1.4108,
+      "step": 178
+    },
+    {
+      "epoch": 0.34373499759961595,
+      "grad_norm": 0.848930385570575,
+      "learning_rate": 6.81992337164751e-06,
+      "loss": 1.4827,
+      "step": 179
+    },
+    {
+      "epoch": 0.3456553048487758,
+      "grad_norm": 0.7602028733722221,
+      "learning_rate": 6.858237547892721e-06,
+      "loss": 1.3414,
+      "step": 180
+    },
+    {
+      "epoch": 0.3475756120979357,
+      "grad_norm": 0.8288114979771807,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 1.2727,
+      "step": 181
+    },
+    {
+      "epoch": 0.34949591934709556,
+      "grad_norm": 0.7693760683257455,
+      "learning_rate": 6.934865900383142e-06,
+      "loss": 1.3706,
+      "step": 182
+    },
+    {
+      "epoch": 0.3514162265962554,
+      "grad_norm": 0.7808563776887848,
+      "learning_rate": 6.973180076628353e-06,
+      "loss": 1.4723,
+      "step": 183
+    },
+    {
+      "epoch": 0.35333653384541525,
+      "grad_norm": 0.8734769267258163,
+      "learning_rate": 7.011494252873564e-06,
+      "loss": 1.2891,
+      "step": 184
+    },
+    {
+      "epoch": 0.3552568410945751,
+      "grad_norm": 0.8867229876837837,
+      "learning_rate": 7.049808429118774e-06,
+      "loss": 1.4359,
+      "step": 185
+    },
+    {
+      "epoch": 0.357177148343735,
+      "grad_norm": 0.8349083935516824,
+      "learning_rate": 7.088122605363985e-06,
+      "loss": 1.5565,
+      "step": 186
+    },
+    {
+      "epoch": 0.35909745559289485,
+      "grad_norm": 0.7400284977493244,
+      "learning_rate": 7.126436781609196e-06,
+      "loss": 1.456,
+      "step": 187
+    },
+    {
+      "epoch": 0.3610177628420547,
+      "grad_norm": 0.6439707610744461,
+      "learning_rate": 7.1647509578544075e-06,
+      "loss": 1.2718,
+      "step": 188
+    },
+    {
+      "epoch": 0.3629380700912146,
+      "grad_norm": 0.757428349385491,
+      "learning_rate": 7.2030651340996175e-06,
+      "loss": 1.3176,
+      "step": 189
+    },
+    {
+      "epoch": 0.36485837734037446,
+      "grad_norm": 0.6979336476743889,
+      "learning_rate": 7.241379310344828e-06,
+      "loss": 1.1578,
+      "step": 190
+    },
+    {
+      "epoch": 0.36677868458953433,
+      "grad_norm": 0.7318936685975646,
+      "learning_rate": 7.279693486590039e-06,
+      "loss": 1.5306,
+      "step": 191
+    },
+    {
+      "epoch": 0.3686989918386942,
+      "grad_norm": 0.6968684758048505,
+      "learning_rate": 7.318007662835249e-06,
+      "loss": 1.3176,
+      "step": 192
+    },
+    {
+      "epoch": 0.37061929908785407,
+      "grad_norm": 0.8360636142316585,
+      "learning_rate": 7.35632183908046e-06,
+      "loss": 1.6162,
+      "step": 193
+    },
+    {
+      "epoch": 0.37253960633701394,
+      "grad_norm": 0.6842163000752954,
+      "learning_rate": 7.394636015325672e-06,
+      "loss": 1.3775,
+      "step": 194
+    },
+    {
+      "epoch": 0.3744599135861738,
+      "grad_norm": 0.7405931672623491,
+      "learning_rate": 7.4329501915708825e-06,
+      "loss": 1.4155,
+      "step": 195
+    },
+    {
+      "epoch": 0.3763802208353337,
+      "grad_norm": 0.6724666872249973,
+      "learning_rate": 7.4712643678160925e-06,
+      "loss": 1.6041,
+      "step": 196
+    },
+    {
+      "epoch": 0.3783005280844935,
+      "grad_norm": 0.6600816017352631,
+      "learning_rate": 7.509578544061303e-06,
+      "loss": 1.3836,
+      "step": 197
+    },
+    {
+      "epoch": 0.38022083533365336,
+      "grad_norm": 0.7294649653837092,
+      "learning_rate": 7.547892720306514e-06,
+      "loss": 1.3381,
+      "step": 198
+    },
+    {
+      "epoch": 0.38214114258281323,
+      "grad_norm": 0.6849006152291969,
+      "learning_rate": 7.586206896551724e-06,
+      "loss": 1.3448,
+      "step": 199
+    },
+    {
+      "epoch": 0.3840614498319731,
+      "grad_norm": 0.6413698682933802,
+      "learning_rate": 7.624521072796936e-06,
+      "loss": 1.1691,
+      "step": 200
+    },
+    {
+      "epoch": 0.38598175708113297,
+      "grad_norm": 0.6761946714227771,
+      "learning_rate": 7.662835249042147e-06,
+      "loss": 1.2401,
+      "step": 201
+    },
+    {
+      "epoch": 0.38790206433029284,
+      "grad_norm": 0.6604987264007839,
+      "learning_rate": 7.701149425287356e-06,
+      "loss": 1.4297,
+      "step": 202
+    },
+    {
+      "epoch": 0.3898223715794527,
+      "grad_norm": 0.6655444711020334,
+      "learning_rate": 7.739463601532567e-06,
+      "loss": 1.2069,
+      "step": 203
+    },
+    {
+      "epoch": 0.3917426788286126,
+      "grad_norm": 0.716356344463303,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.3456,
+      "step": 204
+    },
+    {
+      "epoch": 0.39366298607777245,
+      "grad_norm": 0.7371288713977551,
+      "learning_rate": 7.81609195402299e-06,
+      "loss": 1.3331,
+      "step": 205
+    },
+    {
+      "epoch": 0.3955832933269323,
+      "grad_norm": 0.6679678204725035,
+      "learning_rate": 7.854406130268199e-06,
+      "loss": 1.4027,
+      "step": 206
+    },
+    {
+      "epoch": 0.3975036005760922,
+      "grad_norm": 0.6255494545199866,
+      "learning_rate": 7.89272030651341e-06,
+      "loss": 1.4204,
+      "step": 207
+    },
+    {
+      "epoch": 0.39942390782525206,
+      "grad_norm": 0.7776077277070255,
+      "learning_rate": 7.93103448275862e-06,
+      "loss": 1.4262,
+      "step": 208
+    },
+    {
+      "epoch": 0.4013442150744119,
+      "grad_norm": 0.7495735486802019,
+      "learning_rate": 7.969348659003832e-06,
+      "loss": 1.3803,
+      "step": 209
+    },
+    {
+      "epoch": 0.4032645223235718,
+      "grad_norm": 0.7359991087372613,
+      "learning_rate": 8.007662835249042e-06,
+      "loss": 1.3407,
+      "step": 210
+    },
+    {
+      "epoch": 0.40518482957273166,
+      "grad_norm": 0.74502169700267,
+      "learning_rate": 8.045977011494253e-06,
+      "loss": 1.2109,
+      "step": 211
+    },
+    {
+      "epoch": 0.4071051368218915,
+      "grad_norm": 0.672919730166899,
+      "learning_rate": 8.084291187739464e-06,
+      "loss": 1.4416,
+      "step": 212
+    },
+    {
+      "epoch": 0.40902544407105135,
+      "grad_norm": 0.9039099669663537,
+      "learning_rate": 8.122605363984675e-06,
+      "loss": 1.5869,
+      "step": 213
+    },
+    {
+      "epoch": 0.4109457513202112,
+      "grad_norm": 0.609809507238103,
+      "learning_rate": 8.160919540229886e-06,
+      "loss": 1.392,
+      "step": 214
+    },
+    {
+      "epoch": 0.4128660585693711,
+      "grad_norm": 0.7209905089139385,
+      "learning_rate": 8.199233716475097e-06,
+      "loss": 1.5286,
+      "step": 215
+    },
+    {
+      "epoch": 0.41478636581853096,
+      "grad_norm": 0.7641098257228572,
+      "learning_rate": 8.237547892720307e-06,
+      "loss": 1.4556,
+      "step": 216
+    },
+    {
+      "epoch": 0.4167066730676908,
+      "grad_norm": 0.6671329519012665,
+      "learning_rate": 8.275862068965518e-06,
+      "loss": 1.4011,
+      "step": 217
+    },
+    {
+      "epoch": 0.4186269803168507,
+      "grad_norm": 0.7571002797488336,
+      "learning_rate": 8.31417624521073e-06,
+      "loss": 1.3907,
+      "step": 218
+    },
+    {
+      "epoch": 0.42054728756601056,
+      "grad_norm": 0.6765159050714699,
+      "learning_rate": 8.35249042145594e-06,
+      "loss": 1.411,
+      "step": 219
+    },
+    {
+      "epoch": 0.42246759481517043,
+      "grad_norm": 0.7631812758174655,
+      "learning_rate": 8.390804597701149e-06,
+      "loss": 1.4568,
+      "step": 220
+    },
+    {
+      "epoch": 0.4243879020643303,
+      "grad_norm": 0.819638920429698,
+      "learning_rate": 8.429118773946362e-06,
+      "loss": 1.2607,
+      "step": 221
+    },
+    {
+      "epoch": 0.42630820931349017,
+      "grad_norm": 0.6856567916211197,
+      "learning_rate": 8.467432950191573e-06,
+      "loss": 1.3665,
+      "step": 222
+    },
+    {
+      "epoch": 0.42822851656265004,
+      "grad_norm": 0.756851173923122,
+      "learning_rate": 8.505747126436782e-06,
+      "loss": 1.4053,
+      "step": 223
+    },
+    {
+      "epoch": 0.4301488238118099,
+      "grad_norm": 0.7234111087737882,
+      "learning_rate": 8.544061302681992e-06,
+      "loss": 1.2986,
+      "step": 224
+    },
+    {
+      "epoch": 0.4320691310609698,
+      "grad_norm": 0.7822272702322975,
+      "learning_rate": 8.582375478927203e-06,
+      "loss": 1.4298,
+      "step": 225
+    },
+    {
+      "epoch": 0.4339894383101296,
+      "grad_norm": 0.6810248686630753,
+      "learning_rate": 8.620689655172414e-06,
+      "loss": 1.3399,
+      "step": 226
+    },
+    {
+      "epoch": 0.43590974555928946,
+      "grad_norm": 0.8051128970700839,
+      "learning_rate": 8.659003831417625e-06,
+      "loss": 1.5267,
+      "step": 227
+    },
+    {
+      "epoch": 0.43783005280844933,
+      "grad_norm": 0.732357170450308,
+      "learning_rate": 8.697318007662836e-06,
+      "loss": 1.4766,
+      "step": 228
+    },
+    {
+      "epoch": 0.4397503600576092,
+      "grad_norm": 0.8096174835706853,
+      "learning_rate": 8.735632183908047e-06,
+      "loss": 1.5061,
+      "step": 229
+    },
+    {
+      "epoch": 0.44167066730676907,
+      "grad_norm": 0.7498363788557091,
+      "learning_rate": 8.773946360153257e-06,
+      "loss": 1.5328,
+      "step": 230
+    },
+    {
+      "epoch": 0.44359097455592894,
+      "grad_norm": 0.6639308793432624,
+      "learning_rate": 8.812260536398468e-06,
+      "loss": 1.5014,
+      "step": 231
+    },
+    {
+      "epoch": 0.4455112818050888,
+      "grad_norm": 0.7955581619122697,
+      "learning_rate": 8.85057471264368e-06,
+      "loss": 1.4619,
+      "step": 232
+    },
+    {
+      "epoch": 0.4474315890542487,
+      "grad_norm": 0.6665415125116451,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.4401,
+      "step": 233
+    },
+    {
+      "epoch": 0.44935189630340855,
+      "grad_norm": 0.7497106975893845,
+      "learning_rate": 8.9272030651341e-06,
+      "loss": 1.4444,
+      "step": 234
+    },
+    {
+      "epoch": 0.4512722035525684,
+      "grad_norm": 0.7655250611774667,
+      "learning_rate": 8.965517241379312e-06,
+      "loss": 1.4183,
+      "step": 235
+    },
+    {
+      "epoch": 0.4531925108017283,
+      "grad_norm": 0.7069725743129521,
+      "learning_rate": 9.003831417624522e-06,
+      "loss": 1.4695,
+      "step": 236
+    },
+    {
+      "epoch": 0.45511281805088816,
+      "grad_norm": 0.7033810098540295,
+      "learning_rate": 9.042145593869732e-06,
+      "loss": 1.432,
+      "step": 237
+    },
+    {
+      "epoch": 0.457033125300048,
+      "grad_norm": 0.7657511117560273,
+      "learning_rate": 9.080459770114942e-06,
+      "loss": 1.3106,
+      "step": 238
+    },
+    {
+      "epoch": 0.4589534325492079,
+      "grad_norm": 0.6734934664683107,
+      "learning_rate": 9.118773946360155e-06,
+      "loss": 1.2657,
+      "step": 239
+    },
+    {
+      "epoch": 0.46087373979836777,
+      "grad_norm": 0.6793414583903499,
+      "learning_rate": 9.157088122605364e-06,
+      "loss": 1.2823,
+      "step": 240
+    },
+    {
+      "epoch": 0.4627940470475276,
+      "grad_norm": 0.6868551209136969,
+      "learning_rate": 9.195402298850575e-06,
+      "loss": 1.1981,
+      "step": 241
+    },
+    {
+      "epoch": 0.46471435429668745,
+      "grad_norm": 0.7999498220323118,
+      "learning_rate": 9.233716475095786e-06,
+      "loss": 1.3222,
+      "step": 242
+    },
+    {
+      "epoch": 0.4666346615458473,
+      "grad_norm": 0.6765218175071678,
+      "learning_rate": 9.272030651340997e-06,
+      "loss": 1.3159,
+      "step": 243
+    },
+    {
+      "epoch": 0.4685549687950072,
+      "grad_norm": 0.7633174577339077,
+      "learning_rate": 9.310344827586207e-06,
+      "loss": 1.4493,
+      "step": 244
+    },
+    {
+      "epoch": 0.47047527604416706,
+      "grad_norm": 0.7434731700432043,
+      "learning_rate": 9.348659003831418e-06,
+      "loss": 1.3771,
+      "step": 245
+    },
+    {
+      "epoch": 0.4723955832933269,
+      "grad_norm": 0.7887594763056447,
+      "learning_rate": 9.386973180076629e-06,
+      "loss": 1.2978,
+      "step": 246
+    },
+    {
+      "epoch": 0.4743158905424868,
+      "grad_norm": 0.6816410513843812,
+      "learning_rate": 9.42528735632184e-06,
+      "loss": 1.3221,
+      "step": 247
+    },
+    {
+      "epoch": 0.47623619779164666,
+      "grad_norm": 0.6726044542699641,
+      "learning_rate": 9.46360153256705e-06,
+      "loss": 1.3954,
+      "step": 248
+    },
+    {
+      "epoch": 0.47815650504080653,
+      "grad_norm": 0.8209401718350966,
+      "learning_rate": 9.501915708812262e-06,
+      "loss": 1.2964,
+      "step": 249
+    },
+    {
+      "epoch": 0.4800768122899664,
+      "grad_norm": 0.7150467022402414,
+      "learning_rate": 9.54022988505747e-06,
+      "loss": 1.2462,
+      "step": 250
+    },
+    {
+      "epoch": 0.4819971195391263,
+      "grad_norm": 0.7473286809494867,
+      "learning_rate": 9.578544061302683e-06,
+      "loss": 1.3987,
+      "step": 251
+    },
+    {
+      "epoch": 0.48391742678828614,
+      "grad_norm": 0.8919512518954593,
+      "learning_rate": 9.616858237547894e-06,
+      "loss": 1.3271,
+      "step": 252
+    },
+    {
+      "epoch": 0.485837734037446,
+      "grad_norm": 0.7235436092130889,
+      "learning_rate": 9.655172413793105e-06,
+      "loss": 1.4006,
+      "step": 253
+    },
+    {
+      "epoch": 0.4877580412866059,
+      "grad_norm": 0.7496985395380509,
+      "learning_rate": 9.693486590038314e-06,
+      "loss": 1.3485,
+      "step": 254
+    },
+    {
+      "epoch": 0.4896783485357657,
+      "grad_norm": 0.7558104853132458,
+      "learning_rate": 9.731800766283525e-06,
+      "loss": 1.4354,
+      "step": 255
+    },
+    {
+      "epoch": 0.49159865578492556,
+      "grad_norm": 0.8026084646949407,
+      "learning_rate": 9.770114942528738e-06,
+      "loss": 1.3458,
+      "step": 256
+    },
+    {
+      "epoch": 0.49351896303408543,
+      "grad_norm": 0.6871351528584623,
+      "learning_rate": 9.808429118773947e-06,
+      "loss": 1.2849,
+      "step": 257
+    },
+    {
+      "epoch": 0.4954392702832453,
+      "grad_norm": 0.6733156882936058,
+      "learning_rate": 9.846743295019157e-06,
+      "loss": 1.3416,
+      "step": 258
+    },
+    {
+      "epoch": 0.4973595775324052,
+      "grad_norm": 0.8040078483269378,
+      "learning_rate": 9.885057471264368e-06,
+      "loss": 1.3933,
+      "step": 259
+    },
+    {
+      "epoch": 0.49927988478156504,
+      "grad_norm": 0.7019620968242437,
+      "learning_rate": 9.923371647509579e-06,
+      "loss": 1.4092,
+      "step": 260
+    },
+    {
+      "epoch": 0.501200192030725,
+      "grad_norm": 0.7993119421084648,
+      "learning_rate": 9.96168582375479e-06,
+      "loss": 1.4496,
+      "step": 261
+    },
+    {
+      "epoch": 0.5031204992798848,
+      "grad_norm": 0.7500420731214819,
+      "learning_rate": 1e-05,
+      "loss": 1.275,
+      "step": 262
+    },
+    {
+      "epoch": 0.5050408065290446,
+      "grad_norm": 0.7096092331830312,
+      "learning_rate": 9.999995509192137e-06,
+      "loss": 1.4075,
+      "step": 263
+    },
+    {
+      "epoch": 0.5069611137782045,
+      "grad_norm": 0.6880277184241935,
+      "learning_rate": 9.999982036776617e-06,
+      "loss": 1.3853,
+      "step": 264
+    },
+    {
+      "epoch": 0.5088814210273643,
+      "grad_norm": 0.6887794428820988,
+      "learning_rate": 9.999959582777638e-06,
+      "loss": 1.2526,
+      "step": 265
+    },
+    {
+      "epoch": 0.5108017282765243,
+      "grad_norm": 0.6962486538004352,
+      "learning_rate": 9.999928147235536e-06,
+      "loss": 1.385,
+      "step": 266
+    },
+    {
+      "epoch": 0.5127220355256841,
+      "grad_norm": 0.7314717827625464,
+      "learning_rate": 9.99988773020678e-06,
+      "loss": 1.5364,
+      "step": 267
+    },
+    {
+      "epoch": 0.514642342774844,
+      "grad_norm": 0.6755436676174023,
+      "learning_rate": 9.99983833176397e-06,
+      "loss": 1.1625,
+      "step": 268
+    },
+    {
+      "epoch": 0.5165626500240038,
+      "grad_norm": 0.6799257698916303,
+      "learning_rate": 9.999779951995845e-06,
+      "loss": 1.3743,
+      "step": 269
+    },
+    {
+      "epoch": 0.5184829572731637,
+      "grad_norm": 0.6967760972692357,
+      "learning_rate": 9.99971259100727e-06,
+      "loss": 1.4619,
+      "step": 270
+    },
+    {
+      "epoch": 0.5204032645223235,
+      "grad_norm": 0.6876004824004736,
+      "learning_rate": 9.99963624891925e-06,
+      "loss": 1.2302,
+      "step": 271
+    },
+    {
+      "epoch": 0.5223235717714835,
+      "grad_norm": 0.741525025016876,
+      "learning_rate": 9.999550925868919e-06,
+      "loss": 1.4944,
+      "step": 272
+    },
+    {
+      "epoch": 0.5242438790206433,
+      "grad_norm": 0.7044462970993084,
+      "learning_rate": 9.999456622009545e-06,
+      "loss": 1.3233,
+      "step": 273
+    },
+    {
+      "epoch": 0.5261641862698032,
+      "grad_norm": 0.7147463675731596,
+      "learning_rate": 9.999353337510526e-06,
+      "loss": 1.4728,
+      "step": 274
+    },
+    {
+      "epoch": 0.528084493518963,
+      "grad_norm": 0.7580689885973482,
+      "learning_rate": 9.9992410725574e-06,
+      "loss": 1.3894,
+      "step": 275
+    },
+    {
+      "epoch": 0.530004800768123,
+      "grad_norm": 0.627159916011877,
+      "learning_rate": 9.999119827351824e-06,
+      "loss": 1.1499,
+      "step": 276
+    },
+    {
+      "epoch": 0.5319251080172828,
+      "grad_norm": 0.7347721430138173,
+      "learning_rate": 9.998989602111599e-06,
+      "loss": 1.3863,
+      "step": 277
+    },
+    {
+      "epoch": 0.5338454152664426,
+      "grad_norm": 0.7561542947388183,
+      "learning_rate": 9.99885039707065e-06,
+      "loss": 1.4094,
+      "step": 278
+    },
+    {
+      "epoch": 0.5357657225156025,
+      "grad_norm": 0.7257428578657052,
+      "learning_rate": 9.998702212479031e-06,
+      "loss": 1.3558,
+      "step": 279
+    },
+    {
+      "epoch": 0.5376860297647623,
+      "grad_norm": 0.7156710718415759,
+      "learning_rate": 9.998545048602938e-06,
+      "loss": 1.3932,
+      "step": 280
+    },
+    {
+      "epoch": 0.5396063370139222,
+      "grad_norm": 0.8324536861856755,
+      "learning_rate": 9.998378905724677e-06,
+      "loss": 1.4631,
+      "step": 281
+    },
+    {
+      "epoch": 0.5415266442630821,
+      "grad_norm": 0.769865333129953,
+      "learning_rate": 9.998203784142701e-06,
+      "loss": 1.4778,
+      "step": 282
+    },
+    {
+      "epoch": 0.543446951512242,
+      "grad_norm": 0.7558894226395197,
+      "learning_rate": 9.998019684171585e-06,
+      "loss": 1.3401,
+      "step": 283
+    },
+    {
+      "epoch": 0.5453672587614018,
+      "grad_norm": 0.745453902467428,
+      "learning_rate": 9.997826606142031e-06,
+      "loss": 1.376,
+      "step": 284
+    },
+    {
+      "epoch": 0.5472875660105617,
+      "grad_norm": 0.8525307677948906,
+      "learning_rate": 9.997624550400869e-06,
+      "loss": 1.3269,
+      "step": 285
+    },
+    {
+      "epoch": 0.5492078732597215,
+      "grad_norm": 0.8993675673573773,
+      "learning_rate": 9.997413517311055e-06,
+      "loss": 1.3973,
+      "step": 286
+    },
+    {
+      "epoch": 0.5511281805088815,
+      "grad_norm": 0.7637767723846134,
+      "learning_rate": 9.997193507251676e-06,
+      "loss": 1.4107,
+      "step": 287
+    },
+    {
+      "epoch": 0.5530484877580413,
+      "grad_norm": 0.7334145684503066,
+      "learning_rate": 9.996964520617938e-06,
+      "loss": 1.2548,
+      "step": 288
+    },
+    {
+      "epoch": 0.5549687950072012,
+      "grad_norm": 0.8286763467838726,
+      "learning_rate": 9.996726557821177e-06,
+      "loss": 1.3915,
+      "step": 289
+    },
+    {
+      "epoch": 0.556889102256361,
+      "grad_norm": 0.786223417485439,
+      "learning_rate": 9.996479619288853e-06,
+      "loss": 1.2113,
+      "step": 290
+    },
+    {
+      "epoch": 0.5588094095055209,
+      "grad_norm": 0.6629161249458739,
+      "learning_rate": 9.996223705464542e-06,
+      "loss": 1.383,
+      "step": 291
+    },
+    {
+      "epoch": 0.5607297167546808,
+      "grad_norm": 0.7763568743151492,
+      "learning_rate": 9.995958816807951e-06,
+      "loss": 1.511,
+      "step": 292
+    },
+    {
+      "epoch": 0.5626500240038406,
+      "grad_norm": 0.8706771128343518,
+      "learning_rate": 9.995684953794905e-06,
+      "loss": 1.301,
+      "step": 293
+    },
+    {
+      "epoch": 0.5645703312530005,
+      "grad_norm": 0.7186979734065922,
+      "learning_rate": 9.995402116917353e-06,
+      "loss": 1.3137,
+      "step": 294
+    },
+    {
+      "epoch": 0.5664906385021603,
+      "grad_norm": 0.7197477949946243,
+      "learning_rate": 9.995110306683358e-06,
+      "loss": 1.3924,
+      "step": 295
+    },
+    {
+      "epoch": 0.5684109457513202,
+      "grad_norm": 0.7990094094001562,
+      "learning_rate": 9.994809523617109e-06,
+      "loss": 1.3595,
+      "step": 296
+    },
+    {
+      "epoch": 0.57033125300048,
+      "grad_norm": 0.7548256910113419,
+      "learning_rate": 9.994499768258905e-06,
+      "loss": 1.2627,
+      "step": 297
+    },
+    {
+      "epoch": 0.57225156024964,
+      "grad_norm": 0.6080847660246027,
+      "learning_rate": 9.994181041165169e-06,
+      "loss": 1.233,
+      "step": 298
+    },
+    {
+      "epoch": 0.5741718674987998,
+      "grad_norm": 0.7522747291038051,
+      "learning_rate": 9.99385334290844e-06,
+      "loss": 1.4473,
+      "step": 299
+    },
+    {
+      "epoch": 0.5760921747479597,
+      "grad_norm": 0.693071662167599,
+      "learning_rate": 9.993516674077367e-06,
+      "loss": 1.3112,
+      "step": 300
+    },
+    {
+      "epoch": 0.5780124819971195,
+      "grad_norm": 0.7289575683988432,
+      "learning_rate": 9.993171035276717e-06,
+      "loss": 1.4447,
+      "step": 301
+    },
+    {
+      "epoch": 0.5799327892462794,
+      "grad_norm": 0.7219332107019164,
+      "learning_rate": 9.992816427127367e-06,
+      "loss": 1.3059,
+      "step": 302
+    },
+    {
+      "epoch": 0.5818530964954393,
+      "grad_norm": 0.7884436507160677,
+      "learning_rate": 9.992452850266313e-06,
+      "loss": 1.4828,
+      "step": 303
+    },
+    {
+      "epoch": 0.5837734037445992,
+      "grad_norm": 0.6835750513473184,
+      "learning_rate": 9.992080305346652e-06,
+      "loss": 1.3351,
+      "step": 304
+    },
+    {
+      "epoch": 0.585693710993759,
+      "grad_norm": 0.7455682427352298,
+      "learning_rate": 9.991698793037596e-06,
+      "loss": 1.3393,
+      "step": 305
+    },
+    {
+      "epoch": 0.5876140182429188,
+      "grad_norm": 0.6344234674402865,
+      "learning_rate": 9.991308314024466e-06,
+      "loss": 1.1186,
+      "step": 306
+    },
+    {
+      "epoch": 0.5895343254920787,
+      "grad_norm": 0.6463680302231726,
+      "learning_rate": 9.990908869008685e-06,
+      "loss": 1.3536,
+      "step": 307
+    },
+    {
+      "epoch": 0.5914546327412386,
+      "grad_norm": 0.7763640180054763,
+      "learning_rate": 9.99050045870779e-06,
+      "loss": 1.4034,
+      "step": 308
+    },
+    {
+      "epoch": 0.5933749399903985,
+      "grad_norm": 0.7729679972786335,
+      "learning_rate": 9.990083083855413e-06,
+      "loss": 1.518,
+      "step": 309
+    },
+    {
+      "epoch": 0.5952952472395583,
+      "grad_norm": 0.7293910429882318,
+      "learning_rate": 9.9896567452013e-06,
+      "loss": 1.4168,
+      "step": 310
+    },
+    {
+      "epoch": 0.5972155544887182,
+      "grad_norm": 0.6938721572185663,
+      "learning_rate": 9.989221443511286e-06,
+      "loss": 1.5813,
+      "step": 311
+    },
+    {
+      "epoch": 0.599135861737878,
+      "grad_norm": 0.6823348506968784,
+      "learning_rate": 9.98877717956732e-06,
+      "loss": 1.396,
+      "step": 312
+    },
+    {
+      "epoch": 0.601056168987038,
+      "grad_norm": 0.6470598419144085,
+      "learning_rate": 9.988323954167438e-06,
+      "loss": 1.3138,
+      "step": 313
+    },
+    {
+      "epoch": 0.6029764762361978,
+      "grad_norm": 0.6955870872672735,
+      "learning_rate": 9.987861768125783e-06,
+      "loss": 1.4043,
+      "step": 314
+    },
+    {
+      "epoch": 0.6048967834853577,
+      "grad_norm": 0.7788214111440995,
+      "learning_rate": 9.98739062227259e-06,
+      "loss": 1.5305,
+      "step": 315
+    },
+    {
+      "epoch": 0.6068170907345175,
+      "grad_norm": 0.6598071275711943,
+      "learning_rate": 9.986910517454188e-06,
+      "loss": 1.185,
+      "step": 316
+    },
+    {
+      "epoch": 0.6087373979836774,
+      "grad_norm": 0.7478469808999688,
+      "learning_rate": 9.986421454533001e-06,
+      "loss": 1.2875,
+      "step": 317
+    },
+    {
+      "epoch": 0.6106577052328372,
+      "grad_norm": 0.7102386979345604,
+      "learning_rate": 9.985923434387545e-06,
+      "loss": 1.2623,
+      "step": 318
+    },
+    {
+      "epoch": 0.6125780124819972,
+      "grad_norm": 0.8159684281974479,
+      "learning_rate": 9.985416457912423e-06,
+      "loss": 1.3859,
+      "step": 319
+    },
+    {
+      "epoch": 0.614498319731157,
+      "grad_norm": 0.7359831696017346,
+      "learning_rate": 9.984900526018331e-06,
+      "loss": 1.3127,
+      "step": 320
+    },
+    {
+      "epoch": 0.6164186269803168,
+      "grad_norm": 0.7406050107164467,
+      "learning_rate": 9.984375639632047e-06,
+      "loss": 1.3526,
+      "step": 321
+    },
+    {
+      "epoch": 0.6183389342294767,
+      "grad_norm": 0.7171563092825125,
+      "learning_rate": 9.98384179969644e-06,
+      "loss": 1.3718,
+      "step": 322
+    },
+    {
+      "epoch": 0.6202592414786365,
+      "grad_norm": 0.7444992833186831,
+      "learning_rate": 9.983299007170454e-06,
+      "loss": 1.2605,
+      "step": 323
+    },
+    {
+      "epoch": 0.6221795487277965,
+      "grad_norm": 0.6718703991955216,
+      "learning_rate": 9.982747263029123e-06,
+      "loss": 1.2696,
+      "step": 324
+    },
+    {
+      "epoch": 0.6240998559769563,
+      "grad_norm": 0.6953480856589247,
+      "learning_rate": 9.982186568263558e-06,
+      "loss": 1.402,
+      "step": 325
+    },
+    {
+      "epoch": 0.6260201632261162,
+      "grad_norm": 0.6814410132323387,
+      "learning_rate": 9.981616923880948e-06,
+      "loss": 1.357,
+      "step": 326
+    },
+    {
+      "epoch": 0.627940470475276,
+      "grad_norm": 0.6789399695802566,
+      "learning_rate": 9.981038330904556e-06,
+      "loss": 1.3563,
+      "step": 327
+    },
+    {
+      "epoch": 0.6298607777244359,
+      "grad_norm": 0.7084278817620212,
+      "learning_rate": 9.980450790373724e-06,
+      "loss": 1.3983,
+      "step": 328
+    },
+    {
+      "epoch": 0.6317810849735958,
+      "grad_norm": 0.7561797459599714,
+      "learning_rate": 9.979854303343866e-06,
+      "loss": 1.3998,
+      "step": 329
+    },
+    {
+      "epoch": 0.6337013922227557,
+      "grad_norm": 0.7063013474407709,
+      "learning_rate": 9.979248870886463e-06,
+      "loss": 1.4128,
+      "step": 330
+    },
+    {
+      "epoch": 0.6356216994719155,
+      "grad_norm": 0.7285839877290075,
+      "learning_rate": 9.978634494089066e-06,
+      "loss": 1.5026,
+      "step": 331
+    },
+    {
+      "epoch": 0.6375420067210754,
+      "grad_norm": 0.7631063764422019,
+      "learning_rate": 9.9780111740553e-06,
+      "loss": 1.3353,
+      "step": 332
+    },
+    {
+      "epoch": 0.6394623139702352,
+      "grad_norm": 0.8022733412976211,
+      "learning_rate": 9.977378911904843e-06,
+      "loss": 1.3017,
+      "step": 333
+    },
+    {
+      "epoch": 0.6413826212193952,
+      "grad_norm": 0.7397584519382538,
+      "learning_rate": 9.976737708773445e-06,
+      "loss": 1.4274,
+      "step": 334
+    },
+    {
+      "epoch": 0.643302928468555,
+      "grad_norm": 0.7498254808267637,
+      "learning_rate": 9.976087565812913e-06,
+      "loss": 1.2821,
+      "step": 335
+    },
+    {
+      "epoch": 0.6452232357177148,
+      "grad_norm": 0.8440233129581642,
+      "learning_rate": 9.975428484191117e-06,
+      "loss": 1.507,
+      "step": 336
+    },
+    {
+      "epoch": 0.6471435429668747,
+      "grad_norm": 0.7140420729824992,
+      "learning_rate": 9.974760465091975e-06,
+      "loss": 1.3541,
+      "step": 337
+    },
+    {
+      "epoch": 0.6490638502160345,
+      "grad_norm": 0.7935959679049022,
+      "learning_rate": 9.974083509715471e-06,
+      "loss": 1.3704,
+      "step": 338
+    },
+    {
+      "epoch": 0.6509841574651944,
+      "grad_norm": 0.8128531878052235,
+      "learning_rate": 9.973397619277631e-06,
+      "loss": 1.3941,
+      "step": 339
+    },
+    {
+      "epoch": 0.6529044647143543,
+      "grad_norm": 0.7087809281812139,
+      "learning_rate": 9.972702795010539e-06,
+      "loss": 1.3024,
+      "step": 340
+    },
+    {
+      "epoch": 0.6548247719635142,
+      "grad_norm": 0.8779292688681924,
+      "learning_rate": 9.971999038162322e-06,
+      "loss": 1.275,
+      "step": 341
+    },
+    {
+      "epoch": 0.656745079212674,
+      "grad_norm": 0.8321554450658822,
+      "learning_rate": 9.971286349997155e-06,
+      "loss": 1.4001,
+      "step": 342
+    },
+    {
+      "epoch": 0.6586653864618339,
+      "grad_norm": 0.6785835670278665,
+      "learning_rate": 9.970564731795259e-06,
+      "loss": 1.3563,
+      "step": 343
+    },
+    {
+      "epoch": 0.6605856937109937,
+      "grad_norm": 0.8154880274231641,
+      "learning_rate": 9.96983418485289e-06,
+      "loss": 1.2015,
+      "step": 344
+    },
+    {
+      "epoch": 0.6625060009601537,
+      "grad_norm": 0.7877809686842095,
+      "learning_rate": 9.969094710482345e-06,
+      "loss": 1.3176,
+      "step": 345
+    },
+    {
+      "epoch": 0.6644263082093135,
+      "grad_norm": 0.737869545031943,
+      "learning_rate": 9.968346310011965e-06,
+      "loss": 1.2681,
+      "step": 346
+    },
+    {
+      "epoch": 0.6663466154584734,
+      "grad_norm": 0.8367016706595558,
+      "learning_rate": 9.967588984786113e-06,
+      "loss": 1.3146,
+      "step": 347
+    },
+    {
+      "epoch": 0.6682669227076332,
+      "grad_norm": 0.7157484065111132,
+      "learning_rate": 9.966822736165194e-06,
+      "loss": 1.2279,
+      "step": 348
+    },
+    {
+      "epoch": 0.6701872299567931,
+      "grad_norm": 0.8340115611171157,
+      "learning_rate": 9.966047565525636e-06,
+      "loss": 1.346,
+      "step": 349
+    },
+    {
+      "epoch": 0.672107537205953,
+      "grad_norm": 0.7964501684625079,
+      "learning_rate": 9.965263474259896e-06,
+      "loss": 1.2604,
+      "step": 350
+    },
+    {
+      "epoch": 0.6740278444551128,
+      "grad_norm": 0.7644952492186728,
+      "learning_rate": 9.964470463776457e-06,
+      "loss": 1.2808,
+      "step": 351
+    },
+    {
+      "epoch": 0.6759481517042727,
