diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9231 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9796616612811253,
+  "eval_steps": 500,
+  "global_step": 1314,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001520623455616803,
+      "grad_norm": 3.80404000128578,
+      "learning_rate": 9.033333333333334e-07,
+      "loss": 1.3234,
+      "step": 1
+    },
+    {
+      "epoch": 0.003041246911233606,
+      "grad_norm": 3.7593589398056593,
+      "learning_rate": 1.8066666666666668e-06,
+      "loss": 1.2935,
+      "step": 2
+    },
+    {
+      "epoch": 0.0045618703668504085,
+      "grad_norm": 3.5532664758635017,
+      "learning_rate": 2.7100000000000003e-06,
+      "loss": 1.2481,
+      "step": 3
+    },
+    {
+      "epoch": 0.006082493822467212,
+      "grad_norm": 3.48298105333948,
+      "learning_rate": 3.6133333333333336e-06,
+      "loss": 1.2786,
+      "step": 4
+    },
+    {
+      "epoch": 0.0076031172780840145,
+      "grad_norm": 2.7761816462967768,
+      "learning_rate": 4.516666666666667e-06,
+      "loss": 1.233,
+      "step": 5
+    },
+    {
+      "epoch": 0.009123740733700817,
+      "grad_norm": 1.7543427633433306,
+      "learning_rate": 5.420000000000001e-06,
+      "loss": 1.2459,
+      "step": 6
+    },
+    {
+      "epoch": 0.01064436418931762,
+      "grad_norm": 1.6976304839039194,
+      "learning_rate": 6.3233333333333335e-06,
+      "loss": 1.1905,
+      "step": 7
+    },
+    {
+      "epoch": 0.012164987644934424,
+      "grad_norm": 3.06752449258415,
+      "learning_rate": 7.226666666666667e-06,
+      "loss": 1.1417,
+      "step": 8
+    },
+    {
+      "epoch": 0.013685611100551226,
+      "grad_norm": 3.445950798936361,
+      "learning_rate": 8.13e-06,
+      "loss": 1.1746,
+      "step": 9
+    },
+    {
+      "epoch": 0.015206234556168029,
+      "grad_norm": 3.088218157818855,
+      "learning_rate": 9.033333333333334e-06,
+      "loss": 1.1209,
+      "step": 10
+    },
+    {
+      "epoch": 0.01672685801178483,
+      "grad_norm": 3.3669650804349973,
+      "learning_rate": 9.936666666666666e-06,
+      "loss": 1.138,
+      "step": 11
+    },
+    {
+      "epoch": 0.018247481467401634,
+      "grad_norm": 3.1041727231133174,
+      "learning_rate": 1.0840000000000001e-05,
+      "loss": 1.1188,
+      "step": 12
+    },
+    {
+      "epoch": 0.019768104923018438,
+      "grad_norm": 2.21205722327462,
+      "learning_rate": 1.1743333333333335e-05,
+      "loss": 1.0873,
+      "step": 13
+    },
+    {
+      "epoch": 0.02128872837863524,
+      "grad_norm": 1.32813912161683,
+      "learning_rate": 1.2646666666666667e-05,
+      "loss": 1.1163,
+      "step": 14
+    },
+    {
+      "epoch": 0.022809351834252044,
+      "grad_norm": 2.3535108175852097,
+      "learning_rate": 1.355e-05,
+      "loss": 1.1189,
+      "step": 15
+    },
+    {
+      "epoch": 0.024329975289868848,
+      "grad_norm": 2.8032523790378274,
+      "learning_rate": 1.4453333333333334e-05,
+      "loss": 1.0621,
+      "step": 16
+    },
+    {
+      "epoch": 0.025850598745485648,
+      "grad_norm": 1.8636564282922823,
+      "learning_rate": 1.5356666666666668e-05,
+      "loss": 1.0059,
+      "step": 17
+    },
+    {
+      "epoch": 0.02737122220110245,
+      "grad_norm": 1.198430471743954,
+      "learning_rate": 1.626e-05,
+      "loss": 1.0572,
+      "step": 18
+    },
+    {
+      "epoch": 0.028891845656719255,
+      "grad_norm": 1.3874306913597059,
+      "learning_rate": 1.7163333333333332e-05,
+      "loss": 1.024,
+      "step": 19
+    },
+    {
+      "epoch": 0.030412469112336058,
+      "grad_norm": 1.1392396226185426,
+      "learning_rate": 1.8066666666666668e-05,
+      "loss": 1.0426,
+      "step": 20
+    },
+    {
+      "epoch": 0.03193309256795286,
+      "grad_norm": 0.7881813145609246,
+      "learning_rate": 1.897e-05,
+      "loss": 1.0382,
+      "step": 21
+    },
+    {
+      "epoch": 0.03345371602356966,
+      "grad_norm": 1.0470837153138448,
+      "learning_rate": 1.987333333333333e-05,
+      "loss": 0.9916,
+      "step": 22
+    },
+    {
+      "epoch": 0.03497433947918647,
+      "grad_norm": 0.9265643204523712,
+      "learning_rate": 2.077666666666667e-05,
+      "loss": 1.0348,
+      "step": 23
+    },
+    {
+      "epoch": 0.03649496293480327,
+      "grad_norm": 0.8620032399568974,
+      "learning_rate": 2.1680000000000002e-05,
+      "loss": 0.9893,
+      "step": 24
+    },
+    {
+      "epoch": 0.038015586390420075,
+      "grad_norm": 0.963710426135712,
+      "learning_rate": 2.2583333333333335e-05,
+      "loss": 0.9931,
+      "step": 25
+    },
+    {
+      "epoch": 0.039536209846036875,
+      "grad_norm": 0.7104087725888792,
+      "learning_rate": 2.348666666666667e-05,
+      "loss": 0.9723,
+      "step": 26
+    },
+    {
+      "epoch": 0.041056833301653675,
+      "grad_norm": 0.8206318828097199,
+      "learning_rate": 2.4390000000000002e-05,
+      "loss": 1.0048,
+      "step": 27
+    },
+    {
+      "epoch": 0.04257745675727048,
+      "grad_norm": 0.88460359130528,
+      "learning_rate": 2.5293333333333334e-05,
+      "loss": 0.9686,
+      "step": 28
+    },
+    {
+      "epoch": 0.04409808021288728,
+      "grad_norm": 0.6842646410238233,
+      "learning_rate": 2.619666666666667e-05,
+      "loss": 0.9942,
+      "step": 29
+    },
+    {
+      "epoch": 0.04561870366850409,
+      "grad_norm": 0.9420397284384021,
+      "learning_rate": 2.71e-05,
+      "loss": 0.9619,
+      "step": 30
+    },
+    {
+      "epoch": 0.04713932712412089,
+      "grad_norm": 0.7311688429092981,
+      "learning_rate": 2.7099959441789884e-05,
+      "loss": 0.9722,
+      "step": 31
+    },
+    {
+      "epoch": 0.048659950579737696,
+      "grad_norm": 0.8897905291941917,
+      "learning_rate": 2.7099837767402332e-05,
+      "loss": 0.9767,
+      "step": 32
+    },
+    {
+      "epoch": 0.050180574035354496,
+      "grad_norm": 0.7198270553694668,
+      "learning_rate": 2.709963497756574e-05,
+      "loss": 0.9209,
+      "step": 33
+    },
+    {
+      "epoch": 0.051701197490971296,
+      "grad_norm": 0.6437197076975016,
+      "learning_rate": 2.7099351073494104e-05,
+      "loss": 0.954,
+      "step": 34
+    },
+    {
+      "epoch": 0.0532218209465881,
+      "grad_norm": 0.6539693647049317,
+      "learning_rate": 2.7098986056886998e-05,
+      "loss": 0.9281,
+      "step": 35
+    },
+    {
+      "epoch": 0.0547424444022049,
+      "grad_norm": 0.7438647990016182,
+      "learning_rate": 2.7098539929929583e-05,
+      "loss": 0.9198,
+      "step": 36
+    },
+    {
+      "epoch": 0.05626306785782171,
+      "grad_norm": 0.6710026252513194,
+      "learning_rate": 2.7098012695292568e-05,
+      "loss": 0.9535,
+      "step": 37
+    },
+    {
+      "epoch": 0.05778369131343851,
+      "grad_norm": 0.7336846915639782,
+      "learning_rate": 2.709740435613222e-05,
+      "loss": 0.9188,
+      "step": 38
+    },
+    {
+      "epoch": 0.05930431476905531,
+      "grad_norm": 0.7107571686807039,
+      "learning_rate": 2.709671491609034e-05,
+      "loss": 0.9464,
+      "step": 39
+    },
+    {
+      "epoch": 0.060824938224672116,
+      "grad_norm": 0.6815590773111158,
+      "learning_rate": 2.7095944379294215e-05,
+      "loss": 0.9652,
+      "step": 40
+    },
+    {
+      "epoch": 0.062345561680288916,
+      "grad_norm": 0.7639644922873411,
+      "learning_rate": 2.709509275035663e-05,
+      "loss": 0.9305,
+      "step": 41
+    },
+    {
+      "epoch": 0.06386618513590572,
+      "grad_norm": 0.6137196824616831,
+      "learning_rate": 2.709416003437583e-05,
+      "loss": 0.8947,
+      "step": 42
+    },
+    {
+      "epoch": 0.06538680859152253,
+      "grad_norm": 0.730883182696482,
+      "learning_rate": 2.709314623693546e-05,
+      "loss": 0.9284,
+      "step": 43
+    },
+    {
+      "epoch": 0.06690743204713932,
+      "grad_norm": 0.6378775889130142,
+      "learning_rate": 2.7092051364104584e-05,
+      "loss": 0.9319,
+      "step": 44
+    },
+    {
+      "epoch": 0.06842805550275613,
+      "grad_norm": 0.6881392029525518,
+      "learning_rate": 2.709087542243759e-05,
+      "loss": 0.9561,
+      "step": 45
+    },
+    {
+      "epoch": 0.06994867895837294,
+      "grad_norm": 0.7653476020607418,
+      "learning_rate": 2.708961841897421e-05,
+      "loss": 0.9147,
+      "step": 46
+    },
+    {
+      "epoch": 0.07146930241398973,
+      "grad_norm": 0.6329992439002645,
+      "learning_rate": 2.7088280361239425e-05,
+      "loss": 0.9097,
+      "step": 47
+    },
+    {
+      "epoch": 0.07298992586960654,
+      "grad_norm": 0.7176283996505511,
+      "learning_rate": 2.7086861257243455e-05,
+      "loss": 0.923,
+      "step": 48
+    },
+    {
+      "epoch": 0.07451054932522334,
+      "grad_norm": 0.5868244553340308,
+      "learning_rate": 2.7085361115481697e-05,
+      "loss": 0.9099,
+      "step": 49
+    },
+    {
+      "epoch": 0.07603117278084015,
+      "grad_norm": 0.6598413586849197,
+      "learning_rate": 2.7083779944934685e-05,
+      "loss": 0.9198,
+      "step": 50
+    },
+    {
+      "epoch": 0.07755179623645694,
+      "grad_norm": 0.5578106414522218,
+      "learning_rate": 2.7082117755068008e-05,
+      "loss": 0.8798,
+      "step": 51
+    },
+    {
+      "epoch": 0.07907241969207375,
+      "grad_norm": 0.5382560015493899,
+      "learning_rate": 2.708037455583229e-05,
+      "loss": 0.9423,
+      "step": 52
+    },
+    {
+      "epoch": 0.08059304314769056,
+      "grad_norm": 0.7104523737465288,
+      "learning_rate": 2.7078550357663116e-05,
+      "loss": 0.8525,
+      "step": 53
+    },
+    {
+      "epoch": 0.08211366660330735,
+      "grad_norm": 0.6283073524726329,
+      "learning_rate": 2.7076645171480954e-05,
+      "loss": 0.8691,
+      "step": 54
+    },
+    {
+      "epoch": 0.08363429005892416,
+      "grad_norm": 0.6650416502799579,
+      "learning_rate": 2.7074659008691105e-05,
+      "loss": 0.867,
+      "step": 55
+    },
+    {
+      "epoch": 0.08515491351454096,
+      "grad_norm": 0.7744652466782505,
+      "learning_rate": 2.707259188118364e-05,
+      "loss": 0.9067,
+      "step": 56
+    },
+    {
+      "epoch": 0.08667553697015777,
+      "grad_norm": 0.8615866504032589,
+      "learning_rate": 2.7070443801333323e-05,
+      "loss": 0.888,
+      "step": 57
+    },
+    {
+      "epoch": 0.08819616042577456,
+      "grad_norm": 0.7552923225662669,
+      "learning_rate": 2.706821478199952e-05,
+      "loss": 0.9394,
+      "step": 58
+    },
+    {
+      "epoch": 0.08971678388139137,
+      "grad_norm": 0.8827458802779125,
+      "learning_rate": 2.706590483652616e-05,
+      "loss": 0.9333,
+      "step": 59
+    },
+    {
+      "epoch": 0.09123740733700818,
+      "grad_norm": 0.5735989763886916,
+      "learning_rate": 2.7063513978741612e-05,
+      "loss": 0.9045,
+      "step": 60
+    },
+    {
+      "epoch": 0.09275803079262497,
+      "grad_norm": 0.935081852925912,
+      "learning_rate": 2.706104222295863e-05,
+      "loss": 0.9103,
+      "step": 61
+    },
+    {
+      "epoch": 0.09427865424824178,
+      "grad_norm": 0.974887379533721,
+      "learning_rate": 2.7058489583974263e-05,
+      "loss": 0.8588,
+      "step": 62
+    },
+    {
+      "epoch": 0.09579927770385858,
+      "grad_norm": 0.8802808107178965,
+      "learning_rate": 2.7055856077069762e-05,
+      "loss": 0.9196,
+      "step": 63
+    },
+    {
+      "epoch": 0.09731990115947539,
+      "grad_norm": 0.7021620798712934,
+      "learning_rate": 2.7053141718010486e-05,
+      "loss": 0.891,
+      "step": 64
+    },
+    {
+      "epoch": 0.09884052461509218,
+      "grad_norm": 1.088284062839966,
+      "learning_rate": 2.7050346523045816e-05,
+      "loss": 0.8944,
+      "step": 65
+    },
+    {
+      "epoch": 0.10036114807070899,
+      "grad_norm": 0.6749292391517889,
+      "learning_rate": 2.7047470508909053e-05,
+      "loss": 0.8815,
+      "step": 66
+    },
+    {
+      "epoch": 0.1018817715263258,
+      "grad_norm": 0.8723921836119483,
+      "learning_rate": 2.704451369281731e-05,
+      "loss": 0.859,
+      "step": 67
+    },
+    {
+      "epoch": 0.10340239498194259,
+      "grad_norm": 0.7760246702094213,
+      "learning_rate": 2.7041476092471437e-05,
+      "loss": 0.9184,
+      "step": 68
+    },
+    {
+      "epoch": 0.1049230184375594,
+      "grad_norm": 0.7139493353356939,
+      "learning_rate": 2.7038357726055864e-05,
+      "loss": 0.8692,
+      "step": 69
+    },
+    {
+      "epoch": 0.1064436418931762,
+      "grad_norm": 0.6739017082503604,
+      "learning_rate": 2.7035158612238555e-05,
+      "loss": 0.8609,
+      "step": 70
+    },
+    {
+      "epoch": 0.10796426534879301,
+      "grad_norm": 0.7433509963897333,
+      "learning_rate": 2.7031878770170844e-05,
+      "loss": 0.9063,
+      "step": 71
+    },
+    {
+      "epoch": 0.1094848888044098,
+      "grad_norm": 0.6089113038384066,
+      "learning_rate": 2.7028518219487355e-05,
+      "loss": 0.8703,
+      "step": 72
+    },
+    {
+      "epoch": 0.11100551226002661,
+      "grad_norm": 0.7618672475143956,
+      "learning_rate": 2.7025076980305847e-05,
+      "loss": 0.882,
+      "step": 73
+    },
+    {
+      "epoch": 0.11252613571564342,
+      "grad_norm": 0.5112910612081544,
+      "learning_rate": 2.7021555073227146e-05,
+      "loss": 0.8726,
+      "step": 74
+    },
+    {
+      "epoch": 0.11404675917126021,
+      "grad_norm": 0.6949396608818647,
+      "learning_rate": 2.701795251933497e-05,
+      "loss": 0.8671,
+      "step": 75
+    },
+    {
+      "epoch": 0.11556738262687702,
+      "grad_norm": 0.741408465947176,
+      "learning_rate": 2.7014269340195837e-05,
+      "loss": 0.8816,
+      "step": 76
+    },
+    {
+      "epoch": 0.11708800608249383,
+      "grad_norm": 0.6798649172316688,
+      "learning_rate": 2.7010505557858927e-05,
+      "loss": 0.87,
+      "step": 77
+    },
+    {
+      "epoch": 0.11860862953811062,
+      "grad_norm": 0.7927523181373072,
+      "learning_rate": 2.7006661194855928e-05,
+      "loss": 0.8749,
+      "step": 78
+    },
+    {
+      "epoch": 0.12012925299372743,
+      "grad_norm": 0.5635726177412718,
+      "learning_rate": 2.7002736274200943e-05,
+      "loss": 0.8472,
+      "step": 79
+    },
+    {
+      "epoch": 0.12164987644934423,
+      "grad_norm": 0.6854198399728353,
+      "learning_rate": 2.699873081939032e-05,
+      "loss": 0.8673,
+      "step": 80
+    },
+    {
+      "epoch": 0.12317049990496104,
+      "grad_norm": 0.692082771269964,
+      "learning_rate": 2.6994644854402514e-05,
+      "loss": 0.887,
+      "step": 81
+    },
+    {
+      "epoch": 0.12469112336057783,
+      "grad_norm": 0.5839628354594621,
+      "learning_rate": 2.6990478403697964e-05,
+      "loss": 0.8566,
+      "step": 82
+    },
+    {
+      "epoch": 0.12621174681619465,
+      "grad_norm": 0.5365274637626306,
+      "learning_rate": 2.698623149221892e-05,
+      "loss": 0.8246,
+      "step": 83
+    },
+    {
+      "epoch": 0.12773237027181145,
+      "grad_norm": 0.597903169885847,
+      "learning_rate": 2.6981904145389317e-05,
+      "loss": 0.8331,
+      "step": 84
+    },
+    {
+      "epoch": 0.12925299372742824,
+      "grad_norm": 0.5825252240476892,
+      "learning_rate": 2.697749638911461e-05,
+      "loss": 0.8496,
+      "step": 85
+    },
+    {
+      "epoch": 0.13077361718304506,
+      "grad_norm": 0.7057959507531225,
+      "learning_rate": 2.697300824978161e-05,
+      "loss": 0.9079,
+      "step": 86
+    },
+    {
+      "epoch": 0.13229424063866185,
+      "grad_norm": 0.5509650257453511,
+      "learning_rate": 2.6968439754258348e-05,
+      "loss": 0.8447,
+      "step": 87
+    },
+    {
+      "epoch": 0.13381486409427865,
+      "grad_norm": 0.6392433098525124,
+      "learning_rate": 2.6963790929893908e-05,
+      "loss": 0.8563,
+      "step": 88
+    },
+    {
+      "epoch": 0.13533548754989547,
+      "grad_norm": 0.6150296701472108,
+      "learning_rate": 2.695906180451825e-05,
+      "loss": 0.8746,
+      "step": 89
+    },
+    {
+      "epoch": 0.13685611100551226,
+      "grad_norm": 0.5243071440543693,
+      "learning_rate": 2.6954252406442054e-05,
+      "loss": 0.8462,
+      "step": 90
+    },
+    {
+      "epoch": 0.13837673446112905,
+      "grad_norm": 0.5930951445222234,
+      "learning_rate": 2.6949362764456548e-05,
+      "loss": 0.8579,
+      "step": 91
+    },
+    {
+      "epoch": 0.13989735791674587,
+      "grad_norm": 0.5384200127653291,
+      "learning_rate": 2.694439290783334e-05,
+      "loss": 0.8442,
+      "step": 92
+    },
+    {
+      "epoch": 0.14141798137236267,
+      "grad_norm": 0.4526509726759922,
+      "learning_rate": 2.693934286632423e-05,
+      "loss": 0.84,
+      "step": 93
+    },
+    {
+      "epoch": 0.14293860482797946,
+      "grad_norm": 0.5941168775185707,
+      "learning_rate": 2.6934212670161057e-05,
+      "loss": 0.8715,
+      "step": 94
+    },
+    {
+      "epoch": 0.14445922828359628,
+      "grad_norm": 0.5044642393405714,
+      "learning_rate": 2.6929002350055486e-05,
+      "loss": 0.8815,
+      "step": 95
+    },
+    {
+      "epoch": 0.14597985173921307,
+      "grad_norm": 0.5805569476986123,
+      "learning_rate": 2.6923711937198847e-05,
+      "loss": 0.8678,
+      "step": 96
+    },
+    {
+      "epoch": 0.14750047519482987,
+      "grad_norm": 0.4790647294772535,
+      "learning_rate": 2.6918341463261945e-05,
+      "loss": 0.8557,
+      "step": 97
+    },
+    {
+      "epoch": 0.1490210986504467,
+      "grad_norm": 0.5484601209066985,
+      "learning_rate": 2.691289096039486e-05,
+      "loss": 0.8565,
+      "step": 98
+    },
+    {
+      "epoch": 0.15054172210606348,
+      "grad_norm": 0.5095445371601812,
+      "learning_rate": 2.6907360461226763e-05,
+      "loss": 0.8785,
+      "step": 99
+    },
+    {
+      "epoch": 0.1520623455616803,
+      "grad_norm": 0.5604825268976987,
+      "learning_rate": 2.6901749998865718e-05,
+      "loss": 0.891,
+      "step": 100
+    },
+    {
+      "epoch": 0.1535829690172971,
+      "grad_norm": 0.49509934406592176,
+      "learning_rate": 2.6896059606898493e-05,
+      "loss": 0.8456,
+      "step": 101
+    },
+    {
+      "epoch": 0.1551035924729139,
+      "grad_norm": 0.6476921749987207,
+      "learning_rate": 2.6890289319390343e-05,
+      "loss": 0.8217,
+      "step": 102
+    },
+    {
+      "epoch": 0.1566242159285307,
+      "grad_norm": 0.4780896251811207,
+      "learning_rate": 2.688443917088481e-05,
+      "loss": 0.8711,
+      "step": 103
+    },
+    {
+      "epoch": 0.1581448393841475,
+      "grad_norm": 0.6545666658477076,
+      "learning_rate": 2.687850919640353e-05,
+      "loss": 0.8444,
+      "step": 104
+    },
+    {
+      "epoch": 0.1596654628397643,
+      "grad_norm": 0.61000177257977,
+      "learning_rate": 2.687249943144601e-05,
+      "loss": 0.8433,
+      "step": 105
+    },
+    {
+      "epoch": 0.16118608629538111,
+      "grad_norm": 0.5126707787512214,
+      "learning_rate": 2.6866409911989412e-05,
+      "loss": 0.839,
+      "step": 106
+    },
+    {
+      "epoch": 0.1627067097509979,
+      "grad_norm": 0.4882732798086082,
+      "learning_rate": 2.6860240674488355e-05,
+      "loss": 0.8098,
+      "step": 107
+    },
+    {
+      "epoch": 0.1642273332066147,
+      "grad_norm": 0.5554067699143529,
+      "learning_rate": 2.685399175587468e-05,
+      "loss": 0.8406,
+      "step": 108
+    },
+    {
+      "epoch": 0.16574795666223152,
+      "grad_norm": 0.530718125213625,
+      "learning_rate": 2.6847663193557236e-05,
+      "loss": 0.8281,
+      "step": 109
+    },
+    {
+      "epoch": 0.16726858011784831,
+      "grad_norm": 0.4949823541094598,
+      "learning_rate": 2.684125502542165e-05,
+      "loss": 0.8457,
+      "step": 110
+    },
+    {
+      "epoch": 0.1687892035734651,
+      "grad_norm": 0.556274418077457,
+      "learning_rate": 2.683476728983012e-05,
+      "loss": 0.8543,
+      "step": 111
+    },
+    {
+      "epoch": 0.17030982702908193,
+      "grad_norm": 0.45222659620728906,
+      "learning_rate": 2.682820002562116e-05,
+      "loss": 0.8348,
+      "step": 112
+    },
+    {
+      "epoch": 0.17183045048469872,
+      "grad_norm": 0.4911581618988163,
+      "learning_rate": 2.6821553272109377e-05,
+      "loss": 0.8276,
+      "step": 113
+    },
+    {
+      "epoch": 0.17335107394031554,
+      "grad_norm": 0.5259633420118487,
+      "learning_rate": 2.6814827069085237e-05,
+      "loss": 0.8319,
+      "step": 114
+    },
+    {
+      "epoch": 0.17487169739593234,
+      "grad_norm": 0.49901119248436054,
+      "learning_rate": 2.6808021456814832e-05,
+      "loss": 0.8608,
+      "step": 115
+    },
+    {
+      "epoch": 0.17639232085154913,
+      "grad_norm": 0.44652937902498396,
+      "learning_rate": 2.6801136476039637e-05,
+      "loss": 0.8363,
+      "step": 116
+    },
+    {
+      "epoch": 0.17791294430716595,
+      "grad_norm": 0.4814399942165516,
+      "learning_rate": 2.6794172167976247e-05,
+      "loss": 0.8291,
+      "step": 117
+    },
+    {
+      "epoch": 0.17943356776278274,
+      "grad_norm": 0.4877382938090122,
+      "learning_rate": 2.6787128574316158e-05,
+      "loss": 0.8547,
+      "step": 118
+    },
+    {
+      "epoch": 0.18095419121839953,
+      "grad_norm": 0.6191110354751792,
+      "learning_rate": 2.6780005737225512e-05,
+      "loss": 0.81,
+      "step": 119
+    },
+    {
+      "epoch": 0.18247481467401636,
+      "grad_norm": 0.5437167197085974,
+      "learning_rate": 2.677280369934482e-05,
+      "loss": 0.8393,
+      "step": 120
+    },
+    {
+      "epoch": 0.18399543812963315,
+      "grad_norm": 0.5297828079815319,
+      "learning_rate": 2.676552250378873e-05,
+      "loss": 0.7954,
+      "step": 121
+    },
+    {
+      "epoch": 0.18551606158524994,
+      "grad_norm": 0.5514706250971693,
+      "learning_rate": 2.6758162194145783e-05,
+      "loss": 0.815,
+      "step": 122
+    },
+    {
+      "epoch": 0.18703668504086676,
+      "grad_norm": 0.5843944412164658,
+      "learning_rate": 2.6750722814478098e-05,
+      "loss": 0.8276,
+      "step": 123
+    },
+    {
+      "epoch": 0.18855730849648356,
+      "grad_norm": 0.7942439470942368,
+      "learning_rate": 2.6743204409321177e-05,
+      "loss": 0.8254,
+      "step": 124
+    },
+    {
+      "epoch": 0.19007793195210035,
+      "grad_norm": 0.6174070370639455,
+      "learning_rate": 2.673560702368358e-05,
+      "loss": 0.8549,
+      "step": 125
+    },
+    {
+      "epoch": 0.19159855540771717,
+      "grad_norm": 0.49426597944430567,
+      "learning_rate": 2.6727930703046695e-05,
+      "loss": 0.8447,
+      "step": 126
+    },
+    {
+      "epoch": 0.19311917886333396,
+      "grad_norm": 0.4989733325059912,
+      "learning_rate": 2.6720175493364437e-05,
+      "loss": 0.8499,
+      "step": 127
+    },
+    {
+      "epoch": 0.19463980231895078,
+      "grad_norm": 0.5741663400348602,
+      "learning_rate": 2.6712341441063006e-05,
+      "loss": 0.8675,
+      "step": 128
+    },
+    {
+      "epoch": 0.19616042577456758,
+      "grad_norm": 0.532503832446581,
+      "learning_rate": 2.6704428593040568e-05,
+      "loss": 0.8053,
+      "step": 129
+    },
+    {
+      "epoch": 0.19768104923018437,
+      "grad_norm": 0.47559928839451404,
+      "learning_rate": 2.6696436996667005e-05,
+      "loss": 0.8075,
+      "step": 130
+    },
+    {
+      "epoch": 0.1992016726858012,
+      "grad_norm": 0.5372739975864717,
+      "learning_rate": 2.6688366699783625e-05,
+      "loss": 0.8565,
+      "step": 131
+    },
+    {
+      "epoch": 0.20072229614141798,
+      "grad_norm": 0.5575627816877122,
+      "learning_rate": 2.6680217750702874e-05,
+      "loss": 0.8082,
+      "step": 132
+    },
+    {
+      "epoch": 0.20224291959703478,
+      "grad_norm": 0.5868207344499796,
+      "learning_rate": 2.6671990198208038e-05,
+      "loss": 0.781,
+      "step": 133
+    },
+    {
+      "epoch": 0.2037635430526516,
+      "grad_norm": 0.4362747143153947,
+      "learning_rate": 2.6663684091552962e-05,
+      "loss": 0.817,
+      "step": 134
+    },
+    {
+      "epoch": 0.2052841665082684,
+      "grad_norm": 0.5793319639019973,
+      "learning_rate": 2.6655299480461753e-05,
+      "loss": 0.8414,
+      "step": 135
+    },
+    {
+      "epoch": 0.20680478996388518,
+      "grad_norm": 0.5002998693635706,
+      "learning_rate": 2.6646836415128478e-05,
+      "loss": 0.8399,
+      "step": 136
+    },
+    {
+      "epoch": 0.208325413419502,
+      "grad_norm": 0.5183039332160528,
+      "learning_rate": 2.6638294946216876e-05,
+      "loss": 0.8388,
+      "step": 137
+    },
+    {
+      "epoch": 0.2098460368751188,
+      "grad_norm": 0.5333906191333849,
+      "learning_rate": 2.6629675124860034e-05,
+      "loss": 0.8579,
+      "step": 138
+    },
+    {
+      "epoch": 0.2113666603307356,
+      "grad_norm": 0.5678231978891924,
+      "learning_rate": 2.662097700266009e-05,
+      "loss": 0.7973,
+      "step": 139
+    },
+    {
+      "epoch": 0.2128872837863524,
+      "grad_norm": 0.6779338312838604,
+      "learning_rate": 2.6612200631687935e-05,
+      "loss": 0.7991,
+      "step": 140
+    },
+    {
+      "epoch": 0.2144079072419692,
+      "grad_norm": 0.4549445363589815,
+      "learning_rate": 2.6603346064482896e-05,
+      "loss": 0.8342,
+      "step": 141
+    },
+    {
+      "epoch": 0.21592853069758602,
+      "grad_norm": 0.5534663745815431,
+      "learning_rate": 2.6594413354052406e-05,
+      "loss": 0.8486,
+      "step": 142
+    },
+    {
+      "epoch": 0.21744915415320282,
+      "grad_norm": 0.5263202593856873,
+      "learning_rate": 2.6585402553871707e-05,
+      "loss": 0.8162,
+      "step": 143
+    },
+    {
+      "epoch": 0.2189697776088196,
+      "grad_norm": 0.5005224606123282,
+      "learning_rate": 2.6576313717883517e-05,
+      "loss": 0.8157,
+      "step": 144
+    },
+    {
+      "epoch": 0.22049040106443643,
+      "grad_norm": 0.4408661056131368,
+      "learning_rate": 2.6567146900497715e-05,
+      "loss": 0.8073,
+      "step": 145
+    },
+    {
+      "epoch": 0.22201102452005322,
+      "grad_norm": 0.5697511997765187,
+      "learning_rate": 2.655790215659101e-05,
+      "loss": 0.814,
+      "step": 146
+    },
+    {
+      "epoch": 0.22353164797567002,
+      "grad_norm": 0.5227769257193289,
+      "learning_rate": 2.654857954150661e-05,
+      "loss": 0.8437,
+      "step": 147
+    },
+    {
+      "epoch": 0.22505227143128684,
+      "grad_norm": 0.5335084066857467,
+      "learning_rate": 2.6539179111053904e-05,
+      "loss": 0.8122,
+      "step": 148
+    },
+    {
+      "epoch": 0.22657289488690363,
+      "grad_norm": 0.6845924526755846,
+      "learning_rate": 2.6529700921508117e-05,
+      "loss": 0.8109,
+      "step": 149
+    },
+    {
+      "epoch": 0.22809351834252042,
+      "grad_norm": 0.5497202515145223,
+      "learning_rate": 2.652014502960997e-05,
+      "loss": 0.8482,
+      "step": 150
+    },
+    {
+      "epoch": 0.22961414179813724,
+      "grad_norm": 0.5925506572004898,
+      "learning_rate": 2.651051149256535e-05,
+      "loss": 0.8182,
+      "step": 151
+    },
+    {
+      "epoch": 0.23113476525375404,
+      "grad_norm": 0.4748040759396778,
+      "learning_rate": 2.6500800368044956e-05,
+      "loss": 0.8236,
+      "step": 152
+    },
+    {
+      "epoch": 0.23265538870937083,
+      "grad_norm": 0.4949711715480288,
+      "learning_rate": 2.6491011714183972e-05,
+      "loss": 0.8044,
+      "step": 153
+    },
+    {
+      "epoch": 0.23417601216498765,
+      "grad_norm": 0.4942784715771247,
+      "learning_rate": 2.6481145589581697e-05,
+      "loss": 0.8288,
+      "step": 154
+    },
+    {
+      "epoch": 0.23569663562060444,
+      "grad_norm": 0.5323652293681927,
+      "learning_rate": 2.647120205330121e-05,
+      "loss": 0.7997,
+      "step": 155
+    },
+    {
+      "epoch": 0.23721725907622124,
+      "grad_norm": 0.6476575452897515,
+      "learning_rate": 2.646118116486901e-05,
+      "loss": 0.8077,
+      "step": 156
+    },
+    {
+      "epoch": 0.23873788253183806,
+      "grad_norm": 0.47184034820772325,
+      "learning_rate": 2.6451082984274666e-05,
+      "loss": 0.8292,
+      "step": 157
+    },
+    {
+      "epoch": 0.24025850598745485,
+      "grad_norm": 0.6205069464172376,
+      "learning_rate": 2.6440907571970438e-05,
+      "loss": 0.8393,
+      "step": 158
+    },
+    {
+      "epoch": 0.24177912944307167,
+      "grad_norm": 0.5209484434921768,
+      "learning_rate": 2.6430654988870954e-05,
+      "loss": 0.8245,
+      "step": 159
+    },
+    {
+      "epoch": 0.24329975289868846,
+      "grad_norm": 0.5933368833938595,
+      "learning_rate": 2.6420325296352796e-05,
+      "loss": 0.8197,
+      "step": 160
+    },
+    {
+      "epoch": 0.24482037635430526,
+      "grad_norm": 0.4788998696836293,
+      "learning_rate": 2.6409918556254172e-05,
+      "loss": 0.7937,
+      "step": 161
+    },
+    {
+      "epoch": 0.24634099980992208,
+      "grad_norm": 0.567878399230218,
+      "learning_rate": 2.639943483087453e-05,
+      "loss": 0.8109,
+      "step": 162
+    },
+    {
+      "epoch": 0.24786162326553887,
+      "grad_norm": 0.507261963638979,
+      "learning_rate": 2.6388874182974187e-05,
+      "loss": 0.8218,
+      "step": 163
+    },
+    {
+      "epoch": 0.24938224672115566,
+      "grad_norm": 0.5864902720090787,
+      "learning_rate": 2.637823667577395e-05,
+      "loss": 0.8176,
