diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9492 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.995008319467554,
+  "eval_steps": 500,
+  "global_step": 1350,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0022185246810870773,
+      "grad_norm": 0.06655312329530716,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.691,
+      "step": 1
+    },
+    {
+      "epoch": 0.004437049362174155,
+      "grad_norm": 0.0848173201084137,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.5981,
+      "step": 2
+    },
+    {
+      "epoch": 0.0066555740432612314,
+      "grad_norm": 0.0829235315322876,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.6839,
+      "step": 3
+    },
+    {
+      "epoch": 0.00887409872434831,
+      "grad_norm": 0.0681910365819931,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.5322,
+      "step": 4
+    },
+    {
+      "epoch": 0.011092623405435386,
+      "grad_norm": 0.05567406862974167,
+      "learning_rate": 1.111111111111111e-05,
+      "loss": 0.5908,
+      "step": 5
+    },
+    {
+      "epoch": 0.013311148086522463,
+      "grad_norm": 0.07490547746419907,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5486,
+      "step": 6
+    },
+    {
+      "epoch": 0.01552967276760954,
+      "grad_norm": 0.08135199546813965,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.5219,
+      "step": 7
+    },
+    {
+      "epoch": 0.01774819744869662,
+      "grad_norm": 0.07295224815607071,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.7032,
+      "step": 8
+    },
+    {
+      "epoch": 0.019966722129783693,
+      "grad_norm": 0.12396921962499619,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.5505,
+      "step": 9
+    },
+    {
+      "epoch": 0.022185246810870772,
+      "grad_norm": 0.05305679887533188,
+      "learning_rate": 2.222222222222222e-05,
+      "loss": 0.748,
+      "step": 10
+    },
+    {
+      "epoch": 0.024403771491957847,
+      "grad_norm": 0.2674309015274048,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 0.9136,
+      "step": 11
+    },
+    {
+      "epoch": 0.026622296173044926,
+      "grad_norm": 0.07705745100975037,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5435,
+      "step": 12
+    },
+    {
+      "epoch": 0.028840820854132,
+      "grad_norm": 0.09026669710874557,
+      "learning_rate": 2.8888888888888885e-05,
+      "loss": 0.68,
+      "step": 13
+    },
+    {
+      "epoch": 0.03105934553521908,
+      "grad_norm": 0.06608552485704422,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 0.8521,
+      "step": 14
+    },
+    {
+      "epoch": 0.033277870216306155,
+      "grad_norm": 0.09559015929698944,
+      "learning_rate": 3.333333333333333e-05,
+      "loss": 0.6049,
+      "step": 15
+    },
+    {
+      "epoch": 0.03549639489739324,
+      "grad_norm": 0.0808415338397026,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 0.6825,
+      "step": 16
+    },
+    {
+      "epoch": 0.03771491957848031,
+      "grad_norm": 0.10304183512926102,
+      "learning_rate": 3.777777777777777e-05,
+      "loss": 0.6709,
+      "step": 17
+    },
+    {
+      "epoch": 0.03993344425956739,
+      "grad_norm": 0.07773031294345856,
+      "learning_rate": 3.9999999999999996e-05,
+      "loss": 0.5693,
+      "step": 18
+    },
+    {
+      "epoch": 0.04215196894065446,
+      "grad_norm": 0.0743400976061821,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.5603,
+      "step": 19
+    },
+    {
+      "epoch": 0.044370493621741544,
+      "grad_norm": 0.1096852645277977,
+      "learning_rate": 4.444444444444444e-05,
+      "loss": 0.5969,
+      "step": 20
+    },
+    {
+      "epoch": 0.04658901830282862,
+      "grad_norm": 0.10413377732038498,
+      "learning_rate": 4.6666666666666665e-05,
+      "loss": 0.6698,
+      "step": 21
+    },
+    {
+      "epoch": 0.048807542983915694,
+      "grad_norm": 0.08424200862646103,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.6524,
+      "step": 22
+    },
+    {
+      "epoch": 0.051026067665002776,
+      "grad_norm": 0.1243722140789032,
+      "learning_rate": 5.111111111111111e-05,
+      "loss": 0.6866,
+      "step": 23
+    },
+    {
+      "epoch": 0.05324459234608985,
+      "grad_norm": 0.06986111402511597,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.5264,
+      "step": 24
+    },
+    {
+      "epoch": 0.05546311702717693,
+      "grad_norm": 0.16437886655330658,
+      "learning_rate": 5.5555555555555545e-05,
+      "loss": 0.7156,
+      "step": 25
+    },
+    {
+      "epoch": 0.057681641708264,
+      "grad_norm": 0.16125601530075073,
+      "learning_rate": 5.777777777777777e-05,
+      "loss": 0.6229,
+      "step": 26
+    },
+    {
+      "epoch": 0.059900166389351084,
+      "grad_norm": 0.10645366460084915,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 0.7767,
+      "step": 27
+    },
+    {
+      "epoch": 0.06211869107043816,
+      "grad_norm": 0.13458511233329773,
+      "learning_rate": 6.222222222222222e-05,
+      "loss": 0.5092,
+      "step": 28
+    },
+    {
+      "epoch": 0.06433721575152523,
+      "grad_norm": 0.14062701165676117,
+      "learning_rate": 6.444444444444444e-05,
+      "loss": 0.6853,
+      "step": 29
+    },
+    {
+      "epoch": 0.06655574043261231,
+      "grad_norm": 0.12555214762687683,
+      "learning_rate": 6.666666666666666e-05,
+      "loss": 0.5628,
+      "step": 30
+    },
+    {
+      "epoch": 0.06877426511369938,
+      "grad_norm": 0.10760708898305893,
+      "learning_rate": 6.888888888888888e-05,
+      "loss": 0.5968,
+      "step": 31
+    },
+    {
+      "epoch": 0.07099278979478647,
+      "grad_norm": 0.14426910877227783,
+      "learning_rate": 7.11111111111111e-05,
+      "loss": 0.7116,
+      "step": 32
+    },
+    {
+      "epoch": 0.07321131447587355,
+      "grad_norm": 0.2234889268875122,
+      "learning_rate": 7.333333333333332e-05,
+      "loss": 0.64,
+      "step": 33
+    },
+    {
+      "epoch": 0.07542983915696062,
+      "grad_norm": 0.15983794629573822,
+      "learning_rate": 7.555555555555554e-05,
+      "loss": 0.6568,
+      "step": 34
+    },
+    {
+      "epoch": 0.0776483638380477,
+      "grad_norm": 0.22704409062862396,
+      "learning_rate": 7.777777777777777e-05,
+      "loss": 0.5403,
+      "step": 35
+    },
+    {
+      "epoch": 0.07986688851913477,
+      "grad_norm": 0.17770637571811676,
+      "learning_rate": 7.999999999999999e-05,
+      "loss": 0.6068,
+      "step": 36
+    },
+    {
+      "epoch": 0.08208541320022185,
+      "grad_norm": 0.16450530290603638,
+      "learning_rate": 8.222222222222222e-05,
+      "loss": 0.6893,
+      "step": 37
+    },
+    {
+      "epoch": 0.08430393788130892,
+      "grad_norm": 0.19376608729362488,
+      "learning_rate": 8.444444444444444e-05,
+      "loss": 0.685,
+      "step": 38
+    },
+    {
+      "epoch": 0.08652246256239601,
+      "grad_norm": 0.318289577960968,
+      "learning_rate": 8.666666666666665e-05,
+      "loss": 0.5627,
+      "step": 39
+    },
+    {
+      "epoch": 0.08874098724348309,
+      "grad_norm": 0.20991887152194977,
+      "learning_rate": 8.888888888888888e-05,
+      "loss": 0.6492,
+      "step": 40
+    },
+    {
+      "epoch": 0.09095951192457016,
+      "grad_norm": 0.1718834936618805,
+      "learning_rate": 9.11111111111111e-05,
+      "loss": 0.5709,
+      "step": 41
+    },
+    {
+      "epoch": 0.09317803660565724,
+      "grad_norm": 0.24201668798923492,
+      "learning_rate": 9.333333333333333e-05,
+      "loss": 0.5678,
+      "step": 42
+    },
+    {
+      "epoch": 0.09539656128674431,
+      "grad_norm": 0.20458950102329254,
+      "learning_rate": 9.555555555555555e-05,
+      "loss": 0.6549,
+      "step": 43
+    },
+    {
+      "epoch": 0.09761508596783139,
+      "grad_norm": 0.17686423659324646,
+      "learning_rate": 9.777777777777778e-05,
+      "loss": 0.5986,
+      "step": 44
+    },
+    {
+      "epoch": 0.09983361064891846,
+      "grad_norm": 0.33893463015556335,
+      "learning_rate": 9.999999999999999e-05,
+      "loss": 0.6574,
+      "step": 45
+    },
+    {
+      "epoch": 0.10205213533000555,
+      "grad_norm": 0.6172243356704712,
+      "learning_rate": 0.00010222222222222222,
+      "loss": 0.6031,
+      "step": 46
+    },
+    {
+      "epoch": 0.10427066001109263,
+      "grad_norm": 0.20055758953094482,
+      "learning_rate": 0.00010444444444444444,
+      "loss": 0.6988,
+      "step": 47
+    },
+    {
+      "epoch": 0.1064891846921797,
+      "grad_norm": 0.2371675670146942,
+      "learning_rate": 0.00010666666666666667,
+      "loss": 0.5508,
+      "step": 48
+    },
+    {
+      "epoch": 0.10870770937326678,
+      "grad_norm": 0.2302960753440857,
+      "learning_rate": 0.00010888888888888889,
+      "loss": 0.7228,
+      "step": 49
+    },
+    {
+      "epoch": 0.11092623405435385,
+      "grad_norm": 0.2170049101114273,
+      "learning_rate": 0.00011111111111111109,
+      "loss": 0.6357,
+      "step": 50
+    },
+    {
+      "epoch": 0.11314475873544093,
+      "grad_norm": 0.2509506642818451,
+      "learning_rate": 0.00011333333333333331,
+      "loss": 0.6594,
+      "step": 51
+    },
+    {
+      "epoch": 0.115363283416528,
+      "grad_norm": 0.15898850560188293,
+      "learning_rate": 0.00011555555555555554,
+      "loss": 0.692,
+      "step": 52
+    },
+    {
+      "epoch": 0.11758180809761509,
+      "grad_norm": 0.21606065332889557,
+      "learning_rate": 0.00011777777777777776,
+      "loss": 0.7252,
+      "step": 53
+    },
+    {
+      "epoch": 0.11980033277870217,
+      "grad_norm": 0.20650148391723633,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 0.755,
+      "step": 54
+    },
+    {
+      "epoch": 0.12201885745978924,
+      "grad_norm": 0.13989175856113434,
+      "learning_rate": 0.00012222222222222221,
+      "loss": 0.4785,
+      "step": 55
+    },
+    {
+      "epoch": 0.12423738214087632,
+      "grad_norm": 0.18587812781333923,
+      "learning_rate": 0.00012444444444444444,
+      "loss": 0.5323,
+      "step": 56
+    },
+    {
+      "epoch": 0.1264559068219634,
+      "grad_norm": 0.20608311891555786,
+      "learning_rate": 0.00012666666666666666,
+      "loss": 0.5244,
+      "step": 57
+    },
+    {
+      "epoch": 0.12867443150305047,
+      "grad_norm": 0.24105136096477509,
+      "learning_rate": 0.0001288888888888889,
+      "loss": 0.6089,
+      "step": 58
+    },
+    {
+      "epoch": 0.13089295618413754,
+      "grad_norm": 0.21293887495994568,
+      "learning_rate": 0.00013111111111111111,
+      "loss": 0.5534,
+      "step": 59
+    },
+    {
+      "epoch": 0.13311148086522462,
+      "grad_norm": 0.14354591071605682,
+      "learning_rate": 0.0001333333333333333,
+      "loss": 0.4224,
+      "step": 60
+    },
+    {
+      "epoch": 0.1353300055463117,
+      "grad_norm": 0.18001911044120789,
+      "learning_rate": 0.00013555555555555554,
+      "loss": 0.6599,
+      "step": 61
+    },
+    {
+      "epoch": 0.13754853022739877,
+      "grad_norm": 0.1596135050058365,
+      "learning_rate": 0.00013777777777777776,
+      "loss": 0.5842,
+      "step": 62
+    },
+    {
+      "epoch": 0.13976705490848584,
+      "grad_norm": 0.14428000152111053,
+      "learning_rate": 0.00014,
+      "loss": 0.5899,
+      "step": 63
+    },
+    {
+      "epoch": 0.14198557958957295,
+      "grad_norm": 0.1802503913640976,
+      "learning_rate": 0.0001422222222222222,
+      "loss": 0.5461,
+      "step": 64
+    },
+    {
+      "epoch": 0.14420410427066002,
+      "grad_norm": 0.20101633667945862,
+      "learning_rate": 0.0001444444444444444,
+      "loss": 0.6596,
+      "step": 65
+    },
+    {
+      "epoch": 0.1464226289517471,
+      "grad_norm": 0.1304934024810791,
+      "learning_rate": 0.00014666666666666664,
+      "loss": 0.5568,
+      "step": 66
+    },
+    {
+      "epoch": 0.14864115363283417,
+      "grad_norm": 0.17323292791843414,
+      "learning_rate": 0.00014888888888888886,
+      "loss": 0.6239,
+      "step": 67
+    },
+    {
+      "epoch": 0.15085967831392125,
+      "grad_norm": 0.11214841157197952,
+      "learning_rate": 0.00015111111111111109,
+      "loss": 0.4209,
+      "step": 68
+    },
+    {
+      "epoch": 0.15307820299500832,
+      "grad_norm": 0.14260365068912506,
+      "learning_rate": 0.0001533333333333333,
+      "loss": 0.6109,
+      "step": 69
+    },
+    {
+      "epoch": 0.1552967276760954,
+      "grad_norm": 0.2211947739124298,
+      "learning_rate": 0.00015555555555555554,
+      "loss": 0.5866,
+      "step": 70
+    },
+    {
+      "epoch": 0.15751525235718247,
+      "grad_norm": 0.12835100293159485,
+      "learning_rate": 0.00015777777777777776,
+      "loss": 0.5529,
+      "step": 71
+    },
+    {
+      "epoch": 0.15973377703826955,
+      "grad_norm": 0.18953093886375427,
+      "learning_rate": 0.00015999999999999999,
+      "loss": 0.5406,
+      "step": 72
+    },
+    {
+      "epoch": 0.16195230171935662,
+      "grad_norm": 0.14358682930469513,
+      "learning_rate": 0.0001622222222222222,
+      "loss": 0.5417,
+      "step": 73
+    },
+    {
+      "epoch": 0.1641708264004437,
+      "grad_norm": 0.13438697159290314,
+      "learning_rate": 0.00016444444444444444,
+      "loss": 0.5565,
+      "step": 74
+    },
+    {
+      "epoch": 0.16638935108153077,
+      "grad_norm": 0.18689267337322235,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 0.5589,
+      "step": 75
+    },
+    {
+      "epoch": 0.16860787576261785,
+      "grad_norm": 0.1448279321193695,
+      "learning_rate": 0.00016888888888888889,
+      "loss": 0.8148,
+      "step": 76
+    },
+    {
+      "epoch": 0.17082640044370492,
+      "grad_norm": 0.1968618482351303,
+      "learning_rate": 0.0001711111111111111,
+      "loss": 0.6502,
+      "step": 77
+    },
+    {
+      "epoch": 0.17304492512479203,
+      "grad_norm": 0.09278809279203415,
+      "learning_rate": 0.0001733333333333333,
+      "loss": 0.5163,
+      "step": 78
+    },
+    {
+      "epoch": 0.1752634498058791,
+      "grad_norm": 0.17510953545570374,
+      "learning_rate": 0.00017555555555555553,
+      "loss": 0.6569,
+      "step": 79
+    },
+    {
+      "epoch": 0.17748197448696618,
+      "grad_norm": 0.14331524074077606,
+      "learning_rate": 0.00017777777777777776,
+      "loss": 0.5921,
+      "step": 80
+    },
+    {
+      "epoch": 0.17970049916805325,
+      "grad_norm": 0.15239472687244415,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.6383,
+      "step": 81
+    },
+    {
+      "epoch": 0.18191902384914033,
+      "grad_norm": 0.11516226828098297,
+      "learning_rate": 0.0001822222222222222,
+      "loss": 0.547,
+      "step": 82
+    },
+    {
+      "epoch": 0.1841375485302274,
+      "grad_norm": 0.25776439905166626,
+      "learning_rate": 0.00018444444444444443,
+      "loss": 0.6201,
+      "step": 83
+    },
+    {
+      "epoch": 0.18635607321131448,
+      "grad_norm": 0.14512823522090912,
+      "learning_rate": 0.00018666666666666666,
+      "loss": 0.498,
+      "step": 84
+    },
+    {
+      "epoch": 0.18857459789240155,
+      "grad_norm": 0.24660624563694,
+      "learning_rate": 0.00018888888888888888,
+      "loss": 0.6041,
+      "step": 85
+    },
+    {
+      "epoch": 0.19079312257348863,
+      "grad_norm": 0.1534428745508194,
+      "learning_rate": 0.0001911111111111111,
+      "loss": 0.5178,
+      "step": 86
+    },
+    {
+      "epoch": 0.1930116472545757,
+      "grad_norm": 0.139539435505867,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 0.6308,
+      "step": 87
+    },
+    {
+      "epoch": 0.19523017193566278,
+      "grad_norm": 0.1345309317111969,
+      "learning_rate": 0.00019555555555555556,
+      "loss": 0.6087,
+      "step": 88
+    },
+    {
+      "epoch": 0.19744869661674985,
+      "grad_norm": 0.19402587413787842,
+      "learning_rate": 0.00019777777777777776,
+      "loss": 0.601,
+      "step": 89
+    },
+    {
+      "epoch": 0.19966722129783693,
+      "grad_norm": 0.15261350572109222,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 0.6467,
+      "step": 90
+    },
+    {
+      "epoch": 0.201885745978924,
+      "grad_norm": 0.17507652938365936,
+      "learning_rate": 0.0002022222222222222,
+      "loss": 0.809,
+      "step": 91
+    },
+    {
+      "epoch": 0.2041042706600111,
+      "grad_norm": 0.13561736047267914,
+      "learning_rate": 0.00020444444444444443,
+      "loss": 0.6154,
+      "step": 92
+    },
+    {
+      "epoch": 0.20632279534109818,
+      "grad_norm": 0.15078841149806976,
+      "learning_rate": 0.00020666666666666666,
+      "loss": 0.7217,
+      "step": 93
+    },
+    {
+      "epoch": 0.20854132002218526,
+      "grad_norm": 0.17303484678268433,
+      "learning_rate": 0.00020888888888888888,
+      "loss": 0.6409,
+      "step": 94
+    },
+    {
+      "epoch": 0.21075984470327233,
+      "grad_norm": 0.16978564858436584,
+      "learning_rate": 0.0002111111111111111,
+      "loss": 0.6521,
+      "step": 95
+    },
+    {
+      "epoch": 0.2129783693843594,
+      "grad_norm": 0.15561431646347046,
+      "learning_rate": 0.00021333333333333333,
+      "loss": 0.7608,
+      "step": 96
+    },
+    {
+      "epoch": 0.21519689406544648,
+      "grad_norm": 0.135170117020607,
+      "learning_rate": 0.00021555555555555556,
+      "loss": 0.5334,
+      "step": 97
+    },
+    {
+      "epoch": 0.21741541874653356,
+      "grad_norm": 0.10681459307670593,
+      "learning_rate": 0.00021777777777777778,
+      "loss": 0.5696,
+      "step": 98
+    },
+    {
+      "epoch": 0.21963394342762063,
+      "grad_norm": 0.12983569502830505,
+      "learning_rate": 0.00021999999999999995,
+      "loss": 0.6973,
+      "step": 99
+    },
+    {
+      "epoch": 0.2218524681087077,
+      "grad_norm": 0.15038356184959412,
+      "learning_rate": 0.00022222222222222218,
+      "loss": 0.7459,
+      "step": 100
+    },
+    {
+      "epoch": 0.22407099278979478,
+      "grad_norm": 0.14898072183132172,
+      "learning_rate": 0.0002244444444444444,
+      "loss": 0.509,
+      "step": 101
+    },
+    {
+      "epoch": 0.22628951747088186,
+      "grad_norm": 0.1406606286764145,
+      "learning_rate": 0.00022666666666666663,
+      "loss": 0.6088,
+      "step": 102
+    },
+    {
+      "epoch": 0.22850804215196893,
+      "grad_norm": 0.12118898332118988,
+      "learning_rate": 0.00022888888888888885,
+      "loss": 0.6214,
+      "step": 103
+    },
+    {
+      "epoch": 0.230726566833056,
+      "grad_norm": 0.19789718091487885,
+      "learning_rate": 0.00023111111111111108,
+      "loss": 0.5827,
+      "step": 104
+    },
+    {
+      "epoch": 0.23294509151414308,
+      "grad_norm": 0.13276802003383636,
+      "learning_rate": 0.0002333333333333333,
+      "loss": 0.5885,
+      "step": 105
+    },
+    {
+      "epoch": 0.23516361619523019,
+      "grad_norm": 0.12432726472616196,
+      "learning_rate": 0.00023555555555555553,
+      "loss": 0.4902,
+      "step": 106
+    },
+    {
+      "epoch": 0.23738214087631726,
+      "grad_norm": 0.11484728008508682,
+      "learning_rate": 0.00023777777777777775,
+      "loss": 0.5546,
+      "step": 107
+    },
+    {
+      "epoch": 0.23960066555740434,
+      "grad_norm": 0.10037735104560852,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 0.6113,
+      "step": 108
+    },
+    {
+      "epoch": 0.2418191902384914,
+      "grad_norm": 0.09788448363542557,
+      "learning_rate": 0.0002422222222222222,
+      "loss": 0.6901,
+      "step": 109
+    },
+    {
+      "epoch": 0.24403771491957849,
+      "grad_norm": 0.13254740834236145,
+      "learning_rate": 0.00024444444444444443,
+      "loss": 0.6713,
+      "step": 110
+    },
+    {
+      "epoch": 0.24625623960066556,
+      "grad_norm": 0.10041166096925735,
+      "learning_rate": 0.0002466666666666666,
+      "loss": 0.4563,
+      "step": 111
+    },
+    {
+      "epoch": 0.24847476428175264,
+      "grad_norm": 0.1512630730867386,
+      "learning_rate": 0.0002488888888888889,
+      "loss": 0.5337,
+      "step": 112
+    },
+    {
+      "epoch": 0.2506932889628397,
+      "grad_norm": 0.2391219288110733,
+      "learning_rate": 0.0002511111111111111,
+      "loss": 0.629,
+      "step": 113
+    },
+    {
+      "epoch": 0.2529118136439268,
+      "grad_norm": 0.1287817507982254,
+      "learning_rate": 0.00025333333333333333,
+      "loss": 0.7481,
+      "step": 114
+    },
+    {
+      "epoch": 0.25513033832501386,
+      "grad_norm": 0.10246129333972931,
+      "learning_rate": 0.00025555555555555553,
+      "loss": 0.5754,
+      "step": 115
+    },
+    {
+      "epoch": 0.25734886300610094,
+      "grad_norm": 0.18561941385269165,
+      "learning_rate": 0.0002577777777777778,
+      "loss": 0.6913,
+      "step": 116
+    },
+    {
+      "epoch": 0.259567387687188,
+      "grad_norm": 0.25689896941185,
+      "learning_rate": 0.00026,
+      "loss": 0.6341,
+      "step": 117
+    },
+    {
+      "epoch": 0.2617859123682751,
+      "grad_norm": 0.13846208155155182,
+      "learning_rate": 0.00026222222222222223,
+      "loss": 0.5807,
+      "step": 118
+    },
+    {
+      "epoch": 0.26400443704936216,
+      "grad_norm": 0.14928129315376282,
+      "learning_rate": 0.00026444444444444443,
+      "loss": 0.7073,
+      "step": 119
+    },
+    {
+      "epoch": 0.26622296173044924,
+      "grad_norm": 0.1953628808259964,
+      "learning_rate": 0.0002666666666666666,
+      "loss": 0.8507,
+      "step": 120
+    },
+    {
+      "epoch": 0.2684414864115363,
+      "grad_norm": 0.15973518788814545,
+      "learning_rate": 0.0002688888888888889,
+      "loss": 0.5935,
+      "step": 121
+    },
+    {
+      "epoch": 0.2706600110926234,
+      "grad_norm": 0.12474338710308075,
+      "learning_rate": 0.0002711111111111111,
+      "loss": 0.7048,
+      "step": 122
+    },
+    {
+      "epoch": 0.27287853577371046,
+      "grad_norm": 0.11041238903999329,
+      "learning_rate": 0.00027333333333333333,
+      "loss": 0.5023,
+      "step": 123
+    },
+    {
+      "epoch": 0.27509706045479754,
+      "grad_norm": 0.14531372487545013,
+      "learning_rate": 0.0002755555555555555,
+      "loss": 0.6571,
+      "step": 124
+    },
+    {
+      "epoch": 0.2773155851358846,
+      "grad_norm": 0.140780508518219,
+      "learning_rate": 0.0002777777777777778,
+      "loss": 0.6461,
+      "step": 125
+    },
+    {
+      "epoch": 0.2795341098169717,
+      "grad_norm": 0.12663206458091736,
+      "learning_rate": 0.00028,
+      "loss": 0.4774,
+      "step": 126
+    },
+    {
+      "epoch": 0.2817526344980588,
+      "grad_norm": 0.20010946691036224,
+      "learning_rate": 0.00028222222222222223,
+      "loss": 0.5319,
+      "step": 127
+    },
+    {
+      "epoch": 0.2839711591791459,
+      "grad_norm": 0.11548768728971481,
+      "learning_rate": 0.0002844444444444444,
+      "loss": 0.5586,
+      "step": 128
+    },
+    {
+      "epoch": 0.28618968386023297,
+      "grad_norm": 0.1455412209033966,
+      "learning_rate": 0.0002866666666666667,
+      "loss": 0.6739,
+      "step": 129
+    },
+    {
+      "epoch": 0.28840820854132004,
+      "grad_norm": 0.12119387835264206,
+      "learning_rate": 0.0002888888888888888,
+      "loss": 0.6065,
+      "step": 130
+    },
+    {
+      "epoch": 0.2906267332224071,
+      "grad_norm": 0.20408546924591064,
+      "learning_rate": 0.0002911111111111111,
+      "loss": 0.6575,
+      "step": 131
+    },
+    {
+      "epoch": 0.2928452579034942,
+      "grad_norm": 0.16425302624702454,
+      "learning_rate": 0.00029333333333333327,
+      "loss": 0.467,
+      "step": 132
+    },
+    {
+      "epoch": 0.29506378258458127,
+      "grad_norm": 0.17163507640361786,
+      "learning_rate": 0.0002955555555555555,
+      "loss": 0.6764,
+      "step": 133
+    },
+    {
+      "epoch": 0.29728230726566834,
+      "grad_norm": 0.13018175959587097,
+      "learning_rate": 0.0002977777777777777,
+      "loss": 0.5528,
+      "step": 134
+    },
+    {
+      "epoch": 0.2995008319467554,
+      "grad_norm": 0.2961178719997406,
+      "learning_rate": 0.0003,
+      "loss": 0.4905,
+      "step": 135
+    },
+    {
+      "epoch": 0.3017193566278425,
+      "grad_norm": 0.16747187077999115,
+      "learning_rate": 0.00029975308641975307,
+      "loss": 0.4539,
+      "step": 136
+    },
+    {
+      "epoch": 0.30393788130892957,
+      "grad_norm": 0.17348533868789673,
+      "learning_rate": 0.00029950617283950616,
+      "loss": 0.7489,
+      "step": 137
+    },
+    {
+      "epoch": 0.30615640599001664,
+      "grad_norm": 0.14018985629081726,
+      "learning_rate": 0.00029925925925925926,
+      "loss": 0.6111,
+      "step": 138
+    },
+    {
+      "epoch": 0.3083749306711037,
+      "grad_norm": 0.1378738284111023,
+      "learning_rate": 0.0002990123456790123,
+      "loss": 0.5809,
+      "step": 139
+    },
+    {
+      "epoch": 0.3105934553521908,
+      "grad_norm": 0.1586250215768814,
+      "learning_rate": 0.0002987654320987654,
+      "loss": 0.6304,
+      "step": 140
+    },
+    {
+      "epoch": 0.31281198003327787,
+      "grad_norm": 0.17072919011116028,
+      "learning_rate": 0.0002985185185185185,
+      "loss": 0.5313,
+      "step": 141
+    },
+    {
+      "epoch": 0.31503050471436495,
+      "grad_norm": 0.1692097932100296,