+      "grad_norm": 0.715875104396104,
+      "learning_rate": 9.96366853549982e-06,
+      "loss": 1.3779,
+      "step": 352
+    },
+    {
+      "epoch": 0.6778684589534325,
+      "grad_norm": 0.6658897567313692,
+      "learning_rate": 9.962857690870507e-06,
+      "loss": 1.3005,
+      "step": 353
+    },
+    {
+      "epoch": 0.6797887662025924,
+      "grad_norm": 0.7869576623216241,
+      "learning_rate": 9.962037931345058e-06,
+      "loss": 1.3772,
+      "step": 354
+    },
+    {
+      "epoch": 0.6817090734517522,
+      "grad_norm": 0.9271972422960602,
+      "learning_rate": 9.96120925839603e-06,
+      "loss": 1.346,
+      "step": 355
+    },
+    {
+      "epoch": 0.6836293807009122,
+      "grad_norm": 0.7660166493856924,
+      "learning_rate": 9.96037167351198e-06,
+      "loss": 1.2745,
+      "step": 356
+    },
+    {
+      "epoch": 0.685549687950072,
+      "grad_norm": 0.6278698318719681,
+      "learning_rate": 9.959525178197484e-06,
+      "loss": 1.1849,
+      "step": 357
+    },
+    {
+      "epoch": 0.6874699951992319,
+      "grad_norm": 0.8957448047534449,
+      "learning_rate": 9.958669773973124e-06,
+      "loss": 1.47,
+      "step": 358
+    },
+    {
+      "epoch": 0.6893903024483917,
+      "grad_norm": 0.7893290798004599,
+      "learning_rate": 9.95780546237548e-06,
+      "loss": 1.4411,
+      "step": 359
+    },
+    {
+      "epoch": 0.6913106096975516,
+      "grad_norm": 0.6674442252477479,
+      "learning_rate": 9.956932244957135e-06,
+      "loss": 1.3468,
+      "step": 360
+    },
+    {
+      "epoch": 0.6932309169467115,
+      "grad_norm": 0.8637622685527974,
+      "learning_rate": 9.95605012328667e-06,
+      "loss": 1.4951,
+      "step": 361
+    },
+    {
+      "epoch": 0.6951512241958714,
+      "grad_norm": 0.7032170307148827,
+      "learning_rate": 9.95515909894866e-06,
+      "loss": 1.2168,
+      "step": 362
+    },
+    {
+      "epoch": 0.6970715314450312,
+      "grad_norm": 0.7604809716338964,
+      "learning_rate": 9.954259173543671e-06,
+      "loss": 1.3405,
+      "step": 363
+    },
+    {
+      "epoch": 0.6989918386941911,
+      "grad_norm": 0.7356596357988364,
+      "learning_rate": 9.953350348688264e-06,
+      "loss": 1.4851,
+      "step": 364
+    },
+    {
+      "epoch": 0.7009121459433509,
+      "grad_norm": 0.7106099387583814,
+      "learning_rate": 9.952432626014979e-06,
+      "loss": 1.4105,
+      "step": 365
+    },
+    {
+      "epoch": 0.7028324531925108,
+      "grad_norm": 0.7680103674421779,
+      "learning_rate": 9.951506007172344e-06,
+      "loss": 1.3767,
+      "step": 366
+    },
+    {
+      "epoch": 0.7047527604416707,
+      "grad_norm": 0.7045547217675817,
+      "learning_rate": 9.950570493824864e-06,
+      "loss": 1.3534,
+      "step": 367
+    },
+    {
+      "epoch": 0.7066730676908305,
+      "grad_norm": 0.8618867023288301,
+      "learning_rate": 9.949626087653026e-06,
+      "loss": 1.5932,
+      "step": 368
+    },
+    {
+      "epoch": 0.7085933749399904,
+      "grad_norm": 0.7514120877501722,
+      "learning_rate": 9.948672790353287e-06,
+      "loss": 1.4226,
+      "step": 369
+    },
+    {
+      "epoch": 0.7105136821891502,
+      "grad_norm": 0.7707655494026907,
+      "learning_rate": 9.947710603638078e-06,
+      "loss": 1.3086,
+      "step": 370
+    },
+    {
+      "epoch": 0.7124339894383102,
+      "grad_norm": 0.7353314917000675,
+      "learning_rate": 9.946739529235797e-06,
+      "loss": 1.3498,
+      "step": 371
+    },
+    {
+      "epoch": 0.71435429668747,
+      "grad_norm": 0.7677899312066072,
+      "learning_rate": 9.945759568890804e-06,
+      "loss": 1.337,
+      "step": 372
+    },
+    {
+      "epoch": 0.7162746039366299,
+      "grad_norm": 0.7323170602281932,
+      "learning_rate": 9.944770724363428e-06,
+      "loss": 1.2262,
+      "step": 373
+    },
+    {
+      "epoch": 0.7181949111857897,
+      "grad_norm": 0.7030503301748048,
+      "learning_rate": 9.943772997429955e-06,
+      "loss": 1.2604,
+      "step": 374
+    },
+    {
+      "epoch": 0.7201152184349496,
+      "grad_norm": 0.8803804845996765,
+      "learning_rate": 9.942766389882621e-06,
+      "loss": 1.3465,
+      "step": 375
+    },
+    {
+      "epoch": 0.7220355256841094,
+      "grad_norm": 0.765754153505594,
+      "learning_rate": 9.94175090352962e-06,
+      "loss": 1.4785,
+      "step": 376
+    },
+    {
+      "epoch": 0.7239558329332694,
+      "grad_norm": 0.7412100725496786,
+      "learning_rate": 9.940726540195093e-06,
+      "loss": 1.3886,
+      "step": 377
+    },
+    {
+      "epoch": 0.7258761401824292,
+      "grad_norm": 0.7352092180670398,
+      "learning_rate": 9.939693301719131e-06,
+      "loss": 1.3787,
+      "step": 378
+    },
+    {
+      "epoch": 0.727796447431589,
+      "grad_norm": 0.7081810984489154,
+      "learning_rate": 9.93865118995776e-06,
+      "loss": 1.2855,
+      "step": 379
+    },
+    {
+      "epoch": 0.7297167546807489,
+      "grad_norm": 0.721692280601312,
+      "learning_rate": 9.937600206782951e-06,
+      "loss": 1.2581,
+      "step": 380
+    },
+    {
+      "epoch": 0.7316370619299087,
+      "grad_norm": 0.7219716174107607,
+      "learning_rate": 9.93654035408261e-06,
+      "loss": 1.3989,
+      "step": 381
+    },
+    {
+      "epoch": 0.7335573691790687,
+      "grad_norm": 0.808844014227327,
+      "learning_rate": 9.935471633760572e-06,
+      "loss": 1.489,
+      "step": 382
+    },
+    {
+      "epoch": 0.7354776764282285,
+      "grad_norm": 0.6931394591191726,
+      "learning_rate": 9.934394047736608e-06,
+      "loss": 1.3596,
+      "step": 383
+    },
+    {
+      "epoch": 0.7373979836773884,
+      "grad_norm": 0.7255320444997646,
+      "learning_rate": 9.93330759794641e-06,
+      "loss": 1.2593,
+      "step": 384
+    },
+    {
+      "epoch": 0.7393182909265482,
+      "grad_norm": 0.6469865776133421,
+      "learning_rate": 9.932212286341591e-06,
+      "loss": 1.4305,
+      "step": 385
+    },
+    {
+      "epoch": 0.7412385981757081,
+      "grad_norm": 0.7276641692049547,
+      "learning_rate": 9.931108114889685e-06,
+      "loss": 1.531,
+      "step": 386
+    },
+    {
+      "epoch": 0.743158905424868,
+      "grad_norm": 0.7064363862608019,
+      "learning_rate": 9.929995085574142e-06,
+      "loss": 1.3905,
+      "step": 387
+    },
+    {
+      "epoch": 0.7450792126740279,
+      "grad_norm": 0.7331138015593877,
+      "learning_rate": 9.928873200394323e-06,
+      "loss": 1.3649,
+      "step": 388
+    },
+    {
+      "epoch": 0.7469995199231877,
+      "grad_norm": 0.7324112634125343,
+      "learning_rate": 9.927742461365493e-06,
+      "loss": 1.4049,
+      "step": 389
+    },
+    {
+      "epoch": 0.7489198271723476,
+      "grad_norm": 0.7448582260656762,
+      "learning_rate": 9.926602870518826e-06,
+      "loss": 1.3451,
+      "step": 390
+    },
+    {
+      "epoch": 0.7508401344215074,
+      "grad_norm": 0.7290925508892867,
+      "learning_rate": 9.925454429901397e-06,
+      "loss": 1.265,
+      "step": 391
+    },
+    {
+      "epoch": 0.7527604416706674,
+      "grad_norm": 0.8652107575311744,
+      "learning_rate": 9.924297141576176e-06,
+      "loss": 1.2601,
+      "step": 392
+    },
+    {
+      "epoch": 0.7546807489198272,
+      "grad_norm": 0.8112589543786329,
+      "learning_rate": 9.923131007622027e-06,
+      "loss": 1.3949,
+      "step": 393
+    },
+    {
+      "epoch": 0.756601056168987,
+      "grad_norm": 0.7564370059278668,
+      "learning_rate": 9.9219560301337e-06,
+      "loss": 1.5582,
+      "step": 394
+    },
+    {
+      "epoch": 0.7585213634181469,
+      "grad_norm": 0.7217323314506744,
+      "learning_rate": 9.920772211221841e-06,
+      "loss": 1.3385,
+      "step": 395
+    },
+    {
+      "epoch": 0.7604416706673067,
+      "grad_norm": 0.7276996906484348,
+      "learning_rate": 9.919579553012964e-06,
+      "loss": 1.3778,
+      "step": 396
+    },
+    {
+      "epoch": 0.7623619779164666,
+      "grad_norm": 0.7823149650395126,
+      "learning_rate": 9.918378057649474e-06,
+      "loss": 1.3767,
+      "step": 397
+    },
+    {
+      "epoch": 0.7642822851656265,
+      "grad_norm": 0.6630175095699308,
+      "learning_rate": 9.917167727289641e-06,
+      "loss": 1.4844,
+      "step": 398
+    },
+    {
+      "epoch": 0.7662025924147864,
+      "grad_norm": 0.7344567983746931,
+      "learning_rate": 9.915948564107611e-06,
+      "loss": 1.3379,
+      "step": 399
+    },
+    {
+      "epoch": 0.7681228996639462,
+      "grad_norm": 0.7466071399468908,
+      "learning_rate": 9.914720570293397e-06,
+      "loss": 1.3972,
+      "step": 400
+    },
+    {
+      "epoch": 0.7700432069131061,
+      "grad_norm": 0.6949526386517407,
+      "learning_rate": 9.913483748052871e-06,
+      "loss": 1.4014,
+      "step": 401
+    },
+    {
+      "epoch": 0.7719635141622659,
+      "grad_norm": 0.6613334975361209,
+      "learning_rate": 9.912238099607763e-06,
+      "loss": 1.2069,
+      "step": 402
+    },
+    {
+      "epoch": 0.7738838214114259,
+      "grad_norm": 0.6632714829710001,
+      "learning_rate": 9.910983627195665e-06,
+      "loss": 1.2427,
+      "step": 403
+    },
+    {
+      "epoch": 0.7758041286605857,
+      "grad_norm": 0.6899922310091848,
+      "learning_rate": 9.90972033307001e-06,
+      "loss": 1.4041,
+      "step": 404
+    },
+    {
+      "epoch": 0.7777244359097456,
+      "grad_norm": 0.6259530173512385,
+      "learning_rate": 9.908448219500087e-06,
+      "loss": 1.0889,
+      "step": 405
+    },
+    {
+      "epoch": 0.7796447431589054,
+      "grad_norm": 0.7856357851084043,
+      "learning_rate": 9.90716728877102e-06,
+      "loss": 1.4446,
+      "step": 406
+    },
+    {
+      "epoch": 0.7815650504080653,
+      "grad_norm": 0.6391414313005859,
+      "learning_rate": 9.905877543183776e-06,
+      "loss": 1.3569,
+      "step": 407
+    },
+    {
+      "epoch": 0.7834853576572252,
+      "grad_norm": 0.7168868905941123,
+      "learning_rate": 9.904578985055151e-06,
+      "loss": 1.3422,
+      "step": 408
+    },
+    {
+      "epoch": 0.785405664906385,
+      "grad_norm": 0.7244557046923163,
+      "learning_rate": 9.903271616717782e-06,
+      "loss": 1.4439,
+      "step": 409
+    },
+    {
+      "epoch": 0.7873259721555449,
+      "grad_norm": 0.7135679454347851,
+      "learning_rate": 9.901955440520121e-06,
+      "loss": 1.417,
+      "step": 410
+    },
+    {
+      "epoch": 0.7892462794047047,
+      "grad_norm": 0.6975305523738247,
+      "learning_rate": 9.900630458826443e-06,
+      "loss": 1.2732,
+      "step": 411
+    },
+    {
+      "epoch": 0.7911665866538646,
+      "grad_norm": 0.6844791536520864,
+      "learning_rate": 9.89929667401685e-06,
+      "loss": 1.1994,
+      "step": 412
+    },
+    {
+      "epoch": 0.7930868939030244,
+      "grad_norm": 0.6995208409953875,
+      "learning_rate": 9.897954088487245e-06,
+      "loss": 1.44,
+      "step": 413
+    },
+    {
+      "epoch": 0.7950072011521844,
+      "grad_norm": 0.6664160171541003,
+      "learning_rate": 9.896602704649348e-06,
+      "loss": 1.3604,
+      "step": 414
+    },
+    {
+      "epoch": 0.7969275084013442,
+      "grad_norm": 0.6554390013099766,
+      "learning_rate": 9.89524252493068e-06,
+      "loss": 1.3002,
+      "step": 415
+    },
+    {
+      "epoch": 0.7988478156505041,
+      "grad_norm": 0.7714848495681179,
+      "learning_rate": 9.893873551774561e-06,
+      "loss": 1.4559,
+      "step": 416
+    },
+    {
+      "epoch": 0.8007681228996639,
+      "grad_norm": 0.684531285979124,
+      "learning_rate": 9.892495787640117e-06,
+      "loss": 1.4116,
+      "step": 417
+    },
+    {
+      "epoch": 0.8026884301488239,
+      "grad_norm": 0.625351185162249,
+      "learning_rate": 9.891109235002248e-06,
+      "loss": 1.2968,
+      "step": 418
+    },
+    {
+      "epoch": 0.8046087373979837,
+      "grad_norm": 0.6072425157093212,
+      "learning_rate": 9.889713896351658e-06,
+      "loss": 1.2834,
+      "step": 419
+    },
+    {
+      "epoch": 0.8065290446471436,
+      "grad_norm": 0.7557266979543664,
+      "learning_rate": 9.888309774194822e-06,
+      "loss": 1.2581,
+      "step": 420
+    },
+    {
+      "epoch": 0.8084493518963034,
+      "grad_norm": 0.6948893024278046,
+      "learning_rate": 9.886896871053996e-06,
+      "loss": 1.3472,
+      "step": 421
+    },
+    {
+      "epoch": 0.8103696591454633,
+      "grad_norm": 0.782279217584078,
+      "learning_rate": 9.885475189467217e-06,
+      "loss": 1.3546,
+      "step": 422
+    },
+    {
+      "epoch": 0.8122899663946231,
+      "grad_norm": 0.7188760215338773,
+      "learning_rate": 9.884044731988278e-06,
+      "loss": 1.3683,
+      "step": 423
+    },
+    {
+      "epoch": 0.814210273643783,
+      "grad_norm": 0.7215722020243497,
+      "learning_rate": 9.882605501186747e-06,
+      "loss": 1.2629,
+      "step": 424
+    },
+    {
+      "epoch": 0.8161305808929429,
+      "grad_norm": 0.699083447265363,
+      "learning_rate": 9.881157499647944e-06,
+      "loss": 1.3218,
+      "step": 425
+    },
+    {
+      "epoch": 0.8180508881421027,
+      "grad_norm": 0.724984065762376,
+      "learning_rate": 9.87970072997295e-06,
+      "loss": 1.3034,
+      "step": 426
+    },
+    {
+      "epoch": 0.8199711953912626,
+      "grad_norm": 0.7049961948332424,
+      "learning_rate": 9.878235194778594e-06,
+      "loss": 1.4015,
+      "step": 427
+    },
+    {
+      "epoch": 0.8218915026404224,
+      "grad_norm": 0.6961558656551843,
+      "learning_rate": 9.87676089669745e-06,
+      "loss": 1.3292,
+      "step": 428
+    },
+    {
+      "epoch": 0.8238118098895824,
+      "grad_norm": 0.6755568609982437,
+      "learning_rate": 9.875277838377835e-06,
+      "loss": 1.2485,
+      "step": 429
+    },
+    {
+      "epoch": 0.8257321171387422,
+      "grad_norm": 0.731506928442002,
+      "learning_rate": 9.8737860224838e-06,
+      "loss": 1.371,
+      "step": 430
+    },
+    {
+      "epoch": 0.8276524243879021,
+      "grad_norm": 0.8321982922227138,
+      "learning_rate": 9.872285451695128e-06,
+      "loss": 1.3981,
+      "step": 431
+    },
+    {
+      "epoch": 0.8295727316370619,
+      "grad_norm": 0.7315030508325402,
+      "learning_rate": 9.87077612870733e-06,
+      "loss": 1.3311,
+      "step": 432
+    },
+    {
+      "epoch": 0.8314930388862218,
+      "grad_norm": 0.798773303498576,
+      "learning_rate": 9.869258056231638e-06,
+      "loss": 1.3727,
+      "step": 433
+    },
+    {
+      "epoch": 0.8334133461353816,
+      "grad_norm": 0.651844540018506,
+      "learning_rate": 9.867731236995e-06,
+      "loss": 1.3471,
+      "step": 434
+    },
+    {
+      "epoch": 0.8353336533845416,
+      "grad_norm": 0.6771670988304741,
+      "learning_rate": 9.866195673740076e-06,
+      "loss": 1.3032,
+      "step": 435
+    },
+    {
+      "epoch": 0.8372539606337014,
+      "grad_norm": 0.8611651792236157,
+      "learning_rate": 9.864651369225236e-06,
+      "loss": 1.2559,
+      "step": 436
+    },
+    {
+      "epoch": 0.8391742678828612,
+      "grad_norm": 0.7436061953882284,
+      "learning_rate": 9.863098326224546e-06,
+      "loss": 1.3166,
+      "step": 437
+    },
+    {
+      "epoch": 0.8410945751320211,
+      "grad_norm": 0.7409544469403813,
+      "learning_rate": 9.86153654752778e-06,
+      "loss": 1.4249,
+      "step": 438
+    },
+    {
+      "epoch": 0.8430148823811809,
+      "grad_norm": 0.7184845856939001,
+      "learning_rate": 9.859966035940391e-06,
+      "loss": 1.3899,
+      "step": 439
+    },
+    {
+      "epoch": 0.8449351896303409,
+      "grad_norm": 0.7569549756216626,
+      "learning_rate": 9.858386794283527e-06,
+      "loss": 1.4622,
+      "step": 440
+    },
+    {
+      "epoch": 0.8468554968795007,
+      "grad_norm": 0.7230751877755235,
+      "learning_rate": 9.856798825394017e-06,
+      "loss": 1.3834,
+      "step": 441
+    },
+    {
+      "epoch": 0.8487758041286606,
+      "grad_norm": 0.7861385241652444,
+      "learning_rate": 9.855202132124367e-06,
+      "loss": 1.4005,
+      "step": 442
+    },
+    {
+      "epoch": 0.8506961113778204,
+      "grad_norm": 0.654961628331442,
+      "learning_rate": 9.853596717342751e-06,
+      "loss": 1.2536,
+      "step": 443
+    },
+    {
+      "epoch": 0.8526164186269803,
+      "grad_norm": 0.7338664793979522,
+      "learning_rate": 9.851982583933015e-06,
+      "loss": 1.4289,
+      "step": 444
+    },
+    {
+      "epoch": 0.8545367258761402,
+      "grad_norm": 0.647102767057043,
+      "learning_rate": 9.850359734794664e-06,
+      "loss": 1.3167,
+      "step": 445
+    },
+    {
+      "epoch": 0.8564570331253001,
+      "grad_norm": 0.7940838292459027,
+      "learning_rate": 9.84872817284286e-06,
+      "loss": 1.4604,
+      "step": 446
+    },
+    {
+      "epoch": 0.8583773403744599,
+      "grad_norm": 0.7447428727236874,
+      "learning_rate": 9.847087901008415e-06,
+      "loss": 1.392,
+      "step": 447
+    },
+    {
+      "epoch": 0.8602976476236198,
+      "grad_norm": 0.7643508165383381,
+      "learning_rate": 9.845438922237787e-06,
+      "loss": 1.3017,
+      "step": 448
+    },
+    {
+      "epoch": 0.8622179548727796,
+      "grad_norm": 0.7617962346454569,
+      "learning_rate": 9.843781239493076e-06,
+      "loss": 1.6087,
+      "step": 449
+    },
+    {
+      "epoch": 0.8641382621219396,
+      "grad_norm": 0.7385042271210234,
+      "learning_rate": 9.842114855752013e-06,
+      "loss": 1.287,
+      "step": 450
+    },
+    {
+      "epoch": 0.8660585693710994,
+      "grad_norm": 0.7904769194673538,
+      "learning_rate": 9.840439774007963e-06,
+      "loss": 1.3847,
+      "step": 451
+    },
+    {
+      "epoch": 0.8679788766202592,
+      "grad_norm": 0.7796917727237478,
+      "learning_rate": 9.838755997269917e-06,
+      "loss": 1.4052,
+      "step": 452
+    },
+    {
+      "epoch": 0.8698991838694191,
+      "grad_norm": 0.6685557701211076,
+      "learning_rate": 9.837063528562479e-06,
+      "loss": 1.2191,
+      "step": 453
+    },
+    {
+      "epoch": 0.8718194911185789,
+      "grad_norm": 0.682007015556512,
+      "learning_rate": 9.835362370925868e-06,
+      "loss": 1.4041,
+      "step": 454
+    },
+    {
+      "epoch": 0.8737397983677389,
+      "grad_norm": 0.6809420760471568,
+      "learning_rate": 9.833652527415918e-06,
+      "loss": 1.1179,
+      "step": 455
+    },
+    {
+      "epoch": 0.8756601056168987,
+      "grad_norm": 0.7496451496119366,
+      "learning_rate": 9.831934001104056e-06,
+      "loss": 1.1863,
+      "step": 456
+    },
+    {
+      "epoch": 0.8775804128660586,
+      "grad_norm": 0.7958820882337777,
+      "learning_rate": 9.830206795077313e-06,
+      "loss": 1.4097,
+      "step": 457
+    },
+    {
+      "epoch": 0.8795007201152184,
+      "grad_norm": 0.7760893359826682,
+      "learning_rate": 9.828470912438308e-06,
+      "loss": 1.4765,
+      "step": 458
+    },
+    {
+      "epoch": 0.8814210273643783,
+      "grad_norm": 0.7181251163054755,
+      "learning_rate": 9.826726356305248e-06,
+      "loss": 1.2336,
+      "step": 459
+    },
+    {
+      "epoch": 0.8833413346135381,
+      "grad_norm": 0.7130113363618711,
+      "learning_rate": 9.824973129811919e-06,
+      "loss": 1.3356,
+      "step": 460
+    },
+    {
+      "epoch": 0.8852616418626981,
+      "grad_norm": 0.6976783021258673,
+      "learning_rate": 9.823211236107684e-06,
+      "loss": 1.5375,
+      "step": 461
+    },
+    {
+      "epoch": 0.8871819491118579,
+      "grad_norm": 0.6494482314881563,
+      "learning_rate": 9.82144067835747e-06,
+      "loss": 1.4216,
+      "step": 462
+    },
+    {
+      "epoch": 0.8891022563610178,
+      "grad_norm": 0.6799368531582445,
+      "learning_rate": 9.819661459741774e-06,
+      "loss": 1.2607,
+      "step": 463
+    },
+    {
+      "epoch": 0.8910225636101776,
+      "grad_norm": 0.7200924252996468,
+      "learning_rate": 9.817873583456646e-06,
+      "loss": 1.3954,
+      "step": 464
+    },
+    {
+      "epoch": 0.8929428708593375,
+      "grad_norm": 0.723830426362561,
+      "learning_rate": 9.816077052713689e-06,
+      "loss": 1.2634,
+      "step": 465
+    },
+    {
+      "epoch": 0.8948631781084974,
+      "grad_norm": 0.6795402172591166,
+      "learning_rate": 9.814271870740054e-06,
+      "loss": 1.288,
+      "step": 466
+    },
+    {
+      "epoch": 0.8967834853576572,
+      "grad_norm": 0.7798610485610813,
+      "learning_rate": 9.812458040778433e-06,
+      "loss": 1.3816,
+      "step": 467
+    },
+    {
+      "epoch": 0.8987037926068171,
+      "grad_norm": 0.6598552167407624,
+      "learning_rate": 9.810635566087046e-06,
+      "loss": 1.3142,
+      "step": 468
+    },
+    {
+      "epoch": 0.9006240998559769,
+      "grad_norm": 0.8267440234447042,
+      "learning_rate": 9.808804449939649e-06,
+      "loss": 1.3952,
+      "step": 469
+    },
+    {
+      "epoch": 0.9025444071051368,
+      "grad_norm": 0.6831251782753334,
+      "learning_rate": 9.806964695625521e-06,
+      "loss": 1.3507,
+      "step": 470
+    },
+    {
+      "epoch": 0.9044647143542967,
+      "grad_norm": 0.6913550521559437,
+      "learning_rate": 9.80511630644945e-06,
+      "loss": 1.3858,
+      "step": 471
+    },
+    {
+      "epoch": 0.9063850216034566,
+      "grad_norm": 0.651357091385134,
+      "learning_rate": 9.803259285731744e-06,
+      "loss": 1.3119,
+      "step": 472
+    },
+    {
+      "epoch": 0.9083053288526164,
+      "grad_norm": 0.6876771805281661,
+      "learning_rate": 9.801393636808213e-06,
+      "loss": 1.405,
+      "step": 473
+    },
+    {
+      "epoch": 0.9102256361017763,
+      "grad_norm": 0.6704161705700665,
+      "learning_rate": 9.79951936303016e-06,
+      "loss": 1.1645,
+      "step": 474
+    },
+    {
+      "epoch": 0.9121459433509361,
+      "grad_norm": 0.8403757549868232,
+      "learning_rate": 9.797636467764392e-06,
+      "loss": 1.3374,
+      "step": 475
+    },
+    {
+      "epoch": 0.914066250600096,
+      "grad_norm": 0.6976996239643926,
+      "learning_rate": 9.795744954393193e-06,
+      "loss": 1.2789,
+      "step": 476
+    },
+    {
+      "epoch": 0.9159865578492559,
+      "grad_norm": 0.7007105159698541,
+      "learning_rate": 9.793844826314338e-06,
+      "loss": 1.2513,
+      "step": 477
+    },
+    {
+      "epoch": 0.9179068650984158,
+      "grad_norm": 0.8046443692914484,
+      "learning_rate": 9.791936086941065e-06,
+      "loss": 1.4276,
+      "step": 478
+    },
+    {
+      "epoch": 0.9198271723475756,
+      "grad_norm": 0.6689604084398844,
+      "learning_rate": 9.790018739702091e-06,
+      "loss": 1.1329,
+      "step": 479
+    },
+    {
+      "epoch": 0.9217474795967355,
+      "grad_norm": 0.7418530237185074,
+      "learning_rate": 9.788092788041589e-06,
+      "loss": 1.2312,
+      "step": 480
+    },
+    {
+      "epoch": 0.9236677868458953,
+      "grad_norm": 0.7344158824384261,
+      "learning_rate": 9.78615823541919e-06,
+      "loss": 1.5327,
+      "step": 481
+    },
+    {
+      "epoch": 0.9255880940950552,
+      "grad_norm": 0.7029451657520306,
+      "learning_rate": 9.784215085309977e-06,
+      "loss": 1.3297,
+      "step": 482
+    },
+    {
+      "epoch": 0.9275084013442151,
+      "grad_norm": 0.8107964882673727,
+      "learning_rate": 9.782263341204477e-06,
+      "loss": 1.2561,
+      "step": 483
+    },
+    {
+      "epoch": 0.9294287085933749,
+      "grad_norm": 0.7114545290156662,
+      "learning_rate": 9.78030300660865e-06,
+      "loss": 1.3645,
+      "step": 484
+    },
+    {
+      "epoch": 0.9313490158425348,
+      "grad_norm": 0.7402003366232237,
+      "learning_rate": 9.77833408504389e-06,
+      "loss": 1.3642,
+      "step": 485
+    },
+    {
+      "epoch": 0.9332693230916946,
+      "grad_norm": 0.7109408774118504,
+      "learning_rate": 9.77635658004702e-06,
+      "loss": 1.3221,
+      "step": 486
+    },
+    {
+      "epoch": 0.9351896303408546,
+      "grad_norm": 0.7433358040489166,
+      "learning_rate": 9.774370495170276e-06,
+      "loss": 1.4449,
+      "step": 487
+    },
+    {
+      "epoch": 0.9371099375900144,
+      "grad_norm": 0.6785055655567391,
+      "learning_rate": 9.772375833981306e-06,
+      "loss": 1.3555,
+      "step": 488
+    },
+    {
+      "epoch": 0.9390302448391743,
+      "grad_norm": 0.76870851128488,
+      "learning_rate": 9.770372600063172e-06,
+      "loss": 1.284,
+      "step": 489
+    },
+    {
+      "epoch": 0.9409505520883341,
+      "grad_norm": 0.7344357851176699,
+      "learning_rate": 9.768360797014325e-06,
+      "loss": 1.2853,
+      "step": 490
+    },
+    {
+      "epoch": 0.942870859337494,
+      "grad_norm": 0.8470942465410728,
+      "learning_rate": 9.766340428448614e-06,
+      "loss": 1.3829,
+      "step": 491
+    },
+    {
+      "epoch": 0.9447911665866539,
+      "grad_norm": 0.7211389944931649,
+      "learning_rate": 9.764311497995272e-06,
+      "loss": 1.2677,
+      "step": 492
+    },
+    {
+      "epoch": 0.9467114738358138,
+      "grad_norm": 0.7084359929065828,
+      "learning_rate": 9.762274009298918e-06,
+      "loss": 1.2434,
+      "step": 493
+    },
+    {
+      "epoch": 0.9486317810849736,
+      "grad_norm": 0.7689934246592068,
+      "learning_rate": 9.760227966019537e-06,
+      "loss": 1.4095,
+      "step": 494
+    },
+    {
+      "epoch": 0.9505520883341335,
+      "grad_norm": 0.7773642092371199,
+      "learning_rate": 9.758173371832485e-06,
+      "loss": 1.3244,
+      "step": 495
+    },
+    {
+      "epoch": 0.9524723955832933,
+      "grad_norm": 0.6978701658115153,
+      "learning_rate": 9.756110230428476e-06,
+      "loss": 1.2787,
+      "step": 496
+    },
+    {
+      "epoch": 0.9543927028324531,
+      "grad_norm": 0.6910966359494893,
+      "learning_rate": 9.75403854551358e-06,
+      "loss": 1.3348,
+      "step": 497
+    },
+    {
+      "epoch": 0.9563130100816131,
+      "grad_norm": 0.732636720833676,
+      "learning_rate": 9.751958320809213e-06,
+      "loss": 1.2403,
+      "step": 498
+    },
+    {
+      "epoch": 0.9582333173307729,
+      "grad_norm": 0.7804889809056719,
+      "learning_rate": 9.749869560052128e-06,
+      "loss": 1.1905,
+      "step": 499
+    },
+    {
+      "epoch": 0.9601536245799328,
+      "grad_norm": 0.7286628977028098,
+      "learning_rate": 9.747772266994418e-06,
+      "loss": 1.3252,
+      "step": 500
+    },
+    {
+      "epoch": 0.9620739318290926,
+      "grad_norm": 0.730360159535326,
+      "learning_rate": 9.745666445403496e-06,
+      "loss": 1.4712,
+      "step": 501
+    },
+    {
+      "epoch": 0.9639942390782525,
+      "grad_norm": 0.6650898577066633,
+      "learning_rate": 9.7435520990621e-06,
+      "loss": 1.2945,
+      "step": 502
+    },
+    {
+      "epoch": 0.9659145463274124,
+      "grad_norm": 0.6813965501305161,
+      "learning_rate": 9.741429231768278e-06,
+      "loss": 1.3214,
+      "step": 503
+    },
+    {
+      "epoch": 0.9678348535765723,
+      "grad_norm": 0.8240692929170976,
+      "learning_rate": 9.739297847335387e-06,
+      "loss": 1.4367,
+      "step": 504
+    },
+    {
+      "epoch": 0.9697551608257321,
+      "grad_norm": 0.7664659172540482,
+      "learning_rate": 9.73715794959208e-06,
+      "loss": 1.2429,
+      "step": 505
+    },
+    {
+      "epoch": 0.971675468074892,
+      "grad_norm": 0.663273604561841,
+      "learning_rate": 9.735009542382308e-06,
+      "loss": 1.2678,
+      "step": 506
+    },
+    {
+      "epoch": 0.9735957753240518,
+      "grad_norm": 0.7286317666999398,
+      "learning_rate": 9.732852629565302e-06,
+      "loss": 1.351,
+      "step": 507
+    },
+    {
+      "epoch": 0.9755160825732118,
+      "grad_norm": 0.7222364516570275,
+      "learning_rate": 9.730687215015576e-06,
+      "loss": 1.3875,
+      "step": 508
+    },
+    {
+      "epoch": 0.9774363898223716,
+      "grad_norm": 0.792789498600007,
+      "learning_rate": 9.728513302622911e-06,
+      "loss": 1.4158,
+      "step": 509
+    },
+    {
+      "epoch": 0.9793566970715314,
+      "grad_norm": 0.6898048543889689,
+      "learning_rate": 9.72633089629236e-06,
+      "loss": 1.3018,
+      "step": 510
+    },
+    {
+      "epoch": 0.9812770043206913,
+      "grad_norm": 0.7080789810250435,
+      "learning_rate": 9.72413999994423e-06,
+      "loss": 1.2951,
+      "step": 511
+    },
+    {
+      "epoch": 0.9831973115698511,
+      "grad_norm": 0.6471793883594157,
+      "learning_rate": 9.721940617514076e-06,
+      "loss": 1.1768,
+      "step": 512
+    },
+    {
+      "epoch": 0.985117618819011,
+      "grad_norm": 0.6848848680839071,
+      "learning_rate": 9.719732752952702e-06,
+      "loss": 1.262,
+      "step": 513
+    },
+    {
+      "epoch": 0.9870379260681709,
+      "grad_norm": 0.7903965882462866,
+      "learning_rate": 9.717516410226144e-06,
+      "loss": 1.4717,
+      "step": 514
+    },
+    {
+      "epoch": 0.9889582333173308,
+      "grad_norm": 0.7404310033314039,
+      "learning_rate": 9.715291593315672e-06,
+      "loss": 1.3879,
+      "step": 515
+    },
+    {
+      "epoch": 0.9908785405664906,
+      "grad_norm": 0.735452133325044,
+      "learning_rate": 9.713058306217776e-06,
+      "loss": 1.3079,
+      "step": 516
+    },
+    {
+      "epoch": 0.9927988478156505,
+      "grad_norm": 0.8130352152653534,
+      "learning_rate": 9.710816552944157e-06,
+      "loss": 1.434,
+      "step": 517
+    },
+    {
+      "epoch": 0.9947191550648103,
+      "grad_norm": 0.7502971580652452,
+      "learning_rate": 9.708566337521736e-06,
+      "loss": 1.3013,
+      "step": 518
+    },
+    {
+      "epoch": 0.9966394623139703,
+      "grad_norm": 0.6582057806093718,
+      "learning_rate": 9.70630766399262e-06,
+      "loss": 1.2994,
+      "step": 519
+    },
+    {
+      "epoch": 0.9985597695631301,
+      "grad_norm": 0.7398007707770013,
+      "learning_rate": 9.70404053641412e-06,
+      "loss": 1.3135,
+      "step": 520
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7398007707770013,
+      "learning_rate": 9.701764958858729e-06,
+      "loss": 1.1265,
+      "step": 521
+    },
+    {
+      "epoch": 1.0019203072491598,
+      "grad_norm": 0.914948928730533,
+      "learning_rate": 9.69948093541412e-06,
+      "loss": 1.1908,
+      "step": 522
+    },
+    {
+      "epoch": 1.0038406144983196,
+      "grad_norm": 0.7588233450162211,
+      "learning_rate": 9.697188470183137e-06,
+      "loss": 1.2765,
+      "step": 523
+    },
+    {
+      "epoch": 1.0057609217474797,
+      "grad_norm": 0.6899163901372742,
+      "learning_rate": 9.694887567283786e-06,
+      "loss": 1.262,
+      "step": 524
+    },
+    {
+      "epoch": 1.0076812289966395,
+      "grad_norm": 0.7729265974237742,
+      "learning_rate": 9.692578230849237e-06,
+      "loss": 1.3718,
+      "step": 525