+      "step": 164
+    },
+    {
+      "epoch": 0.2509028701767725,
+      "grad_norm": 0.5918504624886162,
+      "learning_rate": 2.636752237295474e-05,
+      "loss": 0.83,
+      "step": 165
+    },
+    {
+      "epoch": 0.2524234936323893,
+      "grad_norm": 0.43858309995443784,
+      "learning_rate": 2.6356731338657212e-05,
+      "loss": 0.8123,
+      "step": 166
+    },
+    {
+      "epoch": 0.25394411708800607,
+      "grad_norm": 0.49034964682512816,
+      "learning_rate": 2.6345863637481374e-05,
+      "loss": 0.8174,
+      "step": 167
+    },
+    {
+      "epoch": 0.2554647405436229,
+      "grad_norm": 0.5837987779382616,
+      "learning_rate": 2.6334919334486195e-05,
+      "loss": 0.8298,
+      "step": 168
+    },
+    {
+      "epoch": 0.2569853639992397,
+      "grad_norm": 0.5281077280878048,
+      "learning_rate": 2.6323898495189215e-05,
+      "loss": 0.8175,
+      "step": 169
+    },
+    {
+      "epoch": 0.2585059874548565,
+      "grad_norm": 0.5014520173664685,
+      "learning_rate": 2.6312801185566158e-05,
+      "loss": 0.799,
+      "step": 170
+    },
+    {
+      "epoch": 0.2600266109104733,
+      "grad_norm": 0.6803320146493841,
+      "learning_rate": 2.6301627472050526e-05,
+      "loss": 0.7996,
+      "step": 171
+    },
+    {
+      "epoch": 0.2615472343660901,
+      "grad_norm": 0.7395949415978587,
+      "learning_rate": 2.629037742153322e-05,
+      "loss": 0.8107,
+      "step": 172
+    },
+    {
+      "epoch": 0.2630678578217069,
+      "grad_norm": 0.4513389629082933,
+      "learning_rate": 2.6279051101362122e-05,
+      "loss": 0.7952,
+      "step": 173
+    },
+    {
+      "epoch": 0.2645884812773237,
+      "grad_norm": 0.6486009072322451,
+      "learning_rate": 2.62676485793417e-05,
+      "loss": 0.7874,
+      "step": 174
+    },
+    {
+      "epoch": 0.2661091047329405,
+      "grad_norm": 0.5502319884395829,
+      "learning_rate": 2.62561699237326e-05,
+      "loss": 0.7963,
+      "step": 175
+    },
+    {
+      "epoch": 0.2676297281885573,
+      "grad_norm": 0.5997728345326634,
+      "learning_rate": 2.624461520325124e-05,
+      "loss": 0.7972,
+      "step": 176
+    },
+    {
+      "epoch": 0.2691503516441741,
+      "grad_norm": 0.7905728427904212,
+      "learning_rate": 2.62329844870694e-05,
+      "loss": 0.8029,
+      "step": 177
+    },
+    {
+      "epoch": 0.27067097509979093,
+      "grad_norm": 0.6068096518092964,
+      "learning_rate": 2.62212778448138e-05,
+      "loss": 0.8434,
+      "step": 178
+    },
+    {
+      "epoch": 0.2721915985554077,
+      "grad_norm": 0.8084200216220121,
+      "learning_rate": 2.620949534656568e-05,
+      "loss": 0.8005,
+      "step": 179
+    },
+    {
+      "epoch": 0.2737122220110245,
+      "grad_norm": 0.4950225377554484,
+      "learning_rate": 2.61976370628604e-05,
+      "loss": 0.8261,
+      "step": 180
+    },
+    {
+      "epoch": 0.27523284546664134,
+      "grad_norm": 0.5184301676475981,
+      "learning_rate": 2.618570306468701e-05,
+      "loss": 0.8239,
+      "step": 181
+    },
+    {
+      "epoch": 0.2767534689222581,
+      "grad_norm": 0.601710109353413,
+      "learning_rate": 2.61736934234878e-05,
+      "loss": 0.7888,
+      "step": 182
+    },
+    {
+      "epoch": 0.2782740923778749,
+      "grad_norm": 0.5736633456413043,
+      "learning_rate": 2.616160821115792e-05,
+      "loss": 0.8098,
+      "step": 183
+    },
+    {
+      "epoch": 0.27979471583349175,
+      "grad_norm": 0.5690928538172545,
+      "learning_rate": 2.6149447500044904e-05,
+      "loss": 0.8294,
+      "step": 184
+    },
+    {
+      "epoch": 0.2813153392891085,
+      "grad_norm": 0.6275365250577597,
+      "learning_rate": 2.6137211362948256e-05,
+      "loss": 0.8386,
+      "step": 185
+    },
+    {
+      "epoch": 0.28283596274472533,
+      "grad_norm": 0.6701221466806971,
+      "learning_rate": 2.612489987311903e-05,
+      "loss": 0.791,
+      "step": 186
+    },
+    {
+      "epoch": 0.28435658620034215,
+      "grad_norm": 0.7648465329508422,
+      "learning_rate": 2.6112513104259354e-05,
+      "loss": 0.8125,
+      "step": 187
+    },
+    {
+      "epoch": 0.2858772096559589,
+      "grad_norm": 0.44806704127100555,
+      "learning_rate": 2.6100051130522028e-05,
+      "loss": 0.7849,
+      "step": 188
+    },
+    {
+      "epoch": 0.28739783311157574,
+      "grad_norm": 0.7123885835323628,
+      "learning_rate": 2.608751402651006e-05,
+      "loss": 0.7607,
+      "step": 189
+    },
+    {
+      "epoch": 0.28891845656719256,
+      "grad_norm": 0.4934771024884215,
+      "learning_rate": 2.6074901867276204e-05,
+      "loss": 0.8035,
+      "step": 190
+    },
+    {
+      "epoch": 0.2904390800228093,
+      "grad_norm": 0.8095025862539972,
+      "learning_rate": 2.6062214728322555e-05,
+      "loss": 0.7829,
+      "step": 191
+    },
+    {
+      "epoch": 0.29195970347842615,
+      "grad_norm": 0.5050611453741563,
+      "learning_rate": 2.6049452685600052e-05,
+      "loss": 0.7977,
+      "step": 192
+    },
+    {
+      "epoch": 0.29348032693404297,
+      "grad_norm": 0.7210248910137731,
+      "learning_rate": 2.6036615815508043e-05,
+      "loss": 0.7956,
+      "step": 193
+    },
+    {
+      "epoch": 0.29500095038965973,
+      "grad_norm": 0.5038427749270625,
+      "learning_rate": 2.6023704194893835e-05,
+      "loss": 0.7938,
+      "step": 194
+    },
+    {
+      "epoch": 0.29652157384527655,
+      "grad_norm": 0.6126219303001782,
+      "learning_rate": 2.6010717901052224e-05,
+      "loss": 0.7922,
+      "step": 195
+    },
+    {
+      "epoch": 0.2980421973008934,
+      "grad_norm": 0.5362006682961364,
+      "learning_rate": 2.5997657011725022e-05,
+      "loss": 0.8688,
+      "step": 196
+    },
+    {
+      "epoch": 0.2995628207565102,
+      "grad_norm": 0.5017573346892746,
+      "learning_rate": 2.598452160510062e-05,
+      "loss": 0.7833,
+      "step": 197
+    },
+    {
+      "epoch": 0.30108344421212696,
+      "grad_norm": 0.5086171325358533,
+      "learning_rate": 2.597131175981349e-05,
+      "loss": 0.8173,
+      "step": 198
+    },
+    {
+      "epoch": 0.3026040676677438,
+      "grad_norm": 0.46242832959563607,
+      "learning_rate": 2.5958027554943737e-05,
+      "loss": 0.8228,
+      "step": 199
+    },
+    {
+      "epoch": 0.3041246911233606,
+      "grad_norm": 0.5365620033072362,
+      "learning_rate": 2.594466907001661e-05,
+      "loss": 0.8468,
+      "step": 200
+    },
+    {
+      "epoch": 0.30564531457897737,
+      "grad_norm": 0.4536900282460799,
+      "learning_rate": 2.5931236385002023e-05,
+      "loss": 0.7659,
+      "step": 201
+    },
+    {
+      "epoch": 0.3071659380345942,
+      "grad_norm": 0.5349495426115096,
+      "learning_rate": 2.591772958031411e-05,
+      "loss": 0.8031,
+      "step": 202
+    },
+    {
+      "epoch": 0.308686561490211,
+      "grad_norm": 0.5171572447957907,
+      "learning_rate": 2.5904148736810704e-05,
+      "loss": 0.8192,
+      "step": 203
+    },
+    {
+      "epoch": 0.3102071849458278,
+      "grad_norm": 0.543619906524382,
+      "learning_rate": 2.5890493935792855e-05,
+      "loss": 0.8088,
+      "step": 204
+    },
+    {
+      "epoch": 0.3117278084014446,
+      "grad_norm": 0.5471779246235068,
+      "learning_rate": 2.5876765259004386e-05,
+      "loss": 0.775,
+      "step": 205
+    },
+    {
+      "epoch": 0.3132484318570614,
+      "grad_norm": 0.5530162050934202,
+      "learning_rate": 2.5862962788631346e-05,
+      "loss": 0.7886,
+      "step": 206
+    },
+    {
+      "epoch": 0.3147690553126782,
+      "grad_norm": 0.5564374353682834,
+      "learning_rate": 2.5849086607301564e-05,
+      "loss": 0.7822,
+      "step": 207
+    },
+    {
+      "epoch": 0.316289678768295,
+      "grad_norm": 0.50158890267917,
+      "learning_rate": 2.583513679808412e-05,
+      "loss": 0.7857,
+      "step": 208
+    },
+    {
+      "epoch": 0.3178103022239118,
+      "grad_norm": 0.5514584077868472,
+      "learning_rate": 2.582111344448888e-05,
+      "loss": 0.7834,
+      "step": 209
+    },
+    {
+      "epoch": 0.3193309256795286,
+      "grad_norm": 0.4984807668723581,
+      "learning_rate": 2.580701663046597e-05,
+      "loss": 0.7979,
+      "step": 210
+    },
+    {
+      "epoch": 0.3208515491351454,
+      "grad_norm": 0.5779043543552743,
+      "learning_rate": 2.579284644040529e-05,
+      "loss": 0.8365,
+      "step": 211
+    },
+    {
+      "epoch": 0.32237217259076223,
+      "grad_norm": 0.4914952548115715,
+      "learning_rate": 2.5778602959135996e-05,
+      "loss": 0.8082,
+      "step": 212
+    },
+    {
+      "epoch": 0.323892796046379,
+      "grad_norm": 0.45463188187952797,
+      "learning_rate": 2.5764286271925997e-05,
+      "loss": 0.8002,
+      "step": 213
+    },
+    {
+      "epoch": 0.3254134195019958,
+      "grad_norm": 0.5405224256712096,
+      "learning_rate": 2.5749896464481448e-05,
+      "loss": 0.8177,
+      "step": 214
+    },
+    {
+      "epoch": 0.32693404295761264,
+      "grad_norm": 0.4945710149798116,
+      "learning_rate": 2.5735433622946242e-05,
+      "loss": 0.8202,
+      "step": 215
+    },
+    {
+      "epoch": 0.3284546664132294,
+      "grad_norm": 0.5193387642723397,
+      "learning_rate": 2.572089783390148e-05,
+      "loss": 0.7878,
+      "step": 216
+    },
+    {
+      "epoch": 0.3299752898688462,
+      "grad_norm": 0.43644483818152885,
+      "learning_rate": 2.5706289184364962e-05,
+      "loss": 0.8465,
+      "step": 217
+    },
+    {
+      "epoch": 0.33149591332446304,
+      "grad_norm": 0.6429429755905944,
+      "learning_rate": 2.569160776179065e-05,
+      "loss": 0.793,
+      "step": 218
+    },
+    {
+      "epoch": 0.3330165367800798,
+      "grad_norm": 0.4881784207711263,
+      "learning_rate": 2.567685365406819e-05,
+      "loss": 0.7787,
+      "step": 219
+    },
+    {
+      "epoch": 0.33453716023569663,
+      "grad_norm": 0.5903797222010695,
+      "learning_rate": 2.566202694952232e-05,
+      "loss": 0.8071,
+      "step": 220
+    },
+    {
+      "epoch": 0.33605778369131345,
+      "grad_norm": 0.4935742650802179,
+      "learning_rate": 2.5647127736912397e-05,
+      "loss": 0.8128,
+      "step": 221
+    },
+    {
+      "epoch": 0.3375784071469302,
+      "grad_norm": 0.6206075855689506,
+      "learning_rate": 2.5632156105431836e-05,
+      "loss": 0.8134,
+      "step": 222
+    },
+    {
+      "epoch": 0.33909903060254704,
+      "grad_norm": 0.4693335100552506,
+      "learning_rate": 2.561711214470759e-05,
+      "loss": 0.7877,
+      "step": 223
+    },
+    {
+      "epoch": 0.34061965405816386,
+      "grad_norm": 0.5541352157938085,
+      "learning_rate": 2.560199594479959e-05,
+      "loss": 0.793,
+      "step": 224
+    },
+    {
+      "epoch": 0.3421402775137807,
+      "grad_norm": 0.5119311192678664,
+      "learning_rate": 2.5586807596200247e-05,
+      "loss": 0.8076,
+      "step": 225
+    },
+    {
+      "epoch": 0.34366090096939744,
+      "grad_norm": 0.44821567228976084,
+      "learning_rate": 2.5571547189833876e-05,
+      "loss": 0.7676,
+      "step": 226
+    },
+    {
+      "epoch": 0.34518152442501426,
+      "grad_norm": 0.5108021271320223,
+      "learning_rate": 2.555621481705616e-05,
+      "loss": 0.7761,
+      "step": 227
+    },
+    {
+      "epoch": 0.3467021478806311,
+      "grad_norm": 0.520419930384087,
+      "learning_rate": 2.5540810569653606e-05,
+      "loss": 0.8113,
+      "step": 228
+    },
+    {
+      "epoch": 0.34822277133624785,
+      "grad_norm": 0.49172520245402473,
+      "learning_rate": 2.5525334539843e-05,
+      "loss": 0.8049,
+      "step": 229
+    },
+    {
+      "epoch": 0.34974339479186467,
+      "grad_norm": 0.5350433416575631,
+      "learning_rate": 2.550978682027084e-05,
+      "loss": 0.7918,
+      "step": 230
+    },
+    {
+      "epoch": 0.3512640182474815,
+      "grad_norm": 0.5521639416583181,
+      "learning_rate": 2.5494167504012807e-05,
+      "loss": 0.7705,
+      "step": 231
+    },
+    {
+      "epoch": 0.35278464170309826,
+      "grad_norm": 0.4924879256498397,
+      "learning_rate": 2.547847668457318e-05,
+      "loss": 0.8116,
+      "step": 232
+    },
+    {
+      "epoch": 0.3543052651587151,
+      "grad_norm": 0.5466037905256877,
+      "learning_rate": 2.5462714455884287e-05,
+      "loss": 0.7808,
+      "step": 233
+    },
+    {
+      "epoch": 0.3558258886143319,
+      "grad_norm": 0.5030091896084465,
+      "learning_rate": 2.5446880912305954e-05,
+      "loss": 0.7968,
+      "step": 234
+    },
+    {
+      "epoch": 0.35734651206994866,
+      "grad_norm": 0.47695683680901585,
+      "learning_rate": 2.543097614862492e-05,
+      "loss": 0.7777,
+      "step": 235
+    },
+    {
+      "epoch": 0.3588671355255655,
+      "grad_norm": 0.49323354270252745,
+      "learning_rate": 2.5415000260054287e-05,
+      "loss": 0.8114,
+      "step": 236
+    },
+    {
+      "epoch": 0.3603877589811823,
+      "grad_norm": 0.5547918130480031,
+      "learning_rate": 2.5398953342232942e-05,
+      "loss": 0.7959,
+      "step": 237
+    },
+    {
+      "epoch": 0.36190838243679907,
+      "grad_norm": 0.42592099953025825,
+      "learning_rate": 2.5382835491224977e-05,
+      "loss": 0.8067,
+      "step": 238
+    },
+    {
+      "epoch": 0.3634290058924159,
+      "grad_norm": 0.5473160997053151,
+      "learning_rate": 2.536664680351913e-05,
+      "loss": 0.8149,
+      "step": 239
+    },
+    {
+      "epoch": 0.3649496293480327,
+      "grad_norm": 0.5161750461424985,
+      "learning_rate": 2.5350387376028207e-05,
+      "loss": 0.7456,
+      "step": 240
+    },
+    {
+      "epoch": 0.3664702528036495,
+      "grad_norm": 0.562540015047346,
+      "learning_rate": 2.5334057306088472e-05,
+      "loss": 0.8161,
+      "step": 241
+    },
+    {
+      "epoch": 0.3679908762592663,
+      "grad_norm": 0.5331918164095757,
+      "learning_rate": 2.5317656691459103e-05,
+      "loss": 0.7834,
+      "step": 242
+    },
+    {
+      "epoch": 0.3695114997148831,
+      "grad_norm": 0.5704962131477223,
+      "learning_rate": 2.530118563032159e-05,
+      "loss": 0.7881,
+      "step": 243
+    },
+    {
+      "epoch": 0.3710321231704999,
+      "grad_norm": 0.5208634783184293,
+      "learning_rate": 2.5284644221279147e-05,
+      "loss": 0.8138,
+      "step": 244
+    },
+    {
+      "epoch": 0.3725527466261167,
+      "grad_norm": 0.5211185784219639,
+      "learning_rate": 2.526803256335611e-05,
+      "loss": 0.831,
+      "step": 245
+    },
+    {
+      "epoch": 0.3740733700817335,
+      "grad_norm": 0.5045487224701021,
+      "learning_rate": 2.525135075599738e-05,
+      "loss": 0.8156,
+      "step": 246
+    },
+    {
+      "epoch": 0.3755939935373503,
+      "grad_norm": 0.5069878820838861,
+      "learning_rate": 2.523459889906778e-05,
+      "loss": 0.7955,
+      "step": 247
+    },
+    {
+      "epoch": 0.3771146169929671,
+      "grad_norm": 0.4949818081031886,
+      "learning_rate": 2.5217777092851496e-05,
+      "loss": 0.7941,
+      "step": 248
+    },
+    {
+      "epoch": 0.37863524044858393,
+      "grad_norm": 0.47385420044302173,
+      "learning_rate": 2.5200885438051454e-05,
+      "loss": 0.7897,
+      "step": 249
+    },
+    {
+      "epoch": 0.3801558639042007,
+      "grad_norm": 0.4584812154097583,
+      "learning_rate": 2.5183924035788738e-05,
+      "loss": 0.8149,
+      "step": 250
+    },
+    {
+      "epoch": 0.3816764873598175,
+      "grad_norm": 0.40766128893251813,
+      "learning_rate": 2.5166892987601954e-05,
+      "loss": 0.7965,
+      "step": 251
+    },
+    {
+      "epoch": 0.38319711081543434,
+      "grad_norm": 0.4464504741734316,
+      "learning_rate": 2.5149792395446655e-05,
+      "loss": 0.7793,
+      "step": 252
+    },
+    {
+      "epoch": 0.3847177342710511,
+      "grad_norm": 0.42738515223990925,
+      "learning_rate": 2.5132622361694704e-05,
+      "loss": 0.7957,
+      "step": 253
+    },
+    {
+      "epoch": 0.3862383577266679,
+      "grad_norm": 0.4980456405593658,
+      "learning_rate": 2.5115382989133686e-05,
+      "loss": 0.8032,
+      "step": 254
+    },
+    {
+      "epoch": 0.38775898118228475,
+      "grad_norm": 0.5114356446133503,
+      "learning_rate": 2.5098074380966263e-05,
+      "loss": 0.781,
+      "step": 255
+    },
+    {
+      "epoch": 0.38927960463790157,
+      "grad_norm": 0.4141552486101403,
+      "learning_rate": 2.508069664080959e-05,
+      "loss": 0.7593,
+      "step": 256
+    },
+    {
+      "epoch": 0.39080022809351833,
+      "grad_norm": 0.43458143129055027,
+      "learning_rate": 2.5063249872694662e-05,
+      "loss": 0.7879,
+      "step": 257
+    },
+    {
+      "epoch": 0.39232085154913515,
+      "grad_norm": 0.45941178211274825,
+      "learning_rate": 2.504573418106572e-05,
+      "loss": 0.7947,
+      "step": 258
+    },
+    {
+      "epoch": 0.393841475004752,
+      "grad_norm": 0.49731812944358156,
+      "learning_rate": 2.5028149670779597e-05,
+      "loss": 0.7796,
+      "step": 259
+    },
+    {
+      "epoch": 0.39536209846036874,
+      "grad_norm": 0.4021875079520183,
+      "learning_rate": 2.5010496447105118e-05,
+      "loss": 0.7844,
+      "step": 260
+    },
+    {
+      "epoch": 0.39688272191598556,
+      "grad_norm": 0.5734524869531662,
+      "learning_rate": 2.4992774615722457e-05,
+      "loss": 0.7739,
+      "step": 261
+    },
+    {
+      "epoch": 0.3984033453716024,
+      "grad_norm": 0.4767545528986454,
+      "learning_rate": 2.4974984282722498e-05,
+      "loss": 0.7787,
+      "step": 262
+    },
+    {
+      "epoch": 0.39992396882721915,
+      "grad_norm": 0.45569173851803857,
+      "learning_rate": 2.4957125554606207e-05,
+      "loss": 0.7957,
+      "step": 263
+    },
+    {
+      "epoch": 0.40144459228283597,
+      "grad_norm": 0.4583616284261655,
+      "learning_rate": 2.4939198538284008e-05,
+      "loss": 0.7903,
+      "step": 264
+    },
+    {
+      "epoch": 0.4029652157384528,
+      "grad_norm": 0.44137521284735437,
+      "learning_rate": 2.4921203341075102e-05,
+      "loss": 0.8021,
+      "step": 265
+    },
+    {
+      "epoch": 0.40448583919406955,
+      "grad_norm": 0.4648974770143655,
+      "learning_rate": 2.4903140070706876e-05,
+      "loss": 0.7739,
+      "step": 266
+    },
+    {
+      "epoch": 0.4060064626496864,
+      "grad_norm": 0.521001369014178,
+      "learning_rate": 2.4885008835314228e-05,
+      "loss": 0.7837,
+      "step": 267
+    },
+    {
+      "epoch": 0.4075270861053032,
+      "grad_norm": 0.4601570641740828,
+      "learning_rate": 2.4866809743438915e-05,
+      "loss": 0.7891,
+      "step": 268
+    },
+    {
+      "epoch": 0.40904770956091996,
+      "grad_norm": 0.42883010607831595,
+      "learning_rate": 2.484854290402893e-05,
+      "loss": 0.8002,
+      "step": 269
+    },
+    {
+      "epoch": 0.4105683330165368,
+      "grad_norm": 0.5048746687096498,
+      "learning_rate": 2.4830208426437816e-05,
+      "loss": 0.8067,
+      "step": 270
+    },
+    {
+      "epoch": 0.4120889564721536,
+      "grad_norm": 0.4373843602878167,
+      "learning_rate": 2.4811806420424043e-05,
+      "loss": 0.7661,
+      "step": 271
+    },
+    {
+      "epoch": 0.41360957992777037,
+      "grad_norm": 0.4924768994796526,
+      "learning_rate": 2.479333699615033e-05,
+      "loss": 0.8032,
+      "step": 272
+    },
+    {
+      "epoch": 0.4151302033833872,
+      "grad_norm": 0.5495367191198085,
+      "learning_rate": 2.477480026418299e-05,
+      "loss": 0.7823,
+      "step": 273
+    },
+    {
+      "epoch": 0.416650826839004,
+      "grad_norm": 0.4439519855217945,
+      "learning_rate": 2.4756196335491273e-05,
+      "loss": 0.7655,
+      "step": 274
+    },
+    {
+      "epoch": 0.4181714502946208,
+      "grad_norm": 0.5010590403262122,
+      "learning_rate": 2.47375253214467e-05,
+      "loss": 0.7547,
+      "step": 275
+    },
+    {
+      "epoch": 0.4196920737502376,
+      "grad_norm": 0.4482088714543724,
+      "learning_rate": 2.4718787333822395e-05,
+      "loss": 0.7649,
+      "step": 276
+    },
+    {
+      "epoch": 0.4212126972058544,
+      "grad_norm": 0.40410880639654234,
+      "learning_rate": 2.469998248479242e-05,
+      "loss": 0.7669,
+      "step": 277
+    },
+    {
+      "epoch": 0.4227333206614712,
+      "grad_norm": 0.4569274893231673,
+      "learning_rate": 2.468111088693109e-05,
+      "loss": 0.7526,
+      "step": 278
+    },
+    {
+      "epoch": 0.424253944117088,
+      "grad_norm": 0.42296277869189824,
+      "learning_rate": 2.4662172653212313e-05,
+      "loss": 0.7727,
+      "step": 279
+    },
+    {
+      "epoch": 0.4257745675727048,
+      "grad_norm": 0.4835330143189326,
+      "learning_rate": 2.464316789700891e-05,
+      "loss": 0.7633,
+      "step": 280
+    },
+    {
+      "epoch": 0.4272951910283216,
+      "grad_norm": 0.4287243627654868,
+      "learning_rate": 2.462409673209194e-05,
+      "loss": 0.7836,
+      "step": 281
+    },
+    {
+      "epoch": 0.4288158144839384,
+      "grad_norm": 0.5472776086861091,
+      "learning_rate": 2.4604959272630002e-05,
+      "loss": 0.778,
+      "step": 282
+    },
+    {
+      "epoch": 0.4303364379395552,
+      "grad_norm": 0.43700006561477783,
+      "learning_rate": 2.4585755633188585e-05,
+      "loss": 0.7657,
+      "step": 283
+    },
+    {
+      "epoch": 0.43185706139517205,
+      "grad_norm": 0.5233246056259679,
+      "learning_rate": 2.4566485928729338e-05,
+      "loss": 0.7875,
+      "step": 284
+    },
+    {
+      "epoch": 0.4333776848507888,
+      "grad_norm": 0.4739436376744406,
+      "learning_rate": 2.454715027460942e-05,
+      "loss": 0.7769,
+      "step": 285
+    },
+    {
+      "epoch": 0.43489830830640563,
+      "grad_norm": 0.466587612255068,
+      "learning_rate": 2.4527748786580785e-05,
+      "loss": 0.7763,
+      "step": 286
+    },
+    {
+      "epoch": 0.43641893176202246,
+      "grad_norm": 0.5917227864064121,
+      "learning_rate": 2.4508281580789508e-05,
+      "loss": 0.7643,
+      "step": 287
+    },
+    {
+      "epoch": 0.4379395552176392,
+      "grad_norm": 0.47768374233206695,
+      "learning_rate": 2.448874877377508e-05,
+      "loss": 0.7776,
+      "step": 288
+    },
+    {
+      "epoch": 0.43946017867325604,
+      "grad_norm": 0.527002493093892,
+      "learning_rate": 2.4469150482469695e-05,
+      "loss": 0.7796,
+      "step": 289
+    },
+    {
+      "epoch": 0.44098080212887286,
+      "grad_norm": 0.4662349412142309,
+      "learning_rate": 2.4449486824197585e-05,
+      "loss": 0.7385,
+      "step": 290
+    },
+    {
+      "epoch": 0.4425014255844896,
+      "grad_norm": 0.4732365292282727,
+      "learning_rate": 2.4429757916674292e-05,
+      "loss": 0.7967,
+      "step": 291
+    },
+    {
+      "epoch": 0.44402204904010645,
+      "grad_norm": 0.47222270479478673,
+      "learning_rate": 2.440996387800596e-05,
+      "loss": 0.7682,
+      "step": 292
+    },
+    {
+      "epoch": 0.44554267249572327,
+      "grad_norm": 0.47686692962034305,
+      "learning_rate": 2.439010482668865e-05,
+      "loss": 0.7611,
+      "step": 293
+    },
+    {
+      "epoch": 0.44706329595134003,
+      "grad_norm": 0.4008859391749161,
+      "learning_rate": 2.4370180881607614e-05,
+      "loss": 0.7566,
+      "step": 294
+    },
+    {
+      "epoch": 0.44858391940695685,
+      "grad_norm": 0.48140611283685464,
+      "learning_rate": 2.4350192162036584e-05,
+      "loss": 0.8217,
+      "step": 295
+    },
+    {
+      "epoch": 0.4501045428625737,
+      "grad_norm": 0.4560651975700714,
+      "learning_rate": 2.4330138787637062e-05,
+      "loss": 0.7976,
+      "step": 296
+    },
+    {
+      "epoch": 0.45162516631819044,
+      "grad_norm": 0.4034722549460792,
+      "learning_rate": 2.431002087845761e-05,
+      "loss": 0.7679,
+      "step": 297
+    },
+    {
+      "epoch": 0.45314578977380726,
+      "grad_norm": 0.4807251922703316,
+      "learning_rate": 2.4289838554933113e-05,
+      "loss": 0.7984,
+      "step": 298
+    },
+    {
+      "epoch": 0.4546664132294241,
+      "grad_norm": 0.4746736626277042,
+      "learning_rate": 2.4269591937884083e-05,
+      "loss": 0.7794,
+      "step": 299
+    },
+    {
+      "epoch": 0.45618703668504085,
+      "grad_norm": 0.47543886824696496,
+      "learning_rate": 2.4249281148515904e-05,
+      "loss": 0.7865,
+      "step": 300
+    },
+    {
+      "epoch": 0.45770766014065767,
+      "grad_norm": 0.4982520457747873,
+      "learning_rate": 2.4228906308418148e-05,
+      "loss": 0.7653,
+      "step": 301
+    },
+    {
+      "epoch": 0.4592282835962745,
+      "grad_norm": 0.4541992280820245,
+      "learning_rate": 2.42084675395638e-05,
+      "loss": 0.8036,
+      "step": 302
+    },
+    {
+      "epoch": 0.46074890705189125,
+      "grad_norm": 0.4933234210123189,
+      "learning_rate": 2.4187964964308566e-05,
+      "loss": 0.7764,
+      "step": 303
+    },
+    {
+      "epoch": 0.4622695305075081,
+      "grad_norm": 0.46059558913855275,
+      "learning_rate": 2.4167398705390122e-05,
+      "loss": 0.7736,
+      "step": 304
+    },
+    {
+      "epoch": 0.4637901539631249,
+      "grad_norm": 0.5168855999894718,
+      "learning_rate": 2.414676888592739e-05,
+      "loss": 0.8157,
+      "step": 305
+    },
+    {
+      "epoch": 0.46531077741874166,
+      "grad_norm": 0.46112787242313547,
+      "learning_rate": 2.4126075629419772e-05,
+      "loss": 0.767,
+      "step": 306
+    },
+    {
+      "epoch": 0.4668314008743585,
+      "grad_norm": 0.4302292629253772,
+      "learning_rate": 2.4105319059746465e-05,
+      "loss": 0.7775,
+      "step": 307
+    },
+    {
+      "epoch": 0.4683520243299753,
+      "grad_norm": 0.4152046813472176,
+      "learning_rate": 2.4084499301165655e-05,
+      "loss": 0.8137,
+      "step": 308
+    },
+    {
+      "epoch": 0.46987264778559207,
+      "grad_norm": 0.4876150328408346,
+      "learning_rate": 2.4063616478313835e-05,
+      "loss": 0.7809,
+      "step": 309
+    },
+    {
+      "epoch": 0.4713932712412089,
+      "grad_norm": 0.47894882147037643,
+      "learning_rate": 2.4042670716205003e-05,
+      "loss": 0.8071,
+      "step": 310
+    },
+    {
+      "epoch": 0.4729138946968257,
+      "grad_norm": 0.45314893137296913,
+      "learning_rate": 2.4021662140229957e-05,
+      "loss": 0.7734,
+      "step": 311
+    },
+    {
+      "epoch": 0.4744345181524425,
+      "grad_norm": 0.4693692990401656,
+      "learning_rate": 2.4000590876155523e-05,
+      "loss": 0.8049,
+      "step": 312
+    },
+    {
+      "epoch": 0.4759551416080593,
+      "grad_norm": 0.48467692671175333,
+      "learning_rate": 2.3979457050123804e-05,
+      "loss": 0.7605,
+      "step": 313
+    },
+    {
+      "epoch": 0.4774757650636761,
+      "grad_norm": 0.4222969005662164,
+      "learning_rate": 2.3958260788651426e-05,
+      "loss": 0.7788,
+      "step": 314
+    },
+    {
+      "epoch": 0.47899638851929294,
+      "grad_norm": 0.5542291365632019,
+      "learning_rate": 2.3937002218628792e-05,
+      "loss": 0.7928,
+      "step": 315
+    },
+    {
+      "epoch": 0.4805170119749097,
+      "grad_norm": 0.392436349975769,
+      "learning_rate": 2.3915681467319292e-05,
+      "loss": 0.7607,
+      "step": 316
+    },
+    {
+      "epoch": 0.4820376354305265,
+      "grad_norm": 0.49739152408387094,
+      "learning_rate": 2.3894298662358583e-05,
+      "loss": 0.7614,
+      "step": 317
+    },
+    {
+      "epoch": 0.48355825888614334,
+      "grad_norm": 0.433372030563949,
+      "learning_rate": 2.3872853931753787e-05,
+      "loss": 0.7652,
+      "step": 318
+    },
+    {
+      "epoch": 0.4850788823417601,
+      "grad_norm": 0.46102099141447744,
+      "learning_rate": 2.385134740388275e-05,
+      "loss": 0.774,
+      "step": 319
+    },
+    {
+      "epoch": 0.48659950579737693,
+      "grad_norm": 0.4638491872936722,
+      "learning_rate": 2.382977920749327e-05,
+      "loss": 0.7739,
+      "step": 320
+    },
+    {
+      "epoch": 0.48812012925299375,
+      "grad_norm": 0.5020784828944407,
+      "learning_rate": 2.3808149471702295e-05,
+      "loss": 0.7631,
+      "step": 321
+    },
+    {
+      "epoch": 0.4896407527086105,
+      "grad_norm": 0.42204949679328635,
+      "learning_rate": 2.3786458325995214e-05,
+      "loss": 0.7823,
+      "step": 322
+    },
+    {
+      "epoch": 0.49116137616422734,
+      "grad_norm": 0.5055396085814678,
+      "learning_rate": 2.3764705900225004e-05,
+      "loss": 0.7578,
+      "step": 323
+    },
+    {
+      "epoch": 0.49268199961984416,
+      "grad_norm": 0.47833233111711587,
+      "learning_rate": 2.374289232461152e-05,
+      "loss": 0.7545,
+      "step": 324
+    },
+    {
+      "epoch": 0.4942026230754609,
+      "grad_norm": 0.3848734118646103,
+      "learning_rate": 2.372101772974068e-05,
+      "loss": 0.7141,
+      "step": 325
+    },
+    {
+      "epoch": 0.49572324653107774,
+      "grad_norm": 0.5367848465213906,
+      "learning_rate": 2.3699082246563687e-05,
+      "loss": 0.7789,
+      "step": 326
+    },
+    {
+      "epoch": 0.49724386998669456,
+      "grad_norm": 0.5461740930935474,
+      "learning_rate": 2.367708600639625e-05,
+      "loss": 0.7457,
+      "step": 327
+    },
+    {
+      "epoch": 0.49876449344231133,
+      "grad_norm": 0.4247642096240754,
+      "learning_rate": 2.3655029140917794e-05,
+      "loss": 0.76,
+      "step": 328
+    },
+    {