+      "learning_rate": 0.0002982716049382716,
+      "loss": 0.5874,
+      "step": 142
+    },
+    {
+      "epoch": 0.317249029395452,
+      "grad_norm": 0.1669483631849289,
+      "learning_rate": 0.0002980246913580247,
+      "loss": 0.6706,
+      "step": 143
+    },
+    {
+      "epoch": 0.3194675540765391,
+      "grad_norm": 0.15582531690597534,
+      "learning_rate": 0.0002977777777777777,
+      "loss": 0.5096,
+      "step": 144
+    },
+    {
+      "epoch": 0.32168607875762617,
+      "grad_norm": 0.17224912345409393,
+      "learning_rate": 0.00029753086419753087,
+      "loss": 0.5856,
+      "step": 145
+    },
+    {
+      "epoch": 0.32390460343871325,
+      "grad_norm": 0.13735519349575043,
+      "learning_rate": 0.0002972839506172839,
+      "loss": 0.5881,
+      "step": 146
+    },
+    {
+      "epoch": 0.3261231281198003,
+      "grad_norm": 0.10966519266366959,
+      "learning_rate": 0.000297037037037037,
+      "loss": 0.4357,
+      "step": 147
+    },
+    {
+      "epoch": 0.3283416528008874,
+      "grad_norm": 0.13613063097000122,
+      "learning_rate": 0.0002967901234567901,
+      "loss": 0.5431,
+      "step": 148
+    },
+    {
+      "epoch": 0.33056017748197447,
+      "grad_norm": 0.14249297976493835,
+      "learning_rate": 0.0002965432098765432,
+      "loss": 0.6296,
+      "step": 149
+    },
+    {
+      "epoch": 0.33277870216306155,
+      "grad_norm": 0.14954577386379242,
+      "learning_rate": 0.00029629629629629624,
+      "loss": 0.6063,
+      "step": 150
+    },
+    {
+      "epoch": 0.3349972268441486,
+      "grad_norm": 0.11578797549009323,
+      "learning_rate": 0.00029604938271604933,
+      "loss": 0.5696,
+      "step": 151
+    },
+    {
+      "epoch": 0.3372157515252357,
+      "grad_norm": 0.2894980013370514,
+      "learning_rate": 0.0002958024691358025,
+      "loss": 0.6303,
+      "step": 152
+    },
+    {
+      "epoch": 0.33943427620632277,
+      "grad_norm": 0.13511399924755096,
+      "learning_rate": 0.0002955555555555555,
+      "loss": 0.619,
+      "step": 153
+    },
+    {
+      "epoch": 0.34165280088740985,
+      "grad_norm": 0.1340877115726471,
+      "learning_rate": 0.0002953086419753086,
+      "loss": 0.6378,
+      "step": 154
+    },
+    {
+      "epoch": 0.343871325568497,
+      "grad_norm": 0.18187831342220306,
+      "learning_rate": 0.0002950617283950617,
+      "loss": 0.6676,
+      "step": 155
+    },
+    {
+      "epoch": 0.34608985024958405,
+      "grad_norm": 0.12415005266666412,
+      "learning_rate": 0.0002948148148148148,
+      "loss": 0.545,
+      "step": 156
+    },
+    {
+      "epoch": 0.3483083749306711,
+      "grad_norm": 0.2772611975669861,
+      "learning_rate": 0.00029456790123456785,
+      "loss": 0.5896,
+      "step": 157
+    },
+    {
+      "epoch": 0.3505268996117582,
+      "grad_norm": 0.1492634266614914,
+      "learning_rate": 0.00029432098765432095,
+      "loss": 0.6274,
+      "step": 158
+    },
+    {
+      "epoch": 0.3527454242928453,
+      "grad_norm": 0.14092473685741425,
+      "learning_rate": 0.00029407407407407404,
+      "loss": 0.5004,
+      "step": 159
+    },
+    {
+      "epoch": 0.35496394897393235,
+      "grad_norm": 0.1523817777633667,
+      "learning_rate": 0.00029382716049382714,
+      "loss": 0.4755,
+      "step": 160
+    },
+    {
+      "epoch": 0.35718247365501943,
+      "grad_norm": 0.13400276005268097,
+      "learning_rate": 0.00029358024691358023,
+      "loss": 0.5868,
+      "step": 161
+    },
+    {
+      "epoch": 0.3594009983361065,
+      "grad_norm": 0.15949207544326782,
+      "learning_rate": 0.00029333333333333327,
+      "loss": 0.6626,
+      "step": 162
+    },
+    {
+      "epoch": 0.3616195230171936,
+      "grad_norm": 0.1505206823348999,
+      "learning_rate": 0.0002930864197530864,
+      "loss": 0.5697,
+      "step": 163
+    },
+    {
+      "epoch": 0.36383804769828065,
+      "grad_norm": 0.10767590254545212,
+      "learning_rate": 0.00029283950617283946,
+      "loss": 0.5119,
+      "step": 164
+    },
+    {
+      "epoch": 0.36605657237936773,
+      "grad_norm": 0.13458192348480225,
+      "learning_rate": 0.00029259259259259256,
+      "loss": 0.5265,
+      "step": 165
+    },
+    {
+      "epoch": 0.3682750970604548,
+      "grad_norm": 0.1502276062965393,
+      "learning_rate": 0.00029234567901234565,
+      "loss": 0.6426,
+      "step": 166
+    },
+    {
+      "epoch": 0.3704936217415419,
+      "grad_norm": 0.1290915459394455,
+      "learning_rate": 0.00029209876543209875,
+      "loss": 0.5738,
+      "step": 167
+    },
+    {
+      "epoch": 0.37271214642262895,
+      "grad_norm": 0.10446428507566452,
+      "learning_rate": 0.00029185185185185184,
+      "loss": 0.5432,
+      "step": 168
+    },
+    {
+      "epoch": 0.37493067110371603,
+      "grad_norm": 0.17166462540626526,
+      "learning_rate": 0.0002916049382716049,
+      "loss": 0.4576,
+      "step": 169
+    },
+    {
+      "epoch": 0.3771491957848031,
+      "grad_norm": 0.14254140853881836,
+      "learning_rate": 0.000291358024691358,
+      "loss": 0.6232,
+      "step": 170
+    },
+    {
+      "epoch": 0.3793677204658902,
+      "grad_norm": 0.14636781811714172,
+      "learning_rate": 0.0002911111111111111,
+      "loss": 0.676,
+      "step": 171
+    },
+    {
+      "epoch": 0.38158624514697725,
+      "grad_norm": 0.16013355553150177,
+      "learning_rate": 0.00029086419753086417,
+      "loss": 0.528,
+      "step": 172
+    },
+    {
+      "epoch": 0.38380476982806433,
+      "grad_norm": 0.16806840896606445,
+      "learning_rate": 0.00029061728395061726,
+      "loss": 0.6279,
+      "step": 173
+    },
+    {
+      "epoch": 0.3860232945091514,
+      "grad_norm": 0.21191895008087158,
+      "learning_rate": 0.00029037037037037036,
+      "loss": 0.6768,
+      "step": 174
+    },
+    {
+      "epoch": 0.3882418191902385,
+      "grad_norm": 0.16243013739585876,
+      "learning_rate": 0.00029012345679012345,
+      "loss": 0.5368,
+      "step": 175
+    },
+    {
+      "epoch": 0.39046034387132555,
+      "grad_norm": 0.14273805916309357,
+      "learning_rate": 0.0002898765432098765,
+      "loss": 0.589,
+      "step": 176
+    },
+    {
+      "epoch": 0.39267886855241263,
+      "grad_norm": 0.1802993267774582,
+      "learning_rate": 0.0002896296296296296,
+      "loss": 0.631,
+      "step": 177
+    },
+    {
+      "epoch": 0.3948973932334997,
+      "grad_norm": 0.1407935917377472,
+      "learning_rate": 0.0002893827160493827,
+      "loss": 0.5616,
+      "step": 178
+    },
+    {
+      "epoch": 0.3971159179145868,
+      "grad_norm": 0.1350397765636444,
+      "learning_rate": 0.0002891358024691358,
+      "loss": 0.6642,
+      "step": 179
+    },
+    {
+      "epoch": 0.39933444259567386,
+      "grad_norm": 0.15307499468326569,
+      "learning_rate": 0.0002888888888888888,
+      "loss": 0.6019,
+      "step": 180
+    },
+    {
+      "epoch": 0.40155296727676093,
+      "grad_norm": 0.18532080948352814,
+      "learning_rate": 0.00028864197530864197,
+      "loss": 0.628,
+      "step": 181
+    },
+    {
+      "epoch": 0.403771491957848,
+      "grad_norm": 0.1410764455795288,
+      "learning_rate": 0.00028839506172839507,
+      "loss": 0.5383,
+      "step": 182
+    },
+    {
+      "epoch": 0.4059900166389351,
+      "grad_norm": 0.1594809889793396,
+      "learning_rate": 0.0002881481481481481,
+      "loss": 0.5296,
+      "step": 183
+    },
+    {
+      "epoch": 0.4082085413200222,
+      "grad_norm": 0.14522376656532288,
+      "learning_rate": 0.0002879012345679012,
+      "loss": 0.6139,
+      "step": 184
+    },
+    {
+      "epoch": 0.4104270660011093,
+      "grad_norm": 0.11685585975646973,
+      "learning_rate": 0.0002876543209876543,
+      "loss": 0.5107,
+      "step": 185
+    },
+    {
+      "epoch": 0.41264559068219636,
+      "grad_norm": 0.2598012685775757,
+      "learning_rate": 0.0002874074074074074,
+      "loss": 0.6578,
+      "step": 186
+    },
+    {
+      "epoch": 0.41486411536328344,
+      "grad_norm": 0.26201024651527405,
+      "learning_rate": 0.00028716049382716043,
+      "loss": 0.5823,
+      "step": 187
+    },
+    {
+      "epoch": 0.4170826400443705,
+      "grad_norm": 0.14132912456989288,
+      "learning_rate": 0.00028691358024691353,
+      "loss": 0.5657,
+      "step": 188
+    },
+    {
+      "epoch": 0.4193011647254576,
+      "grad_norm": 0.19780173897743225,
+      "learning_rate": 0.0002866666666666667,
+      "loss": 0.7732,
+      "step": 189
+    },
+    {
+      "epoch": 0.42151968940654466,
+      "grad_norm": 0.1830815225839615,
+      "learning_rate": 0.0002864197530864197,
+      "loss": 0.5737,
+      "step": 190
+    },
+    {
+      "epoch": 0.42373821408763174,
+      "grad_norm": 0.1100027784705162,
+      "learning_rate": 0.0002861728395061728,
+      "loss": 0.499,
+      "step": 191
+    },
+    {
+      "epoch": 0.4259567387687188,
+      "grad_norm": 0.14626961946487427,
+      "learning_rate": 0.0002859259259259259,
+      "loss": 0.6307,
+      "step": 192
+    },
+    {
+      "epoch": 0.4281752634498059,
+      "grad_norm": 0.11621357500553131,
+      "learning_rate": 0.000285679012345679,
+      "loss": 0.5992,
+      "step": 193
+    },
+    {
+      "epoch": 0.43039378813089296,
+      "grad_norm": 0.15311074256896973,
+      "learning_rate": 0.00028543209876543204,
+      "loss": 0.4552,
+      "step": 194
+    },
+    {
+      "epoch": 0.43261231281198004,
+      "grad_norm": 0.17864878475666046,
+      "learning_rate": 0.00028518518518518514,
+      "loss": 0.4991,
+      "step": 195
+    },
+    {
+      "epoch": 0.4348308374930671,
+      "grad_norm": 0.17827802896499634,
+      "learning_rate": 0.00028493827160493824,
+      "loss": 0.795,
+      "step": 196
+    },
+    {
+      "epoch": 0.4370493621741542,
+      "grad_norm": 0.1697523295879364,
+      "learning_rate": 0.00028469135802469133,
+      "loss": 0.6736,
+      "step": 197
+    },
+    {
+      "epoch": 0.43926788685524126,
+      "grad_norm": 0.13543234765529633,
+      "learning_rate": 0.0002844444444444444,
+      "loss": 0.5707,
+      "step": 198
+    },
+    {
+      "epoch": 0.44148641153632834,
+      "grad_norm": 0.18668019771575928,
+      "learning_rate": 0.0002841975308641975,
+      "loss": 0.6865,
+      "step": 199
+    },
+    {
+      "epoch": 0.4437049362174154,
+      "grad_norm": 0.10790561884641647,
+      "learning_rate": 0.0002839506172839506,
+      "loss": 0.6167,
+      "step": 200
+    },
+    {
+      "epoch": 0.4459234608985025,
+      "grad_norm": 0.1593407690525055,
+      "learning_rate": 0.00028370370370370366,
+      "loss": 0.6068,
+      "step": 201
+    },
+    {
+      "epoch": 0.44814198557958956,
+      "grad_norm": 0.1326252520084381,
+      "learning_rate": 0.00028345679012345675,
+      "loss": 0.4725,
+      "step": 202
+    },
+    {
+      "epoch": 0.45036051026067664,
+      "grad_norm": 0.11908572167158127,
+      "learning_rate": 0.00028320987654320985,
+      "loss": 0.5911,
+      "step": 203
+    },
+    {
+      "epoch": 0.4525790349417637,
+      "grad_norm": 0.16995583474636078,
+      "learning_rate": 0.00028296296296296294,
+      "loss": 0.8292,
+      "step": 204
+    },
+    {
+      "epoch": 0.4547975596228508,
+      "grad_norm": 0.17582087218761444,
+      "learning_rate": 0.00028271604938271604,
+      "loss": 0.7103,
+      "step": 205
+    },
+    {
+      "epoch": 0.45701608430393786,
+      "grad_norm": 0.1507999300956726,
+      "learning_rate": 0.0002824691358024691,
+      "loss": 0.5931,
+      "step": 206
+    },
+    {
+      "epoch": 0.45923460898502494,
+      "grad_norm": 0.14072640240192413,
+      "learning_rate": 0.00028222222222222223,
+      "loss": 0.7009,
+      "step": 207
+    },
+    {
+      "epoch": 0.461453133666112,
+      "grad_norm": 0.1515844464302063,
+      "learning_rate": 0.00028197530864197527,
+      "loss": 0.5559,
+      "step": 208
+    },
+    {
+      "epoch": 0.4636716583471991,
+      "grad_norm": 0.16681179404258728,
+      "learning_rate": 0.00028172839506172836,
+      "loss": 0.4486,
+      "step": 209
+    },
+    {
+      "epoch": 0.46589018302828616,
+      "grad_norm": 0.15187156200408936,
+      "learning_rate": 0.00028148148148148146,
+      "loss": 0.6349,
+      "step": 210
+    },
+    {
+      "epoch": 0.46810870770937324,
+      "grad_norm": 0.11201658099889755,
+      "learning_rate": 0.00028123456790123455,
+      "loss": 0.5163,
+      "step": 211
+    },
+    {
+      "epoch": 0.47032723239046037,
+      "grad_norm": 0.1587248593568802,
+      "learning_rate": 0.00028098765432098765,
+      "loss": 0.6233,
+      "step": 212
+    },
+    {
+      "epoch": 0.47254575707154745,
+      "grad_norm": 0.12440529465675354,
+      "learning_rate": 0.0002807407407407407,
+      "loss": 0.6035,
+      "step": 213
+    },
+    {
+      "epoch": 0.4747642817526345,
+      "grad_norm": 0.20593607425689697,
+      "learning_rate": 0.0002804938271604938,
+      "loss": 0.5973,
+      "step": 214
+    },
+    {
+      "epoch": 0.4769828064337216,
+      "grad_norm": 0.15497110784053802,
+      "learning_rate": 0.0002802469135802469,
+      "loss": 0.5888,
+      "step": 215
+    },
+    {
+      "epoch": 0.47920133111480867,
+      "grad_norm": 0.17547713220119476,
+      "learning_rate": 0.00028,
+      "loss": 0.6078,
+      "step": 216
+    },
+    {
+      "epoch": 0.48141985579589575,
+      "grad_norm": 0.15233361721038818,
+      "learning_rate": 0.00027975308641975307,
+      "loss": 0.5656,
+      "step": 217
+    },
+    {
+      "epoch": 0.4836383804769828,
+      "grad_norm": 0.15334013104438782,
+      "learning_rate": 0.00027950617283950617,
+      "loss": 0.575,
+      "step": 218
+    },
+    {
+      "epoch": 0.4858569051580699,
+      "grad_norm": 0.16986176371574402,
+      "learning_rate": 0.00027925925925925926,
+      "loss": 0.5374,
+      "step": 219
+    },
+    {
+      "epoch": 0.48807542983915697,
+      "grad_norm": 0.17319795489311218,
+      "learning_rate": 0.0002790123456790123,
+      "loss": 0.5557,
+      "step": 220
+    },
+    {
+      "epoch": 0.49029395452024405,
+      "grad_norm": 0.1536441296339035,
+      "learning_rate": 0.0002787654320987654,
+      "loss": 0.6566,
+      "step": 221
+    },
+    {
+      "epoch": 0.4925124792013311,
+      "grad_norm": 0.1585821807384491,
+      "learning_rate": 0.0002785185185185185,
+      "loss": 0.6118,
+      "step": 222
+    },
+    {
+      "epoch": 0.4947310038824182,
+      "grad_norm": 0.10872151702642441,
+      "learning_rate": 0.0002782716049382716,
+      "loss": 0.704,
+      "step": 223
+    },
+    {
+      "epoch": 0.49694952856350527,
+      "grad_norm": 0.1618778556585312,
+      "learning_rate": 0.00027802469135802463,
+      "loss": 0.6819,
+      "step": 224
+    },
+    {
+      "epoch": 0.49916805324459235,
+      "grad_norm": 0.15477105975151062,
+      "learning_rate": 0.0002777777777777778,
+      "loss": 0.5439,
+      "step": 225
+    },
+    {
+      "epoch": 0.5013865779256794,
+      "grad_norm": 0.1944153904914856,
+      "learning_rate": 0.00027753086419753087,
+      "loss": 0.7084,
+      "step": 226
+    },
+    {
+      "epoch": 0.5036051026067665,
+      "grad_norm": 0.12949158251285553,
+      "learning_rate": 0.0002772839506172839,
+      "loss": 0.7889,
+      "step": 227
+    },
+    {
+      "epoch": 0.5058236272878536,
+      "grad_norm": 0.13340526819229126,
+      "learning_rate": 0.000277037037037037,
+      "loss": 0.5542,
+      "step": 228
+    },
+    {
+      "epoch": 0.5080421519689406,
+      "grad_norm": 0.14349310100078583,
+      "learning_rate": 0.0002767901234567901,
+      "loss": 0.6564,
+      "step": 229
+    },
+    {
+      "epoch": 0.5102606766500277,
+      "grad_norm": 0.12724518775939941,
+      "learning_rate": 0.0002765432098765432,
+      "loss": 0.6158,
+      "step": 230
+    },
+    {
+      "epoch": 0.5124792013311148,
+      "grad_norm": 0.19005486369132996,
+      "learning_rate": 0.00027629629629629624,
+      "loss": 0.4797,
+      "step": 231
+    },
+    {
+      "epoch": 0.5146977260122019,
+      "grad_norm": 0.14671306312084198,
+      "learning_rate": 0.00027604938271604934,
+      "loss": 0.6126,
+      "step": 232
+    },
+    {
+      "epoch": 0.516916250693289,
+      "grad_norm": 0.14581987261772156,
+      "learning_rate": 0.00027580246913580243,
+      "loss": 0.6458,
+      "step": 233
+    },
+    {
+      "epoch": 0.519134775374376,
+      "grad_norm": 0.1615772247314453,
+      "learning_rate": 0.0002755555555555555,
+      "loss": 0.5287,
+      "step": 234
+    },
+    {
+      "epoch": 0.5213533000554631,
+      "grad_norm": 0.18958322703838348,
+      "learning_rate": 0.0002753086419753086,
+      "loss": 0.4864,
+      "step": 235
+    },
+    {
+      "epoch": 0.5235718247365502,
+      "grad_norm": 0.13866981863975525,
+      "learning_rate": 0.0002750617283950617,
+      "loss": 0.6034,
+      "step": 236
+    },
+    {
+      "epoch": 0.5257903494176372,
+      "grad_norm": 0.1341233104467392,
+      "learning_rate": 0.0002748148148148148,
+      "loss": 0.5942,
+      "step": 237
+    },
+    {
+      "epoch": 0.5280088740987243,
+      "grad_norm": 0.15827366709709167,
+      "learning_rate": 0.00027456790123456785,
+      "loss": 0.6605,
+      "step": 238
+    },
+    {
+      "epoch": 0.5302273987798114,
+      "grad_norm": 0.15706217288970947,
+      "learning_rate": 0.00027432098765432095,
+      "loss": 0.5094,
+      "step": 239
+    },
+    {
+      "epoch": 0.5324459234608985,
+      "grad_norm": 0.12293560802936554,
+      "learning_rate": 0.00027407407407407404,
+      "loss": 0.5242,
+      "step": 240
+    },
+    {
+      "epoch": 0.5346644481419855,
+      "grad_norm": 0.19408516585826874,
+      "learning_rate": 0.00027382716049382714,
+      "loss": 0.6095,
+      "step": 241
+    },
+    {
+      "epoch": 0.5368829728230726,
+      "grad_norm": 0.15772481262683868,
+      "learning_rate": 0.00027358024691358023,
+      "loss": 0.5054,
+      "step": 242
+    },
+    {
+      "epoch": 0.5391014975041597,
+      "grad_norm": 0.1341150850057602,
+      "learning_rate": 0.00027333333333333333,
+      "loss": 0.5087,
+      "step": 243
+    },
+    {
+      "epoch": 0.5413200221852468,
+      "grad_norm": 0.1397915482521057,
+      "learning_rate": 0.0002730864197530864,
+      "loss": 0.5429,
+      "step": 244
+    },
+    {
+      "epoch": 0.5435385468663338,
+      "grad_norm": 0.2070881426334381,
+      "learning_rate": 0.00027283950617283946,
+      "loss": 0.611,
+      "step": 245
+    },
+    {
+      "epoch": 0.5457570715474209,
+      "grad_norm": 0.17550469934940338,
+      "learning_rate": 0.00027259259259259256,
+      "loss": 0.4507,
+      "step": 246
+    },
+    {
+      "epoch": 0.547975596228508,
+      "grad_norm": 0.12947581708431244,
+      "learning_rate": 0.00027234567901234565,
+      "loss": 0.5985,
+      "step": 247
+    },
+    {
+      "epoch": 0.5501941209095951,
+      "grad_norm": 0.18589860200881958,
+      "learning_rate": 0.00027209876543209875,
+      "loss": 0.4692,
+      "step": 248
+    },
+    {
+      "epoch": 0.5524126455906821,
+      "grad_norm": 0.10877890139818192,
+      "learning_rate": 0.00027185185185185184,
+      "loss": 0.5262,
+      "step": 249
+    },
+    {
+      "epoch": 0.5546311702717692,
+      "grad_norm": 0.1709548830986023,
+      "learning_rate": 0.0002716049382716049,
+      "loss": 0.5077,
+      "step": 250
+    },
+    {
+      "epoch": 0.5568496949528563,
+      "grad_norm": 0.1793052852153778,
+      "learning_rate": 0.00027135802469135803,
+      "loss": 0.6491,
+      "step": 251
+    },
+    {
+      "epoch": 0.5590682196339434,
+      "grad_norm": 0.1529259979724884,
+      "learning_rate": 0.0002711111111111111,
+      "loss": 0.5269,
+      "step": 252
+    },
+    {
+      "epoch": 0.5612867443150305,
+      "grad_norm": 0.12397223711013794,
+      "learning_rate": 0.00027086419753086417,
+      "loss": 0.5769,
+      "step": 253
+    },
+    {
+      "epoch": 0.5635052689961176,
+      "grad_norm": 0.15325616300106049,
+      "learning_rate": 0.00027061728395061727,
+      "loss": 0.6829,
+      "step": 254
+    },
+    {
+      "epoch": 0.5657237936772047,
+      "grad_norm": 0.12900419533252716,
+      "learning_rate": 0.00027037037037037036,
+      "loss": 0.5254,
+      "step": 255
+    },
+    {
+      "epoch": 0.5679423183582918,
+      "grad_norm": 0.1374083012342453,
+      "learning_rate": 0.00027012345679012346,
+      "loss": 0.5078,
+      "step": 256
+    },
+    {
+      "epoch": 0.5701608430393789,
+      "grad_norm": 0.17645208537578583,
+      "learning_rate": 0.0002698765432098765,
+      "loss": 0.5763,
+      "step": 257
+    },
+    {
+      "epoch": 0.5723793677204659,
+      "grad_norm": 0.2143390029668808,
+      "learning_rate": 0.0002696296296296296,
+      "loss": 0.6764,
+      "step": 258
+    },
+    {
+      "epoch": 0.574597892401553,
+      "grad_norm": 0.13414160907268524,
+      "learning_rate": 0.0002693827160493827,
+      "loss": 0.5826,
+      "step": 259
+    },
+    {
+      "epoch": 0.5768164170826401,
+      "grad_norm": 0.12180149555206299,
+      "learning_rate": 0.0002691358024691358,
+      "loss": 0.6006,
+      "step": 260
+    },
+    {
+      "epoch": 0.5790349417637272,
+      "grad_norm": 0.21711668372154236,
+      "learning_rate": 0.0002688888888888889,
+      "loss": 0.7095,
+      "step": 261
+    },
+    {
+      "epoch": 0.5812534664448142,
+      "grad_norm": 0.16011695563793182,
+      "learning_rate": 0.00026864197530864197,
+      "loss": 0.4274,
+      "step": 262
+    },
+    {
+      "epoch": 0.5834719911259013,
+      "grad_norm": 0.14460201561450958,
+      "learning_rate": 0.000268395061728395,
+      "loss": 0.5741,
+      "step": 263
+    },
+    {
+      "epoch": 0.5856905158069884,
+      "grad_norm": 0.24207746982574463,
+      "learning_rate": 0.0002681481481481481,
+      "loss": 0.5238,
+      "step": 264
+    },
+    {
+      "epoch": 0.5879090404880755,
+      "grad_norm": 0.16883164644241333,
+      "learning_rate": 0.0002679012345679012,
+      "loss": 0.5461,
+      "step": 265
+    },
+    {
+      "epoch": 0.5901275651691625,
+      "grad_norm": 0.15376055240631104,
+      "learning_rate": 0.0002676543209876543,
+      "loss": 0.6751,
+      "step": 266
+    },
+    {
+      "epoch": 0.5923460898502496,
+      "grad_norm": 0.15175005793571472,
+      "learning_rate": 0.0002674074074074074,
+      "loss": 0.6095,
+      "step": 267
+    },
+    {
+      "epoch": 0.5945646145313367,
+      "grad_norm": 0.16412338614463806,
+      "learning_rate": 0.00026716049382716043,
+      "loss": 0.648,
+      "step": 268
+    },
+    {
+      "epoch": 0.5967831392124238,
+      "grad_norm": 0.13986684381961823,
+      "learning_rate": 0.0002669135802469136,
+      "loss": 0.6963,
+      "step": 269
+    },
+    {
+      "epoch": 0.5990016638935108,
+      "grad_norm": 0.19838367402553558,
+      "learning_rate": 0.0002666666666666666,
+      "loss": 0.6468,
+      "step": 270
+    },
+    {
+      "epoch": 0.6012201885745979,
+      "grad_norm": 0.22167545557022095,
+      "learning_rate": 0.0002664197530864197,
+      "loss": 0.6232,
+      "step": 271
+    },
+    {
+      "epoch": 0.603438713255685,
+      "grad_norm": 0.13891999423503876,
+      "learning_rate": 0.0002661728395061728,
+      "loss": 0.4408,
+      "step": 272
+    },
+    {
+      "epoch": 0.6056572379367721,
+      "grad_norm": 0.17830371856689453,
+      "learning_rate": 0.0002659259259259259,
+      "loss": 0.6186,
+      "step": 273
+    },
+    {
+      "epoch": 0.6078757626178591,
+      "grad_norm": 0.14621852338314056,
+      "learning_rate": 0.000265679012345679,
+      "loss": 0.7086,
+      "step": 274
+    },
+    {
+      "epoch": 0.6100942872989462,
+      "grad_norm": 0.11097393184900284,
+      "learning_rate": 0.00026543209876543205,
+      "loss": 0.4441,
+      "step": 275
+    },
+    {
+      "epoch": 0.6123128119800333,
+      "grad_norm": 0.14078162610530853,
+      "learning_rate": 0.00026518518518518514,
+      "loss": 0.6217,
+      "step": 276
+    },
+    {
+      "epoch": 0.6145313366611204,
+      "grad_norm": 0.1096246987581253,
+      "learning_rate": 0.00026493827160493824,
+      "loss": 0.5945,
+      "step": 277
+    },
+    {
+      "epoch": 0.6167498613422074,
+      "grad_norm": 0.1713704764842987,
+      "learning_rate": 0.00026469135802469133,
+      "loss": 0.5195,
+      "step": 278
+    },
+    {
+      "epoch": 0.6189683860232945,
+      "grad_norm": 0.12063216418027878,
+      "learning_rate": 0.00026444444444444443,
+      "loss": 0.5141,
+      "step": 279
+    },
+    {
+      "epoch": 0.6211869107043816,
+      "grad_norm": 0.15517669916152954,
+      "learning_rate": 0.0002641975308641975,
+      "loss": 0.4937,
+      "step": 280
+    },
+    {
+      "epoch": 0.6234054353854687,
+      "grad_norm": 0.13027925789356232,
+      "learning_rate": 0.0002639506172839506,
+      "loss": 0.5323,
+      "step": 281
+    },
+    {
+      "epoch": 0.6256239600665557,
+      "grad_norm": 0.2706521451473236,
+      "learning_rate": 0.00026370370370370366,
+      "loss": 0.5467,
+      "step": 282
+    },
+    {
+      "epoch": 0.6278424847476428,
+      "grad_norm": 0.11003948003053665,