+    },
+    {
+      "epoch": 1.0096015362457993,
+      "grad_norm": 0.76124735458907,
+      "learning_rate": 9.690260465027802e-06,
+      "loss": 1.3091,
+      "step": 526
+    },
+    {
+      "epoch": 1.011521843494959,
+      "grad_norm": 0.8261719411142489,
+      "learning_rate": 9.687934273982934e-06,
+      "loss": 1.2777,
+      "step": 527
+    },
+    {
+      "epoch": 1.0134421507441191,
+      "grad_norm": 0.6830001963579748,
+      "learning_rate": 9.68559966189323e-06,
+      "loss": 1.2425,
+      "step": 528
+    },
+    {
+      "epoch": 1.015362457993279,
+      "grad_norm": 0.7028896372705892,
+      "learning_rate": 9.683256632952402e-06,
+      "loss": 1.3801,
+      "step": 529
+    },
+    {
+      "epoch": 1.0172827652424388,
+      "grad_norm": 0.8453182259497111,
+      "learning_rate": 9.680905191369293e-06,
+      "loss": 1.335,
+      "step": 530
+    },
+    {
+      "epoch": 1.0192030724915986,
+      "grad_norm": 0.7691091692417015,
+      "learning_rate": 9.678545341367846e-06,
+      "loss": 1.2368,
+      "step": 531
+    },
+    {
+      "epoch": 1.0211233797407586,
+      "grad_norm": 0.6781623931537163,
+      "learning_rate": 9.67617708718712e-06,
+      "loss": 1.2796,
+      "step": 532
+    },
+    {
+      "epoch": 1.0230436869899184,
+      "grad_norm": 0.7367258690005868,
+      "learning_rate": 9.67380043308126e-06,
+      "loss": 1.2936,
+      "step": 533
+    },
+    {
+      "epoch": 1.0249639942390782,
+      "grad_norm": 0.691600698798815,
+      "learning_rate": 9.671415383319507e-06,
+      "loss": 1.5315,
+      "step": 534
+    },
+    {
+      "epoch": 1.026884301488238,
+      "grad_norm": 0.7722544599357118,
+      "learning_rate": 9.66902194218618e-06,
+      "loss": 1.3265,
+      "step": 535
+    },
+    {
+      "epoch": 1.0288046087373979,
+      "grad_norm": 0.7907265691669353,
+      "learning_rate": 9.666620113980673e-06,
+      "loss": 1.5249,
+      "step": 536
+    },
+    {
+      "epoch": 1.030724915986558,
+      "grad_norm": 0.8296730708074218,
+      "learning_rate": 9.664209903017447e-06,
+      "loss": 1.36,
+      "step": 537
+    },
+    {
+      "epoch": 1.0326452232357177,
+      "grad_norm": 0.6882651452466539,
+      "learning_rate": 9.661791313626019e-06,
+      "loss": 1.1086,
+      "step": 538
+    },
+    {
+      "epoch": 1.0345655304848775,
+      "grad_norm": 0.7450250373498115,
+      "learning_rate": 9.659364350150955e-06,
+      "loss": 1.1367,
+      "step": 539
+    },
+    {
+      "epoch": 1.0364858377340374,
+      "grad_norm": 0.7120678514024298,
+      "learning_rate": 9.656929016951869e-06,
+      "loss": 1.1975,
+      "step": 540
+    },
+    {
+      "epoch": 1.0384061449831974,
+      "grad_norm": 0.771727566325184,
+      "learning_rate": 9.654485318403404e-06,
+      "loss": 1.1875,
+      "step": 541
+    },
+    {
+      "epoch": 1.0403264522323572,
+      "grad_norm": 0.6961127537941473,
+      "learning_rate": 9.652033258895233e-06,
+      "loss": 1.2588,
+      "step": 542
+    },
+    {
+      "epoch": 1.042246759481517,
+      "grad_norm": 0.7479292355314183,
+      "learning_rate": 9.649572842832048e-06,
+      "loss": 1.4162,
+      "step": 543
+    },
+    {
+      "epoch": 1.0441670667306768,
+      "grad_norm": 0.8317819644739988,
+      "learning_rate": 9.64710407463355e-06,
+      "loss": 1.2725,
+      "step": 544
+    },
+    {
+      "epoch": 1.0460873739798369,
+      "grad_norm": 0.7604838992261496,
+      "learning_rate": 9.644626958734447e-06,
+      "loss": 1.1814,
+      "step": 545
+    },
+    {
+      "epoch": 1.0480076812289967,
+      "grad_norm": 0.7495453422521224,
+      "learning_rate": 9.642141499584436e-06,
+      "loss": 1.2463,
+      "step": 546
+    },
+    {
+      "epoch": 1.0499279884781565,
+      "grad_norm": 0.7152271852476584,
+      "learning_rate": 9.639647701648206e-06,
+      "loss": 1.3762,
+      "step": 547
+    },
+    {
+      "epoch": 1.0518482957273163,
+      "grad_norm": 0.6919477745781931,
+      "learning_rate": 9.637145569405426e-06,
+      "loss": 1.2951,
+      "step": 548
+    },
+    {
+      "epoch": 1.0537686029764763,
+      "grad_norm": 0.5919052540740619,
+      "learning_rate": 9.634635107350731e-06,
+      "loss": 1.1825,
+      "step": 549
+    },
+    {
+      "epoch": 1.0556889102256362,
+      "grad_norm": 0.6527783671699664,
+      "learning_rate": 9.632116319993726e-06,
+      "loss": 1.4155,
+      "step": 550
+    },
+    {
+      "epoch": 1.057609217474796,
+      "grad_norm": 0.8062261860338987,
+      "learning_rate": 9.629589211858963e-06,
+      "loss": 1.2512,
+      "step": 551
+    },
+    {
+      "epoch": 1.0595295247239558,
+      "grad_norm": 0.8008581715050265,
+      "learning_rate": 9.627053787485944e-06,
+      "loss": 1.3264,
+      "step": 552
+    },
+    {
+      "epoch": 1.0614498319731156,
+      "grad_norm": 0.6973877309767292,
+      "learning_rate": 9.624510051429116e-06,
+      "loss": 1.4041,
+      "step": 553
+    },
+    {
+      "epoch": 1.0633701392222756,
+      "grad_norm": 0.8158929404093554,
+      "learning_rate": 9.621958008257848e-06,
+      "loss": 1.3745,
+      "step": 554
+    },
+    {
+      "epoch": 1.0652904464714354,
+      "grad_norm": 0.6809292868930115,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 1.1379,
+      "step": 555
+    },
+    {
+      "epoch": 1.0672107537205953,
+      "grad_norm": 0.7614958232951388,
+      "learning_rate": 9.616829018924083e-06,
+      "loss": 1.4064,
+      "step": 556
+    },
+    {
+      "epoch": 1.069131060969755,
+      "grad_norm": 0.7494950717531818,
+      "learning_rate": 9.614252081974908e-06,
+      "loss": 1.4378,
+      "step": 557
+    },
+    {
+      "epoch": 1.0710513682189151,
+      "grad_norm": 0.6009709835822509,
+      "learning_rate": 9.61166685633792e-06,
+      "loss": 1.306,
+      "step": 558
+    },
+    {
+      "epoch": 1.072971675468075,
+      "grad_norm": 0.7353167635763869,
+      "learning_rate": 9.609073346657021e-06,
+      "loss": 1.3607,
+      "step": 559
+    },
+    {
+      "epoch": 1.0748919827172347,
+      "grad_norm": 0.9417037901973762,
+      "learning_rate": 9.606471557590992e-06,
+      "loss": 1.3322,
+      "step": 560
+    },
+    {
+      "epoch": 1.0768122899663946,
+      "grad_norm": 0.6799455391858547,
+      "learning_rate": 9.603861493813486e-06,
+      "loss": 1.2997,
+      "step": 561
+    },
+    {
+      "epoch": 1.0787325972155546,
+      "grad_norm": 0.7062230147048892,
+      "learning_rate": 9.601243160013023e-06,
+      "loss": 1.2617,
+      "step": 562
+    },
+    {
+      "epoch": 1.0806529044647144,
+      "grad_norm": 0.7674724288187535,
+      "learning_rate": 9.598616560892977e-06,
+      "loss": 1.3903,
+      "step": 563
+    },
+    {
+      "epoch": 1.0825732117138742,
+      "grad_norm": 0.683909615499401,
+      "learning_rate": 9.595981701171564e-06,
+      "loss": 1.0946,
+      "step": 564
+    },
+    {
+      "epoch": 1.084493518963034,
+      "grad_norm": 0.7398808578213143,
+      "learning_rate": 9.593338585581848e-06,
+      "loss": 1.1715,
+      "step": 565
+    },
+    {
+      "epoch": 1.086413826212194,
+      "grad_norm": 0.7047962556591744,
+      "learning_rate": 9.59068721887172e-06,
+      "loss": 1.2756,
+      "step": 566
+    },
+    {
+      "epoch": 1.0883341334613539,
+      "grad_norm": 0.6315097473402969,
+      "learning_rate": 9.588027605803886e-06,
+      "loss": 1.2555,
+      "step": 567
+    },
+    {
+      "epoch": 1.0902544407105137,
+      "grad_norm": 0.7234467460996638,
+      "learning_rate": 9.585359751155874e-06,
+      "loss": 1.3359,
+      "step": 568
+    },
+    {
+      "epoch": 1.0921747479596735,
+      "grad_norm": 0.6076408944277064,
+      "learning_rate": 9.582683659720013e-06,
+      "loss": 1.2367,
+      "step": 569
+    },
+    {
+      "epoch": 1.0940950552088333,
+      "grad_norm": 0.6174139463787864,
+      "learning_rate": 9.579999336303427e-06,
+      "loss": 1.3522,
+      "step": 570
+    },
+    {
+      "epoch": 1.0960153624579934,
+      "grad_norm": 0.7480007037028495,
+      "learning_rate": 9.577306785728028e-06,
+      "loss": 1.2621,
+      "step": 571
+    },
+    {
+      "epoch": 1.0979356697071532,
+      "grad_norm": 0.7643115891903187,
+      "learning_rate": 9.574606012830509e-06,
+      "loss": 1.2505,
+      "step": 572
+    },
+    {
+      "epoch": 1.099855976956313,
+      "grad_norm": 0.6563832097104907,
+      "learning_rate": 9.571897022462329e-06,
+      "loss": 1.2947,
+      "step": 573
+    },
+    {
+      "epoch": 1.1017762842054728,
+      "grad_norm": 0.7178248398921613,
+      "learning_rate": 9.569179819489712e-06,
+      "loss": 1.3066,
+      "step": 574
+    },
+    {
+      "epoch": 1.1036965914546328,
+      "grad_norm": 0.7083396283957663,
+      "learning_rate": 9.566454408793628e-06,
+      "loss": 1.2919,
+      "step": 575
+    },
+    {
+      "epoch": 1.1056168987037926,
+      "grad_norm": 0.6599770996680595,
+      "learning_rate": 9.563720795269801e-06,
+      "loss": 1.3526,
+      "step": 576
+    },
+    {
+      "epoch": 1.1075372059529525,
+      "grad_norm": 0.708272003033926,
+      "learning_rate": 9.560978983828682e-06,
+      "loss": 1.243,
+      "step": 577
+    },
+    {
+      "epoch": 1.1094575132021123,
+      "grad_norm": 0.7309704918989066,
+      "learning_rate": 9.558228979395448e-06,
+      "loss": 1.164,
+      "step": 578
+    },
+    {
+      "epoch": 1.1113778204512723,
+      "grad_norm": 0.6947735355279264,
+      "learning_rate": 9.55547078691e-06,
+      "loss": 1.2518,
+      "step": 579
+    },
+    {
+      "epoch": 1.1132981277004321,
+      "grad_norm": 0.6205776923415179,
+      "learning_rate": 9.552704411326938e-06,
+      "loss": 1.264,
+      "step": 580
+    },
+    {
+      "epoch": 1.115218434949592,
+      "grad_norm": 0.6848751463529882,
+      "learning_rate": 9.549929857615571e-06,
+      "loss": 1.3857,
+      "step": 581
+    },
+    {
+      "epoch": 1.1171387421987518,
+      "grad_norm": 0.6873124432568535,
+      "learning_rate": 9.547147130759894e-06,
+      "loss": 1.2951,
+      "step": 582
+    },
+    {
+      "epoch": 1.1190590494479116,
+      "grad_norm": 0.6699975273598144,
+      "learning_rate": 9.54435623575858e-06,
+      "loss": 1.2551,
+      "step": 583
+    },
+    {
+      "epoch": 1.1209793566970716,
+      "grad_norm": 0.6779630282881215,
+      "learning_rate": 9.541557177624978e-06,
+      "loss": 1.3882,
+      "step": 584
+    },
+    {
+      "epoch": 1.1228996639462314,
+      "grad_norm": 0.6922291738327386,
+      "learning_rate": 9.538749961387106e-06,
+      "loss": 1.3379,
+      "step": 585
+    },
+    {
+      "epoch": 1.1248199711953912,
+      "grad_norm": 0.6743878536243224,
+      "learning_rate": 9.535934592087627e-06,
+      "loss": 1.243,
+      "step": 586
+    },
+    {
+      "epoch": 1.126740278444551,
+      "grad_norm": 0.7100328653945479,
+      "learning_rate": 9.533111074783857e-06,
+      "loss": 1.3178,
+      "step": 587
+    },
+    {
+      "epoch": 1.128660585693711,
+      "grad_norm": 0.604519259408001,
+      "learning_rate": 9.530279414547743e-06,
+      "loss": 1.3051,
+      "step": 588
+    },
+    {
+      "epoch": 1.130580892942871,
+      "grad_norm": 0.6526520665140377,
+      "learning_rate": 9.527439616465864e-06,
+      "loss": 1.4662,
+      "step": 589
+    },
+    {
+      "epoch": 1.1325012001920307,
+      "grad_norm": 0.6906072243725673,
+      "learning_rate": 9.524591685639414e-06,
+      "loss": 1.3682,
+      "step": 590
+    },
+    {
+      "epoch": 1.1344215074411905,
+      "grad_norm": 0.677843087192056,
+      "learning_rate": 9.521735627184197e-06,
+      "loss": 1.2735,
+      "step": 591
+    },
+    {
+      "epoch": 1.1363418146903506,
+      "grad_norm": 0.6367565663261762,
+      "learning_rate": 9.518871446230616e-06,
+      "loss": 1.3467,
+      "step": 592
+    },
+    {
+      "epoch": 1.1382621219395104,
+      "grad_norm": 0.8296312537408668,
+      "learning_rate": 9.515999147923666e-06,
+      "loss": 1.3932,
+      "step": 593
+    },
+    {
+      "epoch": 1.1401824291886702,
+      "grad_norm": 0.6513452900246463,
+      "learning_rate": 9.513118737422926e-06,
+      "loss": 1.2055,
+      "step": 594
+    },
+    {
+      "epoch": 1.14210273643783,
+      "grad_norm": 0.6216514451209604,
+      "learning_rate": 9.51023021990254e-06,
+      "loss": 1.2417,
+      "step": 595
+    },
+    {
+      "epoch": 1.1440230436869898,
+      "grad_norm": 0.7531692439045267,
+      "learning_rate": 9.50733360055122e-06,
+      "loss": 1.2386,
+      "step": 596
+    },
+    {
+      "epoch": 1.1459433509361499,
+      "grad_norm": 0.683513142501149,
+      "learning_rate": 9.50442888457223e-06,
+      "loss": 1.2326,
+      "step": 597
+    },
+    {
+      "epoch": 1.1478636581853097,
+      "grad_norm": 0.6594339187726981,
+      "learning_rate": 9.501516077183381e-06,
+      "loss": 1.2825,
+      "step": 598
+    },
+    {
+      "epoch": 1.1497839654344695,
+      "grad_norm": 0.7622581730719366,
+      "learning_rate": 9.498595183617014e-06,
+      "loss": 1.2091,
+      "step": 599
+    },
+    {
+      "epoch": 1.1517042726836293,
+      "grad_norm": 0.7667615881022617,
+      "learning_rate": 9.495666209119998e-06,
+      "loss": 1.4263,
+      "step": 600
+    },
+    {
+      "epoch": 1.1536245799327893,
+      "grad_norm": 0.6614445792795562,
+      "learning_rate": 9.492729158953717e-06,
+      "loss": 1.3555,
+      "step": 601
+    },
+    {
+      "epoch": 1.1555448871819491,
+      "grad_norm": 0.6775001666189906,
+      "learning_rate": 9.489784038394065e-06,
+      "loss": 1.2491,
+      "step": 602
+    },
+    {
+      "epoch": 1.157465194431109,
+      "grad_norm": 0.7044086225702691,
+      "learning_rate": 9.486830852731428e-06,
+      "loss": 1.2622,
+      "step": 603
+    },
+    {
+      "epoch": 1.1593855016802688,
+      "grad_norm": 0.7464552469199642,
+      "learning_rate": 9.48386960727068e-06,
+      "loss": 1.3274,
+      "step": 604
+    },
+    {
+      "epoch": 1.1613058089294288,
+      "grad_norm": 0.832385639864367,
+      "learning_rate": 9.48090030733118e-06,
+      "loss": 1.2453,
+      "step": 605
+    },
+    {
+      "epoch": 1.1632261161785886,
+      "grad_norm": 0.6252343436809756,
+      "learning_rate": 9.477922958246747e-06,
+      "loss": 1.1933,
+      "step": 606
+    },
+    {
+      "epoch": 1.1651464234277484,
+      "grad_norm": 0.7355703889077223,
+      "learning_rate": 9.47493756536566e-06,
+      "loss": 1.306,
+      "step": 607
+    },
+    {
+      "epoch": 1.1670667306769082,
+      "grad_norm": 0.7273603661247848,
+      "learning_rate": 9.471944134050652e-06,
+      "loss": 1.3707,
+      "step": 608
+    },
+    {
+      "epoch": 1.168987037926068,
+      "grad_norm": 0.7562042075996765,
+      "learning_rate": 9.468942669678893e-06,
+      "loss": 1.3688,
+      "step": 609
+    },
+    {
+      "epoch": 1.170907345175228,
+      "grad_norm": 0.802848759502404,
+      "learning_rate": 9.465933177641981e-06,
+      "loss": 1.2193,
+      "step": 610
+    },
+    {
+      "epoch": 1.172827652424388,
+      "grad_norm": 0.7609509108594381,
+      "learning_rate": 9.46291566334594e-06,
+      "loss": 1.3736,
+      "step": 611
+    },
+    {
+      "epoch": 1.1747479596735477,
+      "grad_norm": 0.8308878738963801,
+      "learning_rate": 9.459890132211198e-06,
+      "loss": 1.3762,
+      "step": 612
+    },
+    {
+      "epoch": 1.1766682669227075,
+      "grad_norm": 0.6532589559736274,
+      "learning_rate": 9.456856589672587e-06,
+      "loss": 1.3125,
+      "step": 613
+    },
+    {
+      "epoch": 1.1785885741718676,
+      "grad_norm": 0.7153467048815468,
+      "learning_rate": 9.453815041179329e-06,
+      "loss": 1.4605,
+      "step": 614
+    },
+    {
+      "epoch": 1.1805088814210274,
+      "grad_norm": 0.8574229555342932,
+      "learning_rate": 9.450765492195027e-06,
+      "loss": 1.2833,
+      "step": 615
+    },
+    {
+      "epoch": 1.1824291886701872,
+      "grad_norm": 0.7052192727678307,
+      "learning_rate": 9.44770794819766e-06,
+      "loss": 1.426,
+      "step": 616
+    },
+    {
+      "epoch": 1.184349495919347,
+      "grad_norm": 0.8135394020685827,
+      "learning_rate": 9.444642414679563e-06,
+      "loss": 1.3768,
+      "step": 617
+    },
+    {
+      "epoch": 1.186269803168507,
+      "grad_norm": 0.6225214032963996,
+      "learning_rate": 9.441568897147423e-06,
+      "loss": 1.3118,
+      "step": 618
+    },
+    {
+      "epoch": 1.1881901104176669,
+      "grad_norm": 0.72856086645004,
+      "learning_rate": 9.438487401122277e-06,
+      "loss": 1.2549,
+      "step": 619
+    },
+    {
+      "epoch": 1.1901104176668267,
+      "grad_norm": 0.7806208477609391,
+      "learning_rate": 9.435397932139478e-06,
+      "loss": 1.2322,
+      "step": 620
+    },
+    {
+      "epoch": 1.1920307249159865,
+      "grad_norm": 0.6519805710347557,
+      "learning_rate": 9.432300495748719e-06,
+      "loss": 1.3982,
+      "step": 621
+    },
+    {
+      "epoch": 1.1939510321651463,
+      "grad_norm": 0.7771841737007339,
+      "learning_rate": 9.429195097513993e-06,
+      "loss": 1.3656,
+      "step": 622
+    },
+    {
+      "epoch": 1.1958713394143063,
+      "grad_norm": 0.7437011183267965,
+      "learning_rate": 9.426081743013599e-06,
+      "loss": 1.222,
+      "step": 623
+    },
+    {
+      "epoch": 1.1977916466634662,
+      "grad_norm": 0.5875165599100549,
+      "learning_rate": 9.422960437840128e-06,
+      "loss": 1.1133,
+      "step": 624
+    },
+    {
+      "epoch": 1.199711953912626,
+      "grad_norm": 0.7624140053435047,
+      "learning_rate": 9.419831187600453e-06,
+      "loss": 1.3283,
+      "step": 625
+    },
+    {
+      "epoch": 1.201632261161786,
+      "grad_norm": 0.7653770887737222,
+      "learning_rate": 9.416693997915717e-06,
+      "loss": 1.1884,
+      "step": 626
+    },
+    {
+      "epoch": 1.2035525684109458,
+      "grad_norm": 0.6152692591706712,
+      "learning_rate": 9.41354887442133e-06,
+      "loss": 1.2304,
+      "step": 627
+    },
+    {
+      "epoch": 1.2054728756601056,
+      "grad_norm": 0.8185718374516386,
+      "learning_rate": 9.410395822766946e-06,
+      "loss": 1.4138,
+      "step": 628
+    },
+    {
+      "epoch": 1.2073931829092654,
+      "grad_norm": 0.7082210656201997,
+      "learning_rate": 9.407234848616467e-06,
+      "loss": 1.2941,
+      "step": 629
+    },
+    {
+      "epoch": 1.2093134901584253,
+      "grad_norm": 0.6232135092488823,
+      "learning_rate": 9.404065957648023e-06,
+      "loss": 1.2503,
+      "step": 630
+    },
+    {
+      "epoch": 1.2112337974075853,
+      "grad_norm": 0.654358423334644,
+      "learning_rate": 9.400889155553969e-06,
+      "loss": 1.2953,
+      "step": 631
+    },
+    {
+      "epoch": 1.2131541046567451,
+      "grad_norm": 0.8666882012026393,
+      "learning_rate": 9.397704448040865e-06,
+      "loss": 1.2951,
+      "step": 632
+    },
+    {
+      "epoch": 1.215074411905905,
+      "grad_norm": 0.799703690690942,
+      "learning_rate": 9.394511840829474e-06,
+      "loss": 1.1688,
+      "step": 633
+    },
+    {
+      "epoch": 1.2169947191550647,
+      "grad_norm": 0.5744242469127047,
+      "learning_rate": 9.391311339654755e-06,
+      "loss": 1.361,
+      "step": 634
+    },
+    {
+      "epoch": 1.2189150264042246,
+      "grad_norm": 0.6674313237003483,
+      "learning_rate": 9.388102950265836e-06,
+      "loss": 1.1743,
+      "step": 635
+    },
+    {
+      "epoch": 1.2208353336533846,
+      "grad_norm": 0.7039450941038129,
+      "learning_rate": 9.384886678426027e-06,
+      "loss": 1.3431,
+      "step": 636
+    },
+    {
+      "epoch": 1.2227556409025444,
+      "grad_norm": 0.7752002897514472,
+      "learning_rate": 9.381662529912787e-06,
+      "loss": 1.3558,
+      "step": 637
+    },
+    {
+      "epoch": 1.2246759481517042,
+      "grad_norm": 0.7005182546751737,
+      "learning_rate": 9.378430510517732e-06,
+      "loss": 1.4033,
+      "step": 638
+    },
+    {
+      "epoch": 1.2265962554008643,
+      "grad_norm": 0.7688364382975947,
+      "learning_rate": 9.37519062604661e-06,
+      "loss": 1.3613,
+      "step": 639
+    },
+    {
+      "epoch": 1.228516562650024,
+      "grad_norm": 0.7239757068116605,
+      "learning_rate": 9.371942882319306e-06,
+      "loss": 1.3262,
+      "step": 640
+    },
+    {
+      "epoch": 1.2304368698991839,
+      "grad_norm": 0.6873674276530499,
+      "learning_rate": 9.36868728516981e-06,
+      "loss": 1.1821,
+      "step": 641
+    },
+    {
+      "epoch": 1.2323571771483437,
+      "grad_norm": 0.6742489136796876,
+      "learning_rate": 9.36542384044623e-06,
+      "loss": 1.1827,
+      "step": 642
+    },
+    {
+      "epoch": 1.2342774843975035,
+      "grad_norm": 0.76768397108436,
+      "learning_rate": 9.36215255401077e-06,
+      "loss": 1.2645,
+      "step": 643
+    },
+    {
+      "epoch": 1.2361977916466635,
+      "grad_norm": 0.7270486835964114,
+      "learning_rate": 9.358873431739712e-06,
+      "loss": 1.2607,
+      "step": 644
+    },
+    {
+      "epoch": 1.2381180988958234,
+      "grad_norm": 0.8519540490456053,
+      "learning_rate": 9.355586479523424e-06,
+      "loss": 1.3387,
+      "step": 645
+    },
+    {
+      "epoch": 1.2400384061449832,
+      "grad_norm": 0.6797527920764758,
+      "learning_rate": 9.352291703266332e-06,
+      "loss": 1.3279,
+      "step": 646
+    },
+    {
+      "epoch": 1.241958713394143,
+      "grad_norm": 0.8368086299915656,
+      "learning_rate": 9.348989108886919e-06,
+      "loss": 1.3217,
+      "step": 647
+    },
+    {
+      "epoch": 1.243879020643303,
+      "grad_norm": 0.8539520255198279,
+      "learning_rate": 9.345678702317711e-06,
+      "loss": 1.3196,
+      "step": 648
+    },
+    {
+      "epoch": 1.2457993278924628,
+      "grad_norm": 0.6860987908944224,
+      "learning_rate": 9.342360489505271e-06,
+      "loss": 1.4192,
+      "step": 649
+    },
+    {
+      "epoch": 1.2477196351416227,
+      "grad_norm": 0.8205531903946733,
+      "learning_rate": 9.339034476410177e-06,
+      "loss": 1.2021,
+      "step": 650
+    },
+    {
+      "epoch": 1.2496399423907825,
+      "grad_norm": 0.9081433696653918,
+      "learning_rate": 9.335700669007028e-06,
+      "loss": 1.2448,
+      "step": 651
+    },
+    {
+      "epoch": 1.2515602496399425,
+      "grad_norm": 0.7071653317141182,
+      "learning_rate": 9.332359073284417e-06,
+      "loss": 1.2858,
+      "step": 652
+    },
+    {
+      "epoch": 1.2534805568891023,
+      "grad_norm": 0.7411138132312242,
+      "learning_rate": 9.329009695244929e-06,
+      "loss": 1.2792,
+      "step": 653
+    },
+    {
+      "epoch": 1.2554008641382621,
+      "grad_norm": 0.8302739758280621,
+      "learning_rate": 9.32565254090513e-06,
+      "loss": 1.2798,
+      "step": 654
+    },
+    {
+      "epoch": 1.257321171387422,
+      "grad_norm": 0.7748066850549846,
+      "learning_rate": 9.322287616295553e-06,
+      "loss": 1.372,
+      "step": 655
+    },
+    {
+      "epoch": 1.2592414786365818,
+      "grad_norm": 0.7053895497064686,
+      "learning_rate": 9.318914927460694e-06,
+      "loss": 1.3919,
+      "step": 656
+    },
+    {
+      "epoch": 1.2611617858857418,
+      "grad_norm": 0.7970812172979578,
+      "learning_rate": 9.315534480458986e-06,
+      "loss": 1.3171,
+      "step": 657
+    },
+    {
+      "epoch": 1.2630820931349016,
+      "grad_norm": 0.6944180261511493,
+      "learning_rate": 9.312146281362811e-06,
+      "loss": 1.0936,
+      "step": 658
+    },
+    {
+      "epoch": 1.2650024003840614,
+      "grad_norm": 0.6378844061470125,
+      "learning_rate": 9.308750336258463e-06,
+      "loss": 1.3324,
+      "step": 659
+    },
+    {
+      "epoch": 1.2669227076332212,
+      "grad_norm": 0.8429774002872888,
+      "learning_rate": 9.30534665124616e-06,
+      "loss": 1.2359,
+      "step": 660
+    },
+    {
+      "epoch": 1.268843014882381,
+      "grad_norm": 0.8204171557741506,
+      "learning_rate": 9.30193523244002e-06,
+      "loss": 1.3692,
+      "step": 661
+    },
+    {
+      "epoch": 1.270763322131541,
+      "grad_norm": 0.8025145383476096,
+      "learning_rate": 9.298516085968052e-06,
+      "loss": 1.1787,
+      "step": 662
+    },
+    {
+      "epoch": 1.272683629380701,
+      "grad_norm": 0.7124288281643562,
+      "learning_rate": 9.295089217972152e-06,
+      "loss": 1.2906,
+      "step": 663
+    },
+    {
+      "epoch": 1.2746039366298607,
+      "grad_norm": 0.7525796363802068,
+      "learning_rate": 9.291654634608079e-06,
+      "loss": 1.297,
+      "step": 664
+    },
+    {
+      "epoch": 1.2765242438790207,
+      "grad_norm": 1.067363938892021,
+      "learning_rate": 9.288212342045453e-06,
+      "loss": 1.3022,
+      "step": 665
+    },
+    {
+      "epoch": 1.2784445511281806,
+      "grad_norm": 0.6619389553270122,
+      "learning_rate": 9.284762346467749e-06,
+      "loss": 1.2192,
+      "step": 666
+    },
+    {
+      "epoch": 1.2803648583773404,
+      "grad_norm": 0.7985420894282017,
+      "learning_rate": 9.281304654072267e-06,
+      "loss": 1.2289,
+      "step": 667
+    },
+    {
+      "epoch": 1.2822851656265002,
+      "grad_norm": 0.6623460682542177,
+      "learning_rate": 9.277839271070146e-06,
+      "loss": 1.1139,
+      "step": 668
+    },
+    {
+      "epoch": 1.28420547287566,
+      "grad_norm": 0.702752052754151,
+      "learning_rate": 9.274366203686332e-06,
+      "loss": 1.1201,
+      "step": 669
+    },
+    {
+      "epoch": 1.28612578012482,
+      "grad_norm": 0.6289400458286583,
+      "learning_rate": 9.270885458159576e-06,
+      "loss": 1.2891,
+      "step": 670
+    },
+    {
+      "epoch": 1.2880460873739799,
+      "grad_norm": 0.8427578486969821,
+      "learning_rate": 9.267397040742419e-06,
+      "loss": 1.3096,
+      "step": 671
+    },
+    {
+      "epoch": 1.2899663946231397,
+      "grad_norm": 0.6682673731700085,
+      "learning_rate": 9.263900957701191e-06,
+      "loss": 1.3436,
+      "step": 672
+    },
+    {
+      "epoch": 1.2918867018722997,
+      "grad_norm": 0.5915568132233988,
+      "learning_rate": 9.260397215315982e-06,
+      "loss": 1.177,
+      "step": 673
+    },
+    {
+      "epoch": 1.2938070091214593,
+      "grad_norm": 0.845875785609105,
+      "learning_rate": 9.25688581988065e-06,
+      "loss": 1.3084,
+      "step": 674
+    },
+    {
+      "epoch": 1.2957273163706193,
+      "grad_norm": 0.7518298927765613,
+      "learning_rate": 9.253366777702793e-06,
+      "loss": 1.3525,
+      "step": 675
+    },
+    {
+      "epoch": 1.2976476236197791,
+      "grad_norm": 0.6317348605399149,
+      "learning_rate": 9.249840095103748e-06,
+      "loss": 1.4346,
+      "step": 676
+    },
+    {
+      "epoch": 1.299567930868939,
+      "grad_norm": 0.7065670849481425,
+      "learning_rate": 9.246305778418578e-06,
+      "loss": 1.2933,
+      "step": 677
+    },
+    {
+      "epoch": 1.301488238118099,
+      "grad_norm": 0.6836131464449633,
+      "learning_rate": 9.242763833996058e-06,
+      "loss": 1.2998,
+      "step": 678
+    },
+    {
+      "epoch": 1.3034085453672588,
+      "grad_norm": 0.6170451881755116,
+      "learning_rate": 9.239214268198662e-06,
+      "loss": 1.3884,
+      "step": 679
+    },
+    {
+      "epoch": 1.3053288526164186,
+      "grad_norm": 0.8701684854560547,
+      "learning_rate": 9.235657087402561e-06,
+      "loss": 1.2995,
+      "step": 680
+    },
+    {
+      "epoch": 1.3072491598655784,
+      "grad_norm": 0.695844689357289,
+      "learning_rate": 9.232092297997598e-06,
+      "loss": 1.2732,
+      "step": 681
+    },
+    {
+      "epoch": 1.3091694671147383,
+      "grad_norm": 0.7103405815755808,
+      "learning_rate": 9.228519906387287e-06,
+      "loss": 1.3625,
+      "step": 682
+    },
+    {
+      "epoch": 1.3110897743638983,
+      "grad_norm": 0.7065126258973163,
+      "learning_rate": 9.2249399189888e-06,
+      "loss": 1.3358,
+      "step": 683
+    },
+    {
+      "epoch": 1.313010081613058,
+      "grad_norm": 0.703504561635419,
+      "learning_rate": 9.22135234223295e-06,
+      "loss": 1.3198,
+      "step": 684
+    },
+    {
+      "epoch": 1.314930388862218,
+      "grad_norm": 0.6758713111088065,
+      "learning_rate": 9.217757182564185e-06,
+      "loss": 1.2968,
+      "step": 685
+    },
+    {
+      "epoch": 1.316850696111378,
+      "grad_norm": 0.6586940502760603,
+      "learning_rate": 9.214154446440571e-06,
+      "loss": 1.3567,
+      "step": 686
+    },
+    {
+      "epoch": 1.3187710033605378,
+      "grad_norm": 0.6916983456158645,
+      "learning_rate": 9.210544140333787e-06,
+      "loss": 1.2495,
+      "step": 687
+    },
+    {
+      "epoch": 1.3206913106096976,
+      "grad_norm": 0.6687033750412827,
+      "learning_rate": 9.206926270729112e-06,
+      "loss": 1.3326,
+      "step": 688
+    },
+    {
+      "epoch": 1.3226116178588574,
+      "grad_norm": 0.6514229288911735,
+      "learning_rate": 9.203300844125407e-06,
+      "loss": 1.3184,
+      "step": 689
+    },
+    {
+      "epoch": 1.3245319251080172,
+      "grad_norm": 0.7367890594020262,
+      "learning_rate": 9.199667867035111e-06,
+      "loss": 1.1066,
+      "step": 690
+    },
+    {
+      "epoch": 1.3264522323571772,
+      "grad_norm": 0.6791993316929129,
+      "learning_rate": 9.196027345984223e-06,
+      "loss": 1.2166,
+      "step": 691
+    },
+    {
+      "epoch": 1.328372539606337,
+      "grad_norm": 0.6186559224575634,
+      "learning_rate": 9.192379287512294e-06,
+      "loss": 1.3028,
+      "step": 692
+    },
+    {
+      "epoch": 1.3302928468554969,
+      "grad_norm": 0.7034503705343318,
+      "learning_rate": 9.188723698172421e-06,
+      "loss": 1.5347,
+      "step": 693
+    },
+    {
+      "epoch": 1.3322131541046567,
+      "grad_norm": 0.6229250085704403,
+      "learning_rate": 9.185060584531218e-06,
+      "loss": 1.4723,
+      "step": 694
+    },
+    {
+      "epoch": 1.3341334613538165,
+      "grad_norm": 0.7534919743366831,
+      "learning_rate": 9.181389953168825e-06,
+      "loss": 1.1994,
+      "step": 695
+    },
+    {
+      "epoch": 1.3360537686029765,
+      "grad_norm": 0.7107991436376052,
+      "learning_rate": 9.17771181067888e-06,
+      "loss": 1.2875,
+      "step": 696
+    },
+    {
+      "epoch": 1.3379740758521363,
+      "grad_norm": 0.6090769881805889,
+      "learning_rate": 9.174026163668516e-06,
+      "loss": 1.1811,
+      "step": 697
+    },
+    {
+      "epoch": 1.3398943831012962,
+      "grad_norm": 0.5976408741436412,
+      "learning_rate": 9.170333018758345e-06,
+      "loss": 1.2105,
+      "step": 698
+    },
+    {
+      "epoch": 1.3418146903504562,
+      "grad_norm": 0.7155288651414388,
+      "learning_rate": 9.166632382582452e-06,