+      "epoch": 0.5002851168979282,
+      "grad_norm": 0.44998842461109745,
+      "learning_rate": 2.3632911782170683e-05,
+      "loss": 0.7793,
+      "step": 329
+    },
+    {
+      "epoch": 0.501805740353545,
+      "grad_norm": 0.47053242929737366,
+      "learning_rate": 2.3610734062559418e-05,
+      "loss": 0.7858,
+      "step": 330
+    },
+    {
+      "epoch": 0.5033263638091617,
+      "grad_norm": 0.43924849365849383,
+      "learning_rate": 2.3588496114849843e-05,
+      "loss": 0.8032,
+      "step": 331
+    },
+    {
+      "epoch": 0.5048469872647786,
+      "grad_norm": 0.4355576929843276,
+      "learning_rate": 2.3566198072168355e-05,
+      "loss": 0.7691,
+      "step": 332
+    },
+    {
+      "epoch": 0.5063676107203954,
+      "grad_norm": 0.4498593288508342,
+      "learning_rate": 2.354384006800112e-05,
+      "loss": 0.7538,
+      "step": 333
+    },
+    {
+      "epoch": 0.5078882341760121,
+      "grad_norm": 0.42469702355418737,
+      "learning_rate": 2.352142223619323e-05,
+      "loss": 0.7594,
+      "step": 334
+    },
+    {
+      "epoch": 0.509408857631629,
+      "grad_norm": 0.5132890157597293,
+      "learning_rate": 2.3498944710947973e-05,
+      "loss": 0.7693,
+      "step": 335
+    },
+    {
+      "epoch": 0.5109294810872458,
+      "grad_norm": 0.43927719097563,
+      "learning_rate": 2.3476407626825953e-05,
+      "loss": 0.7783,
+      "step": 336
+    },
+    {
+      "epoch": 0.5124501045428626,
+      "grad_norm": 0.4416642559264683,
+      "learning_rate": 2.3453811118744345e-05,
+      "loss": 0.7683,
+      "step": 337
+    },
+    {
+      "epoch": 0.5139707279984794,
+      "grad_norm": 0.44247651582101677,
+      "learning_rate": 2.3431155321976045e-05,
+      "loss": 0.746,
+      "step": 338
+    },
+    {
+      "epoch": 0.5154913514540962,
+      "grad_norm": 0.4243771777807823,
+      "learning_rate": 2.340844037214889e-05,
+      "loss": 0.7735,
+      "step": 339
+    },
+    {
+      "epoch": 0.517011974909713,
+      "grad_norm": 0.45520920364804773,
+      "learning_rate": 2.3385666405244823e-05,
+      "loss": 0.7605,
+      "step": 340
+    },
+    {
+      "epoch": 0.5185325983653298,
+      "grad_norm": 0.41908969886233616,
+      "learning_rate": 2.3362833557599105e-05,
+      "loss": 0.7767,
+      "step": 341
+    },
+    {
+      "epoch": 0.5200532218209466,
+      "grad_norm": 0.41422212342426945,
+      "learning_rate": 2.333994196589947e-05,
+      "loss": 0.7858,
+      "step": 342
+    },
+    {
+      "epoch": 0.5215738452765634,
+      "grad_norm": 0.45155134128630714,
+      "learning_rate": 2.331699176718533e-05,
+      "loss": 0.7436,
+      "step": 343
+    },
+    {
+      "epoch": 0.5230944687321802,
+      "grad_norm": 0.4034165723189431,
+      "learning_rate": 2.3293983098846937e-05,
+      "loss": 0.7615,
+      "step": 344
+    },
+    {
+      "epoch": 0.524615092187797,
+      "grad_norm": 0.41147189894163816,
+      "learning_rate": 2.3270916098624567e-05,
+      "loss": 0.7498,
+      "step": 345
+    },
+    {
+      "epoch": 0.5261357156434138,
+      "grad_norm": 0.4234882343930261,
+      "learning_rate": 2.3247790904607715e-05,
+      "loss": 0.7078,
+      "step": 346
+    },
+    {
+      "epoch": 0.5276563390990306,
+      "grad_norm": 0.4238992343706461,
+      "learning_rate": 2.322460765523423e-05,
+      "loss": 0.769,
+      "step": 347
+    },
+    {
+      "epoch": 0.5291769625546474,
+      "grad_norm": 0.4491430678713249,
+      "learning_rate": 2.3201366489289512e-05,
+      "loss": 0.7742,
+      "step": 348
+    },
+    {
+      "epoch": 0.5306975860102642,
+      "grad_norm": 0.4445057795157137,
+      "learning_rate": 2.317806754590568e-05,
+      "loss": 0.7798,
+      "step": 349
+    },
+    {
+      "epoch": 0.532218209465881,
+      "grad_norm": 0.45149251584132893,
+      "learning_rate": 2.3154710964560735e-05,
+      "loss": 0.7705,
+      "step": 350
+    },
+    {
+      "epoch": 0.5337388329214978,
+      "grad_norm": 0.4379914224804398,
+      "learning_rate": 2.313129688507772e-05,
+      "loss": 0.7465,
+      "step": 351
+    },
+    {
+      "epoch": 0.5352594563771146,
+      "grad_norm": 0.5240644813586705,
+      "learning_rate": 2.3107825447623893e-05,
+      "loss": 0.7792,
+      "step": 352
+    },
+    {
+      "epoch": 0.5367800798327315,
+      "grad_norm": 0.43614968464292103,
+      "learning_rate": 2.3084296792709876e-05,
+      "loss": 0.8012,
+      "step": 353
+    },
+    {
+      "epoch": 0.5383007032883482,
+      "grad_norm": 0.5238927538180428,
+      "learning_rate": 2.3060711061188828e-05,
+      "loss": 0.7121,
+      "step": 354
+    },
+    {
+      "epoch": 0.539821326743965,
+      "grad_norm": 0.44343055771039847,
+      "learning_rate": 2.303706839425559e-05,
+      "loss": 0.7597,
+      "step": 355
+    },
+    {
+      "epoch": 0.5413419501995819,
+      "grad_norm": 0.4760425163011581,
+      "learning_rate": 2.301336893344584e-05,
+      "loss": 0.7686,
+      "step": 356
+    },
+    {
+      "epoch": 0.5428625736551986,
+      "grad_norm": 0.3864387482026591,
+      "learning_rate": 2.2989612820635268e-05,
+      "loss": 0.7634,
+      "step": 357
+    },
+    {
+      "epoch": 0.5443831971108154,
+      "grad_norm": 0.4454248548842092,
+      "learning_rate": 2.2965800198038684e-05,
+      "loss": 0.7462,
+      "step": 358
+    },
+    {
+      "epoch": 0.5459038205664323,
+      "grad_norm": 0.4045116377861445,
+      "learning_rate": 2.2941931208209206e-05,
+      "loss": 0.7538,
+      "step": 359
+    },
+    {
+      "epoch": 0.547424444022049,
+      "grad_norm": 0.45148513549461083,
+      "learning_rate": 2.291800599403739e-05,
+      "loss": 0.7576,
+      "step": 360
+    },
+    {
+      "epoch": 0.5489450674776658,
+      "grad_norm": 0.4144212534792813,
+      "learning_rate": 2.289402469875037e-05,
+      "loss": 0.7679,
+      "step": 361
+    },
+    {
+      "epoch": 0.5504656909332827,
+      "grad_norm": 0.3949666631774562,
+      "learning_rate": 2.2869987465911015e-05,
+      "loss": 0.7334,
+      "step": 362
+    },
+    {
+      "epoch": 0.5519863143888994,
+      "grad_norm": 0.592549458633791,
+      "learning_rate": 2.284589443941706e-05,
+      "loss": 0.766,
+      "step": 363
+    },
+    {
+      "epoch": 0.5535069378445162,
+      "grad_norm": 0.3625351248308574,
+      "learning_rate": 2.2821745763500237e-05,
+      "loss": 0.7502,
+      "step": 364
+    },
+    {
+      "epoch": 0.5550275613001331,
+      "grad_norm": 0.55689868686449,
+      "learning_rate": 2.279754158272543e-05,
+      "loss": 0.741,
+      "step": 365
+    },
+    {
+      "epoch": 0.5565481847557499,
+      "grad_norm": 0.4232933895876237,
+      "learning_rate": 2.2773282041989803e-05,
+      "loss": 0.7203,
+      "step": 366
+    },
+    {
+      "epoch": 0.5580688082113666,
+      "grad_norm": 0.47439658622550684,
+      "learning_rate": 2.274896728652192e-05,
+      "loss": 0.7413,
+      "step": 367
+    },
+    {
+      "epoch": 0.5595894316669835,
+      "grad_norm": 0.4229854780360866,
+      "learning_rate": 2.2724597461880887e-05,
+      "loss": 0.7919,
+      "step": 368
+    },
+    {
+      "epoch": 0.5611100551226003,
+      "grad_norm": 0.449535321441458,
+      "learning_rate": 2.2700172713955486e-05,
+      "loss": 0.7871,
+      "step": 369
+    },
+    {
+      "epoch": 0.562630678578217,
+      "grad_norm": 0.42277942967988674,
+      "learning_rate": 2.267569318896329e-05,
+      "loss": 0.7492,
+      "step": 370
+    },
+    {
+      "epoch": 0.5641513020338339,
+      "grad_norm": 0.4008266318587021,
+      "learning_rate": 2.2651159033449787e-05,
+      "loss": 0.7599,
+      "step": 371
+    },
+    {
+      "epoch": 0.5656719254894507,
+      "grad_norm": 0.46433746216794397,
+      "learning_rate": 2.2626570394287524e-05,
+      "loss": 0.769,
+      "step": 372
+    },
+    {
+      "epoch": 0.5671925489450674,
+      "grad_norm": 0.4052331130445469,
+      "learning_rate": 2.260192741867519e-05,
+      "loss": 0.7584,
+      "step": 373
+    },
+    {
+      "epoch": 0.5687131724006843,
+      "grad_norm": 0.4069059419780127,
+      "learning_rate": 2.2577230254136776e-05,
+      "loss": 0.7605,
+      "step": 374
+    },
+    {
+      "epoch": 0.5702337958563011,
+      "grad_norm": 0.45966108042874787,
+      "learning_rate": 2.2552479048520657e-05,
+      "loss": 0.771,
+      "step": 375
+    },
+    {
+      "epoch": 0.5717544193119178,
+      "grad_norm": 0.4230352997839189,
+      "learning_rate": 2.2527673949998737e-05,
+      "loss": 0.7859,
+      "step": 376
+    },
+    {
+      "epoch": 0.5732750427675347,
+      "grad_norm": 0.42063368235914506,
+      "learning_rate": 2.250281510706553e-05,
+      "loss": 0.7962,
+      "step": 377
+    },
+    {
+      "epoch": 0.5747956662231515,
+      "grad_norm": 0.41245979398339844,
+      "learning_rate": 2.2477902668537307e-05,
+      "loss": 0.7692,
+      "step": 378
+    },
+    {
+      "epoch": 0.5763162896787682,
+      "grad_norm": 0.42587671502291896,
+      "learning_rate": 2.2452936783551174e-05,
+      "loss": 0.7659,
+      "step": 379
+    },
+    {
+      "epoch": 0.5778369131343851,
+      "grad_norm": 0.39899767427234584,
+      "learning_rate": 2.2427917601564194e-05,
+      "loss": 0.7576,
+      "step": 380
+    },
+    {
+      "epoch": 0.5793575365900019,
+      "grad_norm": 0.4504820473523934,
+      "learning_rate": 2.2402845272352495e-05,
+      "loss": 0.7657,
+      "step": 381
+    },
+    {
+      "epoch": 0.5808781600456187,
+      "grad_norm": 0.41018097177091434,
+      "learning_rate": 2.237771994601036e-05,
+      "loss": 0.7477,
+      "step": 382
+    },
+    {
+      "epoch": 0.5823987835012355,
+      "grad_norm": 0.38305001213351025,
+      "learning_rate": 2.235254177294934e-05,
+      "loss": 0.7232,
+      "step": 383
+    },
+    {
+      "epoch": 0.5839194069568523,
+      "grad_norm": 0.47095376013594553,
+      "learning_rate": 2.232731090389736e-05,
+      "loss": 0.7741,
+      "step": 384
+    },
+    {
+      "epoch": 0.5854400304124691,
+      "grad_norm": 0.4636124674958436,
+      "learning_rate": 2.2302027489897786e-05,
+      "loss": 0.7517,
+      "step": 385
+    },
+    {
+      "epoch": 0.5869606538680859,
+      "grad_norm": 0.44034211588944,
+      "learning_rate": 2.2276691682308565e-05,
+      "loss": 0.7418,
+      "step": 386
+    },
+    {
+      "epoch": 0.5884812773237027,
+      "grad_norm": 0.4181155110302704,
+      "learning_rate": 2.2251303632801276e-05,
+      "loss": 0.7565,
+      "step": 387
+    },
+    {
+      "epoch": 0.5900019007793195,
+      "grad_norm": 0.44045436172545194,
+      "learning_rate": 2.2225863493360255e-05,
+      "loss": 0.743,
+      "step": 388
+    },
+    {
+      "epoch": 0.5915225242349363,
+      "grad_norm": 0.3847833966152017,
+      "learning_rate": 2.220037141628167e-05,
+      "loss": 0.7338,
+      "step": 389
+    },
+    {
+      "epoch": 0.5930431476905531,
+      "grad_norm": 0.42280183460806886,
+      "learning_rate": 2.2174827554172604e-05,
+      "loss": 0.7704,
+      "step": 390
+    },
+    {
+      "epoch": 0.59456377114617,
+      "grad_norm": 0.4636797659898542,
+      "learning_rate": 2.2149232059950155e-05,
+      "loss": 0.8102,
+      "step": 391
+    },
+    {
+      "epoch": 0.5960843946017867,
+      "grad_norm": 0.41898988989505653,
+      "learning_rate": 2.2123585086840513e-05,
+      "loss": 0.7743,
+      "step": 392
+    },
+    {
+      "epoch": 0.5976050180574035,
+      "grad_norm": 0.4669187287489928,
+      "learning_rate": 2.2097886788378042e-05,
+      "loss": 0.7809,
+      "step": 393
+    },
+    {
+      "epoch": 0.5991256415130204,
+      "grad_norm": 0.3896079438766509,
+      "learning_rate": 2.2072137318404363e-05,
+      "loss": 0.7482,
+      "step": 394
+    },
+    {
+      "epoch": 0.6006462649686372,
+      "grad_norm": 0.49028722719405726,
+      "learning_rate": 2.2046336831067438e-05,
+      "loss": 0.7412,
+      "step": 395
+    },
+    {
+      "epoch": 0.6021668884242539,
+      "grad_norm": 0.40246044806361353,
+      "learning_rate": 2.2020485480820634e-05,
+      "loss": 0.736,
+      "step": 396
+    },
+    {
+      "epoch": 0.6036875118798708,
+      "grad_norm": 0.39677182146473855,
+      "learning_rate": 2.199458342242181e-05,
+      "loss": 0.7436,
+      "step": 397
+    },
+    {
+      "epoch": 0.6052081353354876,
+      "grad_norm": 0.42836586963405954,
+      "learning_rate": 2.196863081093239e-05,
+      "loss": 0.7649,
+      "step": 398
+    },
+    {
+      "epoch": 0.6067287587911043,
+      "grad_norm": 0.4469293995373356,
+      "learning_rate": 2.1942627801716423e-05,
+      "loss": 0.7442,
+      "step": 399
+    },
+    {
+      "epoch": 0.6082493822467212,
+      "grad_norm": 0.4004274410012105,
+      "learning_rate": 2.1916574550439677e-05,
+      "loss": 0.7694,
+      "step": 400
+    },
+    {
+      "epoch": 0.609770005702338,
+      "grad_norm": 0.4128648841658959,
+      "learning_rate": 2.189047121306867e-05,
+      "loss": 0.7628,
+      "step": 401
+    },
+    {
+      "epoch": 0.6112906291579547,
+      "grad_norm": 0.397595260719658,
+      "learning_rate": 2.186431794586978e-05,
+      "loss": 0.7335,
+      "step": 402
+    },
+    {
+      "epoch": 0.6128112526135716,
+      "grad_norm": 0.39655951051855476,
+      "learning_rate": 2.1838114905408268e-05,
+      "loss": 0.7488,
+      "step": 403
+    },
+    {
+      "epoch": 0.6143318760691884,
+      "grad_norm": 0.3954068854911475,
+      "learning_rate": 2.1811862248547378e-05,
+      "loss": 0.7732,
+      "step": 404
+    },
+    {
+      "epoch": 0.6158524995248051,
+      "grad_norm": 0.4029750154489033,
+      "learning_rate": 2.1785560132447365e-05,
+      "loss": 0.7635,
+      "step": 405
+    },
+    {
+      "epoch": 0.617373122980422,
+      "grad_norm": 0.4988589982990038,
+      "learning_rate": 2.1759208714564575e-05,
+      "loss": 0.7482,
+      "step": 406
+    },
+    {
+      "epoch": 0.6188937464360388,
+      "grad_norm": 0.3830796517420403,
+      "learning_rate": 2.1732808152650507e-05,
+      "loss": 0.7059,
+      "step": 407
+    },
+    {
+      "epoch": 0.6204143698916555,
+      "grad_norm": 0.4928324719853511,
+      "learning_rate": 2.1706358604750843e-05,
+      "loss": 0.75,
+      "step": 408
+    },
+    {
+      "epoch": 0.6219349933472724,
+      "grad_norm": 0.41872118978417633,
+      "learning_rate": 2.1679860229204526e-05,
+      "loss": 0.7799,
+      "step": 409
+    },
+    {
+      "epoch": 0.6234556168028892,
+      "grad_norm": 0.44112153018658945,
+      "learning_rate": 2.1653313184642798e-05,
+      "loss": 0.7554,
+      "step": 410
+    },
+    {
+      "epoch": 0.624976240258506,
+      "grad_norm": 0.44207424430007414,
+      "learning_rate": 2.162671762998826e-05,
+      "loss": 0.7303,
+      "step": 411
+    },
+    {
+      "epoch": 0.6264968637141228,
+      "grad_norm": 0.4287487064749124,
+      "learning_rate": 2.1600073724453913e-05,
+      "loss": 0.7311,
+      "step": 412
+    },
+    {
+      "epoch": 0.6280174871697396,
+      "grad_norm": 0.4795008086486292,
+      "learning_rate": 2.157338162754221e-05,
+      "loss": 0.7538,
+      "step": 413
+    },
+    {
+      "epoch": 0.6295381106253564,
+      "grad_norm": 0.4116045721062874,
+      "learning_rate": 2.1546641499044105e-05,
+      "loss": 0.7526,
+      "step": 414
+    },
+    {
+      "epoch": 0.6310587340809732,
+      "grad_norm": 0.4601608034979367,
+      "learning_rate": 2.1519853499038075e-05,
+      "loss": 0.7891,
+      "step": 415
+    },
+    {
+      "epoch": 0.63257935753659,
+      "grad_norm": 0.43000044965092804,
+      "learning_rate": 2.14930177878892e-05,
+      "loss": 0.7347,
+      "step": 416
+    },
+    {
+      "epoch": 0.6340999809922068,
+      "grad_norm": 0.43136823831520454,
+      "learning_rate": 2.1466134526248154e-05,
+      "loss": 0.778,
+      "step": 417
+    },
+    {
+      "epoch": 0.6356206044478236,
+      "grad_norm": 0.40082789389914825,
+      "learning_rate": 2.1439203875050295e-05,
+      "loss": 0.7435,
+      "step": 418
+    },
+    {
+      "epoch": 0.6371412279034404,
+      "grad_norm": 0.46281961581743375,
+      "learning_rate": 2.1412225995514658e-05,
+      "loss": 0.7644,
+      "step": 419
+    },
+    {
+      "epoch": 0.6386618513590572,
+      "grad_norm": 0.39180679623310466,
+      "learning_rate": 2.1385201049143025e-05,
+      "loss": 0.7401,
+      "step": 420
+    },
+    {
+      "epoch": 0.640182474814674,
+      "grad_norm": 0.44410256733578407,
+      "learning_rate": 2.135812919771892e-05,
+      "loss": 0.749,
+      "step": 421
+    },
+    {
+      "epoch": 0.6417030982702908,
+      "grad_norm": 0.4332510457584124,
+      "learning_rate": 2.133101060330668e-05,
+      "loss": 0.724,
+      "step": 422
+    },
+    {
+      "epoch": 0.6432237217259076,
+      "grad_norm": 0.434836178784241,
+      "learning_rate": 2.1303845428250453e-05,
+      "loss": 0.7905,
+      "step": 423
+    },
+    {
+      "epoch": 0.6447443451815245,
+      "grad_norm": 0.5203348960832226,
+      "learning_rate": 2.127663383517326e-05,
+      "loss": 0.7775,
+      "step": 424
+    },
+    {
+      "epoch": 0.6462649686371412,
+      "grad_norm": 0.4239061363149275,
+      "learning_rate": 2.1249375986975975e-05,
+      "loss": 0.7547,
+      "step": 425
+    },
+    {
+      "epoch": 0.647785592092758,
+      "grad_norm": 0.5531850395502146,
+      "learning_rate": 2.1222072046836397e-05,
+      "loss": 0.7562,
+      "step": 426
+    },
+    {
+      "epoch": 0.6493062155483749,
+      "grad_norm": 0.44407831554002236,
+      "learning_rate": 2.119472217820824e-05,
+      "loss": 0.7563,
+      "step": 427
+    },
+    {
+      "epoch": 0.6508268390039916,
+      "grad_norm": 0.46682309596014615,
+      "learning_rate": 2.1167326544820175e-05,
+      "loss": 0.757,
+      "step": 428
+    },
+    {
+      "epoch": 0.6523474624596084,
+      "grad_norm": 0.4356058035250766,
+      "learning_rate": 2.1139885310674832e-05,
+      "loss": 0.7598,
+      "step": 429
+    },
+    {
+      "epoch": 0.6538680859152253,
+      "grad_norm": 0.4201883171891972,
+      "learning_rate": 2.1112398640047842e-05,
+      "loss": 0.7347,
+      "step": 430
+    },
+    {
+      "epoch": 0.655388709370842,
+      "grad_norm": 0.42395223035483165,
+      "learning_rate": 2.1084866697486813e-05,
+      "loss": 0.7414,
+      "step": 431
+    },
+    {
+      "epoch": 0.6569093328264588,
+      "grad_norm": 0.39048036764414096,
+      "learning_rate": 2.1057289647810407e-05,
+      "loss": 0.7338,
+      "step": 432
+    },
+    {
+      "epoch": 0.6584299562820757,
+      "grad_norm": 0.44299976473368746,
+      "learning_rate": 2.1029667656107295e-05,
+      "loss": 0.7791,
+      "step": 433
+    },
+    {
+      "epoch": 0.6599505797376924,
+      "grad_norm": 0.500893872138014,
+      "learning_rate": 2.100200088773519e-05,
+      "loss": 0.7543,
+      "step": 434
+    },
+    {
+      "epoch": 0.6614712031933092,
+      "grad_norm": 0.49372215416444076,
+      "learning_rate": 2.0974289508319867e-05,
+      "loss": 0.7676,
+      "step": 435
+    },
+    {
+      "epoch": 0.6629918266489261,
+      "grad_norm": 0.46705701804279404,
+      "learning_rate": 2.094653368375416e-05,
+      "loss": 0.7631,
+      "step": 436
+    },
+    {
+      "epoch": 0.6645124501045429,
+      "grad_norm": 0.5152291637276656,
+      "learning_rate": 2.0918733580196976e-05,
+      "loss": 0.778,
+      "step": 437
+    },
+    {
+      "epoch": 0.6660330735601596,
+      "grad_norm": 0.4597895523784997,
+      "learning_rate": 2.089088936407228e-05,
+      "loss": 0.7411,
+      "step": 438
+    },
+    {
+      "epoch": 0.6675536970157765,
+      "grad_norm": 0.43780405947061785,
+      "learning_rate": 2.0863001202068135e-05,
+      "loss": 0.7234,
+      "step": 439
+    },
+    {
+      "epoch": 0.6690743204713933,
+      "grad_norm": 0.48252738543172174,
+      "learning_rate": 2.083506926113568e-05,
+      "loss": 0.7537,
+      "step": 440
+    },
+    {
+      "epoch": 0.67059494392701,
+      "grad_norm": 0.386483976594262,
+      "learning_rate": 2.080709370848812e-05,
+      "loss": 0.7385,
+      "step": 441
+    },
+    {
+      "epoch": 0.6721155673826269,
+      "grad_norm": 0.4545211695758658,
+      "learning_rate": 2.077907471159975e-05,
+      "loss": 0.7669,
+      "step": 442
+    },
+    {
+      "epoch": 0.6736361908382437,
+      "grad_norm": 0.4105864152211242,
+      "learning_rate": 2.075101243820494e-05,
+      "loss": 0.748,
+      "step": 443
+    },
+    {
+      "epoch": 0.6751568142938604,
+      "grad_norm": 0.4504033111976225,
+      "learning_rate": 2.0722907056297144e-05,
+      "loss": 0.7364,
+      "step": 444
+    },
+    {
+      "epoch": 0.6766774377494773,
+      "grad_norm": 0.43800556294426285,
+      "learning_rate": 2.0694758734127862e-05,
+      "loss": 0.7816,
+      "step": 445
+    },
+    {
+      "epoch": 0.6781980612050941,
+      "grad_norm": 0.4525088971872918,
+      "learning_rate": 2.066656764020567e-05,
+      "loss": 0.7795,
+      "step": 446
+    },
+    {
+      "epoch": 0.6797186846607108,
+      "grad_norm": 0.48236560996254046,
+      "learning_rate": 2.0638333943295188e-05,
+      "loss": 0.8176,
+      "step": 447
+    },
+    {
+      "epoch": 0.6812393081163277,
+      "grad_norm": 0.41583212690425186,
+      "learning_rate": 2.0610057812416076e-05,
+      "loss": 0.7233,
+      "step": 448
+    },
+    {
+      "epoch": 0.6827599315719445,
+      "grad_norm": 0.45116571513530795,
+      "learning_rate": 2.0581739416842023e-05,
+      "loss": 0.7537,
+      "step": 449
+    },
+    {
+      "epoch": 0.6842805550275614,
+      "grad_norm": 0.3988713943698696,
+      "learning_rate": 2.0553378926099744e-05,
+      "loss": 0.7815,
+      "step": 450
+    },
+    {
+      "epoch": 0.6858011784831781,
+      "grad_norm": 0.4119407206884566,
+      "learning_rate": 2.0524976509967945e-05,
+      "loss": 0.7884,
+      "step": 451
+    },
+    {
+      "epoch": 0.6873218019387949,
+      "grad_norm": 0.4826174722158836,
+      "learning_rate": 2.0496532338476314e-05,
+      "loss": 0.7514,
+      "step": 452
+    },
+    {
+      "epoch": 0.6888424253944118,
+      "grad_norm": 0.4442915093907275,
+      "learning_rate": 2.046804658190451e-05,
+      "loss": 0.7724,
+      "step": 453
+    },
+    {
+      "epoch": 0.6903630488500285,
+      "grad_norm": 0.4078747197054875,
+      "learning_rate": 2.0439519410781135e-05,
+      "loss": 0.7422,
+      "step": 454
+    },
+    {
+      "epoch": 0.6918836723056453,
+      "grad_norm": 0.40352123173925075,
+      "learning_rate": 2.0410950995882723e-05,
+      "loss": 0.7473,
+      "step": 455
+    },
+    {
+      "epoch": 0.6934042957612622,
+      "grad_norm": 0.37505514036015986,
+      "learning_rate": 2.0382341508232713e-05,
+      "loss": 0.7315,
+      "step": 456
+    },
+    {
+      "epoch": 0.6949249192168789,
+      "grad_norm": 0.4001378862906619,
+      "learning_rate": 2.0353691119100415e-05,
+      "loss": 0.7503,
+      "step": 457
+    },
+    {
+      "epoch": 0.6964455426724957,
+      "grad_norm": 0.4227756951676847,
+      "learning_rate": 2.0325e-05,
+      "loss": 0.7604,
+      "step": 458
+    },
+    {
+      "epoch": 0.6979661661281126,
+      "grad_norm": 0.3771305192233914,
+      "learning_rate": 2.029626832268947e-05,
+      "loss": 0.7336,
+      "step": 459
+    },
+    {
+      "epoch": 0.6994867895837293,
+      "grad_norm": 0.36251000512548065,
+      "learning_rate": 2.0267496259169623e-05,
+      "loss": 0.7139,
+      "step": 460
+    },
+    {
+      "epoch": 0.7010074130393461,
+      "grad_norm": 0.3633369727991047,
+      "learning_rate": 2.023868398168302e-05,
+      "loss": 0.7254,
+      "step": 461
+    },
+    {
+      "epoch": 0.702528036494963,
+      "grad_norm": 0.3722032410661908,
+      "learning_rate": 2.0209831662712973e-05,
+      "loss": 0.7291,
+      "step": 462
+    },
+    {
+      "epoch": 0.7040486599505797,
+      "grad_norm": 0.39237895415677215,
+      "learning_rate": 2.0180939474982496e-05,
+      "loss": 0.7807,
+      "step": 463
+    },
+    {
+      "epoch": 0.7055692834061965,
+      "grad_norm": 0.38128737748687397,
+      "learning_rate": 2.015200759145327e-05,
+      "loss": 0.7525,
+      "step": 464
+    },
+    {
+      "epoch": 0.7070899068618134,
+      "grad_norm": 0.39388751320764215,
+      "learning_rate": 2.0123036185324624e-05,
+      "loss": 0.765,
+      "step": 465
+    },
+    {
+      "epoch": 0.7086105303174302,
+      "grad_norm": 0.4161404920966044,
+      "learning_rate": 2.0094025430032477e-05,
+      "loss": 0.7746,
+      "step": 466
+    },
+    {
+      "epoch": 0.7101311537730469,
+      "grad_norm": 0.38567602060145806,
+      "learning_rate": 2.006497549924831e-05,
+      "loss": 0.7525,
+      "step": 467
+    },
+    {
+      "epoch": 0.7116517772286638,
+      "grad_norm": 0.3878678822202346,
+      "learning_rate": 2.0035886566878142e-05,
+      "loss": 0.7363,
+      "step": 468
+    },
+    {
+      "epoch": 0.7131724006842806,
+      "grad_norm": 0.4189479849973627,
+      "learning_rate": 2.0006758807061447e-05,
+      "loss": 0.7202,
+      "step": 469
+    },
+    {
+      "epoch": 0.7146930241398973,
+      "grad_norm": 0.35897122699509787,
+      "learning_rate": 1.997759239417016e-05,
+      "loss": 0.711,
+      "step": 470
+    },
+    {
+      "epoch": 0.7162136475955142,
+      "grad_norm": 0.3719817122737225,
+      "learning_rate": 1.9948387502807602e-05,
+      "loss": 0.7329,
+      "step": 471
+    },
+    {
+      "epoch": 0.717734271051131,
+      "grad_norm": 0.38587499281174487,
+      "learning_rate": 1.991914430780744e-05,
+      "loss": 0.7414,
+      "step": 472
+    },
+    {
+      "epoch": 0.7192548945067477,
+      "grad_norm": 0.40031974274913,
+      "learning_rate": 1.988986298423265e-05,
+      "loss": 0.7578,
+      "step": 473
+    },
+    {
+      "epoch": 0.7207755179623646,
+      "grad_norm": 0.46518346734697724,
+      "learning_rate": 1.986054370737446e-05,
+      "loss": 0.7399,
+      "step": 474
+    },
+    {
+      "epoch": 0.7222961414179814,
+      "grad_norm": 0.34848519411774526,
+      "learning_rate": 1.983118665275131e-05,
+      "loss": 0.7102,
+      "step": 475
+    },
+    {
+      "epoch": 0.7238167648735981,
+      "grad_norm": 0.3891532280462324,
+      "learning_rate": 1.9801791996107777e-05,
+      "loss": 0.7307,
+      "step": 476
+    },
+    {
+      "epoch": 0.725337388329215,
+      "grad_norm": 0.4460610364235613,
+      "learning_rate": 1.9772359913413563e-05,
+      "loss": 0.7933,
+      "step": 477
+    },
+    {
+      "epoch": 0.7268580117848318,
+      "grad_norm": 0.3931330019630469,
+      "learning_rate": 1.9742890580862405e-05,
+      "loss": 0.7626,
+      "step": 478
+    },
+    {
+      "epoch": 0.7283786352404485,
+      "grad_norm": 0.3869696043446492,
+      "learning_rate": 1.9713384174871036e-05,
+      "loss": 0.7652,
+      "step": 479
+    },
+    {
+      "epoch": 0.7298992586960654,
+      "grad_norm": 0.4142673906264668,
+      "learning_rate": 1.9683840872078135e-05,
+      "loss": 0.7683,
+      "step": 480
+    },
+    {
+      "epoch": 0.7314198821516822,
+      "grad_norm": 0.3758360055485875,
+      "learning_rate": 1.965426084934325e-05,
+      "loss": 0.7352,
+      "step": 481
+    },
+    {
+      "epoch": 0.732940505607299,
+      "grad_norm": 0.3957928284148903,
+      "learning_rate": 1.9624644283745763e-05,
+      "loss": 0.7709,
+      "step": 482
+    },
+    {
+      "epoch": 0.7344611290629158,
+      "grad_norm": 0.4166452839891987,
+      "learning_rate": 1.959499135258381e-05,
+      "loss": 0.7428,
+      "step": 483
+    },
+    {
+      "epoch": 0.7359817525185326,
+      "grad_norm": 0.3930427010522597,
+      "learning_rate": 1.9565302233373233e-05,
+      "loss": 0.7239,
+      "step": 484
+    },
+    {
+      "epoch": 0.7375023759741494,
+      "grad_norm": 0.3601178498038616,
+      "learning_rate": 1.9535577103846507e-05,
+      "loss": 0.7099,
+      "step": 485
+    },
+    {
+      "epoch": 0.7390229994297662,
+      "grad_norm": 0.4753582216812286,
+      "learning_rate": 1.950581614195169e-05,
+      "loss": 0.749,
+      "step": 486
+    },
+    {
+      "epoch": 0.740543622885383,
+      "grad_norm": 0.3946913425771556,
+      "learning_rate": 1.9476019525851328e-05,
+      "loss": 0.7461,
+      "step": 487
+    },
+    {
+      "epoch": 0.7420642463409998,
+      "grad_norm": 0.408093102930642,
+      "learning_rate": 1.9446187433921438e-05,
+      "loss": 0.7433,
+      "step": 488
+    },
+    {
+      "epoch": 0.7435848697966166,
+      "grad_norm": 0.4409320643561924,
+      "learning_rate": 1.941632004475039e-05,
+      "loss": 0.7419,
+      "step": 489
+    },
+    {
+      "epoch": 0.7451054932522334,
+      "grad_norm": 0.4107498993063142,
+      "learning_rate": 1.938641753713787e-05,
+      "loss": 0.7663,
+      "step": 490
+    },
+    {
+      "epoch": 0.7466261167078502,
+      "grad_norm": 0.3992120160222559,
+      "learning_rate": 1.9356480090093787e-05,
+      "loss": 0.7549,
+      "step": 491
+    },
+    {
+      "epoch": 0.748146740163467,
+      "grad_norm": 0.4323109678553807,
+      "learning_rate": 1.9326507882837228e-05,
+      "loss": 0.7578,
+      "step": 492
+    },
+    {
+      "epoch": 0.7496673636190838,
+      "grad_norm": 0.4188967295796775,
+      "learning_rate": 1.929650109479536e-05,