+      "learning_rate": 0.00026345679012345675,
+      "loss": 0.4158,
+      "step": 283
+    },
+    {
+      "epoch": 0.6300610094287299,
+      "grad_norm": 0.13010843098163605,
+      "learning_rate": 0.00026320987654320985,
+      "loss": 0.5091,
+      "step": 284
+    },
+    {
+      "epoch": 0.632279534109817,
+      "grad_norm": 0.18734945356845856,
+      "learning_rate": 0.00026296296296296294,
+      "loss": 0.5741,
+      "step": 285
+    },
+    {
+      "epoch": 0.634498058790904,
+      "grad_norm": 0.15195336937904358,
+      "learning_rate": 0.00026271604938271604,
+      "loss": 0.5592,
+      "step": 286
+    },
+    {
+      "epoch": 0.6367165834719911,
+      "grad_norm": 0.15096698701381683,
+      "learning_rate": 0.00026246913580246913,
+      "loss": 0.5681,
+      "step": 287
+    },
+    {
+      "epoch": 0.6389351081530782,
+      "grad_norm": 0.3738124668598175,
+      "learning_rate": 0.00026222222222222223,
+      "loss": 0.5584,
+      "step": 288
+    },
+    {
+      "epoch": 0.6411536328341653,
+      "grad_norm": 0.11052624881267548,
+      "learning_rate": 0.00026197530864197527,
+      "loss": 0.6455,
+      "step": 289
+    },
+    {
+      "epoch": 0.6433721575152523,
+      "grad_norm": 0.31686267256736755,
+      "learning_rate": 0.00026172839506172837,
+      "loss": 0.6811,
+      "step": 290
+    },
+    {
+      "epoch": 0.6455906821963394,
+      "grad_norm": 0.16095049679279327,
+      "learning_rate": 0.00026148148148148146,
+      "loss": 0.6555,
+      "step": 291
+    },
+    {
+      "epoch": 0.6478092068774265,
+      "grad_norm": 0.14359760284423828,
+      "learning_rate": 0.00026123456790123456,
+      "loss": 0.6258,
+      "step": 292
+    },
+    {
+      "epoch": 0.6500277315585136,
+      "grad_norm": 0.13574042916297913,
+      "learning_rate": 0.0002609876543209876,
+      "loss": 0.5826,
+      "step": 293
+    },
+    {
+      "epoch": 0.6522462562396006,
+      "grad_norm": 0.1617458313703537,
+      "learning_rate": 0.0002607407407407407,
+      "loss": 0.5448,
+      "step": 294
+    },
+    {
+      "epoch": 0.6544647809206877,
+      "grad_norm": 0.1457916796207428,
+      "learning_rate": 0.00026049382716049384,
+      "loss": 0.4572,
+      "step": 295
+    },
+    {
+      "epoch": 0.6566833056017748,
+      "grad_norm": 0.1726388782262802,
+      "learning_rate": 0.0002602469135802469,
+      "loss": 0.6818,
+      "step": 296
+    },
+    {
+      "epoch": 0.6589018302828619,
+      "grad_norm": 0.1685604304075241,
+      "learning_rate": 0.00026,
+      "loss": 0.5009,
+      "step": 297
+    },
+    {
+      "epoch": 0.6611203549639489,
+      "grad_norm": 0.13495516777038574,
+      "learning_rate": 0.00025975308641975307,
+      "loss": 0.5254,
+      "step": 298
+    },
+    {
+      "epoch": 0.663338879645036,
+      "grad_norm": 0.15805339813232422,
+      "learning_rate": 0.00025950617283950617,
+      "loss": 0.7498,
+      "step": 299
+    },
+    {
+      "epoch": 0.6655574043261231,
+      "grad_norm": 0.15871265530586243,
+      "learning_rate": 0.0002592592592592592,
+      "loss": 0.663,
+      "step": 300
+    },
+    {
+      "epoch": 0.6677759290072102,
+      "grad_norm": 0.2122025191783905,
+      "learning_rate": 0.0002590123456790123,
+      "loss": 0.5082,
+      "step": 301
+    },
+    {
+      "epoch": 0.6699944536882972,
+      "grad_norm": 0.1571735441684723,
+      "learning_rate": 0.0002587654320987654,
+      "loss": 0.6083,
+      "step": 302
+    },
+    {
+      "epoch": 0.6722129783693843,
+      "grad_norm": 0.12305045872926712,
+      "learning_rate": 0.0002585185185185185,
+      "loss": 0.5205,
+      "step": 303
+    },
+    {
+      "epoch": 0.6744315030504714,
+      "grad_norm": 0.12740938365459442,
+      "learning_rate": 0.0002582716049382716,
+      "loss": 0.4626,
+      "step": 304
+    },
+    {
+      "epoch": 0.6766500277315585,
+      "grad_norm": 0.15358245372772217,
+      "learning_rate": 0.0002580246913580247,
+      "loss": 0.5887,
+      "step": 305
+    },
+    {
+      "epoch": 0.6788685524126455,
+      "grad_norm": 0.14395689964294434,
+      "learning_rate": 0.0002577777777777778,
+      "loss": 0.4883,
+      "step": 306
+    },
+    {
+      "epoch": 0.6810870770937326,
+      "grad_norm": 0.1655711531639099,
+      "learning_rate": 0.0002575308641975308,
+      "loss": 0.6266,
+      "step": 307
+    },
+    {
+      "epoch": 0.6833056017748197,
+      "grad_norm": 0.1656055599451065,
+      "learning_rate": 0.0002572839506172839,
+      "loss": 0.6647,
+      "step": 308
+    },
+    {
+      "epoch": 0.6855241264559068,
+      "grad_norm": 0.14409437775611877,
+      "learning_rate": 0.000257037037037037,
+      "loss": 0.5409,
+      "step": 309
+    },
+    {
+      "epoch": 0.687742651136994,
+      "grad_norm": 0.18999314308166504,
+      "learning_rate": 0.0002567901234567901,
+      "loss": 0.4875,
+      "step": 310
+    },
+    {
+      "epoch": 0.689961175818081,
+      "grad_norm": 0.16352809965610504,
+      "learning_rate": 0.0002565432098765432,
+      "loss": 0.684,
+      "step": 311
+    },
+    {
+      "epoch": 0.6921797004991681,
+      "grad_norm": 0.1448543816804886,
+      "learning_rate": 0.00025629629629629624,
+      "loss": 0.5774,
+      "step": 312
+    },
+    {
+      "epoch": 0.6943982251802552,
+      "grad_norm": 0.15293769538402557,
+      "learning_rate": 0.0002560493827160494,
+      "loss": 0.5496,
+      "step": 313
+    },
+    {
+      "epoch": 0.6966167498613423,
+      "grad_norm": 0.17188560962677002,
+      "learning_rate": 0.00025580246913580243,
+      "loss": 0.5907,
+      "step": 314
+    },
+    {
+      "epoch": 0.6988352745424293,
+      "grad_norm": 0.11369536817073822,
+      "learning_rate": 0.00025555555555555553,
+      "loss": 0.5516,
+      "step": 315
+    },
+    {
+      "epoch": 0.7010537992235164,
+      "grad_norm": 0.15182070434093475,
+      "learning_rate": 0.0002553086419753086,
+      "loss": 0.504,
+      "step": 316
+    },
+    {
+      "epoch": 0.7032723239046035,
+      "grad_norm": 0.12096603959798813,
+      "learning_rate": 0.0002550617283950617,
+      "loss": 0.4251,
+      "step": 317
+    },
+    {
+      "epoch": 0.7054908485856906,
+      "grad_norm": 0.11515217274427414,
+      "learning_rate": 0.0002548148148148148,
+      "loss": 0.4051,
+      "step": 318
+    },
+    {
+      "epoch": 0.7077093732667776,
+      "grad_norm": 0.15343734622001648,
+      "learning_rate": 0.00025456790123456785,
+      "loss": 0.6203,
+      "step": 319
+    },
+    {
+      "epoch": 0.7099278979478647,
+      "grad_norm": 0.20248335599899292,
+      "learning_rate": 0.00025432098765432095,
+      "loss": 0.6725,
+      "step": 320
+    },
+    {
+      "epoch": 0.7121464226289518,
+      "grad_norm": 0.12475546449422836,
+      "learning_rate": 0.00025407407407407404,
+      "loss": 0.5038,
+      "step": 321
+    },
+    {
+      "epoch": 0.7143649473100389,
+      "grad_norm": 0.17478933930397034,
+      "learning_rate": 0.00025382716049382714,
+      "loss": 0.6241,
+      "step": 322
+    },
+    {
+      "epoch": 0.7165834719911259,
+      "grad_norm": 0.1438637673854828,
+      "learning_rate": 0.00025358024691358023,
+      "loss": 0.5478,
+      "step": 323
+    },
+    {
+      "epoch": 0.718801996672213,
+      "grad_norm": 0.31659042835235596,
+      "learning_rate": 0.00025333333333333333,
+      "loss": 0.6046,
+      "step": 324
+    },
+    {
+      "epoch": 0.7210205213533001,
+      "grad_norm": 0.1448049694299698,
+      "learning_rate": 0.0002530864197530864,
+      "loss": 0.4721,
+      "step": 325
+    },
+    {
+      "epoch": 0.7232390460343872,
+      "grad_norm": 0.118183933198452,
+      "learning_rate": 0.00025283950617283947,
+      "loss": 0.5822,
+      "step": 326
+    },
+    {
+      "epoch": 0.7254575707154742,
+      "grad_norm": 0.14460495114326477,
+      "learning_rate": 0.00025259259259259256,
+      "loss": 0.6391,
+      "step": 327
+    },
+    {
+      "epoch": 0.7276760953965613,
+      "grad_norm": 0.1279022991657257,
+      "learning_rate": 0.00025234567901234566,
+      "loss": 0.6033,
+      "step": 328
+    },
+    {
+      "epoch": 0.7298946200776484,
+      "grad_norm": 0.13242857158184052,
+      "learning_rate": 0.00025209876543209875,
+      "loss": 0.493,
+      "step": 329
+    },
+    {
+      "epoch": 0.7321131447587355,
+      "grad_norm": 0.14890001714229584,
+      "learning_rate": 0.0002518518518518518,
+      "loss": 0.555,
+      "step": 330
+    },
+    {
+      "epoch": 0.7343316694398225,
+      "grad_norm": 0.13545982539653778,
+      "learning_rate": 0.00025160493827160494,
+      "loss": 0.6882,
+      "step": 331
+    },
+    {
+      "epoch": 0.7365501941209096,
+      "grad_norm": 0.1517675220966339,
+      "learning_rate": 0.00025135802469135804,
+      "loss": 0.5805,
+      "step": 332
+    },
+    {
+      "epoch": 0.7387687188019967,
+      "grad_norm": 0.1199595257639885,
+      "learning_rate": 0.0002511111111111111,
+      "loss": 0.4633,
+      "step": 333
+    },
+    {
+      "epoch": 0.7409872434830838,
+      "grad_norm": 0.16254772245883942,
+      "learning_rate": 0.00025086419753086417,
+      "loss": 0.6124,
+      "step": 334
+    },
+    {
+      "epoch": 0.7432057681641708,
+      "grad_norm": 0.17287039756774902,
+      "learning_rate": 0.00025061728395061727,
+      "loss": 0.6125,
+      "step": 335
+    },
+    {
+      "epoch": 0.7454242928452579,
+      "grad_norm": 0.13678033649921417,
+      "learning_rate": 0.00025037037037037036,
+      "loss": 0.6131,
+      "step": 336
+    },
+    {
+      "epoch": 0.747642817526345,
+      "grad_norm": 0.15491366386413574,
+      "learning_rate": 0.0002501234567901234,
+      "loss": 0.6468,
+      "step": 337
+    },
+    {
+      "epoch": 0.7498613422074321,
+      "grad_norm": 0.15261207520961761,
+      "learning_rate": 0.0002498765432098765,
+      "loss": 0.6164,
+      "step": 338
+    },
+    {
+      "epoch": 0.7520798668885191,
+      "grad_norm": 0.187705859541893,
+      "learning_rate": 0.00024962962962962965,
+      "loss": 0.6478,
+      "step": 339
+    },
+    {
+      "epoch": 0.7542983915696062,
+      "grad_norm": 0.12091420590877533,
+      "learning_rate": 0.0002493827160493827,
+      "loss": 0.6777,
+      "step": 340
+    },
+    {
+      "epoch": 0.7565169162506933,
+      "grad_norm": 0.17506422102451324,
+      "learning_rate": 0.0002491358024691358,
+      "loss": 0.5777,
+      "step": 341
+    },
+    {
+      "epoch": 0.7587354409317804,
+      "grad_norm": 0.18164187669754028,
+      "learning_rate": 0.0002488888888888889,
+      "loss": 0.4314,
+      "step": 342
+    },
+    {
+      "epoch": 0.7609539656128674,
+      "grad_norm": 0.13189102709293365,
+      "learning_rate": 0.000248641975308642,
+      "loss": 0.4374,
+      "step": 343
+    },
+    {
+      "epoch": 0.7631724902939545,
+      "grad_norm": 0.17622314393520355,
+      "learning_rate": 0.000248395061728395,
+      "loss": 0.6155,
+      "step": 344
+    },
+    {
+      "epoch": 0.7653910149750416,
+      "grad_norm": 0.16227610409259796,
+      "learning_rate": 0.0002481481481481481,
+      "loss": 0.6174,
+      "step": 345
+    },
+    {
+      "epoch": 0.7676095396561287,
+      "grad_norm": 0.12783291935920715,
+      "learning_rate": 0.0002479012345679012,
+      "loss": 0.591,
+      "step": 346
+    },
+    {
+      "epoch": 0.7698280643372157,
+      "grad_norm": 0.17881222069263458,
+      "learning_rate": 0.0002476543209876543,
+      "loss": 0.4954,
+      "step": 347
+    },
+    {
+      "epoch": 0.7720465890183028,
+      "grad_norm": 0.17209431529045105,
+      "learning_rate": 0.0002474074074074074,
+      "loss": 0.5502,
+      "step": 348
+    },
+    {
+      "epoch": 0.7742651136993899,
+      "grad_norm": 0.1927851438522339,
+      "learning_rate": 0.0002471604938271605,
+      "loss": 0.5006,
+      "step": 349
+    },
+    {
+      "epoch": 0.776483638380477,
+      "grad_norm": 0.197723850607872,
+      "learning_rate": 0.0002469135802469136,
+      "loss": 0.7457,
+      "step": 350
+    },
+    {
+      "epoch": 0.778702163061564,
+      "grad_norm": 0.11731608211994171,
+      "learning_rate": 0.0002466666666666666,
+      "loss": 0.4393,
+      "step": 351
+    },
+    {
+      "epoch": 0.7809206877426511,
+      "grad_norm": 0.16354243457317352,
+      "learning_rate": 0.0002464197530864197,
+      "loss": 0.5846,
+      "step": 352
+    },
+    {
+      "epoch": 0.7831392124237382,
+      "grad_norm": 0.1469251960515976,
+      "learning_rate": 0.0002461728395061728,
+      "loss": 0.5755,
+      "step": 353
+    },
+    {
+      "epoch": 0.7853577371048253,
+      "grad_norm": 0.15552709996700287,
+      "learning_rate": 0.0002459259259259259,
+      "loss": 0.5872,
+      "step": 354
+    },
+    {
+      "epoch": 0.7875762617859123,
+      "grad_norm": 0.22125008702278137,
+      "learning_rate": 0.000245679012345679,
+      "loss": 0.7383,
+      "step": 355
+    },
+    {
+      "epoch": 0.7897947864669994,
+      "grad_norm": 0.15643565356731415,
+      "learning_rate": 0.00024543209876543205,
+      "loss": 0.658,
+      "step": 356
+    },
+    {
+      "epoch": 0.7920133111480865,
+      "grad_norm": 0.16107241809368134,
+      "learning_rate": 0.0002451851851851852,
+      "loss": 0.6308,
+      "step": 357
+    },
+    {
+      "epoch": 0.7942318358291736,
+      "grad_norm": 0.16499176621437073,
+      "learning_rate": 0.00024493827160493824,
+      "loss": 0.5454,
+      "step": 358
+    },
+    {
+      "epoch": 0.7964503605102606,
+      "grad_norm": 0.1468290537595749,
+      "learning_rate": 0.00024469135802469133,
+      "loss": 0.5487,
+      "step": 359
+    },
+    {
+      "epoch": 0.7986688851913477,
+      "grad_norm": 0.20611970126628876,
+      "learning_rate": 0.00024444444444444443,
+      "loss": 0.6921,
+      "step": 360
+    },
+    {
+      "epoch": 0.8008874098724348,
+      "grad_norm": 0.13404445350170135,
+      "learning_rate": 0.0002441975308641975,
+      "loss": 0.5745,
+      "step": 361
+    },
+    {
+      "epoch": 0.8031059345535219,
+      "grad_norm": 0.1558988243341446,
+      "learning_rate": 0.00024395061728395062,
+      "loss": 0.4636,
+      "step": 362
+    },
+    {
+      "epoch": 0.8053244592346089,
+      "grad_norm": 0.13059605658054352,
+      "learning_rate": 0.0002437037037037037,
+      "loss": 0.5475,
+      "step": 363
+    },
+    {
+      "epoch": 0.807542983915696,
+      "grad_norm": 0.21544645726680756,
+      "learning_rate": 0.00024345679012345678,
+      "loss": 0.7376,
+      "step": 364
+    },
+    {
+      "epoch": 0.8097615085967831,
+      "grad_norm": 0.16309435665607452,
+      "learning_rate": 0.00024320987654320985,
+      "loss": 0.7372,
+      "step": 365
+    },
+    {
+      "epoch": 0.8119800332778702,
+      "grad_norm": 0.12765438854694366,
+      "learning_rate": 0.00024296296296296295,
+      "loss": 0.4409,
+      "step": 366
+    },
+    {
+      "epoch": 0.8141985579589573,
+      "grad_norm": 0.17903602123260498,
+      "learning_rate": 0.00024271604938271601,
+      "loss": 0.7466,
+      "step": 367
+    },
+    {
+      "epoch": 0.8164170826400444,
+      "grad_norm": 0.13719666004180908,
+      "learning_rate": 0.0002424691358024691,
+      "loss": 0.567,
+      "step": 368
+    },
+    {
+      "epoch": 0.8186356073211315,
+      "grad_norm": 0.19879184663295746,
+      "learning_rate": 0.0002422222222222222,
+      "loss": 0.7069,
+      "step": 369
+    },
+    {
+      "epoch": 0.8208541320022186,
+      "grad_norm": 0.13216494023799896,
+      "learning_rate": 0.00024197530864197527,
+      "loss": 0.628,
+      "step": 370
+    },
+    {
+      "epoch": 0.8230726566833056,
+      "grad_norm": 0.13406598567962646,
+      "learning_rate": 0.0002417283950617284,
+      "loss": 0.5219,
+      "step": 371
+    },
+    {
+      "epoch": 0.8252911813643927,
+      "grad_norm": 0.1698460876941681,
+      "learning_rate": 0.00024148148148148146,
+      "loss": 0.6288,
+      "step": 372
+    },
+    {
+      "epoch": 0.8275097060454798,
+      "grad_norm": 0.14814548194408417,
+      "learning_rate": 0.00024123456790123456,
+      "loss": 0.5459,
+      "step": 373
+    },
+    {
+      "epoch": 0.8297282307265669,
+      "grad_norm": 0.16568443179130554,
+      "learning_rate": 0.00024098765432098763,
+      "loss": 0.5363,
+      "step": 374
+    },
+    {
+      "epoch": 0.831946755407654,
+      "grad_norm": 0.1569383293390274,
+      "learning_rate": 0.00024074074074074072,
+      "loss": 0.5092,
+      "step": 375
+    },
+    {
+      "epoch": 0.834165280088741,
+      "grad_norm": 0.130343958735466,
+      "learning_rate": 0.0002404938271604938,
+      "loss": 0.6982,
+      "step": 376
+    },
+    {
+      "epoch": 0.8363838047698281,
+      "grad_norm": 0.1652912199497223,
+      "learning_rate": 0.00024024691358024688,
+      "loss": 0.6444,
+      "step": 377
+    },
+    {
+      "epoch": 0.8386023294509152,
+      "grad_norm": 0.15600495040416718,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 0.8094,
+      "step": 378
+    },
+    {
+      "epoch": 0.8408208541320022,
+      "grad_norm": 0.15927466750144958,
+      "learning_rate": 0.00023975308641975305,
+      "loss": 0.6015,
+      "step": 379
+    },
+    {
+      "epoch": 0.8430393788130893,
+      "grad_norm": 0.15467892587184906,
+      "learning_rate": 0.00023950617283950617,
+      "loss": 0.5174,
+      "step": 380
+    },
+    {
+      "epoch": 0.8452579034941764,
+      "grad_norm": 0.19386357069015503,
+      "learning_rate": 0.00023925925925925924,
+      "loss": 0.5798,
+      "step": 381
+    },
+    {
+      "epoch": 0.8474764281752635,
+      "grad_norm": 0.16425317525863647,
+      "learning_rate": 0.00023901234567901233,
+      "loss": 0.5176,
+      "step": 382
+    },
+    {
+      "epoch": 0.8496949528563505,
+      "grad_norm": 0.1314079910516739,
+      "learning_rate": 0.0002387654320987654,
+      "loss": 0.5921,
+      "step": 383
+    },
+    {
+      "epoch": 0.8519134775374376,
+      "grad_norm": 0.14600609242916107,
+      "learning_rate": 0.0002385185185185185,
+      "loss": 0.416,
+      "step": 384
+    },
+    {
+      "epoch": 0.8541320022185247,
+      "grad_norm": 0.15176311135292053,
+      "learning_rate": 0.0002382716049382716,
+      "loss": 0.5023,
+      "step": 385
+    },
+    {
+      "epoch": 0.8563505268996118,
+      "grad_norm": 0.24551711976528168,
+      "learning_rate": 0.00023802469135802466,
+      "loss": 0.5944,
+      "step": 386
+    },
+    {
+      "epoch": 0.8585690515806988,
+      "grad_norm": 0.1287483423948288,
+      "learning_rate": 0.00023777777777777775,
+      "loss": 0.8383,
+      "step": 387
+    },
+    {
+      "epoch": 0.8607875762617859,
+      "grad_norm": 0.12329156696796417,
+      "learning_rate": 0.00023753086419753082,
+      "loss": 0.511,
+      "step": 388
+    },
+    {
+      "epoch": 0.863006100942873,
+      "grad_norm": 0.21289831399917603,
+      "learning_rate": 0.00023728395061728394,
+      "loss": 0.5418,
+      "step": 389
+    },
+    {
+      "epoch": 0.8652246256239601,
+      "grad_norm": 0.15550296008586884,
+      "learning_rate": 0.000237037037037037,
+      "loss": 0.6062,
+      "step": 390
+    },
+    {
+      "epoch": 0.8674431503050472,
+      "grad_norm": 0.15629637241363525,
+      "learning_rate": 0.0002367901234567901,
+      "loss": 0.6096,
+      "step": 391
+    },
+    {
+      "epoch": 0.8696616749861342,
+      "grad_norm": 0.12582623958587646,
+      "learning_rate": 0.0002365432098765432,
+      "loss": 0.4946,
+      "step": 392
+    },
+    {
+      "epoch": 0.8718801996672213,
+      "grad_norm": 0.13614597916603088,
+      "learning_rate": 0.00023629629629629627,
+      "loss": 0.5564,
+      "step": 393
+    },
+    {
+      "epoch": 0.8740987243483084,
+      "grad_norm": 0.12428078800439835,
+      "learning_rate": 0.00023604938271604937,
+      "loss": 0.5287,
+      "step": 394
+    },
+    {
+      "epoch": 0.8763172490293955,
+      "grad_norm": 0.15186212956905365,
+      "learning_rate": 0.00023580246913580243,
+      "loss": 0.5419,
+      "step": 395
+    },
+    {
+      "epoch": 0.8785357737104825,
+      "grad_norm": 0.13204239308834076,
+      "learning_rate": 0.00023555555555555553,
+      "loss": 0.5711,
+      "step": 396
+    },
+    {
+      "epoch": 0.8807542983915696,
+      "grad_norm": 0.16621756553649902,
+      "learning_rate": 0.0002353086419753086,
+      "loss": 0.6289,
+      "step": 397
+    },
+    {
+      "epoch": 0.8829728230726567,
+      "grad_norm": 0.15927202999591827,
+      "learning_rate": 0.00023506172839506172,
+      "loss": 0.6292,
+      "step": 398
+    },
+    {
+      "epoch": 0.8851913477537438,
+      "grad_norm": 0.17237892746925354,
+      "learning_rate": 0.0002348148148148148,
+      "loss": 0.7016,
+      "step": 399
+    },
+    {
+      "epoch": 0.8874098724348308,
+      "grad_norm": 0.15591222047805786,
+      "learning_rate": 0.00023456790123456788,
+      "loss": 0.6015,
+      "step": 400
+    },
+    {
+      "epoch": 0.8896283971159179,
+      "grad_norm": 0.16728012263774872,
+      "learning_rate": 0.00023432098765432098,
+      "loss": 0.4788,
+      "step": 401
+    },
+    {
+      "epoch": 0.891846921797005,
+      "grad_norm": 0.1853276789188385,
+      "learning_rate": 0.00023407407407407405,
+      "loss": 0.5656,
+      "step": 402
+    },
+    {
+      "epoch": 0.894065446478092,
+      "grad_norm": 0.1460583359003067,
+      "learning_rate": 0.00023382716049382714,
+      "loss": 0.519,
+      "step": 403
+    },
+    {
+      "epoch": 0.8962839711591791,
+      "grad_norm": 0.16378501057624817,
+      "learning_rate": 0.0002335802469135802,
+      "loss": 0.5851,
+      "step": 404
+    },
+    {
+      "epoch": 0.8985024958402662,
+      "grad_norm": 0.1604446917772293,
+      "learning_rate": 0.0002333333333333333,
+      "loss": 0.5659,
+      "step": 405
+    },
+    {
+      "epoch": 0.9007210205213533,
+      "grad_norm": 0.13852907717227936,
+      "learning_rate": 0.00023308641975308637,
+      "loss": 0.3854,
+      "step": 406
+    },
+    {
+      "epoch": 0.9029395452024404,
+      "grad_norm": 0.18803678452968597,
+      "learning_rate": 0.0002328395061728395,
+      "loss": 0.5673,
+      "step": 407
+    },
+    {
+      "epoch": 0.9051580698835274,
+      "grad_norm": 0.20272214710712433,
+      "learning_rate": 0.0002325925925925926,
+      "loss": 0.8044,
+      "step": 408
+    },
+    {
+      "epoch": 0.9073765945646145,
+      "grad_norm": 0.13501562178134918,
+      "learning_rate": 0.00023234567901234566,
+      "loss": 0.6492,
+      "step": 409
+    },
+    {
+      "epoch": 0.9095951192457016,
+      "grad_norm": 0.14694491028785706,
+      "learning_rate": 0.00023209876543209875,
+      "loss": 0.6265,
+      "step": 410
+    },
+    {
+      "epoch": 0.9118136439267887,
+      "grad_norm": 0.20001384615898132,
+      "learning_rate": 0.00023185185185185182,
+      "loss": 0.5348,
+      "step": 411
+    },
+    {
+      "epoch": 0.9140321686078757,
+      "grad_norm": 0.17962579429149628,
+      "learning_rate": 0.00023160493827160492,
+      "loss": 0.5998,
+      "step": 412
+    },
+    {
+      "epoch": 0.9162506932889628,
+      "grad_norm": 0.2047513872385025,
+      "learning_rate": 0.00023135802469135798,
+      "loss": 0.4863,
+      "step": 413
+    },
+    {
+      "epoch": 0.9184692179700499,
+      "grad_norm": 0.15945269167423248,
+      "learning_rate": 0.00023111111111111108,
+      "loss": 0.5937,
+      "step": 414
+    },
+    {
+      "epoch": 0.920687742651137,
+      "grad_norm": 0.15113690495491028,
+      "learning_rate": 0.0002308641975308642,
+      "loss": 0.5121,
+      "step": 415
+    },
+    {
+      "epoch": 0.922906267332224,
+      "grad_norm": 0.13143697381019592,
+      "learning_rate": 0.00023061728395061727,
+      "loss": 0.5742,
+      "step": 416
+    },
+    {
+      "epoch": 0.9251247920133111,
+      "grad_norm": 0.2231714129447937,
+      "learning_rate": 0.00023037037037037036,
+      "loss": 0.5725,
+      "step": 417
+    },
+    {
+      "epoch": 0.9273433166943982,
+      "grad_norm": 0.17646975815296173,
+      "learning_rate": 0.00023012345679012343,
+      "loss": 0.6401,
+      "step": 418
+    },
+    {
+      "epoch": 0.9295618413754853,
+      "grad_norm": 0.21490518748760223,
+      "learning_rate": 0.00022987654320987653,
+      "loss": 0.639,
+      "step": 419
+    },
+    {
+      "epoch": 0.9317803660565723,
+      "grad_norm": 0.14946721494197845,
+      "learning_rate": 0.0002296296296296296,
+      "loss": 0.5441,
+      "step": 420
+    },
+    {
+      "epoch": 0.9339988907376594,
+      "grad_norm": 0.2597978413105011,
+      "learning_rate": 0.0002293827160493827,
+      "loss": 0.5753,
+      "step": 421
+    },
+    {
+      "epoch": 0.9362174154187465,
+      "grad_norm": 0.12574319541454315,
+      "learning_rate": 0.00022913580246913579,
+      "loss": 0.6766,
+      "step": 422
+    },
+    {
+      "epoch": 0.9384359400998337,
+      "grad_norm": 0.1399654597043991,
+      "learning_rate": 0.00022888888888888885,
+      "loss": 0.6271,
+      "step": 423
+    },
+    {
+      "epoch": 0.9406544647809207,
+      "grad_norm": 0.16189031302928925,