+      "loss": 1.2097,
+      "step": 699
+    },
+    {
+      "epoch": 1.343734997599616,
+      "grad_norm": 0.6340468084444298,
+      "learning_rate": 9.162924261788372e-06,
+      "loss": 1.2756,
+      "step": 700
+    },
+    {
+      "epoch": 1.3456553048487758,
+      "grad_norm": 0.62226057586666,
+      "learning_rate": 9.159208663037088e-06,
+      "loss": 1.3445,
+      "step": 701
+    },
+    {
+      "epoch": 1.3475756120979356,
+      "grad_norm": 0.75460068826842,
+      "learning_rate": 9.15548559300302e-06,
+      "loss": 1.4983,
+      "step": 702
+    },
+    {
+      "epoch": 1.3494959193470955,
+      "grad_norm": 0.6610830578070283,
+      "learning_rate": 9.151755058374e-06,
+      "loss": 1.4167,
+      "step": 703
+    },
+    {
+      "epoch": 1.3514162265962555,
+      "grad_norm": 1.0148103056939377,
+      "learning_rate": 9.148017065851276e-06,
+      "loss": 1.152,
+      "step": 704
+    },
+    {
+      "epoch": 1.3533365338454153,
+      "grad_norm": 0.7157508626391539,
+      "learning_rate": 9.144271622149492e-06,
+      "loss": 1.1954,
+      "step": 705
+    },
+    {
+      "epoch": 1.3552568410945751,
+      "grad_norm": 0.6035040897309075,
+      "learning_rate": 9.140518733996672e-06,
+      "loss": 1.2952,
+      "step": 706
+    },
+    {
+      "epoch": 1.357177148343735,
+      "grad_norm": 0.7909130402584547,
+      "learning_rate": 9.136758408134219e-06,
+      "loss": 1.2898,
+      "step": 707
+    },
+    {
+      "epoch": 1.3590974555928947,
+      "grad_norm": 0.7293448393510418,
+      "learning_rate": 9.13299065131689e-06,
+      "loss": 1.2905,
+      "step": 708
+    },
+    {
+      "epoch": 1.3610177628420548,
+      "grad_norm": 0.630031286073638,
+      "learning_rate": 9.129215470312798e-06,
+      "loss": 1.3821,
+      "step": 709
+    },
+    {
+      "epoch": 1.3629380700912146,
+      "grad_norm": 0.6260025482348084,
+      "learning_rate": 9.125432871903383e-06,
+      "loss": 1.2427,
+      "step": 710
+    },
+    {
+      "epoch": 1.3648583773403744,
+      "grad_norm": 0.6759927808317432,
+      "learning_rate": 9.121642862883418e-06,
+      "loss": 1.3635,
+      "step": 711
+    },
+    {
+      "epoch": 1.3667786845895344,
+      "grad_norm": 0.6875922592092826,
+      "learning_rate": 9.117845450060983e-06,
+      "loss": 1.3153,
+      "step": 712
+    },
+    {
+      "epoch": 1.3686989918386943,
+      "grad_norm": 0.6975079984965866,
+      "learning_rate": 9.114040640257457e-06,
+      "loss": 1.3284,
+      "step": 713
+    },
+    {
+      "epoch": 1.370619299087854,
+      "grad_norm": 0.6542547355789712,
+      "learning_rate": 9.11022844030751e-06,
+      "loss": 1.268,
+      "step": 714
+    },
+    {
+      "epoch": 1.3725396063370139,
+      "grad_norm": 0.7131124321593435,
+      "learning_rate": 9.106408857059083e-06,
+      "loss": 1.3038,
+      "step": 715
+    },
+    {
+      "epoch": 1.3744599135861737,
+      "grad_norm": 0.6621678872162269,
+      "learning_rate": 9.102581897373385e-06,
+      "loss": 1.3846,
+      "step": 716
+    },
+    {
+      "epoch": 1.3763802208353337,
+      "grad_norm": 0.6528497631975559,
+      "learning_rate": 9.098747568124867e-06,
+      "loss": 1.391,
+      "step": 717
+    },
+    {
+      "epoch": 1.3783005280844935,
+      "grad_norm": 0.7950540458654306,
+      "learning_rate": 9.09490587620123e-06,
+      "loss": 1.2342,
+      "step": 718
+    },
+    {
+      "epoch": 1.3802208353336534,
+      "grad_norm": 0.6661090729260767,
+      "learning_rate": 9.09105682850339e-06,
+      "loss": 1.1368,
+      "step": 719
+    },
+    {
+      "epoch": 1.3821411425828132,
+      "grad_norm": 0.6896287298603911,
+      "learning_rate": 9.08720043194548e-06,
+      "loss": 1.2482,
+      "step": 720
+    },
+    {
+      "epoch": 1.384061449831973,
+      "grad_norm": 0.6469933404131487,
+      "learning_rate": 9.083336693454835e-06,
+      "loss": 1.2464,
+      "step": 721
+    },
+    {
+      "epoch": 1.385981757081133,
+      "grad_norm": 0.6564169816340895,
+      "learning_rate": 9.079465619971979e-06,
+      "loss": 1.3078,
+      "step": 722
+    },
+    {
+      "epoch": 1.3879020643302928,
+      "grad_norm": 0.6191833636580572,
+      "learning_rate": 9.075587218450611e-06,
+      "loss": 1.1747,
+      "step": 723
+    },
+    {
+      "epoch": 1.3898223715794527,
+      "grad_norm": 0.6183363215998144,
+      "learning_rate": 9.071701495857593e-06,
+      "loss": 1.3765,
+      "step": 724
+    },
+    {
+      "epoch": 1.3917426788286127,
+      "grad_norm": 0.7166010853381539,
+      "learning_rate": 9.067808459172935e-06,
+      "loss": 1.1739,
+      "step": 725
+    },
+    {
+      "epoch": 1.3936629860777725,
+      "grad_norm": 0.5814057362039444,
+      "learning_rate": 9.063908115389794e-06,
+      "loss": 1.2789,
+      "step": 726
+    },
+    {
+      "epoch": 1.3955832933269323,
+      "grad_norm": 0.7121122914175096,
+      "learning_rate": 9.060000471514447e-06,
+      "loss": 1.4757,
+      "step": 727
+    },
+    {
+      "epoch": 1.3975036005760921,
+      "grad_norm": 0.705583390032838,
+      "learning_rate": 9.056085534566283e-06,
+      "loss": 1.1249,
+      "step": 728
+    },
+    {
+      "epoch": 1.399423907825252,
+      "grad_norm": 0.6107217100732334,
+      "learning_rate": 9.052163311577795e-06,
+      "loss": 1.2457,
+      "step": 729
+    },
+    {
+      "epoch": 1.401344215074412,
+      "grad_norm": 0.6570510132698529,
+      "learning_rate": 9.048233809594561e-06,
+      "loss": 1.3098,
+      "step": 730
+    },
+    {
+      "epoch": 1.4032645223235718,
+      "grad_norm": 0.6242960521987202,
+      "learning_rate": 9.04429703567524e-06,
+      "loss": 1.3533,
+      "step": 731
+    },
+    {
+      "epoch": 1.4051848295727316,
+      "grad_norm": 0.699339430879501,
+      "learning_rate": 9.040352996891549e-06,
+      "loss": 1.5123,
+      "step": 732
+    },
+    {
+      "epoch": 1.4071051368218914,
+      "grad_norm": 0.8344946761901371,
+      "learning_rate": 9.036401700328255e-06,
+      "loss": 1.3611,
+      "step": 733
+    },
+    {
+      "epoch": 1.4090254440710512,
+      "grad_norm": 0.5789608462391438,
+      "learning_rate": 9.032443153083163e-06,
+      "loss": 1.2094,
+      "step": 734
+    },
+    {
+      "epoch": 1.4109457513202113,
+      "grad_norm": 0.7135604127502478,
+      "learning_rate": 9.028477362267103e-06,
+      "loss": 1.3675,
+      "step": 735
+    },
+    {
+      "epoch": 1.412866058569371,
+      "grad_norm": 0.657978923325545,
+      "learning_rate": 9.024504335003918e-06,
+      "loss": 1.1803,
+      "step": 736
+    },
+    {
+      "epoch": 1.414786365818531,
+      "grad_norm": 0.6166786992689153,
+      "learning_rate": 9.02052407843045e-06,
+      "loss": 1.3075,
+      "step": 737
+    },
+    {
+      "epoch": 1.416706673067691,
+      "grad_norm": 0.6363515202021924,
+      "learning_rate": 9.016536599696524e-06,
+      "loss": 1.4076,
+      "step": 738
+    },
+    {
+      "epoch": 1.4186269803168507,
+      "grad_norm": 0.7118481237155734,
+      "learning_rate": 9.01254190596494e-06,
+      "loss": 1.3571,
+      "step": 739
+    },
+    {
+      "epoch": 1.4205472875660106,
+      "grad_norm": 0.6600459827422428,
+      "learning_rate": 9.00854000441146e-06,
+      "loss": 1.1144,
+      "step": 740
+    },
+    {
+      "epoch": 1.4224675948151704,
+      "grad_norm": 0.6239591484123759,
+      "learning_rate": 9.004530902224793e-06,
+      "loss": 1.3528,
+      "step": 741
+    },
+    {
+      "epoch": 1.4243879020643302,
+      "grad_norm": 0.7100163193828547,
+      "learning_rate": 9.00051460660658e-06,
+      "loss": 1.31,
+      "step": 742
+    },
+    {
+      "epoch": 1.4263082093134902,
+      "grad_norm": 0.7325087503328771,
+      "learning_rate": 8.996491124771387e-06,
+      "loss": 1.306,
+      "step": 743
+    },
+    {
+      "epoch": 1.42822851656265,
+      "grad_norm": 0.730452846870962,
+      "learning_rate": 8.992460463946689e-06,
+      "loss": 1.2024,
+      "step": 744
+    },
+    {
+      "epoch": 1.4301488238118099,
+      "grad_norm": 0.7161540737322926,
+      "learning_rate": 8.988422631372854e-06,
+      "loss": 1.1938,
+      "step": 745
+    },
+    {
+      "epoch": 1.43206913106097,
+      "grad_norm": 0.6561434676104188,
+      "learning_rate": 8.98437763430313e-06,
+      "loss": 1.1944,
+      "step": 746
+    },
+    {
+      "epoch": 1.4339894383101295,
+      "grad_norm": 0.6297905970632676,
+      "learning_rate": 8.980325480003647e-06,
+      "loss": 1.3503,
+      "step": 747
+    },
+    {
+      "epoch": 1.4359097455592895,
+      "grad_norm": 0.7625277635565622,
+      "learning_rate": 8.976266175753376e-06,
+      "loss": 1.34,
+      "step": 748
+    },
+    {
+      "epoch": 1.4378300528084493,
+      "grad_norm": 0.6672733475659698,
+      "learning_rate": 8.972199728844144e-06,
+      "loss": 1.1897,
+      "step": 749
+    },
+    {
+      "epoch": 1.4397503600576091,
+      "grad_norm": 0.7199838468969945,
+      "learning_rate": 8.968126146580602e-06,
+      "loss": 1.3659,
+      "step": 750
+    },
+    {
+      "epoch": 1.4416706673067692,
+      "grad_norm": 0.6953976515544105,
+      "learning_rate": 8.96404543628022e-06,
+      "loss": 1.3271,
+      "step": 751
+    },
+    {
+      "epoch": 1.443590974555929,
+      "grad_norm": 0.6846915648125165,
+      "learning_rate": 8.959957605273274e-06,
+      "loss": 1.1573,
+      "step": 752
+    },
+    {
+      "epoch": 1.4455112818050888,
+      "grad_norm": 0.6841588862649848,
+      "learning_rate": 8.955862660902828e-06,
+      "loss": 1.3383,
+      "step": 753
+    },
+    {
+      "epoch": 1.4474315890542486,
+      "grad_norm": 0.6213000908457852,
+      "learning_rate": 8.951760610524725e-06,
+      "loss": 1.1511,
+      "step": 754
+    },
+    {
+      "epoch": 1.4493518963034084,
+      "grad_norm": 0.5816520616610011,
+      "learning_rate": 8.947651461507573e-06,
+      "loss": 1.2896,
+      "step": 755
+    },
+    {
+      "epoch": 1.4512722035525685,
+      "grad_norm": 0.6834825786713268,
+      "learning_rate": 8.943535221232731e-06,
+      "loss": 1.1905,
+      "step": 756
+    },
+    {
+      "epoch": 1.4531925108017283,
+      "grad_norm": 0.7390087469911581,
+      "learning_rate": 8.939411897094298e-06,
+      "loss": 1.3174,
+      "step": 757
+    },
+    {
+      "epoch": 1.455112818050888,
+      "grad_norm": 0.7196135019950182,
+      "learning_rate": 8.935281496499098e-06,
+      "loss": 1.2327,
+      "step": 758
+    },
+    {
+      "epoch": 1.4570331253000481,
+      "grad_norm": 0.5860396585624823,
+      "learning_rate": 8.931144026866662e-06,
+      "loss": 1.2348,
+      "step": 759
+    },
+    {
+      "epoch": 1.458953432549208,
+      "grad_norm": 0.7918382564660377,
+      "learning_rate": 8.926999495629225e-06,
+      "loss": 1.1791,
+      "step": 760
+    },
+    {
+      "epoch": 1.4608737397983678,
+      "grad_norm": 0.7055490148273642,
+      "learning_rate": 8.922847910231702e-06,
+      "loss": 1.1687,
+      "step": 761
+    },
+    {
+      "epoch": 1.4627940470475276,
+      "grad_norm": 0.5524065552778338,
+      "learning_rate": 8.918689278131684e-06,
+      "loss": 1.3717,
+      "step": 762
+    },
+    {
+      "epoch": 1.4647143542966874,
+      "grad_norm": 0.7460708719531967,
+      "learning_rate": 8.914523606799416e-06,
+      "loss": 1.2305,
+      "step": 763
+    },
+    {
+      "epoch": 1.4666346615458474,
+      "grad_norm": 0.6951265014515928,
+      "learning_rate": 8.910350903717793e-06,
+      "loss": 1.3857,
+      "step": 764
+    },
+    {
+      "epoch": 1.4685549687950072,
+      "grad_norm": 0.6742950089132507,
+      "learning_rate": 8.906171176382336e-06,
+      "loss": 1.2732,
+      "step": 765
+    },
+    {
+      "epoch": 1.470475276044167,
+      "grad_norm": 0.7672591308011247,
+      "learning_rate": 8.901984432301185e-06,
+      "loss": 1.3741,
+      "step": 766
+    },
+    {
+      "epoch": 1.4723955832933269,
+      "grad_norm": 0.6895255447617727,
+      "learning_rate": 8.897790678995088e-06,
+      "loss": 1.1914,
+      "step": 767
+    },
+    {
+      "epoch": 1.4743158905424867,
+      "grad_norm": 0.5905849417666214,
+      "learning_rate": 8.893589923997379e-06,
+      "loss": 1.2328,
+      "step": 768
+    },
+    {
+      "epoch": 1.4762361977916467,
+      "grad_norm": 0.7661433188154492,
+      "learning_rate": 8.889382174853971e-06,
+      "loss": 1.4402,
+      "step": 769
+    },
+    {
+      "epoch": 1.4781565050408065,
+      "grad_norm": 0.7226773436729304,
+      "learning_rate": 8.885167439123343e-06,
+      "loss": 1.3498,
+      "step": 770
+    },
+    {
+      "epoch": 1.4800768122899663,
+      "grad_norm": 0.645854499389175,
+      "learning_rate": 8.880945724376522e-06,
+      "loss": 1.0738,
+      "step": 771
+    },
+    {
+      "epoch": 1.4819971195391264,
+      "grad_norm": 0.730924402691002,
+      "learning_rate": 8.876717038197072e-06,
+      "loss": 1.2628,
+      "step": 772
+    },
+    {
+      "epoch": 1.4839174267882862,
+      "grad_norm": 0.7007233981122745,
+      "learning_rate": 8.872481388181076e-06,
+      "loss": 1.1873,
+      "step": 773
+    },
+    {
+      "epoch": 1.485837734037446,
+      "grad_norm": 0.5889909988446032,
+      "learning_rate": 8.868238781937137e-06,
+      "loss": 1.0944,
+      "step": 774
+    },
+    {
+      "epoch": 1.4877580412866058,
+      "grad_norm": 0.7107956135191795,
+      "learning_rate": 8.863989227086343e-06,
+      "loss": 1.3349,
+      "step": 775
+    },
+    {
+      "epoch": 1.4896783485357656,
+      "grad_norm": 0.6175391436776866,
+      "learning_rate": 8.859732731262268e-06,
+      "loss": 1.2646,
+      "step": 776
+    },
+    {
+      "epoch": 1.4915986557849257,
+      "grad_norm": 0.65271111460711,
+      "learning_rate": 8.855469302110952e-06,
+      "loss": 1.2263,
+      "step": 777
+    },
+    {
+      "epoch": 1.4935189630340855,
+      "grad_norm": 0.6583187935159579,
+      "learning_rate": 8.851198947290895e-06,
+      "loss": 1.202,
+      "step": 778
+    },
+    {
+      "epoch": 1.4954392702832453,
+      "grad_norm": 0.5781288868057943,
+      "learning_rate": 8.846921674473033e-06,
+      "loss": 1.3714,
+      "step": 779
+    },
+    {
+      "epoch": 1.4973595775324051,
+      "grad_norm": 0.6783395545902824,
+      "learning_rate": 8.842637491340728e-06,
+      "loss": 1.3051,
+      "step": 780
+    },
+    {
+      "epoch": 1.499279884781565,
+      "grad_norm": 0.7190117162268456,
+      "learning_rate": 8.83834640558976e-06,
+      "loss": 1.1348,
+      "step": 781
+    },
+    {
+      "epoch": 1.501200192030725,
+      "grad_norm": 0.7371660449929316,
+      "learning_rate": 8.834048424928305e-06,
+      "loss": 1.2488,
+      "step": 782
+    },
+    {
+      "epoch": 1.5031204992798848,
+      "grad_norm": 0.6734459491576749,
+      "learning_rate": 8.829743557076926e-06,
+      "loss": 1.2673,
+      "step": 783
+    },
+    {
+      "epoch": 1.5050408065290446,
+      "grad_norm": 0.6032860650966215,
+      "learning_rate": 8.825431809768554e-06,
+      "loss": 1.2917,
+      "step": 784
+    },
+    {
+      "epoch": 1.5069611137782046,
+      "grad_norm": 0.7510759569269795,
+      "learning_rate": 8.821113190748482e-06,
+      "loss": 1.4579,
+      "step": 785
+    },
+    {
+      "epoch": 1.5088814210273642,
+      "grad_norm": 0.6955345602964695,
+      "learning_rate": 8.816787707774347e-06,
+      "loss": 1.2052,
+      "step": 786
+    },
+    {
+      "epoch": 1.5108017282765243,
+      "grad_norm": 0.699131477732091,
+      "learning_rate": 8.812455368616112e-06,
+      "loss": 1.2347,
+      "step": 787
+    },
+    {
+      "epoch": 1.512722035525684,
+      "grad_norm": 0.6752955040693549,
+      "learning_rate": 8.808116181056059e-06,
+      "loss": 1.2042,
+      "step": 788
+    },
+    {
+      "epoch": 1.5146423427748439,
+      "grad_norm": 0.7495344136959898,
+      "learning_rate": 8.803770152888771e-06,
+      "loss": 1.2847,
+      "step": 789
+    },
+    {
+      "epoch": 1.516562650024004,
+      "grad_norm": 0.6746945912508688,
+      "learning_rate": 8.799417291921117e-06,
+      "loss": 1.3633,
+      "step": 790
+    },
+    {
+      "epoch": 1.5184829572731637,
+      "grad_norm": 0.7701161584892516,
+      "learning_rate": 8.795057605972247e-06,
+      "loss": 1.3794,
+      "step": 791
+    },
+    {
+      "epoch": 1.5204032645223235,
+      "grad_norm": 0.7268566540300971,
+      "learning_rate": 8.790691102873558e-06,
+      "loss": 1.1022,
+      "step": 792
+    },
+    {
+      "epoch": 1.5223235717714836,
+      "grad_norm": 0.7419238096069151,
+      "learning_rate": 8.786317790468708e-06,
+      "loss": 1.3445,
+      "step": 793
+    },
+    {
+      "epoch": 1.5242438790206432,
+      "grad_norm": 0.5804158480082398,
+      "learning_rate": 8.781937676613577e-06,
+      "loss": 1.5153,
+      "step": 794
+    },
+    {
+      "epoch": 1.5261641862698032,
+      "grad_norm": 0.816197319588985,
+      "learning_rate": 8.777550769176263e-06,
+      "loss": 1.2333,
+      "step": 795
+    },
+    {
+      "epoch": 1.528084493518963,
+      "grad_norm": 0.7546662492454415,
+      "learning_rate": 8.77315707603707e-06,
+      "loss": 1.2277,
+      "step": 796
+    },
+    {
+      "epoch": 1.5300048007681228,
+      "grad_norm": 0.580844842754435,
+      "learning_rate": 8.76875660508849e-06,
+      "loss": 1.403,
+      "step": 797
+    },
+    {
+      "epoch": 1.5319251080172829,
+      "grad_norm": 0.6741408076571399,
+      "learning_rate": 8.764349364235197e-06,
+      "loss": 1.3271,
+      "step": 798
+    },
+    {
+      "epoch": 1.5338454152664425,
+      "grad_norm": 0.779691874848022,
+      "learning_rate": 8.759935361394011e-06,
+      "loss": 1.3694,
+      "step": 799
+    },
+    {
+      "epoch": 1.5357657225156025,
+      "grad_norm": 0.7069960035792014,
+      "learning_rate": 8.755514604493912e-06,
+      "loss": 1.1364,
+      "step": 800
+    },
+    {
+      "epoch": 1.5376860297647623,
+      "grad_norm": 0.6332924636304982,
+      "learning_rate": 8.751087101476008e-06,
+      "loss": 1.4002,
+      "step": 801
+    },
+    {
+      "epoch": 1.5396063370139221,
+      "grad_norm": 0.7044744362416293,
+      "learning_rate": 8.746652860293523e-06,
+      "loss": 1.1917,
+      "step": 802
+    },
+    {
+      "epoch": 1.5415266442630822,
+      "grad_norm": 0.7812830705790423,
+      "learning_rate": 8.742211888911788e-06,
+      "loss": 1.4769,
+      "step": 803
+    },
+    {
+      "epoch": 1.543446951512242,
+      "grad_norm": 0.675685814114717,
+      "learning_rate": 8.737764195308226e-06,
+      "loss": 1.2728,
+      "step": 804
+    },
+    {
+      "epoch": 1.5453672587614018,
+      "grad_norm": 0.6820193005764167,
+      "learning_rate": 8.733309787472328e-06,
+      "loss": 1.3128,
+      "step": 805
+    },
+    {
+      "epoch": 1.5472875660105618,
+      "grad_norm": 0.7153026774691486,
+      "learning_rate": 8.72884867340565e-06,
+      "loss": 1.1845,
+      "step": 806
+    },
+    {
+      "epoch": 1.5492078732597214,
+      "grad_norm": 0.6525337231248045,
+      "learning_rate": 8.724380861121795e-06,
+      "loss": 1.2251,
+      "step": 807
+    },
+    {
+      "epoch": 1.5511281805088815,
+      "grad_norm": 0.6540180109670667,
+      "learning_rate": 8.7199063586464e-06,
+      "loss": 1.2874,
+      "step": 808
+    },
+    {
+      "epoch": 1.5530484877580413,
+      "grad_norm": 0.7635326405494673,
+      "learning_rate": 8.715425174017113e-06,
+      "loss": 1.358,
+      "step": 809
+    },
+    {
+      "epoch": 1.554968795007201,
+      "grad_norm": 0.6242507146581721,
+      "learning_rate": 8.710937315283594e-06,
+      "loss": 1.2419,
+      "step": 810
+    },
+    {
+      "epoch": 1.5568891022563611,
+      "grad_norm": 0.744752759091194,
+      "learning_rate": 8.706442790507486e-06,
+      "loss": 1.3055,
+      "step": 811
+    },
+    {
+      "epoch": 1.558809409505521,
+      "grad_norm": 0.6654168047572514,
+      "learning_rate": 8.701941607762407e-06,
+      "loss": 1.4469,
+      "step": 812
+    },
+    {
+      "epoch": 1.5607297167546808,
+      "grad_norm": 0.7586952812790637,
+      "learning_rate": 8.697433775133934e-06,
+      "loss": 1.2091,
+      "step": 813
+    },
+    {
+      "epoch": 1.5626500240038406,
+      "grad_norm": 0.7505624695472355,
+      "learning_rate": 8.692919300719596e-06,
+      "loss": 1.1346,
+      "step": 814
+    },
+    {
+      "epoch": 1.5645703312530004,
+      "grad_norm": 0.6150274934828531,
+      "learning_rate": 8.688398192628844e-06,
+      "loss": 1.2218,
+      "step": 815
+    },
+    {
+      "epoch": 1.5664906385021604,
+      "grad_norm": 0.6830054941296687,
+      "learning_rate": 8.68387045898305e-06,
+      "loss": 1.1967,
+      "step": 816
+    },
+    {
+      "epoch": 1.5684109457513202,
+      "grad_norm": 0.6697744853587582,
+      "learning_rate": 8.679336107915487e-06,
+      "loss": 1.2602,
+      "step": 817
+    },
+    {
+      "epoch": 1.57033125300048,
+      "grad_norm": 0.6949578343964008,
+      "learning_rate": 8.674795147571318e-06,
+      "loss": 1.2928,
+      "step": 818
+    },
+    {
+      "epoch": 1.57225156024964,
+      "grad_norm": 0.633917787386206,
+      "learning_rate": 8.670247586107567e-06,
+      "loss": 1.2756,
+      "step": 819
+    },
+    {
+      "epoch": 1.5741718674987997,
+      "grad_norm": 0.7247388852385745,
+      "learning_rate": 8.665693431693132e-06,
+      "loss": 1.2376,
+      "step": 820
+    },
+    {
+      "epoch": 1.5760921747479597,
+      "grad_norm": 0.6370568157248708,
+      "learning_rate": 8.661132692508742e-06,
+      "loss": 1.2305,
+      "step": 821
+    },
+    {
+      "epoch": 1.5780124819971195,
+      "grad_norm": 0.5773275861749614,
+      "learning_rate": 8.656565376746959e-06,
+      "loss": 1.3996,
+      "step": 822
+    },
+    {
+      "epoch": 1.5799327892462793,
+      "grad_norm": 0.7055843854327438,
+      "learning_rate": 8.65199149261216e-06,
+      "loss": 1.2102,
+      "step": 823
+    },
+    {
+      "epoch": 1.5818530964954394,
+      "grad_norm": 0.7568845107850198,
+      "learning_rate": 8.647411048320515e-06,
+      "loss": 1.14,
+      "step": 824
+    },
+    {
+      "epoch": 1.5837734037445992,
+      "grad_norm": 0.6353739121479085,
+      "learning_rate": 8.642824052099986e-06,
+      "loss": 1.383,
+      "step": 825
+    },
+    {
+      "epoch": 1.585693710993759,
+      "grad_norm": 0.7245216800702469,
+      "learning_rate": 8.638230512190298e-06,
+      "loss": 1.3269,
+      "step": 826
+    },
+    {
+      "epoch": 1.5876140182429188,
+      "grad_norm": 0.6456003754892913,
+      "learning_rate": 8.633630436842933e-06,
+      "loss": 1.391,
+      "step": 827
+    },
+    {
+      "epoch": 1.5895343254920786,
+      "grad_norm": 0.764262777332275,
+      "learning_rate": 8.629023834321113e-06,
+      "loss": 1.3651,
+      "step": 828
+    },
+    {
+      "epoch": 1.5914546327412387,
+      "grad_norm": 0.6827790863280228,
+      "learning_rate": 8.624410712899787e-06,
+      "loss": 1.2995,
+      "step": 829
+    },
+    {
+      "epoch": 1.5933749399903985,
+      "grad_norm": 0.6862806810660659,
+      "learning_rate": 8.619791080865609e-06,
+      "loss": 1.1752,
+      "step": 830
+    },
+    {
+      "epoch": 1.5952952472395583,
+      "grad_norm": 0.6926687316972496,
+      "learning_rate": 8.615164946516935e-06,
+      "loss": 1.2555,
+      "step": 831
+    },
+    {
+      "epoch": 1.5972155544887183,
+      "grad_norm": 0.6251419290492334,
+      "learning_rate": 8.61053231816379e-06,
+      "loss": 1.163,
+      "step": 832
+    },
+    {
+      "epoch": 1.599135861737878,
+      "grad_norm": 0.692086428022269,
+      "learning_rate": 8.605893204127877e-06,
+      "loss": 1.324,
+      "step": 833
+    },
+    {
+      "epoch": 1.601056168987038,
+      "grad_norm": 0.7778455904448239,
+      "learning_rate": 8.601247612742545e-06,
+      "loss": 1.2905,
+      "step": 834
+    },
+    {
+      "epoch": 1.6029764762361978,
+      "grad_norm": 0.6196302763793925,
+      "learning_rate": 8.596595552352773e-06,
+      "loss": 1.274,
+      "step": 835
+    },
+    {
+      "epoch": 1.6048967834853576,
+      "grad_norm": 0.6237845300699906,
+      "learning_rate": 8.591937031315167e-06,
+      "loss": 1.2982,
+      "step": 836
+    },
+    {
+      "epoch": 1.6068170907345176,
+      "grad_norm": 0.613978703840158,
+      "learning_rate": 8.587272057997937e-06,
+      "loss": 1.2335,
+      "step": 837
+    },
+    {
+      "epoch": 1.6087373979836774,
+      "grad_norm": 0.6686343059058948,
+      "learning_rate": 8.58260064078088e-06,
+      "loss": 1.3702,
+      "step": 838
+    },
+    {
+      "epoch": 1.6106577052328372,
+      "grad_norm": 0.759769684426153,
+      "learning_rate": 8.577922788055375e-06,
+      "loss": 1.2766,
+      "step": 839
+    },
+    {
+      "epoch": 1.6125780124819973,
+      "grad_norm": 0.6245956368253657,
+      "learning_rate": 8.573238508224351e-06,
+      "loss": 1.2805,
+      "step": 840
+    },
+    {
+      "epoch": 1.6144983197311569,
+      "grad_norm": 0.7290629551761605,
+      "learning_rate": 8.568547809702294e-06,
+      "loss": 1.1846,
+      "step": 841
+    },
+    {
+      "epoch": 1.616418626980317,
+      "grad_norm": 0.6779816447540971,
+      "learning_rate": 8.563850700915211e-06,
+      "loss": 1.3086,
+      "step": 842
+    },
+    {
+      "epoch": 1.6183389342294767,
+      "grad_norm": 0.6318966650515423,
+      "learning_rate": 8.55914719030063e-06,
+      "loss": 1.131,
+      "step": 843
+    },
+    {
+      "epoch": 1.6202592414786365,
+      "grad_norm": 0.6427129904533472,
+      "learning_rate": 8.554437286307573e-06,
+      "loss": 1.3493,
+      "step": 844
+    },
+    {
+      "epoch": 1.6221795487277966,
+      "grad_norm": 0.7080851195930218,
+      "learning_rate": 8.54972099739655e-06,
+      "loss": 1.2393,
+      "step": 845
+    },
+    {
+      "epoch": 1.6240998559769562,
+      "grad_norm": 0.6160771656855515,
+      "learning_rate": 8.544998332039543e-06,
+      "loss": 1.3681,
+      "step": 846
+    },
+    {
+      "epoch": 1.6260201632261162,
+      "grad_norm": 0.8454044975860124,
+      "learning_rate": 8.540269298719981e-06,
+      "loss": 1.2243,
+      "step": 847
+    },
+    {
+      "epoch": 1.627940470475276,
+      "grad_norm": 0.7375528021953657,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 1.325,
+      "step": 848
+    },
+    {
+      "epoch": 1.6298607777244358,
+      "grad_norm": 0.6216999198337684,
+      "learning_rate": 8.530792162184112e-06,
+      "loss": 1.3877,
+      "step": 849
+    },
+    {
+      "epoch": 1.6317810849735959,
+      "grad_norm": 0.6818891969657709,
+      "learning_rate": 8.526044075991801e-06,
+      "loss": 1.1877,
+      "step": 850
+    },
+    {
+      "epoch": 1.6337013922227557,
+      "grad_norm": 0.7427712661902629,
+      "learning_rate": 8.521289655884908e-06,
+      "loss": 1.1278,
+      "step": 851
+    },
+    {
+      "epoch": 1.6356216994719155,
+      "grad_norm": 0.5952544225101838,
+      "learning_rate": 8.516528910403906e-06,
+      "loss": 1.4064,
+      "step": 852
+    },
+    {
+      "epoch": 1.6375420067210755,
+      "grad_norm": 0.6848120275660302,
+      "learning_rate": 8.51176184810063e-06,
+      "loss": 1.3963,
+      "step": 853
+    },
+    {
+      "epoch": 1.6394623139702351,
+      "grad_norm": 0.7063980789331809,
+      "learning_rate": 8.506988477538267e-06,
+      "loss": 1.1821,
+      "step": 854
+    },
+    {
+      "epoch": 1.6413826212193952,
+      "grad_norm": 0.6446077020839321,
+      "learning_rate": 8.502208807291332e-06,
+      "loss": 1.3422,
+      "step": 855
+    },
+    {
+      "epoch": 1.643302928468555,
+      "grad_norm": 0.6388455910147647,
+      "learning_rate": 8.497422845945658e-06,
+      "loss": 1.2511,
+      "step": 856
+    },
+    {
+      "epoch": 1.6452232357177148,
+      "grad_norm": 0.7399303493707643,
+      "learning_rate": 8.492630602098377e-06,
+      "loss": 1.2217,
+      "step": 857
+    },
+    {
+      "epoch": 1.6471435429668748,
+      "grad_norm": 0.7045086476609864,
+      "learning_rate": 8.487832084357908e-06,
+      "loss": 1.3431,
+      "step": 858
+    },
+    {
+      "epoch": 1.6490638502160344,
+      "grad_norm": 0.7039565820761484,
+      "learning_rate": 8.483027301343942e-06,
+      "loss": 1.3048,
+      "step": 859
+    },
+    {
+      "epoch": 1.6509841574651944,
+      "grad_norm": 0.6338362458948208,
+      "learning_rate": 8.478216261687417e-06,
+      "loss": 1.2741,
+      "step": 860
+    },
+    {
+      "epoch": 1.6529044647143543,
+      "grad_norm": 0.7079854924528219,
+      "learning_rate": 8.473398974030519e-06,
+      "loss": 1.317,
+      "step": 861
+    },
+    {
+      "epoch": 1.654824771963514,
+      "grad_norm": 0.7544349465014712,
+      "learning_rate": 8.468575447026653e-06,
+      "loss": 1.2428,
+      "step": 862
+    },
+    {
+      "epoch": 1.656745079212674,
+      "grad_norm": 0.7579460936381943,
+      "learning_rate": 8.463745689340428e-06,
+      "loss": 1.327,
+      "step": 863
+    },
+    {
+      "epoch": 1.658665386461834,
+      "grad_norm": 0.6500673020764594,
+      "learning_rate": 8.458909709647653e-06,
+      "loss": 1.3508,
+      "step": 864
+    },
+    {
+      "epoch": 1.6605856937109937,
+      "grad_norm": 0.5897561762214972,
+      "learning_rate": 8.45406751663531e-06,
+      "loss": 1.5225,
+      "step": 865
+    },
+    {
+      "epoch": 1.6625060009601538,
+      "grad_norm": 0.7959327705311353,
+      "learning_rate": 8.449219119001543e-06,
+      "loss": 1.3729,
+      "step": 866
+    },
+    {
+      "epoch": 1.6644263082093134,
+      "grad_norm": 0.7331153774645073,
+      "learning_rate": 8.444364525455642e-06,
+      "loss": 1.2854,
+      "step": 867
+    },
+    {
+      "epoch": 1.6663466154584734,
+      "grad_norm": 0.5928127084001641,
+      "learning_rate": 8.43950374471802e-06,
+      "loss": 1.3122,
+      "step": 868
+    },
+    {
+      "epoch": 1.6682669227076332,
+      "grad_norm": 0.6755707034773414,
+      "learning_rate": 8.434636785520218e-06,
+      "loss": 1.2862,
+      "step": 869
+    },
+    {
+      "epoch": 1.670187229956793,
+      "grad_norm": 0.6339666901160925,
+      "learning_rate": 8.42976365660486e-06,
+      "loss": 1.3128,
+      "step": 870
+    },
+    {
+      "epoch": 1.672107537205953,
+      "grad_norm": 0.678020909918651,
+      "learning_rate": 8.424884366725665e-06,
+      "loss": 1.2833,
+      "step": 871
+    },
+    {
+      "epoch": 1.6740278444551127,
+      "grad_norm": 0.643027666772922,
+      "learning_rate": 8.419998924647412e-06,
+      "loss": 1.2969,
+      "step": 872
+    },
+    {
+      "epoch": 1.6759481517042727,