+      "loss": 0.7623,
+      "step": 493
+    },
+    {
+      "epoch": 0.7511879870747006,
+      "grad_norm": 0.4016307344915274,
+      "learning_rate": 1.9266459905602376e-05,
+      "loss": 0.7504,
+      "step": 494
+    },
+    {
+      "epoch": 0.7527086105303175,
+      "grad_norm": 0.45431926716757953,
+      "learning_rate": 1.9236384495098395e-05,
+      "loss": 0.7569,
+      "step": 495
+    },
+    {
+      "epoch": 0.7542292339859342,
+      "grad_norm": 0.37461201818259915,
+      "learning_rate": 1.9206275043328413e-05,
+      "loss": 0.7228,
+      "step": 496
+    },
+    {
+      "epoch": 0.755749857441551,
+      "grad_norm": 0.44795894634744293,
+      "learning_rate": 1.9176131730541206e-05,
+      "loss": 0.7204,
+      "step": 497
+    },
+    {
+      "epoch": 0.7572704808971679,
+      "grad_norm": 0.3739926769832927,
+      "learning_rate": 1.9145954737188265e-05,
+      "loss": 0.7123,
+      "step": 498
+    },
+    {
+      "epoch": 0.7587911043527846,
+      "grad_norm": 0.39711092135289616,
+      "learning_rate": 1.911574424392269e-05,
+      "loss": 0.7532,
+      "step": 499
+    },
+    {
+      "epoch": 0.7603117278084014,
+      "grad_norm": 0.3754721725582843,
+      "learning_rate": 1.9085500431598156e-05,
+      "loss": 0.6882,
+      "step": 500
+    },
+    {
+      "epoch": 0.7618323512640183,
+      "grad_norm": 0.39959794535350796,
+      "learning_rate": 1.9055223481267776e-05,
+      "loss": 0.7479,
+      "step": 501
+    },
+    {
+      "epoch": 0.763352974719635,
+      "grad_norm": 0.35365524692767664,
+      "learning_rate": 1.9024913574183046e-05,
+      "loss": 0.7487,
+      "step": 502
+    },
+    {
+      "epoch": 0.7648735981752518,
+      "grad_norm": 0.4166202806824158,
+      "learning_rate": 1.899457089179276e-05,
+      "loss": 0.7574,
+      "step": 503
+    },
+    {
+      "epoch": 0.7663942216308687,
+      "grad_norm": 0.3915634338240503,
+      "learning_rate": 1.896419561574193e-05,
+      "loss": 0.7325,
+      "step": 504
+    },
+    {
+      "epoch": 0.7679148450864854,
+      "grad_norm": 0.48602884460303647,
+      "learning_rate": 1.8933787927870668e-05,
+      "loss": 0.7286,
+      "step": 505
+    },
+    {
+      "epoch": 0.7694354685421022,
+      "grad_norm": 0.36368946637812705,
+      "learning_rate": 1.890334801021313e-05,
+      "loss": 0.7562,
+      "step": 506
+    },
+    {
+      "epoch": 0.7709560919977191,
+      "grad_norm": 0.45475514908964265,
+      "learning_rate": 1.8872876044996417e-05,
+      "loss": 0.7414,
+      "step": 507
+    },
+    {
+      "epoch": 0.7724767154533358,
+      "grad_norm": 0.4086589491103268,
+      "learning_rate": 1.8842372214639475e-05,
+      "loss": 0.7661,
+      "step": 508
+    },
+    {
+      "epoch": 0.7739973389089527,
+      "grad_norm": 0.4038130105419086,
+      "learning_rate": 1.8811836701752004e-05,
+      "loss": 0.733,
+      "step": 509
+    },
+    {
+      "epoch": 0.7755179623645695,
+      "grad_norm": 0.4234636977161692,
+      "learning_rate": 1.8781269689133386e-05,
+      "loss": 0.8073,
+      "step": 510
+    },
+    {
+      "epoch": 0.7770385858201863,
+      "grad_norm": 0.395363248052582,
+      "learning_rate": 1.8750671359771563e-05,
+      "loss": 0.7298,
+      "step": 511
+    },
+    {
+      "epoch": 0.7785592092758031,
+      "grad_norm": 0.49129541089929685,
+      "learning_rate": 1.8720041896841943e-05,
+      "loss": 0.7118,
+      "step": 512
+    },
+    {
+      "epoch": 0.7800798327314199,
+      "grad_norm": 0.41442318450837484,
+      "learning_rate": 1.868938148370634e-05,
+      "loss": 0.7451,
+      "step": 513
+    },
+    {
+      "epoch": 0.7816004561870367,
+      "grad_norm": 0.4332314377021711,
+      "learning_rate": 1.8658690303911834e-05,
+      "loss": 0.7456,
+      "step": 514
+    },
+    {
+      "epoch": 0.7831210796426535,
+      "grad_norm": 0.49756006836526884,
+      "learning_rate": 1.8627968541189676e-05,
+      "loss": 0.7449,
+      "step": 515
+    },
+    {
+      "epoch": 0.7846417030982703,
+      "grad_norm": 0.39053028376926463,
+      "learning_rate": 1.8597216379454236e-05,
+      "loss": 0.7353,
+      "step": 516
+    },
+    {
+      "epoch": 0.7861623265538871,
+      "grad_norm": 0.42683269902895027,
+      "learning_rate": 1.8566434002801828e-05,
+      "loss": 0.73,
+      "step": 517
+    },
+    {
+      "epoch": 0.787682950009504,
+      "grad_norm": 0.4414935000692991,
+      "learning_rate": 1.8535621595509665e-05,
+      "loss": 0.7222,
+      "step": 518
+    },
+    {
+      "epoch": 0.7892035734651207,
+      "grad_norm": 0.36090118649581926,
+      "learning_rate": 1.8504779342034735e-05,
+      "loss": 0.7206,
+      "step": 519
+    },
+    {
+      "epoch": 0.7907241969207375,
+      "grad_norm": 0.44956102255887004,
+      "learning_rate": 1.8473907427012702e-05,
+      "loss": 0.7809,
+      "step": 520
+    },
+    {
+      "epoch": 0.7922448203763544,
+      "grad_norm": 0.400049613455352,
+      "learning_rate": 1.8443006035256794e-05,
+      "loss": 0.7165,
+      "step": 521
+    },
+    {
+      "epoch": 0.7937654438319711,
+      "grad_norm": 0.4096379673462462,
+      "learning_rate": 1.8412075351756694e-05,
+      "loss": 0.7188,
+      "step": 522
+    },
+    {
+      "epoch": 0.7952860672875879,
+      "grad_norm": 0.3813329658819688,
+      "learning_rate": 1.8381115561677448e-05,
+      "loss": 0.7434,
+      "step": 523
+    },
+    {
+      "epoch": 0.7968066907432048,
+      "grad_norm": 0.3883504678303308,
+      "learning_rate": 1.835012685035834e-05,
+      "loss": 0.7538,
+      "step": 524
+    },
+    {
+      "epoch": 0.7983273141988215,
+      "grad_norm": 0.4810224536969676,
+      "learning_rate": 1.831910940331179e-05,
+      "loss": 0.732,
+      "step": 525
+    },
+    {
+      "epoch": 0.7998479376544383,
+      "grad_norm": 0.4274875955939722,
+      "learning_rate": 1.828806340622226e-05,
+      "loss": 0.7654,
+      "step": 526
+    },
+    {
+      "epoch": 0.8013685611100552,
+      "grad_norm": 0.6167789867453826,
+      "learning_rate": 1.8256989044945094e-05,
+      "loss": 0.723,
+      "step": 527
+    },
+    {
+      "epoch": 0.8028891845656719,
+      "grad_norm": 0.38651635512015386,
+      "learning_rate": 1.8225886505505468e-05,
+      "loss": 0.7301,
+      "step": 528
+    },
+    {
+      "epoch": 0.8044098080212887,
+      "grad_norm": 0.5105848984644641,
+      "learning_rate": 1.8194755974097218e-05,
+      "loss": 0.7379,
+      "step": 529
+    },
+    {
+      "epoch": 0.8059304314769056,
+      "grad_norm": 0.4337444604279403,
+      "learning_rate": 1.8163597637081783e-05,
+      "loss": 0.7438,
+      "step": 530
+    },
+    {
+      "epoch": 0.8074510549325223,
+      "grad_norm": 0.38311547081568653,
+      "learning_rate": 1.8132411680987025e-05,
+      "loss": 0.7166,
+      "step": 531
+    },
+    {
+      "epoch": 0.8089716783881391,
+      "grad_norm": 0.40940875061446746,
+      "learning_rate": 1.8101198292506176e-05,
+      "loss": 0.7377,
+      "step": 532
+    },
+    {
+      "epoch": 0.810492301843756,
+      "grad_norm": 0.39069303091376745,
+      "learning_rate": 1.8069957658496668e-05,
+      "loss": 0.7362,
+      "step": 533
+    },
+    {
+      "epoch": 0.8120129252993727,
+      "grad_norm": 0.4034863997233365,
+      "learning_rate": 1.8038689965979057e-05,
+      "loss": 0.7845,
+      "step": 534
+    },
+    {
+      "epoch": 0.8135335487549895,
+      "grad_norm": 0.3743182233611382,
+      "learning_rate": 1.8007395402135864e-05,
+      "loss": 0.7249,
+      "step": 535
+    },
+    {
+      "epoch": 0.8150541722106064,
+      "grad_norm": 0.4151612507277004,
+      "learning_rate": 1.7976074154310488e-05,
+      "loss": 0.7501,
+      "step": 536
+    },
+    {
+      "epoch": 0.8165747956662232,
+      "grad_norm": 0.4052560608313417,
+      "learning_rate": 1.7944726410006063e-05,
+      "loss": 0.7397,
+      "step": 537
+    },
+    {
+      "epoch": 0.8180954191218399,
+      "grad_norm": 0.4204385753908119,
+      "learning_rate": 1.7913352356884348e-05,
+      "loss": 0.7269,
+      "step": 538
+    },
+    {
+      "epoch": 0.8196160425774568,
+      "grad_norm": 0.36912065852511794,
+      "learning_rate": 1.7881952182764595e-05,
+      "loss": 0.7216,
+      "step": 539
+    },
+    {
+      "epoch": 0.8211366660330736,
+      "grad_norm": 0.4091528689771071,
+      "learning_rate": 1.785052607562243e-05,
+      "loss": 0.741,
+      "step": 540
+    },
+    {
+      "epoch": 0.8226572894886903,
+      "grad_norm": 0.3712296997315177,
+      "learning_rate": 1.7819074223588718e-05,
+      "loss": 0.7361,
+      "step": 541
+    },
+    {
+      "epoch": 0.8241779129443072,
+      "grad_norm": 0.41163387712501454,
+      "learning_rate": 1.7787596814948465e-05,
+      "loss": 0.7153,
+      "step": 542
+    },
+    {
+      "epoch": 0.825698536399924,
+      "grad_norm": 0.3881722683934304,
+      "learning_rate": 1.7756094038139645e-05,
+      "loss": 0.7354,
+      "step": 543
+    },
+    {
+      "epoch": 0.8272191598555407,
+      "grad_norm": 0.37426692455200733,
+      "learning_rate": 1.7724566081752113e-05,
+      "loss": 0.7132,
+      "step": 544
+    },
+    {
+      "epoch": 0.8287397833111576,
+      "grad_norm": 0.36537784686372354,
+      "learning_rate": 1.7693013134526457e-05,
+      "loss": 0.6791,
+      "step": 545
+    },
+    {
+      "epoch": 0.8302604067667744,
+      "grad_norm": 0.3789388855291124,
+      "learning_rate": 1.7661435385352875e-05,
+      "loss": 0.7158,
+      "step": 546
+    },
+    {
+      "epoch": 0.8317810302223911,
+      "grad_norm": 0.41030137404819805,
+      "learning_rate": 1.762983302327003e-05,
+      "loss": 0.6648,
+      "step": 547
+    },
+    {
+      "epoch": 0.833301653678008,
+      "grad_norm": 0.3866957065876247,
+      "learning_rate": 1.7598206237463935e-05,
+      "loss": 0.7383,
+      "step": 548
+    },
+    {
+      "epoch": 0.8348222771336248,
+      "grad_norm": 0.42152702106156975,
+      "learning_rate": 1.7566555217266816e-05,
+      "loss": 0.7486,
+      "step": 549
+    },
+    {
+      "epoch": 0.8363429005892415,
+      "grad_norm": 0.3837180537567676,
+      "learning_rate": 1.753488015215598e-05,
+      "loss": 0.7139,
+      "step": 550
+    },
+    {
+      "epoch": 0.8378635240448584,
+      "grad_norm": 0.45359893466752715,
+      "learning_rate": 1.750318123175267e-05,
+      "loss": 0.7415,
+      "step": 551
+    },
+    {
+      "epoch": 0.8393841475004752,
+      "grad_norm": 0.37167124021677417,
+      "learning_rate": 1.7471458645820944e-05,
+      "loss": 0.7341,
+      "step": 552
+    },
+    {
+      "epoch": 0.840904770956092,
+      "grad_norm": 0.386261837626807,
+      "learning_rate": 1.7439712584266523e-05,
+      "loss": 0.7753,
+      "step": 553
+    },
+    {
+      "epoch": 0.8424253944117088,
+      "grad_norm": 0.3607815357285951,
+      "learning_rate": 1.740794323713568e-05,
+      "loss": 0.7295,
+      "step": 554
+    },
+    {
+      "epoch": 0.8439460178673256,
+      "grad_norm": 0.3805870265080064,
+      "learning_rate": 1.7376150794614067e-05,
+      "loss": 0.742,
+      "step": 555
+    },
+    {
+      "epoch": 0.8454666413229424,
+      "grad_norm": 0.3634857699838318,
+      "learning_rate": 1.7344335447025617e-05,
+      "loss": 0.7405,
+      "step": 556
+    },
+    {
+      "epoch": 0.8469872647785592,
+      "grad_norm": 0.3830792827233582,
+      "learning_rate": 1.731249738483136e-05,
+      "loss": 0.7593,
+      "step": 557
+    },
+    {
+      "epoch": 0.848507888234176,
+      "grad_norm": 0.37680323864124654,
+      "learning_rate": 1.728063679862832e-05,
+      "loss": 0.7374,
+      "step": 558
+    },
+    {
+      "epoch": 0.8500285116897928,
+      "grad_norm": 0.3812378150946191,
+      "learning_rate": 1.724875387914836e-05,
+      "loss": 0.7317,
+      "step": 559
+    },
+    {
+      "epoch": 0.8515491351454096,
+      "grad_norm": 0.3955303602601571,
+      "learning_rate": 1.7216848817257034e-05,
+      "loss": 0.7558,
+      "step": 560
+    },
+    {
+      "epoch": 0.8530697586010264,
+      "grad_norm": 0.36089368326825794,
+      "learning_rate": 1.718492180395245e-05,
+      "loss": 0.7124,
+      "step": 561
+    },
+    {
+      "epoch": 0.8545903820566432,
+      "grad_norm": 0.39264324951073387,
+      "learning_rate": 1.715297303036414e-05,
+      "loss": 0.6976,
+      "step": 562
+    },
+    {
+      "epoch": 0.85611100551226,
+      "grad_norm": 0.3583167479634998,
+      "learning_rate": 1.712100268775188e-05,
+      "loss": 0.7459,
+      "step": 563
+    },
+    {
+      "epoch": 0.8576316289678768,
+      "grad_norm": 0.3652537723370093,
+      "learning_rate": 1.708901096750459e-05,
+      "loss": 0.752,
+      "step": 564
+    },
+    {
+      "epoch": 0.8591522524234936,
+      "grad_norm": 0.35829414089305417,
+      "learning_rate": 1.7056998061139157e-05,
+      "loss": 0.7293,
+      "step": 565
+    },
+    {
+      "epoch": 0.8606728758791105,
+      "grad_norm": 0.41471406367802255,
+      "learning_rate": 1.70249641602993e-05,
+      "loss": 0.7134,
+      "step": 566
+    },
+    {
+      "epoch": 0.8621934993347272,
+      "grad_norm": 0.40497506621492396,
+      "learning_rate": 1.6992909456754408e-05,
+      "loss": 0.7266,
+      "step": 567
+    },
+    {
+      "epoch": 0.8637141227903441,
+      "grad_norm": 0.5118453814476593,
+      "learning_rate": 1.6960834142398424e-05,
+      "loss": 0.7524,
+      "step": 568
+    },
+    {
+      "epoch": 0.8652347462459609,
+      "grad_norm": 0.4027525684201806,
+      "learning_rate": 1.6928738409248668e-05,
+      "loss": 0.7209,
+      "step": 569
+    },
+    {
+      "epoch": 0.8667553697015776,
+      "grad_norm": 0.39420637097855954,
+      "learning_rate": 1.689662244944469e-05,
+      "loss": 0.7445,
+      "step": 570
+    },
+    {
+      "epoch": 0.8682759931571945,
+      "grad_norm": 0.4106148870965668,
+      "learning_rate": 1.686448645524714e-05,
+      "loss": 0.7655,
+      "step": 571
+    },
+    {
+      "epoch": 0.8697966166128113,
+      "grad_norm": 0.3579948523808814,
+      "learning_rate": 1.6832330619036584e-05,
+      "loss": 0.7165,
+      "step": 572
+    },
+    {
+      "epoch": 0.871317240068428,
+      "grad_norm": 0.3751959854555261,
+      "learning_rate": 1.6800155133312387e-05,
+      "loss": 0.7708,
+      "step": 573
+    },
+    {
+      "epoch": 0.8728378635240449,
+      "grad_norm": 0.39336551188405533,
+      "learning_rate": 1.6767960190691538e-05,
+      "loss": 0.7161,
+      "step": 574
+    },
+    {
+      "epoch": 0.8743584869796617,
+      "grad_norm": 0.38785678050667954,
+      "learning_rate": 1.6735745983907504e-05,
+      "loss": 0.719,
+      "step": 575
+    },
+    {
+      "epoch": 0.8758791104352784,
+      "grad_norm": 0.3696861272192392,
+      "learning_rate": 1.6703512705809085e-05,
+      "loss": 0.7434,
+      "step": 576
+    },
+    {
+      "epoch": 0.8773997338908953,
+      "grad_norm": 0.3807058540190109,
+      "learning_rate": 1.667126054935923e-05,
+      "loss": 0.7336,
+      "step": 577
+    },
+    {
+      "epoch": 0.8789203573465121,
+      "grad_norm": 0.3907235180627425,
+      "learning_rate": 1.6638989707633927e-05,
+      "loss": 0.723,
+      "step": 578
+    },
+    {
+      "epoch": 0.8804409808021288,
+      "grad_norm": 0.39300541368827063,
+      "learning_rate": 1.6606700373821005e-05,
+      "loss": 0.7313,
+      "step": 579
+    },
+    {
+      "epoch": 0.8819616042577457,
+      "grad_norm": 0.3994037757594598,
+      "learning_rate": 1.6574392741219003e-05,
+      "loss": 0.7366,
+      "step": 580
+    },
+    {
+      "epoch": 0.8834822277133625,
+      "grad_norm": 0.35949981685128496,
+      "learning_rate": 1.6542067003236e-05,
+      "loss": 0.7288,
+      "step": 581
+    },
+    {
+      "epoch": 0.8850028511689793,
+      "grad_norm": 0.3987872705670089,
+      "learning_rate": 1.6509723353388463e-05,
+      "loss": 0.6909,
+      "step": 582
+    },
+    {
+      "epoch": 0.8865234746245961,
+      "grad_norm": 0.3829862590465747,
+      "learning_rate": 1.6477361985300094e-05,
+      "loss": 0.6989,
+      "step": 583
+    },
+    {
+      "epoch": 0.8880440980802129,
+      "grad_norm": 0.39080192305445455,
+      "learning_rate": 1.6444983092700657e-05,
+      "loss": 0.7357,
+      "step": 584
+    },
+    {
+      "epoch": 0.8895647215358297,
+      "grad_norm": 0.39244835712820425,
+      "learning_rate": 1.6412586869424823e-05,
+      "loss": 0.6906,
+      "step": 585
+    },
+    {
+      "epoch": 0.8910853449914465,
+      "grad_norm": 0.38987649746021225,
+      "learning_rate": 1.6380173509411024e-05,
+      "loss": 0.7655,
+      "step": 586
+    },
+    {
+      "epoch": 0.8926059684470633,
+      "grad_norm": 0.3792483881544404,
+      "learning_rate": 1.634774320670027e-05,
+      "loss": 0.7414,
+      "step": 587
+    },
+    {
+      "epoch": 0.8941265919026801,
+      "grad_norm": 0.37786695217123767,
+      "learning_rate": 1.6315296155435003e-05,
+      "loss": 0.7523,
+      "step": 588
+    },
+    {
+      "epoch": 0.8956472153582969,
+      "grad_norm": 0.40039846555450576,
+      "learning_rate": 1.6282832549857926e-05,
+      "loss": 0.7358,
+      "step": 589
+    },
+    {
+      "epoch": 0.8971678388139137,
+      "grad_norm": 0.3928470737436745,
+      "learning_rate": 1.6250352584310847e-05,
+      "loss": 0.755,
+      "step": 590
+    },
+    {
+      "epoch": 0.8986884622695305,
+      "grad_norm": 0.3903264232496443,
+      "learning_rate": 1.6217856453233505e-05,
+      "loss": 0.7334,
+      "step": 591
+    },
+    {
+      "epoch": 0.9002090857251474,
+      "grad_norm": 0.3685085038233517,
+      "learning_rate": 1.6185344351162427e-05,
+      "loss": 0.7137,
+      "step": 592
+    },
+    {
+      "epoch": 0.9017297091807641,
+      "grad_norm": 0.39206546198749437,
+      "learning_rate": 1.6152816472729734e-05,
+      "loss": 0.7451,
+      "step": 593
+    },
+    {
+      "epoch": 0.9032503326363809,
+      "grad_norm": 0.379717976951851,
+      "learning_rate": 1.6120273012662e-05,
+      "loss": 0.7153,
+      "step": 594
+    },
+    {
+      "epoch": 0.9047709560919978,
+      "grad_norm": 0.38621564008551273,
+      "learning_rate": 1.608771416577908e-05,
+      "loss": 0.7559,
+      "step": 595
+    },
+    {
+      "epoch": 0.9062915795476145,
+      "grad_norm": 0.4060366023992974,
+      "learning_rate": 1.6055140126992933e-05,
+      "loss": 0.7386,
+      "step": 596
+    },
+    {
+      "epoch": 0.9078122030032313,
+      "grad_norm": 0.35024041522069665,
+      "learning_rate": 1.6022551091306466e-05,
+      "loss": 0.7248,
+      "step": 597
+    },
+    {
+      "epoch": 0.9093328264588482,
+      "grad_norm": 0.4429630612789365,
+      "learning_rate": 1.5989947253812372e-05,
+      "loss": 0.7344,
+      "step": 598
+    },
+    {
+      "epoch": 0.9108534499144649,
+      "grad_norm": 0.35160112786541303,
+      "learning_rate": 1.5957328809691947e-05,
+      "loss": 0.7136,
+      "step": 599
+    },
+    {
+      "epoch": 0.9123740733700817,
+      "grad_norm": 0.3607594118044688,
+      "learning_rate": 1.592469595421393e-05,
+      "loss": 0.7152,
+      "step": 600
+    },
+    {
+      "epoch": 0.9138946968256986,
+      "grad_norm": 0.3888604747890697,
+      "learning_rate": 1.589204888273333e-05,
+      "loss": 0.7214,
+      "step": 601
+    },
+    {
+      "epoch": 0.9154153202813153,
+      "grad_norm": 0.36939791749863715,
+      "learning_rate": 1.585938779069027e-05,
+      "loss": 0.724,
+      "step": 602
+    },
+    {
+      "epoch": 0.9169359437369321,
+      "grad_norm": 0.3492211353535467,
+      "learning_rate": 1.5826712873608794e-05,
+      "loss": 0.7059,
+      "step": 603
+    },
+    {
+      "epoch": 0.918456567192549,
+      "grad_norm": 0.3485676743224654,
+      "learning_rate": 1.5794024327095722e-05,
+      "loss": 0.6882,
+      "step": 604
+    },
+    {
+      "epoch": 0.9199771906481657,
+      "grad_norm": 0.34637043053576727,
+      "learning_rate": 1.5761322346839444e-05,
+      "loss": 0.7329,
+      "step": 605
+    },
+    {
+      "epoch": 0.9214978141037825,
+      "grad_norm": 0.39153636395112856,
+      "learning_rate": 1.5728607128608797e-05,
+      "loss": 0.7039,
+      "step": 606
+    },
+    {
+      "epoch": 0.9230184375593994,
+      "grad_norm": 0.36173214969406103,
+      "learning_rate": 1.5695878868251847e-05,
+      "loss": 0.7238,
+      "step": 607
+    },
+    {
+      "epoch": 0.9245390610150162,
+      "grad_norm": 0.3761817399411723,
+      "learning_rate": 1.5663137761694748e-05,
+      "loss": 0.7287,
+      "step": 608
+    },
+    {
+      "epoch": 0.9260596844706329,
+      "grad_norm": 0.35432511798173993,
+      "learning_rate": 1.5630384004940545e-05,
+      "loss": 0.7169,
+      "step": 609
+    },
+    {
+      "epoch": 0.9275803079262498,
+      "grad_norm": 0.37885488266265444,
+      "learning_rate": 1.559761779406803e-05,
+      "loss": 0.7456,
+      "step": 610
+    },
+    {
+      "epoch": 0.9291009313818666,
+      "grad_norm": 0.3834340488104107,
+      "learning_rate": 1.5564839325230532e-05,
+      "loss": 0.7145,
+      "step": 611
+    },
+    {
+      "epoch": 0.9306215548374833,
+      "grad_norm": 0.36202133079807264,
+      "learning_rate": 1.5532048794654782e-05,
+      "loss": 0.7408,
+      "step": 612
+    },
+    {
+      "epoch": 0.9321421782931002,
+      "grad_norm": 0.39335695813338395,
+      "learning_rate": 1.549924639863969e-05,
+      "loss": 0.7403,
+      "step": 613
+    },
+    {
+      "epoch": 0.933662801748717,
+      "grad_norm": 0.3933661187801479,
+      "learning_rate": 1.546643233355523e-05,
+      "loss": 0.7082,
+      "step": 614
+    },
+    {
+      "epoch": 0.9351834252043337,
+      "grad_norm": 0.37654144207241497,
+      "learning_rate": 1.5433606795841207e-05,
+      "loss": 0.7435,
+      "step": 615
+    },
+    {
+      "epoch": 0.9367040486599506,
+      "grad_norm": 0.44483184731571473,
+      "learning_rate": 1.5400769982006128e-05,
+      "loss": 0.7156,
+      "step": 616
+    },
+    {
+      "epoch": 0.9382246721155674,
+      "grad_norm": 0.42682697285522647,
+      "learning_rate": 1.536792208862598e-05,
+      "loss": 0.696,
+      "step": 617
+    },
+    {
+      "epoch": 0.9397452955711841,
+      "grad_norm": 0.40979396640326593,
+      "learning_rate": 1.5335063312343095e-05,
+      "loss": 0.727,
+      "step": 618
+    },
+    {
+      "epoch": 0.941265919026801,
+      "grad_norm": 0.3960793962289382,
+      "learning_rate": 1.530219384986495e-05,
+      "loss": 0.7032,
+      "step": 619
+    },
+    {
+      "epoch": 0.9427865424824178,
+      "grad_norm": 0.3857008168238225,
+      "learning_rate": 1.5269313897962994e-05,
+      "loss": 0.7502,
+      "step": 620
+    },
+    {
+      "epoch": 0.9443071659380345,
+      "grad_norm": 0.3982600098061033,
+      "learning_rate": 1.5236423653471463e-05,
+      "loss": 0.7666,
+      "step": 621
+    },
+    {
+      "epoch": 0.9458277893936514,
+      "grad_norm": 0.41685987364222565,
+      "learning_rate": 1.5203523313286227e-05,
+      "loss": 0.7036,
+      "step": 622
+    },
+    {
+      "epoch": 0.9473484128492682,
+      "grad_norm": 0.3959161081756824,
+      "learning_rate": 1.5170613074363573e-05,
+      "loss": 0.7311,
+      "step": 623
+    },
+    {
+      "epoch": 0.948869036304885,
+      "grad_norm": 0.36920330872863855,
+      "learning_rate": 1.5137693133719065e-05,
+      "loss": 0.7039,
+      "step": 624
+    },
+    {
+      "epoch": 0.9503896597605018,
+      "grad_norm": 0.39113501294384384,
+      "learning_rate": 1.5104763688426331e-05,
+      "loss": 0.7506,
+      "step": 625
+    },
+    {
+      "epoch": 0.9519102832161186,
+      "grad_norm": 0.39144353205492227,
+      "learning_rate": 1.5071824935615908e-05,
+      "loss": 0.6886,
+      "step": 626
+    },
+    {
+      "epoch": 0.9534309066717354,
+      "grad_norm": 0.36594301273514485,
+      "learning_rate": 1.5038877072474041e-05,
+      "loss": 0.7444,
+      "step": 627
+    },
+    {
+      "epoch": 0.9549515301273522,
+      "grad_norm": 0.3954444253170849,
+      "learning_rate": 1.500592029624153e-05,
+      "loss": 0.7147,
+      "step": 628
+    },
+    {
+      "epoch": 0.956472153582969,
+      "grad_norm": 0.3773327611421358,
+      "learning_rate": 1.4972954804212516e-05,
+      "loss": 0.7453,
+      "step": 629
+    },
+    {
+      "epoch": 0.9579927770385859,
+      "grad_norm": 0.41652271596518264,
+      "learning_rate": 1.4939980793733329e-05,
+      "loss": 0.7437,
+      "step": 630
+    },
+    {
+      "epoch": 0.9595134004942026,
+      "grad_norm": 0.4094773012771971,
+      "learning_rate": 1.4906998462201288e-05,
+      "loss": 0.7654,
+      "step": 631
+    },
+    {
+      "epoch": 0.9610340239498194,
+      "grad_norm": 0.3944195978868427,
+      "learning_rate": 1.4874008007063523e-05,
+      "loss": 0.7452,
+      "step": 632
+    },
+    {
+      "epoch": 0.9625546474054363,
+      "grad_norm": 0.3707638628881551,
+      "learning_rate": 1.48410096258158e-05,
+      "loss": 0.7042,
+      "step": 633
+    },
+    {
+      "epoch": 0.964075270861053,
+      "grad_norm": 0.3755039123645331,
+      "learning_rate": 1.4808003516001344e-05,
+      "loss": 0.7177,
+      "step": 634
+    },
+    {
+      "epoch": 0.9655958943166698,
+      "grad_norm": 0.3723560524603749,
+      "learning_rate": 1.4774989875209624e-05,
+      "loss": 0.7553,
+      "step": 635
+    },
+    {
+      "epoch": 0.9671165177722867,
+      "grad_norm": 0.3651623249941728,
+      "learning_rate": 1.474196890107521e-05,
+      "loss": 0.7404,
+      "step": 636
+    },
+    {
+      "epoch": 0.9686371412279035,
+      "grad_norm": 0.37966360420328005,
+      "learning_rate": 1.4708940791276568e-05,
+      "loss": 0.7159,
+      "step": 637
+    },
+    {
+      "epoch": 0.9701577646835202,
+      "grad_norm": 0.36168885071485135,
+      "learning_rate": 1.4675905743534884e-05,
+      "loss": 0.7541,
+      "step": 638
+    },
+    {
+      "epoch": 0.9716783881391371,
+      "grad_norm": 0.36095700915855466,
+      "learning_rate": 1.464286395561287e-05,
+      "loss": 0.7092,
+      "step": 639
+    },
+    {
+      "epoch": 0.9731990115947539,
+      "grad_norm": 0.43805492394151224,
+      "learning_rate": 1.46098156253136e-05,
+      "loss": 0.7362,
+      "step": 640
+    },
+    {
+      "epoch": 0.9747196350503706,
+      "grad_norm": 0.36203353195893573,
+      "learning_rate": 1.4576760950479295e-05,
+      "loss": 0.7106,
+      "step": 641
+    },
+    {
+      "epoch": 0.9762402585059875,
+      "grad_norm": 0.40877067220155616,
+      "learning_rate": 1.454370012899018e-05,
+      "loss": 0.7026,
+      "step": 642
+    },
+    {
+      "epoch": 0.9777608819616043,
+      "grad_norm": 0.40374065798245223,
+      "learning_rate": 1.4510633358763258e-05,
+      "loss": 0.7182,
+      "step": 643
+    },
+    {
+      "epoch": 0.979281505417221,
+      "grad_norm": 0.3659907217224348,
+      "learning_rate": 1.4477560837751156e-05,
+      "loss": 0.713,
+      "step": 644
+    },
+    {
+      "epoch": 0.9808021288728379,
+      "grad_norm": 0.3622601843846982,
+      "learning_rate": 1.4444482763940919e-05,
+      "loss": 0.7327,
+      "step": 645
+    },
+    {
+      "epoch": 0.9823227523284547,
+      "grad_norm": 0.3885583144262136,
+      "learning_rate": 1.441139933535284e-05,
+      "loss": 0.6978,
+      "step": 646
+    },
+    {
+      "epoch": 0.9838433757840714,
+      "grad_norm": 0.37410385831599374,
+      "learning_rate": 1.437831075003926e-05,
+      "loss": 0.724,
+      "step": 647
+    },
+    {
+      "epoch": 0.9853639992396883,
+      "grad_norm": 0.3696274971269831,
+      "learning_rate": 1.4345217206083407e-05,
+      "loss": 0.6965,
+      "step": 648
+    },
+    {
+      "epoch": 0.9868846226953051,
+      "grad_norm": 0.41889852029503616,
+      "learning_rate": 1.4312118901598172e-05,
+      "loss": 0.7678,
+      "step": 649
+    },
+    {
+      "epoch": 0.9884052461509218,
+      "grad_norm": 0.3429687194640586,
+      "learning_rate": 1.4279016034724961e-05,
+      "loss": 0.7047,
+      "step": 650
+    },
+    {
+      "epoch": 0.9899258696065387,
+      "grad_norm": 0.3736752188164251,
+      "learning_rate": 1.4245908803632486e-05,
+      "loss": 0.7362,
+      "step": 651
+    },
+    {
+      "epoch": 0.9914464930621555,
+      "grad_norm": 0.38265917497289204,
+      "learning_rate": 1.4212797406515585e-05,
+      "loss": 0.7225,
+      "step": 652
+    },
+    {
+      "epoch": 0.9929671165177723,
+      "grad_norm": 0.37332104277853684,
+      "learning_rate": 1.417968204159404e-05,
+      "loss": 0.7312,
+      "step": 653
+    },
+    {
+      "epoch": 0.9944877399733891,
+      "grad_norm": 0.42232449088131074,
+      "learning_rate": 1.4146562907111377e-05,
+      "loss": 0.7095,
+      "step": 654
+    },
+    {
+      "epoch": 0.9960083634290059,
+      "grad_norm": 0.3868095376144605,
+      "learning_rate": 1.4113440201333701e-05,
+      "loss": 0.6992,
+      "step": 655
+    },
+    {
+      "epoch": 0.9975289868846227,
+      "grad_norm": 0.42385907348580026,
+      "learning_rate": 1.4080314122548488e-05,
+      "loss": 0.7347,
+      "step": 656
+    },
+    {
+      "epoch": 0.9990496103402395,
+      "grad_norm": 0.41348925205041664,
+      "learning_rate": 1.4047184869063401e-05,
+      "loss": 0.7441,
+      "step": 657
+    },
+    {
+      "epoch": 1.0005702337958564,