+      "learning_rate": 0.00022864197530864198,
+      "loss": 0.6221,
+      "step": 424
+    },
+    {
+      "epoch": 0.9428729894620078,
+      "grad_norm": 0.14537374675273895,
+      "learning_rate": 0.00022839506172839504,
+      "loss": 0.6281,
+      "step": 425
+    },
+    {
+      "epoch": 0.9450915141430949,
+      "grad_norm": 0.20549283921718597,
+      "learning_rate": 0.00022814814814814814,
+      "loss": 0.6839,
+      "step": 426
+    },
+    {
+      "epoch": 0.947310038824182,
+      "grad_norm": 0.13587121665477753,
+      "learning_rate": 0.0002279012345679012,
+      "loss": 0.5587,
+      "step": 427
+    },
+    {
+      "epoch": 0.949528563505269,
+      "grad_norm": 0.16577410697937012,
+      "learning_rate": 0.0002276543209876543,
+      "loss": 0.6343,
+      "step": 428
+    },
+    {
+      "epoch": 0.9517470881863561,
+      "grad_norm": 0.15042836964130402,
+      "learning_rate": 0.00022740740740740737,
+      "loss": 0.6241,
+      "step": 429
+    },
+    {
+      "epoch": 0.9539656128674432,
+      "grad_norm": 0.15503482520580292,
+      "learning_rate": 0.00022716049382716047,
+      "loss": 0.4745,
+      "step": 430
+    },
+    {
+      "epoch": 0.9561841375485303,
+      "grad_norm": 0.1614726334810257,
+      "learning_rate": 0.00022691358024691356,
+      "loss": 0.5726,
+      "step": 431
+    },
+    {
+      "epoch": 0.9584026622296173,
+      "grad_norm": 0.161455899477005,
+      "learning_rate": 0.00022666666666666663,
+      "loss": 0.6528,
+      "step": 432
+    },
+    {
+      "epoch": 0.9606211869107044,
+      "grad_norm": 0.1562987118959427,
+      "learning_rate": 0.00022641975308641975,
+      "loss": 0.6265,
+      "step": 433
+    },
+    {
+      "epoch": 0.9628397115917915,
+      "grad_norm": 0.16343088448047638,
+      "learning_rate": 0.00022617283950617282,
+      "loss": 0.7354,
+      "step": 434
+    },
+    {
+      "epoch": 0.9650582362728786,
+      "grad_norm": 0.19845885038375854,
+      "learning_rate": 0.00022592592592592591,
+      "loss": 0.6119,
+      "step": 435
+    },
+    {
+      "epoch": 0.9672767609539656,
+      "grad_norm": 0.12249289453029633,
+      "learning_rate": 0.00022567901234567898,
+      "loss": 0.5743,
+      "step": 436
+    },
+    {
+      "epoch": 0.9694952856350527,
+      "grad_norm": 0.1397278904914856,
+      "learning_rate": 0.00022543209876543208,
+      "loss": 0.5532,
+      "step": 437
+    },
+    {
+      "epoch": 0.9717138103161398,
+      "grad_norm": 0.13337703049182892,
+      "learning_rate": 0.00022518518518518517,
+      "loss": 0.4743,
+      "step": 438
+    },
+    {
+      "epoch": 0.9739323349972269,
+      "grad_norm": 0.142564058303833,
+      "learning_rate": 0.00022493827160493824,
+      "loss": 0.5764,
+      "step": 439
+    },
+    {
+      "epoch": 0.9761508596783139,
+      "grad_norm": 0.18556907773017883,
+      "learning_rate": 0.00022469135802469134,
+      "loss": 0.7607,
+      "step": 440
+    },
+    {
+      "epoch": 0.978369384359401,
+      "grad_norm": 0.14716467261314392,
+      "learning_rate": 0.0002244444444444444,
+      "loss": 0.5718,
+      "step": 441
+    },
+    {
+      "epoch": 0.9805879090404881,
+      "grad_norm": 0.15531912446022034,
+      "learning_rate": 0.00022419753086419753,
+      "loss": 0.5241,
+      "step": 442
+    },
+    {
+      "epoch": 0.9828064337215752,
+      "grad_norm": 0.16878224909305573,
+      "learning_rate": 0.0002239506172839506,
+      "loss": 0.7798,
+      "step": 443
+    },
+    {
+      "epoch": 0.9850249584026622,
+      "grad_norm": 0.17022493481636047,
+      "learning_rate": 0.0002237037037037037,
+      "loss": 0.5077,
+      "step": 444
+    },
+    {
+      "epoch": 0.9872434830837493,
+      "grad_norm": 0.12562663853168488,
+      "learning_rate": 0.00022345679012345678,
+      "loss": 0.5432,
+      "step": 445
+    },
+    {
+      "epoch": 0.9894620077648364,
+      "grad_norm": 0.17453064024448395,
+      "learning_rate": 0.00022320987654320985,
+      "loss": 0.6335,
+      "step": 446
+    },
+    {
+      "epoch": 0.9916805324459235,
+      "grad_norm": 0.14868859946727753,
+      "learning_rate": 0.00022296296296296295,
+      "loss": 0.5022,
+      "step": 447
+    },
+    {
+      "epoch": 0.9938990571270105,
+      "grad_norm": 0.1363925188779831,
+      "learning_rate": 0.00022271604938271602,
+      "loss": 0.5482,
+      "step": 448
+    },
+    {
+      "epoch": 0.9961175818080976,
+      "grad_norm": 0.30763643980026245,
+      "learning_rate": 0.0002224691358024691,
+      "loss": 0.6983,
+      "step": 449
+    },
+    {
+      "epoch": 0.9983361064891847,
+      "grad_norm": 0.1627693772315979,
+      "learning_rate": 0.00022222222222222218,
+      "loss": 0.6178,
+      "step": 450
+    },
+    {
+      "epoch": 1.0005546311702718,
+      "grad_norm": 0.13489562273025513,
+      "learning_rate": 0.0002219753086419753,
+      "loss": 0.7188,
+      "step": 451
+    },
+    {
+      "epoch": 1.0027731558513588,
+      "grad_norm": 0.14810210466384888,
+      "learning_rate": 0.0002217283950617284,
+      "loss": 0.6466,
+      "step": 452
+    },
+    {
+      "epoch": 1.004991680532446,
+      "grad_norm": 0.16299809515476227,
+      "learning_rate": 0.00022148148148148146,
+      "loss": 0.5169,
+      "step": 453
+    },
+    {
+      "epoch": 1.007210205213533,
+      "grad_norm": 0.10054440051317215,
+      "learning_rate": 0.00022123456790123456,
+      "loss": 0.4629,
+      "step": 454
+    },
+    {
+      "epoch": 1.00942872989462,
+      "grad_norm": 0.13299618661403656,
+      "learning_rate": 0.00022098765432098763,
+      "loss": 0.4716,
+      "step": 455
+    },
+    {
+      "epoch": 1.0116472545757071,
+      "grad_norm": 0.13289417326450348,
+      "learning_rate": 0.00022074074074074072,
+      "loss": 0.6378,
+      "step": 456
+    },
+    {
+      "epoch": 1.0138657792567942,
+      "grad_norm": 0.1984228938817978,
+      "learning_rate": 0.0002204938271604938,
+      "loss": 0.4398,
+      "step": 457
+    },
+    {
+      "epoch": 1.0160843039378813,
+      "grad_norm": 0.12393958866596222,
+      "learning_rate": 0.00022024691358024689,
+      "loss": 0.4637,
+      "step": 458
+    },
+    {
+      "epoch": 1.0183028286189684,
+      "grad_norm": 0.16184231638908386,
+      "learning_rate": 0.00021999999999999995,
+      "loss": 0.5114,
+      "step": 459
+    },
+    {
+      "epoch": 1.0205213533000554,
+      "grad_norm": 0.1200922355055809,
+      "learning_rate": 0.00021975308641975308,
+      "loss": 0.4138,
+      "step": 460
+    },
+    {
+      "epoch": 1.0227398779811425,
+      "grad_norm": 0.1476788967847824,
+      "learning_rate": 0.00021950617283950617,
+      "loss": 0.4778,
+      "step": 461
+    },
+    {
+      "epoch": 1.0249584026622296,
+      "grad_norm": 0.13434381783008575,
+      "learning_rate": 0.00021925925925925924,
+      "loss": 0.6,
+      "step": 462
+    },
+    {
+      "epoch": 1.0271769273433167,
+      "grad_norm": 0.12179043143987656,
+      "learning_rate": 0.00021901234567901233,
+      "loss": 0.4164,
+      "step": 463
+    },
+    {
+      "epoch": 1.0293954520244037,
+      "grad_norm": 0.15724028646945953,
+      "learning_rate": 0.0002187654320987654,
+      "loss": 0.4854,
+      "step": 464
+    },
+    {
+      "epoch": 1.0316139767054908,
+      "grad_norm": 0.1674419641494751,
+      "learning_rate": 0.0002185185185185185,
+      "loss": 0.5142,
+      "step": 465
+    },
+    {
+      "epoch": 1.033832501386578,
+      "grad_norm": 0.15710777044296265,
+      "learning_rate": 0.00021827160493827157,
+      "loss": 0.5381,
+      "step": 466
+    },
+    {
+      "epoch": 1.036051026067665,
+      "grad_norm": 0.19710059463977814,
+      "learning_rate": 0.00021802469135802466,
+      "loss": 0.5977,
+      "step": 467
+    },
+    {
+      "epoch": 1.038269550748752,
+      "grad_norm": 0.23621946573257446,
+      "learning_rate": 0.00021777777777777778,
+      "loss": 0.6843,
+      "step": 468
+    },
+    {
+      "epoch": 1.0404880754298391,
+      "grad_norm": 0.15305998921394348,
+      "learning_rate": 0.00021753086419753085,
+      "loss": 0.5181,
+      "step": 469
+    },
+    {
+      "epoch": 1.0427066001109262,
+      "grad_norm": 0.17592258751392365,
+      "learning_rate": 0.00021728395061728395,
+      "loss": 0.4635,
+      "step": 470
+    },
+    {
+      "epoch": 1.0449251247920133,
+      "grad_norm": 0.1380111128091812,
+      "learning_rate": 0.00021703703703703701,
+      "loss": 0.5177,
+      "step": 471
+    },
+    {
+      "epoch": 1.0471436494731003,
+      "grad_norm": 0.1673838347196579,
+      "learning_rate": 0.0002167901234567901,
+      "loss": 0.6196,
+      "step": 472
+    },
+    {
+      "epoch": 1.0493621741541874,
+      "grad_norm": 0.1779494434595108,
+      "learning_rate": 0.00021654320987654318,
+      "loss": 0.5441,
+      "step": 473
+    },
+    {
+      "epoch": 1.0515806988352745,
+      "grad_norm": 0.15355642139911652,
+      "learning_rate": 0.00021629629629629627,
+      "loss": 0.6082,
+      "step": 474
+    },
+    {
+      "epoch": 1.0537992235163616,
+      "grad_norm": 0.21717680990695953,
+      "learning_rate": 0.00021604938271604937,
+      "loss": 0.534,
+      "step": 475
+    },
+    {
+      "epoch": 1.0560177481974486,
+      "grad_norm": 0.20280776917934418,
+      "learning_rate": 0.00021580246913580244,
+      "loss": 0.5772,
+      "step": 476
+    },
+    {
+      "epoch": 1.0582362728785357,
+      "grad_norm": 0.16771502792835236,
+      "learning_rate": 0.00021555555555555556,
+      "loss": 0.3689,
+      "step": 477
+    },
+    {
+      "epoch": 1.0604547975596228,
+      "grad_norm": 0.15932482481002808,
+      "learning_rate": 0.00021530864197530863,
+      "loss": 0.4394,
+      "step": 478
+    },
+    {
+      "epoch": 1.0626733222407099,
+      "grad_norm": 0.15166310966014862,
+      "learning_rate": 0.00021506172839506172,
+      "loss": 0.4707,
+      "step": 479
+    },
+    {
+      "epoch": 1.064891846921797,
+      "grad_norm": 0.2376459687948227,
+      "learning_rate": 0.0002148148148148148,
+      "loss": 0.6653,
+      "step": 480
+    },
+    {
+      "epoch": 1.067110371602884,
+      "grad_norm": 0.1822025328874588,
+      "learning_rate": 0.00021456790123456788,
+      "loss": 0.4809,
+      "step": 481
+    },
+    {
+      "epoch": 1.069328896283971,
+      "grad_norm": 0.15670810639858246,
+      "learning_rate": 0.00021432098765432095,
+      "loss": 0.6046,
+      "step": 482
+    },
+    {
+      "epoch": 1.0715474209650582,
+      "grad_norm": 0.18272632360458374,
+      "learning_rate": 0.00021407407407407405,
+      "loss": 0.7191,
+      "step": 483
+    },
+    {
+      "epoch": 1.0737659456461452,
+      "grad_norm": 0.18568427860736847,
+      "learning_rate": 0.00021382716049382714,
+      "loss": 0.5737,
+      "step": 484
+    },
+    {
+      "epoch": 1.0759844703272323,
+      "grad_norm": 0.2356289178133011,
+      "learning_rate": 0.0002135802469135802,
+      "loss": 0.4498,
+      "step": 485
+    },
+    {
+      "epoch": 1.0782029950083194,
+      "grad_norm": 0.14072103798389435,
+      "learning_rate": 0.00021333333333333333,
+      "loss": 0.4929,
+      "step": 486
+    },
+    {
+      "epoch": 1.0804215196894065,
+      "grad_norm": 0.22377143800258636,
+      "learning_rate": 0.0002130864197530864,
+      "loss": 0.4213,
+      "step": 487
+    },
+    {
+      "epoch": 1.0826400443704935,
+      "grad_norm": 0.15749813616275787,
+      "learning_rate": 0.0002128395061728395,
+      "loss": 0.592,
+      "step": 488
+    },
+    {
+      "epoch": 1.0848585690515806,
+      "grad_norm": 0.1740855723619461,
+      "learning_rate": 0.00021259259259259256,
+      "loss": 0.8171,
+      "step": 489
+    },
+    {
+      "epoch": 1.0870770937326677,
+      "grad_norm": 0.16143059730529785,
+      "learning_rate": 0.00021234567901234566,
+      "loss": 0.6868,
+      "step": 490
+    },
+    {
+      "epoch": 1.0892956184137548,
+      "grad_norm": 0.22964100539684296,
+      "learning_rate": 0.00021209876543209875,
+      "loss": 0.73,
+      "step": 491
+    },
+    {
+      "epoch": 1.0915141430948418,
+      "grad_norm": 0.12745749950408936,
+      "learning_rate": 0.00021185185185185182,
+      "loss": 0.5179,
+      "step": 492
+    },
+    {
+      "epoch": 1.093732667775929,
+      "grad_norm": 0.1981736719608307,
+      "learning_rate": 0.00021160493827160492,
+      "loss": 0.6771,
+      "step": 493
+    },
+    {
+      "epoch": 1.095951192457016,
+      "grad_norm": 0.20737721025943756,
+      "learning_rate": 0.00021135802469135799,
+      "loss": 0.563,
+      "step": 494
+    },
+    {
+      "epoch": 1.098169717138103,
+      "grad_norm": 0.1787654459476471,
+      "learning_rate": 0.0002111111111111111,
+      "loss": 0.4357,
+      "step": 495
+    },
+    {
+      "epoch": 1.1003882418191901,
+      "grad_norm": 0.1449396312236786,
+      "learning_rate": 0.00021086419753086418,
+      "loss": 0.6308,
+      "step": 496
+    },
+    {
+      "epoch": 1.1026067665002772,
+      "grad_norm": 0.1525779515504837,
+      "learning_rate": 0.00021061728395061727,
+      "loss": 0.5908,
+      "step": 497
+    },
+    {
+      "epoch": 1.1048252911813643,
+      "grad_norm": 0.15887467563152313,
+      "learning_rate": 0.00021037037037037037,
+      "loss": 0.5314,
+      "step": 498
+    },
+    {
+      "epoch": 1.1070438158624514,
+      "grad_norm": 0.2404686063528061,
+      "learning_rate": 0.00021012345679012343,
+      "loss": 0.5637,
+      "step": 499
+    },
+    {
+      "epoch": 1.1092623405435384,
+      "grad_norm": 0.1815386265516281,
+      "learning_rate": 0.00020987654320987653,
+      "loss": 0.5867,
+      "step": 500
+    },
+    {
+      "epoch": 1.1114808652246255,
+      "grad_norm": 0.20449085533618927,
+      "learning_rate": 0.0002096296296296296,
+      "loss": 0.4907,
+      "step": 501
+    },
+    {
+      "epoch": 1.1136993899057126,
+      "grad_norm": 0.16751590371131897,
+      "learning_rate": 0.0002093827160493827,
+      "loss": 0.4946,
+      "step": 502
+    },
+    {
+      "epoch": 1.1159179145867997,
+      "grad_norm": 0.15466222167015076,
+      "learning_rate": 0.00020913580246913576,
+      "loss": 0.5879,
+      "step": 503
+    },
+    {
+      "epoch": 1.1181364392678868,
+      "grad_norm": 0.22958067059516907,
+      "learning_rate": 0.00020888888888888888,
+      "loss": 0.5836,
+      "step": 504
+    },
+    {
+      "epoch": 1.1203549639489738,
+      "grad_norm": 0.15312771499156952,
+      "learning_rate": 0.00020864197530864198,
+      "loss": 0.5132,
+      "step": 505
+    },
+    {
+      "epoch": 1.122573488630061,
+      "grad_norm": 0.18419933319091797,
+      "learning_rate": 0.00020839506172839505,
+      "loss": 0.4928,
+      "step": 506
+    },
+    {
+      "epoch": 1.124792013311148,
+      "grad_norm": 0.17035695910453796,
+      "learning_rate": 0.00020814814814814814,
+      "loss": 0.5838,
+      "step": 507
+    },
+    {
+      "epoch": 1.127010537992235,
+      "grad_norm": 0.19350723922252655,
+      "learning_rate": 0.0002079012345679012,
+      "loss": 0.5087,
+      "step": 508
+    },
+    {
+      "epoch": 1.1292290626733221,
+      "grad_norm": 0.15138496458530426,
+      "learning_rate": 0.0002076543209876543,
+      "loss": 0.4444,
+      "step": 509
+    },
+    {
+      "epoch": 1.1314475873544092,
+      "grad_norm": 0.16679179668426514,
+      "learning_rate": 0.00020740740740740737,
+      "loss": 0.6148,
+      "step": 510
+    },
+    {
+      "epoch": 1.1336661120354963,
+      "grad_norm": 0.1691301017999649,
+      "learning_rate": 0.00020716049382716047,
+      "loss": 0.6405,
+      "step": 511
+    },
+    {
+      "epoch": 1.1358846367165834,
+      "grad_norm": 0.19456829130649567,
+      "learning_rate": 0.00020691358024691354,
+      "loss": 0.7303,
+      "step": 512
+    },
+    {
+      "epoch": 1.1381031613976704,
+      "grad_norm": 0.192270427942276,
+      "learning_rate": 0.00020666666666666666,
+      "loss": 0.6111,
+      "step": 513
+    },
+    {
+      "epoch": 1.1403216860787575,
+      "grad_norm": 0.1941678822040558,
+      "learning_rate": 0.00020641975308641975,
+      "loss": 0.4374,
+      "step": 514
+    },
+    {
+      "epoch": 1.1425402107598448,
+      "grad_norm": 0.2405455857515335,
+      "learning_rate": 0.00020617283950617282,
+      "loss": 0.6127,
+      "step": 515
+    },
+    {
+      "epoch": 1.1447587354409319,
+      "grad_norm": 0.17750173807144165,
+      "learning_rate": 0.00020592592592592592,
+      "loss": 0.5969,
+      "step": 516
+    },
+    {
+      "epoch": 1.146977260122019,
+      "grad_norm": 0.20301392674446106,
+      "learning_rate": 0.00020567901234567898,
+      "loss": 0.57,
+      "step": 517
+    },
+    {
+      "epoch": 1.149195784803106,
+      "grad_norm": 0.24414236843585968,
+      "learning_rate": 0.00020543209876543208,
+      "loss": 0.5735,
+      "step": 518
+    },
+    {
+      "epoch": 1.151414309484193,
+      "grad_norm": 0.21640996634960175,
+      "learning_rate": 0.00020518518518518515,
+      "loss": 0.6374,
+      "step": 519
+    },
+    {
+      "epoch": 1.1536328341652802,
+      "grad_norm": 0.26867935061454773,
+      "learning_rate": 0.00020493827160493824,
+      "loss": 0.4562,
+      "step": 520
+    },
+    {
+      "epoch": 1.1558513588463672,
+      "grad_norm": 0.14391440153121948,
+      "learning_rate": 0.00020469135802469136,
+      "loss": 0.5391,
+      "step": 521
+    },
+    {
+      "epoch": 1.1580698835274543,
+      "grad_norm": 0.13912127912044525,
+      "learning_rate": 0.00020444444444444443,
+      "loss": 0.5713,
+      "step": 522
+    },
+    {
+      "epoch": 1.1602884082085414,
+      "grad_norm": 0.16650395095348358,
+      "learning_rate": 0.00020419753086419753,
+      "loss": 0.5275,
+      "step": 523
+    },
+    {
+      "epoch": 1.1625069328896285,
+      "grad_norm": 0.18671400845050812,
+      "learning_rate": 0.0002039506172839506,
+      "loss": 0.4663,
+      "step": 524
+    },
+    {
+      "epoch": 1.1647254575707155,
+      "grad_norm": 0.21383784711360931,
+      "learning_rate": 0.0002037037037037037,
+      "loss": 0.6219,
+      "step": 525
+    },
+    {
+      "epoch": 1.1669439822518026,
+      "grad_norm": 0.16587862372398376,
+      "learning_rate": 0.00020345679012345676,
+      "loss": 0.612,
+      "step": 526
+    },
+    {
+      "epoch": 1.1691625069328897,
+      "grad_norm": 0.16420531272888184,
+      "learning_rate": 0.00020320987654320985,
+      "loss": 0.4516,
+      "step": 527
+    },
+    {
+      "epoch": 1.1713810316139768,
+      "grad_norm": 0.15526890754699707,
+      "learning_rate": 0.00020296296296296295,
+      "loss": 0.5065,
+      "step": 528
+    },
+    {
+      "epoch": 1.1735995562950639,
+      "grad_norm": 0.21110498905181885,
+      "learning_rate": 0.00020271604938271602,
+      "loss": 0.667,
+      "step": 529
+    },
+    {
+      "epoch": 1.175818080976151,
+      "grad_norm": 0.17298609018325806,
+      "learning_rate": 0.00020246913580246914,
+      "loss": 0.4872,
+      "step": 530
+    },
+    {
+      "epoch": 1.178036605657238,
+      "grad_norm": 0.2367205172777176,
+      "learning_rate": 0.0002022222222222222,
+      "loss": 0.5769,
+      "step": 531
+    },
+    {
+      "epoch": 1.180255130338325,
+      "grad_norm": 0.18996378779411316,
+      "learning_rate": 0.0002019753086419753,
+      "loss": 0.5112,
+      "step": 532
+    },
+    {
+      "epoch": 1.1824736550194122,
+      "grad_norm": 0.31726616621017456,
+      "learning_rate": 0.00020172839506172837,
+      "loss": 0.5464,
+      "step": 533
+    },
+    {
+      "epoch": 1.1846921797004992,
+      "grad_norm": 0.15378770232200623,
+      "learning_rate": 0.00020148148148148147,
+      "loss": 0.4171,
+      "step": 534
+    },
+    {
+      "epoch": 1.1869107043815863,
+      "grad_norm": 0.1976374238729477,
+      "learning_rate": 0.00020123456790123456,
+      "loss": 0.5448,
+      "step": 535
+    },
+    {
+      "epoch": 1.1891292290626734,
+      "grad_norm": 0.1982663869857788,
+      "learning_rate": 0.00020098765432098763,
+      "loss": 0.4871,
+      "step": 536
+    },
+    {
+      "epoch": 1.1913477537437605,
+      "grad_norm": 0.21939164400100708,
+      "learning_rate": 0.00020074074074074072,
+      "loss": 0.5285,
+      "step": 537
+    },
+    {
+      "epoch": 1.1935662784248475,
+      "grad_norm": 0.2253114879131317,
+      "learning_rate": 0.0002004938271604938,
+      "loss": 0.6108,
+      "step": 538
+    },
+    {
+      "epoch": 1.1957848031059346,
+      "grad_norm": 0.18163472414016724,
+      "learning_rate": 0.00020024691358024691,
+      "loss": 0.4937,
+      "step": 539
+    },
+    {
+      "epoch": 1.1980033277870217,
+      "grad_norm": 0.4652941823005676,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 0.524,
+      "step": 540
+    },
+    {
+      "epoch": 1.2002218524681088,
+      "grad_norm": 0.15802715718746185,
+      "learning_rate": 0.00019975308641975308,
+      "loss": 0.5249,
+      "step": 541
+    },
+    {
+      "epoch": 1.2024403771491958,
+      "grad_norm": 0.1567433774471283,
+      "learning_rate": 0.00019950617283950615,
+      "loss": 0.4851,
+      "step": 542
+    },
+    {
+      "epoch": 1.204658901830283,
+      "grad_norm": 0.16850852966308594,
+      "learning_rate": 0.00019925925925925924,
+      "loss": 0.5605,
+      "step": 543
+    },
+    {
+      "epoch": 1.20687742651137,
+      "grad_norm": 0.20014235377311707,
+      "learning_rate": 0.00019901234567901234,
+      "loss": 0.452,
+      "step": 544
+    },
+    {
+      "epoch": 1.209095951192457,
+      "grad_norm": 0.19419488310813904,
+      "learning_rate": 0.0001987654320987654,
+      "loss": 0.481,
+      "step": 545
+    },
+    {
+      "epoch": 1.2113144758735441,
+      "grad_norm": 0.18302187323570251,
+      "learning_rate": 0.0001985185185185185,
+      "loss": 0.412,
+      "step": 546
+    },
+    {
+      "epoch": 1.2135330005546312,
+      "grad_norm": 0.19586671888828278,
+      "learning_rate": 0.00019827160493827157,
+      "loss": 0.5595,
+      "step": 547
+    },
+    {
+      "epoch": 1.2157515252357183,
+      "grad_norm": 0.22752362489700317,
+      "learning_rate": 0.0001980246913580247,
+      "loss": 0.4988,
+      "step": 548
+    },
+    {
+      "epoch": 1.2179700499168054,
+      "grad_norm": 0.20472568273544312,
+      "learning_rate": 0.00019777777777777776,
+      "loss": 0.6569,
+      "step": 549
+    },
+    {
+      "epoch": 1.2201885745978924,
+      "grad_norm": 0.1821528822183609,
+      "learning_rate": 0.00019753086419753085,
+      "loss": 0.6787,
+      "step": 550
+    },
+    {
+      "epoch": 1.2224070992789795,
+      "grad_norm": 0.21358664333820343,
+      "learning_rate": 0.00019728395061728395,
+      "loss": 0.6277,
+      "step": 551
+    },
+    {
+      "epoch": 1.2246256239600666,
+      "grad_norm": 0.2970047891139984,
+      "learning_rate": 0.00019703703703703702,
+      "loss": 0.5562,
+      "step": 552
+    },
+    {
+      "epoch": 1.2268441486411537,
+      "grad_norm": 0.16585226356983185,
+      "learning_rate": 0.0001967901234567901,
+      "loss": 0.5253,
+      "step": 553
+    },
+    {
+      "epoch": 1.2290626733222407,
+      "grad_norm": 0.16079090535640717,
+      "learning_rate": 0.00019654320987654318,
+      "loss": 0.5555,
+      "step": 554
+    },
+    {
+      "epoch": 1.2312811980033278,
+      "grad_norm": 0.17712751030921936,
+      "learning_rate": 0.00019629629629629627,
+      "loss": 0.6251,
+      "step": 555
+    },
+    {
+      "epoch": 1.2334997226844149,
+      "grad_norm": 0.16909275949001312,
+      "learning_rate": 0.00019604938271604934,
+      "loss": 0.6163,
+      "step": 556
+    },
+    {
+      "epoch": 1.235718247365502,
+      "grad_norm": 0.18430787324905396,
+      "learning_rate": 0.00019580246913580246,
+      "loss": 0.5248,
+      "step": 557
+    },
+    {
+      "epoch": 1.237936772046589,
+      "grad_norm": 0.17066165804862976,
+      "learning_rate": 0.00019555555555555556,
+      "loss": 0.4817,
+      "step": 558
+    },
+    {
+      "epoch": 1.240155296727676,
+      "grad_norm": 0.15731728076934814,
+      "learning_rate": 0.00019530864197530863,
+      "loss": 0.5386,
+      "step": 559
+    },
+    {
+      "epoch": 1.2423738214087632,
+      "grad_norm": 0.16297486424446106,
+      "learning_rate": 0.00019506172839506172,
+      "loss": 0.5151,
+      "step": 560
+    },
+    {
+      "epoch": 1.2445923460898503,
+      "grad_norm": 0.18040071427822113,
+      "learning_rate": 0.0001948148148148148,
+      "loss": 0.4657,
+      "step": 561
+    },
+    {
+      "epoch": 1.2468108707709373,
+      "grad_norm": 0.19494326412677765,
+      "learning_rate": 0.00019456790123456789,
+      "loss": 0.5594,
+      "step": 562
+    },
+    {
+      "epoch": 1.2490293954520244,
+      "grad_norm": 0.17954906821250916,
+      "learning_rate": 0.00019432098765432095,
+      "loss": 0.6626,
+      "step": 563
+    },
+    {
+      "epoch": 1.2512479201331115,
+      "grad_norm": 0.19580917060375214,
+      "learning_rate": 0.00019407407407407405,
+      "loss": 0.5956,
+      "step": 564
+    },
+    {
+      "epoch": 1.2534664448141986,
+      "grad_norm": 0.1645064800977707,