+      "grad_norm": 0.6961417219596171,
+      "learning_rate": 8.415107339145933e-06,
+      "loss": 1.1697,
+      "step": 873
+    },
+    {
+      "epoch": 1.6778684589534325,
+      "grad_norm": 0.5982028382247672,
+      "learning_rate": 8.4102096190081e-06,
+      "loss": 1.3004,
+      "step": 874
+    },
+    {
+      "epoch": 1.6797887662025923,
+      "grad_norm": 0.7682354786227164,
+      "learning_rate": 8.4053057730318e-06,
+      "loss": 1.3747,
+      "step": 875
+    },
+    {
+      "epoch": 1.6817090734517524,
+      "grad_norm": 0.6716740624063096,
+      "learning_rate": 8.400395810025922e-06,
+      "loss": 1.2719,
+      "step": 876
+    },
+    {
+      "epoch": 1.6836293807009122,
+      "grad_norm": 0.6440524699748738,
+      "learning_rate": 8.395479738810347e-06,
+      "loss": 1.3379,
+      "step": 877
+    },
+    {
+      "epoch": 1.685549687950072,
+      "grad_norm": 0.6920927110189645,
+      "learning_rate": 8.39055756821593e-06,
+      "loss": 1.2993,
+      "step": 878
+    },
+    {
+      "epoch": 1.687469995199232,
+      "grad_norm": 0.6987825363370698,
+      "learning_rate": 8.385629307084477e-06,
+      "loss": 1.2867,
+      "step": 879
+    },
+    {
+      "epoch": 1.6893903024483916,
+      "grad_norm": 0.7765961352229686,
+      "learning_rate": 8.38069496426874e-06,
+      "loss": 1.2875,
+      "step": 880
+    },
+    {
+      "epoch": 1.6913106096975516,
+      "grad_norm": 0.5996561276864317,
+      "learning_rate": 8.375754548632392e-06,
+      "loss": 1.4783,
+      "step": 881
+    },
+    {
+      "epoch": 1.6932309169467115,
+      "grad_norm": 0.5850694559829233,
+      "learning_rate": 8.370808069050016e-06,
+      "loss": 1.1561,
+      "step": 882
+    },
+    {
+      "epoch": 1.6951512241958713,
+      "grad_norm": 0.7456789269050706,
+      "learning_rate": 8.365855534407089e-06,
+      "loss": 1.2655,
+      "step": 883
+    },
+    {
+      "epoch": 1.6970715314450313,
+      "grad_norm": 0.6083339409975786,
+      "learning_rate": 8.360896953599962e-06,
+      "loss": 1.1948,
+      "step": 884
+    },
+    {
+      "epoch": 1.6989918386941911,
+      "grad_norm": 0.9171462954821785,
+      "learning_rate": 8.355932335535849e-06,
+      "loss": 1.2125,
+      "step": 885
+    },
+    {
+      "epoch": 1.700912145943351,
+      "grad_norm": 0.6521085415420599,
+      "learning_rate": 8.350961689132808e-06,
+      "loss": 1.4241,
+      "step": 886
+    },
+    {
+      "epoch": 1.7028324531925108,
+      "grad_norm": 0.7910874600723566,
+      "learning_rate": 8.345985023319727e-06,
+      "loss": 1.1417,
+      "step": 887
+    },
+    {
+      "epoch": 1.7047527604416706,
+      "grad_norm": 0.6027394559994305,
+      "learning_rate": 8.341002347036304e-06,
+      "loss": 1.2801,
+      "step": 888
+    },
+    {
+      "epoch": 1.7066730676908306,
+      "grad_norm": 0.6966016112269007,
+      "learning_rate": 8.336013669233039e-06,
+      "loss": 1.271,
+      "step": 889
+    },
+    {
+      "epoch": 1.7085933749399904,
+      "grad_norm": 0.6919363199470899,
+      "learning_rate": 8.331018998871207e-06,
+      "loss": 1.2261,
+      "step": 890
+    },
+    {
+      "epoch": 1.7105136821891502,
+      "grad_norm": 0.6706001120414775,
+      "learning_rate": 8.32601834492285e-06,
+      "loss": 1.176,
+      "step": 891
+    },
+    {
+      "epoch": 1.7124339894383103,
+      "grad_norm": 0.6297425533497251,
+      "learning_rate": 8.32101171637076e-06,
+      "loss": 1.0596,
+      "step": 892
+    },
+    {
+      "epoch": 1.7143542966874699,
+      "grad_norm": 0.7344756503897651,
+      "learning_rate": 8.315999122208459e-06,
+      "loss": 1.2274,
+      "step": 893
+    },
+    {
+      "epoch": 1.71627460393663,
+      "grad_norm": 0.5546082526816237,
+      "learning_rate": 8.310980571440184e-06,
+      "loss": 1.2394,
+      "step": 894
+    },
+    {
+      "epoch": 1.7181949111857897,
+      "grad_norm": 0.7256143287268254,
+      "learning_rate": 8.305956073080879e-06,
+      "loss": 1.2171,
+      "step": 895
+    },
+    {
+      "epoch": 1.7201152184349495,
+      "grad_norm": 0.6381053874642392,
+      "learning_rate": 8.300925636156159e-06,
+      "loss": 1.3499,
+      "step": 896
+    },
+    {
+      "epoch": 1.7220355256841096,
+      "grad_norm": 0.705514992532687,
+      "learning_rate": 8.295889269702322e-06,
+      "loss": 1.3576,
+      "step": 897
+    },
+    {
+      "epoch": 1.7239558329332694,
+      "grad_norm": 0.6870236039044634,
+      "learning_rate": 8.290846982766305e-06,
+      "loss": 1.3549,
+      "step": 898
+    },
+    {
+      "epoch": 1.7258761401824292,
+      "grad_norm": 0.6054537907728821,
+      "learning_rate": 8.285798784405685e-06,
+      "loss": 1.4147,
+      "step": 899
+    },
+    {
+      "epoch": 1.727796447431589,
+      "grad_norm": 0.6774991278621717,
+      "learning_rate": 8.28074468368866e-06,
+      "loss": 1.2747,
+      "step": 900
+    },
+    {
+      "epoch": 1.7297167546807488,
+      "grad_norm": 0.5809303428546913,
+      "learning_rate": 8.275684689694025e-06,
+      "loss": 1.2324,
+      "step": 901
+    },
+    {
+      "epoch": 1.7316370619299088,
+      "grad_norm": 0.6418772583066495,
+      "learning_rate": 8.270618811511166e-06,
+      "loss": 1.3204,
+      "step": 902
+    },
+    {
+      "epoch": 1.7335573691790687,
+      "grad_norm": 0.6916575011270142,
+      "learning_rate": 8.265547058240038e-06,
+      "loss": 1.2656,
+      "step": 903
+    },
+    {
+      "epoch": 1.7354776764282285,
+      "grad_norm": 0.6254607447916599,
+      "learning_rate": 8.260469438991147e-06,
+      "loss": 1.5506,
+      "step": 904
+    },
+    {
+      "epoch": 1.7373979836773885,
+      "grad_norm": 0.7560305369852807,
+      "learning_rate": 8.255385962885541e-06,
+      "loss": 1.2419,
+      "step": 905
+    },
+    {
+      "epoch": 1.739318290926548,
+      "grad_norm": 0.6276013332706113,
+      "learning_rate": 8.250296639054782e-06,
+      "loss": 1.1115,
+      "step": 906
+    },
+    {
+      "epoch": 1.7412385981757081,
+      "grad_norm": 0.6544438928109518,
+      "learning_rate": 8.245201476640943e-06,
+      "loss": 1.3512,
+      "step": 907
+    },
+    {
+      "epoch": 1.743158905424868,
+      "grad_norm": 0.5929397616630802,
+      "learning_rate": 8.240100484796581e-06,
+      "loss": 1.3889,
+      "step": 908
+    },
+    {
+      "epoch": 1.7450792126740278,
+      "grad_norm": 0.7204827062155562,
+      "learning_rate": 8.234993672684724e-06,
+      "loss": 1.2227,
+      "step": 909
+    },
+    {
+      "epoch": 1.7469995199231878,
+      "grad_norm": 0.6671792020030981,
+      "learning_rate": 8.229881049478859e-06,
+      "loss": 1.3579,
+      "step": 910
+    },
+    {
+      "epoch": 1.7489198271723476,
+      "grad_norm": 0.6522101763399759,
+      "learning_rate": 8.224762624362909e-06,
+      "loss": 1.4075,
+      "step": 911
+    },
+    {
+      "epoch": 1.7508401344215074,
+      "grad_norm": 0.6373133273160264,
+      "learning_rate": 8.21963840653122e-06,
+      "loss": 1.2467,
+      "step": 912
+    },
+    {
+      "epoch": 1.7527604416706675,
+      "grad_norm": 0.6100626574590131,
+      "learning_rate": 8.214508405188543e-06,
+      "loss": 1.3204,
+      "step": 913
+    },
+    {
+      "epoch": 1.754680748919827,
+      "grad_norm": 0.7382960675500586,
+      "learning_rate": 8.209372629550018e-06,
+      "loss": 1.3108,
+      "step": 914
+    },
+    {
+      "epoch": 1.756601056168987,
+      "grad_norm": 0.6147124939284418,
+      "learning_rate": 8.204231088841156e-06,
+      "loss": 1.301,
+      "step": 915
+    },
+    {
+      "epoch": 1.758521363418147,
+      "grad_norm": 0.6584981237032568,
+      "learning_rate": 8.199083792297828e-06,
+      "loss": 1.2281,
+      "step": 916
+    },
+    {
+      "epoch": 1.7604416706673067,
+      "grad_norm": 0.6195212578955458,
+      "learning_rate": 8.193930749166239e-06,
+      "loss": 1.3111,
+      "step": 917
+    },
+    {
+      "epoch": 1.7623619779164668,
+      "grad_norm": 0.6466090662113759,
+      "learning_rate": 8.188771968702924e-06,
+      "loss": 1.1323,
+      "step": 918
+    },
+    {
+      "epoch": 1.7642822851656264,
+      "grad_norm": 0.7163500412541859,
+      "learning_rate": 8.183607460174716e-06,
+      "loss": 1.2133,
+      "step": 919
+    },
+    {
+      "epoch": 1.7662025924147864,
+      "grad_norm": 0.6355256943730633,
+      "learning_rate": 8.178437232858743e-06,
+      "loss": 1.2225,
+      "step": 920
+    },
+    {
+      "epoch": 1.7681228996639462,
+      "grad_norm": 0.6133836361205134,
+      "learning_rate": 8.173261296042401e-06,
+      "loss": 1.2216,
+      "step": 921
+    },
+    {
+      "epoch": 1.770043206913106,
+      "grad_norm": 0.6259716243382641,
+      "learning_rate": 8.168079659023349e-06,
+      "loss": 1.1422,
+      "step": 922
+    },
+    {
+      "epoch": 1.771963514162266,
+      "grad_norm": 0.6770902691471423,
+      "learning_rate": 8.162892331109483e-06,
+      "loss": 1.321,
+      "step": 923
+    },
+    {
+      "epoch": 1.7738838214114259,
+      "grad_norm": 0.637568102850358,
+      "learning_rate": 8.157699321618912e-06,
+      "loss": 1.1804,
+      "step": 924
+    },
+    {
+      "epoch": 1.7758041286605857,
+      "grad_norm": 0.7382626387455697,
+      "learning_rate": 8.152500639879968e-06,
+      "loss": 1.4325,
+      "step": 925
+    },
+    {
+      "epoch": 1.7777244359097457,
+      "grad_norm": 0.6818352440888269,
+      "learning_rate": 8.147296295231158e-06,
+      "loss": 1.2576,
+      "step": 926
+    },
+    {
+      "epoch": 1.7796447431589053,
+      "grad_norm": 0.7236454601947588,
+      "learning_rate": 8.14208629702117e-06,
+      "loss": 1.2064,
+      "step": 927
+    },
+    {
+      "epoch": 1.7815650504080653,
+      "grad_norm": 0.6339212840273679,
+      "learning_rate": 8.136870654608842e-06,
+      "loss": 1.3321,
+      "step": 928
+    },
+    {
+      "epoch": 1.7834853576572252,
+      "grad_norm": 0.6741918786895994,
+      "learning_rate": 8.131649377363154e-06,
+      "loss": 1.2423,
+      "step": 929
+    },
+    {
+      "epoch": 1.785405664906385,
+      "grad_norm": 0.7124925197818013,
+      "learning_rate": 8.126422474663205e-06,
+      "loss": 1.2692,
+      "step": 930
+    },
+    {
+      "epoch": 1.787325972155545,
+      "grad_norm": 0.6741223242032481,
+      "learning_rate": 8.121189955898209e-06,
+      "loss": 1.2554,
+      "step": 931
+    },
+    {
+      "epoch": 1.7892462794047046,
+      "grad_norm": 0.6802865384435456,
+      "learning_rate": 8.11595183046745e-06,
+      "loss": 1.3452,
+      "step": 932
+    },
+    {
+      "epoch": 1.7911665866538646,
+      "grad_norm": 0.7320772893126046,
+      "learning_rate": 8.1107081077803e-06,
+      "loss": 1.3645,
+      "step": 933
+    },
+    {
+      "epoch": 1.7930868939030244,
+      "grad_norm": 0.7052645982217766,
+      "learning_rate": 8.105458797256178e-06,
+      "loss": 1.1873,
+      "step": 934
+    },
+    {
+      "epoch": 1.7950072011521843,
+      "grad_norm": 0.6249372982539962,
+      "learning_rate": 8.100203908324545e-06,
+      "loss": 1.2649,
+      "step": 935
+    },
+    {
+      "epoch": 1.7969275084013443,
+      "grad_norm": 0.6873208027555705,
+      "learning_rate": 8.094943450424874e-06,
+      "loss": 1.3771,
+      "step": 936
+    },
+    {
+      "epoch": 1.7988478156505041,
+      "grad_norm": 0.6238917875853156,
+      "learning_rate": 8.089677433006651e-06,
+      "loss": 1.3356,
+      "step": 937
+    },
+    {
+      "epoch": 1.800768122899664,
+      "grad_norm": 0.7232238085832342,
+      "learning_rate": 8.084405865529345e-06,
+      "loss": 1.3674,
+      "step": 938
+    },
+    {
+      "epoch": 1.802688430148824,
+      "grad_norm": 0.5989499559453365,
+      "learning_rate": 8.079128757462393e-06,
+      "loss": 1.2039,
+      "step": 939
+    },
+    {
+      "epoch": 1.8046087373979836,
+      "grad_norm": 0.5909026058023332,
+      "learning_rate": 8.07384611828519e-06,
+      "loss": 1.0623,
+      "step": 940
+    },
+    {
+      "epoch": 1.8065290446471436,
+      "grad_norm": 0.7105011473328626,
+      "learning_rate": 8.068557957487058e-06,
+      "loss": 1.2987,
+      "step": 941
+    },
+    {
+      "epoch": 1.8084493518963034,
+      "grad_norm": 0.6532626538849133,
+      "learning_rate": 8.063264284567245e-06,
+      "loss": 1.3902,
+      "step": 942
+    },
+    {
+      "epoch": 1.8103696591454632,
+      "grad_norm": 0.6938980954824397,
+      "learning_rate": 8.057965109034898e-06,
+      "loss": 1.4046,
+      "step": 943
+    },
+    {
+      "epoch": 1.8122899663946233,
+      "grad_norm": 0.6873736259677121,
+      "learning_rate": 8.052660440409049e-06,
+      "loss": 1.3805,
+      "step": 944
+    },
+    {
+      "epoch": 1.8142102736437828,
+      "grad_norm": 0.696256394032348,
+      "learning_rate": 8.047350288218597e-06,
+      "loss": 1.2767,
+      "step": 945
+    },
+    {
+      "epoch": 1.8161305808929429,
+      "grad_norm": 0.682639831162432,
+      "learning_rate": 8.042034662002291e-06,
+      "loss": 1.3496,
+      "step": 946
+    },
+    {
+      "epoch": 1.8180508881421027,
+      "grad_norm": 0.5954020484262338,
+      "learning_rate": 8.036713571308711e-06,
+      "loss": 1.3287,
+      "step": 947
+    },
+    {
+      "epoch": 1.8199711953912625,
+      "grad_norm": 0.594228242920655,
+      "learning_rate": 8.031387025696262e-06,
+      "loss": 1.2077,
+      "step": 948
+    },
+    {
+      "epoch": 1.8218915026404225,
+      "grad_norm": 0.6239403127487786,
+      "learning_rate": 8.026055034733136e-06,
+      "loss": 1.1723,
+      "step": 949
+    },
+    {
+      "epoch": 1.8238118098895824,
+      "grad_norm": 0.8472497042145998,
+      "learning_rate": 8.020717607997311e-06,
+      "loss": 1.2776,
+      "step": 950
+    },
+    {
+      "epoch": 1.8257321171387422,
+      "grad_norm": 0.7073477206436827,
+      "learning_rate": 8.015374755076533e-06,
+      "loss": 1.2585,
+      "step": 951
+    },
+    {
+      "epoch": 1.8276524243879022,
+      "grad_norm": 0.6337785378456356,
+      "learning_rate": 8.010026485568292e-06,
+      "loss": 1.1437,
+      "step": 952
+    },
+    {
+      "epoch": 1.8295727316370618,
+      "grad_norm": 0.6413104778228611,
+      "learning_rate": 8.004672809079808e-06,
+      "loss": 1.3064,
+      "step": 953
+    },
+    {
+      "epoch": 1.8314930388862218,
+      "grad_norm": 0.6969466642249281,
+      "learning_rate": 7.999313735228012e-06,
+      "loss": 1.1853,
+      "step": 954
+    },
+    {
+      "epoch": 1.8334133461353816,
+      "grad_norm": 0.617228891044283,
+      "learning_rate": 7.993949273639535e-06,
+      "loss": 1.3008,
+      "step": 955
+    },
+    {
+      "epoch": 1.8353336533845415,
+      "grad_norm": 0.5988865953262722,
+      "learning_rate": 7.988579433950682e-06,
+      "loss": 1.1226,
+      "step": 956
+    },
+    {
+      "epoch": 1.8372539606337015,
+      "grad_norm": 0.6387895790199921,
+      "learning_rate": 7.983204225807421e-06,
+      "loss": 1.2028,
+      "step": 957
+    },
+    {
+      "epoch": 1.839174267882861,
+      "grad_norm": 0.6697886686110965,
+      "learning_rate": 7.977823658865364e-06,
+      "loss": 1.1313,
+      "step": 958
+    },
+    {
+      "epoch": 1.8410945751320211,
+      "grad_norm": 0.6747989966586785,
+      "learning_rate": 7.972437742789746e-06,
+      "loss": 1.2433,
+      "step": 959
+    },
+    {
+      "epoch": 1.843014882381181,
+      "grad_norm": 0.6640266478824588,
+      "learning_rate": 7.967046487255412e-06,
+      "loss": 1.2806,
+      "step": 960
+    },
+    {
+      "epoch": 1.8449351896303408,
+      "grad_norm": 0.6268462710779998,
+      "learning_rate": 7.9616499019468e-06,
+      "loss": 1.3466,
+      "step": 961
+    },
+    {
+      "epoch": 1.8468554968795008,
+      "grad_norm": 0.6891610903475727,
+      "learning_rate": 7.956247996557924e-06,
+      "loss": 1.2218,
+      "step": 962
+    },
+    {
+      "epoch": 1.8487758041286606,
+      "grad_norm": 0.6382707748643591,
+      "learning_rate": 7.950840780792348e-06,
+      "loss": 1.4222,
+      "step": 963
+    },
+    {
+      "epoch": 1.8506961113778204,
+      "grad_norm": 0.597235927390851,
+      "learning_rate": 7.94542826436318e-06,
+      "loss": 1.3666,
+      "step": 964
+    },
+    {
+      "epoch": 1.8526164186269805,
+      "grad_norm": 0.5958380189432244,
+      "learning_rate": 7.940010456993048e-06,
+      "loss": 1.2513,
+      "step": 965
+    },
+    {
+      "epoch": 1.85453672587614,
+      "grad_norm": 0.8207312265214936,
+      "learning_rate": 7.934587368414085e-06,
+      "loss": 1.2384,
+      "step": 966
+    },
+    {
+      "epoch": 1.8564570331253,
+      "grad_norm": 0.6383139862183063,
+      "learning_rate": 7.929159008367913e-06,
+      "loss": 1.3282,
+      "step": 967
+    },
+    {
+      "epoch": 1.85837734037446,
+      "grad_norm": 0.6473414488061252,
+      "learning_rate": 7.923725386605617e-06,
+      "loss": 1.369,
+      "step": 968
+    },
+    {
+      "epoch": 1.8602976476236197,
+      "grad_norm": 0.6461249289533425,
+      "learning_rate": 7.918286512887738e-06,
+      "loss": 1.3708,
+      "step": 969
+    },
+    {
+      "epoch": 1.8622179548727797,
+      "grad_norm": 0.6996074686318671,
+      "learning_rate": 7.912842396984256e-06,
+      "loss": 1.301,
+      "step": 970
+    },
+    {
+      "epoch": 1.8641382621219396,
+      "grad_norm": 0.747195100156215,
+      "learning_rate": 7.907393048674554e-06,
+      "loss": 1.3042,
+      "step": 971
+    },
+    {
+      "epoch": 1.8660585693710994,
+      "grad_norm": 0.6487597764927422,
+      "learning_rate": 7.901938477747428e-06,
+      "loss": 1.3622,
+      "step": 972
+    },
+    {
+      "epoch": 1.8679788766202592,
+      "grad_norm": 0.6311909241077927,
+      "learning_rate": 7.89647869400105e-06,
+      "loss": 1.2718,
+      "step": 973
+    },
+    {
+      "epoch": 1.869899183869419,
+      "grad_norm": 0.8165044894363437,
+      "learning_rate": 7.891013707242953e-06,
+      "loss": 1.3604,
+      "step": 974
+    },
+    {
+      "epoch": 1.871819491118579,
+      "grad_norm": 0.6283536050272592,
+      "learning_rate": 7.885543527290023e-06,
+      "loss": 1.3357,
+      "step": 975
+    },
+    {
+      "epoch": 1.8737397983677389,
+      "grad_norm": 0.6286945845306801,
+      "learning_rate": 7.880068163968467e-06,
+      "loss": 1.3898,
+      "step": 976
+    },
+    {
+      "epoch": 1.8756601056168987,
+      "grad_norm": 0.7789702794360976,
+      "learning_rate": 7.874587627113809e-06,
+      "loss": 1.3092,
+      "step": 977
+    },
+    {
+      "epoch": 1.8775804128660587,
+      "grad_norm": 0.8489003200129025,
+      "learning_rate": 7.869101926570864e-06,
+      "loss": 1.2689,
+      "step": 978
+    },
+    {
+      "epoch": 1.8795007201152183,
+      "grad_norm": 0.7248282650889785,
+      "learning_rate": 7.863611072193721e-06,
+      "loss": 1.1416,
+      "step": 979
+    },
+    {
+      "epoch": 1.8814210273643783,
+      "grad_norm": 0.6171058372697622,
+      "learning_rate": 7.858115073845733e-06,
+      "loss": 1.2812,
+      "step": 980
+    },
+    {
+      "epoch": 1.8833413346135381,
+      "grad_norm": 0.6593534507567731,
+      "learning_rate": 7.852613941399487e-06,
+      "loss": 1.1924,
+      "step": 981
+    },
+    {
+      "epoch": 1.885261641862698,
+      "grad_norm": 0.7799530933319198,
+      "learning_rate": 7.847107684736792e-06,
+      "loss": 1.313,
+      "step": 982
+    },
+    {
+      "epoch": 1.887181949111858,
+      "grad_norm": 0.6386650101481485,
+      "learning_rate": 7.841596313748668e-06,
+      "loss": 1.291,
+      "step": 983
+    },
+    {
+      "epoch": 1.8891022563610178,
+      "grad_norm": 0.6172368641032806,
+      "learning_rate": 7.836079838335317e-06,
+      "loss": 1.1656,
+      "step": 984
+    },
+    {
+      "epoch": 1.8910225636101776,
+      "grad_norm": 0.6451182909345482,
+      "learning_rate": 7.830558268406109e-06,
+      "loss": 1.2536,
+      "step": 985
+    },
+    {
+      "epoch": 1.8929428708593377,
+      "grad_norm": 0.7371994656615729,
+      "learning_rate": 7.825031613879572e-06,
+      "loss": 1.2996,
+      "step": 986
+    },
+    {
+      "epoch": 1.8948631781084972,
+      "grad_norm": 0.6712446179245939,
+      "learning_rate": 7.81949988468336e-06,
+      "loss": 1.2534,
+      "step": 987
+    },
+    {
+      "epoch": 1.8967834853576573,
+      "grad_norm": 0.6169395012762885,
+      "learning_rate": 7.813963090754248e-06,
+      "loss": 1.2692,
+      "step": 988
+    },
+    {
+      "epoch": 1.898703792606817,
+      "grad_norm": 0.7540727018335017,
+      "learning_rate": 7.808421242038107e-06,
+      "loss": 1.4006,
+      "step": 989
+    },
+    {
+      "epoch": 1.900624099855977,
+      "grad_norm": 0.6610596828013748,
+      "learning_rate": 7.802874348489887e-06,
+      "loss": 1.2828,
+      "step": 990
+    },
+    {
+      "epoch": 1.902544407105137,
+      "grad_norm": 0.5961190818156942,
+      "learning_rate": 7.797322420073602e-06,
+      "loss": 1.2956,
+      "step": 991
+    },
+    {
+      "epoch": 1.9044647143542965,
+      "grad_norm": 0.6595286029545728,
+      "learning_rate": 7.791765466762308e-06,
+      "loss": 1.4421,
+      "step": 992
+    },
+    {
+      "epoch": 1.9063850216034566,
+      "grad_norm": 0.6296949876578616,
+      "learning_rate": 7.786203498538094e-06,
+      "loss": 1.352,
+      "step": 993
+    },
+    {
+      "epoch": 1.9083053288526164,
+      "grad_norm": 0.6073618042792093,
+      "learning_rate": 7.780636525392047e-06,
+      "loss": 1.2247,
+      "step": 994
+    },
+    {
+      "epoch": 1.9102256361017762,
+      "grad_norm": 0.6617452539336137,
+      "learning_rate": 7.775064557324251e-06,
+      "loss": 1.2457,
+      "step": 995
+    },
+    {
+      "epoch": 1.9121459433509362,
+      "grad_norm": 0.6536089335974361,
+      "learning_rate": 7.769487604343761e-06,
+      "loss": 1.0969,
+      "step": 996
+    },
+    {
+      "epoch": 1.914066250600096,
+      "grad_norm": 0.6568241606553313,
+      "learning_rate": 7.76390567646859e-06,
+      "loss": 1.2838,
+      "step": 997
+    },
+    {
+      "epoch": 1.9159865578492559,
+      "grad_norm": 0.6286384085429085,
+      "learning_rate": 7.758318783725678e-06,
+      "loss": 1.1126,
+      "step": 998
+    },
+    {
+      "epoch": 1.917906865098416,
+      "grad_norm": 0.6059610606241428,
+      "learning_rate": 7.752726936150895e-06,
+      "loss": 1.3271,
+      "step": 999
+    },
+    {
+      "epoch": 1.9198271723475755,
+      "grad_norm": 0.6494214620118615,
+      "learning_rate": 7.747130143789006e-06,
+      "loss": 1.4887,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9217474795967355,
+      "grad_norm": 0.6087877694738525,
+      "learning_rate": 7.741528416693656e-06,
+      "loss": 1.2577,
+      "step": 1001
+    },
+    {
+      "epoch": 1.9236677868458953,
+      "grad_norm": 0.6037766820835285,
+      "learning_rate": 7.73592176492736e-06,
+      "loss": 1.3029,
+      "step": 1002
+    },
+    {
+      "epoch": 1.9255880940950552,
+      "grad_norm": 0.642573465220044,
+      "learning_rate": 7.73031019856147e-06,
+      "loss": 1.3448,
+      "step": 1003
+    },
+    {
+      "epoch": 1.9275084013442152,
+      "grad_norm": 0.7197151370634155,
+      "learning_rate": 7.724693727676181e-06,
+      "loss": 1.2893,
+      "step": 1004
+    },
+    {
+      "epoch": 1.9294287085933748,
+      "grad_norm": 0.7163014753258418,
+      "learning_rate": 7.719072362360482e-06,
+      "loss": 1.2781,
+      "step": 1005
+    },
+    {
+      "epoch": 1.9313490158425348,
+      "grad_norm": 0.6339725691672796,
+      "learning_rate": 7.71344611271217e-06,
+      "loss": 1.2689,
+      "step": 1006
+    },
+    {
+      "epoch": 1.9332693230916946,
+      "grad_norm": 0.6255558475512056,
+      "learning_rate": 7.707814988837798e-06,
+      "loss": 1.405,
+      "step": 1007
+    },
+    {
+      "epoch": 1.9351896303408544,
+      "grad_norm": 0.7454922422525622,
+      "learning_rate": 7.702179000852693e-06,
+      "loss": 1.4645,
+      "step": 1008
+    },
+    {
+      "epoch": 1.9371099375900145,
+      "grad_norm": 0.6652688810974042,
+      "learning_rate": 7.696538158880905e-06,
+      "loss": 1.2935,
+      "step": 1009
+    },
+    {
+      "epoch": 1.9390302448391743,
+      "grad_norm": 0.7472100254077061,
+      "learning_rate": 7.69089247305521e-06,
+      "loss": 1.1844,
+      "step": 1010
+    },
+    {
+      "epoch": 1.9409505520883341,
+      "grad_norm": 0.6281178136625541,
+      "learning_rate": 7.685241953517084e-06,
+      "loss": 1.2398,
+      "step": 1011
+    },
+    {
+      "epoch": 1.9428708593374941,
+      "grad_norm": 0.7255326505427592,
+      "learning_rate": 7.679586610416689e-06,
+      "loss": 1.169,
+      "step": 1012
+    },
+    {
+      "epoch": 1.9447911665866537,
+      "grad_norm": 0.6110290026538456,
+      "learning_rate": 7.673926453912846e-06,
+      "loss": 1.2312,
+      "step": 1013
+    },
+    {
+      "epoch": 1.9467114738358138,
+      "grad_norm": 0.5927911874653067,
+      "learning_rate": 7.668261494173024e-06,
+      "loss": 1.4202,
+      "step": 1014
+    },
+    {
+      "epoch": 1.9486317810849736,
+      "grad_norm": 0.8083622338160641,
+      "learning_rate": 7.662591741373324e-06,
+      "loss": 1.1196,
+      "step": 1015
+    },
+    {
+      "epoch": 1.9505520883341334,
+      "grad_norm": 0.5787538666088754,
+      "learning_rate": 7.656917205698452e-06,
+      "loss": 1.185,
+      "step": 1016
+    },
+    {
+      "epoch": 1.9524723955832934,
+      "grad_norm": 0.5986617852156635,
+      "learning_rate": 7.65123789734171e-06,
+      "loss": 1.3589,
+      "step": 1017
+    },
+    {
+      "epoch": 1.954392702832453,
+      "grad_norm": 0.7092631074096405,
+      "learning_rate": 7.64555382650497e-06,
+      "loss": 1.447,
+      "step": 1018
+    },
+    {
+      "epoch": 1.956313010081613,
+      "grad_norm": 0.6945189144617475,
+      "learning_rate": 7.639865003398659e-06,
+      "loss": 1.38,
+      "step": 1019
+    },
+    {
+      "epoch": 1.9582333173307729,
+      "grad_norm": 0.7154207337450551,
+      "learning_rate": 7.634171438241745e-06,
+      "loss": 1.0459,
+      "step": 1020
+    },
+    {
+      "epoch": 1.9601536245799327,
+      "grad_norm": 0.5671478613388374,
+      "learning_rate": 7.628473141261704e-06,
+      "loss": 1.2583,
+      "step": 1021
+    },
+    {
+      "epoch": 1.9620739318290927,
+      "grad_norm": 0.674217778263447,
+      "learning_rate": 7.622770122694526e-06,
+      "loss": 1.2944,
+      "step": 1022
+    },
+    {
+      "epoch": 1.9639942390782525,
+      "grad_norm": 0.6035487994586758,
+      "learning_rate": 7.617062392784672e-06,
+      "loss": 1.2888,
+      "step": 1023
+    },
+    {
+      "epoch": 1.9659145463274124,
+      "grad_norm": 0.715233119055766,
+      "learning_rate": 7.61134996178507e-06,
+      "loss": 1.4473,
+      "step": 1024
+    },
+    {
+      "epoch": 1.9678348535765724,
+      "grad_norm": 0.6781200282810221,
+      "learning_rate": 7.605632839957091e-06,
+      "loss": 1.4216,
+      "step": 1025
+    },
+    {
+      "epoch": 1.969755160825732,
+      "grad_norm": 0.7348434559650923,
+      "learning_rate": 7.599911037570533e-06,
+      "loss": 1.296,
+      "step": 1026
+    },
+    {
+      "epoch": 1.971675468074892,
+      "grad_norm": 0.7174502974503351,
+      "learning_rate": 7.594184564903605e-06,
+      "loss": 1.4189,
+      "step": 1027
+    },
+    {
+      "epoch": 1.9735957753240518,
+      "grad_norm": 0.6554843561695753,
+      "learning_rate": 7.588453432242899e-06,
+      "loss": 1.1638,
+      "step": 1028
+    },
+    {
+      "epoch": 1.9755160825732117,
+      "grad_norm": 0.6193717120764839,
+      "learning_rate": 7.5827176498833844e-06,
+      "loss": 1.2138,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9774363898223717,
+      "grad_norm": 0.6230337020055144,
+      "learning_rate": 7.576977228128377e-06,
+      "loss": 1.496,
+      "step": 1030
+    },
+    {
+      "epoch": 1.9793566970715313,
+      "grad_norm": 0.7516886536978838,
+      "learning_rate": 7.57123217728953e-06,
+      "loss": 1.3157,
+      "step": 1031
+    },
+    {
+      "epoch": 1.9812770043206913,
+      "grad_norm": 0.7082303577635574,
+      "learning_rate": 7.5654825076868124e-06,
+      "loss": 1.1099,
+      "step": 1032
+    },
+    {
+      "epoch": 1.9831973115698511,
+      "grad_norm": 0.630636438554523,
+      "learning_rate": 7.559728229648489e-06,
+      "loss": 1.3774,
+      "step": 1033
+    },
+    {
+      "epoch": 1.985117618819011,
+      "grad_norm": 0.6242705516766209,
+      "learning_rate": 7.553969353511099e-06,
+      "loss": 1.2796,
+      "step": 1034
+    },
+    {
+      "epoch": 1.987037926068171,
+      "grad_norm": 0.6929200704156835,
+      "learning_rate": 7.5482058896194476e-06,
+      "loss": 1.1909,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9889582333173308,
+      "grad_norm": 0.674947851456176,
+      "learning_rate": 7.5424378483265795e-06,
+      "loss": 1.2089,
+      "step": 1036
+    },
+    {
+      "epoch": 1.9908785405664906,
+      "grad_norm": 0.6655740470860053,
+      "learning_rate": 7.536665239993759e-06,
+      "loss": 1.3069,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9927988478156506,
+      "grad_norm": 0.7234506954218087,
+      "learning_rate": 7.5308880749904576e-06,
+      "loss": 1.3131,
+      "step": 1038
+    },
+    {
+      "epoch": 1.9947191550648102,
+      "grad_norm": 0.6336720269563768,
+      "learning_rate": 7.525106363694328e-06,
+      "loss": 1.2551,
+      "step": 1039
+    },
+    {
+      "epoch": 1.9966394623139703,
+      "grad_norm": 0.6654745873292405,
+      "learning_rate": 7.519320116491195e-06,
+      "loss": 1.2598,
+      "step": 1040
+    },
+    {
+      "epoch": 1.99855976956313,
+      "grad_norm": 0.6416056547121506,
+      "learning_rate": 7.513529343775025e-06,
+      "loss": 1.3895,
+      "step": 1041
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7381453276807621,
+      "learning_rate": 7.50773405594792e-06,
+      "loss": 1.1004,
+      "step": 1042
+    },
+    {
+      "epoch": 2.00192030724916,
+      "grad_norm": 0.6984585189454183,
+      "learning_rate": 7.50193426342009e-06,
+      "loss": 1.2338,
+      "step": 1043
+    },
+    {
+      "epoch": 2.0038406144983196,
+      "grad_norm": 0.6860005251072618,
+      "learning_rate": 7.496129976609833e-06,
+      "loss": 1.1611,
+      "step": 1044
+    },
+    {
+      "epoch": 2.0057609217474797,
+      "grad_norm": 0.6710341040166462,
+      "learning_rate": 7.490321205943526e-06,
+      "loss": 1.1453,
+      "step": 1045
+    },
+    {
+      "epoch": 2.0076812289966393,
+      "grad_norm": 0.6881057248251186,
+      "learning_rate": 7.484507961855599e-06,