+      "grad_norm": 0.35648966659894604,
+      "learning_rate": 1.4014052639205127e-05,
+      "loss": 0.6707,
+      "step": 658
+    },
+    {
+      "epoch": 1.002090857251473,
+      "grad_norm": 0.34244915908174534,
+      "learning_rate": 1.3980917631318145e-05,
+      "loss": 0.7027,
+      "step": 659
+    },
+    {
+      "epoch": 1.00361148070709,
+      "grad_norm": 0.3670346028136332,
+      "learning_rate": 1.394778004376359e-05,
+      "loss": 0.7363,
+      "step": 660
+    },
+    {
+      "epoch": 1.0051321041627068,
+      "grad_norm": 0.391603629750881,
+      "learning_rate": 1.3914640074918023e-05,
+      "loss": 0.7351,
+      "step": 661
+    },
+    {
+      "epoch": 1.0066527276183235,
+      "grad_norm": 0.3716653098931945,
+      "learning_rate": 1.3881497923172269e-05,
+      "loss": 0.7441,
+      "step": 662
+    },
+    {
+      "epoch": 1.0081733510739403,
+      "grad_norm": 0.3755372180582686,
+      "learning_rate": 1.3848353786930213e-05,
+      "loss": 0.6965,
+      "step": 663
+    },
+    {
+      "epoch": 1.0096939745295572,
+      "grad_norm": 0.381844726209517,
+      "learning_rate": 1.3815207864607634e-05,
+      "loss": 0.7302,
+      "step": 664
+    },
+    {
+      "epoch": 1.0112145979851739,
+      "grad_norm": 0.36415027113970666,
+      "learning_rate": 1.3782060354630988e-05,
+      "loss": 0.6935,
+      "step": 665
+    },
+    {
+      "epoch": 1.0127352214407908,
+      "grad_norm": 0.3814827703352279,
+      "learning_rate": 1.3748911455436241e-05,
+      "loss": 0.7288,
+      "step": 666
+    },
+    {
+      "epoch": 1.0142558448964076,
+      "grad_norm": 0.3695679755631074,
+      "learning_rate": 1.371576136546768e-05,
+      "loss": 0.7069,
+      "step": 667
+    },
+    {
+      "epoch": 1.0157764683520243,
+      "grad_norm": 0.3692625124863184,
+      "learning_rate": 1.3682610283176713e-05,
+      "loss": 0.7308,
+      "step": 668
+    },
+    {
+      "epoch": 1.0172970918076412,
+      "grad_norm": 0.3647179500902266,
+      "learning_rate": 1.3649458407020692e-05,
+      "loss": 0.724,
+      "step": 669
+    },
+    {
+      "epoch": 1.0003801558639043,
+      "grad_norm": 0.39033436102991464,
+      "learning_rate": 1.3616305935461726e-05,
+      "loss": 0.6831,
+      "step": 670
+    },
+    {
+      "epoch": 1.001900779319521,
+      "grad_norm": 0.6228269584548538,
+      "learning_rate": 1.358315306696548e-05,
+      "loss": 0.5697,
+      "step": 671
+    },
+    {
+      "epoch": 1.0034214027751378,
+      "grad_norm": 0.4482616727348143,
+      "learning_rate": 1.355e-05,
+      "loss": 0.5192,
+      "step": 672
+    },
+    {
+      "epoch": 1.0049420262307547,
+      "grad_norm": 0.5196584857248853,
+      "learning_rate": 1.3516846933034525e-05,
+      "loss": 0.5425,
+      "step": 673
+    },
+    {
+      "epoch": 1.0064626496863713,
+      "grad_norm": 0.5838517226989333,
+      "learning_rate": 1.3483694064538279e-05,
+      "loss": 0.5638,
+      "step": 674
+    },
+    {
+      "epoch": 1.0079832731419882,
+      "grad_norm": 0.4333646030218863,
+      "learning_rate": 1.3450541592979308e-05,
+      "loss": 0.5361,
+      "step": 675
+    },
+    {
+      "epoch": 1.009503896597605,
+      "grad_norm": 0.46529778581028425,
+      "learning_rate": 1.341738971682329e-05,
+      "loss": 0.5338,
+      "step": 676
+    },
+    {
+      "epoch": 1.0110245200532217,
+      "grad_norm": 0.50417271156346,
+      "learning_rate": 1.3384238634532324e-05,
+      "loss": 0.5341,
+      "step": 677
+    },
+    {
+      "epoch": 1.0125451435088386,
+      "grad_norm": 0.4846246525284431,
+      "learning_rate": 1.3351088544563762e-05,
+      "loss": 0.5454,
+      "step": 678
+    },
+    {
+      "epoch": 1.0140657669644555,
+      "grad_norm": 0.43151729793082855,
+      "learning_rate": 1.3317939645369014e-05,
+      "loss": 0.5149,
+      "step": 679
+    },
+    {
+      "epoch": 1.0155863904200721,
+      "grad_norm": 0.45346621189906927,
+      "learning_rate": 1.3284792135392369e-05,
+      "loss": 0.5241,
+      "step": 680
+    },
+    {
+      "epoch": 1.017107013875689,
+      "grad_norm": 0.4529580634485115,
+      "learning_rate": 1.3251646213069789e-05,
+      "loss": 0.5252,
+      "step": 681
+    },
+    {
+      "epoch": 1.018627637331306,
+      "grad_norm": 0.4309687856541669,
+      "learning_rate": 1.3218502076827738e-05,
+      "loss": 0.5041,
+      "step": 682
+    },
+    {
+      "epoch": 1.0201482607869226,
+      "grad_norm": 0.40932019500719957,
+      "learning_rate": 1.3185359925081977e-05,
+      "loss": 0.5131,
+      "step": 683
+    },
+    {
+      "epoch": 1.0216688842425394,
+      "grad_norm": 0.43569657302929155,
+      "learning_rate": 1.3152219956236414e-05,
+      "loss": 0.5058,
+      "step": 684
+    },
+    {
+      "epoch": 1.0231895076981563,
+      "grad_norm": 0.41224009605600054,
+      "learning_rate": 1.3119082368681858e-05,
+      "loss": 0.5102,
+      "step": 685
+    },
+    {
+      "epoch": 1.024710131153773,
+      "grad_norm": 0.4633671152656241,
+      "learning_rate": 1.3085947360794881e-05,
+      "loss": 0.5538,
+      "step": 686
+    },
+    {
+      "epoch": 1.0262307546093898,
+      "grad_norm": 0.3981390104088507,
+      "learning_rate": 1.3052815130936598e-05,
+      "loss": 0.5053,
+      "step": 687
+    },
+    {
+      "epoch": 1.0277513780650067,
+      "grad_norm": 0.42462762773917034,
+      "learning_rate": 1.3019685877451516e-05,
+      "loss": 0.5308,
+      "step": 688
+    },
+    {
+      "epoch": 1.0292720015206234,
+      "grad_norm": 0.3928387548640261,
+      "learning_rate": 1.29865597986663e-05,
+      "loss": 0.4768,
+      "step": 689
+    },
+    {
+      "epoch": 1.0307926249762402,
+      "grad_norm": 0.4320276733077232,
+      "learning_rate": 1.2953437092888625e-05,
+      "loss": 0.5221,
+      "step": 690
+    },
+    {
+      "epoch": 1.0323132484318571,
+      "grad_norm": 0.4334542028887193,
+      "learning_rate": 1.2920317958405962e-05,
+      "loss": 0.5129,
+      "step": 691
+    },
+    {
+      "epoch": 1.0338338718874738,
+      "grad_norm": 0.40146788598973115,
+      "learning_rate": 1.2887202593484417e-05,
+      "loss": 0.5052,
+      "step": 692
+    },
+    {
+      "epoch": 1.0353544953430907,
+      "grad_norm": 0.41287405977658365,
+      "learning_rate": 1.2854091196367518e-05,
+      "loss": 0.5425,
+      "step": 693
+    },
+    {
+      "epoch": 1.0368751187987075,
+      "grad_norm": 0.3708390977190173,
+      "learning_rate": 1.2820983965275043e-05,
+      "loss": 0.5238,
+      "step": 694
+    },
+    {
+      "epoch": 1.0383957422543242,
+      "grad_norm": 0.35512557723314947,
+      "learning_rate": 1.278788109840183e-05,
+      "loss": 0.5096,
+      "step": 695
+    },
+    {
+      "epoch": 1.039916365709941,
+      "grad_norm": 0.41469287495730844,
+      "learning_rate": 1.2754782793916598e-05,
+      "loss": 0.5159,
+      "step": 696
+    },
+    {
+      "epoch": 1.041436989165558,
+      "grad_norm": 0.39113759999764336,
+      "learning_rate": 1.2721689249960743e-05,
+      "loss": 0.5184,
+      "step": 697
+    },
+    {
+      "epoch": 1.0429576126211746,
+      "grad_norm": 0.41576137745186253,
+      "learning_rate": 1.2688600664647168e-05,
+      "loss": 0.5251,
+      "step": 698
+    },
+    {
+      "epoch": 1.0444782360767915,
+      "grad_norm": 0.37659361938427355,
+      "learning_rate": 1.2655517236059083e-05,
+      "loss": 0.511,
+      "step": 699
+    },
+    {
+      "epoch": 1.0459988595324083,
+      "grad_norm": 0.36191609676818887,
+      "learning_rate": 1.2622439162248846e-05,
+      "loss": 0.5142,
+      "step": 700
+    },
+    {
+      "epoch": 1.047519482988025,
+      "grad_norm": 0.3501769260644456,
+      "learning_rate": 1.2589366641236745e-05,
+      "loss": 0.5086,
+      "step": 701
+    },
+    {
+      "epoch": 1.0490401064436419,
+      "grad_norm": 0.374193201836417,
+      "learning_rate": 1.2556299871009825e-05,
+      "loss": 0.5251,
+      "step": 702
+    },
+    {
+      "epoch": 1.0505607298992587,
+      "grad_norm": 0.40667780389565045,
+      "learning_rate": 1.2523239049520706e-05,
+      "loss": 0.5133,
+      "step": 703
+    },
+    {
+      "epoch": 1.0520813533548754,
+      "grad_norm": 0.3484682044458917,
+      "learning_rate": 1.2490184374686406e-05,
+      "loss": 0.5078,
+      "step": 704
+    },
+    {
+      "epoch": 1.0536019768104923,
+      "grad_norm": 0.3406587720902772,
+      "learning_rate": 1.2457136044387133e-05,
+      "loss": 0.5107,
+      "step": 705
+    },
+    {
+      "epoch": 1.0551226002661092,
+      "grad_norm": 0.35983619595762045,
+      "learning_rate": 1.242409425646512e-05,
+      "loss": 0.5349,
+      "step": 706
+    },
+    {
+      "epoch": 1.0566432237217258,
+      "grad_norm": 0.3722826497059545,
+      "learning_rate": 1.2391059208723433e-05,
+      "loss": 0.5118,
+      "step": 707
+    },
+    {
+      "epoch": 1.0581638471773427,
+      "grad_norm": 0.367794414326388,
+      "learning_rate": 1.2358031098924792e-05,
+      "loss": 0.5012,
+      "step": 708
+    },
+    {
+      "epoch": 1.0596844706329596,
+      "grad_norm": 0.379706737726683,
+      "learning_rate": 1.232501012479038e-05,
+      "loss": 0.5234,
+      "step": 709
+    },
+    {
+      "epoch": 1.0612050940885762,
+      "grad_norm": 0.3808569060578708,
+      "learning_rate": 1.2291996483998664e-05,
+      "loss": 0.4976,
+      "step": 710
+    },
+    {
+      "epoch": 1.062725717544193,
+      "grad_norm": 0.4052638293016256,
+      "learning_rate": 1.2258990374184197e-05,
+      "loss": 0.5115,
+      "step": 711
+    },
+    {
+      "epoch": 1.06424634099981,
+      "grad_norm": 0.3884487585023922,
+      "learning_rate": 1.222599199293648e-05,
+      "loss": 0.5469,
+      "step": 712
+    },
+    {
+      "epoch": 1.0657669644554266,
+      "grad_norm": 0.3523224705049633,
+      "learning_rate": 1.2193001537798717e-05,
+      "loss": 0.5031,
+      "step": 713
+    },
+    {
+      "epoch": 1.0672875879110435,
+      "grad_norm": 0.4035860752243172,
+      "learning_rate": 1.2160019206266676e-05,
+      "loss": 0.5036,
+      "step": 714
+    },
+    {
+      "epoch": 1.0688082113666604,
+      "grad_norm": 0.35467473723128,
+      "learning_rate": 1.2127045195787486e-05,
+      "loss": 0.5182,
+      "step": 715
+    },
+    {
+      "epoch": 1.070328834822277,
+      "grad_norm": 0.41329760884730493,
+      "learning_rate": 1.2094079703758474e-05,
+      "loss": 0.5212,
+      "step": 716
+    },
+    {
+      "epoch": 1.071849458277894,
+      "grad_norm": 0.36806047789304813,
+      "learning_rate": 1.2061122927525962e-05,
+      "loss": 0.5148,
+      "step": 717
+    },
+    {
+      "epoch": 1.0733700817335108,
+      "grad_norm": 0.4473641673792849,
+      "learning_rate": 1.20281750643841e-05,
+      "loss": 0.5217,
+      "step": 718
+    },
+    {
+      "epoch": 1.0748907051891274,
+      "grad_norm": 0.35449349576414274,
+      "learning_rate": 1.1995236311573668e-05,
+      "loss": 0.5057,
+      "step": 719
+    },
+    {
+      "epoch": 1.0764113286447443,
+      "grad_norm": 0.37802608508785335,
+      "learning_rate": 1.1962306866280938e-05,
+      "loss": 0.5127,
+      "step": 720
+    },
+    {
+      "epoch": 1.0779319521003612,
+      "grad_norm": 0.38606300562371665,
+      "learning_rate": 1.192938692563643e-05,
+      "loss": 0.5225,
+      "step": 721
+    },
+    {
+      "epoch": 1.0794525755559778,
+      "grad_norm": 0.3548827181233496,
+      "learning_rate": 1.1896476686713781e-05,
+      "loss": 0.5097,
+      "step": 722
+    },
+    {
+      "epoch": 1.0809731990115947,
+      "grad_norm": 0.38704197291443254,
+      "learning_rate": 1.1863576346528536e-05,
+      "loss": 0.5147,
+      "step": 723
+    },
+    {
+      "epoch": 1.0824938224672116,
+      "grad_norm": 0.3788637749591371,
+      "learning_rate": 1.1830686102037011e-05,
+      "loss": 0.526,
+      "step": 724
+    },
+    {
+      "epoch": 1.0840144459228283,
+      "grad_norm": 0.36699155176089776,
+      "learning_rate": 1.1797806150135052e-05,
+      "loss": 0.5074,
+      "step": 725
+    },
+    {
+      "epoch": 1.0855350693784451,
+      "grad_norm": 0.398207632535489,
+      "learning_rate": 1.1764936687656908e-05,
+      "loss": 0.5263,
+      "step": 726
+    },
+    {
+      "epoch": 1.087055692834062,
+      "grad_norm": 0.34386679094163725,
+      "learning_rate": 1.1732077911374022e-05,
+      "loss": 0.5292,
+      "step": 727
+    },
+    {
+      "epoch": 1.0885763162896787,
+      "grad_norm": 0.3898697838147865,
+      "learning_rate": 1.1699230017993875e-05,
+      "loss": 0.5157,
+      "step": 728
+    },
+    {
+      "epoch": 1.0900969397452955,
+      "grad_norm": 0.35036102017636284,
+      "learning_rate": 1.1666393204158793e-05,
+      "loss": 0.5103,
+      "step": 729
+    },
+    {
+      "epoch": 1.0916175632009124,
+      "grad_norm": 0.3766316653297432,
+      "learning_rate": 1.1633567666444775e-05,
+      "loss": 0.5355,
+      "step": 730
+    },
+    {
+      "epoch": 1.093138186656529,
+      "grad_norm": 0.3395395584528322,
+      "learning_rate": 1.1600753601360308e-05,
+      "loss": 0.4922,
+      "step": 731
+    },
+    {
+      "epoch": 1.094658810112146,
+      "grad_norm": 0.36396997728904185,
+      "learning_rate": 1.1567951205345224e-05,
+      "loss": 0.5219,
+      "step": 732
+    },
+    {
+      "epoch": 1.0961794335677628,
+      "grad_norm": 0.3565103395663914,
+      "learning_rate": 1.153516067476947e-05,
+      "loss": 0.5089,
+      "step": 733
+    },
+    {
+      "epoch": 1.0977000570233795,
+      "grad_norm": 0.3671854484970251,
+      "learning_rate": 1.1502382205931974e-05,
+      "loss": 0.5259,
+      "step": 734
+    },
+    {
+      "epoch": 1.0992206804789963,
+      "grad_norm": 0.35699046104294285,
+      "learning_rate": 1.1469615995059455e-05,
+      "loss": 0.5199,
+      "step": 735
+    },
+    {
+      "epoch": 1.1007413039346132,
+      "grad_norm": 0.35814749552055475,
+      "learning_rate": 1.1436862238305256e-05,
+      "loss": 0.4951,
+      "step": 736
+    },
+    {
+      "epoch": 1.1022619273902299,
+      "grad_norm": 0.37766411014866846,
+      "learning_rate": 1.1404121131748156e-05,
+      "loss": 0.5133,
+      "step": 737
+    },
+    {
+      "epoch": 1.1037825508458468,
+      "grad_norm": 0.38876585742040404,
+      "learning_rate": 1.1371392871391207e-05,
+      "loss": 0.5358,
+      "step": 738
+    },
+    {
+      "epoch": 1.1053031743014636,
+      "grad_norm": 0.4211502488863111,
+      "learning_rate": 1.1338677653160557e-05,
+      "loss": 0.5083,
+      "step": 739
+    },
+    {
+      "epoch": 1.1068237977570803,
+      "grad_norm": 0.3864747484032165,
+      "learning_rate": 1.1305975672904284e-05,
+      "loss": 0.5258,
+      "step": 740
+    },
+    {
+      "epoch": 1.1083444212126972,
+      "grad_norm": 0.39083288027419083,
+      "learning_rate": 1.1273287126391206e-05,
+      "loss": 0.4843,
+      "step": 741
+    },
+    {
+      "epoch": 1.109865044668314,
+      "grad_norm": 0.3910337272002525,
+      "learning_rate": 1.1240612209309734e-05,
+      "loss": 0.5479,
+      "step": 742
+    },
+    {
+      "epoch": 1.111385668123931,
+      "grad_norm": 0.37980182130941675,
+      "learning_rate": 1.1207951117266668e-05,
+      "loss": 0.4885,
+      "step": 743
+    },
+    {
+      "epoch": 1.1129062915795476,
+      "grad_norm": 0.3697135253054006,
+      "learning_rate": 1.1175304045786073e-05,
+      "loss": 0.523,
+      "step": 744
+    },
+    {
+      "epoch": 1.1144269150351644,
+      "grad_norm": 0.3772904905798372,
+      "learning_rate": 1.1142671190308056e-05,
+      "loss": 0.5237,
+      "step": 745
+    },
+    {
+      "epoch": 1.1159475384907813,
+      "grad_norm": 0.4056017049635124,
+      "learning_rate": 1.1110052746187631e-05,
+      "loss": 0.5137,
+      "step": 746
+    },
+    {
+      "epoch": 1.117468161946398,
+      "grad_norm": 0.35112819485311203,
+      "learning_rate": 1.1077448908693534e-05,
+      "loss": 0.5038,
+      "step": 747
+    },
+    {
+      "epoch": 1.1189887854020149,
+      "grad_norm": 0.38291495750095184,
+      "learning_rate": 1.104485987300707e-05,
+      "loss": 0.5084,
+      "step": 748
+    },
+    {
+      "epoch": 1.1205094088576317,
+      "grad_norm": 0.3922969549725083,
+      "learning_rate": 1.1012285834220924e-05,
+      "loss": 0.5092,
+      "step": 749
+    },
+    {
+      "epoch": 1.1220300323132484,
+      "grad_norm": 0.40549020760259696,
+      "learning_rate": 1.0979726987338002e-05,
+      "loss": 0.5204,
+      "step": 750
+    },
+    {
+      "epoch": 1.1235506557688653,
+      "grad_norm": 0.40395038468147576,
+      "learning_rate": 1.0947183527270267e-05,
+      "loss": 0.5227,
+      "step": 751
+    },
+    {
+      "epoch": 1.1250712792244821,
+      "grad_norm": 0.41236903215126236,
+      "learning_rate": 1.0914655648837576e-05,
+      "loss": 0.5319,
+      "step": 752
+    },
+    {
+      "epoch": 1.1265919026800988,
+      "grad_norm": 0.3677643398357805,
+      "learning_rate": 1.0882143546766496e-05,
+      "loss": 0.5038,
+      "step": 753
+    },
+    {
+      "epoch": 1.1281125261357157,
+      "grad_norm": 0.3828445757904712,
+      "learning_rate": 1.084964741568916e-05,
+      "loss": 0.5194,
+      "step": 754
+    },
+    {
+      "epoch": 1.1296331495913325,
+      "grad_norm": 0.39209056252387725,
+      "learning_rate": 1.0817167450142074e-05,
+      "loss": 0.5013,
+      "step": 755
+    },
+    {
+      "epoch": 1.1311537730469492,
+      "grad_norm": 0.3680446404406516,
+      "learning_rate": 1.0784703844564998e-05,
+      "loss": 0.5231,
+      "step": 756
+    },
+    {
+      "epoch": 1.132674396502566,
+      "grad_norm": 0.3703842374744353,
+      "learning_rate": 1.0752256793299732e-05,
+      "loss": 0.5153,
+      "step": 757
+    },
+    {
+      "epoch": 1.134195019958183,
+      "grad_norm": 0.37037413568566663,
+      "learning_rate": 1.071982649058898e-05,
+      "loss": 0.5124,
+      "step": 758
+    },
+    {
+      "epoch": 1.1357156434137996,
+      "grad_norm": 0.34295639916690035,
+      "learning_rate": 1.0687413130575176e-05,
+      "loss": 0.5132,
+      "step": 759
+    },
+    {
+      "epoch": 1.1372362668694165,
+      "grad_norm": 0.39357330789283446,
+      "learning_rate": 1.0655016907299346e-05,
+      "loss": 0.5288,
+      "step": 760
+    },
+    {
+      "epoch": 1.1387568903250334,
+      "grad_norm": 0.3690059684924214,
+      "learning_rate": 1.0622638014699909e-05,
+      "loss": 0.4947,
+      "step": 761
+    },
+    {
+      "epoch": 1.14027751378065,
+      "grad_norm": 0.3797018296519779,
+      "learning_rate": 1.0590276646611539e-05,
+      "loss": 0.5021,
+      "step": 762
+    },
+    {
+      "epoch": 1.1417981372362669,
+      "grad_norm": 0.37238784723531165,
+      "learning_rate": 1.0557932996764002e-05,
+      "loss": 0.5111,
+      "step": 763
+    },
+    {
+      "epoch": 1.1433187606918838,
+      "grad_norm": 0.34262491030815395,
+      "learning_rate": 1.0525607258781e-05,
+      "loss": 0.5095,
+      "step": 764
+    },
+    {
+      "epoch": 1.1448393841475004,
+      "grad_norm": 0.3516626249819126,
+      "learning_rate": 1.0493299626178997e-05,
+      "loss": 0.5256,
+      "step": 765
+    },
+    {
+      "epoch": 1.1463600076031173,
+      "grad_norm": 0.36336127776946636,
+      "learning_rate": 1.0461010292366076e-05,
+      "loss": 0.5163,
+      "step": 766
+    },
+    {
+      "epoch": 1.1478806310587342,
+      "grad_norm": 0.3660245007970849,
+      "learning_rate": 1.0428739450640768e-05,
+      "loss": 0.5025,
+      "step": 767
+    },
+    {
+      "epoch": 1.1494012545143508,
+      "grad_norm": 0.36947456530920786,
+      "learning_rate": 1.039648729419092e-05,
+      "loss": 0.5331,
+      "step": 768
+    },
+    {
+      "epoch": 1.1509218779699677,
+      "grad_norm": 0.35068301401351004,
+      "learning_rate": 1.0364254016092499e-05,
+      "loss": 0.4979,
+      "step": 769
+    },
+    {
+      "epoch": 1.1524425014255846,
+      "grad_norm": 0.3711143416445139,
+      "learning_rate": 1.033203980930847e-05,
+      "loss": 0.5349,
+      "step": 770
+    },
+    {
+      "epoch": 1.1539631248812012,
+      "grad_norm": 0.3739635788047117,
+      "learning_rate": 1.0299844866687616e-05,
+      "loss": 0.4975,
+      "step": 771
+    },
+    {
+      "epoch": 1.155483748336818,
+      "grad_norm": 0.33767131159596325,
+      "learning_rate": 1.0267669380963419e-05,
+      "loss": 0.4986,
+      "step": 772
+    },
+    {
+      "epoch": 1.157004371792435,
+      "grad_norm": 0.4087303964655199,
+      "learning_rate": 1.0235513544752866e-05,
+      "loss": 0.4995,
+      "step": 773
+    },
+    {
+      "epoch": 1.1585249952480516,
+      "grad_norm": 0.3669384792321335,
+      "learning_rate": 1.0203377550555312e-05,
+      "loss": 0.5033,
+      "step": 774
+    },
+    {
+      "epoch": 1.1600456187036685,
+      "grad_norm": 0.3582614117799248,
+      "learning_rate": 1.0171261590751334e-05,
+      "loss": 0.5068,
+      "step": 775
+    },
+    {
+      "epoch": 1.1615662421592854,
+      "grad_norm": 0.3548960213582146,
+      "learning_rate": 1.0139165857601578e-05,
+      "loss": 0.5219,
+      "step": 776
+    },
+    {
+      "epoch": 1.163086865614902,
+      "grad_norm": 0.35769530792522997,
+      "learning_rate": 1.0107090543245597e-05,
+      "loss": 0.4991,
+      "step": 777
+    },
+    {
+      "epoch": 1.164607489070519,
+      "grad_norm": 0.3762621019947577,
+      "learning_rate": 1.0075035839700708e-05,
+      "loss": 0.5318,
+      "step": 778
+    },
+    {
+      "epoch": 1.1661281125261358,
+      "grad_norm": 0.3428064272464147,
+      "learning_rate": 1.0043001938860842e-05,
+      "loss": 0.5168,
+      "step": 779
+    },
+    {
+      "epoch": 1.1676487359817525,
+      "grad_norm": 0.3740132674321784,
+      "learning_rate": 1.001098903249541e-05,
+      "loss": 0.512,
+      "step": 780
+    },
+    {
+      "epoch": 1.1691693594373693,
+      "grad_norm": 0.37086418652390696,
+      "learning_rate": 9.978997312248124e-06,
+      "loss": 0.5046,
+      "step": 781
+    },
+    {
+      "epoch": 1.1706899828929862,
+      "grad_norm": 0.34389709834427135,
+      "learning_rate": 9.947026969635868e-06,
+      "loss": 0.499,
+      "step": 782
+    },
+    {
+      "epoch": 1.1722106063486029,
+      "grad_norm": 0.36925894599740416,
+      "learning_rate": 9.915078196047548e-06,
+      "loss": 0.5269,
+      "step": 783
+    },
+    {
+      "epoch": 1.1737312298042197,
+      "grad_norm": 0.34734014615226194,
+      "learning_rate": 9.883151182742967e-06,
+      "loss": 0.5159,
+      "step": 784
+    },
+    {
+      "epoch": 1.1752518532598366,
+      "grad_norm": 0.38858923268368156,
+      "learning_rate": 9.851246120851641e-06,
+      "loss": 0.5599,
+      "step": 785
+    },
+    {
+      "epoch": 1.1767724767154533,
+      "grad_norm": 0.36352194480464795,
+      "learning_rate": 9.819363201371682e-06,
+      "loss": 0.5077,
+      "step": 786
+    },
+    {
+      "epoch": 1.1782931001710701,
+      "grad_norm": 0.38508119384888934,
+      "learning_rate": 9.78750261516864e-06,
+      "loss": 0.5455,
+      "step": 787
+    },
+    {
+      "epoch": 1.179813723626687,
+      "grad_norm": 0.35527440995181414,
+      "learning_rate": 9.755664552974386e-06,
+      "loss": 0.5128,
+      "step": 788
+    },
+    {
+      "epoch": 1.1813343470823037,
+      "grad_norm": 0.3524825961134292,
+      "learning_rate": 9.723849205385932e-06,
+      "loss": 0.5062,
+      "step": 789
+    },
+    {
+      "epoch": 1.1828549705379205,
+      "grad_norm": 0.4044926970167433,
+      "learning_rate": 9.692056762864325e-06,
+      "loss": 0.5412,
+      "step": 790
+    },
+    {
+      "epoch": 1.1843755939935374,
+      "grad_norm": 0.35710970899973166,
+      "learning_rate": 9.660287415733477e-06,
+      "loss": 0.52,
+      "step": 791
+    },
+    {
+      "epoch": 1.185896217449154,
+      "grad_norm": 0.37803033804258834,
+      "learning_rate": 9.62854135417906e-06,
+      "loss": 0.5052,
+      "step": 792
+    },
+    {
+      "epoch": 1.187416840904771,
+      "grad_norm": 0.39768355225535407,
+      "learning_rate": 9.596818768247333e-06,
+      "loss": 0.5092,
+      "step": 793
+    },
+    {
+      "epoch": 1.1889374643603878,
+      "grad_norm": 0.348711884248085,
+      "learning_rate": 9.565119847844027e-06,
+      "loss": 0.5267,
+      "step": 794
+    },
+    {
+      "epoch": 1.1904580878160045,
+      "grad_norm": 0.3834819895765528,
+      "learning_rate": 9.533444782733183e-06,
+      "loss": 0.5037,
+      "step": 795
+    },
+    {
+      "epoch": 1.1919787112716214,
+      "grad_norm": 0.3715822681968465,
+      "learning_rate": 9.501793762536068e-06,
+      "loss": 0.5308,
+      "step": 796
+    },
+    {
+      "epoch": 1.1934993347272382,
+      "grad_norm": 0.3652977625843937,
+      "learning_rate": 9.470166976729976e-06,
+      "loss": 0.5372,
+      "step": 797
+    },
+    {
+      "epoch": 1.195019958182855,
+      "grad_norm": 0.3364456352273384,
+      "learning_rate": 9.43856461464713e-06,
+      "loss": 0.4963,
+      "step": 798
+    },
+    {
+      "epoch": 1.1965405816384718,
+      "grad_norm": 0.3472569441190444,
+      "learning_rate": 9.40698686547354e-06,
+      "loss": 0.5284,
+      "step": 799
+    },
+    {
+      "epoch": 1.1980612050940886,
+      "grad_norm": 0.34440523348480445,
+      "learning_rate": 9.375433918247888e-06,
+      "loss": 0.4989,
+      "step": 800
+    },
+    {
+      "epoch": 1.1995818285497053,
+      "grad_norm": 0.348206283208153,
+      "learning_rate": 9.343905961860358e-06,
+      "loss": 0.5226,
+      "step": 801
+    },
+    {
+      "epoch": 1.2011024520053222,
+      "grad_norm": 0.3505525203415447,
+      "learning_rate": 9.31240318505154e-06,
+      "loss": 0.5047,
+      "step": 802
+    },
+    {
+      "epoch": 1.202623075460939,
+      "grad_norm": 0.3515111190341964,
+      "learning_rate": 9.280925776411278e-06,
+      "loss": 0.5127,
+      "step": 803
+    },
+    {
+      "epoch": 1.2041436989165557,
+      "grad_norm": 0.3620692993419406,
+      "learning_rate": 9.249473924377573e-06,
+      "loss": 0.5044,
+      "step": 804
+    },
+    {
+      "epoch": 1.2056643223721726,
+      "grad_norm": 0.3659564097222614,
+      "learning_rate": 9.218047817235408e-06,
+      "loss": 0.5319,
+      "step": 805
+    },
+    {
+      "epoch": 1.2071849458277895,
+      "grad_norm": 0.33548451693694653,
+      "learning_rate": 9.186647643115659e-06,
+      "loss": 0.4965,
+      "step": 806
+    },
+    {
+      "epoch": 1.2087055692834061,
+      "grad_norm": 0.37688953100329553,
+      "learning_rate": 9.15527358999394e-06,
+      "loss": 0.5453,
+      "step": 807
+    },
+    {
+      "epoch": 1.210226192739023,
+      "grad_norm": 0.3428883181534561,
+      "learning_rate": 9.123925845689515e-06,
+      "loss": 0.5392,
+      "step": 808
+    },
+    {
+      "epoch": 1.2117468161946399,
+      "grad_norm": 0.3431087734831433,
+      "learning_rate": 9.092604597864141e-06,
+      "loss": 0.5118,
+      "step": 809
+    },
+    {
+      "epoch": 1.2132674396502565,
+      "grad_norm": 0.35251065259476605,
+      "learning_rate": 9.06131003402095e-06,
+      "loss": 0.5054,
+      "step": 810
+    },
+    {
+      "epoch": 1.2147880631058734,
+      "grad_norm": 0.3496035139299239,
+      "learning_rate": 9.030042341503332e-06,
+      "loss": 0.5084,
+      "step": 811
+    },
+    {
+      "epoch": 1.2163086865614903,
+      "grad_norm": 0.33786656336389453,
+      "learning_rate": 8.998801707493827e-06,
+      "loss": 0.5199,
+      "step": 812
+    },
+    {
+      "epoch": 1.217829310017107,
+      "grad_norm": 0.3668696340788812,
+      "learning_rate": 8.967588319012977e-06,
+      "loss": 0.4874,
+      "step": 813
+    },
+    {
+      "epoch": 1.2193499334727238,
+      "grad_norm": 0.34867802829711625,
+      "learning_rate": 8.936402362918224e-06,
+      "loss": 0.4939,
+      "step": 814
+    },
+    {
+      "epoch": 1.2208705569283407,
+      "grad_norm": 0.33881823564824526,
+      "learning_rate": 8.90524402590278e-06,
+      "loss": 0.5354,
+      "step": 815
+    },
+    {
+      "epoch": 1.2223911803839573,
+      "grad_norm": 0.3395916132176579,
+      "learning_rate": 8.874113494494537e-06,
+      "loss": 0.5017,
+      "step": 816
+    },
+    {
+      "epoch": 1.2239118038395742,
+      "grad_norm": 0.33632837523359915,
+      "learning_rate": 8.843010955054909e-06,
+      "loss": 0.5006,
+      "step": 817
+    },
+    {
+      "epoch": 1.225432427295191,
+      "grad_norm": 0.33436403561258093,
+      "learning_rate": 8.811936593777748e-06,
+      "loss": 0.5052,
+      "step": 818
+    },
+    {
+      "epoch": 1.2269530507508077,
+      "grad_norm": 0.3224896536759886,
+      "learning_rate": 8.780890596688211e-06,
+      "loss": 0.4981,
+      "step": 819
+    },
+    {
+      "epoch": 1.2284736742064246,
+      "grad_norm": 0.34938066443230276,
+      "learning_rate": 8.749873149641665e-06,
+      "loss": 0.5203,
+      "step": 820
+    },
+    {
+      "epoch": 1.2299942976620415,
+      "grad_norm": 0.345563582040282,
+      "learning_rate": 8.718884438322557e-06,
+      "loss": 0.5162,
+      "step": 821
+    },
+    {
+      "epoch": 1.2315149211176581,
+      "grad_norm": 0.3428900758147224,
+      "learning_rate": 8.687924648243311e-06,
+      "loss": 0.4962,