+      "learning_rate": 0.00019382716049382712,
+      "loss": 0.504,
+      "step": 565
+    },
+    {
+      "epoch": 1.2556849694952856,
+      "grad_norm": 0.18674708902835846,
+      "learning_rate": 0.00019358024691358024,
+      "loss": 0.6956,
+      "step": 566
+    },
+    {
+      "epoch": 1.2579034941763727,
+      "grad_norm": 0.15892228484153748,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 0.5789,
+      "step": 567
+    },
+    {
+      "epoch": 1.2601220188574598,
+      "grad_norm": 0.15037551522254944,
+      "learning_rate": 0.0001930864197530864,
+      "loss": 0.5831,
+      "step": 568
+    },
+    {
+      "epoch": 1.2623405435385469,
+      "grad_norm": 0.1267644315958023,
+      "learning_rate": 0.0001928395061728395,
+      "loss": 0.4957,
+      "step": 569
+    },
+    {
+      "epoch": 1.264559068219634,
+      "grad_norm": 0.16408823430538177,
+      "learning_rate": 0.00019259259259259257,
+      "loss": 0.4809,
+      "step": 570
+    },
+    {
+      "epoch": 1.266777592900721,
+      "grad_norm": 0.17988921701908112,
+      "learning_rate": 0.00019234567901234566,
+      "loss": 0.5296,
+      "step": 571
+    },
+    {
+      "epoch": 1.268996117581808,
+      "grad_norm": 0.20517736673355103,
+      "learning_rate": 0.00019209876543209873,
+      "loss": 0.5896,
+      "step": 572
+    },
+    {
+      "epoch": 1.2712146422628952,
+      "grad_norm": 0.192581444978714,
+      "learning_rate": 0.00019185185185185182,
+      "loss": 0.6183,
+      "step": 573
+    },
+    {
+      "epoch": 1.2734331669439822,
+      "grad_norm": 0.20094512403011322,
+      "learning_rate": 0.00019160493827160495,
+      "loss": 0.541,
+      "step": 574
+    },
+    {
+      "epoch": 1.2756516916250693,
+      "grad_norm": 0.19196541607379913,
+      "learning_rate": 0.00019135802469135801,
+      "loss": 0.6753,
+      "step": 575
+    },
+    {
+      "epoch": 1.2778702163061564,
+      "grad_norm": 0.1585908979177475,
+      "learning_rate": 0.0001911111111111111,
+      "loss": 0.4932,
+      "step": 576
+    },
+    {
+      "epoch": 1.2800887409872435,
+      "grad_norm": 0.1960020810365677,
+      "learning_rate": 0.00019086419753086418,
+      "loss": 0.5741,
+      "step": 577
+    },
+    {
+      "epoch": 1.2823072656683305,
+      "grad_norm": 0.15070077776908875,
+      "learning_rate": 0.00019061728395061727,
+      "loss": 0.5684,
+      "step": 578
+    },
+    {
+      "epoch": 1.2845257903494176,
+      "grad_norm": 0.15428100526332855,
+      "learning_rate": 0.00019037037037037034,
+      "loss": 0.4767,
+      "step": 579
+    },
+    {
+      "epoch": 1.2867443150305047,
+      "grad_norm": 0.2502506375312805,
+      "learning_rate": 0.00019012345679012344,
+      "loss": 0.6736,
+      "step": 580
+    },
+    {
+      "epoch": 1.2889628397115918,
+      "grad_norm": 0.15290825068950653,
+      "learning_rate": 0.00018987654320987653,
+      "loss": 0.6992,
+      "step": 581
+    },
+    {
+      "epoch": 1.2911813643926788,
+      "grad_norm": 0.16831225156784058,
+      "learning_rate": 0.0001896296296296296,
+      "loss": 0.4834,
+      "step": 582
+    },
+    {
+      "epoch": 1.293399889073766,
+      "grad_norm": 0.1888023018836975,
+      "learning_rate": 0.00018938271604938272,
+      "loss": 0.4217,
+      "step": 583
+    },
+    {
+      "epoch": 1.295618413754853,
+      "grad_norm": 0.21119245886802673,
+      "learning_rate": 0.0001891358024691358,
+      "loss": 0.5473,
+      "step": 584
+    },
+    {
+      "epoch": 1.29783693843594,
+      "grad_norm": 0.20793120563030243,
+      "learning_rate": 0.00018888888888888888,
+      "loss": 0.6137,
+      "step": 585
+    },
+    {
+      "epoch": 1.3000554631170271,
+      "grad_norm": 0.19353698194026947,
+      "learning_rate": 0.00018864197530864195,
+      "loss": 0.4787,
+      "step": 586
+    },
+    {
+      "epoch": 1.3022739877981142,
+      "grad_norm": 0.1978445202112198,
+      "learning_rate": 0.00018839506172839505,
+      "loss": 0.4305,
+      "step": 587
+    },
+    {
+      "epoch": 1.3044925124792013,
+      "grad_norm": 0.1779300570487976,
+      "learning_rate": 0.00018814814814814814,
+      "loss": 0.5827,
+      "step": 588
+    },
+    {
+      "epoch": 1.3067110371602884,
+      "grad_norm": 0.1878674179315567,
+      "learning_rate": 0.0001879012345679012,
+      "loss": 0.5607,
+      "step": 589
+    },
+    {
+      "epoch": 1.3089295618413754,
+      "grad_norm": 0.19992372393608093,
+      "learning_rate": 0.0001876543209876543,
+      "loss": 0.5274,
+      "step": 590
+    },
+    {
+      "epoch": 1.3111480865224625,
+      "grad_norm": 0.17217043042182922,
+      "learning_rate": 0.00018740740740740737,
+      "loss": 0.4119,
+      "step": 591
+    },
+    {
+      "epoch": 1.3133666112035496,
+      "grad_norm": 0.30168217420578003,
+      "learning_rate": 0.0001871604938271605,
+      "loss": 0.6624,
+      "step": 592
+    },
+    {
+      "epoch": 1.3155851358846367,
+      "grad_norm": 0.15776364505290985,
+      "learning_rate": 0.00018691358024691356,
+      "loss": 0.5043,
+      "step": 593
+    },
+    {
+      "epoch": 1.3178036605657237,
+      "grad_norm": 0.19596697390079498,
+      "learning_rate": 0.00018666666666666666,
+      "loss": 0.7099,
+      "step": 594
+    },
+    {
+      "epoch": 1.3200221852468108,
+      "grad_norm": 0.16606877744197845,
+      "learning_rate": 0.00018641975308641973,
+      "loss": 0.7678,
+      "step": 595
+    },
+    {
+      "epoch": 1.3222407099278979,
+      "grad_norm": 0.17500612139701843,
+      "learning_rate": 0.00018617283950617282,
+      "loss": 0.4475,
+      "step": 596
+    },
+    {
+      "epoch": 1.324459234608985,
+      "grad_norm": 0.18675503134727478,
+      "learning_rate": 0.00018592592592592592,
+      "loss": 0.4391,
+      "step": 597
+    },
+    {
+      "epoch": 1.326677759290072,
+      "grad_norm": 0.20592869818210602,
+      "learning_rate": 0.00018567901234567899,
+      "loss": 0.5251,
+      "step": 598
+    },
+    {
+      "epoch": 1.328896283971159,
+      "grad_norm": 0.18670955300331116,
+      "learning_rate": 0.00018543209876543208,
+      "loss": 0.3769,
+      "step": 599
+    },
+    {
+      "epoch": 1.3311148086522462,
+      "grad_norm": 0.19756749272346497,
+      "learning_rate": 0.00018518518518518515,
+      "loss": 0.6092,
+      "step": 600
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.20875504612922668,
+      "learning_rate": 0.00018493827160493827,
+      "loss": 0.641,
+      "step": 601
+    },
+    {
+      "epoch": 1.3355518580144203,
+      "grad_norm": 0.2367030531167984,
+      "learning_rate": 0.00018469135802469134,
+      "loss": 0.6319,
+      "step": 602
+    },
+    {
+      "epoch": 1.3377703826955074,
+      "grad_norm": 0.1682172417640686,
+      "learning_rate": 0.00018444444444444443,
+      "loss": 0.4942,
+      "step": 603
+    },
+    {
+      "epoch": 1.3399889073765945,
+      "grad_norm": 0.17947247624397278,
+      "learning_rate": 0.00018419753086419753,
+      "loss": 0.5567,
+      "step": 604
+    },
+    {
+      "epoch": 1.3422074320576816,
+      "grad_norm": 0.20006439089775085,
+      "learning_rate": 0.0001839506172839506,
+      "loss": 0.5632,
+      "step": 605
+    },
+    {
+      "epoch": 1.3444259567387689,
+      "grad_norm": 0.2004607766866684,
+      "learning_rate": 0.0001837037037037037,
+      "loss": 0.6892,
+      "step": 606
+    },
+    {
+      "epoch": 1.346644481419856,
+      "grad_norm": 0.18117636442184448,
+      "learning_rate": 0.00018345679012345676,
+      "loss": 0.6597,
+      "step": 607
+    },
+    {
+      "epoch": 1.348863006100943,
+      "grad_norm": 0.1746131330728531,
+      "learning_rate": 0.00018320987654320986,
+      "loss": 0.6014,
+      "step": 608
+    },
+    {
+      "epoch": 1.35108153078203,
+      "grad_norm": 0.16680322587490082,
+      "learning_rate": 0.00018296296296296292,
+      "loss": 0.6714,
+      "step": 609
+    },
+    {
+      "epoch": 1.3533000554631172,
+      "grad_norm": 0.1810801476240158,
+      "learning_rate": 0.00018271604938271605,
+      "loss": 0.5316,
+      "step": 610
+    },
+    {
+      "epoch": 1.3555185801442042,
+      "grad_norm": 0.15907979011535645,
+      "learning_rate": 0.00018246913580246914,
+      "loss": 0.3918,
+      "step": 611
+    },
+    {
+      "epoch": 1.3577371048252913,
+      "grad_norm": 0.241806760430336,
+      "learning_rate": 0.0001822222222222222,
+      "loss": 0.5185,
+      "step": 612
+    },
+    {
+      "epoch": 1.3599556295063784,
+      "grad_norm": 0.20513887703418732,
+      "learning_rate": 0.0001819753086419753,
+      "loss": 0.5772,
+      "step": 613
+    },
+    {
+      "epoch": 1.3621741541874655,
+      "grad_norm": 0.17053154110908508,
+      "learning_rate": 0.00018172839506172837,
+      "loss": 0.5316,
+      "step": 614
+    },
+    {
+      "epoch": 1.3643926788685525,
+      "grad_norm": 0.17985829710960388,
+      "learning_rate": 0.00018148148148148147,
+      "loss": 0.4912,
+      "step": 615
+    },
+    {
+      "epoch": 1.3666112035496396,
+      "grad_norm": 0.176662415266037,
+      "learning_rate": 0.00018123456790123454,
+      "loss": 0.4729,
+      "step": 616
+    },
+    {
+      "epoch": 1.3688297282307267,
+      "grad_norm": 0.24574291706085205,
+      "learning_rate": 0.00018098765432098763,
+      "loss": 0.6835,
+      "step": 617
+    },
+    {
+      "epoch": 1.3710482529118138,
+      "grad_norm": 0.16730976104736328,
+      "learning_rate": 0.00018074074074074075,
+      "loss": 0.5616,
+      "step": 618
+    },
+    {
+      "epoch": 1.3732667775929008,
+      "grad_norm": 0.22502173483371735,
+      "learning_rate": 0.00018049382716049382,
+      "loss": 0.5315,
+      "step": 619
+    },
+    {
+      "epoch": 1.375485302273988,
+      "grad_norm": 0.23504018783569336,
+      "learning_rate": 0.00018024691358024692,
+      "loss": 0.4912,
+      "step": 620
+    },
+    {
+      "epoch": 1.377703826955075,
+      "grad_norm": 0.17786575853824615,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.669,
+      "step": 621
+    },
+    {
+      "epoch": 1.379922351636162,
+      "grad_norm": 0.16418124735355377,
+      "learning_rate": 0.00017975308641975308,
+      "loss": 0.4058,
+      "step": 622
+    },
+    {
+      "epoch": 1.3821408763172491,
+      "grad_norm": 0.23493215441703796,
+      "learning_rate": 0.00017950617283950615,
+      "loss": 0.6309,
+      "step": 623
+    },
+    {
+      "epoch": 1.3843594009983362,
+      "grad_norm": 0.17955836653709412,
+      "learning_rate": 0.00017925925925925924,
+      "loss": 0.5325,
+      "step": 624
+    },
+    {
+      "epoch": 1.3865779256794233,
+      "grad_norm": 0.19326384365558624,
+      "learning_rate": 0.0001790123456790123,
+      "loss": 0.5129,
+      "step": 625
+    },
+    {
+      "epoch": 1.3887964503605104,
+      "grad_norm": 0.17727799713611603,
+      "learning_rate": 0.0001787654320987654,
+      "loss": 0.3935,
+      "step": 626
+    },
+    {
+      "epoch": 1.3910149750415974,
+      "grad_norm": 0.1752282977104187,
+      "learning_rate": 0.00017851851851851853,
+      "loss": 0.5223,
+      "step": 627
+    },
+    {
+      "epoch": 1.3932334997226845,
+      "grad_norm": 0.19551189243793488,
+      "learning_rate": 0.0001782716049382716,
+      "loss": 0.455,
+      "step": 628
+    },
+    {
+      "epoch": 1.3954520244037716,
+      "grad_norm": 0.29861530661582947,
+      "learning_rate": 0.0001780246913580247,
+      "loss": 0.529,
+      "step": 629
+    },
+    {
+      "epoch": 1.3976705490848587,
+      "grad_norm": 0.19856123626232147,
+      "learning_rate": 0.00017777777777777776,
+      "loss": 0.4982,
+      "step": 630
+    },
+    {
+      "epoch": 1.3998890737659457,
+      "grad_norm": 0.2507825195789337,
+      "learning_rate": 0.00017753086419753085,
+      "loss": 0.5123,
+      "step": 631
+    },
+    {
+      "epoch": 1.4021075984470328,
+      "grad_norm": 0.19093061983585358,
+      "learning_rate": 0.00017728395061728392,
+      "loss": 0.4846,
+      "step": 632
+    },
+    {
+      "epoch": 1.4043261231281199,
+      "grad_norm": 0.19345994293689728,
+      "learning_rate": 0.00017703703703703702,
+      "loss": 0.5509,
+      "step": 633
+    },
+    {
+      "epoch": 1.406544647809207,
+      "grad_norm": 0.2659754157066345,
+      "learning_rate": 0.0001767901234567901,
+      "loss": 0.5395,
+      "step": 634
+    },
+    {
+      "epoch": 1.408763172490294,
+      "grad_norm": 0.21121443808078766,
+      "learning_rate": 0.00017654320987654318,
+      "loss": 0.4654,
+      "step": 635
+    },
+    {
+      "epoch": 1.410981697171381,
+      "grad_norm": 0.20511746406555176,
+      "learning_rate": 0.0001762962962962963,
+      "loss": 0.5407,
+      "step": 636
+    },
+    {
+      "epoch": 1.4132002218524682,
+      "grad_norm": 0.1723611056804657,
+      "learning_rate": 0.00017604938271604937,
+      "loss": 0.4959,
+      "step": 637
+    },
+    {
+      "epoch": 1.4154187465335553,
+      "grad_norm": 0.22937335073947906,
+      "learning_rate": 0.00017580246913580247,
+      "loss": 0.5831,
+      "step": 638
+    },
+    {
+      "epoch": 1.4176372712146423,
+      "grad_norm": 0.28765061497688293,
+      "learning_rate": 0.00017555555555555553,
+      "loss": 0.589,
+      "step": 639
+    },
+    {
+      "epoch": 1.4198557958957294,
+      "grad_norm": 0.27344661951065063,
+      "learning_rate": 0.00017530864197530863,
+      "loss": 0.5078,
+      "step": 640
+    },
+    {
+      "epoch": 1.4220743205768165,
+      "grad_norm": 0.21969613432884216,
+      "learning_rate": 0.00017506172839506172,
+      "loss": 0.5247,
+      "step": 641
+    },
+    {
+      "epoch": 1.4242928452579036,
+      "grad_norm": 0.18761585652828217,
+      "learning_rate": 0.0001748148148148148,
+      "loss": 0.4898,
+      "step": 642
+    },
+    {
+      "epoch": 1.4265113699389906,
+      "grad_norm": 0.20467689633369446,
+      "learning_rate": 0.0001745679012345679,
+      "loss": 0.4361,
+      "step": 643
+    },
+    {
+      "epoch": 1.4287298946200777,
+      "grad_norm": 0.24104191362857819,
+      "learning_rate": 0.00017432098765432096,
+      "loss": 0.6436,
+      "step": 644
+    },
+    {
+      "epoch": 1.4309484193011648,
+      "grad_norm": 0.18173179030418396,
+      "learning_rate": 0.00017407407407407408,
+      "loss": 0.5,
+      "step": 645
+    },
+    {
+      "epoch": 1.4331669439822519,
+      "grad_norm": 0.20782898366451263,
+      "learning_rate": 0.00017382716049382715,
+      "loss": 0.496,
+      "step": 646
+    },
+    {
+      "epoch": 1.435385468663339,
+      "grad_norm": 0.16304484009742737,
+      "learning_rate": 0.00017358024691358024,
+      "loss": 0.4977,
+      "step": 647
+    },
+    {
+      "epoch": 1.437603993344426,
+      "grad_norm": 0.21319276094436646,
+      "learning_rate": 0.0001733333333333333,
+      "loss": 0.454,
+      "step": 648
+    },
+    {
+      "epoch": 1.439822518025513,
+      "grad_norm": 0.17866571247577667,
+      "learning_rate": 0.0001730864197530864,
+      "loss": 0.5325,
+      "step": 649
+    },
+    {
+      "epoch": 1.4420410427066002,
+      "grad_norm": 0.22857487201690674,
+      "learning_rate": 0.0001728395061728395,
+      "loss": 0.4672,
+      "step": 650
+    },
+    {
+      "epoch": 1.4442595673876872,
+      "grad_norm": 0.2302277386188507,
+      "learning_rate": 0.00017259259259259257,
+      "loss": 0.542,
+      "step": 651
+    },
+    {
+      "epoch": 1.4464780920687743,
+      "grad_norm": 0.2412882149219513,
+      "learning_rate": 0.00017234567901234566,
+      "loss": 0.5419,
+      "step": 652
+    },
+    {
+      "epoch": 1.4486966167498614,
+      "grad_norm": 0.22220225632190704,
+      "learning_rate": 0.00017209876543209873,
+      "loss": 0.4274,
+      "step": 653
+    },
+    {
+      "epoch": 1.4509151414309485,
+      "grad_norm": 0.22964227199554443,
+      "learning_rate": 0.00017185185185185185,
+      "loss": 0.4719,
+      "step": 654
+    },
+    {
+      "epoch": 1.4531336661120355,
+      "grad_norm": 0.1967972368001938,
+      "learning_rate": 0.00017160493827160492,
+      "loss": 0.5533,
+      "step": 655
+    },
+    {
+      "epoch": 1.4553521907931226,
+      "grad_norm": 0.2344427853822708,
+      "learning_rate": 0.00017135802469135802,
+      "loss": 0.5138,
+      "step": 656
+    },
+    {
+      "epoch": 1.4575707154742097,
+      "grad_norm": 0.23079179227352142,
+      "learning_rate": 0.0001711111111111111,
+      "loss": 0.4738,
+      "step": 657
+    },
+    {
+      "epoch": 1.4597892401552968,
+      "grad_norm": 0.17954687774181366,
+      "learning_rate": 0.00017086419753086418,
+      "loss": 0.5514,
+      "step": 658
+    },
+    {
+      "epoch": 1.4620077648363838,
+      "grad_norm": 0.20924286544322968,
+      "learning_rate": 0.00017061728395061727,
+      "loss": 0.7017,
+      "step": 659
+    },
+    {
+      "epoch": 1.464226289517471,
+      "grad_norm": 0.21974889934062958,
+      "learning_rate": 0.00017037037037037034,
+      "loss": 0.4529,
+      "step": 660
+    },
+    {
+      "epoch": 1.466444814198558,
+      "grad_norm": 0.1946646124124527,
+      "learning_rate": 0.00017012345679012344,
+      "loss": 0.7267,
+      "step": 661
+    },
+    {
+      "epoch": 1.468663338879645,
+      "grad_norm": 0.19824987649917603,
+      "learning_rate": 0.0001698765432098765,
+      "loss": 0.6157,
+      "step": 662
+    },
+    {
+      "epoch": 1.4708818635607321,
+      "grad_norm": 0.17034663259983063,
+      "learning_rate": 0.00016962962962962963,
+      "loss": 0.531,
+      "step": 663
+    },
+    {
+      "epoch": 1.4731003882418192,
+      "grad_norm": 0.21619726717472076,
+      "learning_rate": 0.00016938271604938272,
+      "loss": 0.5661,
+      "step": 664
+    },
+    {
+      "epoch": 1.4753189129229063,
+      "grad_norm": 0.19192738831043243,
+      "learning_rate": 0.0001691358024691358,
+      "loss": 0.681,
+      "step": 665
+    },
+    {
+      "epoch": 1.4775374376039934,
+      "grad_norm": 0.18317005038261414,
+      "learning_rate": 0.00016888888888888889,
+      "loss": 0.6724,
+      "step": 666
+    },
+    {
+      "epoch": 1.4797559622850804,
+      "grad_norm": 0.18956716358661652,
+      "learning_rate": 0.00016864197530864195,
+      "loss": 0.5378,
+      "step": 667
+    },
+    {
+      "epoch": 1.4819744869661675,
+      "grad_norm": 0.21412551403045654,
+      "learning_rate": 0.00016839506172839505,
+      "loss": 0.3959,
+      "step": 668
+    },
+    {
+      "epoch": 1.4841930116472546,
+      "grad_norm": 0.20469453930854797,
+      "learning_rate": 0.00016814814814814812,
+      "loss": 0.6003,
+      "step": 669
+    },
+    {
+      "epoch": 1.4864115363283417,
+      "grad_norm": 0.2250760942697525,
+      "learning_rate": 0.0001679012345679012,
+      "loss": 0.5332,
+      "step": 670
+    },
+    {
+      "epoch": 1.4886300610094287,
+      "grad_norm": 0.1805281788110733,
+      "learning_rate": 0.00016765432098765433,
+      "loss": 0.5309,
+      "step": 671
+    },
+    {
+      "epoch": 1.4908485856905158,
+      "grad_norm": 0.1872943639755249,
+      "learning_rate": 0.0001674074074074074,
+      "loss": 0.5402,
+      "step": 672
+    },
+    {
+      "epoch": 1.493067110371603,
+      "grad_norm": 0.18290308117866516,
+      "learning_rate": 0.0001671604938271605,
+      "loss": 0.5007,
+      "step": 673
+    },
+    {
+      "epoch": 1.49528563505269,
+      "grad_norm": 0.2077639251947403,
+      "learning_rate": 0.00016691358024691357,
+      "loss": 0.4931,
+      "step": 674
+    },
+    {
+      "epoch": 1.497504159733777,
+      "grad_norm": 0.16692917048931122,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 0.4469,
+      "step": 675
+    },
+    {
+      "epoch": 1.4997226844148641,
+      "grad_norm": 0.22428935766220093,
+      "learning_rate": 0.00016641975308641973,
+      "loss": 0.4909,
+      "step": 676
+    },
+    {
+      "epoch": 1.5019412090959512,
+      "grad_norm": 0.19021223485469818,
+      "learning_rate": 0.00016617283950617282,
+      "loss": 0.6502,
+      "step": 677
+    },
+    {
+      "epoch": 1.5041597337770383,
+      "grad_norm": 0.19795015454292297,
+      "learning_rate": 0.0001659259259259259,
+      "loss": 0.3924,
+      "step": 678
+    },
+    {
+      "epoch": 1.5063782584581253,
+      "grad_norm": 0.1813489943742752,
+      "learning_rate": 0.000165679012345679,
+      "loss": 0.5022,
+      "step": 679
+    },
+    {
+      "epoch": 1.5085967831392124,
+      "grad_norm": 0.19167926907539368,
+      "learning_rate": 0.0001654320987654321,
+      "loss": 0.4748,
+      "step": 680
+    },
+    {
+      "epoch": 1.5108153078202995,
+      "grad_norm": 0.26904281973838806,
+      "learning_rate": 0.00016518518518518518,
+      "loss": 0.5973,
+      "step": 681
+    },
+    {
+      "epoch": 1.5130338325013866,
+      "grad_norm": 0.4328569769859314,
+      "learning_rate": 0.00016493827160493827,
+      "loss": 0.7961,
+      "step": 682
+    },
+    {
+      "epoch": 1.5152523571824736,
+      "grad_norm": 0.198623925447464,
+      "learning_rate": 0.00016469135802469134,
+      "loss": 0.53,
+      "step": 683
+    },
+    {
+      "epoch": 1.5174708818635607,
+      "grad_norm": 0.27176323533058167,
+      "learning_rate": 0.00016444444444444444,
+      "loss": 0.5479,
+      "step": 684
+    },
+    {
+      "epoch": 1.5196894065446478,
+      "grad_norm": 0.22740134596824646,
+      "learning_rate": 0.0001641975308641975,
+      "loss": 0.605,
+      "step": 685
+    },
+    {
+      "epoch": 1.5219079312257349,
+      "grad_norm": 0.2270364612340927,
+      "learning_rate": 0.0001639506172839506,
+      "loss": 0.4176,
+      "step": 686
+    },
+    {
+      "epoch": 1.524126455906822,
+      "grad_norm": 0.2676932215690613,
+      "learning_rate": 0.0001637037037037037,
+      "loss": 0.5932,
+      "step": 687
+    },
+    {
+      "epoch": 1.526344980587909,
+      "grad_norm": 0.14646005630493164,
+      "learning_rate": 0.00016345679012345676,
+      "loss": 0.6876,
+      "step": 688
+    },
+    {
+      "epoch": 1.528563505268996,
+      "grad_norm": 0.18707947432994843,
+      "learning_rate": 0.00016320987654320988,
+      "loss": 0.4818,
+      "step": 689
+    },
+    {
+      "epoch": 1.5307820299500832,
+      "grad_norm": 0.17231060564517975,
+      "learning_rate": 0.00016296296296296295,
+      "loss": 0.4963,
+      "step": 690
+    },
+    {
+      "epoch": 1.5330005546311702,
+      "grad_norm": 0.17558452486991882,
+      "learning_rate": 0.00016271604938271605,
+      "loss": 0.5289,
+      "step": 691
+    },
+    {
+      "epoch": 1.5352190793122573,
+      "grad_norm": 0.21872323751449585,
+      "learning_rate": 0.00016246913580246912,
+      "loss": 0.5779,
+      "step": 692
+    },
+    {
+      "epoch": 1.5374376039933444,
+      "grad_norm": 0.19022437930107117,
+      "learning_rate": 0.0001622222222222222,
+      "loss": 0.5951,
+      "step": 693
+    },
+    {
+      "epoch": 1.5396561286744315,
+      "grad_norm": 0.21465063095092773,
+      "learning_rate": 0.0001619753086419753,
+      "loss": 0.6435,
+      "step": 694
+    },
+    {
+      "epoch": 1.5418746533555185,
+      "grad_norm": 0.1980956494808197,
+      "learning_rate": 0.00016172839506172837,
+      "loss": 0.547,
+      "step": 695
+    },
+    {
+      "epoch": 1.5440931780366056,
+      "grad_norm": 0.207494854927063,
+      "learning_rate": 0.00016148148148148147,
+      "loss": 0.5973,
+      "step": 696
+    },
+    {
+      "epoch": 1.5463117027176927,
+      "grad_norm": 0.22936350107192993,
+      "learning_rate": 0.00016123456790123454,
+      "loss": 0.5249,
+      "step": 697
+    },
+    {
+      "epoch": 1.5485302273987798,
+      "grad_norm": 0.1833280324935913,
+      "learning_rate": 0.00016098765432098766,
+      "loss": 0.4906,
+      "step": 698
+    },
+    {
+      "epoch": 1.5507487520798668,
+      "grad_norm": 0.20271070301532745,
+      "learning_rate": 0.00016074074074074073,
+      "loss": 0.6244,
+      "step": 699
+    },
+    {
+      "epoch": 1.552967276760954,
+      "grad_norm": 0.20530453324317932,
+      "learning_rate": 0.00016049382716049382,
+      "loss": 0.473,
+      "step": 700
+    },
+    {
+      "epoch": 1.555185801442041,
+      "grad_norm": 0.22996342182159424,
+      "learning_rate": 0.00016024691358024692,
+      "loss": 0.5962,
+      "step": 701
+    },
+    {
+      "epoch": 1.557404326123128,
+      "grad_norm": 0.23241150379180908,
+      "learning_rate": 0.00015999999999999999,
+      "loss": 0.6106,
+      "step": 702
+    },
+    {
+      "epoch": 1.5596228508042151,
+      "grad_norm": 0.24837328493595123,
+      "learning_rate": 0.00015975308641975308,
+      "loss": 0.7441,
+      "step": 703
+    },
+    {
+      "epoch": 1.5618413754853022,
+      "grad_norm": 0.17307542264461517,
+      "learning_rate": 0.00015950617283950615,
+      "loss": 0.5957,
+      "step": 704
+    },
+    {
+      "epoch": 1.5640599001663893,
+      "grad_norm": 0.27853789925575256,
+      "learning_rate": 0.00015925925925925924,
+      "loss": 0.545,
+      "step": 705
+    },
+    {
+      "epoch": 1.5662784248474764,
+      "grad_norm": 0.21932540833950043,
+      "learning_rate": 0.0001590123456790123,
+      "loss": 0.526,
+      "step": 706