+      "loss": 1.3585,
+      "step": 1046
+    },
+    {
+      "epoch": 2.0096015362457993,
+      "grad_norm": 0.7149692590237099,
+      "learning_rate": 7.478690254788515e-06,
+      "loss": 1.3108,
+      "step": 1047
+    },
+    {
+      "epoch": 2.0115218434949593,
+      "grad_norm": 0.5040575852249538,
+      "learning_rate": 7.472868095192758e-06,
+      "loss": 1.2666,
+      "step": 1048
+    },
+    {
+      "epoch": 2.013442150744119,
+      "grad_norm": 0.6668343259353099,
+      "learning_rate": 7.467041493526806e-06,
+      "loss": 1.3017,
+      "step": 1049
+    },
+    {
+      "epoch": 2.015362457993279,
+      "grad_norm": 0.6213130267441712,
+      "learning_rate": 7.46121046025712e-06,
+      "loss": 1.3463,
+      "step": 1050
+    },
+    {
+      "epoch": 2.017282765242439,
+      "grad_norm": 0.5465921578009497,
+      "learning_rate": 7.4553750058581186e-06,
+      "loss": 1.1172,
+      "step": 1051
+    },
+    {
+      "epoch": 2.0192030724915986,
+      "grad_norm": 0.7058326054289525,
+      "learning_rate": 7.449535140812164e-06,
+      "loss": 1.0875,
+      "step": 1052
+    },
+    {
+      "epoch": 2.0211233797407586,
+      "grad_norm": 0.6224285545020704,
+      "learning_rate": 7.443690875609543e-06,
+      "loss": 1.2569,
+      "step": 1053
+    },
+    {
+      "epoch": 2.023043686989918,
+      "grad_norm": 0.5891570753306204,
+      "learning_rate": 7.437842220748441e-06,
+      "loss": 1.2506,
+      "step": 1054
+    },
+    {
+      "epoch": 2.0249639942390782,
+      "grad_norm": 0.7187156173332934,
+      "learning_rate": 7.431989186734935e-06,
+      "loss": 1.1866,
+      "step": 1055
+    },
+    {
+      "epoch": 2.0268843014882383,
+      "grad_norm": 0.656805950152783,
+      "learning_rate": 7.4261317840829635e-06,
+      "loss": 1.2794,
+      "step": 1056
+    },
+    {
+      "epoch": 2.028804608737398,
+      "grad_norm": 0.6565140993598239,
+      "learning_rate": 7.420270023314315e-06,
+      "loss": 1.4964,
+      "step": 1057
+    },
+    {
+      "epoch": 2.030724915986558,
+      "grad_norm": 0.6674673642468827,
+      "learning_rate": 7.414403914958607e-06,
+      "loss": 1.1226,
+      "step": 1058
+    },
+    {
+      "epoch": 2.0326452232357175,
+      "grad_norm": 0.6383574520864321,
+      "learning_rate": 7.408533469553264e-06,
+      "loss": 1.276,
+      "step": 1059
+    },
+    {
+      "epoch": 2.0345655304848775,
+      "grad_norm": 0.7154715475803043,
+      "learning_rate": 7.402658697643504e-06,
+      "loss": 1.2388,
+      "step": 1060
+    },
+    {
+      "epoch": 2.0364858377340376,
+      "grad_norm": 0.5223568124972519,
+      "learning_rate": 7.396779609782316e-06,
+      "loss": 1.2159,
+      "step": 1061
+    },
+    {
+      "epoch": 2.038406144983197,
+      "grad_norm": 0.5739489933046834,
+      "learning_rate": 7.390896216530442e-06,
+      "loss": 1.2659,
+      "step": 1062
+    },
+    {
+      "epoch": 2.040326452232357,
+      "grad_norm": 0.6737444843346947,
+      "learning_rate": 7.385008528456357e-06,
+      "loss": 1.3207,
+      "step": 1063
+    },
+    {
+      "epoch": 2.0422467594815172,
+      "grad_norm": 0.5760491490696834,
+      "learning_rate": 7.379116556136251e-06,
+      "loss": 1.2992,
+      "step": 1064
+    },
+    {
+      "epoch": 2.044167066730677,
+      "grad_norm": 0.656419511670537,
+      "learning_rate": 7.373220310154008e-06,
+      "loss": 1.2658,
+      "step": 1065
+    },
+    {
+      "epoch": 2.046087373979837,
+      "grad_norm": 0.6150154633631958,
+      "learning_rate": 7.367319801101196e-06,
+      "loss": 1.3401,
+      "step": 1066
+    },
+    {
+      "epoch": 2.0480076812289965,
+      "grad_norm": 1.021222823586059,
+      "learning_rate": 7.361415039577033e-06,
+      "loss": 1.2002,
+      "step": 1067
+    },
+    {
+      "epoch": 2.0499279884781565,
+      "grad_norm": 0.5854320877750915,
+      "learning_rate": 7.355506036188379e-06,
+      "loss": 1.1323,
+      "step": 1068
+    },
+    {
+      "epoch": 2.0518482957273165,
+      "grad_norm": 0.7216867130918194,
+      "learning_rate": 7.349592801549715e-06,
+      "loss": 1.2601,
+      "step": 1069
+    },
+    {
+      "epoch": 2.053768602976476,
+      "grad_norm": 0.6194054541902049,
+      "learning_rate": 7.343675346283118e-06,
+      "loss": 1.2808,
+      "step": 1070
+    },
+    {
+      "epoch": 2.055688910225636,
+      "grad_norm": 0.9039301133482249,
+      "learning_rate": 7.337753681018251e-06,
+      "loss": 1.2842,
+      "step": 1071
+    },
+    {
+      "epoch": 2.0576092174747957,
+      "grad_norm": 0.5876501457594527,
+      "learning_rate": 7.331827816392341e-06,
+      "loss": 1.221,
+      "step": 1072
+    },
+    {
+      "epoch": 2.059529524723956,
+      "grad_norm": 0.6740935986291116,
+      "learning_rate": 7.325897763050155e-06,
+      "loss": 1.267,
+      "step": 1073
+    },
+    {
+      "epoch": 2.061449831973116,
+      "grad_norm": 0.6337403720742513,
+      "learning_rate": 7.319963531643983e-06,
+      "loss": 1.2348,
+      "step": 1074
+    },
+    {
+      "epoch": 2.0633701392222754,
+      "grad_norm": 0.6354665055103681,
+      "learning_rate": 7.3140251328336234e-06,
+      "loss": 1.2987,
+      "step": 1075
+    },
+    {
+      "epoch": 2.0652904464714354,
+      "grad_norm": 0.5933193811451509,
+      "learning_rate": 7.308082577286359e-06,
+      "loss": 1.2048,
+      "step": 1076
+    },
+    {
+      "epoch": 2.0672107537205955,
+      "grad_norm": 0.7623792540828787,
+      "learning_rate": 7.3021358756769425e-06,
+      "loss": 1.2341,
+      "step": 1077
+    },
+    {
+      "epoch": 2.069131060969755,
+      "grad_norm": 0.6015392053110737,
+      "learning_rate": 7.296185038687566e-06,
+      "loss": 1.2542,
+      "step": 1078
+    },
+    {
+      "epoch": 2.071051368218915,
+      "grad_norm": 0.6955251812505019,
+      "learning_rate": 7.290230077007863e-06,
+      "loss": 1.1749,
+      "step": 1079
+    },
+    {
+      "epoch": 2.0729716754680747,
+      "grad_norm": 0.6087755249687018,
+      "learning_rate": 7.284271001334862e-06,
+      "loss": 1.2355,
+      "step": 1080
+    },
+    {
+      "epoch": 2.0748919827172347,
+      "grad_norm": 0.5589403698261416,
+      "learning_rate": 7.278307822372992e-06,
+      "loss": 1.2345,
+      "step": 1081
+    },
+    {
+      "epoch": 2.0768122899663948,
+      "grad_norm": 0.6219252458630142,
+      "learning_rate": 7.272340550834049e-06,
+      "loss": 1.2623,
+      "step": 1082
+    },
+    {
+      "epoch": 2.0787325972155544,
+      "grad_norm": 0.6568214702123812,
+      "learning_rate": 7.266369197437182e-06,
+      "loss": 1.2551,
+      "step": 1083
+    },
+    {
+      "epoch": 2.0806529044647144,
+      "grad_norm": 0.6644980261981873,
+      "learning_rate": 7.26039377290887e-06,
+      "loss": 1.2986,
+      "step": 1084
+    },
+    {
+      "epoch": 2.082573211713874,
+      "grad_norm": 0.652931545403877,
+      "learning_rate": 7.254414287982907e-06,
+      "loss": 1.2229,
+      "step": 1085
+    },
+    {
+      "epoch": 2.084493518963034,
+      "grad_norm": 0.5965090727981581,
+      "learning_rate": 7.24843075340038e-06,
+      "loss": 1.187,
+      "step": 1086
+    },
+    {
+      "epoch": 2.086413826212194,
+      "grad_norm": 0.6636455612362417,
+      "learning_rate": 7.242443179909649e-06,
+      "loss": 1.2377,
+      "step": 1087
+    },
+    {
+      "epoch": 2.0883341334613537,
+      "grad_norm": 0.5706280085675597,
+      "learning_rate": 7.236451578266334e-06,
+      "loss": 1.3363,
+      "step": 1088
+    },
+    {
+      "epoch": 2.0902544407105137,
+      "grad_norm": 0.5452970868510953,
+      "learning_rate": 7.230455959233284e-06,
+      "loss": 1.2301,
+      "step": 1089
+    },
+    {
+      "epoch": 2.0921747479596737,
+      "grad_norm": 0.732832750259065,
+      "learning_rate": 7.224456333580574e-06,
+      "loss": 1.2697,
+      "step": 1090
+    },
+    {
+      "epoch": 2.0940950552088333,
+      "grad_norm": 0.5776781009920395,
+      "learning_rate": 7.218452712085464e-06,
+      "loss": 1.1108,
+      "step": 1091
+    },
+    {
+      "epoch": 2.0960153624579934,
+      "grad_norm": 0.7340170720293495,
+      "learning_rate": 7.212445105532402e-06,
+      "loss": 1.4598,
+      "step": 1092
+    },
+    {
+      "epoch": 2.097935669707153,
+      "grad_norm": 0.5608573831708831,
+      "learning_rate": 7.206433524712989e-06,
+      "loss": 1.3114,
+      "step": 1093
+    },
+    {
+      "epoch": 2.099855976956313,
+      "grad_norm": 0.6123136642885363,
+      "learning_rate": 7.200417980425969e-06,
+      "loss": 1.176,
+      "step": 1094
+    },
+    {
+      "epoch": 2.101776284205473,
+      "grad_norm": 0.6263758820886606,
+      "learning_rate": 7.1943984834771995e-06,
+      "loss": 1.3055,
+      "step": 1095
+    },
+    {
+      "epoch": 2.1036965914546326,
+      "grad_norm": 0.6301511028435693,
+      "learning_rate": 7.188375044679645e-06,
+      "loss": 1.2084,
+      "step": 1096
+    },
+    {
+      "epoch": 2.1056168987037926,
+      "grad_norm": 0.5961686912489383,
+      "learning_rate": 7.182347674853349e-06,
+      "loss": 1.1919,
+      "step": 1097
+    },
+    {
+      "epoch": 2.1075372059529527,
+      "grad_norm": 0.6383316351170043,
+      "learning_rate": 7.176316384825414e-06,
+      "loss": 1.2242,
+      "step": 1098
+    },
+    {
+      "epoch": 2.1094575132021123,
+      "grad_norm": 0.6832418991450104,
+      "learning_rate": 7.170281185429986e-06,
+      "loss": 1.2969,
+      "step": 1099
+    },
+    {
+      "epoch": 2.1113778204512723,
+      "grad_norm": 0.7003707945546503,
+      "learning_rate": 7.164242087508232e-06,
+      "loss": 1.3452,
+      "step": 1100
+    },
+    {
+      "epoch": 2.113298127700432,
+      "grad_norm": 0.6524324557488448,
+      "learning_rate": 7.1581991019083255e-06,
+      "loss": 1.2907,
+      "step": 1101
+    },
+    {
+      "epoch": 2.115218434949592,
+      "grad_norm": 0.6346709626357138,
+      "learning_rate": 7.152152239485419e-06,
+      "loss": 1.2925,
+      "step": 1102
+    },
+    {
+      "epoch": 2.117138742198752,
+      "grad_norm": 0.6027369910884686,
+      "learning_rate": 7.1461015111016365e-06,
+      "loss": 1.2191,
+      "step": 1103
+    },
+    {
+      "epoch": 2.1190590494479116,
+      "grad_norm": 0.7317096771295276,
+      "learning_rate": 7.140046927626034e-06,
+      "loss": 1.2868,
+      "step": 1104
+    },
+    {
+      "epoch": 2.1209793566970716,
+      "grad_norm": 0.664283102964385,
+      "learning_rate": 7.1339884999346065e-06,
+      "loss": 1.2537,
+      "step": 1105
+    },
+    {
+      "epoch": 2.122899663946231,
+      "grad_norm": 0.5617222908710969,
+      "learning_rate": 7.127926238910243e-06,
+      "loss": 1.203,
+      "step": 1106
+    },
+    {
+      "epoch": 2.1248199711953912,
+      "grad_norm": 0.6937789147849563,
+      "learning_rate": 7.121860155442727e-06,
+      "loss": 1.18,
+      "step": 1107
+    },
+    {
+      "epoch": 2.1267402784445513,
+      "grad_norm": 0.6090841223763318,
+      "learning_rate": 7.115790260428704e-06,
+      "loss": 1.2649,
+      "step": 1108
+    },
+    {
+      "epoch": 2.128660585693711,
+      "grad_norm": 0.5913292372532395,
+      "learning_rate": 7.109716564771663e-06,
+      "loss": 1.3238,
+      "step": 1109
+    },
+    {
+      "epoch": 2.130580892942871,
+      "grad_norm": 0.6373338530440186,
+      "learning_rate": 7.103639079381931e-06,
+      "loss": 1.3145,
+      "step": 1110
+    },
+    {
+      "epoch": 2.132501200192031,
+      "grad_norm": 0.7356254332331955,
+      "learning_rate": 7.09755781517663e-06,
+      "loss": 1.3062,
+      "step": 1111
+    },
+    {
+      "epoch": 2.1344215074411905,
+      "grad_norm": 0.6354999779911574,
+      "learning_rate": 7.091472783079677e-06,
+      "loss": 1.2691,
+      "step": 1112
+    },
+    {
+      "epoch": 2.1363418146903506,
+      "grad_norm": 0.634333016129095,
+      "learning_rate": 7.085383994021757e-06,
+      "loss": 1.2463,
+      "step": 1113
+    },
+    {
+      "epoch": 2.13826212193951,
+      "grad_norm": 0.628640631103715,
+      "learning_rate": 7.079291458940302e-06,
+      "loss": 1.3555,
+      "step": 1114
+    },
+    {
+      "epoch": 2.14018242918867,
+      "grad_norm": 0.5878511095609504,
+      "learning_rate": 7.073195188779474e-06,
+      "loss": 1.2021,
+      "step": 1115
+    },
+    {
+      "epoch": 2.1421027364378302,
+      "grad_norm": 0.6234889678557616,
+      "learning_rate": 7.067095194490143e-06,
+      "loss": 1.2481,
+      "step": 1116
+    },
+    {
+      "epoch": 2.14402304368699,
+      "grad_norm": 0.5849361230078369,
+      "learning_rate": 7.060991487029872e-06,
+      "loss": 1.148,
+      "step": 1117
+    },
+    {
+      "epoch": 2.14594335093615,
+      "grad_norm": 0.5677121596010603,
+      "learning_rate": 7.0548840773628915e-06,
+      "loss": 1.2142,
+      "step": 1118
+    },
+    {
+      "epoch": 2.1478636581853094,
+      "grad_norm": 0.6134358379808856,
+      "learning_rate": 7.0487729764600824e-06,
+      "loss": 1.0703,
+      "step": 1119
+    },
+    {
+      "epoch": 2.1497839654344695,
+      "grad_norm": 0.5806561061052112,
+      "learning_rate": 7.042658195298956e-06,
+      "loss": 1.3193,
+      "step": 1120
+    },
+    {
+      "epoch": 2.1517042726836295,
+      "grad_norm": 0.6169966362309864,
+      "learning_rate": 7.036539744863636e-06,
+      "loss": 1.266,
+      "step": 1121
+    },
+    {
+      "epoch": 2.153624579932789,
+      "grad_norm": 0.6281321250720722,
+      "learning_rate": 7.030417636144836e-06,
+      "loss": 1.3127,
+      "step": 1122
+    },
+    {
+      "epoch": 2.155544887181949,
+      "grad_norm": 0.552552267873573,
+      "learning_rate": 7.024291880139843e-06,
+      "loss": 1.1968,
+      "step": 1123
+    },
+    {
+      "epoch": 2.157465194431109,
+      "grad_norm": 0.7742167483021233,
+      "learning_rate": 7.018162487852494e-06,
+      "loss": 1.2514,
+      "step": 1124
+    },
+    {
+      "epoch": 2.1593855016802688,
+      "grad_norm": 0.6102221138856729,
+      "learning_rate": 7.012029470293156e-06,
+      "loss": 1.2888,
+      "step": 1125
+    },
+    {
+      "epoch": 2.161305808929429,
+      "grad_norm": 0.678988077797767,
+      "learning_rate": 7.0058928384787115e-06,
+      "loss": 1.2372,
+      "step": 1126
+    },
+    {
+      "epoch": 2.1632261161785884,
+      "grad_norm": 0.6737189312331883,
+      "learning_rate": 6.999752603432534e-06,
+      "loss": 1.1838,
+      "step": 1127
+    },
+    {
+      "epoch": 2.1651464234277484,
+      "grad_norm": 0.596052318398853,
+      "learning_rate": 6.993608776184473e-06,
+      "loss": 1.2675,
+      "step": 1128
+    },
+    {
+      "epoch": 2.1670667306769085,
+      "grad_norm": 0.5778382747735666,
+      "learning_rate": 6.987461367770825e-06,
+      "loss": 1.1558,
+      "step": 1129
+    },
+    {
+      "epoch": 2.168987037926068,
+      "grad_norm": 0.7612550327008405,
+      "learning_rate": 6.9813103892343205e-06,
+      "loss": 1.1798,
+      "step": 1130
+    },
+    {
+      "epoch": 2.170907345175228,
+      "grad_norm": 0.6250451986476963,
+      "learning_rate": 6.975155851624107e-06,
+      "loss": 1.2171,
+      "step": 1131
+    },
+    {
+      "epoch": 2.172827652424388,
+      "grad_norm": 0.617676347733231,
+      "learning_rate": 6.968997765995722e-06,
+      "loss": 1.2002,
+      "step": 1132
+    },
+    {
+      "epoch": 2.1747479596735477,
+      "grad_norm": 0.7215322900975272,
+      "learning_rate": 6.962836143411077e-06,
+      "loss": 1.1815,
+      "step": 1133
+    },
+    {
+      "epoch": 2.1766682669227078,
+      "grad_norm": 0.6889525154465694,
+      "learning_rate": 6.956670994938438e-06,
+      "loss": 1.2466,
+      "step": 1134
+    },
+    {
+      "epoch": 2.1785885741718674,
+      "grad_norm": 0.5751161226801118,
+      "learning_rate": 6.9505023316524024e-06,
+      "loss": 1.2167,
+      "step": 1135
+    },
+    {
+      "epoch": 2.1805088814210274,
+      "grad_norm": 0.6513915776200747,
+      "learning_rate": 6.944330164633886e-06,
+      "loss": 1.3228,
+      "step": 1136
+    },
+    {
+      "epoch": 2.1824291886701874,
+      "grad_norm": 0.6003888012093858,
+      "learning_rate": 6.938154504970092e-06,
+      "loss": 1.2462,
+      "step": 1137
+    },
+    {
+      "epoch": 2.184349495919347,
+      "grad_norm": 0.6206783890336667,
+      "learning_rate": 6.931975363754502e-06,
+      "loss": 1.291,
+      "step": 1138
+    },
+    {
+      "epoch": 2.186269803168507,
+      "grad_norm": 0.7554272133869872,
+      "learning_rate": 6.92579275208685e-06,
+      "loss": 1.1497,
+      "step": 1139
+    },
+    {
+      "epoch": 2.1881901104176666,
+      "grad_norm": 0.6236041362503416,
+      "learning_rate": 6.9196066810731055e-06,
+      "loss": 1.2785,
+      "step": 1140
+    },
+    {
+      "epoch": 2.1901104176668267,
+      "grad_norm": 0.5902897052882286,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 1.2064,
+      "step": 1141
+    },
+    {
+      "epoch": 2.1920307249159867,
+      "grad_norm": 0.6788328193144818,
+      "learning_rate": 6.90722420546226e-06,
+      "loss": 1.201,
+      "step": 1142
+    },
+    {
+      "epoch": 2.1939510321651463,
+      "grad_norm": 0.7102736450080036,
+      "learning_rate": 6.901027823108088e-06,
+      "loss": 1.1767,
+      "step": 1143
+    },
+    {
+      "epoch": 2.1958713394143063,
+      "grad_norm": 0.6002222918018697,
+      "learning_rate": 6.894828025893636e-06,
+      "loss": 1.2062,
+      "step": 1144
+    },
+    {
+      "epoch": 2.1977916466634664,
+      "grad_norm": 0.5307888776119263,
+      "learning_rate": 6.888624824955746e-06,
+      "loss": 1.1376,
+      "step": 1145
+    },
+    {
+      "epoch": 2.199711953912626,
+      "grad_norm": 0.6991995097343485,
+      "learning_rate": 6.882418231437371e-06,
+      "loss": 1.2353,
+      "step": 1146
+    },
+    {
+      "epoch": 2.201632261161786,
+      "grad_norm": 0.5955303781294875,
+      "learning_rate": 6.87620825648756e-06,
+      "loss": 1.14,
+      "step": 1147
+    },
+    {
+      "epoch": 2.2035525684109456,
+      "grad_norm": 0.6099852257845632,
+      "learning_rate": 6.869994911261429e-06,
+      "loss": 1.354,
+      "step": 1148
+    },
+    {
+      "epoch": 2.2054728756601056,
+      "grad_norm": 0.6184213161381115,
+      "learning_rate": 6.863778206920161e-06,
+      "loss": 1.228,
+      "step": 1149
+    },
+    {
+      "epoch": 2.2073931829092657,
+      "grad_norm": 0.6632479032393596,
+      "learning_rate": 6.8575581546309614e-06,
+      "loss": 1.19,
+      "step": 1150
+    },
+    {
+      "epoch": 2.2093134901584253,
+      "grad_norm": 0.736023733055779,
+      "learning_rate": 6.851334765567054e-06,
+      "loss": 1.2977,
+      "step": 1151
+    },
+    {
+      "epoch": 2.2112337974075853,
+      "grad_norm": 0.5801625495939454,
+      "learning_rate": 6.8451080509076594e-06,
+      "loss": 1.3641,
+      "step": 1152
+    },
+    {
+      "epoch": 2.213154104656745,
+      "grad_norm": 0.5864932647879217,
+      "learning_rate": 6.838878021837968e-06,
+      "loss": 1.0784,
+      "step": 1153
+    },
+    {
+      "epoch": 2.215074411905905,
+      "grad_norm": 0.6816487618025293,
+      "learning_rate": 6.832644689549124e-06,
+      "loss": 1.3253,
+      "step": 1154
+    },
+    {
+      "epoch": 2.216994719155065,
+      "grad_norm": 0.685898447688162,
+      "learning_rate": 6.826408065238208e-06,
+      "loss": 1.3172,
+      "step": 1155
+    },
+    {
+      "epoch": 2.2189150264042246,
+      "grad_norm": 0.6181285906434999,
+      "learning_rate": 6.820168160108211e-06,
+      "loss": 1.3543,
+      "step": 1156
+    },
+    {
+      "epoch": 2.2208353336533846,
+      "grad_norm": 0.6573999090601903,
+      "learning_rate": 6.813924985368021e-06,
+      "loss": 1.1342,
+      "step": 1157
+    },
+    {
+      "epoch": 2.2227556409025446,
+      "grad_norm": 0.6345506761790236,
+      "learning_rate": 6.807678552232397e-06,
+      "loss": 1.3285,
+      "step": 1158
+    },
+    {
+      "epoch": 2.224675948151704,
+      "grad_norm": 0.5910829243317203,
+      "learning_rate": 6.8014288719219505e-06,
+      "loss": 1.2601,
+      "step": 1159
+    },
+    {
+      "epoch": 2.2265962554008643,
+      "grad_norm": 0.5952533335690797,
+      "learning_rate": 6.795175955663127e-06,
+      "loss": 1.2067,
+      "step": 1160
+    },
+    {
+      "epoch": 2.228516562650024,
+      "grad_norm": 0.5720846452241152,
+      "learning_rate": 6.788919814688183e-06,
+      "loss": 1.2387,
+      "step": 1161
+    },
+    {
+      "epoch": 2.230436869899184,
+      "grad_norm": 0.6051635760894251,
+      "learning_rate": 6.782660460235174e-06,
+      "loss": 1.198,
+      "step": 1162
+    },
+    {
+      "epoch": 2.232357177148344,
+      "grad_norm": 0.6808334105131638,
+      "learning_rate": 6.776397903547919e-06,
+      "loss": 1.2227,
+      "step": 1163
+    },
+    {
+      "epoch": 2.2342774843975035,
+      "grad_norm": 0.6930823511105796,
+      "learning_rate": 6.770132155875994e-06,
+      "loss": 1.2159,
+      "step": 1164
+    },
+    {
+      "epoch": 2.2361977916466635,
+      "grad_norm": 0.6416443437211603,
+      "learning_rate": 6.76386322847471e-06,
+      "loss": 1.3446,
+      "step": 1165
+    },
+    {
+      "epoch": 2.238118098895823,
+      "grad_norm": 0.7018085526370589,
+      "learning_rate": 6.757591132605082e-06,
+      "loss": 1.3883,
+      "step": 1166
+    },
+    {
+      "epoch": 2.240038406144983,
+      "grad_norm": 0.5811582821637894,
+      "learning_rate": 6.7513158795338245e-06,
+      "loss": 1.2509,
+      "step": 1167
+    },
+    {
+      "epoch": 2.241958713394143,
+      "grad_norm": 0.6921861009803321,
+      "learning_rate": 6.745037480533316e-06,
+      "loss": 1.2685,
+      "step": 1168
+    },
+    {
+      "epoch": 2.243879020643303,
+      "grad_norm": 0.6003605885065617,
+      "learning_rate": 6.738755946881593e-06,
+      "loss": 1.1512,
+      "step": 1169
+    },
+    {
+      "epoch": 2.245799327892463,
+      "grad_norm": 0.6330910082110218,
+      "learning_rate": 6.73247128986232e-06,
+      "loss": 1.3939,
+      "step": 1170
+    },
+    {
+      "epoch": 2.247719635141623,
+      "grad_norm": 0.7009718702070067,
+      "learning_rate": 6.72618352076477e-06,
+      "loss": 1.3504,
+      "step": 1171
+    },
+    {
+      "epoch": 2.2496399423907825,
+      "grad_norm": 0.6243298363689899,
+      "learning_rate": 6.7198926508838095e-06,
+      "loss": 1.2077,
+      "step": 1172
+    },
+    {
+      "epoch": 2.2515602496399425,
+      "grad_norm": 0.665060474987669,
+      "learning_rate": 6.713598691519873e-06,
+      "loss": 1.2854,
+      "step": 1173
+    },
+    {
+      "epoch": 2.253480556889102,
+      "grad_norm": 0.6795425121016735,
+      "learning_rate": 6.707301653978945e-06,
+      "loss": 1.211,
+      "step": 1174
+    },
+    {
+      "epoch": 2.255400864138262,
+      "grad_norm": 0.649787483941008,
+      "learning_rate": 6.701001549572541e-06,
+      "loss": 1.2618,
+      "step": 1175
+    },
+    {
+      "epoch": 2.257321171387422,
+      "grad_norm": 0.5530616366843604,
+      "learning_rate": 6.694698389617684e-06,
+      "loss": 1.189,
+      "step": 1176
+    },
+    {
+      "epoch": 2.2592414786365818,
+      "grad_norm": 0.7578591256854189,
+      "learning_rate": 6.688392185436884e-06,
+      "loss": 1.1721,
+      "step": 1177
+    },
+    {
+      "epoch": 2.261161785885742,
+      "grad_norm": 0.681600411779859,
+      "learning_rate": 6.682082948358125e-06,
+      "loss": 1.1017,
+      "step": 1178
+    },
+    {
+      "epoch": 2.2630820931349014,
+      "grad_norm": 0.5753391550196724,
+      "learning_rate": 6.675770689714832e-06,
+      "loss": 1.2545,
+      "step": 1179
+    },
+    {
+      "epoch": 2.2650024003840614,
+      "grad_norm": 0.5926485691603264,
+      "learning_rate": 6.6694554208458665e-06,
+      "loss": 1.3251,
+      "step": 1180
+    },
+    {
+      "epoch": 2.2669227076332215,
+      "grad_norm": 0.7701074371021729,
+      "learning_rate": 6.663137153095487e-06,
+      "loss": 1.3517,
+      "step": 1181
+    },
+    {
+      "epoch": 2.268843014882381,
+      "grad_norm": 0.6037580306390564,
+      "learning_rate": 6.6568158978133455e-06,
+      "loss": 1.1961,
+      "step": 1182
+    },
+    {
+      "epoch": 2.270763322131541,
+      "grad_norm": 0.6242293746700862,
+      "learning_rate": 6.65049166635446e-06,
+      "loss": 1.2815,
+      "step": 1183
+    },
+    {
+      "epoch": 2.272683629380701,
+      "grad_norm": 0.7092181079591579,
+      "learning_rate": 6.644164470079193e-06,
+      "loss": 1.259,
+      "step": 1184
+    },
+    {
+      "epoch": 2.2746039366298607,
+      "grad_norm": 0.7292906796102022,
+      "learning_rate": 6.637834320353235e-06,
+      "loss": 1.3405,
+      "step": 1185
+    },
+    {
+      "epoch": 2.2765242438790207,
+      "grad_norm": 0.5858173584780139,
+      "learning_rate": 6.63150122854758e-06,
+      "loss": 1.2043,
+      "step": 1186
+    },
+    {
+      "epoch": 2.2784445511281803,
+      "grad_norm": 0.6644872468479983,
+      "learning_rate": 6.625165206038504e-06,
+      "loss": 1.3984,
+      "step": 1187
+    },
+    {
+      "epoch": 2.2803648583773404,
+      "grad_norm": 0.6617391253904775,
+      "learning_rate": 6.6188262642075566e-06,
+      "loss": 1.2627,
+      "step": 1188
+    },
+    {
+      "epoch": 2.2822851656265004,
+      "grad_norm": 0.6415220686748404,
+      "learning_rate": 6.612484414441522e-06,
+      "loss": 1.2287,
+      "step": 1189
+    },
+    {
+      "epoch": 2.28420547287566,
+      "grad_norm": 0.6597124470851262,
+      "learning_rate": 6.606139668132412e-06,
+      "loss": 1.295,
+      "step": 1190
+    },
+    {
+      "epoch": 2.28612578012482,
+      "grad_norm": 0.6003215660383565,
+      "learning_rate": 6.599792036677444e-06,
+      "loss": 1.286,
+      "step": 1191
+    },
+    {
+      "epoch": 2.2880460873739796,
+      "grad_norm": 0.6684003644336577,
+      "learning_rate": 6.593441531479011e-06,
+      "loss": 1.1995,
+      "step": 1192
+    },
+    {
+      "epoch": 2.2899663946231397,
+      "grad_norm": 0.613380897872638,
+      "learning_rate": 6.587088163944677e-06,
+      "loss": 1.2017,
+      "step": 1193
+    },
+    {
+      "epoch": 2.2918867018722997,
+      "grad_norm": 0.6759779333103918,
+      "learning_rate": 6.5807319454871385e-06,
+      "loss": 1.3439,
+      "step": 1194
+    },
+    {
+      "epoch": 2.2938070091214593,
+      "grad_norm": 0.6209490809831877,
+      "learning_rate": 6.574372887524221e-06,
+      "loss": 1.4075,
+      "step": 1195
+    },
+    {
+      "epoch": 2.2957273163706193,
+      "grad_norm": 0.7296562782355499,
+      "learning_rate": 6.568011001478846e-06,
+      "loss": 1.376,
+      "step": 1196
+    },
+    {
+      "epoch": 2.2976476236197794,
+      "grad_norm": 0.7580590870766526,
+      "learning_rate": 6.561646298779019e-06,
+      "loss": 1.2304,
+      "step": 1197
+    },
+    {
+      "epoch": 2.299567930868939,
+      "grad_norm": 0.6166792793488481,
+      "learning_rate": 6.5552787908578e-06,
+      "loss": 1.2215,
+      "step": 1198
+    },
+    {
+      "epoch": 2.301488238118099,
+      "grad_norm": 0.6233271317496348,
+      "learning_rate": 6.548908489153292e-06,
+      "loss": 1.2564,
+      "step": 1199
+    },
+    {
+      "epoch": 2.3034085453672586,
+      "grad_norm": 0.6182453476486088,
+      "learning_rate": 6.542535405108614e-06,
+      "loss": 1.0683,
+      "step": 1200
+    },
+    {
+      "epoch": 2.3053288526164186,
+      "grad_norm": 0.6326303682008909,
+      "learning_rate": 6.536159550171888e-06,
+      "loss": 1.2443,
+      "step": 1201
+    },
+    {
+      "epoch": 2.3072491598655787,
+      "grad_norm": 0.6398707602190346,
+      "learning_rate": 6.5297809357962064e-06,
+      "loss": 1.1574,
+      "step": 1202
+    },
+    {
+      "epoch": 2.3091694671147383,
+      "grad_norm": 0.6852341547792379,
+      "learning_rate": 6.523399573439621e-06,
+      "loss": 1.1444,
+      "step": 1203
+    },
+    {
+      "epoch": 2.3110897743638983,
+      "grad_norm": 0.616233805909782,
+      "learning_rate": 6.517015474565127e-06,
+      "loss": 1.2797,
+      "step": 1204
+    },
+    {
+      "epoch": 2.313010081613058,
+      "grad_norm": 0.5644035791459282,
+      "learning_rate": 6.51062865064062e-06,
+      "loss": 1.1528,
+      "step": 1205
+    },
+    {
+      "epoch": 2.314930388862218,
+      "grad_norm": 0.5470958572312058,
+      "learning_rate": 6.5042391131389086e-06,
+      "loss": 1.1724,
+      "step": 1206
+    },
+    {
+      "epoch": 2.316850696111378,
+      "grad_norm": 0.8369300122733229,
+      "learning_rate": 6.49784687353766e-06,
+      "loss": 1.1318,
+      "step": 1207
+    },
+    {
+      "epoch": 2.3187710033605375,
+      "grad_norm": 0.6176806677274713,
+      "learning_rate": 6.4914519433194046e-06,
+      "loss": 1.1819,
+      "step": 1208
+    },
+    {
+      "epoch": 2.3206913106096976,
+      "grad_norm": 0.70239567156242,
+      "learning_rate": 6.485054333971505e-06,
+      "loss": 1.3287,
+      "step": 1209
+    },
+    {
+      "epoch": 2.3226116178588576,
+      "grad_norm": 0.6849156492178405,
+      "learning_rate": 6.4786540569861315e-06,
+      "loss": 1.2436,
+      "step": 1210
+    },
+    {
+      "epoch": 2.324531925108017,
+      "grad_norm": 0.623250772845043,
+      "learning_rate": 6.472251123860252e-06,
+      "loss": 1.0973,
+      "step": 1211
+    },
+    {
+      "epoch": 2.3264522323571772,
+      "grad_norm": 0.6275812836811542,
+      "learning_rate": 6.465845546095605e-06,
+      "loss": 1.1364,
+      "step": 1212
+    },
+    {
+      "epoch": 2.328372539606337,
+      "grad_norm": 0.6215409724184875,
+      "learning_rate": 6.459437335198676e-06,
+      "loss": 1.2874,
+      "step": 1213
+    },
+    {
+      "epoch": 2.330292846855497,
+      "grad_norm": 0.6379824148538079,
+      "learning_rate": 6.453026502680683e-06,
+      "loss": 1.2216,
+      "step": 1214
+    },
+    {
+      "epoch": 2.332213154104657,
+      "grad_norm": 0.6794226960110918,
+      "learning_rate": 6.446613060057551e-06,
+      "loss": 1.3418,
+      "step": 1215
+    },
+    {
+      "epoch": 2.3341334613538165,
+      "grad_norm": 0.6520127896780581,
+      "learning_rate": 6.4401970188499e-06,
+      "loss": 1.1196,
+      "step": 1216
+    },
+    {
+      "epoch": 2.3360537686029765,
+      "grad_norm": 0.7316497491112776,
+      "learning_rate": 6.43377839058301e-06,
+      "loss": 1.2544,
+      "step": 1217
+    },
+    {
+      "epoch": 2.337974075852136,
+      "grad_norm": 0.6291690821103928,
+      "learning_rate": 6.42735718678681e-06,
+      "loss": 1.2161,
+      "step": 1218
+    },
+    {
+      "epoch": 2.339894383101296,
+      "grad_norm": 0.7107157838496435,