+      "step": 822
+    },
+    {
+      "epoch": 1.233035544573275,
+      "grad_norm": 0.34750322921749416,
+      "learning_rate": 8.65699396474321e-06,
+      "loss": 0.5186,
+      "step": 823
+    },
+    {
+      "epoch": 1.234556168028892,
+      "grad_norm": 0.3589082109646499,
+      "learning_rate": 8.626092572987297e-06,
+      "loss": 0.5078,
+      "step": 824
+    },
+    {
+      "epoch": 1.2360767914845086,
+      "grad_norm": 0.3506370697057621,
+      "learning_rate": 8.595220657965268e-06,
+      "loss": 0.5105,
+      "step": 825
+    },
+    {
+      "epoch": 1.2375974149401254,
+      "grad_norm": 0.36011717525565096,
+      "learning_rate": 8.564378404490342e-06,
+      "loss": 0.5181,
+      "step": 826
+    },
+    {
+      "epoch": 1.2391180383957423,
+      "grad_norm": 0.3402637505055028,
+      "learning_rate": 8.533565997198175e-06,
+      "loss": 0.5186,
+      "step": 827
+    },
+    {
+      "epoch": 1.2406386618513592,
+      "grad_norm": 0.3475154384318057,
+      "learning_rate": 8.50278362054577e-06,
+      "loss": 0.4946,
+      "step": 828
+    },
+    {
+      "epoch": 1.2421592853069758,
+      "grad_norm": 0.3590598605292272,
+      "learning_rate": 8.472031458810325e-06,
+      "loss": 0.5011,
+      "step": 829
+    },
+    {
+      "epoch": 1.2436799087625927,
+      "grad_norm": 0.3483002221345324,
+      "learning_rate": 8.441309696088174e-06,
+      "loss": 0.5231,
+      "step": 830
+    },
+    {
+      "epoch": 1.2452005322182096,
+      "grad_norm": 0.3475177777900519,
+      "learning_rate": 8.410618516293662e-06,
+      "loss": 0.5165,
+      "step": 831
+    },
+    {
+      "epoch": 1.2467211556738262,
+      "grad_norm": 0.3633656751272662,
+      "learning_rate": 8.37995810315806e-06,
+      "loss": 0.5441,
+      "step": 832
+    },
+    {
+      "epoch": 1.2482417791294431,
+      "grad_norm": 0.3447799135446017,
+      "learning_rate": 8.349328640228443e-06,
+      "loss": 0.4933,
+      "step": 833
+    },
+    {
+      "epoch": 1.24976240258506,
+      "grad_norm": 0.34981751009747264,
+      "learning_rate": 8.318730310866615e-06,
+      "loss": 0.5008,
+      "step": 834
+    },
+    {
+      "epoch": 1.2512830260406766,
+      "grad_norm": 0.361234032357548,
+      "learning_rate": 8.288163298247994e-06,
+      "loss": 0.4979,
+      "step": 835
+    },
+    {
+      "epoch": 1.2528036494962935,
+      "grad_norm": 0.34834574638448507,
+      "learning_rate": 8.257627785360528e-06,
+      "loss": 0.5041,
+      "step": 836
+    },
+    {
+      "epoch": 1.2543242729519104,
+      "grad_norm": 0.3466954777402415,
+      "learning_rate": 8.227123955003586e-06,
+      "loss": 0.519,
+      "step": 837
+    },
+    {
+      "epoch": 1.255844896407527,
+      "grad_norm": 0.3555313535631245,
+      "learning_rate": 8.196651989786872e-06,
+      "loss": 0.513,
+      "step": 838
+    },
+    {
+      "epoch": 1.257365519863144,
+      "grad_norm": 0.36059188542064946,
+      "learning_rate": 8.166212072129334e-06,
+      "loss": 0.4975,
+      "step": 839
+    },
+    {
+      "epoch": 1.2588861433187608,
+      "grad_norm": 0.33893355712900364,
+      "learning_rate": 8.135804384258074e-06,
+      "loss": 0.4952,
+      "step": 840
+    },
+    {
+      "epoch": 1.2604067667743775,
+      "grad_norm": 0.33361223601614015,
+      "learning_rate": 8.10542910820724e-06,
+      "loss": 0.5061,
+      "step": 841
+    },
+    {
+      "epoch": 1.2619273902299943,
+      "grad_norm": 0.36477092577701276,
+      "learning_rate": 8.07508642581696e-06,
+      "loss": 0.5305,
+      "step": 842
+    },
+    {
+      "epoch": 1.2634480136856112,
+      "grad_norm": 0.3563029319847921,
+      "learning_rate": 8.044776518732229e-06,
+      "loss": 0.4953,
+      "step": 843
+    },
+    {
+      "epoch": 1.2649686371412279,
+      "grad_norm": 0.3480130680776454,
+      "learning_rate": 8.014499568401845e-06,
+      "loss": 0.5053,
+      "step": 844
+    },
+    {
+      "epoch": 1.2664892605968447,
+      "grad_norm": 0.33347701875752966,
+      "learning_rate": 7.98425575607731e-06,
+      "loss": 0.4956,
+      "step": 845
+    },
+    {
+      "epoch": 1.2680098840524616,
+      "grad_norm": 0.33599833733050055,
+      "learning_rate": 7.95404526281174e-06,
+      "loss": 0.5015,
+      "step": 846
+    },
+    {
+      "epoch": 1.2695305075080783,
+      "grad_norm": 0.3384957381349461,
+      "learning_rate": 7.923868269458795e-06,
+      "loss": 0.5062,
+      "step": 847
+    },
+    {
+      "epoch": 1.2710511309636952,
+      "grad_norm": 0.3252174947797439,
+      "learning_rate": 7.893724956671587e-06,
+      "loss": 0.4884,
+      "step": 848
+    },
+    {
+      "epoch": 1.272571754419312,
+      "grad_norm": 0.34540716209835276,
+      "learning_rate": 7.86361550490161e-06,
+      "loss": 0.5045,
+      "step": 849
+    },
+    {
+      "epoch": 1.2740923778749287,
+      "grad_norm": 0.3498563998796463,
+      "learning_rate": 7.833540094397626e-06,
+      "loss": 0.5118,
+      "step": 850
+    },
+    {
+      "epoch": 1.2756130013305456,
+      "grad_norm": 0.3647153038013536,
+      "learning_rate": 7.803498905204638e-06,
+      "loss": 0.5064,
+      "step": 851
+    },
+    {
+      "epoch": 1.2771336247861624,
+      "grad_norm": 0.3508552754714093,
+      "learning_rate": 7.773492117162775e-06,
+      "loss": 0.5035,
+      "step": 852
+    },
+    {
+      "epoch": 1.278654248241779,
+      "grad_norm": 0.33239766666827464,
+      "learning_rate": 7.743519909906217e-06,
+      "loss": 0.4986,
+      "step": 853
+    },
+    {
+      "epoch": 1.280174871697396,
+      "grad_norm": 0.4315070616641703,
+      "learning_rate": 7.713582462862138e-06,
+      "loss": 0.5331,
+      "step": 854
+    },
+    {
+      "epoch": 1.2816954951530128,
+      "grad_norm": 0.36173927943521234,
+      "learning_rate": 7.683679955249609e-06,
+      "loss": 0.5064,
+      "step": 855
+    },
+    {
+      "epoch": 1.2832161186086295,
+      "grad_norm": 0.3487948964838659,
+      "learning_rate": 7.653812566078565e-06,
+      "loss": 0.5061,
+      "step": 856
+    },
+    {
+      "epoch": 1.2847367420642464,
+      "grad_norm": 0.3484239609890833,
+      "learning_rate": 7.623980474148672e-06,
+      "loss": 0.5023,
+      "step": 857
+    },
+    {
+      "epoch": 1.2862573655198632,
+      "grad_norm": 0.36052769608217394,
+      "learning_rate": 7.59418385804832e-06,
+      "loss": 0.5267,
+      "step": 858
+    },
+    {
+      "epoch": 1.28777798897548,
+      "grad_norm": 0.37188400515013914,
+      "learning_rate": 7.564422896153491e-06,
+      "loss": 0.5455,
+      "step": 859
+    },
+    {
+      "epoch": 1.2892986124310968,
+      "grad_norm": 0.34411324307178126,
+      "learning_rate": 7.534697766626769e-06,
+      "loss": 0.5137,
+      "step": 860
+    },
+    {
+      "epoch": 1.2908192358867137,
+      "grad_norm": 0.3364768988197693,
+      "learning_rate": 7.50500864741619e-06,
+      "loss": 0.5209,
+      "step": 861
+    },
+    {
+      "epoch": 1.2923398593423303,
+      "grad_norm": 0.3670085290524646,
+      "learning_rate": 7.475355716254242e-06,
+      "loss": 0.5402,
+      "step": 862
+    },
+    {
+      "epoch": 1.2938604827979472,
+      "grad_norm": 0.33251162694155534,
+      "learning_rate": 7.445739150656753e-06,
+      "loss": 0.4987,
+      "step": 863
+    },
+    {
+      "epoch": 1.295381106253564,
+      "grad_norm": 0.35070274634255805,
+      "learning_rate": 7.416159127921868e-06,
+      "loss": 0.5071,
+      "step": 864
+    },
+    {
+      "epoch": 1.2969017297091807,
+      "grad_norm": 0.3795469724864985,
+      "learning_rate": 7.38661582512897e-06,
+      "loss": 0.5399,
+      "step": 865
+    },
+    {
+      "epoch": 1.2984223531647976,
+      "grad_norm": 0.3396586485316285,
+      "learning_rate": 7.3571094191375995e-06,
+      "loss": 0.5218,
+      "step": 866
+    },
+    {
+      "epoch": 1.2999429766204145,
+      "grad_norm": 0.3360542873886429,
+      "learning_rate": 7.327640086586438e-06,
+      "loss": 0.5034,
+      "step": 867
+    },
+    {
+      "epoch": 1.3014636000760311,
+      "grad_norm": 0.3295788458214383,
+      "learning_rate": 7.298208003892223e-06,
+      "loss": 0.5096,
+      "step": 868
+    },
+    {
+      "epoch": 1.302984223531648,
+      "grad_norm": 0.35244946035448127,
+      "learning_rate": 7.268813347248696e-06,
+      "loss": 0.5085,
+      "step": 869
+    },
+    {
+      "epoch": 1.3045048469872649,
+      "grad_norm": 0.3451919708597768,
+      "learning_rate": 7.239456292625539e-06,
+      "loss": 0.5017,
+      "step": 870
+    },
+    {
+      "epoch": 1.3060254704428815,
+      "grad_norm": 0.3208050987232861,
+      "learning_rate": 7.210137015767349e-06,
+      "loss": 0.5024,
+      "step": 871
+    },
+    {
+      "epoch": 1.3075460938984984,
+      "grad_norm": 0.35131393844950287,
+      "learning_rate": 7.1808556921925585e-06,
+      "loss": 0.529,
+      "step": 872
+    },
+    {
+      "epoch": 1.3090667173541153,
+      "grad_norm": 0.35801843724419535,
+      "learning_rate": 7.1516124971924e-06,
+      "loss": 0.5297,
+      "step": 873
+    },
+    {
+      "epoch": 1.310587340809732,
+      "grad_norm": 0.3401955783986041,
+      "learning_rate": 7.122407605829839e-06,
+      "loss": 0.5035,
+      "step": 874
+    },
+    {
+      "epoch": 1.3121079642653488,
+      "grad_norm": 0.3371522526619987,
+      "learning_rate": 7.093241192938552e-06,
+      "loss": 0.4876,
+      "step": 875
+    },
+    {
+      "epoch": 1.3136285877209657,
+      "grad_norm": 0.35482672035164964,
+      "learning_rate": 7.064113433121862e-06,
+      "loss": 0.5243,
+      "step": 876
+    },
+    {
+      "epoch": 1.3151492111765823,
+      "grad_norm": 0.3535329850445862,
+      "learning_rate": 7.03502450075169e-06,
+      "loss": 0.5081,
+      "step": 877
+    },
+    {
+      "epoch": 1.3166698346321992,
+      "grad_norm": 0.3423103799188622,
+      "learning_rate": 7.00597456996753e-06,
+      "loss": 0.5121,
+      "step": 878
+    },
+    {
+      "epoch": 1.318190458087816,
+      "grad_norm": 0.35140677064268294,
+      "learning_rate": 6.976963814675376e-06,
+      "loss": 0.5171,
+      "step": 879
+    },
+    {
+      "epoch": 1.3197110815434328,
+      "grad_norm": 0.352748609509177,
+      "learning_rate": 6.947992408546731e-06,
+      "loss": 0.535,
+      "step": 880
+    },
+    {
+      "epoch": 1.3212317049990496,
+      "grad_norm": 0.3317879618506482,
+      "learning_rate": 6.919060525017507e-06,
+      "loss": 0.4942,
+      "step": 881
+    },
+    {
+      "epoch": 1.3227523284546665,
+      "grad_norm": 0.3547220499965655,
+      "learning_rate": 6.890168337287031e-06,
+      "loss": 0.5091,
+      "step": 882
+    },
+    {
+      "epoch": 1.3242729519102832,
+      "grad_norm": 0.32077453688103713,
+      "learning_rate": 6.861316018316979e-06,
+      "loss": 0.515,
+      "step": 883
+    },
+    {
+      "epoch": 1.3257935753659,
+      "grad_norm": 0.3399152548809487,
+      "learning_rate": 6.832503740830382e-06,
+      "loss": 0.507,
+      "step": 884
+    },
+    {
+      "epoch": 1.327314198821517,
+      "grad_norm": 0.3318950288383037,
+      "learning_rate": 6.803731677310529e-06,
+      "loss": 0.4966,
+      "step": 885
+    },
+    {
+      "epoch": 1.3288348222771336,
+      "grad_norm": 0.3464596127850353,
+      "learning_rate": 6.775000000000004e-06,
+      "loss": 0.5115,
+      "step": 886
+    },
+    {
+      "epoch": 1.3303554457327504,
+      "grad_norm": 0.34400466984436623,
+      "learning_rate": 6.746308880899589e-06,
+      "loss": 0.5082,
+      "step": 887
+    },
+    {
+      "epoch": 1.3318760691883673,
+      "grad_norm": 0.3700416848449434,
+      "learning_rate": 6.71765849176729e-06,
+      "loss": 0.5302,
+      "step": 888
+    },
+    {
+      "epoch": 1.333396692643984,
+      "grad_norm": 0.333378413090238,
+      "learning_rate": 6.689049004117282e-06,
+      "loss": 0.5021,
+      "step": 889
+    },
+    {
+      "epoch": 1.3349173160996008,
+      "grad_norm": 0.3384873522425988,
+      "learning_rate": 6.660480589218871e-06,
+      "loss": 0.4805,
+      "step": 890
+    },
+    {
+      "epoch": 1.3364379395552177,
+      "grad_norm": 0.34806352457541345,
+      "learning_rate": 6.631953418095496e-06,
+      "loss": 0.4912,
+      "step": 891
+    },
+    {
+      "epoch": 1.3379585630108344,
+      "grad_norm": 0.34876873046591195,
+      "learning_rate": 6.603467661523689e-06,
+      "loss": 0.5082,
+      "step": 892
+    },
+    {
+      "epoch": 1.3394791864664513,
+      "grad_norm": 0.31774849749377526,
+      "learning_rate": 6.5750234900320585e-06,
+      "loss": 0.4818,
+      "step": 893
+    },
+    {
+      "epoch": 1.3409998099220681,
+      "grad_norm": 0.3324374020356446,
+      "learning_rate": 6.546621073900255e-06,
+      "loss": 0.4998,
+      "step": 894
+    },
+    {
+      "epoch": 1.3425204333776848,
+      "grad_norm": 0.3353452020425694,
+      "learning_rate": 6.518260583157976e-06,
+      "loss": 0.4921,
+      "step": 895
+    },
+    {
+      "epoch": 1.3440410568333017,
+      "grad_norm": 0.3493478007336028,
+      "learning_rate": 6.489942187583926e-06,
+      "loss": 0.5116,
+      "step": 896
+    },
+    {
+      "epoch": 1.3455616802889185,
+      "grad_norm": 0.3284929233347785,
+      "learning_rate": 6.461666056704817e-06,
+      "loss": 0.5009,
+      "step": 897
+    },
+    {
+      "epoch": 1.3470823037445352,
+      "grad_norm": 0.34202330071520587,
+      "learning_rate": 6.433432359794333e-06,
+      "loss": 0.5034,
+      "step": 898
+    },
+    {
+      "epoch": 1.348602927200152,
+      "grad_norm": 0.33111021835858967,
+      "learning_rate": 6.405241265872139e-06,
+      "loss": 0.5044,
+      "step": 899
+    },
+    {
+      "epoch": 1.350123550655769,
+      "grad_norm": 0.3354271189663323,
+      "learning_rate": 6.37709294370286e-06,
+      "loss": 0.4976,
+      "step": 900
+    },
+    {
+      "epoch": 1.3516441741113856,
+      "grad_norm": 0.33711612171640787,
+      "learning_rate": 6.3489875617950605e-06,
+      "loss": 0.5018,
+      "step": 901
+    },
+    {
+      "epoch": 1.3531647975670025,
+      "grad_norm": 0.33268125847154223,
+      "learning_rate": 6.320925288400259e-06,
+      "loss": 0.5061,
+      "step": 902
+    },
+    {
+      "epoch": 1.3546854210226194,
+      "grad_norm": 0.3297595687007157,
+      "learning_rate": 6.292906291511883e-06,
+      "loss": 0.5008,
+      "step": 903
+    },
+    {
+      "epoch": 1.356206044478236,
+      "grad_norm": 0.33711090972493357,
+      "learning_rate": 6.2649307388643245e-06,
+      "loss": 0.4899,
+      "step": 904
+    },
+    {
+      "epoch": 1.3577266679338529,
+      "grad_norm": 0.35714761250864835,
+      "learning_rate": 6.236998797931864e-06,
+      "loss": 0.5336,
+      "step": 905
+    },
+    {
+      "epoch": 1.3592472913894698,
+      "grad_norm": 0.33101330668166734,
+      "learning_rate": 6.209110635927724e-06,
+      "loss": 0.5188,
+      "step": 906
+    },
+    {
+      "epoch": 1.3607679148450864,
+      "grad_norm": 0.33231897329769716,
+      "learning_rate": 6.181266419803025e-06,
+      "loss": 0.4992,
+      "step": 907
+    },
+    {
+      "epoch": 1.3622885383007033,
+      "grad_norm": 0.3391798280349641,
+      "learning_rate": 6.153466316245841e-06,
+      "loss": 0.5145,
+      "step": 908
+    },
+    {
+      "epoch": 1.3638091617563202,
+      "grad_norm": 0.3397695996188909,
+      "learning_rate": 6.125710491680132e-06,
+      "loss": 0.5148,
+      "step": 909
+    },
+    {
+      "epoch": 1.3653297852119368,
+      "grad_norm": 0.33259300354324434,
+      "learning_rate": 6.097999112264814e-06,
+      "loss": 0.4983,
+      "step": 910
+    },
+    {
+      "epoch": 1.3668504086675537,
+      "grad_norm": 0.34841156749227775,
+      "learning_rate": 6.070332343892708e-06,
+      "loss": 0.5197,
+      "step": 911
+    },
+    {
+      "epoch": 1.3683710321231706,
+      "grad_norm": 0.3406141056782546,
+      "learning_rate": 6.042710352189592e-06,
+      "loss": 0.49,
+      "step": 912
+    },
+    {
+      "epoch": 1.3698916555787872,
+      "grad_norm": 0.3479407700510544,
+      "learning_rate": 6.01513330251319e-06,
+      "loss": 0.5054,
+      "step": 913
+    },
+    {
+      "epoch": 1.371412279034404,
+      "grad_norm": 0.34636977510682354,
+      "learning_rate": 5.9876013599521654e-06,
+      "loss": 0.5273,
+      "step": 914
+    },
+    {
+      "epoch": 1.372932902490021,
+      "grad_norm": 0.32587606375019723,
+      "learning_rate": 5.9601146893251685e-06,
+      "loss": 0.5032,
+      "step": 915
+    },
+    {
+      "epoch": 1.3744535259456376,
+      "grad_norm": 0.34209096131732425,
+      "learning_rate": 5.932673455179826e-06,
+      "loss": 0.5004,
+      "step": 916
+    },
+    {
+      "epoch": 1.3759741494012545,
+      "grad_norm": 0.3462622805308902,
+      "learning_rate": 5.9052778217917614e-06,
+      "loss": 0.5121,
+      "step": 917
+    },
+    {
+      "epoch": 1.3774947728568714,
+      "grad_norm": 0.3685686058332204,
+      "learning_rate": 5.8779279531636046e-06,
+      "loss": 0.5314,
+      "step": 918
+    },
+    {
+      "epoch": 1.379015396312488,
+      "grad_norm": 0.3515509029896568,
+      "learning_rate": 5.850624013024024e-06,
+      "loss": 0.5215,
+      "step": 919
+    },
+    {
+      "epoch": 1.380536019768105,
+      "grad_norm": 0.32046588423851496,
+      "learning_rate": 5.82336616482674e-06,
+      "loss": 0.4974,
+      "step": 920
+    },
+    {
+      "epoch": 1.3820566432237218,
+      "grad_norm": 0.3808130154691177,
+      "learning_rate": 5.796154571749547e-06,
+      "loss": 0.525,
+      "step": 921
+    },
+    {
+      "epoch": 1.3835772666793384,
+      "grad_norm": 0.3454746636565212,
+      "learning_rate": 5.768989396693324e-06,
+      "loss": 0.4923,
+      "step": 922
+    },
+    {
+      "epoch": 1.3850978901349553,
+      "grad_norm": 0.3321076118691578,
+      "learning_rate": 5.741870802281081e-06,
+      "loss": 0.4797,
+      "step": 923
+    },
+    {
+      "epoch": 1.3866185135905722,
+      "grad_norm": 0.3420626306833587,
+      "learning_rate": 5.714798950856981e-06,
+      "loss": 0.4923,
+      "step": 924
+    },
+    {
+      "epoch": 1.3881391370461889,
+      "grad_norm": 0.3353059399098719,
+      "learning_rate": 5.687774004485342e-06,
+      "loss": 0.5073,
+      "step": 925
+    },
+    {
+      "epoch": 1.3896597605018057,
+      "grad_norm": 0.3257554102906788,
+      "learning_rate": 5.660796124949711e-06,
+      "loss": 0.4942,
+      "step": 926
+    },
+    {
+      "epoch": 1.3911803839574226,
+      "grad_norm": 0.3413851694093131,
+      "learning_rate": 5.633865473751846e-06,
+      "loss": 0.5345,
+      "step": 927
+    },
+    {
+      "epoch": 1.3927010074130393,
+      "grad_norm": 0.33413542250569817,
+      "learning_rate": 5.606982212110806e-06,
+      "loss": 0.5205,
+      "step": 928
+    },
+    {
+      "epoch": 1.3942216308686561,
+      "grad_norm": 0.3413782795261156,
+      "learning_rate": 5.580146500961927e-06,
+      "loss": 0.4914,
+      "step": 929
+    },
+    {
+      "epoch": 1.395742254324273,
+      "grad_norm": 0.34562101413317636,
+      "learning_rate": 5.553358500955903e-06,
+      "loss": 0.5191,
+      "step": 930
+    },
+    {
+      "epoch": 1.3972628777798897,
+      "grad_norm": 0.36463942955800077,
+      "learning_rate": 5.526618372457788e-06,
+      "loss": 0.5031,
+      "step": 931
+    },
+    {
+      "epoch": 1.3987835012355065,
+      "grad_norm": 0.3490946304863948,
+      "learning_rate": 5.49992627554609e-06,
+      "loss": 0.5117,
+      "step": 932
+    },
+    {
+      "epoch": 1.4003041246911234,
+      "grad_norm": 0.35279790273339234,
+      "learning_rate": 5.473282370011742e-06,
+      "loss": 0.5274,
+      "step": 933
+    },
+    {
+      "epoch": 1.40182474814674,
+      "grad_norm": 0.349655925155731,
+      "learning_rate": 5.446686815357206e-06,
+      "loss": 0.4965,
+      "step": 934
+    },
+    {
+      "epoch": 1.403345371602357,
+      "grad_norm": 0.334218280540011,
+      "learning_rate": 5.420139770795477e-06,
+      "loss": 0.5216,
+      "step": 935
+    },
+    {
+      "epoch": 1.4048659950579738,
+      "grad_norm": 0.3630958366505651,
+      "learning_rate": 5.393641395249157e-06,
+      "loss": 0.5125,
+      "step": 936
+    },
+    {
+      "epoch": 1.4063866185135905,
+      "grad_norm": 0.3665354616192356,
+      "learning_rate": 5.367191847349496e-06,
+      "loss": 0.5171,
+      "step": 937
+    },
+    {
+      "epoch": 1.4079072419692074,
+      "grad_norm": 0.3388173032700996,
+      "learning_rate": 5.340791285435425e-06,
+      "loss": 0.5107,
+      "step": 938
+    },
+    {
+      "epoch": 1.4094278654248242,
+      "grad_norm": 0.3408412945651779,
+      "learning_rate": 5.314439867552638e-06,
+      "loss": 0.5092,
+      "step": 939
+    },
+    {
+      "epoch": 1.4109484888804409,
+      "grad_norm": 0.32178831689172704,
+      "learning_rate": 5.288137751452625e-06,
+      "loss": 0.5052,
+      "step": 940
+    },
+    {
+      "epoch": 1.4124691123360578,
+      "grad_norm": 0.3404604279948066,
+      "learning_rate": 5.261885094591735e-06,
+      "loss": 0.4989,
+      "step": 941
+    },
+    {
+      "epoch": 1.4139897357916746,
+      "grad_norm": 0.3528505749791748,
+      "learning_rate": 5.235682054130224e-06,
+      "loss": 0.4883,
+      "step": 942
+    },
+    {
+      "epoch": 1.4155103592472913,
+      "grad_norm": 0.33074842020507783,
+      "learning_rate": 5.209528786931329e-06,
+      "loss": 0.5143,
+      "step": 943
+    },
+    {
+      "epoch": 1.4170309827029082,
+      "grad_norm": 0.33129196603154915,
+      "learning_rate": 5.183425449560322e-06,
+      "loss": 0.5069,
+      "step": 944
+    },
+    {
+      "epoch": 1.418551606158525,
+      "grad_norm": 0.3450555489678585,
+      "learning_rate": 5.1573721982835766e-06,
+      "loss": 0.5194,
+      "step": 945
+    },
+    {
+      "epoch": 1.4200722296141417,
+      "grad_norm": 0.32000469036683216,
+      "learning_rate": 5.131369189067611e-06,
+      "loss": 0.4963,
+      "step": 946
+    },
+    {
+      "epoch": 1.4215928530697586,
+      "grad_norm": 0.3579257724532904,
+      "learning_rate": 5.105416577578189e-06,
+      "loss": 0.5009,
+      "step": 947
+    },
+    {
+      "epoch": 1.4231134765253755,
+      "grad_norm": 0.341889660403509,
+      "learning_rate": 5.079514519179368e-06,
+      "loss": 0.4955,
+      "step": 948
+    },
+    {
+      "epoch": 1.424634099980992,
+      "grad_norm": 0.3375109020144137,
+      "learning_rate": 5.053663168932563e-06,
+      "loss": 0.4925,
+      "step": 949
+    },
+    {
+      "epoch": 1.426154723436609,
+      "grad_norm": 0.3354140119830031,
+      "learning_rate": 5.02786268159564e-06,
+      "loss": 0.4828,
+      "step": 950
+    },
+    {
+      "epoch": 1.4276753468922259,
+      "grad_norm": 0.3411962173320927,
+      "learning_rate": 5.002113211621957e-06,
+      "loss": 0.5059,
+      "step": 951
+    },
+    {
+      "epoch": 1.4291959703478425,
+      "grad_norm": 0.3794888650970929,
+      "learning_rate": 4.97641491315949e-06,
+      "loss": 0.5182,
+      "step": 952
+    },
+    {
+      "epoch": 1.4307165938034594,
+      "grad_norm": 0.3438813382757282,
+      "learning_rate": 4.950767940049846e-06,
+      "loss": 0.5231,
+      "step": 953
+    },
+    {
+      "epoch": 1.4322372172590763,
+      "grad_norm": 0.33702340568088535,
+      "learning_rate": 4.925172445827401e-06,
+      "loss": 0.4898,
+      "step": 954
+    },
+    {
+      "epoch": 1.433757840714693,
+      "grad_norm": 0.33891170335451204,
+      "learning_rate": 4.8996285837183344e-06,
+      "loss": 0.4999,
+      "step": 955
+    },
+    {
+      "epoch": 1.4352784641703098,
+      "grad_norm": 0.3613620365932021,
+      "learning_rate": 4.874136506639747e-06,
+      "loss": 0.5109,
+      "step": 956
+    },
+    {
+      "epoch": 1.4367990876259267,
+      "grad_norm": 0.34896992869532173,
+      "learning_rate": 4.848696367198726e-06,
+      "loss": 0.503,
+      "step": 957
+    },
+    {
+      "epoch": 1.4383197110815433,
+      "grad_norm": 0.3319351812805159,
+      "learning_rate": 4.823308317691442e-06,
+      "loss": 0.5189,
+      "step": 958
+    },
+    {
+      "epoch": 1.4398403345371602,
+      "grad_norm": 0.3319386837420594,
+      "learning_rate": 4.7979725101022175e-06,
+      "loss": 0.4813,
+      "step": 959
+    },
+    {
+      "epoch": 1.441360957992777,
+      "grad_norm": 0.3397974585040773,
+      "learning_rate": 4.772689096102645e-06,
+      "loss": 0.5045,
+      "step": 960
+    },
+    {
+      "epoch": 1.4428815814483937,
+      "grad_norm": 0.31839220001298785,
+      "learning_rate": 4.747458227050663e-06,
+      "loss": 0.4815,
+      "step": 961
+    },
+    {
+      "epoch": 1.4444022049040106,
+      "grad_norm": 0.32870599013036034,
+      "learning_rate": 4.7222800539896445e-06,
+      "loss": 0.5142,
+      "step": 962
+    },
+    {
+      "epoch": 1.4459228283596275,
+      "grad_norm": 0.3342769560881234,
+      "learning_rate": 4.697154727647507e-06,
+      "loss": 0.4943,
+      "step": 963
+    },
+    {
+      "epoch": 1.4474434518152441,
+      "grad_norm": 0.34488780755334697,
+      "learning_rate": 4.672082398435805e-06,
+      "loss": 0.5137,
+      "step": 964
+    },
+    {
+      "epoch": 1.448964075270861,
+      "grad_norm": 0.3153153534150591,
+      "learning_rate": 4.6470632164488295e-06,
+      "loss": 0.4837,
+      "step": 965
+    },
+    {
+      "epoch": 1.450484698726478,
+      "grad_norm": 0.32145993722141425,
+      "learning_rate": 4.622097331462696e-06,
+      "loss": 0.4962,
+      "step": 966
+    },
+    {
+      "epoch": 1.4520053221820945,
+      "grad_norm": 0.33298486778180947,
+      "learning_rate": 4.597184892934472e-06,
+      "loss": 0.5103,
+      "step": 967
+    },
+    {
+      "epoch": 1.4535259456377114,
+      "grad_norm": 0.3450267637519744,
+      "learning_rate": 4.572326050001267e-06,
+      "loss": 0.4909,
+      "step": 968
+    },
+    {
+      "epoch": 1.4550465690933283,
+      "grad_norm": 0.3318326795828598,
+      "learning_rate": 4.547520951479347e-06,
+      "loss": 0.5057,
+      "step": 969
+    },
+    {
+      "epoch": 1.456567192548945,
+      "grad_norm": 0.33679166346861333,
+      "learning_rate": 4.522769745863228e-06,
+      "loss": 0.5057,
+      "step": 970
+    },
+    {
+      "epoch": 1.4580878160045618,
+      "grad_norm": 0.3395760337602621,
+      "learning_rate": 4.498072581324811e-06,
+      "loss": 0.5358,
+      "step": 971
+    },
+    {
+      "epoch": 1.4596084394601787,
+      "grad_norm": 0.3350852965124303,
+      "learning_rate": 4.473429605712482e-06,
+      "loss": 0.516,
+      "step": 972
+    },
+    {
+      "epoch": 1.4611290629157954,
+      "grad_norm": 0.3402104641226832,
+      "learning_rate": 4.448840966550214e-06,
+      "loss": 0.5042,
+      "step": 973
+    },
+    {
+      "epoch": 1.4626496863714122,
+      "grad_norm": 0.35213392448418906,
+      "learning_rate": 4.424306811036717e-06,
+      "loss": 0.5048,
+      "step": 974
+    },
+    {
+      "epoch": 1.4641703098270291,
+      "grad_norm": 0.30760979362082447,
+      "learning_rate": 4.399827286044515e-06,
+      "loss": 0.4976,
+      "step": 975
+    },
+    {
+      "epoch": 1.4656909332826458,
+      "grad_norm": 0.3713309741272073,
+      "learning_rate": 4.375402538119116e-06,
+      "loss": 0.4849,
+      "step": 976
+    },
+    {
+      "epoch": 1.4672115567382626,
+      "grad_norm": 0.32636579757729905,
+      "learning_rate": 4.351032713478084e-06,
+      "loss": 0.5011,
+      "step": 977
+    },
+    {
+      "epoch": 1.4687321801938795,
+      "grad_norm": 0.34427812460067464,
+      "learning_rate": 4.326717958010203e-06,
+      "loss": 0.4985,
+      "step": 978
+    },
+    {
+      "epoch": 1.4702528036494962,
+      "grad_norm": 0.3431253440023036,
+      "learning_rate": 4.302458417274573e-06,
+      "loss": 0.504,
+      "step": 979
+    },
+    {
+      "epoch": 1.471773427105113,
+      "grad_norm": 0.3219320912973509,
+      "learning_rate": 4.278254236499767e-06,
+      "loss": 0.4872,
+      "step": 980
+    },
+    {
+      "epoch": 1.47329405056073,
+      "grad_norm": 0.3259111453940081,
+      "learning_rate": 4.2541055605829445e-06,
+      "loss": 0.4915,
+      "step": 981
+    },
+    {
+      "epoch": 1.4748146740163466,
+      "grad_norm": 0.34356473045787483,
+      "learning_rate": 4.2300125340889885e-06,
+      "loss": 0.5041,
+      "step": 982
+    },
+    {
+      "epoch": 1.4763352974719635,
+      "grad_norm": 0.3302732477989796,
+      "learning_rate": 4.2059753012496334e-06,
+      "loss": 0.4975,
+      "step": 983
+    },
+    {
+      "epoch": 1.4778559209275803,
+      "grad_norm": 0.3139509766944064,
+      "learning_rate": 4.181994005962612e-06,
+      "loss": 0.4987,
+      "step": 984
+    },
+    {
+      "epoch": 1.479376544383197,
+      "grad_norm": 0.33007516038018775,
+      "learning_rate": 4.158068791790798e-06,
+      "loss": 0.5094,
+      "step": 985
+    },
+    {
+      "epoch": 1.4808971678388139,
+      "grad_norm": 0.3314396852932346,
+      "learning_rate": 4.134199801961319e-06,
+      "loss": 0.5214,
+      "step": 986
+    },
+    {
+      "epoch": 1.4824177912944307,
+      "grad_norm": 0.3265821672215369,
+      "learning_rate": 4.110387179364735e-06,
+      "loss": 0.4829,
+      "step": 987
+    },
+    {