+    },
+    {
+      "epoch": 1.5684969495285634,
+      "grad_norm": 0.19056032598018646,
+      "learning_rate": 0.00015876543209876543,
+      "loss": 0.4891,
+      "step": 707
+    },
+    {
+      "epoch": 1.5707154742096505,
+      "grad_norm": 0.18396295607089996,
+      "learning_rate": 0.0001585185185185185,
+      "loss": 0.4133,
+      "step": 708
+    },
+    {
+      "epoch": 1.5729339988907376,
+      "grad_norm": 0.19319258630275726,
+      "learning_rate": 0.0001582716049382716,
+      "loss": 0.5622,
+      "step": 709
+    },
+    {
+      "epoch": 1.5751525235718247,
+      "grad_norm": 0.163610577583313,
+      "learning_rate": 0.0001580246913580247,
+      "loss": 0.4503,
+      "step": 710
+    },
+    {
+      "epoch": 1.5773710482529117,
+      "grad_norm": 0.34292569756507874,
+      "learning_rate": 0.00015777777777777776,
+      "loss": 0.7428,
+      "step": 711
+    },
+    {
+      "epoch": 1.5795895729339988,
+      "grad_norm": 0.18306036293506622,
+      "learning_rate": 0.00015753086419753086,
+      "loss": 0.4663,
+      "step": 712
+    },
+    {
+      "epoch": 1.581808097615086,
+      "grad_norm": 0.17871612310409546,
+      "learning_rate": 0.00015728395061728392,
+      "loss": 0.3816,
+      "step": 713
+    },
+    {
+      "epoch": 1.584026622296173,
+      "grad_norm": 0.23765674233436584,
+      "learning_rate": 0.00015703703703703702,
+      "loss": 0.6424,
+      "step": 714
+    },
+    {
+      "epoch": 1.58624514697726,
+      "grad_norm": 0.18209566175937653,
+      "learning_rate": 0.0001567901234567901,
+      "loss": 0.448,
+      "step": 715
+    },
+    {
+      "epoch": 1.5884636716583471,
+      "grad_norm": 0.19453491270542145,
+      "learning_rate": 0.0001565432098765432,
+      "loss": 0.6436,
+      "step": 716
+    },
+    {
+      "epoch": 1.5906821963394342,
+      "grad_norm": 0.20257550477981567,
+      "learning_rate": 0.0001562962962962963,
+      "loss": 0.4455,
+      "step": 717
+    },
+    {
+      "epoch": 1.5929007210205213,
+      "grad_norm": 0.254025936126709,
+      "learning_rate": 0.00015604938271604937,
+      "loss": 0.5357,
+      "step": 718
+    },
+    {
+      "epoch": 1.5951192457016083,
+      "grad_norm": 0.21836106479167938,
+      "learning_rate": 0.00015580246913580247,
+      "loss": 0.5543,
+      "step": 719
+    },
+    {
+      "epoch": 1.5973377703826954,
+      "grad_norm": 0.22487813234329224,
+      "learning_rate": 0.00015555555555555554,
+      "loss": 0.6309,
+      "step": 720
+    },
+    {
+      "epoch": 1.5995562950637825,
+      "grad_norm": 0.18639123439788818,
+      "learning_rate": 0.00015530864197530863,
+      "loss": 0.5976,
+      "step": 721
+    },
+    {
+      "epoch": 1.6017748197448696,
+      "grad_norm": 0.2182561606168747,
+      "learning_rate": 0.0001550617283950617,
+      "loss": 0.5475,
+      "step": 722
+    },
+    {
+      "epoch": 1.6039933444259566,
+      "grad_norm": 0.266687273979187,
+      "learning_rate": 0.0001548148148148148,
+      "loss": 0.4137,
+      "step": 723
+    },
+    {
+      "epoch": 1.6062118691070437,
+      "grad_norm": 0.16083861887454987,
+      "learning_rate": 0.00015456790123456792,
+      "loss": 0.7401,
+      "step": 724
+    },
+    {
+      "epoch": 1.6084303937881308,
+      "grad_norm": 0.20802190899848938,
+      "learning_rate": 0.00015432098765432098,
+      "loss": 0.4922,
+      "step": 725
+    },
+    {
+      "epoch": 1.6106489184692179,
+      "grad_norm": 0.18486987054347992,
+      "learning_rate": 0.00015407407407407408,
+      "loss": 0.5744,
+      "step": 726
+    },
+    {
+      "epoch": 1.612867443150305,
+      "grad_norm": 0.24239201843738556,
+      "learning_rate": 0.00015382716049382715,
+      "loss": 0.5581,
+      "step": 727
+    },
+    {
+      "epoch": 1.615085967831392,
+      "grad_norm": 0.2472328394651413,
+      "learning_rate": 0.00015358024691358024,
+      "loss": 0.6439,
+      "step": 728
+    },
+    {
+      "epoch": 1.617304492512479,
+      "grad_norm": 0.22095270454883575,
+      "learning_rate": 0.0001533333333333333,
+      "loss": 0.5875,
+      "step": 729
+    },
+    {
+      "epoch": 1.6195230171935662,
+      "grad_norm": 0.20829857885837555,
+      "learning_rate": 0.0001530864197530864,
+      "loss": 0.571,
+      "step": 730
+    },
+    {
+      "epoch": 1.6217415418746532,
+      "grad_norm": 0.18936103582382202,
+      "learning_rate": 0.00015283950617283947,
+      "loss": 0.5189,
+      "step": 731
+    },
+    {
+      "epoch": 1.6239600665557403,
+      "grad_norm": 0.23725983500480652,
+      "learning_rate": 0.00015259259259259257,
+      "loss": 0.636,
+      "step": 732
+    },
+    {
+      "epoch": 1.6261785912368274,
+      "grad_norm": 0.15596434473991394,
+      "learning_rate": 0.0001523456790123457,
+      "loss": 0.4134,
+      "step": 733
+    },
+    {
+      "epoch": 1.6283971159179145,
+      "grad_norm": 0.24175524711608887,
+      "learning_rate": 0.00015209876543209873,
+      "loss": 0.5071,
+      "step": 734
+    },
+    {
+      "epoch": 1.6306156405990015,
+      "grad_norm": 0.17027562856674194,
+      "learning_rate": 0.00015185185185185185,
+      "loss": 0.5318,
+      "step": 735
+    },
+    {
+      "epoch": 1.6328341652800886,
+      "grad_norm": 0.1750715672969818,
+      "learning_rate": 0.00015160493827160492,
+      "loss": 0.5206,
+      "step": 736
+    },
+    {
+      "epoch": 1.6350526899611757,
+      "grad_norm": 0.1808721423149109,
+      "learning_rate": 0.00015135802469135802,
+      "loss": 0.5109,
+      "step": 737
+    },
+    {
+      "epoch": 1.6372712146422628,
+      "grad_norm": 0.18512779474258423,
+      "learning_rate": 0.00015111111111111109,
+      "loss": 0.498,
+      "step": 738
+    },
+    {
+      "epoch": 1.6394897393233498,
+      "grad_norm": 0.17602881789207458,
+      "learning_rate": 0.00015086419753086418,
+      "loss": 0.5301,
+      "step": 739
+    },
+    {
+      "epoch": 1.641708264004437,
+      "grad_norm": 0.20164436101913452,
+      "learning_rate": 0.00015061728395061728,
+      "loss": 0.5673,
+      "step": 740
+    },
+    {
+      "epoch": 1.643926788685524,
+      "grad_norm": 0.2727857232093811,
+      "learning_rate": 0.00015037037037037034,
+      "loss": 0.6057,
+      "step": 741
+    },
+    {
+      "epoch": 1.646145313366611,
+      "grad_norm": 0.17279329895973206,
+      "learning_rate": 0.00015012345679012347,
+      "loss": 0.4077,
+      "step": 742
+    },
+    {
+      "epoch": 1.6483638380476981,
+      "grad_norm": 0.2078826129436493,
+      "learning_rate": 0.00014987654320987653,
+      "loss": 0.6244,
+      "step": 743
+    },
+    {
+      "epoch": 1.6505823627287852,
+      "grad_norm": 0.22430792450904846,
+      "learning_rate": 0.00014962962962962963,
+      "loss": 0.5551,
+      "step": 744
+    },
+    {
+      "epoch": 1.6528008874098723,
+      "grad_norm": 0.2048642486333847,
+      "learning_rate": 0.0001493827160493827,
+      "loss": 0.5601,
+      "step": 745
+    },
+    {
+      "epoch": 1.6550194120909594,
+      "grad_norm": 0.20506982505321503,
+      "learning_rate": 0.0001491358024691358,
+      "loss": 0.5862,
+      "step": 746
+    },
+    {
+      "epoch": 1.6572379367720464,
+      "grad_norm": 0.24828793108463287,
+      "learning_rate": 0.00014888888888888886,
+      "loss": 0.6274,
+      "step": 747
+    },
+    {
+      "epoch": 1.6594564614531335,
+      "grad_norm": 0.17184409499168396,
+      "learning_rate": 0.00014864197530864196,
+      "loss": 0.5141,
+      "step": 748
+    },
+    {
+      "epoch": 1.6616749861342206,
+      "grad_norm": 0.1907324492931366,
+      "learning_rate": 0.00014839506172839505,
+      "loss": 0.4809,
+      "step": 749
+    },
+    {
+      "epoch": 1.6638935108153077,
+      "grad_norm": 0.1639389544725418,
+      "learning_rate": 0.00014814814814814812,
+      "loss": 0.3653,
+      "step": 750
+    },
+    {
+      "epoch": 1.6661120354963947,
+      "grad_norm": 0.17569246888160706,
+      "learning_rate": 0.00014790123456790124,
+      "loss": 0.4398,
+      "step": 751
+    },
+    {
+      "epoch": 1.6683305601774818,
+      "grad_norm": 0.19524939358234406,
+      "learning_rate": 0.0001476543209876543,
+      "loss": 0.4481,
+      "step": 752
+    },
+    {
+      "epoch": 1.670549084858569,
+      "grad_norm": 0.22567732632160187,
+      "learning_rate": 0.0001474074074074074,
+      "loss": 0.5012,
+      "step": 753
+    },
+    {
+      "epoch": 1.672767609539656,
+      "grad_norm": 0.2320326566696167,
+      "learning_rate": 0.00014716049382716047,
+      "loss": 0.5469,
+      "step": 754
+    },
+    {
+      "epoch": 1.674986134220743,
+      "grad_norm": 0.17192989587783813,
+      "learning_rate": 0.00014691358024691357,
+      "loss": 0.5948,
+      "step": 755
+    },
+    {
+      "epoch": 1.6772046589018301,
+      "grad_norm": 0.16983619332313538,
+      "learning_rate": 0.00014666666666666664,
+      "loss": 0.6645,
+      "step": 756
+    },
+    {
+      "epoch": 1.6794231835829172,
+      "grad_norm": 0.20436342060565948,
+      "learning_rate": 0.00014641975308641973,
+      "loss": 0.5705,
+      "step": 757
+    },
+    {
+      "epoch": 1.6816417082640043,
+      "grad_norm": 0.17205803096294403,
+      "learning_rate": 0.00014617283950617283,
+      "loss": 0.5444,
+      "step": 758
+    },
+    {
+      "epoch": 1.6838602329450914,
+      "grad_norm": 0.23433679342269897,
+      "learning_rate": 0.00014592592592592592,
+      "loss": 0.5473,
+      "step": 759
+    },
+    {
+      "epoch": 1.6860787576261784,
+      "grad_norm": 0.23521554470062256,
+      "learning_rate": 0.000145679012345679,
+      "loss": 0.5704,
+      "step": 760
+    },
+    {
+      "epoch": 1.6882972823072657,
+      "grad_norm": 0.18088901042938232,
+      "learning_rate": 0.00014543209876543208,
+      "loss": 0.5538,
+      "step": 761
+    },
+    {
+      "epoch": 1.6905158069883528,
+      "grad_norm": 0.1953285038471222,
+      "learning_rate": 0.00014518518518518518,
+      "loss": 0.5212,
+      "step": 762
+    },
+    {
+      "epoch": 1.6927343316694399,
+      "grad_norm": 0.22369714081287384,
+      "learning_rate": 0.00014493827160493825,
+      "loss": 0.8965,
+      "step": 763
+    },
+    {
+      "epoch": 1.694952856350527,
+      "grad_norm": 0.25895318388938904,
+      "learning_rate": 0.00014469135802469134,
+      "loss": 0.5057,
+      "step": 764
+    },
+    {
+      "epoch": 1.697171381031614,
+      "grad_norm": 0.27749761939048767,
+      "learning_rate": 0.0001444444444444444,
+      "loss": 0.6702,
+      "step": 765
+    },
+    {
+      "epoch": 1.699389905712701,
+      "grad_norm": 0.19398140907287598,
+      "learning_rate": 0.00014419753086419753,
+      "loss": 0.5178,
+      "step": 766
+    },
+    {
+      "epoch": 1.7016084303937882,
+      "grad_norm": 0.2554016411304474,
+      "learning_rate": 0.0001439506172839506,
+      "loss": 0.7326,
+      "step": 767
+    },
+    {
+      "epoch": 1.7038269550748752,
+      "grad_norm": 0.14787107706069946,
+      "learning_rate": 0.0001437037037037037,
+      "loss": 0.4291,
+      "step": 768
+    },
+    {
+      "epoch": 1.7060454797559623,
+      "grad_norm": 0.371913880109787,
+      "learning_rate": 0.00014345679012345676,
+      "loss": 0.5061,
+      "step": 769
+    },
+    {
+      "epoch": 1.7082640044370494,
+      "grad_norm": 0.24057237803936005,
+      "learning_rate": 0.00014320987654320986,
+      "loss": 0.624,
+      "step": 770
+    },
+    {
+      "epoch": 1.7104825291181365,
+      "grad_norm": 0.1722036749124527,
+      "learning_rate": 0.00014296296296296295,
+      "loss": 0.5118,
+      "step": 771
+    },
+    {
+      "epoch": 1.7127010537992235,
+      "grad_norm": 0.2469390332698822,
+      "learning_rate": 0.00014271604938271602,
+      "loss": 0.4961,
+      "step": 772
+    },
+    {
+      "epoch": 1.7149195784803106,
+      "grad_norm": 0.24460850656032562,
+      "learning_rate": 0.00014246913580246912,
+      "loss": 0.6101,
+      "step": 773
+    },
+    {
+      "epoch": 1.7171381031613977,
+      "grad_norm": 0.21602055430412292,
+      "learning_rate": 0.0001422222222222222,
+      "loss": 0.613,
+      "step": 774
+    },
+    {
+      "epoch": 1.7193566278424848,
+      "grad_norm": 0.24315950274467468,
+      "learning_rate": 0.0001419753086419753,
+      "loss": 0.4513,
+      "step": 775
+    },
+    {
+      "epoch": 1.7215751525235718,
+      "grad_norm": 0.20765557885169983,
+      "learning_rate": 0.00014172839506172838,
+      "loss": 0.584,
+      "step": 776
+    },
+    {
+      "epoch": 1.723793677204659,
+      "grad_norm": 0.2053861916065216,
+      "learning_rate": 0.00014148148148148147,
+      "loss": 0.6858,
+      "step": 777
+    },
+    {
+      "epoch": 1.726012201885746,
+      "grad_norm": 0.19838115572929382,
+      "learning_rate": 0.00014123456790123454,
+      "loss": 0.6236,
+      "step": 778
+    },
+    {
+      "epoch": 1.728230726566833,
+      "grad_norm": 0.2241971492767334,
+      "learning_rate": 0.00014098765432098763,
+      "loss": 0.5261,
+      "step": 779
+    },
+    {
+      "epoch": 1.7304492512479202,
+      "grad_norm": 0.20968253910541534,
+      "learning_rate": 0.00014074074074074073,
+      "loss": 0.5665,
+      "step": 780
+    },
+    {
+      "epoch": 1.7326677759290072,
+      "grad_norm": 0.193669393658638,
+      "learning_rate": 0.00014049382716049382,
+      "loss": 0.5106,
+      "step": 781
+    },
+    {
+      "epoch": 1.7348863006100943,
+      "grad_norm": 0.24194596707820892,
+      "learning_rate": 0.0001402469135802469,
+      "loss": 0.5436,
+      "step": 782
+    },
+    {
+      "epoch": 1.7371048252911814,
+      "grad_norm": 0.19737020134925842,
+      "learning_rate": 0.00014,
+      "loss": 0.553,
+      "step": 783
+    },
+    {
+      "epoch": 1.7393233499722685,
+      "grad_norm": 0.25685209035873413,
+      "learning_rate": 0.00013975308641975308,
+      "loss": 0.5263,
+      "step": 784
+    },
+    {
+      "epoch": 1.7415418746533555,
+      "grad_norm": 0.21238987147808075,
+      "learning_rate": 0.00013950617283950615,
+      "loss": 0.5772,
+      "step": 785
+    },
+    {
+      "epoch": 1.7437603993344426,
+      "grad_norm": 0.20661203563213348,
+      "learning_rate": 0.00013925925925925925,
+      "loss": 0.6013,
+      "step": 786
+    },
+    {
+      "epoch": 1.7459789240155297,
+      "grad_norm": 0.20427457988262177,
+      "learning_rate": 0.00013901234567901231,
+      "loss": 0.4453,
+      "step": 787
+    },
+    {
+      "epoch": 1.7481974486966168,
+      "grad_norm": 0.25096967816352844,
+      "learning_rate": 0.00013876543209876544,
+      "loss": 0.6952,
+      "step": 788
+    },
+    {
+      "epoch": 1.7504159733777038,
+      "grad_norm": 0.22071118652820587,
+      "learning_rate": 0.0001385185185185185,
+      "loss": 0.4854,
+      "step": 789
+    },
+    {
+      "epoch": 1.752634498058791,
+      "grad_norm": 0.18470212817192078,
+      "learning_rate": 0.0001382716049382716,
+      "loss": 0.4576,
+      "step": 790
+    },
+    {
+      "epoch": 1.754853022739878,
+      "grad_norm": 0.20831331610679626,
+      "learning_rate": 0.00013802469135802467,
+      "loss": 0.5626,
+      "step": 791
+    },
+    {
+      "epoch": 1.757071547420965,
+      "grad_norm": 0.2043461799621582,
+      "learning_rate": 0.00013777777777777776,
+      "loss": 0.4078,
+      "step": 792
+    },
+    {
+      "epoch": 1.7592900721020521,
+      "grad_norm": 0.3409642279148102,
+      "learning_rate": 0.00013753086419753086,
+      "loss": 0.434,
+      "step": 793
+    },
+    {
+      "epoch": 1.7615085967831392,
+      "grad_norm": 0.24864424765110016,
+      "learning_rate": 0.00013728395061728393,
+      "loss": 0.5556,
+      "step": 794
+    },
+    {
+      "epoch": 1.7637271214642263,
+      "grad_norm": 0.2736254930496216,
+      "learning_rate": 0.00013703703703703702,
+      "loss": 0.5865,
+      "step": 795
+    },
+    {
+      "epoch": 1.7659456461453134,
+      "grad_norm": 0.22004298865795135,
+      "learning_rate": 0.00013679012345679012,
+      "loss": 0.5692,
+      "step": 796
+    },
+    {
+      "epoch": 1.7681641708264004,
+      "grad_norm": 0.19719061255455017,
+      "learning_rate": 0.0001365432098765432,
+      "loss": 0.5514,
+      "step": 797
+    },
+    {
+      "epoch": 1.7703826955074875,
+      "grad_norm": 0.19986668229103088,
+      "learning_rate": 0.00013629629629629628,
+      "loss": 0.5389,
+      "step": 798
+    },
+    {
+      "epoch": 1.7726012201885746,
+      "grad_norm": 0.21329765021800995,
+      "learning_rate": 0.00013604938271604937,
+      "loss": 0.5889,
+      "step": 799
+    },
+    {
+      "epoch": 1.7748197448696617,
+      "grad_norm": 0.213465616106987,
+      "learning_rate": 0.00013580246913580244,
+      "loss": 0.4789,
+      "step": 800
+    },
+    {
+      "epoch": 1.7770382695507487,
+      "grad_norm": 0.1650550812482834,
+      "learning_rate": 0.00013555555555555554,
+      "loss": 0.4791,
+      "step": 801
+    },
+    {
+      "epoch": 1.7792567942318358,
+      "grad_norm": 0.20463810861110687,
+      "learning_rate": 0.00013530864197530863,
+      "loss": 0.4744,
+      "step": 802
+    },
+    {
+      "epoch": 1.7814753189129229,
+      "grad_norm": 0.19727501273155212,
+      "learning_rate": 0.00013506172839506173,
+      "loss": 0.4683,
+      "step": 803
+    },
+    {
+      "epoch": 1.78369384359401,
+      "grad_norm": 0.19071635603904724,
+      "learning_rate": 0.0001348148148148148,
+      "loss": 0.5873,
+      "step": 804
+    },
+    {
+      "epoch": 1.785912368275097,
+      "grad_norm": 0.18468137085437775,
+      "learning_rate": 0.0001345679012345679,
+      "loss": 0.4581,
+      "step": 805
+    },
+    {
+      "epoch": 1.788130892956184,
+      "grad_norm": 0.1673188954591751,
+      "learning_rate": 0.00013432098765432099,
+      "loss": 0.5123,
+      "step": 806
+    },
+    {
+      "epoch": 1.7903494176372712,
+      "grad_norm": 0.26824751496315,
+      "learning_rate": 0.00013407407407407405,
+      "loss": 0.5314,
+      "step": 807
+    },
+    {
+      "epoch": 1.7925679423183583,
+      "grad_norm": 0.22963069379329681,
+      "learning_rate": 0.00013382716049382715,
+      "loss": 0.6202,
+      "step": 808
+    },
+    {
+      "epoch": 1.7947864669994453,
+      "grad_norm": 0.1775522232055664,
+      "learning_rate": 0.00013358024691358022,
+      "loss": 0.4575,
+      "step": 809
+    },
+    {
+      "epoch": 1.7970049916805324,
+      "grad_norm": 0.1774640530347824,
+      "learning_rate": 0.0001333333333333333,
+      "loss": 0.5077,
+      "step": 810
+    },
+    {
+      "epoch": 1.7992235163616195,
+      "grad_norm": 0.16231273114681244,
+      "learning_rate": 0.0001330864197530864,
+      "loss": 0.4898,
+      "step": 811
+    },
+    {
+      "epoch": 1.8014420410427066,
+      "grad_norm": 0.20999057590961456,
+      "learning_rate": 0.0001328395061728395,
+      "loss": 0.5723,
+      "step": 812
+    },
+    {
+      "epoch": 1.8036605657237936,
+      "grad_norm": 0.16892686486244202,
+      "learning_rate": 0.00013259259259259257,
+      "loss": 0.5842,
+      "step": 813
+    },
+    {
+      "epoch": 1.8058790904048807,
+      "grad_norm": 0.23635898530483246,
+      "learning_rate": 0.00013234567901234567,
+      "loss": 0.5519,
+      "step": 814
+    },
+    {
+      "epoch": 1.8080976150859678,
+      "grad_norm": 0.24205300211906433,
+      "learning_rate": 0.00013209876543209876,
+      "loss": 0.4776,
+      "step": 815
+    },
+    {
+      "epoch": 1.8103161397670549,
+      "grad_norm": 0.22460991144180298,
+      "learning_rate": 0.00013185185185185183,
+      "loss": 0.74,
+      "step": 816
+    },
+    {
+      "epoch": 1.8125346644481422,
+      "grad_norm": 0.22930462658405304,
+      "learning_rate": 0.00013160493827160492,
+      "loss": 0.504,
+      "step": 817
+    },
+    {
+      "epoch": 1.8147531891292292,
+      "grad_norm": 0.18085402250289917,
+      "learning_rate": 0.00013135802469135802,
+      "loss": 0.5735,
+      "step": 818
+    },
+    {
+      "epoch": 1.8169717138103163,
+      "grad_norm": 0.2589346170425415,
+      "learning_rate": 0.00013111111111111111,
+      "loss": 0.5317,
+      "step": 819
+    },
+    {
+      "epoch": 1.8191902384914034,
+      "grad_norm": 0.27089110016822815,
+      "learning_rate": 0.00013086419753086418,
+      "loss": 0.5143,
+      "step": 820
+    },
+    {
+      "epoch": 1.8214087631724905,
+      "grad_norm": 0.20884625613689423,
+      "learning_rate": 0.00013061728395061728,
+      "loss": 0.5585,
+      "step": 821
+    },
+    {
+      "epoch": 1.8236272878535775,
+      "grad_norm": 0.251960813999176,
+      "learning_rate": 0.00013037037037037035,
+      "loss": 0.5539,
+      "step": 822
+    },
+    {
+      "epoch": 1.8258458125346646,
+      "grad_norm": 0.24113720655441284,
+      "learning_rate": 0.00013012345679012344,
+      "loss": 0.48,
+      "step": 823
+    },
+    {
+      "epoch": 1.8280643372157517,
+      "grad_norm": 0.25741907954216003,
+      "learning_rate": 0.00012987654320987654,
+      "loss": 0.6535,
+      "step": 824
+    },
+    {
+      "epoch": 1.8302828618968388,
+      "grad_norm": 0.253353476524353,
+      "learning_rate": 0.0001296296296296296,
+      "loss": 0.5608,
+      "step": 825
+    },
+    {
+      "epoch": 1.8325013865779258,
+      "grad_norm": 0.2818291187286377,
+      "learning_rate": 0.0001293827160493827,
+      "loss": 0.604,
+      "step": 826
+    },
+    {
+      "epoch": 1.834719911259013,
+      "grad_norm": 0.20671424269676208,
+      "learning_rate": 0.0001291358024691358,
+      "loss": 0.6396,
+      "step": 827
+    },
+    {
+      "epoch": 1.8369384359401,
+      "grad_norm": 0.25901779532432556,
+      "learning_rate": 0.0001288888888888889,
+      "loss": 0.4614,
+      "step": 828
+    },
+    {
+      "epoch": 1.839156960621187,
+      "grad_norm": 0.19807875156402588,
+      "learning_rate": 0.00012864197530864196,
+      "loss": 0.5888,
+      "step": 829
+    },
+    {
+      "epoch": 1.8413754853022741,
+      "grad_norm": 0.24440619349479675,
+      "learning_rate": 0.00012839506172839505,
+      "loss": 0.5764,
+      "step": 830
+    },
+    {
+      "epoch": 1.8435940099833612,
+      "grad_norm": 0.24335061013698578,
+      "learning_rate": 0.00012814814814814812,
+      "loss": 0.5763,
+      "step": 831
+    },
+    {
+      "epoch": 1.8458125346644483,
+      "grad_norm": 0.21975311636924744,
+      "learning_rate": 0.00012790123456790122,
+      "loss": 0.5549,
+      "step": 832
+    },
+    {
+      "epoch": 1.8480310593455354,
+      "grad_norm": 0.2547023296356201,
+      "learning_rate": 0.0001276543209876543,
+      "loss": 0.6194,
+      "step": 833
+    },
+    {
+      "epoch": 1.8502495840266224,
+      "grad_norm": 0.17079323530197144,
+      "learning_rate": 0.0001274074074074074,
+      "loss": 0.4493,
+      "step": 834
+    },
+    {
+      "epoch": 1.8524681087077095,
+      "grad_norm": 0.26923972368240356,
+      "learning_rate": 0.00012716049382716047,
+      "loss": 0.5406,
+      "step": 835
+    },
+    {
+      "epoch": 1.8546866333887966,
+      "grad_norm": 0.25859659910202026,
+      "learning_rate": 0.00012691358024691357,
+      "loss": 0.6406,
+      "step": 836
+    },
+    {
+      "epoch": 1.8569051580698837,
+      "grad_norm": 0.2729698717594147,
+      "learning_rate": 0.00012666666666666666,
+      "loss": 0.5857,
+      "step": 837
+    },
+    {
+      "epoch": 1.8591236827509707,
+      "grad_norm": 0.24708813428878784,
+      "learning_rate": 0.00012641975308641973,
+      "loss": 0.5313,
+      "step": 838
+    },
+    {
+      "epoch": 1.8613422074320578,
+      "grad_norm": 0.19806735217571259,
+      "learning_rate": 0.00012617283950617283,
+      "loss": 0.4261,
+      "step": 839
+    },
+    {
+      "epoch": 1.8635607321131449,
+      "grad_norm": 0.1822500228881836,
+      "learning_rate": 0.0001259259259259259,
+      "loss": 0.5485,
+      "step": 840
+    },
+    {
+      "epoch": 1.865779256794232,
+      "grad_norm": 0.25701263546943665,
+      "learning_rate": 0.00012567901234567902,
+      "loss": 0.6229,
+      "step": 841
+    },
+    {
+      "epoch": 1.867997781475319,
+      "grad_norm": 0.22260640561580658,
+      "learning_rate": 0.00012543209876543209,
+      "loss": 0.6318,
+      "step": 842
+    },
+    {
+      "epoch": 1.870216306156406,
+      "grad_norm": 0.22851626574993134,
+      "learning_rate": 0.00012518518518518518,
+      "loss": 0.4628,
+      "step": 843
+    },
+    {
+      "epoch": 1.8724348308374932,
+      "grad_norm": 0.24706195294857025,
+      "learning_rate": 0.00012493827160493825,
+      "loss": 0.6725,
+      "step": 844
+    },
+    {
+      "epoch": 1.8746533555185803,
+      "grad_norm": 0.16072241961956024,
+      "learning_rate": 0.00012469135802469134,
+      "loss": 0.4882,
+      "step": 845
+    },
+    {
+      "epoch": 1.8768718801996673,
+      "grad_norm": 0.21493948996067047,
+      "learning_rate": 0.00012444444444444444,
+      "loss": 0.6294,
+      "step": 846
+    },
+    {
+      "epoch": 1.8790904048807544,
+      "grad_norm": 0.20755286514759064,
+      "learning_rate": 0.0001241975308641975,
+      "loss": 0.5049,
+      "step": 847
+    },
+    {