+      "learning_rate": 6.420933418995861e-06,
+      "loss": 1.3376,
+      "step": 1219
+    },
+    {
+      "epoch": 2.341814690350456,
+      "grad_norm": 0.6661525380725084,
+      "learning_rate": 6.414507098749324e-06,
+      "loss": 1.3506,
+      "step": 1220
+    },
+    {
+      "epoch": 2.343734997599616,
+      "grad_norm": 0.6653555103275538,
+      "learning_rate": 6.4080782375909455e-06,
+      "loss": 1.1913,
+      "step": 1221
+    },
+    {
+      "epoch": 2.345655304848776,
+      "grad_norm": 0.6138861767200757,
+      "learning_rate": 6.401646847069038e-06,
+      "loss": 1.2745,
+      "step": 1222
+    },
+    {
+      "epoch": 2.347575612097936,
+      "grad_norm": 0.6543629399503493,
+      "learning_rate": 6.39521293873646e-06,
+      "loss": 1.2057,
+      "step": 1223
+    },
+    {
+      "epoch": 2.3494959193470955,
+      "grad_norm": 0.5774136334197446,
+      "learning_rate": 6.388776524150586e-06,
+      "loss": 1.174,
+      "step": 1224
+    },
+    {
+      "epoch": 2.3514162265962555,
+      "grad_norm": 0.5859347161113456,
+      "learning_rate": 6.3823376148733e-06,
+      "loss": 1.2425,
+      "step": 1225
+    },
+    {
+      "epoch": 2.353336533845415,
+      "grad_norm": 0.6408466768455149,
+      "learning_rate": 6.375896222470961e-06,
+      "loss": 1.1288,
+      "step": 1226
+    },
+    {
+      "epoch": 2.355256841094575,
+      "grad_norm": 0.6146169617139874,
+      "learning_rate": 6.369452358514392e-06,
+      "loss": 1.2214,
+      "step": 1227
+    },
+    {
+      "epoch": 2.357177148343735,
+      "grad_norm": 0.5819190647223798,
+      "learning_rate": 6.363006034578856e-06,
+      "loss": 1.2009,
+      "step": 1228
+    },
+    {
+      "epoch": 2.3590974555928947,
+      "grad_norm": 0.6030278958480552,
+      "learning_rate": 6.356557262244033e-06,
+      "loss": 1.165,
+      "step": 1229
+    },
+    {
+      "epoch": 2.361017762842055,
+      "grad_norm": 0.6491642246517341,
+      "learning_rate": 6.350106053094004e-06,
+      "loss": 1.1608,
+      "step": 1230
+    },
+    {
+      "epoch": 2.3629380700912144,
+      "grad_norm": 0.6463817243208629,
+      "learning_rate": 6.34365241871722e-06,
+      "loss": 1.3024,
+      "step": 1231
+    },
+    {
+      "epoch": 2.3648583773403744,
+      "grad_norm": 0.7103506380741759,
+      "learning_rate": 6.3371963707065e-06,
+      "loss": 1.382,
+      "step": 1232
+    },
+    {
+      "epoch": 2.3667786845895344,
+      "grad_norm": 0.6785469941350024,
+      "learning_rate": 6.330737920658989e-06,
+      "loss": 1.2195,
+      "step": 1233
+    },
+    {
+      "epoch": 2.368698991838694,
+      "grad_norm": 0.5570736910667738,
+      "learning_rate": 6.324277080176151e-06,
+      "loss": 1.2398,
+      "step": 1234
+    },
+    {
+      "epoch": 2.370619299087854,
+      "grad_norm": 0.6701770195276846,
+      "learning_rate": 6.317813860863743e-06,
+      "loss": 1.093,
+      "step": 1235
+    },
+    {
+      "epoch": 2.372539606337014,
+      "grad_norm": 0.6046075353253518,
+      "learning_rate": 6.311348274331797e-06,
+      "loss": 1.325,
+      "step": 1236
+    },
+    {
+      "epoch": 2.3744599135861737,
+      "grad_norm": 0.594000111250696,
+      "learning_rate": 6.304880332194593e-06,
+      "loss": 1.2857,
+      "step": 1237
+    },
+    {
+      "epoch": 2.3763802208353337,
+      "grad_norm": 0.6346049535343968,
+      "learning_rate": 6.2984100460706476e-06,
+      "loss": 1.2374,
+      "step": 1238
+    },
+    {
+      "epoch": 2.3783005280844933,
+      "grad_norm": 0.5748289888747085,
+      "learning_rate": 6.2919374275826835e-06,
+      "loss": 1.3284,
+      "step": 1239
+    },
+    {
+      "epoch": 2.3802208353336534,
+      "grad_norm": 0.6113265562594793,
+      "learning_rate": 6.285462488357618e-06,
+      "loss": 1.1277,
+      "step": 1240
+    },
+    {
+      "epoch": 2.3821411425828134,
+      "grad_norm": 0.6292319853626924,
+      "learning_rate": 6.278985240026532e-06,
+      "loss": 1.1987,
+      "step": 1241
+    },
+    {
+      "epoch": 2.384061449831973,
+      "grad_norm": 0.6263302899230722,
+      "learning_rate": 6.272505694224655e-06,
+      "loss": 1.3001,
+      "step": 1242
+    },
+    {
+      "epoch": 2.385981757081133,
+      "grad_norm": 0.6136706336821629,
+      "learning_rate": 6.26602386259135e-06,
+      "loss": 1.1442,
+      "step": 1243
+    },
+    {
+      "epoch": 2.3879020643302926,
+      "grad_norm": 0.5570855018978512,
+      "learning_rate": 6.259539756770078e-06,
+      "loss": 1.1984,
+      "step": 1244
+    },
+    {
+      "epoch": 2.3898223715794527,
+      "grad_norm": 0.5596547520906855,
+      "learning_rate": 6.253053388408389e-06,
+      "loss": 1.1809,
+      "step": 1245
+    },
+    {
+      "epoch": 2.3917426788286127,
+      "grad_norm": 0.693407976065708,
+      "learning_rate": 6.246564769157895e-06,
+      "loss": 1.3885,
+      "step": 1246
+    },
+    {
+      "epoch": 2.3936629860777723,
+      "grad_norm": 0.7502411766292096,
+      "learning_rate": 6.2400739106742545e-06,
+      "loss": 1.2608,
+      "step": 1247
+    },
+    {
+      "epoch": 2.3955832933269323,
+      "grad_norm": 0.659943311180132,
+      "learning_rate": 6.233580824617147e-06,
+      "loss": 1.1978,
+      "step": 1248
+    },
+    {
+      "epoch": 2.3975036005760924,
+      "grad_norm": 0.5468106201549785,
+      "learning_rate": 6.227085522650253e-06,
+      "loss": 1.2264,
+      "step": 1249
+    },
+    {
+      "epoch": 2.399423907825252,
+      "grad_norm": 0.5859970642991462,
+      "learning_rate": 6.220588016441234e-06,
+      "loss": 1.3923,
+      "step": 1250
+    },
+    {
+      "epoch": 2.401344215074412,
+      "grad_norm": 0.6745672723394359,
+      "learning_rate": 6.214088317661709e-06,
+      "loss": 1.1152,
+      "step": 1251
+    },
+    {
+      "epoch": 2.403264522323572,
+      "grad_norm": 0.6158852563754607,
+      "learning_rate": 6.207586437987241e-06,
+      "loss": 1.3166,
+      "step": 1252
+    },
+    {
+      "epoch": 2.4051848295727316,
+      "grad_norm": 0.5730095887391516,
+      "learning_rate": 6.201082389097302e-06,
+      "loss": 1.1936,
+      "step": 1253
+    },
+    {
+      "epoch": 2.4071051368218916,
+      "grad_norm": 1.4640138016466724,
+      "learning_rate": 6.19457618267527e-06,
+      "loss": 1.2584,
+      "step": 1254
+    },
+    {
+      "epoch": 2.4090254440710512,
+      "grad_norm": 0.6349000539598749,
+      "learning_rate": 6.188067830408393e-06,
+      "loss": 1.224,
+      "step": 1255
+    },
+    {
+      "epoch": 2.4109457513202113,
+      "grad_norm": 0.6311572823390743,
+      "learning_rate": 6.181557343987775e-06,
+      "loss": 1.2915,
+      "step": 1256
+    },
+    {
+      "epoch": 2.412866058569371,
+      "grad_norm": 0.6695682124451278,
+      "learning_rate": 6.175044735108349e-06,
+      "loss": 1.2961,
+      "step": 1257
+    },
+    {
+      "epoch": 2.414786365818531,
+      "grad_norm": 0.6661777691771004,
+      "learning_rate": 6.168530015468872e-06,
+      "loss": 1.1564,
+      "step": 1258
+    },
+    {
+      "epoch": 2.416706673067691,
+      "grad_norm": 0.5769627139546278,
+      "learning_rate": 6.162013196771882e-06,
+      "loss": 1.1191,
+      "step": 1259
+    },
+    {
+      "epoch": 2.4186269803168505,
+      "grad_norm": 0.6696625061542401,
+      "learning_rate": 6.155494290723691e-06,
+      "loss": 1.2594,
+      "step": 1260
+    },
+    {
+      "epoch": 2.4205472875660106,
+      "grad_norm": 0.5824696247611385,
+      "learning_rate": 6.148973309034363e-06,
+      "loss": 1.1747,
+      "step": 1261
+    },
+    {
+      "epoch": 2.4224675948151706,
+      "grad_norm": 0.6813330398855266,
+      "learning_rate": 6.142450263417685e-06,
+      "loss": 1.2173,
+      "step": 1262
+    },
+    {
+      "epoch": 2.42438790206433,
+      "grad_norm": 0.5692154085477008,
+      "learning_rate": 6.135925165591159e-06,
+      "loss": 1.3443,
+      "step": 1263
+    },
+    {
+      "epoch": 2.4263082093134902,
+      "grad_norm": 0.5197059030883115,
+      "learning_rate": 6.129398027275966e-06,
+      "loss": 1.0257,
+      "step": 1264
+    },
+    {
+      "epoch": 2.4282285165626503,
+      "grad_norm": 0.7088728686250311,
+      "learning_rate": 6.122868860196956e-06,
+      "loss": 1.1714,
+      "step": 1265
+    },
+    {
+      "epoch": 2.43014882381181,
+      "grad_norm": 0.5794244685269138,
+      "learning_rate": 6.116337676082623e-06,
+      "loss": 1.2578,
+      "step": 1266
+    },
+    {
+      "epoch": 2.43206913106097,
+      "grad_norm": 0.5659150508785673,
+      "learning_rate": 6.109804486665085e-06,
+      "loss": 1.2957,
+      "step": 1267
+    },
+    {
+      "epoch": 2.4339894383101295,
+      "grad_norm": 0.6333267338148806,
+      "learning_rate": 6.103269303680063e-06,
+      "loss": 1.2752,
+      "step": 1268
+    },
+    {
+      "epoch": 2.4359097455592895,
+      "grad_norm": 0.5950812272774857,
+      "learning_rate": 6.0967321388668534e-06,
+      "loss": 1.2894,
+      "step": 1269
+    },
+    {
+      "epoch": 2.437830052808449,
+      "grad_norm": 0.6350175691173675,
+      "learning_rate": 6.090193003968319e-06,
+      "loss": 1.3265,
+      "step": 1270
+    },
+    {
+      "epoch": 2.439750360057609,
+      "grad_norm": 0.6837493426325877,
+      "learning_rate": 6.08365191073086e-06,
+      "loss": 1.1427,
+      "step": 1271
+    },
+    {
+      "epoch": 2.441670667306769,
+      "grad_norm": 0.5958537398944314,
+      "learning_rate": 6.0771088709043915e-06,
+      "loss": 1.1849,
+      "step": 1272
+    },
+    {
+      "epoch": 2.4435909745559288,
+      "grad_norm": 0.6021156128212082,
+      "learning_rate": 6.070563896242329e-06,
+      "loss": 1.2944,
+      "step": 1273
+    },
+    {
+      "epoch": 2.445511281805089,
+      "grad_norm": 0.6631372287185189,
+      "learning_rate": 6.064016998501563e-06,
+      "loss": 1.1679,
+      "step": 1274
+    },
+    {
+      "epoch": 2.447431589054249,
+      "grad_norm": 0.5836864312172045,
+      "learning_rate": 6.057468189442432e-06,
+      "loss": 1.1259,
+      "step": 1275
+    },
+    {
+      "epoch": 2.4493518963034084,
+      "grad_norm": 0.7125336969548751,
+      "learning_rate": 6.050917480828721e-06,
+      "loss": 1.4025,
+      "step": 1276
+    },
+    {
+      "epoch": 2.4512722035525685,
+      "grad_norm": 0.5999779991511556,
+      "learning_rate": 6.044364884427614e-06,
+      "loss": 1.2283,
+      "step": 1277
+    },
+    {
+      "epoch": 2.4531925108017285,
+      "grad_norm": 0.6228786103662747,
+      "learning_rate": 6.037810412009693e-06,
+      "loss": 1.1781,
+      "step": 1278
+    },
+    {
+      "epoch": 2.455112818050888,
+      "grad_norm": 0.5995133183671691,
+      "learning_rate": 6.031254075348908e-06,
+      "loss": 1.1651,
+      "step": 1279
+    },
+    {
+      "epoch": 2.457033125300048,
+      "grad_norm": 0.5726228213086509,
+      "learning_rate": 6.02469588622256e-06,
+      "loss": 1.2179,
+      "step": 1280
+    },
+    {
+      "epoch": 2.4589534325492077,
+      "grad_norm": 0.6319072116257335,
+      "learning_rate": 6.018135856411275e-06,
+      "loss": 1.0789,
+      "step": 1281
+    },
+    {
+      "epoch": 2.4608737397983678,
+      "grad_norm": 0.5791309734437495,
+      "learning_rate": 6.011573997698985e-06,
+      "loss": 1.233,
+      "step": 1282
+    },
+    {
+      "epoch": 2.4627940470475274,
+      "grad_norm": 0.6145255184440226,
+      "learning_rate": 6.00501032187291e-06,
+      "loss": 1.1686,
+      "step": 1283
+    },
+    {
+      "epoch": 2.4647143542966874,
+      "grad_norm": 0.5379184144373157,
+      "learning_rate": 5.998444840723534e-06,
+      "loss": 1.2602,
+      "step": 1284
+    },
+    {
+      "epoch": 2.4666346615458474,
+      "grad_norm": 0.6167816126301073,
+      "learning_rate": 5.991877566044581e-06,
+      "loss": 1.2921,
+      "step": 1285
+    },
+    {
+      "epoch": 2.468554968795007,
+      "grad_norm": 0.6535144856705375,
+      "learning_rate": 5.985308509633e-06,
+      "loss": 1.2201,
+      "step": 1286
+    },
+    {
+      "epoch": 2.470475276044167,
+      "grad_norm": 0.6860549323986097,
+      "learning_rate": 5.978737683288938e-06,
+      "loss": 1.2078,
+      "step": 1287
+    },
+    {
+      "epoch": 2.472395583293327,
+      "grad_norm": 0.6951627540400147,
+      "learning_rate": 5.972165098815721e-06,
+      "loss": 1.2361,
+      "step": 1288
+    },
+    {
+      "epoch": 2.4743158905424867,
+      "grad_norm": 0.6546505487989865,
+      "learning_rate": 5.965590768019838e-06,
+      "loss": 1.2895,
+      "step": 1289
+    },
+    {
+      "epoch": 2.4762361977916467,
+      "grad_norm": 0.6120831995174033,
+      "learning_rate": 5.959014702710908e-06,
+      "loss": 1.4001,
+      "step": 1290
+    },
+    {
+      "epoch": 2.4781565050408068,
+      "grad_norm": 0.6617422728749986,
+      "learning_rate": 5.952436914701673e-06,
+      "loss": 1.2279,
+      "step": 1291
+    },
+    {
+      "epoch": 2.4800768122899663,
+      "grad_norm": 0.6466808021093433,
+      "learning_rate": 5.945857415807962e-06,
+      "loss": 1.3116,
+      "step": 1292
+    },
+    {
+      "epoch": 2.4819971195391264,
+      "grad_norm": 0.5593849707459436,
+      "learning_rate": 5.939276217848684e-06,
+      "loss": 1.3044,
+      "step": 1293
+    },
+    {
+      "epoch": 2.483917426788286,
+      "grad_norm": 0.694337911560069,
+      "learning_rate": 5.932693332645796e-06,
+      "loss": 1.4405,
+      "step": 1294
+    },
+    {
+      "epoch": 2.485837734037446,
+      "grad_norm": 0.5796450029223225,
+      "learning_rate": 5.926108772024286e-06,
+      "loss": 1.3507,
+      "step": 1295
+    },
+    {
+      "epoch": 2.487758041286606,
+      "grad_norm": 0.6046024795406408,
+      "learning_rate": 5.919522547812155e-06,
+      "loss": 1.286,
+      "step": 1296
+    },
+    {
+      "epoch": 2.4896783485357656,
+      "grad_norm": 0.6493749876717334,
+      "learning_rate": 5.912934671840389e-06,
+      "loss": 1.2146,
+      "step": 1297
+    },
+    {
+      "epoch": 2.4915986557849257,
+      "grad_norm": 0.6796731338841863,
+      "learning_rate": 5.906345155942943e-06,
+      "loss": 1.4317,
+      "step": 1298
+    },
+    {
+      "epoch": 2.4935189630340853,
+      "grad_norm": 0.6860980566749547,
+      "learning_rate": 5.899754011956715e-06,
+      "loss": 1.2145,
+      "step": 1299
+    },
+    {
+      "epoch": 2.4954392702832453,
+      "grad_norm": 0.5565130162971886,
+      "learning_rate": 5.8931612517215305e-06,
+      "loss": 1.2637,
+      "step": 1300
+    },
+    {
+      "epoch": 2.4973595775324053,
+      "grad_norm": 0.6378841349295684,
+      "learning_rate": 5.886566887080117e-06,
+      "loss": 1.0994,
+      "step": 1301
+    },
+    {
+      "epoch": 2.499279884781565,
+      "grad_norm": 0.7116787998039626,
+      "learning_rate": 5.879970929878086e-06,
+      "loss": 1.3812,
+      "step": 1302
+    },
+    {
+      "epoch": 2.501200192030725,
+      "grad_norm": 0.6423843577571763,
+      "learning_rate": 5.873373391963906e-06,
+      "loss": 1.1543,
+      "step": 1303
+    },
+    {
+      "epoch": 2.503120499279885,
+      "grad_norm": 0.5726699800446029,
+      "learning_rate": 5.866774285188887e-06,
+      "loss": 1.3002,
+      "step": 1304
+    },
+    {
+      "epoch": 2.5050408065290446,
+      "grad_norm": 0.6330116873694995,
+      "learning_rate": 5.860173621407157e-06,
+      "loss": 1.239,
+      "step": 1305
+    },
+    {
+      "epoch": 2.5069611137782046,
+      "grad_norm": 0.6372043038208098,
+      "learning_rate": 5.853571412475644e-06,
+      "loss": 1.2213,
+      "step": 1306
+    },
+    {
+      "epoch": 2.5088814210273642,
+      "grad_norm": 0.6247955734407857,
+      "learning_rate": 5.8469676702540454e-06,
+      "loss": 1.2773,
+      "step": 1307
+    },
+    {
+      "epoch": 2.5108017282765243,
+      "grad_norm": 0.5817678497097796,
+      "learning_rate": 5.840362406604818e-06,
+      "loss": 1.2735,
+      "step": 1308
+    },
+    {
+      "epoch": 2.512722035525684,
+      "grad_norm": 0.5520561018680428,
+      "learning_rate": 5.83375563339315e-06,
+      "loss": 1.2459,
+      "step": 1309
+    },
+    {
+      "epoch": 2.514642342774844,
+      "grad_norm": 0.7410450237967919,
+      "learning_rate": 5.82714736248694e-06,
+      "loss": 1.1031,
+      "step": 1310
+    },
+    {
+      "epoch": 2.516562650024004,
+      "grad_norm": 0.5738368012193878,
+      "learning_rate": 5.820537605756778e-06,
+      "loss": 1.3073,
+      "step": 1311
+    },
+    {
+      "epoch": 2.5184829572731635,
+      "grad_norm": 0.5573105806604853,
+      "learning_rate": 5.813926375075924e-06,
+      "loss": 1.2266,
+      "step": 1312
+    },
+    {
+      "epoch": 2.5204032645223235,
+      "grad_norm": 0.6181886080856598,
+      "learning_rate": 5.807313682320285e-06,
+      "loss": 1.1996,
+      "step": 1313
+    },
+    {
+      "epoch": 2.5223235717714836,
+      "grad_norm": 0.6730335062822276,
+      "learning_rate": 5.800699539368391e-06,
+      "loss": 1.2198,
+      "step": 1314
+    },
+    {
+      "epoch": 2.524243879020643,
+      "grad_norm": 0.5605883111414951,
+      "learning_rate": 5.794083958101383e-06,
+      "loss": 1.2109,
+      "step": 1315
+    },
+    {
+      "epoch": 2.526164186269803,
+      "grad_norm": 0.6421642462562456,
+      "learning_rate": 5.7874669504029825e-06,
+      "loss": 1.0927,
+      "step": 1316
+    },
+    {
+      "epoch": 2.5280844935189632,
+      "grad_norm": 0.6376536627209224,
+      "learning_rate": 5.780848528159471e-06,
+      "loss": 1.1857,
+      "step": 1317
+    },
+    {
+      "epoch": 2.530004800768123,
+      "grad_norm": 0.6341601739426433,
+      "learning_rate": 5.774228703259678e-06,
+      "loss": 1.0717,
+      "step": 1318
+    },
+    {
+      "epoch": 2.531925108017283,
+      "grad_norm": 0.5804812110235914,
+      "learning_rate": 5.767607487594944e-06,
+      "loss": 1.2235,
+      "step": 1319
+    },
+    {
+      "epoch": 2.5338454152664425,
+      "grad_norm": 0.6154932839231937,
+      "learning_rate": 5.760984893059115e-06,
+      "loss": 1.3717,
+      "step": 1320
+    },
+    {
+      "epoch": 2.5357657225156025,
+      "grad_norm": 0.7574489368316524,
+      "learning_rate": 5.754360931548509e-06,
+      "loss": 1.2414,
+      "step": 1321
+    },
+    {
+      "epoch": 2.537686029764762,
+      "grad_norm": 0.5677124356572427,
+      "learning_rate": 5.747735614961902e-06,
+      "loss": 1.2608,
+      "step": 1322
+    },
+    {
+      "epoch": 2.539606337013922,
+      "grad_norm": 0.5704134007435095,
+      "learning_rate": 5.741108955200503e-06,
+      "loss": 1.3432,
+      "step": 1323
+    },
+    {
+      "epoch": 2.541526644263082,
+      "grad_norm": 0.5951256027367144,
+      "learning_rate": 5.734480964167935e-06,
+      "loss": 0.9872,
+      "step": 1324
+    },
+    {
+      "epoch": 2.5434469515122418,
+      "grad_norm": 0.6719946916626364,
+      "learning_rate": 5.727851653770211e-06,
+      "loss": 1.25,
+      "step": 1325
+    },
+    {
+      "epoch": 2.545367258761402,
+      "grad_norm": 0.6286872277200853,
+      "learning_rate": 5.721221035915717e-06,
+      "loss": 1.1818,
+      "step": 1326
+    },
+    {
+      "epoch": 2.547287566010562,
+      "grad_norm": 0.5265728959047694,
+      "learning_rate": 5.714589122515182e-06,
+      "loss": 1.3031,
+      "step": 1327
+    },
+    {
+      "epoch": 2.5492078732597214,
+      "grad_norm": 0.6295646479508796,
+      "learning_rate": 5.7079559254816665e-06,
+      "loss": 1.2045,
+      "step": 1328
+    },
+    {
+      "epoch": 2.5511281805088815,
+      "grad_norm": 0.6270913464833281,
+      "learning_rate": 5.701321456730536e-06,
+      "loss": 1.2163,
+      "step": 1329
+    },
+    {
+      "epoch": 2.5530484877580415,
+      "grad_norm": 0.6115400056563143,
+      "learning_rate": 5.694685728179442e-06,
+      "loss": 1.1537,
+      "step": 1330
+    },
+    {
+      "epoch": 2.554968795007201,
+      "grad_norm": 0.6602324772946387,
+      "learning_rate": 5.688048751748296e-06,
+      "loss": 1.1886,
+      "step": 1331
+    },
+    {
+      "epoch": 2.556889102256361,
+      "grad_norm": 0.6285231165205739,
+      "learning_rate": 5.681410539359251e-06,
+      "loss": 1.1562,
+      "step": 1332
+    },
+    {
+      "epoch": 2.558809409505521,
+      "grad_norm": 0.5580091730080278,
+      "learning_rate": 5.6747711029366845e-06,
+      "loss": 1.1137,
+      "step": 1333
+    },
+    {
+      "epoch": 2.5607297167546808,
+      "grad_norm": 0.6155227527633782,
+      "learning_rate": 5.668130454407168e-06,
+      "loss": 1.2266,
+      "step": 1334
+    },
+    {
+      "epoch": 2.5626500240038403,
+      "grad_norm": 0.7208056151770321,
+      "learning_rate": 5.661488605699451e-06,
+      "loss": 1.2748,
+      "step": 1335
+    },
+    {
+      "epoch": 2.5645703312530004,
+      "grad_norm": 0.6358039541543873,
+      "learning_rate": 5.654845568744443e-06,
+      "loss": 1.2893,
+      "step": 1336
+    },
+    {
+      "epoch": 2.5664906385021604,
+      "grad_norm": 0.6022914407629255,
+      "learning_rate": 5.648201355475182e-06,
+      "loss": 1.1789,
+      "step": 1337
+    },
+    {
+      "epoch": 2.56841094575132,
+      "grad_norm": 0.6261507684350669,
+      "learning_rate": 5.641555977826824e-06,
+      "loss": 1.2932,
+      "step": 1338
+    },
+    {
+      "epoch": 2.57033125300048,
+      "grad_norm": 0.6556797218150923,
+      "learning_rate": 5.634909447736614e-06,
+      "loss": 1.4065,
+      "step": 1339
+    },
+    {
+      "epoch": 2.57225156024964,
+      "grad_norm": 0.8088493377456971,
+      "learning_rate": 5.628261777143867e-06,
+      "loss": 1.1762,
+      "step": 1340
+    },
+    {
+      "epoch": 2.5741718674987997,
+      "grad_norm": 0.6106237399013169,
+      "learning_rate": 5.62161297798995e-06,
+      "loss": 1.2155,
+      "step": 1341
+    },
+    {
+      "epoch": 2.5760921747479597,
+      "grad_norm": 0.5620193635243992,
+      "learning_rate": 5.614963062218253e-06,
+      "loss": 1.1267,
+      "step": 1342
+    },
+    {
+      "epoch": 2.5780124819971197,
+      "grad_norm": 0.5916968354702812,
+      "learning_rate": 5.608312041774175e-06,
+      "loss": 1.3556,
+      "step": 1343
+    },
+    {
+      "epoch": 2.5799327892462793,
+      "grad_norm": 0.6222711849579262,
+      "learning_rate": 5.601659928605095e-06,
+      "loss": 1.2217,
+      "step": 1344
+    },
+    {
+      "epoch": 2.5818530964954394,
+      "grad_norm": 0.7012656976911378,
+      "learning_rate": 5.5950067346603595e-06,
+      "loss": 1.2048,
+      "step": 1345
+    },
+    {
+      "epoch": 2.5837734037445994,
+      "grad_norm": 0.5879328784503015,
+      "learning_rate": 5.588352471891259e-06,
+      "loss": 1.3748,
+      "step": 1346
+    },
+    {
+      "epoch": 2.585693710993759,
+      "grad_norm": 0.6321780544880533,
+      "learning_rate": 5.581697152250992e-06,
+      "loss": 1.3289,
+      "step": 1347
+    },
+    {
+      "epoch": 2.5876140182429186,
+      "grad_norm": 0.6604594554850995,
+      "learning_rate": 5.575040787694668e-06,
+      "loss": 1.2861,
+      "step": 1348
+    },
+    {
+      "epoch": 2.5895343254920786,
+      "grad_norm": 0.7738496596525316,
+      "learning_rate": 5.568383390179267e-06,
+      "loss": 1.371,
+      "step": 1349
+    },
+    {
+      "epoch": 2.5914546327412387,
+      "grad_norm": 0.561705763795756,
+      "learning_rate": 5.561724971663628e-06,
+      "loss": 1.4532,
+      "step": 1350
+    },
+    {
+      "epoch": 2.5933749399903983,
+      "grad_norm": 0.6718027044013396,
+      "learning_rate": 5.55506554410842e-06,
+      "loss": 1.276,
+      "step": 1351
+    },
+    {
+      "epoch": 2.5952952472395583,
+      "grad_norm": 0.6972126937496573,
+      "learning_rate": 5.548405119476129e-06,
+      "loss": 1.2157,
+      "step": 1352
+    },
+    {
+      "epoch": 2.5972155544887183,
+      "grad_norm": 0.5943209839227868,
+      "learning_rate": 5.541743709731029e-06,
+      "loss": 1.2271,
+      "step": 1353
+    },
+    {
+      "epoch": 2.599135861737878,
+      "grad_norm": 0.6646420886497804,
+      "learning_rate": 5.535081326839165e-06,
+      "loss": 1.3276,
+      "step": 1354
+    },
+    {
+      "epoch": 2.601056168987038,
+      "grad_norm": 0.6178199961209532,
+      "learning_rate": 5.528417982768328e-06,
+      "loss": 1.2772,
+      "step": 1355
+    },
+    {
+      "epoch": 2.602976476236198,
+      "grad_norm": 0.6774402354328428,
+      "learning_rate": 5.521753689488039e-06,
+      "loss": 1.2322,
+      "step": 1356
+    },
+    {
+      "epoch": 2.6048967834853576,
+      "grad_norm": 0.5815814662153418,
+      "learning_rate": 5.515088458969522e-06,
+      "loss": 1.1769,
+      "step": 1357
+    },
+    {
+      "epoch": 2.6068170907345176,
+      "grad_norm": 0.6922973775030319,
+      "learning_rate": 5.508422303185682e-06,
+      "loss": 1.2962,
+      "step": 1358
+    },
+    {
+      "epoch": 2.6087373979836777,
+      "grad_norm": 0.6665626499511711,
+      "learning_rate": 5.501755234111095e-06,
+      "loss": 1.3545,
+      "step": 1359
+    },
+    {
+      "epoch": 2.6106577052328372,
+      "grad_norm": 0.735542600803809,
+      "learning_rate": 5.495087263721965e-06,
+      "loss": 1.1893,
+      "step": 1360
+    },
+    {
+      "epoch": 2.6125780124819973,
+      "grad_norm": 0.5238670939024578,
+      "learning_rate": 5.488418403996125e-06,
+      "loss": 1.1627,
+      "step": 1361
+    },
+    {
+      "epoch": 2.614498319731157,
+      "grad_norm": 0.6210197741128046,
+      "learning_rate": 5.481748666913001e-06,
+      "loss": 1.1276,
+      "step": 1362
+    },
+    {
+      "epoch": 2.616418626980317,
+      "grad_norm": 0.5963335444059878,
+      "learning_rate": 5.4750780644535975e-06,
+      "loss": 1.0468,
+      "step": 1363
+    },
+    {
+      "epoch": 2.6183389342294765,
+      "grad_norm": 0.628355725486478,
+      "learning_rate": 5.46840660860047e-06,
+      "loss": 1.309,
+      "step": 1364
+    },
+    {
+      "epoch": 2.6202592414786365,
+      "grad_norm": 0.5712468661870007,
+      "learning_rate": 5.461734311337712e-06,
+      "loss": 1.2055,
+      "step": 1365
+    },
+    {
+      "epoch": 2.6221795487277966,
+      "grad_norm": 0.6550614152257059,
+      "learning_rate": 5.455061184650921e-06,
+      "loss": 1.1254,
+      "step": 1366
+    },
+    {
+      "epoch": 2.624099855976956,
+      "grad_norm": 0.5857760293047071,
+      "learning_rate": 5.448387240527195e-06,
+      "loss": 1.2524,
+      "step": 1367
+    },
+    {
+      "epoch": 2.626020163226116,
+      "grad_norm": 0.5996253176790339,
+      "learning_rate": 5.441712490955088e-06,
+      "loss": 1.1753,
+      "step": 1368
+    },
+    {
+      "epoch": 2.6279404704752762,
+      "grad_norm": 0.5192831528485529,
+      "learning_rate": 5.435036947924611e-06,
+      "loss": 1.2642,
+      "step": 1369
+    },
+    {
+      "epoch": 2.629860777724436,
+      "grad_norm": 0.6322581887680471,
+      "learning_rate": 5.4283606234271955e-06,
+      "loss": 1.1797,
+      "step": 1370
+    },
+    {
+      "epoch": 2.631781084973596,
+      "grad_norm": 0.6208687760914947,
+      "learning_rate": 5.421683529455677e-06,
+      "loss": 1.2889,
+      "step": 1371
+    },
+    {
+      "epoch": 2.633701392222756,
+      "grad_norm": 0.6024350882429259,
+      "learning_rate": 5.415005678004277e-06,
+      "loss": 1.2419,
+      "step": 1372
+    },
+    {
+      "epoch": 2.6356216994719155,
+      "grad_norm": 0.5985715149546723,
+      "learning_rate": 5.40832708106857e-06,
+      "loss": 1.3948,
+      "step": 1373
+    },
+    {
+      "epoch": 2.6375420067210755,
+      "grad_norm": 0.6196127560581245,
+      "learning_rate": 5.401647750645477e-06,
+      "loss": 1.4807,
+      "step": 1374
+    },
+    {
+      "epoch": 2.639462313970235,
+      "grad_norm": 0.64681914597301,
+      "learning_rate": 5.394967698733234e-06,
+      "loss": 1.1374,
+      "step": 1375
+    },
+    {
+      "epoch": 2.641382621219395,
+      "grad_norm": 0.4890585689575943,
+      "learning_rate": 5.38828693733137e-06,
+      "loss": 1.1564,
+      "step": 1376
+    },
+    {
+      "epoch": 2.6433029284685547,
+      "grad_norm": 0.5630617563950847,
+      "learning_rate": 5.381605478440696e-06,
+      "loss": 1.2389,
+      "step": 1377
+    },
+    {
+      "epoch": 2.645223235717715,
+      "grad_norm": 0.5788570414678232,
+      "learning_rate": 5.3749233340632676e-06,
+      "loss": 1.2963,
+      "step": 1378
+    },
+    {
+      "epoch": 2.647143542966875,
+      "grad_norm": 0.64567790799067,
+      "learning_rate": 5.368240516202376e-06,
+      "loss": 1.1306,
+      "step": 1379
+    },
+    {
+      "epoch": 2.6490638502160344,
+      "grad_norm": 0.6394046194644902,
+      "learning_rate": 5.3615570368625235e-06,
+      "loss": 1.3908,
+      "step": 1380
+    },
+    {
+      "epoch": 2.6509841574651944,
+      "grad_norm": 0.6542343993204641,
+      "learning_rate": 5.354872908049398e-06,
+      "loss": 1.3043,
+      "step": 1381
+    },
+    {
+      "epoch": 2.6529044647143545,
+      "grad_norm": 0.5928443232243815,
+      "learning_rate": 5.348188141769852e-06,
+      "loss": 1.1948,
+      "step": 1382
+    },
+    {
+      "epoch": 2.654824771963514,
+      "grad_norm": 0.6634119943181676,
+      "learning_rate": 5.34150275003189e-06,
+      "loss": 1.2738,
+      "step": 1383
+    },
+    {
+      "epoch": 2.656745079212674,
+      "grad_norm": 0.6175973847987798,
+      "learning_rate": 5.334816744844633e-06,
+      "loss": 1.2039,
+      "step": 1384
+    },
+    {
+      "epoch": 2.658665386461834,
+      "grad_norm": 0.5328533760172757,
+      "learning_rate": 5.328130138218309e-06,
+      "loss": 1.168,
+      "step": 1385
+    },
+    {
+      "epoch": 2.6605856937109937,
+      "grad_norm": 0.6333132588626341,
+      "learning_rate": 5.3214429421642224e-06,
+      "loss": 1.2574,
+      "step": 1386
+    },
+    {
+      "epoch": 2.6625060009601538,
+      "grad_norm": 0.6141814037172568,
+      "learning_rate": 5.3147551686947385e-06,
+      "loss": 1.3352,
+      "step": 1387
+    },
+    {
+      "epoch": 2.6644263082093134,
+      "grad_norm": 0.6628973771397662,
+      "learning_rate": 5.308066829823261e-06,
+      "loss": 1.164,
+      "step": 1388
+    },
+    {
+      "epoch": 2.6663466154584734,
+      "grad_norm": 0.6856680778100335,
+      "learning_rate": 5.301377937564205e-06,
+      "loss": 1.1983,
+      "step": 1389
+    },
+    {
+      "epoch": 2.668266922707633,
+      "grad_norm": 0.5261090189495043,
+      "learning_rate": 5.294688503932986e-06,
+      "loss": 1.4359,
+      "step": 1390
+    },
+    {
+      "epoch": 2.670187229956793,
+      "grad_norm": 0.6886373426064217,
+      "learning_rate": 5.287998540945987e-06,
+      "loss": 1.271,
+      "step": 1391
+    },
+    {