+      "epoch": 1.4839384147500474,
+      "grad_norm": 0.3332694006250523,
+      "learning_rate": 4.0866310665541576e-06,
+      "loss": 0.5052,
+      "step": 988
+    },
+    {
+      "epoch": 1.4854590382056643,
+      "grad_norm": 0.3490720940072708,
+      "learning_rate": 4.062931605744416e-06,
+      "loss": 0.5092,
+      "step": 989
+    },
+    {
+      "epoch": 1.4869796616612811,
+      "grad_norm": 0.33398988538744084,
+      "learning_rate": 4.0392889388111746e-06,
+      "loss": 0.5188,
+      "step": 990
+    },
+    {
+      "epoch": 1.4885002851168978,
+      "grad_norm": 0.3270648319812971,
+      "learning_rate": 4.015703207290125e-06,
+      "loss": 0.5159,
+      "step": 991
+    },
+    {
+      "epoch": 1.4900209085725147,
+      "grad_norm": 0.32741030567450113,
+      "learning_rate": 3.992174552376112e-06,
+      "loss": 0.5025,
+      "step": 992
+    },
+    {
+      "epoch": 1.4915415320281316,
+      "grad_norm": 0.32463136930033865,
+      "learning_rate": 3.968703114922282e-06,
+      "loss": 0.4912,
+      "step": 993
+    },
+    {
+      "epoch": 1.4930621554837482,
+      "grad_norm": 0.3493187625119753,
+      "learning_rate": 3.945289035439266e-06,
+      "loss": 0.5022,
+      "step": 994
+    },
+    {
+      "epoch": 1.494582778939365,
+      "grad_norm": 0.3393915308815615,
+      "learning_rate": 3.921932454094324e-06,
+      "loss": 0.5121,
+      "step": 995
+    },
+    {
+      "epoch": 1.496103402394982,
+      "grad_norm": 0.3361281345069889,
+      "learning_rate": 3.8986335107104904e-06,
+      "loss": 0.5113,
+      "step": 996
+    },
+    {
+      "epoch": 1.4976240258505986,
+      "grad_norm": 0.33203560338943144,
+      "learning_rate": 3.875392344765772e-06,
+      "loss": 0.4754,
+      "step": 997
+    },
+    {
+      "epoch": 1.4991446493062155,
+      "grad_norm": 0.324328611770242,
+      "learning_rate": 3.852209095392288e-06,
+      "loss": 0.4993,
+      "step": 998
+    },
+    {
+      "epoch": 1.5006652727618324,
+      "grad_norm": 0.3558742459808321,
+      "learning_rate": 3.829083901375433e-06,
+      "loss": 0.5047,
+      "step": 999
+    },
+    {
+      "epoch": 1.502185896217449,
+      "grad_norm": 0.3338340547562902,
+      "learning_rate": 3.8060169011530668e-06,
+      "loss": 0.4759,
+      "step": 1000
+    },
+    {
+      "epoch": 1.5037065196730661,
+      "grad_norm": 0.35120584802546506,
+      "learning_rate": 3.783008232814671e-06,
+      "loss": 0.5294,
+      "step": 1001
+    },
+    {
+      "epoch": 1.5052271431286828,
+      "grad_norm": 0.3353936427085398,
+      "learning_rate": 3.760058034100533e-06,
+      "loss": 0.5284,
+      "step": 1002
+    },
+    {
+      "epoch": 1.5067477665842994,
+      "grad_norm": 0.33688668654855014,
+      "learning_rate": 3.7371664424008973e-06,
+      "loss": 0.4959,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5082683900399165,
+      "grad_norm": 0.32266166581151934,
+      "learning_rate": 3.714333594755177e-06,
+      "loss": 0.4752,
+      "step": 1004
+    },
+    {
+      "epoch": 1.5097890134955332,
+      "grad_norm": 0.33412231708735246,
+      "learning_rate": 3.691559627851118e-06,
+      "loss": 0.5179,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5113096369511498,
+      "grad_norm": 0.3243441675650943,
+      "learning_rate": 3.6688446780239596e-06,
+      "loss": 0.4943,
+      "step": 1006
+    },
+    {
+      "epoch": 1.512830260406767,
+      "grad_norm": 0.327154574269167,
+      "learning_rate": 3.646188881255658e-06,
+      "loss": 0.492,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5143508838623836,
+      "grad_norm": 0.3192294634078481,
+      "learning_rate": 3.623592373174045e-06,
+      "loss": 0.4771,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5158715073180002,
+      "grad_norm": 0.3372732645788767,
+      "learning_rate": 3.601055289052031e-06,
+      "loss": 0.4949,
+      "step": 1009
+    },
+    {
+      "epoch": 1.5173921307736173,
+      "grad_norm": 0.31829723772061774,
+      "learning_rate": 3.578577763806769e-06,
+      "loss": 0.49,
+      "step": 1010
+    },
+    {
+      "epoch": 1.518912754229234,
+      "grad_norm": 0.3374508150505602,
+      "learning_rate": 3.556159931998888e-06,
+      "loss": 0.5138,
+      "step": 1011
+    },
+    {
+      "epoch": 1.5204333776848507,
+      "grad_norm": 0.3368177667255345,
+      "learning_rate": 3.5338019278316426e-06,
+      "loss": 0.5332,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5219540011404677,
+      "grad_norm": 0.35595499914728906,
+      "learning_rate": 3.5115038851501605e-06,
+      "loss": 0.5093,
+      "step": 1013
+    },
+    {
+      "epoch": 1.5234746245960844,
+      "grad_norm": 0.3549641645181648,
+      "learning_rate": 3.4892659374405844e-06,
+      "loss": 0.5251,
+      "step": 1014
+    },
+    {
+      "epoch": 1.524995248051701,
+      "grad_norm": 0.32092817344554087,
+      "learning_rate": 3.467088217829321e-06,
+      "loss": 0.5086,
+      "step": 1015
+    },
+    {
+      "epoch": 1.5265158715073182,
+      "grad_norm": 0.3163937191442778,
+      "learning_rate": 3.4449708590822104e-06,
+      "loss": 0.4997,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5280364949629348,
+      "grad_norm": 0.3204985454805181,
+      "learning_rate": 3.422913993603756e-06,
+      "loss": 0.523,
+      "step": 1017
+    },
+    {
+      "epoch": 1.5295571184185515,
+      "grad_norm": 0.3282138238031523,
+      "learning_rate": 3.4009177534363184e-06,
+      "loss": 0.5096,
+      "step": 1018
+    },
+    {
+      "epoch": 1.5310777418741686,
+      "grad_norm": 0.3243188648838239,
+      "learning_rate": 3.378982270259323e-06,
+      "loss": 0.4974,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5325983653297852,
+      "grad_norm": 0.32762061133060594,
+      "learning_rate": 3.3571076753884802e-06,
+      "loss": 0.5174,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5341189887854019,
+      "grad_norm": 0.3315702913947939,
+      "learning_rate": 3.335294099774996e-06,
+      "loss": 0.5317,
+      "step": 1021
+    },
+    {
+      "epoch": 1.535639612241019,
+      "grad_norm": 0.3146514780388357,
+      "learning_rate": 3.313541674004791e-06,
+      "loss": 0.4823,
+      "step": 1022
+    },
+    {
+      "epoch": 1.5371602356966356,
+      "grad_norm": 0.31324712226441537,
+      "learning_rate": 3.291850528297705e-06,
+      "loss": 0.4936,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5386808591522523,
+      "grad_norm": 0.3220338054628576,
+      "learning_rate": 3.270220792506735e-06,
+      "loss": 0.4897,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5402014826078694,
+      "grad_norm": 0.33327886740248275,
+      "learning_rate": 3.2486525961172487e-06,
+      "loss": 0.5118,
+      "step": 1025
+    },
+    {
+      "epoch": 1.541722106063486,
+      "grad_norm": 0.31839568655177714,
+      "learning_rate": 3.2271460682462175e-06,
+      "loss": 0.4913,
+      "step": 1026
+    },
+    {
+      "epoch": 1.5432427295191027,
+      "grad_norm": 0.3275702952073424,
+      "learning_rate": 3.2057013376414218e-06,
+      "loss": 0.5022,
+      "step": 1027
+    },
+    {
+      "epoch": 1.5447633529747198,
+      "grad_norm": 0.3265904791104775,
+      "learning_rate": 3.1843185326807096e-06,
+      "loss": 0.5013,
+      "step": 1028
+    },
+    {
+      "epoch": 1.5462839764303364,
+      "grad_norm": 0.34562281052781296,
+      "learning_rate": 3.162997781371215e-06,
+      "loss": 0.5245,
+      "step": 1029
+    },
+    {
+      "epoch": 1.547804599885953,
+      "grad_norm": 0.3304005470169874,
+      "learning_rate": 3.141739211348575e-06,
+      "loss": 0.5034,
+      "step": 1030
+    },
+    {
+      "epoch": 1.5493252233415702,
+      "grad_norm": 0.32880228905554193,
+      "learning_rate": 3.1205429498761974e-06,
+      "loss": 0.5015,
+      "step": 1031
+    },
+    {
+      "epoch": 1.5508458467971868,
+      "grad_norm": 0.32336045302160143,
+      "learning_rate": 3.0994091238444775e-06,
+      "loss": 0.511,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5523664702528035,
+      "grad_norm": 0.3311162660651907,
+      "learning_rate": 3.0783378597700453e-06,
+      "loss": 0.5114,
+      "step": 1033
+    },
+    {
+      "epoch": 1.5538870937084206,
+      "grad_norm": 0.3375600053400384,
+      "learning_rate": 3.0573292837949997e-06,
+      "loss": 0.4995,
+      "step": 1034
+    },
+    {
+      "epoch": 1.5554077171640373,
+      "grad_norm": 0.3184801308022672,
+      "learning_rate": 3.036383521686171e-06,
+      "loss": 0.491,
+      "step": 1035
+    },
+    {
+      "epoch": 1.556928340619654,
+      "grad_norm": 0.3453112317336798,
+      "learning_rate": 3.015500698834343e-06,
+      "loss": 0.5348,
+      "step": 1036
+    },
+    {
+      "epoch": 1.558448964075271,
+      "grad_norm": 0.3178400015966866,
+      "learning_rate": 2.9946809402535396e-06,
+      "loss": 0.5184,
+      "step": 1037
+    },
+    {
+      "epoch": 1.5599695875308877,
+      "grad_norm": 0.32719134398419775,
+      "learning_rate": 2.9739243705802274e-06,
+      "loss": 0.515,
+      "step": 1038
+    },
+    {
+      "epoch": 1.5614902109865043,
+      "grad_norm": 0.3205800396648508,
+      "learning_rate": 2.953231114072617e-06,
+      "loss": 0.4964,
+      "step": 1039
+    },
+    {
+      "epoch": 1.5630108344421214,
+      "grad_norm": 0.316255767582057,
+      "learning_rate": 2.9326012946098793e-06,
+      "loss": 0.4772,
+      "step": 1040
+    },
+    {
+      "epoch": 1.564531457897738,
+      "grad_norm": 0.3175652603893164,
+      "learning_rate": 2.9120350356914355e-06,
+      "loss": 0.5086,
+      "step": 1041
+    },
+    {
+      "epoch": 1.5660520813533547,
+      "grad_norm": 0.32582238160345084,
+      "learning_rate": 2.891532460436206e-06,
+      "loss": 0.513,
+      "step": 1042
+    },
+    {
+      "epoch": 1.5675727048089718,
+      "grad_norm": 0.3296575006626612,
+      "learning_rate": 2.8710936915818587e-06,
+      "loss": 0.5023,
+      "step": 1043
+    },
+    {
+      "epoch": 1.5690933282645885,
+      "grad_norm": 0.31611337448881216,
+      "learning_rate": 2.8507188514840983e-06,
+      "loss": 0.5156,
+      "step": 1044
+    },
+    {
+      "epoch": 1.5706139517202051,
+      "grad_norm": 0.31450333160108834,
+      "learning_rate": 2.8304080621159222e-06,
+      "loss": 0.5039,
+      "step": 1045
+    },
+    {
+      "epoch": 1.5721345751758222,
+      "grad_norm": 0.3213166122638692,
+      "learning_rate": 2.8101614450668923e-06,
+      "loss": 0.4891,
+      "step": 1046
+    },
+    {
+      "epoch": 1.5736551986314389,
+      "grad_norm": 0.3410745502394496,
+      "learning_rate": 2.7899791215423954e-06,
+      "loss": 0.5091,
+      "step": 1047
+    },
+    {
+      "epoch": 1.5751758220870555,
+      "grad_norm": 0.3888744291624106,
+      "learning_rate": 2.76986121236294e-06,
+      "loss": 0.5075,
+      "step": 1048
+    },
+    {
+      "epoch": 1.5766964455426726,
+      "grad_norm": 0.32027232788552623,
+      "learning_rate": 2.7498078379634194e-06,
+      "loss": 0.4863,
+      "step": 1049
+    },
+    {
+      "epoch": 1.5782170689982893,
+      "grad_norm": 0.3219460654212709,
+      "learning_rate": 2.7298191183923903e-06,
+      "loss": 0.4919,
+      "step": 1050
+    },
+    {
+      "epoch": 1.579737692453906,
+      "grad_norm": 0.3189327947333556,
+      "learning_rate": 2.709895173311352e-06,
+      "loss": 0.507,
+      "step": 1051
+    },
+    {
+      "epoch": 1.581258315909523,
+      "grad_norm": 0.34438583434356085,
+      "learning_rate": 2.690036121994041e-06,
+      "loss": 0.4899,
+      "step": 1052
+    },
+    {
+      "epoch": 1.5827789393651397,
+      "grad_norm": 0.3398976699129184,
+      "learning_rate": 2.6702420833257135e-06,
+      "loss": 0.5225,
+      "step": 1053
+    },
+    {
+      "epoch": 1.5842995628207563,
+      "grad_norm": 0.3102911393257611,
+      "learning_rate": 2.6505131758024167e-06,
+      "loss": 0.5059,
+      "step": 1054
+    },
+    {
+      "epoch": 1.5858201862763734,
+      "grad_norm": 0.324270478165599,
+      "learning_rate": 2.630849517530311e-06,
+      "loss": 0.4981,
+      "step": 1055
+    },
+    {
+      "epoch": 1.58734080973199,
+      "grad_norm": 0.3546913242376741,
+      "learning_rate": 2.6112512262249243e-06,
+      "loss": 0.5361,
+      "step": 1056
+    },
+    {
+      "epoch": 1.5888614331876068,
+      "grad_norm": 0.34869377738298407,
+      "learning_rate": 2.591718419210495e-06,
+      "loss": 0.5103,
+      "step": 1057
+    },
+    {
+      "epoch": 1.5903820566432239,
+      "grad_norm": 0.3123532728864231,
+      "learning_rate": 2.572251213419218e-06,
+      "loss": 0.4932,
+      "step": 1058
+    },
+    {
+      "epoch": 1.5919026800988405,
+      "grad_norm": 0.32330942665052126,
+      "learning_rate": 2.5528497253905883e-06,
+      "loss": 0.5037,
+      "step": 1059
+    },
+    {
+      "epoch": 1.5934233035544574,
+      "grad_norm": 0.31181836853952816,
+      "learning_rate": 2.5335140712706637e-06,
+      "loss": 0.4866,
+      "step": 1060
+    },
+    {
+      "epoch": 1.5949439270100743,
+      "grad_norm": 0.3030338685118879,
+      "learning_rate": 2.5142443668114184e-06,
+      "loss": 0.4618,
+      "step": 1061
+    },
+    {
+      "epoch": 1.596464550465691,
+      "grad_norm": 0.3270760691359473,
+      "learning_rate": 2.495040727369997e-06,
+      "loss": 0.5345,
+      "step": 1062
+    },
+    {
+      "epoch": 1.5979851739213078,
+      "grad_norm": 0.31027653370341124,
+      "learning_rate": 2.4759032679080646e-06,
+      "loss": 0.4881,
+      "step": 1063
+    },
+    {
+      "epoch": 1.5995057973769247,
+      "grad_norm": 0.3062497695212509,
+      "learning_rate": 2.4568321029910926e-06,
+      "loss": 0.4787,
+      "step": 1064
+    },
+    {
+      "epoch": 1.6010264208325413,
+      "grad_norm": 0.32621381275453964,
+      "learning_rate": 2.4378273467876895e-06,
+      "loss": 0.525,
+      "step": 1065
+    },
+    {
+      "epoch": 1.6025470442881582,
+      "grad_norm": 0.3318069849664119,
+      "learning_rate": 2.4188891130689163e-06,
+      "loss": 0.4909,
+      "step": 1066
+    },
+    {
+      "epoch": 1.604067667743775,
+      "grad_norm": 0.32860192649298986,
+      "learning_rate": 2.4000175152075832e-06,
+      "loss": 0.5047,
+      "step": 1067
+    },
+    {
+      "epoch": 1.6055882911993917,
+      "grad_norm": 0.32129364609495753,
+      "learning_rate": 2.3812126661776048e-06,
+      "loss": 0.49,
+      "step": 1068
+    },
+    {
+      "epoch": 1.6071089146550086,
+      "grad_norm": 0.34479880933344564,
+      "learning_rate": 2.3624746785533015e-06,
+      "loss": 0.5202,
+      "step": 1069
+    },
+    {
+      "epoch": 1.6086295381106255,
+      "grad_norm": 0.32050670560042915,
+      "learning_rate": 2.3438036645087323e-06,
+      "loss": 0.5045,
+      "step": 1070
+    },
+    {
+      "epoch": 1.6101501615662421,
+      "grad_norm": 0.3236770692201634,
+      "learning_rate": 2.325199735817016e-06,
+      "loss": 0.5007,
+      "step": 1071
+    },
+    {
+      "epoch": 1.611670785021859,
+      "grad_norm": 0.3182094882577364,
+      "learning_rate": 2.306663003849674e-06,
+      "loss": 0.4972,
+      "step": 1072
+    },
+    {
+      "epoch": 1.6131914084774759,
+      "grad_norm": 0.31720365603712886,
+      "learning_rate": 2.2881935795759588e-06,
+      "loss": 0.4716,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6147120319330925,
+      "grad_norm": 0.32974314670568317,
+      "learning_rate": 2.2697915735621873e-06,
+      "loss": 0.484,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6162326553887094,
+      "grad_norm": 0.30577586884242175,
+      "learning_rate": 2.2514570959710736e-06,
+      "loss": 0.4735,
+      "step": 1075
+    },
+    {
+      "epoch": 1.6177532788443263,
+      "grad_norm": 0.3197001472064691,
+      "learning_rate": 2.2331902565610855e-06,
+      "loss": 0.5218,
+      "step": 1076
+    },
+    {
+      "epoch": 1.619273902299943,
+      "grad_norm": 0.3166847446707726,
+      "learning_rate": 2.214991164685776e-06,
+      "loss": 0.4903,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6207945257555598,
+      "grad_norm": 0.3195718257381146,
+      "learning_rate": 2.1968599292931262e-06,
+      "loss": 0.4971,
+      "step": 1078
+    },
+    {
+      "epoch": 1.6223151492111767,
+      "grad_norm": 0.31426671771687575,
+      "learning_rate": 2.178796658924904e-06,
+      "loss": 0.5064,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6238357726667934,
+      "grad_norm": 0.3212716102723848,
+      "learning_rate": 2.160801461715996e-06,
+      "loss": 0.5079,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6253563961224102,
+      "grad_norm": 0.3191289490920015,
+      "learning_rate": 2.1428744453937935e-06,
+      "loss": 0.4761,
+      "step": 1081
+    },
+    {
+      "epoch": 1.626877019578027,
+      "grad_norm": 0.3159849608628463,
+      "learning_rate": 2.125015717277504e-06,
+      "loss": 0.5033,
+      "step": 1082
+    },
+    {
+      "epoch": 1.6283976430336438,
+      "grad_norm": 0.31696066878201823,
+      "learning_rate": 2.1072253842775457e-06,
+      "loss": 0.4788,
+      "step": 1083
+    },
+    {
+      "epoch": 1.6299182664892606,
+      "grad_norm": 0.3258225759542187,
+      "learning_rate": 2.089503552894881e-06,
+      "loss": 0.5099,
+      "step": 1084
+    },
+    {
+      "epoch": 1.6314388899448775,
+      "grad_norm": 0.3097721381921143,
+      "learning_rate": 2.0718503292204057e-06,
+      "loss": 0.4837,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6329595134004942,
+      "grad_norm": 0.32471560408203043,
+      "learning_rate": 2.054265818934283e-06,
+      "loss": 0.523,
+      "step": 1086
+    },
+    {
+      "epoch": 1.634480136856111,
+      "grad_norm": 0.30086700392629573,
+      "learning_rate": 2.036750127305341e-06,
+      "loss": 0.4751,
+      "step": 1087
+    },
+    {
+      "epoch": 1.636000760311728,
+      "grad_norm": 0.306302856546338,
+      "learning_rate": 2.0193033591904125e-06,
+      "loss": 0.4734,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6375213837673446,
+      "grad_norm": 0.30581159890991627,
+      "learning_rate": 2.001925619033737e-06,
+      "loss": 0.4881,
+      "step": 1089
+    },
+    {
+      "epoch": 1.6390420072229615,
+      "grad_norm": 0.3148442376719998,
+      "learning_rate": 1.9846170108663187e-06,
+      "loss": 0.4862,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6405626306785783,
+      "grad_norm": 0.3187412303526785,
+      "learning_rate": 1.9673776383052978e-06,
+      "loss": 0.5088,
+      "step": 1091
+    },
+    {
+      "epoch": 1.642083254134195,
+      "grad_norm": 0.3237394151376111,
+      "learning_rate": 1.9502076045533484e-06,
+      "loss": 0.5127,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6436038775898119,
+      "grad_norm": 0.30397789760314514,
+      "learning_rate": 1.9331070123980477e-06,
+      "loss": 0.4906,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6451245010454287,
+      "grad_norm": 0.31316588193164063,
+      "learning_rate": 1.916075964211268e-06,
+      "loss": 0.496,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6466451245010454,
+      "grad_norm": 0.33024840735397576,
+      "learning_rate": 1.8991145619485479e-06,
+      "loss": 0.5203,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6481657479566623,
+      "grad_norm": 0.3409025779210624,
+      "learning_rate": 1.8822229071485085e-06,
+      "loss": 0.5002,
+      "step": 1096
+    },
+    {
+      "epoch": 1.6496863714122791,
+      "grad_norm": 0.3287972930467973,
+      "learning_rate": 1.8654011009322228e-06,
+      "loss": 0.4995,
+      "step": 1097
+    },
+    {
+      "epoch": 1.6512069948678958,
+      "grad_norm": 0.32623077457196087,
+      "learning_rate": 1.8486492440026244e-06,
+      "loss": 0.487,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6527276183235127,
+      "grad_norm": 0.3106687896956913,
+      "learning_rate": 1.83196743664389e-06,
+      "loss": 0.4946,
+      "step": 1099
+    },
+    {
+      "epoch": 1.6542482417791295,
+      "grad_norm": 0.32812705879714626,
+      "learning_rate": 1.8153557787208556e-06,
+      "loss": 0.4903,
+      "step": 1100
+    },
+    {
+      "epoch": 1.6557688652347462,
+      "grad_norm": 0.31890170792842726,
+      "learning_rate": 1.7988143696784124e-06,
+      "loss": 0.4821,
+      "step": 1101
+    },
+    {
+      "epoch": 1.657289488690363,
+      "grad_norm": 0.31284347487731023,
+      "learning_rate": 1.7823433085409e-06,
+      "loss": 0.4931,
+      "step": 1102
+    },
+    {
+      "epoch": 1.65881011214598,
+      "grad_norm": 0.30982765014249525,
+      "learning_rate": 1.7659426939115348e-06,
+      "loss": 0.4683,
+      "step": 1103
+    },
+    {
+      "epoch": 1.6603307356015966,
+      "grad_norm": 0.3164620121768728,
+      "learning_rate": 1.7496126239717968e-06,
+      "loss": 0.5007,
+      "step": 1104
+    },
+    {
+      "epoch": 1.6618513590572135,
+      "grad_norm": 0.3479489881630459,
+      "learning_rate": 1.7333531964808702e-06,
+      "loss": 0.5055,
+      "step": 1105
+    },
+    {
+      "epoch": 1.6633719825128304,
+      "grad_norm": 0.3208880246840334,
+      "learning_rate": 1.7171645087750255e-06,
+      "loss": 0.4876,
+      "step": 1106
+    },
+    {
+      "epoch": 1.664892605968447,
+      "grad_norm": 0.32500628409148635,
+      "learning_rate": 1.7010466577670633e-06,
+      "loss": 0.5185,
+      "step": 1107
+    },
+    {
+      "epoch": 1.666413229424064,
+      "grad_norm": 0.3151870780999753,
+      "learning_rate": 1.684999739945713e-06,
+      "loss": 0.5224,
+      "step": 1108
+    },
+    {
+      "epoch": 1.6679338528796808,
+      "grad_norm": 0.30364068565385566,
+      "learning_rate": 1.669023851375082e-06,
+      "loss": 0.4843,
+      "step": 1109
+    },
+    {
+      "epoch": 1.6694544763352974,
+      "grad_norm": 0.32280756678101397,
+      "learning_rate": 1.6531190876940478e-06,
+      "loss": 0.5129,
+      "step": 1110
+    },
+    {
+      "epoch": 1.6709750997909143,
+      "grad_norm": 0.3058304028983153,
+      "learning_rate": 1.6372855441157166e-06,
+      "loss": 0.4708,
+      "step": 1111
+    },
+    {
+      "epoch": 1.6724957232465312,
+      "grad_norm": 0.34052506206539823,
+      "learning_rate": 1.6215233154268233e-06,
+      "loss": 0.5272,
+      "step": 1112
+    },
+    {
+      "epoch": 1.6740163467021478,
+      "grad_norm": 0.32203401759522543,
+      "learning_rate": 1.6058324959871942e-06,
+      "loss": 0.4975,
+      "step": 1113
+    },
+    {
+      "epoch": 1.6755369701577647,
+      "grad_norm": 0.3192131968898841,
+      "learning_rate": 1.5902131797291616e-06,
+      "loss": 0.4892,
+      "step": 1114
+    },
+    {
+      "epoch": 1.6770575936133816,
+      "grad_norm": 0.30564977758041467,
+      "learning_rate": 1.5746654601570045e-06,
+      "loss": 0.4891,
+      "step": 1115
+    },
+    {
+      "epoch": 1.6785782170689982,
+      "grad_norm": 0.3174124356131291,
+      "learning_rate": 1.559189430346397e-06,
+      "loss": 0.4948,
+      "step": 1116
+    },
+    {
+      "epoch": 1.6800988405246151,
+      "grad_norm": 0.3094722335828649,
+      "learning_rate": 1.5437851829438422e-06,
+      "loss": 0.488,
+      "step": 1117
+    },
+    {
+      "epoch": 1.681619463980232,
+      "grad_norm": 0.3131260119981047,
+      "learning_rate": 1.5284528101661273e-06,
+      "loss": 0.4664,
+      "step": 1118
+    },
+    {
+      "epoch": 1.6831400874358486,
+      "grad_norm": 0.31529735958006433,
+      "learning_rate": 1.5131924037997535e-06,
+      "loss": 0.4951,
+      "step": 1119
+    },
+    {
+      "epoch": 1.6846607108914655,
+      "grad_norm": 0.30173098838134094,
+      "learning_rate": 1.4980040552004121e-06,
+      "loss": 0.4903,
+      "step": 1120
+    },
+    {
+      "epoch": 1.6861813343470824,
+      "grad_norm": 0.30988057448857725,
+      "learning_rate": 1.4828878552924145e-06,
+      "loss": 0.49,
+      "step": 1121
+    },
+    {
+      "epoch": 1.687701957802699,
+      "grad_norm": 0.3077026541913675,
+      "learning_rate": 1.4678438945681649e-06,
+      "loss": 0.4849,
+      "step": 1122
+    },
+    {
+      "epoch": 1.689222581258316,
+      "grad_norm": 0.31664882915707754,
+      "learning_rate": 1.4528722630876041e-06,
+      "loss": 0.4972,
+      "step": 1123
+    },
+    {
+      "epoch": 1.6907432047139328,
+      "grad_norm": 0.31667781766589775,
+      "learning_rate": 1.4379730504776807e-06,
+      "loss": 0.4907,
+      "step": 1124
+    },
+    {
+      "epoch": 1.6922638281695495,
+      "grad_norm": 0.30777745107859716,
+      "learning_rate": 1.4231463459318138e-06,
+      "loss": 0.4791,
+      "step": 1125
+    },
+    {
+      "epoch": 1.6937844516251663,
+      "grad_norm": 0.3802082342644203,
+      "learning_rate": 1.4083922382093484e-06,
+      "loss": 0.4925,
+      "step": 1126
+    },
+    {
+      "epoch": 1.6953050750807832,
+      "grad_norm": 0.319724112059326,
+      "learning_rate": 1.3937108156350437e-06,
+      "loss": 0.5069,
+      "step": 1127
+    },
+    {
+      "epoch": 1.6968256985363999,
+      "grad_norm": 0.32744646943463035,
+      "learning_rate": 1.379102166098519e-06,
+      "loss": 0.4892,
+      "step": 1128
+    },
+    {
+      "epoch": 1.6983463219920167,
+      "grad_norm": 0.33236823986934483,
+      "learning_rate": 1.3645663770537586e-06,
+      "loss": 0.5106,
+      "step": 1129
+    },
+    {
+      "epoch": 1.6998669454476336,
+      "grad_norm": 0.2988679855910158,
+      "learning_rate": 1.350103535518552e-06,
+      "loss": 0.4792,
+      "step": 1130
+    },
+    {
+      "epoch": 1.7013875689032503,
+      "grad_norm": 0.31594423002452454,
+      "learning_rate": 1.3357137280740085e-06,
+      "loss": 0.5113,
+      "step": 1131
+    },
+    {
+      "epoch": 1.7029081923588671,
+      "grad_norm": 0.30589751459809705,
+      "learning_rate": 1.3213970408640052e-06,
+      "loss": 0.4611,
+      "step": 1132
+    },
+    {
+      "epoch": 1.704428815814484,
+      "grad_norm": 0.32052631965810274,
+      "learning_rate": 1.307153559594711e-06,
+      "loss": 0.5103,
+      "step": 1133
+    },
+    {
+      "epoch": 1.7059494392701007,
+      "grad_norm": 0.3293578066194276,
+      "learning_rate": 1.2929833695340287e-06,
+      "loss": 0.4955,
+      "step": 1134
+    },
+    {
+      "epoch": 1.7074700627257176,
+      "grad_norm": 0.30301511320734814,
+      "learning_rate": 1.2788865555111225e-06,
+      "loss": 0.457,
+      "step": 1135
+    },
+    {
+      "epoch": 1.7089906861813344,
+      "grad_norm": 0.31166180214752437,
+      "learning_rate": 1.264863201915883e-06,
+      "loss": 0.4976,
+      "step": 1136
+    },
+    {
+      "epoch": 1.710511309636951,
+      "grad_norm": 0.3163518521392497,
+      "learning_rate": 1.2509133926984408e-06,
+      "loss": 0.4966,
+      "step": 1137
+    },
+    {
+      "epoch": 1.712031933092568,
+      "grad_norm": 0.3540193363362305,
+      "learning_rate": 1.2370372113686581e-06,
+      "loss": 0.5157,
+      "step": 1138
+    },
+    {
+      "epoch": 1.7135525565481848,
+      "grad_norm": 0.3054649208614663,
+      "learning_rate": 1.2232347409956184e-06,
+      "loss": 0.4674,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7150731800038015,
+      "grad_norm": 0.31124066442535125,
+      "learning_rate": 1.2095060642071457e-06,
+      "loss": 0.4922,
+      "step": 1140
+    },
+    {
+      "epoch": 1.7165938034594184,
+      "grad_norm": 0.31790128208883095,
+      "learning_rate": 1.1958512631893e-06,
+      "loss": 0.4973,
+      "step": 1141
+    },
+    {
+      "epoch": 1.7181144269150352,
+      "grad_norm": 0.3165756985087125,
+      "learning_rate": 1.1822704196858915e-06,
+      "loss": 0.503,
+      "step": 1142
+    },
+    {
+      "epoch": 1.719635050370652,
+      "grad_norm": 0.32017884174072114,
+      "learning_rate": 1.1687636149979783e-06,
+      "loss": 0.4865,
+      "step": 1143
+    },
+    {
+      "epoch": 1.7211556738262688,
+      "grad_norm": 0.31138001106614027,
+      "learning_rate": 1.155330929983396e-06,
+      "loss": 0.5083,
+      "step": 1144
+    },
+    {
+      "epoch": 1.7226762972818856,
+      "grad_norm": 0.3178786337486812,
+      "learning_rate": 1.1419724450562654e-06,
+      "loss": 0.4909,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7241969207375023,
+      "grad_norm": 0.30532938139355004,
+      "learning_rate": 1.1286882401865111e-06,
+      "loss": 0.472,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7257175441931192,
+      "grad_norm": 0.3296827200280989,
+      "learning_rate": 1.1154783948993806e-06,
+      "loss": 0.5244,
+      "step": 1147
+    },
+    {
+      "epoch": 1.727238167648736,
+      "grad_norm": 0.31508066429540393,
+      "learning_rate": 1.1023429882749769e-06,
+      "loss": 0.5102,
+      "step": 1148
+    },
+    {
+      "epoch": 1.7287587911043527,
+      "grad_norm": 0.3196487275495865,
+      "learning_rate": 1.0892820989477799e-06,
+      "loss": 0.4768,
+      "step": 1149
+    },
+    {
+      "epoch": 1.7302794145599696,
+      "grad_norm": 0.30207628232666384,
+      "learning_rate": 1.0762958051061657e-06,
+      "loss": 0.4691,
+      "step": 1150
+    },
+    {
+      "epoch": 1.7318000380155865,
+      "grad_norm": 0.3054462008267432,
+      "learning_rate": 1.0633841844919598e-06,
+      "loss": 0.4862,
+      "step": 1151
+    },
+    {