+      "epoch": 1.8813089295618415,
+      "grad_norm": 0.2281264066696167,
+      "learning_rate": 0.0001239506172839506,
+      "loss": 0.5642,
+      "step": 848
+    },
+    {
+      "epoch": 1.8835274542429286,
+      "grad_norm": 0.23598706722259521,
+      "learning_rate": 0.0001237037037037037,
+      "loss": 0.5133,
+      "step": 849
+    },
+    {
+      "epoch": 1.8857459789240156,
+      "grad_norm": 0.21183046698570251,
+      "learning_rate": 0.0001234567901234568,
+      "loss": 0.5497,
+      "step": 850
+    },
+    {
+      "epoch": 1.8879645036051027,
+      "grad_norm": 0.19870619475841522,
+      "learning_rate": 0.00012320987654320986,
+      "loss": 0.5622,
+      "step": 851
+    },
+    {
+      "epoch": 1.8901830282861898,
+      "grad_norm": 0.20340567827224731,
+      "learning_rate": 0.00012296296296296296,
+      "loss": 0.5172,
+      "step": 852
+    },
+    {
+      "epoch": 1.8924015529672769,
+      "grad_norm": 0.21144112944602966,
+      "learning_rate": 0.00012271604938271602,
+      "loss": 0.4332,
+      "step": 853
+    },
+    {
+      "epoch": 1.894620077648364,
+      "grad_norm": 0.21746166050434113,
+      "learning_rate": 0.00012246913580246912,
+      "loss": 0.5181,
+      "step": 854
+    },
+    {
+      "epoch": 1.896838602329451,
+      "grad_norm": 0.16804184019565582,
+      "learning_rate": 0.00012222222222222221,
+      "loss": 0.723,
+      "step": 855
+    },
+    {
+      "epoch": 1.899057127010538,
+      "grad_norm": 0.15586820244789124,
+      "learning_rate": 0.00012197530864197531,
+      "loss": 0.461,
+      "step": 856
+    },
+    {
+      "epoch": 1.9012756516916252,
+      "grad_norm": 0.24398058652877808,
+      "learning_rate": 0.00012172839506172839,
+      "loss": 0.3959,
+      "step": 857
+    },
+    {
+      "epoch": 1.9034941763727122,
+      "grad_norm": 0.2544343173503876,
+      "learning_rate": 0.00012148148148148147,
+      "loss": 0.6398,
+      "step": 858
+    },
+    {
+      "epoch": 1.9057127010537993,
+      "grad_norm": 0.24063482880592346,
+      "learning_rate": 0.00012123456790123455,
+      "loss": 0.7235,
+      "step": 859
+    },
+    {
+      "epoch": 1.9079312257348864,
+      "grad_norm": 0.23648963868618011,
+      "learning_rate": 0.00012098765432098764,
+      "loss": 0.5548,
+      "step": 860
+    },
+    {
+      "epoch": 1.9101497504159735,
+      "grad_norm": 0.1573287397623062,
+      "learning_rate": 0.00012074074074074073,
+      "loss": 0.7682,
+      "step": 861
+    },
+    {
+      "epoch": 1.9123682750970605,
+      "grad_norm": 0.22214579582214355,
+      "learning_rate": 0.00012049382716049381,
+      "loss": 0.4495,
+      "step": 862
+    },
+    {
+      "epoch": 1.9145867997781476,
+      "grad_norm": 0.18273594975471497,
+      "learning_rate": 0.0001202469135802469,
+      "loss": 0.4289,
+      "step": 863
+    },
+    {
+      "epoch": 1.9168053244592347,
+      "grad_norm": 0.2493828684091568,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 0.4771,
+      "step": 864
+    },
+    {
+      "epoch": 1.9190238491403218,
+      "grad_norm": 0.16754016280174255,
+      "learning_rate": 0.00011975308641975308,
+      "loss": 0.5403,
+      "step": 865
+    },
+    {
+      "epoch": 1.9212423738214088,
+      "grad_norm": 0.30242446064949036,
+      "learning_rate": 0.00011950617283950617,
+      "loss": 0.5251,
+      "step": 866
+    },
+    {
+      "epoch": 1.923460898502496,
+      "grad_norm": 0.2416113317012787,
+      "learning_rate": 0.00011925925925925925,
+      "loss": 0.4413,
+      "step": 867
+    },
+    {
+      "epoch": 1.925679423183583,
+      "grad_norm": 0.24770741164684296,
+      "learning_rate": 0.00011901234567901233,
+      "loss": 0.6852,
+      "step": 868
+    },
+    {
+      "epoch": 1.92789794786467,
+      "grad_norm": 0.23942086100578308,
+      "learning_rate": 0.00011876543209876541,
+      "loss": 0.553,
+      "step": 869
+    },
+    {
+      "epoch": 1.9301164725457571,
+      "grad_norm": 0.22944417595863342,
+      "learning_rate": 0.0001185185185185185,
+      "loss": 0.7421,
+      "step": 870
+    },
+    {
+      "epoch": 1.9323349972268442,
+      "grad_norm": 0.23155508935451508,
+      "learning_rate": 0.0001182716049382716,
+      "loss": 0.5334,
+      "step": 871
+    },
+    {
+      "epoch": 1.9345535219079313,
+      "grad_norm": 0.31126219034194946,
+      "learning_rate": 0.00011802469135802468,
+      "loss": 0.5492,
+      "step": 872
+    },
+    {
+      "epoch": 1.9367720465890184,
+      "grad_norm": 0.2361699938774109,
+      "learning_rate": 0.00011777777777777776,
+      "loss": 0.4765,
+      "step": 873
+    },
+    {
+      "epoch": 1.9389905712701054,
+      "grad_norm": 0.21641334891319275,
+      "learning_rate": 0.00011753086419753086,
+      "loss": 0.5521,
+      "step": 874
+    },
+    {
+      "epoch": 1.9412090959511925,
+      "grad_norm": 0.18453632295131683,
+      "learning_rate": 0.00011728395061728394,
+      "loss": 0.6133,
+      "step": 875
+    },
+    {
+      "epoch": 1.9434276206322796,
+      "grad_norm": 0.2211303412914276,
+      "learning_rate": 0.00011703703703703702,
+      "loss": 0.4638,
+      "step": 876
+    },
+    {
+      "epoch": 1.9456461453133667,
+      "grad_norm": 0.19533082842826843,
+      "learning_rate": 0.0001167901234567901,
+      "loss": 0.4209,
+      "step": 877
+    },
+    {
+      "epoch": 1.9478646699944537,
+      "grad_norm": 0.1928904801607132,
+      "learning_rate": 0.00011654320987654319,
+      "loss": 0.4921,
+      "step": 878
+    },
+    {
+      "epoch": 1.9500831946755408,
+      "grad_norm": 0.22104381024837494,
+      "learning_rate": 0.0001162962962962963,
+      "loss": 0.5655,
+      "step": 879
+    },
+    {
+      "epoch": 1.9523017193566279,
+      "grad_norm": 0.1776837855577469,
+      "learning_rate": 0.00011604938271604938,
+      "loss": 0.4926,
+      "step": 880
+    },
+    {
+      "epoch": 1.954520244037715,
+      "grad_norm": 0.1908082813024521,
+      "learning_rate": 0.00011580246913580246,
+      "loss": 0.5771,
+      "step": 881
+    },
+    {
+      "epoch": 1.956738768718802,
+      "grad_norm": 0.17688776552677155,
+      "learning_rate": 0.00011555555555555554,
+      "loss": 0.5462,
+      "step": 882
+    },
+    {
+      "epoch": 1.958957293399889,
+      "grad_norm": 0.22206731140613556,
+      "learning_rate": 0.00011530864197530863,
+      "loss": 0.652,
+      "step": 883
+    },
+    {
+      "epoch": 1.9611758180809762,
+      "grad_norm": 0.2799658179283142,
+      "learning_rate": 0.00011506172839506172,
+      "loss": 0.5306,
+      "step": 884
+    },
+    {
+      "epoch": 1.9633943427620633,
+      "grad_norm": 0.24645838141441345,
+      "learning_rate": 0.0001148148148148148,
+      "loss": 0.5795,
+      "step": 885
+    },
+    {
+      "epoch": 1.9656128674431503,
+      "grad_norm": 0.22399859130382538,
+      "learning_rate": 0.00011456790123456789,
+      "loss": 0.5523,
+      "step": 886
+    },
+    {
+      "epoch": 1.9678313921242374,
+      "grad_norm": 0.20631204545497894,
+      "learning_rate": 0.00011432098765432099,
+      "loss": 0.5419,
+      "step": 887
+    },
+    {
+      "epoch": 1.9700499168053245,
+      "grad_norm": 0.24074740707874298,
+      "learning_rate": 0.00011407407407407407,
+      "loss": 0.5103,
+      "step": 888
+    },
+    {
+      "epoch": 1.9722684414864116,
+      "grad_norm": 0.17942370474338531,
+      "learning_rate": 0.00011382716049382715,
+      "loss": 0.4942,
+      "step": 889
+    },
+    {
+      "epoch": 1.9744869661674986,
+      "grad_norm": 0.23273387551307678,
+      "learning_rate": 0.00011358024691358023,
+      "loss": 0.4786,
+      "step": 890
+    },
+    {
+      "epoch": 1.9767054908485857,
+      "grad_norm": 0.2732599377632141,
+      "learning_rate": 0.00011333333333333331,
+      "loss": 0.6059,
+      "step": 891
+    },
+    {
+      "epoch": 1.9789240155296728,
+      "grad_norm": 0.21639283001422882,
+      "learning_rate": 0.00011308641975308641,
+      "loss": 0.4252,
+      "step": 892
+    },
+    {
+      "epoch": 1.9811425402107599,
+      "grad_norm": 0.2821376323699951,
+      "learning_rate": 0.00011283950617283949,
+      "loss": 0.4855,
+      "step": 893
+    },
+    {
+      "epoch": 1.983361064891847,
+      "grad_norm": 0.18219923973083496,
+      "learning_rate": 0.00011259259259259259,
+      "loss": 0.4889,
+      "step": 894
+    },
+    {
+      "epoch": 1.985579589572934,
+      "grad_norm": 0.2017829716205597,
+      "learning_rate": 0.00011234567901234567,
+      "loss": 0.5836,
+      "step": 895
+    },
+    {
+      "epoch": 1.987798114254021,
+      "grad_norm": 0.2333594560623169,
+      "learning_rate": 0.00011209876543209876,
+      "loss": 0.5363,
+      "step": 896
+    },
+    {
+      "epoch": 1.9900166389351082,
+      "grad_norm": 0.19913619756698608,
+      "learning_rate": 0.00011185185185185184,
+      "loss": 0.4447,
+      "step": 897
+    },
+    {
+      "epoch": 1.9922351636161952,
+      "grad_norm": 0.24440062046051025,
+      "learning_rate": 0.00011160493827160493,
+      "loss": 0.4326,
+      "step": 898
+    },
+    {
+      "epoch": 1.9944536882972823,
+      "grad_norm": 0.19665968418121338,
+      "learning_rate": 0.00011135802469135801,
+      "loss": 0.5448,
+      "step": 899
+    },
+    {
+      "epoch": 1.9966722129783694,
+      "grad_norm": 0.26268884539604187,
+      "learning_rate": 0.00011111111111111109,
+      "loss": 0.6054,
+      "step": 900
+    },
+    {
+      "epoch": 1.9988907376594565,
+      "grad_norm": 0.2307175248861313,
+      "learning_rate": 0.0001108641975308642,
+      "loss": 0.5008,
+      "step": 901
+    },
+    {
+      "epoch": 2.0011092623405435,
+      "grad_norm": 0.21265113353729248,
+      "learning_rate": 0.00011061728395061728,
+      "loss": 0.5523,
+      "step": 902
+    },
+    {
+      "epoch": 2.0033277870216306,
+      "grad_norm": 0.18614155054092407,
+      "learning_rate": 0.00011037037037037036,
+      "loss": 0.3988,
+      "step": 903
+    },
+    {
+      "epoch": 2.0055463117027177,
+      "grad_norm": 0.23707857728004456,
+      "learning_rate": 0.00011012345679012344,
+      "loss": 0.5082,
+      "step": 904
+    },
+    {
+      "epoch": 2.0077648363838048,
+      "grad_norm": 0.20816181600093842,
+      "learning_rate": 0.00010987654320987654,
+      "loss": 0.5473,
+      "step": 905
+    },
+    {
+      "epoch": 2.009983361064892,
+      "grad_norm": 0.21082249283790588,
+      "learning_rate": 0.00010962962962962962,
+      "loss": 0.4002,
+      "step": 906
+    },
+    {
+      "epoch": 2.012201885745979,
+      "grad_norm": 0.1583278626203537,
+      "learning_rate": 0.0001093827160493827,
+      "loss": 0.4936,
+      "step": 907
+    },
+    {
+      "epoch": 2.014420410427066,
+      "grad_norm": 0.17613030970096588,
+      "learning_rate": 0.00010913580246913578,
+      "loss": 0.5362,
+      "step": 908
+    },
+    {
+      "epoch": 2.016638935108153,
+      "grad_norm": 0.2371397465467453,
+      "learning_rate": 0.00010888888888888889,
+      "loss": 0.5425,
+      "step": 909
+    },
+    {
+      "epoch": 2.01885745978924,
+      "grad_norm": 0.21185088157653809,
+      "learning_rate": 0.00010864197530864197,
+      "loss": 0.5046,
+      "step": 910
+    },
+    {
+      "epoch": 2.021075984470327,
+      "grad_norm": 0.18065595626831055,
+      "learning_rate": 0.00010839506172839505,
+      "loss": 0.5274,
+      "step": 911
+    },
+    {
+      "epoch": 2.0232945091514143,
+      "grad_norm": 0.21010708808898926,
+      "learning_rate": 0.00010814814814814814,
+      "loss": 0.439,
+      "step": 912
+    },
+    {
+      "epoch": 2.0255130338325014,
+      "grad_norm": 0.16843914985656738,
+      "learning_rate": 0.00010790123456790122,
+      "loss": 0.3828,
+      "step": 913
+    },
+    {
+      "epoch": 2.0277315585135884,
+      "grad_norm": 0.19600214064121246,
+      "learning_rate": 0.00010765432098765431,
+      "loss": 0.52,
+      "step": 914
+    },
+    {
+      "epoch": 2.0299500831946755,
+      "grad_norm": 0.2053930163383484,
+      "learning_rate": 0.0001074074074074074,
+      "loss": 0.4035,
+      "step": 915
+    },
+    {
+      "epoch": 2.0321686078757626,
+      "grad_norm": 0.18540045619010925,
+      "learning_rate": 0.00010716049382716048,
+      "loss": 0.372,
+      "step": 916
+    },
+    {
+      "epoch": 2.0343871325568497,
+      "grad_norm": 0.2439187914133072,
+      "learning_rate": 0.00010691358024691357,
+      "loss": 0.4003,
+      "step": 917
+    },
+    {
+      "epoch": 2.0366056572379367,
+      "grad_norm": 0.22519932687282562,
+      "learning_rate": 0.00010666666666666667,
+      "loss": 0.3984,
+      "step": 918
+    },
+    {
+      "epoch": 2.038824181919024,
+      "grad_norm": 0.21707755327224731,
+      "learning_rate": 0.00010641975308641975,
+      "loss": 0.6235,
+      "step": 919
+    },
+    {
+      "epoch": 2.041042706600111,
+      "grad_norm": 0.17157481610774994,
+      "learning_rate": 0.00010617283950617283,
+      "loss": 0.3875,
+      "step": 920
+    },
+    {
+      "epoch": 2.043261231281198,
+      "grad_norm": 0.18599830567836761,
+      "learning_rate": 0.00010592592592592591,
+      "loss": 0.4481,
+      "step": 921
+    },
+    {
+      "epoch": 2.045479755962285,
+      "grad_norm": 0.2018834352493286,
+      "learning_rate": 0.00010567901234567899,
+      "loss": 0.4834,
+      "step": 922
+    },
+    {
+      "epoch": 2.047698280643372,
+      "grad_norm": 0.17320840060710907,
+      "learning_rate": 0.00010543209876543209,
+      "loss": 0.4198,
+      "step": 923
+    },
+    {
+      "epoch": 2.049916805324459,
+      "grad_norm": 0.24207624793052673,
+      "learning_rate": 0.00010518518518518518,
+      "loss": 0.3874,
+      "step": 924
+    },
+    {
+      "epoch": 2.0521353300055463,
+      "grad_norm": 0.20677320659160614,
+      "learning_rate": 0.00010493827160493826,
+      "loss": 0.4368,
+      "step": 925
+    },
+    {
+      "epoch": 2.0543538546866333,
+      "grad_norm": 0.21154583990573883,
+      "learning_rate": 0.00010469135802469135,
+      "loss": 0.5391,
+      "step": 926
+    },
+    {
+      "epoch": 2.0565723793677204,
+      "grad_norm": 0.2118653953075409,
+      "learning_rate": 0.00010444444444444444,
+      "loss": 0.5233,
+      "step": 927
+    },
+    {
+      "epoch": 2.0587909040488075,
+      "grad_norm": 0.21823357045650482,
+      "learning_rate": 0.00010419753086419752,
+      "loss": 0.379,
+      "step": 928
+    },
+    {
+      "epoch": 2.0610094287298946,
+      "grad_norm": 0.25995132327079773,
+      "learning_rate": 0.0001039506172839506,
+      "loss": 0.5762,
+      "step": 929
+    },
+    {
+      "epoch": 2.0632279534109816,
+      "grad_norm": 0.2049769163131714,
+      "learning_rate": 0.00010370370370370369,
+      "loss": 0.4606,
+      "step": 930
+    },
+    {
+      "epoch": 2.0654464780920687,
+      "grad_norm": 0.2339644581079483,
+      "learning_rate": 0.00010345679012345677,
+      "loss": 0.6403,
+      "step": 931
+    },
+    {
+      "epoch": 2.067665002773156,
+      "grad_norm": 0.21399492025375366,
+      "learning_rate": 0.00010320987654320988,
+      "loss": 0.3884,
+      "step": 932
+    },
+    {
+      "epoch": 2.069883527454243,
+      "grad_norm": 0.2578800916671753,
+      "learning_rate": 0.00010296296296296296,
+      "loss": 0.5472,
+      "step": 933
+    },
+    {
+      "epoch": 2.07210205213533,
+      "grad_norm": 0.2798822522163391,
+      "learning_rate": 0.00010271604938271604,
+      "loss": 0.5836,
+      "step": 934
+    },
+    {
+      "epoch": 2.074320576816417,
+      "grad_norm": 0.22812147438526154,
+      "learning_rate": 0.00010246913580246912,
+      "loss": 0.434,
+      "step": 935
+    },
+    {
+      "epoch": 2.076539101497504,
+      "grad_norm": 0.23864109814167023,
+      "learning_rate": 0.00010222222222222222,
+      "loss": 0.635,
+      "step": 936
+    },
+    {
+      "epoch": 2.078757626178591,
+      "grad_norm": 0.23195762932300568,
+      "learning_rate": 0.0001019753086419753,
+      "loss": 0.5275,
+      "step": 937
+    },
+    {
+      "epoch": 2.0809761508596782,
+      "grad_norm": 0.23830029368400574,
+      "learning_rate": 0.00010172839506172838,
+      "loss": 0.3405,
+      "step": 938
+    },
+    {
+      "epoch": 2.0831946755407653,
+      "grad_norm": 0.27229878306388855,
+      "learning_rate": 0.00010148148148148147,
+      "loss": 0.4647,
+      "step": 939
+    },
+    {
+      "epoch": 2.0854132002218524,
+      "grad_norm": 0.1947094053030014,
+      "learning_rate": 0.00010123456790123457,
+      "loss": 0.4299,
+      "step": 940
+    },
+    {
+      "epoch": 2.0876317249029395,
+      "grad_norm": 0.24421322345733643,
+      "learning_rate": 0.00010098765432098765,
+      "loss": 0.5578,
+      "step": 941
+    },
+    {
+      "epoch": 2.0898502495840265,
+      "grad_norm": 0.26572397351264954,
+      "learning_rate": 0.00010074074074074073,
+      "loss": 0.5128,
+      "step": 942
+    },
+    {
+      "epoch": 2.0920687742651136,
+      "grad_norm": 0.24750950932502747,
+      "learning_rate": 0.00010049382716049381,
+      "loss": 0.487,
+      "step": 943
+    },
+    {
+      "epoch": 2.0942872989462007,
+      "grad_norm": 0.19441011548042297,
+      "learning_rate": 0.0001002469135802469,
+      "loss": 0.443,
+      "step": 944
+    },
+    {
+      "epoch": 2.0965058236272878,
+      "grad_norm": 0.2350420504808426,
+      "learning_rate": 9.999999999999999e-05,
+      "loss": 0.6314,
+      "step": 945
+    },
+    {
+      "epoch": 2.098724348308375,
+      "grad_norm": 0.31033432483673096,
+      "learning_rate": 9.975308641975307e-05,
+      "loss": 0.4626,
+      "step": 946
+    },
+    {
+      "epoch": 2.100942872989462,
+      "grad_norm": 0.25598734617233276,
+      "learning_rate": 9.950617283950617e-05,
+      "loss": 0.4575,
+      "step": 947
+    },
+    {
+      "epoch": 2.103161397670549,
+      "grad_norm": 0.24734239280223846,
+      "learning_rate": 9.925925925925925e-05,
+      "loss": 0.456,
+      "step": 948
+    },
+    {
+      "epoch": 2.105379922351636,
+      "grad_norm": 0.19811579585075378,
+      "learning_rate": 9.901234567901234e-05,
+      "loss": 0.4467,
+      "step": 949
+    },
+    {
+      "epoch": 2.107598447032723,
+      "grad_norm": 0.2632578909397125,
+      "learning_rate": 9.876543209876543e-05,
+      "loss": 0.4757,
+      "step": 950
+    },
+    {
+      "epoch": 2.10981697171381,
+      "grad_norm": 0.2333948314189911,
+      "learning_rate": 9.851851851851851e-05,
+      "loss": 0.5043,
+      "step": 951
+    },
+    {
+      "epoch": 2.1120354963948973,
+      "grad_norm": 0.2361207753419876,
+      "learning_rate": 9.827160493827159e-05,
+      "loss": 0.4076,
+      "step": 952
+    },
+    {
+      "epoch": 2.1142540210759844,
+      "grad_norm": 0.24212293326854706,
+      "learning_rate": 9.802469135802467e-05,
+      "loss": 0.4518,
+      "step": 953
+    },
+    {
+      "epoch": 2.1164725457570714,
+      "grad_norm": 0.23921775817871094,
+      "learning_rate": 9.777777777777778e-05,
+      "loss": 0.5375,
+      "step": 954
+    },
+    {
+      "epoch": 2.1186910704381585,
+      "grad_norm": 0.2229742854833603,
+      "learning_rate": 9.753086419753086e-05,
+      "loss": 0.3746,
+      "step": 955
+    },
+    {
+      "epoch": 2.1209095951192456,
+      "grad_norm": 0.2603321969509125,
+      "learning_rate": 9.728395061728394e-05,
+      "loss": 0.5167,
+      "step": 956
+    },
+    {
+      "epoch": 2.1231281198003327,
+      "grad_norm": 0.2475040704011917,
+      "learning_rate": 9.703703703703702e-05,
+      "loss": 0.6131,
+      "step": 957
+    },
+    {
+      "epoch": 2.1253466444814197,
+      "grad_norm": 0.23410473763942719,
+      "learning_rate": 9.679012345679012e-05,
+      "loss": 0.5342,
+      "step": 958
+    },
+    {
+      "epoch": 2.127565169162507,
+      "grad_norm": 0.2917667031288147,
+      "learning_rate": 9.65432098765432e-05,
+      "loss": 0.4026,
+      "step": 959
+    },
+    {
+      "epoch": 2.129783693843594,
+      "grad_norm": 0.28873470425605774,
+      "learning_rate": 9.629629629629628e-05,
+      "loss": 0.7325,
+      "step": 960
+    },
+    {
+      "epoch": 2.132002218524681,
+      "grad_norm": 0.238609179854393,
+      "learning_rate": 9.604938271604936e-05,
+      "loss": 0.3637,
+      "step": 961
+    },
+    {
+      "epoch": 2.134220743205768,
+      "grad_norm": 0.2859095633029938,
+      "learning_rate": 9.580246913580247e-05,
+      "loss": 0.7343,
+      "step": 962
+    },
+    {
+      "epoch": 2.136439267886855,
+      "grad_norm": 0.28797978162765503,
+      "learning_rate": 9.555555555555555e-05,
+      "loss": 0.5249,
+      "step": 963
+    },
+    {
+      "epoch": 2.138657792567942,
+      "grad_norm": 0.25537165999412537,
+      "learning_rate": 9.530864197530864e-05,
+      "loss": 0.504,
+      "step": 964
+    },
+    {
+      "epoch": 2.1408763172490293,
+      "grad_norm": 0.24694858491420746,
+      "learning_rate": 9.506172839506172e-05,
+      "loss": 0.5708,
+      "step": 965
+    },
+    {
+      "epoch": 2.1430948419301163,
+      "grad_norm": 0.2794093191623688,
+      "learning_rate": 9.48148148148148e-05,
+      "loss": 0.5427,
+      "step": 966
+    },
+    {
+      "epoch": 2.1453133666112034,
+      "grad_norm": 0.260640412569046,
+      "learning_rate": 9.45679012345679e-05,
+      "loss": 0.6097,
+      "step": 967
+    },
+    {
+      "epoch": 2.1475318912922905,
+      "grad_norm": 0.21139080822467804,
+      "learning_rate": 9.432098765432098e-05,
+      "loss": 0.4341,
+      "step": 968
+    },
+    {
+      "epoch": 2.1497504159733776,
+      "grad_norm": 0.24866950511932373,
+      "learning_rate": 9.407407407407407e-05,
+      "loss": 0.5564,
+      "step": 969
+    },
+    {
+      "epoch": 2.1519689406544646,
+      "grad_norm": 0.2639906406402588,
+      "learning_rate": 9.382716049382715e-05,
+      "loss": 0.584,
+      "step": 970
+    },
+    {
+      "epoch": 2.1541874653355517,
+      "grad_norm": 0.31558966636657715,
+      "learning_rate": 9.358024691358025e-05,
+      "loss": 0.3317,
+      "step": 971
+    },
+    {
+      "epoch": 2.156405990016639,
+      "grad_norm": 0.23257380723953247,
+      "learning_rate": 9.333333333333333e-05,
+      "loss": 0.3756,
+      "step": 972
+    },
+    {
+      "epoch": 2.158624514697726,
+      "grad_norm": 0.2579594850540161,
+      "learning_rate": 9.308641975308641e-05,
+      "loss": 0.5175,
+      "step": 973
+    },
+    {
+      "epoch": 2.160843039378813,
+      "grad_norm": 0.27227920293807983,
+      "learning_rate": 9.283950617283949e-05,
+      "loss": 0.461,
+      "step": 974
+    },
+    {
+      "epoch": 2.1630615640599,
+      "grad_norm": 0.20013847947120667,
+      "learning_rate": 9.259259259259257e-05,
+      "loss": 0.4845,
+      "step": 975
+    },
+    {
+      "epoch": 2.165280088740987,
+      "grad_norm": 0.23085667192935944,
+      "learning_rate": 9.234567901234567e-05,
+      "loss": 0.4663,
+      "step": 976
+    },
+    {
+      "epoch": 2.167498613422074,
+      "grad_norm": 0.20291349291801453,
+      "learning_rate": 9.209876543209876e-05,
+      "loss": 0.3105,
+      "step": 977
+    },
+    {
+      "epoch": 2.1697171381031612,
+      "grad_norm": 0.24949903786182404,
+      "learning_rate": 9.185185185185185e-05,
+      "loss": 0.437,
+      "step": 978
+    },
+    {
+      "epoch": 2.1719356627842483,
+      "grad_norm": 0.271541953086853,
+      "learning_rate": 9.160493827160493e-05,
+      "loss": 0.8248,
+      "step": 979
+    },
+    {
+      "epoch": 2.1741541874653354,
+      "grad_norm": 0.3158898651599884,
+      "learning_rate": 9.135802469135802e-05,
+      "loss": 0.457,
+      "step": 980
+    },
+    {
+      "epoch": 2.1763727121464225,
+      "grad_norm": 0.2580934166908264,
+      "learning_rate": 9.11111111111111e-05,
+      "loss": 0.3432,
+      "step": 981
+    },
+    {
+      "epoch": 2.1785912368275095,
+      "grad_norm": 0.26732900738716125,
+      "learning_rate": 9.086419753086419e-05,
+      "loss": 0.4277,
+      "step": 982
+    },
+    {
+      "epoch": 2.1808097615085966,
+      "grad_norm": 0.4489876329898834,
+      "learning_rate": 9.061728395061727e-05,
+      "loss": 0.6496,
+      "step": 983
+    },
+    {
+      "epoch": 2.1830282861896837,
+      "grad_norm": 0.24634425342082977,
+      "learning_rate": 9.037037037037038e-05,
+      "loss": 0.4576,
+      "step": 984
+    },
+    {
+      "epoch": 2.1852468108707708,
+      "grad_norm": 0.2975209951400757,
+      "learning_rate": 9.012345679012346e-05,
+      "loss": 0.4453,
+      "step": 985
+    },
+    {
+      "epoch": 2.187465335551858,
+      "grad_norm": 0.22377163171768188,
+      "learning_rate": 8.987654320987654e-05,
+      "loss": 0.4955,
+      "step": 986
+    },
+    {
+      "epoch": 2.189683860232945,
+      "grad_norm": 0.2111307829618454,
+      "learning_rate": 8.962962962962962e-05,
+      "loss": 0.3127,
+      "step": 987
+    },
+    {
+      "epoch": 2.191902384914032,
+      "grad_norm": 0.33432328701019287,
+      "learning_rate": 8.93827160493827e-05,
+      "loss": 0.3936,
+      "step": 988
+    },
+    {
+      "epoch": 2.194120909595119,
+      "grad_norm": 0.23631781339645386,