+      "epoch": 2.672107537205953,
+      "grad_norm": 0.7012339259498463,
+      "learning_rate": 5.281308060620543e-06,
+      "loss": 1.2969,
+      "step": 1392
+    },
+    {
+      "epoch": 2.6740278444551127,
+      "grad_norm": 0.7355650125399332,
+      "learning_rate": 5.274617074974918e-06,
+      "loss": 1.1504,
+      "step": 1393
+    },
+    {
+      "epoch": 2.6759481517042727,
+      "grad_norm": 0.5562169304602139,
+      "learning_rate": 5.267925596028285e-06,
+      "loss": 1.0409,
+      "step": 1394
+    },
+    {
+      "epoch": 2.6778684589534327,
+      "grad_norm": 0.5919913846193492,
+      "learning_rate": 5.2612336358007035e-06,
+      "loss": 1.2113,
+      "step": 1395
+    },
+    {
+      "epoch": 2.6797887662025923,
+      "grad_norm": 0.6692107545177737,
+      "learning_rate": 5.2545412063130964e-06,
+      "loss": 1.1718,
+      "step": 1396
+    },
+    {
+      "epoch": 2.6817090734517524,
+      "grad_norm": 0.6418333213055866,
+      "learning_rate": 5.247848319587226e-06,
+      "loss": 1.2882,
+      "step": 1397
+    },
+    {
+      "epoch": 2.6836293807009124,
+      "grad_norm": 0.6934356927705114,
+      "learning_rate": 5.241154987645687e-06,
+      "loss": 1.2716,
+      "step": 1398
+    },
+    {
+      "epoch": 2.685549687950072,
+      "grad_norm": 0.5551620206008213,
+      "learning_rate": 5.234461222511858e-06,
+      "loss": 1.3031,
+      "step": 1399
+    },
+    {
+      "epoch": 2.687469995199232,
+      "grad_norm": 0.6360668296023374,
+      "learning_rate": 5.227767036209911e-06,
+      "loss": 1.1746,
+      "step": 1400
+    },
+    {
+      "epoch": 2.6893903024483916,
+      "grad_norm": 0.6995023935942758,
+      "learning_rate": 5.221072440764765e-06,
+      "loss": 1.1183,
+      "step": 1401
+    },
+    {
+      "epoch": 2.6913106096975516,
+      "grad_norm": 0.5654432672068804,
+      "learning_rate": 5.214377448202075e-06,
+      "loss": 1.2492,
+      "step": 1402
+    },
+    {
+      "epoch": 2.6932309169467112,
+      "grad_norm": 0.6937559082305748,
+      "learning_rate": 5.2076820705482155e-06,
+      "loss": 1.2054,
+      "step": 1403
+    },
+    {
+      "epoch": 2.6951512241958713,
+      "grad_norm": 0.6365688104050511,
+      "learning_rate": 5.200986319830245e-06,
+      "loss": 1.3126,
+      "step": 1404
+    },
+    {
+      "epoch": 2.6970715314450313,
+      "grad_norm": 0.6076408603108523,
+      "learning_rate": 5.194290208075896e-06,
+      "loss": 1.4652,
+      "step": 1405
+    },
+    {
+      "epoch": 2.698991838694191,
+      "grad_norm": 0.7778885537673794,
+      "learning_rate": 5.18759374731355e-06,
+      "loss": 1.346,
+      "step": 1406
+    },
+    {
+      "epoch": 2.700912145943351,
+      "grad_norm": 0.5841648932428964,
+      "learning_rate": 5.180896949572213e-06,
+      "loss": 1.4298,
+      "step": 1407
+    },
+    {
+      "epoch": 2.702832453192511,
+      "grad_norm": 0.6337842674058476,
+      "learning_rate": 5.174199826881498e-06,
+      "loss": 1.2634,
+      "step": 1408
+    },
+    {
+      "epoch": 2.7047527604416706,
+      "grad_norm": 0.5528118794971537,
+      "learning_rate": 5.167502391271603e-06,
+      "loss": 1.1476,
+      "step": 1409
+    },
+    {
+      "epoch": 2.7066730676908306,
+      "grad_norm": 0.6428624442509177,
+      "learning_rate": 5.160804654773286e-06,
+      "loss": 1.1906,
+      "step": 1410
+    },
+    {
+      "epoch": 2.7085933749399906,
+      "grad_norm": 0.5508196505900297,
+      "learning_rate": 5.154106629417845e-06,
+      "loss": 1.1828,
+      "step": 1411
+    },
+    {
+      "epoch": 2.7105136821891502,
+      "grad_norm": 0.5399557474898792,
+      "learning_rate": 5.147408327237099e-06,
+      "loss": 1.1979,
+      "step": 1412
+    },
+    {
+      "epoch": 2.7124339894383103,
+      "grad_norm": 0.5785664255689172,
+      "learning_rate": 5.140709760263364e-06,
+      "loss": 1.2234,
+      "step": 1413
+    },
+    {
+      "epoch": 2.71435429668747,
+      "grad_norm": 0.5862267357113349,
+      "learning_rate": 5.134010940529429e-06,
+      "loss": 1.3416,
+      "step": 1414
+    },
+    {
+      "epoch": 2.71627460393663,
+      "grad_norm": 0.5962477885128445,
+      "learning_rate": 5.127311880068539e-06,
+      "loss": 1.1983,
+      "step": 1415
+    },
+    {
+      "epoch": 2.7181949111857895,
+      "grad_norm": 0.6863189095052216,
+      "learning_rate": 5.1206125909143745e-06,
+      "loss": 1.3491,
+      "step": 1416
+    },
+    {
+      "epoch": 2.7201152184349495,
+      "grad_norm": 0.5885538990134215,
+      "learning_rate": 5.11391308510102e-06,
+      "loss": 1.1574,
+      "step": 1417
+    },
+    {
+      "epoch": 2.7220355256841096,
+      "grad_norm": 0.6104132348781779,
+      "learning_rate": 5.107213374662954e-06,
+      "loss": 1.2853,
+      "step": 1418
+    },
+    {
+      "epoch": 2.723955832933269,
+      "grad_norm": 0.6153215004897434,
+      "learning_rate": 5.100513471635022e-06,
+      "loss": 1.0575,
+      "step": 1419
+    },
+    {
+      "epoch": 2.725876140182429,
+      "grad_norm": 0.6305261136020173,
+      "learning_rate": 5.0938133880524145e-06,
+      "loss": 1.1139,
+      "step": 1420
+    },
+    {
+      "epoch": 2.727796447431589,
+      "grad_norm": 0.6887979401369777,
+      "learning_rate": 5.087113135950646e-06,
+      "loss": 1.2312,
+      "step": 1421
+    },
+    {
+      "epoch": 2.729716754680749,
+      "grad_norm": 0.54068610955665,
+      "learning_rate": 5.080412727365536e-06,
+      "loss": 1.0664,
+      "step": 1422
+    },
+    {
+      "epoch": 2.731637061929909,
+      "grad_norm": 0.6171721436312619,
+      "learning_rate": 5.073712174333182e-06,
+      "loss": 1.3482,
+      "step": 1423
+    },
+    {
+      "epoch": 2.733557369179069,
+      "grad_norm": 0.6148374132384409,
+      "learning_rate": 5.067011488889944e-06,
+      "loss": 1.1996,
+      "step": 1424
+    },
+    {
+      "epoch": 2.7354776764282285,
+      "grad_norm": 0.5915123714502185,
+      "learning_rate": 5.060310683072417e-06,
+      "loss": 1.3484,
+      "step": 1425
+    },
+    {
+      "epoch": 2.7373979836773885,
+      "grad_norm": 0.5320933394573898,
+      "learning_rate": 5.053609768917414e-06,
+      "loss": 1.3501,
+      "step": 1426
+    },
+    {
+      "epoch": 2.739318290926548,
+      "grad_norm": 0.6105600950051874,
+      "learning_rate": 5.0469087584619435e-06,
+      "loss": 1.0733,
+      "step": 1427
+    },
+    {
+      "epoch": 2.741238598175708,
+      "grad_norm": 0.5991959421709251,
+      "learning_rate": 5.040207663743182e-06,
+      "loss": 1.2451,
+      "step": 1428
+    },
+    {
+      "epoch": 2.7431589054248677,
+      "grad_norm": 0.6231899390190498,
+      "learning_rate": 5.033506496798466e-06,
+      "loss": 1.2847,
+      "step": 1429
+    },
+    {
+      "epoch": 2.7450792126740278,
+      "grad_norm": 0.5893989916209164,
+      "learning_rate": 5.026805269665254e-06,
+      "loss": 1.2254,
+      "step": 1430
+    },
+    {
+      "epoch": 2.746999519923188,
+      "grad_norm": 0.6084702626959012,
+      "learning_rate": 5.020103994381114e-06,
+      "loss": 1.2172,
+      "step": 1431
+    },
+    {
+      "epoch": 2.7489198271723474,
+      "grad_norm": 0.6097212003787716,
+      "learning_rate": 5.013402682983705e-06,
+      "loss": 1.1736,
+      "step": 1432
+    },
+    {
+      "epoch": 2.7508401344215074,
+      "grad_norm": 0.6318957281734904,
+      "learning_rate": 5.006701347510745e-06,
+      "loss": 1.3208,
+      "step": 1433
+    },
+    {
+      "epoch": 2.7527604416706675,
+      "grad_norm": 0.5993572834373231,
+      "learning_rate": 5e-06,
+      "loss": 1.3066,
+      "step": 1434
+    },
+    {
+      "epoch": 2.754680748919827,
+      "grad_norm": 0.650769528122097,
+      "learning_rate": 4.9932986524892554e-06,
+      "loss": 1.3586,
+      "step": 1435
+    },
+    {
+      "epoch": 2.756601056168987,
+      "grad_norm": 0.5228860007583129,
+      "learning_rate": 4.986597317016298e-06,
+      "loss": 1.2146,
+      "step": 1436
+    },
+    {
+      "epoch": 2.758521363418147,
+      "grad_norm": 0.6789312600215679,
+      "learning_rate": 4.979896005618887e-06,
+      "loss": 1.1601,
+      "step": 1437
+    },
+    {
+      "epoch": 2.7604416706673067,
+      "grad_norm": 0.6510933300140923,
+      "learning_rate": 4.9731947303347485e-06,
+      "loss": 1.169,
+      "step": 1438
+    },
+    {
+      "epoch": 2.7623619779164668,
+      "grad_norm": 0.4995020628757473,
+      "learning_rate": 4.966493503201537e-06,
+      "loss": 1.17,
+      "step": 1439
+    },
+    {
+      "epoch": 2.7642822851656264,
+      "grad_norm": 0.5852462719837537,
+      "learning_rate": 4.959792336256819e-06,
+      "loss": 1.307,
+      "step": 1440
+    },
+    {
+      "epoch": 2.7662025924147864,
+      "grad_norm": 0.5619937862506597,
+      "learning_rate": 4.953091241538058e-06,
+      "loss": 1.2523,
+      "step": 1441
+    },
+    {
+      "epoch": 2.768122899663946,
+      "grad_norm": 0.5729798655879715,
+      "learning_rate": 4.946390231082586e-06,
+      "loss": 1.1788,
+      "step": 1442
+    },
+    {
+      "epoch": 2.770043206913106,
+      "grad_norm": 0.6120211741613024,
+      "learning_rate": 4.939689316927584e-06,
+      "loss": 1.1841,
+      "step": 1443
+    },
+    {
+      "epoch": 2.771963514162266,
+      "grad_norm": 0.6194227496714729,
+      "learning_rate": 4.932988511110058e-06,
+      "loss": 1.1851,
+      "step": 1444
+    },
+    {
+      "epoch": 2.7738838214114256,
+      "grad_norm": 0.6589157606436065,
+      "learning_rate": 4.926287825666818e-06,
+      "loss": 1.328,
+      "step": 1445
+    },
+    {
+      "epoch": 2.7758041286605857,
+      "grad_norm": 0.5867467759483119,
+      "learning_rate": 4.919587272634466e-06,
+      "loss": 1.2535,
+      "step": 1446
+    },
+    {
+      "epoch": 2.7777244359097457,
+      "grad_norm": 0.6274013769868285,
+      "learning_rate": 4.9128868640493556e-06,
+      "loss": 1.239,
+      "step": 1447
+    },
+    {
+      "epoch": 2.7796447431589053,
+      "grad_norm": 0.5473097320742439,
+      "learning_rate": 4.906186611947587e-06,
+      "loss": 1.2383,
+      "step": 1448
+    },
+    {
+      "epoch": 2.7815650504080653,
+      "grad_norm": 0.6396556685335861,
+      "learning_rate": 4.89948652836498e-06,
+      "loss": 1.2159,
+      "step": 1449
+    },
+    {
+      "epoch": 2.7834853576572254,
+      "grad_norm": 0.5649886726795607,
+      "learning_rate": 4.892786625337047e-06,
+      "loss": 1.2335,
+      "step": 1450
+    },
+    {
+      "epoch": 2.785405664906385,
+      "grad_norm": 0.591568752371345,
+      "learning_rate": 4.886086914898982e-06,
+      "loss": 1.2937,
+      "step": 1451
+    },
+    {
+      "epoch": 2.787325972155545,
+      "grad_norm": 0.6259062421801402,
+      "learning_rate": 4.879387409085628e-06,
+      "loss": 1.3336,
+      "step": 1452
+    },
+    {
+      "epoch": 2.7892462794047046,
+      "grad_norm": 0.6111028684284887,
+      "learning_rate": 4.8726881199314615e-06,
+      "loss": 1.3996,
+      "step": 1453
+    },
+    {
+      "epoch": 2.7911665866538646,
+      "grad_norm": 0.6073121787268624,
+      "learning_rate": 4.865989059470572e-06,
+      "loss": 1.2477,
+      "step": 1454
+    },
+    {
+      "epoch": 2.7930868939030242,
+      "grad_norm": 0.6160489072141255,
+      "learning_rate": 4.859290239736637e-06,
+      "loss": 1.2308,
+      "step": 1455
+    },
+    {
+      "epoch": 2.7950072011521843,
+      "grad_norm": 0.616064300242587,
+      "learning_rate": 4.8525916727629025e-06,
+      "loss": 1.4222,
+      "step": 1456
+    },
+    {
+      "epoch": 2.7969275084013443,
+      "grad_norm": 0.599425508718928,
+      "learning_rate": 4.845893370582156e-06,
+      "loss": 1.2779,
+      "step": 1457
+    },
+    {
+      "epoch": 2.798847815650504,
+      "grad_norm": 0.621016471101322,
+      "learning_rate": 4.839195345226715e-06,
+      "loss": 1.2773,
+      "step": 1458
+    },
+    {
+      "epoch": 2.800768122899664,
+      "grad_norm": 0.6134704585567262,
+      "learning_rate": 4.832497608728398e-06,
+      "loss": 0.9875,
+      "step": 1459
+    },
+    {
+      "epoch": 2.802688430148824,
+      "grad_norm": 0.6274670165366076,
+      "learning_rate": 4.825800173118503e-06,
+      "loss": 1.3622,
+      "step": 1460
+    },
+    {
+      "epoch": 2.8046087373979836,
+      "grad_norm": 0.5740419397746597,
+      "learning_rate": 4.819103050427788e-06,
+      "loss": 1.2185,
+      "step": 1461
+    },
+    {
+      "epoch": 2.8065290446471436,
+      "grad_norm": 0.5571624343029756,
+      "learning_rate": 4.812406252686453e-06,
+      "loss": 1.2336,
+      "step": 1462
+    },
+    {
+      "epoch": 2.8084493518963036,
+      "grad_norm": 0.7028307041043893,
+      "learning_rate": 4.805709791924106e-06,
+      "loss": 1.2159,
+      "step": 1463
+    },
+    {
+      "epoch": 2.810369659145463,
+      "grad_norm": 0.5857678198418313,
+      "learning_rate": 4.799013680169757e-06,
+      "loss": 1.3844,
+      "step": 1464
+    },
+    {
+      "epoch": 2.8122899663946233,
+      "grad_norm": 0.5664125125624209,
+      "learning_rate": 4.792317929451787e-06,
+      "loss": 1.3233,
+      "step": 1465
+    },
+    {
+      "epoch": 2.814210273643783,
+      "grad_norm": 0.6736424995932656,
+      "learning_rate": 4.785622551797926e-06,
+      "loss": 1.174,
+      "step": 1466
+    },
+    {
+      "epoch": 2.816130580892943,
+      "grad_norm": 0.6185112123870505,
+      "learning_rate": 4.778927559235236e-06,
+      "loss": 1.169,
+      "step": 1467
+    },
+    {
+      "epoch": 2.8180508881421025,
+      "grad_norm": 0.6282965696796463,
+      "learning_rate": 4.7722329637900895e-06,
+      "loss": 1.3449,
+      "step": 1468
+    },
+    {
+      "epoch": 2.8199711953912625,
+      "grad_norm": 0.595392473019611,
+      "learning_rate": 4.765538777488143e-06,
+      "loss": 1.3191,
+      "step": 1469
+    },
+    {
+      "epoch": 2.8218915026404225,
+      "grad_norm": 0.6339955968714532,
+      "learning_rate": 4.758845012354314e-06,
+      "loss": 1.1657,
+      "step": 1470
+    },
+    {
+      "epoch": 2.823811809889582,
+      "grad_norm": 0.5884945613572157,
+      "learning_rate": 4.752151680412774e-06,
+      "loss": 1.1951,
+      "step": 1471
+    },
+    {
+      "epoch": 2.825732117138742,
+      "grad_norm": 0.570157509946084,
+      "learning_rate": 4.745458793686906e-06,
+      "loss": 1.0683,
+      "step": 1472
+    },
+    {
+      "epoch": 2.827652424387902,
+      "grad_norm": 0.5701706927749041,
+      "learning_rate": 4.738766364199298e-06,
+      "loss": 1.2678,
+      "step": 1473
+    },
+    {
+      "epoch": 2.829572731637062,
+      "grad_norm": 0.6821992872870765,
+      "learning_rate": 4.732074403971716e-06,
+      "loss": 1.0771,
+      "step": 1474
+    },
+    {
+      "epoch": 2.831493038886222,
+      "grad_norm": 0.5742447400668917,
+      "learning_rate": 4.725382925025085e-06,
+      "loss": 1.3243,
+      "step": 1475
+    },
+    {
+      "epoch": 2.833413346135382,
+      "grad_norm": 0.5073154580413796,
+      "learning_rate": 4.718691939379459e-06,
+      "loss": 1.1815,
+      "step": 1476
+    },
+    {
+      "epoch": 2.8353336533845415,
+      "grad_norm": 0.6281127816737985,
+      "learning_rate": 4.712001459054015e-06,
+      "loss": 1.2478,
+      "step": 1477
+    },
+    {
+      "epoch": 2.8372539606337015,
+      "grad_norm": 0.6280724554540403,
+      "learning_rate": 4.705311496067016e-06,
+      "loss": 1.2767,
+      "step": 1478
+    },
+    {
+      "epoch": 2.839174267882861,
+      "grad_norm": 0.6498477298816885,
+      "learning_rate": 4.698622062435797e-06,
+      "loss": 1.4186,
+      "step": 1479
+    },
+    {
+      "epoch": 2.841094575132021,
+      "grad_norm": 0.6271695380655552,
+      "learning_rate": 4.691933170176741e-06,
+      "loss": 1.2015,
+      "step": 1480
+    },
+    {
+      "epoch": 2.8430148823811807,
+      "grad_norm": 0.5491548417377377,
+      "learning_rate": 4.685244831305262e-06,
+      "loss": 1.2355,
+      "step": 1481
+    },
+    {
+      "epoch": 2.8449351896303408,
+      "grad_norm": 0.5864034248211509,
+      "learning_rate": 4.67855705783578e-06,
+      "loss": 1.341,
+      "step": 1482
+    },
+    {
+      "epoch": 2.846855496879501,
+      "grad_norm": 0.651269297725374,
+      "learning_rate": 4.671869861781692e-06,
+      "loss": 1.2618,
+      "step": 1483
+    },
+    {
+      "epoch": 2.8487758041286604,
+      "grad_norm": 0.5736137573243583,
+      "learning_rate": 4.665183255155367e-06,
+      "loss": 1.1813,
+      "step": 1484
+    },
+    {
+      "epoch": 2.8506961113778204,
+      "grad_norm": 0.5966060659621522,
+      "learning_rate": 4.658497249968111e-06,
+      "loss": 1.2838,
+      "step": 1485
+    },
+    {
+      "epoch": 2.8526164186269805,
+      "grad_norm": 0.6436473495005255,
+      "learning_rate": 4.651811858230149e-06,
+      "loss": 1.1568,
+      "step": 1486
+    },
+    {
+      "epoch": 2.85453672587614,
+      "grad_norm": 0.6006426997501972,
+      "learning_rate": 4.645127091950603e-06,
+      "loss": 1.0614,
+      "step": 1487
+    },
+    {
+      "epoch": 2.8564570331253,
+      "grad_norm": 0.563971575777618,
+      "learning_rate": 4.638442963137478e-06,
+      "loss": 1.1671,
+      "step": 1488
+    },
+    {
+      "epoch": 2.85837734037446,
+      "grad_norm": 0.6062151438031776,
+      "learning_rate": 4.631759483797625e-06,
+      "loss": 1.2848,
+      "step": 1489
+    },
+    {
+      "epoch": 2.8602976476236197,
+      "grad_norm": 0.5577440382869018,
+      "learning_rate": 4.625076665936733e-06,
+      "loss": 0.9997,
+      "step": 1490
+    },
+    {
+      "epoch": 2.8622179548727797,
+      "grad_norm": 0.5757168561642692,
+      "learning_rate": 4.6183945215593065e-06,
+      "loss": 1.2167,
+      "step": 1491
+    },
+    {
+      "epoch": 2.86413826212194,
+      "grad_norm": 0.5328563205215284,
+      "learning_rate": 4.6117130626686304e-06,
+      "loss": 1.1324,
+      "step": 1492
+    },
+    {
+      "epoch": 2.8660585693710994,
+      "grad_norm": 0.5623264525950473,
+      "learning_rate": 4.605032301266768e-06,
+      "loss": 1.1798,
+      "step": 1493
+    },
+    {
+      "epoch": 2.867978876620259,
+      "grad_norm": 0.6044089763292664,
+      "learning_rate": 4.5983522493545246e-06,
+      "loss": 1.2068,
+      "step": 1494
+    },
+    {
+      "epoch": 2.869899183869419,
+      "grad_norm": 0.5847969602032593,
+      "learning_rate": 4.591672918931433e-06,
+      "loss": 1.3445,
+      "step": 1495
+    },
+    {
+      "epoch": 2.871819491118579,
+      "grad_norm": 0.7719499870853559,
+      "learning_rate": 4.584994321995725e-06,
+      "loss": 1.2524,
+      "step": 1496
+    },
+    {
+      "epoch": 2.8737397983677386,
+      "grad_norm": 0.6664338376156864,
+      "learning_rate": 4.578316470544323e-06,
+      "loss": 1.4155,
+      "step": 1497
+    },
+    {
+      "epoch": 2.8756601056168987,
+      "grad_norm": 0.7115327797700147,
+      "learning_rate": 4.571639376572806e-06,
+      "loss": 1.2976,
+      "step": 1498
+    },
+    {
+      "epoch": 2.8775804128660587,
+      "grad_norm": 0.613845802385372,
+      "learning_rate": 4.564963052075391e-06,
+      "loss": 1.1501,
+      "step": 1499
+    },
+    {
+      "epoch": 2.8795007201152183,
+      "grad_norm": 0.581581397000781,
+      "learning_rate": 4.558287509044913e-06,
+      "loss": 1.1418,
+      "step": 1500
+    },
+    {
+      "epoch": 2.8814210273643783,
+      "grad_norm": 0.6661672047569515,
+      "learning_rate": 4.551612759472808e-06,
+      "loss": 1.2158,
+      "step": 1501
+    },
+    {
+      "epoch": 2.8833413346135384,
+      "grad_norm": 0.5811427703283559,
+      "learning_rate": 4.544938815349079e-06,
+      "loss": 1.0787,
+      "step": 1502
+    },
+    {
+      "epoch": 2.885261641862698,
+      "grad_norm": 0.6076712464127186,
+      "learning_rate": 4.53826568866229e-06,
+      "loss": 1.3383,
+      "step": 1503
+    },
+    {
+      "epoch": 2.887181949111858,
+      "grad_norm": 0.6240599236817694,
+      "learning_rate": 4.531593391399532e-06,
+      "loss": 1.1371,
+      "step": 1504
+    },
+    {
+      "epoch": 2.889102256361018,
+      "grad_norm": 0.5834223060081416,
+      "learning_rate": 4.524921935546403e-06,
+      "loss": 1.1745,
+      "step": 1505
+    },
+    {
+      "epoch": 2.8910225636101776,
+      "grad_norm": 0.5520494475127871,
+      "learning_rate": 4.5182513330869996e-06,
+      "loss": 1.2237,
+      "step": 1506
+    },
+    {
+      "epoch": 2.8929428708593377,
+      "grad_norm": 0.6261061301979235,
+      "learning_rate": 4.511581596003876e-06,
+      "loss": 1.2101,
+      "step": 1507
+    },
+    {
+      "epoch": 2.8948631781084972,
+      "grad_norm": 0.5647265473440193,
+      "learning_rate": 4.504912736278038e-06,
+      "loss": 1.3304,
+      "step": 1508
+    },
+    {
+      "epoch": 2.8967834853576573,
+      "grad_norm": 0.5480899790116897,
+      "learning_rate": 4.498244765888907e-06,
+      "loss": 1.1861,
+      "step": 1509
+    },
+    {
+      "epoch": 2.898703792606817,
+      "grad_norm": 0.6563010965058005,
+      "learning_rate": 4.491577696814318e-06,
+      "loss": 1.1922,
+      "step": 1510
+    },
+    {
+      "epoch": 2.900624099855977,
+      "grad_norm": 0.6608401850521457,
+      "learning_rate": 4.484911541030481e-06,
+      "loss": 1.2171,
+      "step": 1511
+    },
+    {
+      "epoch": 2.902544407105137,
+      "grad_norm": 0.579456906409508,
+      "learning_rate": 4.478246310511963e-06,
+      "loss": 1.208,
+      "step": 1512
+    },
+    {
+      "epoch": 2.9044647143542965,
+      "grad_norm": 0.5845531368383728,
+      "learning_rate": 4.471582017231673e-06,
+      "loss": 1.2664,
+      "step": 1513
+    },
+    {
+      "epoch": 2.9063850216034566,
+      "grad_norm": 0.6093436425644907,
+      "learning_rate": 4.464918673160837e-06,
+      "loss": 1.4305,
+      "step": 1514
+    },
+    {
+      "epoch": 2.9083053288526166,
+      "grad_norm": 0.610189940208305,
+      "learning_rate": 4.4582562902689726e-06,
+      "loss": 1.1343,
+      "step": 1515
+    },
+    {
+      "epoch": 2.910225636101776,
+      "grad_norm": 0.5634128751018531,
+      "learning_rate": 4.451594880523872e-06,
+      "loss": 1.2433,
+      "step": 1516
+    },
+    {
+      "epoch": 2.9121459433509362,
+      "grad_norm": 0.5726466147422486,
+      "learning_rate": 4.44493445589158e-06,
+      "loss": 1.3141,
+      "step": 1517
+    },
+    {
+      "epoch": 2.9140662506000963,
+      "grad_norm": 0.623536692319963,
+      "learning_rate": 4.438275028336374e-06,
+      "loss": 1.2369,
+      "step": 1518
+    },
+    {
+      "epoch": 2.915986557849256,
+      "grad_norm": 0.6090348325148056,
+      "learning_rate": 4.431616609820734e-06,
+      "loss": 1.3005,
+      "step": 1519
+    },
+    {
+      "epoch": 2.917906865098416,
+      "grad_norm": 0.5454657429606619,
+      "learning_rate": 4.424959212305334e-06,
+      "loss": 1.1802,
+      "step": 1520
+    },
+    {
+      "epoch": 2.9198271723475755,
+      "grad_norm": 0.5865537935296061,
+      "learning_rate": 4.4183028477490104e-06,
+      "loss": 1.2783,
+      "step": 1521
+    },
+    {
+      "epoch": 2.9217474795967355,
+      "grad_norm": 0.6401005519806768,
+      "learning_rate": 4.411647528108744e-06,
+      "loss": 1.2738,
+      "step": 1522
+    },
+    {
+      "epoch": 2.923667786845895,
+      "grad_norm": 0.654060271373098,
+      "learning_rate": 4.40499326533964e-06,
+      "loss": 1.3858,
+      "step": 1523
+    },
+    {
+      "epoch": 2.925588094095055,
+      "grad_norm": 0.5852133615331452,
+      "learning_rate": 4.398340071394906e-06,
+      "loss": 1.1804,
+      "step": 1524
+    },
+    {
+      "epoch": 2.927508401344215,
+      "grad_norm": 0.593873231371995,
+      "learning_rate": 4.391687958225828e-06,
+      "loss": 1.2319,
+      "step": 1525
+    },
+    {
+      "epoch": 2.929428708593375,
+      "grad_norm": 0.5987318760689788,
+      "learning_rate": 4.385036937781747e-06,
+      "loss": 1.358,
+      "step": 1526
+    },
+    {
+      "epoch": 2.931349015842535,
+      "grad_norm": 0.5920609101980319,
+      "learning_rate": 4.378387022010051e-06,
+      "loss": 1.2592,
+      "step": 1527
+    },
+    {
+      "epoch": 2.933269323091695,
+      "grad_norm": 0.5531012218038981,
+      "learning_rate": 4.371738222856134e-06,
+      "loss": 1.156,
+      "step": 1528
+    },
+    {
+      "epoch": 2.9351896303408544,
+      "grad_norm": 0.7449451350986409,
+      "learning_rate": 4.365090552263388e-06,
+      "loss": 1.342,
+      "step": 1529
+    },
+    {
+      "epoch": 2.9371099375900145,
+      "grad_norm": 0.6024741270458807,
+      "learning_rate": 4.358444022173177e-06,
+      "loss": 1.173,
+      "step": 1530
+    },
+    {
+      "epoch": 2.9390302448391745,
+      "grad_norm": 0.5819701472541976,
+      "learning_rate": 4.351798644524819e-06,
+      "loss": 1.1478,
+      "step": 1531
+    },
+    {
+      "epoch": 2.940950552088334,
+      "grad_norm": 0.6380830987617244,
+      "learning_rate": 4.345154431255559e-06,
+      "loss": 1.1626,
+      "step": 1532
+    },
+    {
+      "epoch": 2.942870859337494,
+      "grad_norm": 0.6189462026330061,
+      "learning_rate": 4.338511394300549e-06,
+      "loss": 1.1721,
+      "step": 1533
+    },
+    {
+      "epoch": 2.9447911665866537,
+      "grad_norm": 0.5954787088500139,
+      "learning_rate": 4.331869545592834e-06,
+      "loss": 1.2768,
+      "step": 1534
+    },
+    {
+      "epoch": 2.9467114738358138,
+      "grad_norm": 0.6099943998376384,
+      "learning_rate": 4.325228897063316e-06,
+      "loss": 1.2215,
+      "step": 1535
+    },
+    {
+      "epoch": 2.9486317810849734,
+      "grad_norm": 0.596414187956695,
+      "learning_rate": 4.318589460640748e-06,
+      "loss": 1.2065,
+      "step": 1536
+    },
+    {
+      "epoch": 2.9505520883341334,
+      "grad_norm": 0.5572935791426047,
+      "learning_rate": 4.311951248251706e-06,
+      "loss": 1.3073,
+      "step": 1537
+    },
+    {
+      "epoch": 2.9524723955832934,
+      "grad_norm": 0.6052886400414235,
+      "learning_rate": 4.30531427182056e-06,
+      "loss": 1.2854,
+      "step": 1538
+    },
+    {
+      "epoch": 2.954392702832453,
+      "grad_norm": 0.7271655672959135,
+      "learning_rate": 4.298678543269463e-06,
+      "loss": 1.2081,
+      "step": 1539
+    },
+    {
+      "epoch": 2.956313010081613,
+      "grad_norm": 0.5378551913802601,
+      "learning_rate": 4.292044074518335e-06,
+      "loss": 1.4137,
+      "step": 1540
+    },
+    {
+      "epoch": 2.958233317330773,
+      "grad_norm": 0.5352069611163749,
+      "learning_rate": 4.2854108774848205e-06,
+      "loss": 1.1978,
+      "step": 1541
+    },
+    {
+      "epoch": 2.9601536245799327,
+      "grad_norm": 0.6477438905633637,
+      "learning_rate": 4.278778964084284e-06,
+      "loss": 1.4337,
+      "step": 1542
+    },
+    {
+      "epoch": 2.9620739318290927,
+      "grad_norm": 0.6350641894866229,
+      "learning_rate": 4.272148346229789e-06,
+      "loss": 1.343,
+      "step": 1543
+    },
+    {
+      "epoch": 2.9639942390782528,
+      "grad_norm": 0.648098579992315,
+      "learning_rate": 4.2655190358320665e-06,
+      "loss": 1.1793,
+      "step": 1544
+    },
+    {
+      "epoch": 2.9659145463274124,
+      "grad_norm": 0.5675147431907982,
+      "learning_rate": 4.2588910447994984e-06,
+      "loss": 1.1826,
+      "step": 1545
+    },
+    {
+      "epoch": 2.9678348535765724,
+      "grad_norm": 0.5562251145463197,
+      "learning_rate": 4.2522643850380985e-06,
+      "loss": 1.3042,
+      "step": 1546
+    },
+    {
+      "epoch": 2.969755160825732,
+      "grad_norm": 0.5951722972940838,
+      "learning_rate": 4.2456390684514934e-06,
+      "loss": 1.1933,
+      "step": 1547
+    },
+    {
+      "epoch": 2.971675468074892,
+      "grad_norm": 0.585992209920053,
+      "learning_rate": 4.239015106940887e-06,
+      "loss": 1.2478,
+      "step": 1548
+    },
+    {
+      "epoch": 2.9735957753240516,
+      "grad_norm": 0.6390443418921725,
+      "learning_rate": 4.2323925124050565e-06,
+      "loss": 1.1196,
+      "step": 1549
+    },
+    {
+      "epoch": 2.9755160825732117,
+      "grad_norm": 0.5159014061938866,
+      "learning_rate": 4.225771296740325e-06,
+      "loss": 1.0836,
+      "step": 1550
+    },
+    {
+      "epoch": 2.9774363898223717,
+      "grad_norm": 0.5977392002236628,
+      "learning_rate": 4.2191514718405294e-06,
+      "loss": 1.2605,
+      "step": 1551
+    },
+    {
+      "epoch": 2.9793566970715313,
+      "grad_norm": 0.6357097877139801,
+      "learning_rate": 4.21253304959702e-06,
+      "loss": 1.3157,
+      "step": 1552
+    },
+    {
+      "epoch": 2.9812770043206913,
+      "grad_norm": 0.6322697400391105,
+      "learning_rate": 4.20591604189862e-06,
+      "loss": 1.2,
+      "step": 1553
+    },
+    {
+      "epoch": 2.9831973115698514,
+      "grad_norm": 0.6867060625705013,
+      "learning_rate": 4.1993004606316114e-06,
+      "loss": 1.2566,
+      "step": 1554
+    },
+    {
+      "epoch": 2.985117618819011,
+      "grad_norm": 0.5429761264375036,
+      "learning_rate": 4.192686317679718e-06,
+      "loss": 1.0437,
+      "step": 1555
+    },
+    {
+      "epoch": 2.987037926068171,
+      "grad_norm": 0.5915388732923772,
+      "learning_rate": 4.186073624924077e-06,
+      "loss": 1.1368,
+      "step": 1556
+    },
+    {
+      "epoch": 2.988958233317331,
+      "grad_norm": 0.589920236068399,
+      "learning_rate": 4.179462394243223e-06,
+      "loss": 0.9987,
+      "step": 1557
+    },
+    {
+      "epoch": 2.9908785405664906,
+      "grad_norm": 0.6024351644712322,
+      "learning_rate": 4.172852637513062e-06,
+      "loss": 1.2706,
+      "step": 1558
+    },
+    {
+      "epoch": 2.9927988478156506,
+      "grad_norm": 0.6159633084221685,
+      "learning_rate": 4.16624436660685e-06,
+      "loss": 1.2461,
+      "step": 1559
+    },
+    {
+      "epoch": 2.9947191550648102,
+      "grad_norm": 0.6171382523271884,
+      "learning_rate": 4.1596375933951835e-06,
+      "loss": 1.4,
+      "step": 1560
+    },
+    {
+      "epoch": 2.9966394623139703,
+      "grad_norm": 0.6212508915364183,
+      "learning_rate": 4.153032329745955e-06,
+      "loss": 1.321,
+      "step": 1561
+    },
+    {
+      "epoch": 2.99855976956313,
+      "grad_norm": 0.6648103018447411,
+      "learning_rate": 4.146428587524358e-06,
+      "loss": 1.1697,
+      "step": 1562
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6571383703715333,
+      "learning_rate": 4.139826378592845e-06,
+      "loss": 1.2437,
+      "step": 1563
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2605,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 271122201362432.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}