+      "epoch": 1.7333206614712031,
+      "grad_norm": 0.3078006467585517,
+      "learning_rate": 1.0505473143999502e-06,
+      "loss": 0.4875,
+      "step": 1152
+    },
+    {
+      "epoch": 1.73484128492682,
+      "grad_norm": 0.3053091170945442,
+      "learning_rate": 1.0377852716774458e-06,
+      "loss": 0.5073,
+      "step": 1153
+    },
+    {
+      "epoch": 1.7363619083824369,
+      "grad_norm": 0.3076690595978604,
+      "learning_rate": 1.0250981327237949e-06,
+      "loss": 0.4949,
+      "step": 1154
+    },
+    {
+      "epoch": 1.7378825318380535,
+      "grad_norm": 0.32895341988079463,
+      "learning_rate": 1.012485973489944e-06,
+      "loss": 0.4841,
+      "step": 1155
+    },
+    {
+      "epoch": 1.7394031552936704,
+      "grad_norm": 0.3253480760618867,
+      "learning_rate": 9.99948869477969e-07,
+      "loss": 0.5004,
+      "step": 1156
+    },
+    {
+      "epoch": 1.7409237787492873,
+      "grad_norm": 0.3259517227944411,
+      "learning_rate": 9.874868957406476e-07,
+      "loss": 0.5077,
+      "step": 1157
+    },
+    {
+      "epoch": 1.742444402204904,
+      "grad_norm": 0.30986669451274373,
+      "learning_rate": 9.751001268809732e-07,
+      "loss": 0.4904,
+      "step": 1158
+    },
+    {
+      "epoch": 1.7439650256605208,
+      "grad_norm": 0.3174807363161579,
+      "learning_rate": 9.627886370517466e-07,
+      "loss": 0.5058,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7454856491161377,
+      "grad_norm": 0.3060614656643119,
+      "learning_rate": 9.505524999551009e-07,
+      "loss": 0.4807,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7470062725717543,
+      "grad_norm": 0.3104685490140639,
+      "learning_rate": 9.383917888420816e-07,
+      "loss": 0.4892,
+      "step": 1161
+    },
+    {
+      "epoch": 1.7485268960273712,
+      "grad_norm": 0.31773900649056236,
+      "learning_rate": 9.263065765122013e-07,
+      "loss": 0.4974,
+      "step": 1162
+    },
+    {
+      "epoch": 1.750047519482988,
+      "grad_norm": 0.3218554742602446,
+      "learning_rate": 9.142969353129946e-07,
+      "loss": 0.4846,
+      "step": 1163
+    },
+    {
+      "epoch": 1.7515681429386047,
+      "grad_norm": 0.3060712871920658,
+      "learning_rate": 9.023629371396e-07,
+      "loss": 0.4787,
+      "step": 1164
+    },
+    {
+      "epoch": 1.7530887663942216,
+      "grad_norm": 0.3057039287255903,
+      "learning_rate": 8.905046534343228e-07,
+      "loss": 0.4888,
+      "step": 1165
+    },
+    {
+      "epoch": 1.7546093898498385,
+      "grad_norm": 0.33483369302675053,
+      "learning_rate": 8.787221551862048e-07,
+      "loss": 0.5001,
+      "step": 1166
+    },
+    {
+      "epoch": 1.7561300133054552,
+      "grad_norm": 0.3104906521034538,
+      "learning_rate": 8.670155129306e-07,
+      "loss": 0.496,
+      "step": 1167
+    },
+    {
+      "epoch": 1.757650636761072,
+      "grad_norm": 0.3073575609788421,
+      "learning_rate": 8.55384796748759e-07,
+      "loss": 0.4905,
+      "step": 1168
+    },
+    {
+      "epoch": 1.759171260216689,
+      "grad_norm": 0.3150067268140591,
+      "learning_rate": 8.438300762674001e-07,
+      "loss": 0.4872,
+      "step": 1169
+    },
+    {
+      "epoch": 1.7606918836723056,
+      "grad_norm": 0.30605216319902684,
+      "learning_rate": 8.323514206583039e-07,
+      "loss": 0.4935,
+      "step": 1170
+    },
+    {
+      "epoch": 1.7622125071279224,
+      "grad_norm": 0.3177954531343794,
+      "learning_rate": 8.20948898637881e-07,
+      "loss": 0.4916,
+      "step": 1171
+    },
+    {
+      "epoch": 1.7637331305835393,
+      "grad_norm": 0.31828297184631565,
+      "learning_rate": 8.09622578466782e-07,
+      "loss": 0.4982,
+      "step": 1172
+    },
+    {
+      "epoch": 1.765253754039156,
+      "grad_norm": 0.3112173769721995,
+      "learning_rate": 7.983725279494758e-07,
+      "loss": 0.4857,
+      "step": 1173
+    },
+    {
+      "epoch": 1.7667743774947728,
+      "grad_norm": 0.31585454715709904,
+      "learning_rate": 7.871988144338443e-07,
+      "loss": 0.4784,
+      "step": 1174
+    },
+    {
+      "epoch": 1.7682950009503897,
+      "grad_norm": 0.32127393064790427,
+      "learning_rate": 7.761015048107861e-07,
+      "loss": 0.5075,
+      "step": 1175
+    },
+    {
+      "epoch": 1.7698156244060064,
+      "grad_norm": 0.30794451953754653,
+      "learning_rate": 7.650806655138031e-07,
+      "loss": 0.5038,
+      "step": 1176
+    },
+    {
+      "epoch": 1.7713362478616232,
+      "grad_norm": 0.31181423286855536,
+      "learning_rate": 7.541363625186259e-07,
+      "loss": 0.5005,
+      "step": 1177
+    },
+    {
+      "epoch": 1.7728568713172401,
+      "grad_norm": 0.32245652816729514,
+      "learning_rate": 7.432686613427891e-07,
+      "loss": 0.5095,
+      "step": 1178
+    },
+    {
+      "epoch": 1.7743774947728568,
+      "grad_norm": 0.3242738181594943,
+      "learning_rate": 7.324776270452638e-07,
+      "loss": 0.52,
+      "step": 1179
+    },
+    {
+      "epoch": 1.7758981182284737,
+      "grad_norm": 0.3110243759601029,
+      "learning_rate": 7.217633242260516e-07,
+      "loss": 0.464,
+      "step": 1180
+    },
+    {
+      "epoch": 1.7774187416840905,
+      "grad_norm": 0.3319624462691413,
+      "learning_rate": 7.111258170258142e-07,
+      "loss": 0.5098,
+      "step": 1181
+    },
+    {
+      "epoch": 1.7789393651397072,
+      "grad_norm": 0.3068597543181992,
+      "learning_rate": 7.005651691254696e-07,
+      "loss": 0.4815,
+      "step": 1182
+    },
+    {
+      "epoch": 1.780459988595324,
+      "grad_norm": 0.3157827076249044,
+      "learning_rate": 6.900814437458295e-07,
+      "loss": 0.5024,
+      "step": 1183
+    },
+    {
+      "epoch": 1.781980612050941,
+      "grad_norm": 0.30638021830468265,
+      "learning_rate": 6.79674703647207e-07,
+      "loss": 0.4798,
+      "step": 1184
+    },
+    {
+      "epoch": 1.7835012355065576,
+      "grad_norm": 0.33125316725002335,
+      "learning_rate": 6.693450111290479e-07,
+      "loss": 0.5069,
+      "step": 1185
+    },
+    {
+      "epoch": 1.7850218589621745,
+      "grad_norm": 0.325386774890156,
+      "learning_rate": 6.59092428029562e-07,
+      "loss": 0.5246,
+      "step": 1186
+    },
+    {
+      "epoch": 1.7865424824177913,
+      "grad_norm": 0.2985022483714369,
+      "learning_rate": 6.489170157253382e-07,
+      "loss": 0.4852,
+      "step": 1187
+    },
+    {
+      "epoch": 1.788063105873408,
+      "grad_norm": 0.3102718525215574,
+      "learning_rate": 6.388188351309907e-07,
+      "loss": 0.5006,
+      "step": 1188
+    },
+    {
+      "epoch": 1.7895837293290249,
+      "grad_norm": 0.31889832653696915,
+      "learning_rate": 6.287979466987909e-07,
+      "loss": 0.49,
+      "step": 1189
+    },
+    {
+      "epoch": 1.7911043527846418,
+      "grad_norm": 0.31148436424823517,
+      "learning_rate": 6.188544104183059e-07,
+      "loss": 0.4733,
+      "step": 1190
+    },
+    {
+      "epoch": 1.7926249762402584,
+      "grad_norm": 0.31317825721031184,
+      "learning_rate": 6.089882858160299e-07,
+      "loss": 0.4879,
+      "step": 1191
+    },
+    {
+      "epoch": 1.7941455996958753,
+      "grad_norm": 0.31818518365850107,
+      "learning_rate": 5.991996319550448e-07,
+      "loss": 0.5198,
+      "step": 1192
+    },
+    {
+      "epoch": 1.7956662231514922,
+      "grad_norm": 0.3120540099643981,
+      "learning_rate": 5.894885074346511e-07,
+      "loss": 0.4969,
+      "step": 1193
+    },
+    {
+      "epoch": 1.7971868466071088,
+      "grad_norm": 0.31290703317847507,
+      "learning_rate": 5.79854970390031e-07,
+      "loss": 0.4902,
+      "step": 1194
+    },
+    {
+      "epoch": 1.7987074700627257,
+      "grad_norm": 0.3128686096799172,
+      "learning_rate": 5.702990784918843e-07,
+      "loss": 0.4837,
+      "step": 1195
+    },
+    {
+      "epoch": 1.8002280935183426,
+      "grad_norm": 0.3152764097292228,
+      "learning_rate": 5.608208889460964e-07,
+      "loss": 0.5082,
+      "step": 1196
+    },
+    {
+      "epoch": 1.8017487169739592,
+      "grad_norm": 0.3238711254005141,
+      "learning_rate": 5.514204584933931e-07,
+      "loss": 0.4932,
+      "step": 1197
+    },
+    {
+      "epoch": 1.803269340429576,
+      "grad_norm": 0.32742976059180334,
+      "learning_rate": 5.420978434089952e-07,
+      "loss": 0.5006,
+      "step": 1198
+    },
+    {
+      "epoch": 1.804789963885193,
+      "grad_norm": 0.3120702271210168,
+      "learning_rate": 5.328530995022905e-07,
+      "loss": 0.5077,
+      "step": 1199
+    },
+    {
+      "epoch": 1.8063105873408096,
+      "grad_norm": 0.3176329082136289,
+      "learning_rate": 5.236862821164854e-07,
+      "loss": 0.5182,
+      "step": 1200
+    },
+    {
+      "epoch": 1.8078312107964265,
+      "grad_norm": 0.3014290469656112,
+      "learning_rate": 5.145974461282964e-07,
+      "loss": 0.484,
+      "step": 1201
+    },
+    {
+      "epoch": 1.8093518342520434,
+      "grad_norm": 0.3214449630653542,
+      "learning_rate": 5.055866459475958e-07,
+      "loss": 0.4808,
+      "step": 1202
+    },
+    {
+      "epoch": 1.81087245770766,
+      "grad_norm": 0.32637831086774916,
+      "learning_rate": 4.966539355171074e-07,
+      "loss": 0.5317,
+      "step": 1203
+    },
+    {
+      "epoch": 1.812393081163277,
+      "grad_norm": 0.3111952143588352,
+      "learning_rate": 4.877993683120647e-07,
+      "loss": 0.48,
+      "step": 1204
+    },
+    {
+      "epoch": 1.8139137046188938,
+      "grad_norm": 0.3159439961230018,
+      "learning_rate": 4.790229973399132e-07,
+      "loss": 0.5355,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8154343280745104,
+      "grad_norm": 0.32606917992981577,
+      "learning_rate": 4.7032487513997023e-07,
+      "loss": 0.4854,
+      "step": 1206
+    },
+    {
+      "epoch": 1.8169549515301273,
+      "grad_norm": 0.3099232220873522,
+      "learning_rate": 4.617050537831259e-07,
+      "loss": 0.4744,
+      "step": 1207
+    },
+    {
+      "epoch": 1.8184755749857442,
+      "grad_norm": 0.3189690329444183,
+      "learning_rate": 4.5316358487152045e-07,
+      "loss": 0.5154,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8199961984413608,
+      "grad_norm": 0.3245800528868221,
+      "learning_rate": 4.447005195382475e-07,
+      "loss": 0.486,
+      "step": 1209
+    },
+    {
+      "epoch": 1.8215168218969777,
+      "grad_norm": 0.3265685397280179,
+      "learning_rate": 4.3631590844703997e-07,
+      "loss": 0.5012,
+      "step": 1210
+    },
+    {
+      "epoch": 1.8230374453525946,
+      "grad_norm": 0.3158110920991978,
+      "learning_rate": 4.2800980179196426e-07,
+      "loss": 0.4975,
+      "step": 1211
+    },
+    {
+      "epoch": 1.8245580688082113,
+      "grad_norm": 0.30590044892564083,
+      "learning_rate": 4.197822492971273e-07,
+      "loss": 0.4761,
+      "step": 1212
+    },
+    {
+      "epoch": 1.8260786922638281,
+      "grad_norm": 0.3296107652675287,
+      "learning_rate": 4.1163330021637403e-07,
+      "loss": 0.5024,
+      "step": 1213
+    },
+    {
+      "epoch": 1.827599315719445,
+      "grad_norm": 0.31081520867244933,
+      "learning_rate": 4.0356300333299696e-07,
+      "loss": 0.4861,
+      "step": 1214
+    },
+    {
+      "epoch": 1.8291199391750617,
+      "grad_norm": 0.31577863324052,
+      "learning_rate": 3.9557140695943544e-07,
+      "loss": 0.5023,
+      "step": 1215
+    },
+    {
+      "epoch": 1.8306405626306785,
+      "grad_norm": 0.30718130485703893,
+      "learning_rate": 3.8765855893699576e-07,
+      "loss": 0.4847,
+      "step": 1216
+    },
+    {
+      "epoch": 1.8321611860862954,
+      "grad_norm": 0.3044770096969053,
+      "learning_rate": 3.798245066355609e-07,
+      "loss": 0.4774,
+      "step": 1217
+    },
+    {
+      "epoch": 1.833681809541912,
+      "grad_norm": 0.3233110025620584,
+      "learning_rate": 3.7206929695330764e-07,
+      "loss": 0.516,
+      "step": 1218
+    },
+    {
+      "epoch": 1.835202432997529,
+      "grad_norm": 0.3184528078482147,
+      "learning_rate": 3.643929763164223e-07,
+      "loss": 0.5002,
+      "step": 1219
+    },
+    {
+      "epoch": 1.8367230564531458,
+      "grad_norm": 0.3162950503495864,
+      "learning_rate": 3.567955906788252e-07,
+      "loss": 0.5208,
+      "step": 1220
+    },
+    {
+      "epoch": 1.8382436799087625,
+      "grad_norm": 0.3032249377512437,
+      "learning_rate": 3.492771855219034e-07,
+      "loss": 0.4749,
+      "step": 1221
+    },
+    {
+      "epoch": 1.8397643033643793,
+      "grad_norm": 0.32645828201108573,
+      "learning_rate": 3.418378058542228e-07,
+      "loss": 0.5212,
+      "step": 1222
+    },
+    {
+      "epoch": 1.8412849268199962,
+      "grad_norm": 0.31381954625492664,
+      "learning_rate": 3.344774962112713e-07,
+      "loss": 0.4697,
+      "step": 1223
+    },
+    {
+      "epoch": 1.8428055502756129,
+      "grad_norm": 0.3163864792118917,
+      "learning_rate": 3.2719630065518326e-07,
+      "loss": 0.5161,
+      "step": 1224
+    },
+    {
+      "epoch": 1.8443261737312298,
+      "grad_norm": 0.29974584672007193,
+      "learning_rate": 3.199942627744915e-07,
+      "loss": 0.4769,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8458467971868466,
+      "grad_norm": 0.32581659125357904,
+      "learning_rate": 3.1287142568384115e-07,
+      "loss": 0.485,
+      "step": 1226
+    },
+    {
+      "epoch": 1.8473674206424633,
+      "grad_norm": 0.32074295034449724,
+      "learning_rate": 3.058278320237554e-07,
+      "loss": 0.4896,
+      "step": 1227
+    },
+    {
+      "epoch": 1.8488880440980802,
+      "grad_norm": 0.30909473328904485,
+      "learning_rate": 2.9886352396036584e-07,
+      "loss": 0.4768,
+      "step": 1228
+    },
+    {
+      "epoch": 1.850408667553697,
+      "grad_norm": 0.32336521053629386,
+      "learning_rate": 2.919785431851674e-07,
+      "loss": 0.4999,
+      "step": 1229
+    },
+    {
+      "epoch": 1.8519292910093137,
+      "grad_norm": 0.305906498433603,
+      "learning_rate": 2.851729309147642e-07,
+      "loss": 0.4998,
+      "step": 1230
+    },
+    {
+      "epoch": 1.8534499144649306,
+      "grad_norm": 0.30991453199000357,
+      "learning_rate": 2.7844672789062724e-07,
+      "loss": 0.5011,
+      "step": 1231
+    },
+    {
+      "epoch": 1.8549705379205474,
+      "grad_norm": 0.314943371794473,
+      "learning_rate": 2.717999743788416e-07,
+      "loss": 0.5096,
+      "step": 1232
+    },
+    {
+      "epoch": 1.856491161376164,
+      "grad_norm": 0.3150834801717786,
+      "learning_rate": 2.652327101698796e-07,
+      "loss": 0.504,
+      "step": 1233
+    },
+    {
+      "epoch": 1.858011784831781,
+      "grad_norm": 0.32620085243083274,
+      "learning_rate": 2.587449745783506e-07,
+      "loss": 0.5239,
+      "step": 1234
+    },
+    {
+      "epoch": 1.8595324082873979,
+      "grad_norm": 0.31548675603667015,
+      "learning_rate": 2.5233680644276673e-07,
+      "loss": 0.4889,
+      "step": 1235
+    },
+    {
+      "epoch": 1.8610530317430145,
+      "grad_norm": 0.32956663567551076,
+      "learning_rate": 2.4600824412532154e-07,
+      "loss": 0.5108,
+      "step": 1236
+    },
+    {
+      "epoch": 1.8625736551986314,
+      "grad_norm": 0.29811803143834026,
+      "learning_rate": 2.3975932551164486e-07,
+      "loss": 0.4616,
+      "step": 1237
+    },
+    {
+      "epoch": 1.8640942786542483,
+      "grad_norm": 0.3118381439138631,
+      "learning_rate": 2.335900880105892e-07,
+      "loss": 0.5051,
+      "step": 1238
+    },
+    {
+      "epoch": 1.865614902109865,
+      "grad_norm": 0.30830068205472616,
+      "learning_rate": 2.27500568553992e-07,
+      "loss": 0.5012,
+      "step": 1239
+    },
+    {
+      "epoch": 1.867135525565482,
+      "grad_norm": 0.304947912980323,
+      "learning_rate": 2.2149080359646954e-07,
+      "loss": 0.4583,
+      "step": 1240
+    },
+    {
+      "epoch": 1.8686561490210987,
+      "grad_norm": 0.3114996653388056,
+      "learning_rate": 2.1556082911518989e-07,
+      "loss": 0.4898,
+      "step": 1241
+    },
+    {
+      "epoch": 1.8701767724767153,
+      "grad_norm": 0.30177416022258724,
+      "learning_rate": 2.0971068060965908e-07,
+      "loss": 0.4965,
+      "step": 1242
+    },
+    {
+      "epoch": 1.8716973959323324,
+      "grad_norm": 0.3309823897680051,
+      "learning_rate": 2.0394039310150775e-07,
+      "loss": 0.5185,
+      "step": 1243
+    },
+    {
+      "epoch": 1.873218019387949,
+      "grad_norm": 0.32115239410578134,
+      "learning_rate": 1.9825000113428182e-07,
+      "loss": 0.5228,
+      "step": 1244
+    },
+    {
+      "epoch": 1.8747386428435657,
+      "grad_norm": 0.3075640176251186,
+      "learning_rate": 1.9263953877323952e-07,
+      "loss": 0.4882,
+      "step": 1245
+    },
+    {
+      "epoch": 1.8762592662991828,
+      "grad_norm": 0.31724674747730425,
+      "learning_rate": 1.8710903960514225e-07,
+      "loss": 0.5207,
+      "step": 1246
+    },
+    {
+      "epoch": 1.8777798897547995,
+      "grad_norm": 0.3164911565117331,
+      "learning_rate": 1.8165853673805754e-07,
+      "loss": 0.5032,
+      "step": 1247
+    },
+    {
+      "epoch": 1.8793005132104161,
+      "grad_norm": 0.30881540155384896,
+      "learning_rate": 1.7628806280115286e-07,
+      "loss": 0.5033,
+      "step": 1248
+    },
+    {
+      "epoch": 1.8808211366660332,
+      "grad_norm": 0.31442427065508066,
+      "learning_rate": 1.7099764994451522e-07,
+      "loss": 0.5076,
+      "step": 1249
+    },
+    {
+      "epoch": 1.8823417601216499,
+      "grad_norm": 0.31278434986007647,
+      "learning_rate": 1.65787329838945e-07,
+      "loss": 0.4989,
+      "step": 1250
+    },
+    {
+      "epoch": 1.8838623835772665,
+      "grad_norm": 0.3025781565220422,
+      "learning_rate": 1.60657133675771e-07,
+      "loss": 0.5014,
+      "step": 1251
+    },
+    {
+      "epoch": 1.8853830070328836,
+      "grad_norm": 0.30236403268840834,
+      "learning_rate": 1.556070921666637e-07,
+      "loss": 0.4653,
+      "step": 1252
+    },
+    {
+      "epoch": 1.8869036304885003,
+      "grad_norm": 0.32144829401638425,
+      "learning_rate": 1.5063723554345349e-07,
+      "loss": 0.5022,
+      "step": 1253
+    },
+    {
+      "epoch": 1.888424253944117,
+      "grad_norm": 0.3081811486598015,
+      "learning_rate": 1.4574759355794846e-07,
+      "loss": 0.468,
+      "step": 1254
+    },
+    {
+      "epoch": 1.889944877399734,
+      "grad_norm": 0.3110080086508686,
+      "learning_rate": 1.4093819548175095e-07,
+      "loss": 0.4821,
+      "step": 1255
+    },
+    {
+      "epoch": 1.8914655008553507,
+      "grad_norm": 0.30999522834801807,
+      "learning_rate": 1.3620907010609206e-07,
+      "loss": 0.475,
+      "step": 1256
+    },
+    {
+      "epoch": 1.8929861243109674,
+      "grad_norm": 0.3244026681645176,
+      "learning_rate": 1.3156024574165258e-07,
+      "loss": 0.4967,
+      "step": 1257
+    },
+    {
+      "epoch": 1.8945067477665845,
+      "grad_norm": 0.30596722710536933,
+      "learning_rate": 1.2699175021839463e-07,
+      "loss": 0.4794,
+      "step": 1258
+    },
+    {
+      "epoch": 1.896027371222201,
+      "grad_norm": 0.33281008993735,
+      "learning_rate": 1.2250361088539598e-07,
+      "loss": 0.5202,
+      "step": 1259
+    },
+    {
+      "epoch": 1.8975479946778178,
+      "grad_norm": 0.3092714395241654,
+      "learning_rate": 1.1809585461068329e-07,
+      "loss": 0.4897,
+      "step": 1260
+    },
+    {
+      "epoch": 1.8990686181334349,
+      "grad_norm": 0.30328721741220116,
+      "learning_rate": 1.1376850778108004e-07,
+      "loss": 0.4734,
+      "step": 1261
+    },
+    {
+      "epoch": 1.9005892415890515,
+      "grad_norm": 0.31671920046718893,
+      "learning_rate": 1.0952159630203802e-07,
+      "loss": 0.495,
+      "step": 1262
+    },
+    {
+      "epoch": 1.9021098650446682,
+      "grad_norm": 0.301311429340991,
+      "learning_rate": 1.0535514559748699e-07,
+      "loss": 0.4879,
+      "step": 1263
+    },
+    {
+      "epoch": 1.9036304885002853,
+      "grad_norm": 0.31829608262735665,
+      "learning_rate": 1.0126918060968266e-07,
+      "loss": 0.5054,
+      "step": 1264
+    },
+    {
+      "epoch": 1.905151111955902,
+      "grad_norm": 0.3087155473121018,
+      "learning_rate": 9.726372579905783e-08,
+      "loss": 0.4972,
+      "step": 1265
+    },
+    {
+      "epoch": 1.9066717354115186,
+      "grad_norm": 0.3080173664665244,
+      "learning_rate": 9.33388051440734e-08,
+      "loss": 0.5339,
+      "step": 1266
+    },
+    {
+      "epoch": 1.9081923588671357,
+      "grad_norm": 0.3282773688288346,
+      "learning_rate": 8.949444214107697e-08,
+      "loss": 0.4775,
+      "step": 1267
+    },
+    {
+      "epoch": 1.9097129823227523,
+      "grad_norm": 0.33035543404288714,
+      "learning_rate": 8.573065980416298e-08,
+      "loss": 0.5071,
+      "step": 1268
+    },
+    {
+      "epoch": 1.911233605778369,
+      "grad_norm": 0.3215706061649872,
+      "learning_rate": 8.204748066503122e-08,
+      "loss": 0.5094,
+      "step": 1269
+    },
+    {
+      "epoch": 1.912754229233986,
+      "grad_norm": 0.2988811222168615,
+      "learning_rate": 7.844492677285754e-08,
+      "loss": 0.4984,
+      "step": 1270
+    },
+    {
+      "epoch": 1.9142748526896027,
+      "grad_norm": 0.31431504366013957,
+      "learning_rate": 7.492301969415391e-08,
+      "loss": 0.5072,
+      "step": 1271
+    },
+    {
+      "epoch": 1.9157954761452194,
+      "grad_norm": 0.3079234038112652,
+      "learning_rate": 7.148178051264804e-08,
+      "loss": 0.4883,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9173160996008365,
+      "grad_norm": 0.30748269412345614,
+      "learning_rate": 6.812122982915408e-08,
+      "loss": 0.5152,
+      "step": 1273
+    },
+    {
+      "epoch": 1.9188367230564531,
+      "grad_norm": 0.3111607368191631,
+      "learning_rate": 6.484138776144467e-08,
+      "loss": 0.4992,
+      "step": 1274
+    },
+    {
+      "epoch": 1.9203573465120698,
+      "grad_norm": 0.3130854739102124,
+      "learning_rate": 6.164227394413516e-08,
+      "loss": 0.4947,
+      "step": 1275
+    },
+    {
+      "epoch": 1.921877969967687,
+      "grad_norm": 0.29735483922959655,
+      "learning_rate": 5.852390752856624e-08,
+      "loss": 0.4717,
+      "step": 1276
+    },
+    {
+      "epoch": 1.9233985934233035,
+      "grad_norm": 0.30465543049913124,
+      "learning_rate": 5.548630718268813e-08,
+      "loss": 0.4977,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9249192168789202,
+      "grad_norm": 0.3133152057012344,
+      "learning_rate": 5.252949109094924e-08,
+      "loss": 0.4839,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9264398403345373,
+      "grad_norm": 0.32301869827083196,
+      "learning_rate": 4.965347695418483e-08,
+      "loss": 0.4755,
+      "step": 1279
+    },
+    {
+      "epoch": 1.927960463790154,
+      "grad_norm": 0.30143456487467507,
+      "learning_rate": 4.685828198951479e-08,
+      "loss": 0.4901,
+      "step": 1280
+    },
+    {
+      "epoch": 1.9294810872457706,
+      "grad_norm": 0.3198896489356579,
+      "learning_rate": 4.4143922930239736e-08,
+      "loss": 0.5014,
+      "step": 1281
+    },
+    {
+      "epoch": 1.9310017107013877,
+      "grad_norm": 0.29774733042477136,
+      "learning_rate": 4.1510416025737274e-08,
+      "loss": 0.4598,
+      "step": 1282
+    },
+    {
+      "epoch": 1.9325223341570044,
+      "grad_norm": 0.3117201437762231,
+      "learning_rate": 3.895777704137174e-08,
+      "loss": 0.4718,
+      "step": 1283
+    },
+    {
+      "epoch": 1.934042957612621,
+      "grad_norm": 0.30970105466061604,
+      "learning_rate": 3.648602125839037e-08,
+      "loss": 0.4925,
+      "step": 1284
+    },
+    {
+      "epoch": 1.9355635810682381,
+      "grad_norm": 0.3032281003832038,
+      "learning_rate": 3.409516347384058e-08,
+      "loss": 0.4876,
+      "step": 1285
+    },
+    {
+      "epoch": 1.9370842045238548,
+      "grad_norm": 0.29947248473886334,
+      "learning_rate": 3.178521800047818e-08,
+      "loss": 0.4652,
+      "step": 1286
+    },
+    {
+      "epoch": 1.9386048279794714,
+      "grad_norm": 0.3101511675568205,
+      "learning_rate": 2.9556198666678676e-08,
+      "loss": 0.487,
+      "step": 1287
+    },
+    {
+      "epoch": 1.9401254514350885,
+      "grad_norm": 0.3031822788379352,
+      "learning_rate": 2.7408118816360456e-08,
+      "loss": 0.4906,
+      "step": 1288
+    },
+    {
+      "epoch": 1.9416460748907052,
+      "grad_norm": 0.3022071103062291,
+      "learning_rate": 2.5340991308897615e-08,
+      "loss": 0.4749,
+      "step": 1289
+    },
+    {
+      "epoch": 1.9431666983463218,
+      "grad_norm": 0.3376492826020699,
+      "learning_rate": 2.3354828519049215e-08,
+      "loss": 0.5193,
+      "step": 1290
+    },
+    {
+      "epoch": 1.944687321801939,
+      "grad_norm": 0.3049763317749277,
+      "learning_rate": 2.144964233688558e-08,
+      "loss": 0.492,
+      "step": 1291
+    },
+    {
+      "epoch": 1.9462079452575556,
+      "grad_norm": 0.31915116281549216,
+      "learning_rate": 1.9625444167708558e-08,
+      "loss": 0.5077,
+      "step": 1292
+    },
+    {
+      "epoch": 1.9477285687131722,
+      "grad_norm": 0.29789807615910413,
+      "learning_rate": 1.7882244931992866e-08,
+      "loss": 0.4802,
+      "step": 1293
+    },
+    {
+      "epoch": 1.9492491921687893,
+      "grad_norm": 0.29936729470728324,
+      "learning_rate": 1.622005506531838e-08,
+      "loss": 0.4683,
+      "step": 1294
+    },
+    {
+      "epoch": 1.950769815624406,
+      "grad_norm": 0.30557270295233097,
+      "learning_rate": 1.4638884518302449e-08,
+      "loss": 0.493,
+      "step": 1295
+    },
+    {
+      "epoch": 1.9522904390800226,
+      "grad_norm": 0.3152644678698385,
+      "learning_rate": 1.313874275654573e-08,
+      "loss": 0.4872,
+      "step": 1296
+    },
+    {
+      "epoch": 1.9538110625356397,
+      "grad_norm": 0.32155409817146613,
+      "learning_rate": 1.1719638760576535e-08,
+      "loss": 0.4898,
+      "step": 1297
+    },
+    {
+      "epoch": 1.9553316859912564,
+      "grad_norm": 0.322441859921911,
+      "learning_rate": 1.0381581025792154e-08,
+      "loss": 0.5123,
+      "step": 1298
+    },
+    {
+      "epoch": 1.9568523094468733,
+      "grad_norm": 0.3075162182309576,
+      "learning_rate": 9.124577562409214e-09,
+      "loss": 0.5075,
+      "step": 1299
+    },
+    {
+      "epoch": 1.9583729329024901,
+      "grad_norm": 0.31214568646689916,
+      "learning_rate": 7.948635895418555e-09,
+      "loss": 0.5012,
+      "step": 1300
+    },
+    {
+      "epoch": 1.9598935563581068,
+      "grad_norm": 0.3160685661400899,
+      "learning_rate": 6.853763064538582e-09,
+      "loss": 0.5009,
+      "step": 1301
+    },
+    {
+      "epoch": 1.9614141798137237,
+      "grad_norm": 0.31359686192724745,
+      "learning_rate": 5.8399656241716505e-09,
+      "loss": 0.4962,
+      "step": 1302
+    },
+    {
+      "epoch": 1.9629348032693406,
+      "grad_norm": 0.31993025547008763,
+      "learning_rate": 4.907249643367956e-09,
+      "loss": 0.4837,
+      "step": 1303
+    },
+    {
+      "epoch": 1.9644554267249572,
+      "grad_norm": 0.30154655369767214,
+      "learning_rate": 4.055620705787927e-09,
+      "loss": 0.4907,
+      "step": 1304
+    },
+    {
+      "epoch": 1.965976050180574,
+      "grad_norm": 0.3082339737443702,
+      "learning_rate": 3.2850839096646157e-09,
+      "loss": 0.49,
+      "step": 1305
+    },
+    {
+      "epoch": 1.967496673636191,
+      "grad_norm": 0.31262848669610493,
+      "learning_rate": 2.5956438677796293e-09,
+      "loss": 0.5033,
+      "step": 1306
+    },
+    {
+      "epoch": 1.9690172970918076,
+      "grad_norm": 0.3078427537941908,
+      "learning_rate": 1.9873047074345453e-09,
+      "loss": 0.4793,
+      "step": 1307
+    },
+    {
+      "epoch": 1.9705379205474245,
+      "grad_norm": 0.3074339790589038,
+      "learning_rate": 1.4600700704193227e-09,
+      "loss": 0.4843,
+      "step": 1308
+    },
+    {
+      "epoch": 1.9720585440030414,
+      "grad_norm": 0.31784494867260554,
+      "learning_rate": 1.0139431130017695e-09,
+      "loss": 0.4984,
+      "step": 1309
+    },
+    {
+      "epoch": 1.973579167458658,
+      "grad_norm": 0.32037784960225285,
+      "learning_rate": 6.489265058974569e-10,
+      "loss": 0.5143,
+      "step": 1310
+    },
+    {
+      "epoch": 1.975099790914275,
+      "grad_norm": 0.3115474055785006,
+      "learning_rate": 3.6502243426069227e-10,
+      "loss": 0.4958,
+      "step": 1311
+    },
+    {
+      "epoch": 1.9766204143698918,
+      "grad_norm": 0.30740455562554947,
+      "learning_rate": 1.622325976709804e-10,
+      "loss": 0.4825,
+      "step": 1312
+    },
+    {
+      "epoch": 1.9781410378255084,
+      "grad_norm": 0.3192120605150713,
+      "learning_rate": 4.055821011798022e-11,
+      "loss": 0.5028,
+      "step": 1313
+    },
+    {
+      "epoch": 1.9796616612811253,
+      "grad_norm": 0.3128442843272166,
+      "learning_rate": 0.0,
+      "loss": 0.4886,
+      "step": 1314
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1314,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 329,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4190576087400448.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}