+      "learning_rate": 8.91358024691358e-05,
+      "loss": 0.4018,
+      "step": 989
+    },
+    {
+      "epoch": 2.196339434276206,
+      "grad_norm": 0.2523908317089081,
+      "learning_rate": 8.888888888888888e-05,
+      "loss": 0.4464,
+      "step": 990
+    },
+    {
+      "epoch": 2.1985579589572932,
+      "grad_norm": 0.24178925156593323,
+      "learning_rate": 8.864197530864196e-05,
+      "loss": 0.457,
+      "step": 991
+    },
+    {
+      "epoch": 2.2007764836383803,
+      "grad_norm": 0.20687422156333923,
+      "learning_rate": 8.839506172839506e-05,
+      "loss": 0.3687,
+      "step": 992
+    },
+    {
+      "epoch": 2.2029950083194674,
+      "grad_norm": 0.28263452649116516,
+      "learning_rate": 8.814814814814815e-05,
+      "loss": 0.8665,
+      "step": 993
+    },
+    {
+      "epoch": 2.2052135330005544,
+      "grad_norm": 0.24008749425411224,
+      "learning_rate": 8.790123456790123e-05,
+      "loss": 0.3297,
+      "step": 994
+    },
+    {
+      "epoch": 2.2074320576816415,
+      "grad_norm": 0.18816936016082764,
+      "learning_rate": 8.765432098765431e-05,
+      "loss": 0.352,
+      "step": 995
+    },
+    {
+      "epoch": 2.2096505823627286,
+      "grad_norm": 0.32737934589385986,
+      "learning_rate": 8.74074074074074e-05,
+      "loss": 0.5611,
+      "step": 996
+    },
+    {
+      "epoch": 2.2118691070438157,
+      "grad_norm": 0.20922309160232544,
+      "learning_rate": 8.716049382716048e-05,
+      "loss": 0.4396,
+      "step": 997
+    },
+    {
+      "epoch": 2.2140876317249027,
+      "grad_norm": 0.2731016278266907,
+      "learning_rate": 8.691358024691357e-05,
+      "loss": 0.5306,
+      "step": 998
+    },
+    {
+      "epoch": 2.21630615640599,
+      "grad_norm": 0.29952049255371094,
+      "learning_rate": 8.666666666666665e-05,
+      "loss": 0.3342,
+      "step": 999
+    },
+    {
+      "epoch": 2.218524681087077,
+      "grad_norm": 0.3391759693622589,
+      "learning_rate": 8.641975308641975e-05,
+      "loss": 0.4333,
+      "step": 1000
+    },
+    {
+      "epoch": 2.220743205768164,
+      "grad_norm": 0.31748226284980774,
+      "learning_rate": 8.617283950617283e-05,
+      "loss": 0.4655,
+      "step": 1001
+    },
+    {
+      "epoch": 2.222961730449251,
+      "grad_norm": 0.22518371045589447,
+      "learning_rate": 8.592592592592593e-05,
+      "loss": 0.5144,
+      "step": 1002
+    },
+    {
+      "epoch": 2.225180255130338,
+      "grad_norm": 0.24096240103244781,
+      "learning_rate": 8.567901234567901e-05,
+      "loss": 0.4571,
+      "step": 1003
+    },
+    {
+      "epoch": 2.227398779811425,
+      "grad_norm": 0.272975891828537,
+      "learning_rate": 8.543209876543209e-05,
+      "loss": 0.387,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2296173044925123,
+      "grad_norm": 0.27608585357666016,
+      "learning_rate": 8.518518518518517e-05,
+      "loss": 0.4335,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2318358291735994,
+      "grad_norm": 0.31609565019607544,
+      "learning_rate": 8.493827160493825e-05,
+      "loss": 0.5056,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2340543538546864,
+      "grad_norm": 0.22396518290042877,
+      "learning_rate": 8.469135802469136e-05,
+      "loss": 0.4071,
+      "step": 1007
+    },
+    {
+      "epoch": 2.2362728785357735,
+      "grad_norm": 0.26825276017189026,
+      "learning_rate": 8.444444444444444e-05,
+      "loss": 0.5391,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2384914032168606,
+      "grad_norm": 0.2613063156604767,
+      "learning_rate": 8.419753086419752e-05,
+      "loss": 0.5135,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2407099278979477,
+      "grad_norm": 0.20980030298233032,
+      "learning_rate": 8.39506172839506e-05,
+      "loss": 0.3927,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2429284525790347,
+      "grad_norm": 0.3035288453102112,
+      "learning_rate": 8.37037037037037e-05,
+      "loss": 0.4384,
+      "step": 1011
+    },
+    {
+      "epoch": 2.245146977260122,
+      "grad_norm": 0.20786920189857483,
+      "learning_rate": 8.345679012345678e-05,
+      "loss": 0.4056,
+      "step": 1012
+    },
+    {
+      "epoch": 2.247365501941209,
+      "grad_norm": 0.2734518349170685,
+      "learning_rate": 8.320987654320986e-05,
+      "loss": 0.4664,
+      "step": 1013
+    },
+    {
+      "epoch": 2.249584026622296,
+      "grad_norm": 0.2648674249649048,
+      "learning_rate": 8.296296296296295e-05,
+      "loss": 0.6017,
+      "step": 1014
+    },
+    {
+      "epoch": 2.251802551303383,
+      "grad_norm": 0.22290760278701782,
+      "learning_rate": 8.271604938271605e-05,
+      "loss": 0.5939,
+      "step": 1015
+    },
+    {
+      "epoch": 2.25402107598447,
+      "grad_norm": 0.26923441886901855,
+      "learning_rate": 8.246913580246914e-05,
+      "loss": 0.522,
+      "step": 1016
+    },
+    {
+      "epoch": 2.256239600665557,
+      "grad_norm": 0.2590654492378235,
+      "learning_rate": 8.222222222222222e-05,
+      "loss": 0.6658,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2584581253466443,
+      "grad_norm": 0.26027846336364746,
+      "learning_rate": 8.19753086419753e-05,
+      "loss": 0.6499,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2606766500277313,
+      "grad_norm": 0.266592800617218,
+      "learning_rate": 8.172839506172838e-05,
+      "loss": 0.609,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2628951747088184,
+      "grad_norm": 0.2997714877128601,
+      "learning_rate": 8.148148148148148e-05,
+      "loss": 0.5473,
+      "step": 1020
+    },
+    {
+      "epoch": 2.2651136993899055,
+      "grad_norm": 0.29512912034988403,
+      "learning_rate": 8.123456790123456e-05,
+      "loss": 0.5437,
+      "step": 1021
+    },
+    {
+      "epoch": 2.2673322240709926,
+      "grad_norm": 0.30122214555740356,
+      "learning_rate": 8.098765432098765e-05,
+      "loss": 0.4236,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2695507487520796,
+      "grad_norm": 0.27383488416671753,
+      "learning_rate": 8.074074074074073e-05,
+      "loss": 0.5202,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2717692734331667,
+      "grad_norm": 0.3243815004825592,
+      "learning_rate": 8.049382716049383e-05,
+      "loss": 0.5157,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2739877981142538,
+      "grad_norm": 0.1904493123292923,
+      "learning_rate": 8.024691358024691e-05,
+      "loss": 0.5468,
+      "step": 1025
+    },
+    {
+      "epoch": 2.276206322795341,
+      "grad_norm": 0.28189200162887573,
+      "learning_rate": 7.999999999999999e-05,
+      "loss": 0.444,
+      "step": 1026
+    },
+    {
+      "epoch": 2.278424847476428,
+      "grad_norm": 0.21269427239894867,
+      "learning_rate": 7.975308641975307e-05,
+      "loss": 0.4751,
+      "step": 1027
+    },
+    {
+      "epoch": 2.280643372157515,
+      "grad_norm": 0.2935750186443329,
+      "learning_rate": 7.950617283950616e-05,
+      "loss": 0.3623,
+      "step": 1028
+    },
+    {
+      "epoch": 2.2828618968386025,
+      "grad_norm": 0.2264406979084015,
+      "learning_rate": 7.925925925925925e-05,
+      "loss": 0.4865,
+      "step": 1029
+    },
+    {
+      "epoch": 2.2850804215196896,
+      "grad_norm": 0.2584576904773712,
+      "learning_rate": 7.901234567901235e-05,
+      "loss": 0.5686,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2872989462007767,
+      "grad_norm": 0.36906859278678894,
+      "learning_rate": 7.876543209876543e-05,
+      "loss": 0.5463,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2895174708818637,
+      "grad_norm": 0.24862168729305267,
+      "learning_rate": 7.851851851851851e-05,
+      "loss": 0.5172,
+      "step": 1032
+    },
+    {
+      "epoch": 2.291735995562951,
+      "grad_norm": 0.28619495034217834,
+      "learning_rate": 7.82716049382716e-05,
+      "loss": 0.364,
+      "step": 1033
+    },
+    {
+      "epoch": 2.293954520244038,
+      "grad_norm": 0.3470110595226288,
+      "learning_rate": 7.802469135802469e-05,
+      "loss": 0.4848,
+      "step": 1034
+    },
+    {
+      "epoch": 2.296173044925125,
+      "grad_norm": 0.231489360332489,
+      "learning_rate": 7.777777777777777e-05,
+      "loss": 0.3884,
+      "step": 1035
+    },
+    {
+      "epoch": 2.298391569606212,
+      "grad_norm": 0.24071519076824188,
+      "learning_rate": 7.753086419753085e-05,
+      "loss": 0.4686,
+      "step": 1036
+    },
+    {
+      "epoch": 2.300610094287299,
+      "grad_norm": 0.2599813640117645,
+      "learning_rate": 7.728395061728396e-05,
+      "loss": 0.393,
+      "step": 1037
+    },
+    {
+      "epoch": 2.302828618968386,
+      "grad_norm": 0.21404802799224854,
+      "learning_rate": 7.703703703703704e-05,
+      "loss": 0.5799,
+      "step": 1038
+    },
+    {
+      "epoch": 2.3050471436494733,
+      "grad_norm": 0.25654473900794983,
+      "learning_rate": 7.679012345679012e-05,
+      "loss": 0.4815,
+      "step": 1039
+    },
+    {
+      "epoch": 2.3072656683305603,
+      "grad_norm": 0.2517067790031433,
+      "learning_rate": 7.65432098765432e-05,
+      "loss": 0.4869,
+      "step": 1040
+    },
+    {
+      "epoch": 2.3094841930116474,
+      "grad_norm": 0.2651129364967346,
+      "learning_rate": 7.629629629629628e-05,
+      "loss": 0.5992,
+      "step": 1041
+    },
+    {
+      "epoch": 2.3117027176927345,
+      "grad_norm": 0.26585349440574646,
+      "learning_rate": 7.604938271604937e-05,
+      "loss": 0.4377,
+      "step": 1042
+    },
+    {
+      "epoch": 2.3139212423738216,
+      "grad_norm": 0.4704441726207733,
+      "learning_rate": 7.580246913580246e-05,
+      "loss": 0.46,
+      "step": 1043
+    },
+    {
+      "epoch": 2.3161397670549086,
+      "grad_norm": 0.295143723487854,
+      "learning_rate": 7.555555555555554e-05,
+      "loss": 0.4813,
+      "step": 1044
+    },
+    {
+      "epoch": 2.3183582917359957,
+      "grad_norm": 0.3856975734233856,
+      "learning_rate": 7.530864197530864e-05,
+      "loss": 0.4784,
+      "step": 1045
+    },
+    {
+      "epoch": 2.320576816417083,
+      "grad_norm": 0.2904879152774811,
+      "learning_rate": 7.506172839506173e-05,
+      "loss": 0.5739,
+      "step": 1046
+    },
+    {
+      "epoch": 2.32279534109817,
+      "grad_norm": 0.270382285118103,
+      "learning_rate": 7.481481481481481e-05,
+      "loss": 0.4238,
+      "step": 1047
+    },
+    {
+      "epoch": 2.325013865779257,
+      "grad_norm": 0.22698435187339783,
+      "learning_rate": 7.45679012345679e-05,
+      "loss": 0.4896,
+      "step": 1048
+    },
+    {
+      "epoch": 2.327232390460344,
+      "grad_norm": 0.3453735113143921,
+      "learning_rate": 7.432098765432098e-05,
+      "loss": 0.5085,
+      "step": 1049
+    },
+    {
+      "epoch": 2.329450915141431,
+      "grad_norm": 0.2840602695941925,
+      "learning_rate": 7.407407407407406e-05,
+      "loss": 0.493,
+      "step": 1050
+    },
+    {
+      "epoch": 2.331669439822518,
+      "grad_norm": 0.2767769992351532,
+      "learning_rate": 7.382716049382715e-05,
+      "loss": 0.5162,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3338879645036053,
+      "grad_norm": 0.2660716772079468,
+      "learning_rate": 7.358024691358024e-05,
+      "loss": 0.382,
+      "step": 1052
+    },
+    {
+      "epoch": 2.3361064891846923,
+      "grad_norm": 0.33769261837005615,
+      "learning_rate": 7.333333333333332e-05,
+      "loss": 0.6231,
+      "step": 1053
+    },
+    {
+      "epoch": 2.3383250138657794,
+      "grad_norm": 0.3232424557209015,
+      "learning_rate": 7.308641975308641e-05,
+      "loss": 0.5082,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3405435385468665,
+      "grad_norm": 0.24822954833507538,
+      "learning_rate": 7.28395061728395e-05,
+      "loss": 0.5166,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3427620632279536,
+      "grad_norm": 0.24538779258728027,
+      "learning_rate": 7.259259259259259e-05,
+      "loss": 0.5491,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3449805879090406,
+      "grad_norm": 0.2786666750907898,
+      "learning_rate": 7.234567901234567e-05,
+      "loss": 0.441,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3471991125901277,
+      "grad_norm": 0.26666828989982605,
+      "learning_rate": 7.209876543209877e-05,
+      "loss": 0.4572,
+      "step": 1058
+    },
+    {
+      "epoch": 2.3494176372712148,
+      "grad_norm": 0.2929750084877014,
+      "learning_rate": 7.185185185185185e-05,
+      "loss": 0.4823,
+      "step": 1059
+    },
+    {
+      "epoch": 2.351636161952302,
+      "grad_norm": 0.27779436111450195,
+      "learning_rate": 7.160493827160493e-05,
+      "loss": 0.3989,
+      "step": 1060
+    },
+    {
+      "epoch": 2.353854686633389,
+      "grad_norm": 0.23364423215389252,
+      "learning_rate": 7.135802469135801e-05,
+      "loss": 0.4281,
+      "step": 1061
+    },
+    {
+      "epoch": 2.356073211314476,
+      "grad_norm": 0.2560143768787384,
+      "learning_rate": 7.11111111111111e-05,
+      "loss": 0.5097,
+      "step": 1062
+    },
+    {
+      "epoch": 2.358291735995563,
+      "grad_norm": 0.24645791947841644,
+      "learning_rate": 7.086419753086419e-05,
+      "loss": 0.4215,
+      "step": 1063
+    },
+    {
+      "epoch": 2.36051026067665,
+      "grad_norm": 0.3272581398487091,
+      "learning_rate": 7.061728395061727e-05,
+      "loss": 0.4775,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3627287853577372,
+      "grad_norm": 0.34457358717918396,
+      "learning_rate": 7.037037037037036e-05,
+      "loss": 0.3729,
+      "step": 1065
+    },
+    {
+      "epoch": 2.3649473100388243,
+      "grad_norm": 0.2311803102493286,
+      "learning_rate": 7.012345679012345e-05,
+      "loss": 0.4713,
+      "step": 1066
+    },
+    {
+      "epoch": 2.3671658347199114,
+      "grad_norm": 0.3108791410923004,
+      "learning_rate": 6.987654320987654e-05,
+      "loss": 0.5174,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3693843594009985,
+      "grad_norm": 0.3345027267932892,
+      "learning_rate": 6.962962962962962e-05,
+      "loss": 0.6065,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3716028840820855,
+      "grad_norm": 0.24299198389053345,
+      "learning_rate": 6.938271604938272e-05,
+      "loss": 0.3747,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3738214087631726,
+      "grad_norm": 0.2534796893596649,
+      "learning_rate": 6.91358024691358e-05,
+      "loss": 0.4053,
+      "step": 1070
+    },
+    {
+      "epoch": 2.3760399334442597,
+      "grad_norm": 0.29387426376342773,
+      "learning_rate": 6.888888888888888e-05,
+      "loss": 0.5884,
+      "step": 1071
+    },
+    {
+      "epoch": 2.3782584581253468,
+      "grad_norm": 0.24638226628303528,
+      "learning_rate": 6.864197530864196e-05,
+      "loss": 0.567,
+      "step": 1072
+    },
+    {
+      "epoch": 2.380476982806434,
+      "grad_norm": 0.23020130395889282,
+      "learning_rate": 6.839506172839506e-05,
+      "loss": 0.5617,
+      "step": 1073
+    },
+    {
+      "epoch": 2.382695507487521,
+      "grad_norm": 0.28502628207206726,
+      "learning_rate": 6.814814814814814e-05,
+      "loss": 0.532,
+      "step": 1074
+    },
+    {
+      "epoch": 2.384914032168608,
+      "grad_norm": 0.318715363740921,
+      "learning_rate": 6.790123456790122e-05,
+      "loss": 0.3711,
+      "step": 1075
+    },
+    {
+      "epoch": 2.387132556849695,
+      "grad_norm": 0.2816349267959595,
+      "learning_rate": 6.765432098765432e-05,
+      "loss": 0.4025,
+      "step": 1076
+    },
+    {
+      "epoch": 2.389351081530782,
+      "grad_norm": 0.26949313282966614,
+      "learning_rate": 6.74074074074074e-05,
+      "loss": 0.6569,
+      "step": 1077
+    },
+    {
+      "epoch": 2.391569606211869,
+      "grad_norm": 0.3079036474227905,
+      "learning_rate": 6.716049382716049e-05,
+      "loss": 0.7052,
+      "step": 1078
+    },
+    {
+      "epoch": 2.3937881308929563,
+      "grad_norm": 0.4732149839401245,
+      "learning_rate": 6.691358024691357e-05,
+      "loss": 0.4576,
+      "step": 1079
+    },
+    {
+      "epoch": 2.3960066555740434,
+      "grad_norm": 0.2551252841949463,
+      "learning_rate": 6.666666666666666e-05,
+      "loss": 0.463,
+      "step": 1080
+    },
+    {
+      "epoch": 2.3982251802551304,
+      "grad_norm": 0.2529161870479584,
+      "learning_rate": 6.641975308641975e-05,
+      "loss": 0.4668,
+      "step": 1081
+    },
+    {
+      "epoch": 2.4004437049362175,
+      "grad_norm": 0.3839707374572754,
+      "learning_rate": 6.617283950617283e-05,
+      "loss": 0.4706,
+      "step": 1082
+    },
+    {
+      "epoch": 2.4026622296173046,
+      "grad_norm": 0.21397164463996887,
+      "learning_rate": 6.592592592592591e-05,
+      "loss": 0.4488,
+      "step": 1083
+    },
+    {
+      "epoch": 2.4048807542983917,
+      "grad_norm": 0.26458674669265747,
+      "learning_rate": 6.567901234567901e-05,
+      "loss": 0.4549,
+      "step": 1084
+    },
+    {
+      "epoch": 2.4070992789794787,
+      "grad_norm": 0.2645561695098877,
+      "learning_rate": 6.543209876543209e-05,
+      "loss": 0.4099,
+      "step": 1085
+    },
+    {
+      "epoch": 2.409317803660566,
+      "grad_norm": 0.26775380969047546,
+      "learning_rate": 6.518518518518517e-05,
+      "loss": 0.4206,
+      "step": 1086
+    },
+    {
+      "epoch": 2.411536328341653,
+      "grad_norm": 0.2704344093799591,
+      "learning_rate": 6.493827160493827e-05,
+      "loss": 0.4124,
+      "step": 1087
+    },
+    {
+      "epoch": 2.41375485302274,
+      "grad_norm": 0.2871568202972412,
+      "learning_rate": 6.469135802469135e-05,
+      "loss": 0.374,
+      "step": 1088
+    },
+    {
+      "epoch": 2.415973377703827,
+      "grad_norm": 0.24475519359111786,
+      "learning_rate": 6.444444444444444e-05,
+      "loss": 0.4005,
+      "step": 1089
+    },
+    {
+      "epoch": 2.418191902384914,
+      "grad_norm": 0.2873852550983429,
+      "learning_rate": 6.419753086419753e-05,
+      "loss": 0.4834,
+      "step": 1090
+    },
+    {
+      "epoch": 2.420410427066001,
+      "grad_norm": 0.24380742013454437,
+      "learning_rate": 6.395061728395061e-05,
+      "loss": 0.5039,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4226289517470883,
+      "grad_norm": 0.23561698198318481,
+      "learning_rate": 6.37037037037037e-05,
+      "loss": 0.3788,
+      "step": 1092
+    },
+    {
+      "epoch": 2.4248474764281753,
+      "grad_norm": 0.31819596886634827,
+      "learning_rate": 6.345679012345678e-05,
+      "loss": 0.4645,
+      "step": 1093
+    },
+    {
+      "epoch": 2.4270660011092624,
+      "grad_norm": 0.3289719521999359,
+      "learning_rate": 6.320987654320987e-05,
+      "loss": 0.5429,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4292845257903495,
+      "grad_norm": 0.34532973170280457,
+      "learning_rate": 6.296296296296295e-05,
+      "loss": 0.3918,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4315030504714366,
+      "grad_norm": 0.2644065320491791,
+      "learning_rate": 6.271604938271604e-05,
+      "loss": 0.528,
+      "step": 1096
+    },
+    {
+      "epoch": 2.4337215751525236,
+      "grad_norm": 0.256035178899765,
+      "learning_rate": 6.246913580246912e-05,
+      "loss": 0.4772,
+      "step": 1097
+    },
+    {
+      "epoch": 2.4359400998336107,
+      "grad_norm": 0.2972947359085083,
+      "learning_rate": 6.222222222222222e-05,
+      "loss": 0.5188,
+      "step": 1098
+    },
+    {
+      "epoch": 2.438158624514698,
+      "grad_norm": 0.2419045865535736,
+      "learning_rate": 6.19753086419753e-05,
+      "loss": 0.4515,
+      "step": 1099
+    },
+    {
+      "epoch": 2.440377149195785,
+      "grad_norm": 0.278051495552063,
+      "learning_rate": 6.17283950617284e-05,
+      "loss": 0.4469,
+      "step": 1100
+    },
+    {
+      "epoch": 2.442595673876872,
+      "grad_norm": 0.2693118751049042,
+      "learning_rate": 6.148148148148148e-05,
+      "loss": 0.4733,
+      "step": 1101
+    },
+    {
+      "epoch": 2.444814198557959,
+      "grad_norm": 0.25968068838119507,
+      "learning_rate": 6.123456790123456e-05,
+      "loss": 0.459,
+      "step": 1102
+    },
+    {
+      "epoch": 2.447032723239046,
+      "grad_norm": 0.2818559408187866,
+      "learning_rate": 6.0987654320987655e-05,
+      "loss": 0.4864,
+      "step": 1103
+    },
+    {
+      "epoch": 2.449251247920133,
+      "grad_norm": 0.24768030643463135,
+      "learning_rate": 6.0740740740740736e-05,
+      "loss": 0.3507,
+      "step": 1104
+    },
+    {
+      "epoch": 2.4514697726012202,
+      "grad_norm": 0.27249452471733093,
+      "learning_rate": 6.049382716049382e-05,
+      "loss": 0.406,
+      "step": 1105
+    },
+    {
+      "epoch": 2.4536882972823073,
+      "grad_norm": 0.2984795868396759,
+      "learning_rate": 6.0246913580246906e-05,
+      "loss": 0.4504,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4559068219633944,
+      "grad_norm": 0.350572407245636,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 0.4131,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4581253466444815,
+      "grad_norm": 0.22387048602104187,
+      "learning_rate": 5.975308641975308e-05,
+      "loss": 0.3973,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4603438713255685,
+      "grad_norm": 0.27799737453460693,
+      "learning_rate": 5.9506172839506165e-05,
+      "loss": 0.306,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4625623960066556,
+      "grad_norm": 0.30395323038101196,
+      "learning_rate": 5.925925925925925e-05,
+      "loss": 0.4097,
+      "step": 1110
+    },
+    {
+      "epoch": 2.4647809206877427,
+      "grad_norm": 0.2956414818763733,
+      "learning_rate": 5.901234567901234e-05,
+      "loss": 0.4217,
+      "step": 1111
+    },
+    {
+      "epoch": 2.4669994453688298,
+      "grad_norm": 0.24104991555213928,
+      "learning_rate": 5.876543209876543e-05,
+      "loss": 0.4049,
+      "step": 1112
+    },
+    {
+      "epoch": 2.469217970049917,
+      "grad_norm": 0.33756014704704285,
+      "learning_rate": 5.851851851851851e-05,
+      "loss": 0.434,
+      "step": 1113
+    },
+    {
+      "epoch": 2.471436494731004,
+      "grad_norm": 0.2566487491130829,
+      "learning_rate": 5.827160493827159e-05,
+      "loss": 0.4902,
+      "step": 1114
+    },
+    {
+      "epoch": 2.473655019412091,
+      "grad_norm": 0.29076656699180603,
+      "learning_rate": 5.802469135802469e-05,
+      "loss": 0.4188,
+      "step": 1115
+    },
+    {
+      "epoch": 2.475873544093178,
+      "grad_norm": 0.3608337938785553,
+      "learning_rate": 5.777777777777777e-05,
+      "loss": 0.559,
+      "step": 1116
+    },
+    {
+      "epoch": 2.478092068774265,
+      "grad_norm": 0.27408990263938904,
+      "learning_rate": 5.753086419753086e-05,
+      "loss": 0.4439,
+      "step": 1117
+    },
+    {
+      "epoch": 2.480310593455352,
+      "grad_norm": 0.264445036649704,
+      "learning_rate": 5.7283950617283946e-05,
+      "loss": 0.4161,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4825291181364393,
+      "grad_norm": 0.280110239982605,
+      "learning_rate": 5.7037037037037035e-05,
+      "loss": 0.504,
+      "step": 1119
+    },
+    {
+      "epoch": 2.4847476428175264,
+      "grad_norm": 0.34883859753608704,
+      "learning_rate": 5.6790123456790116e-05,
+      "loss": 0.518,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4869661674986134,
+      "grad_norm": 0.33939680457115173,
+      "learning_rate": 5.6543209876543205e-05,
+      "loss": 0.416,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4891846921797005,
+      "grad_norm": 0.2898100018501282,
+      "learning_rate": 5.629629629629629e-05,
+      "loss": 0.5552,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4914032168607876,
+      "grad_norm": 0.27850252389907837,
+      "learning_rate": 5.604938271604938e-05,
+      "loss": 0.4675,
+      "step": 1123
+    },
+    {
+      "epoch": 2.4936217415418747,
+      "grad_norm": 0.3024316728115082,
+      "learning_rate": 5.580246913580246e-05,
+      "loss": 0.5366,
+      "step": 1124
+    },
+    {
+      "epoch": 2.4958402662229617,
+      "grad_norm": 0.24996912479400635,
+      "learning_rate": 5.5555555555555545e-05,
+      "loss": 0.536,
+      "step": 1125
+    },
+    {
+      "epoch": 2.498058790904049,
+      "grad_norm": 0.27548402547836304,
+      "learning_rate": 5.530864197530864e-05,
+      "loss": 0.4876,
+      "step": 1126
+    },
+    {
+      "epoch": 2.500277315585136,
+      "grad_norm": 0.2569044232368469,
+      "learning_rate": 5.506172839506172e-05,
+      "loss": 0.4103,
+      "step": 1127
+    },
+    {
+      "epoch": 2.502495840266223,
+      "grad_norm": 0.22059248387813568,
+      "learning_rate": 5.481481481481481e-05,
+      "loss": 0.4744,
+      "step": 1128
+    },
+    {
+      "epoch": 2.50471436494731,
+      "grad_norm": 0.26526302099227905,
+      "learning_rate": 5.456790123456789e-05,
+      "loss": 0.5068,
+      "step": 1129
+    },
+    {
+      "epoch": 2.506932889628397,
+      "grad_norm": 0.2807089388370514,
+      "learning_rate": 5.4320987654320986e-05,
+      "loss": 0.3713,
+      "step": 1130
+    },
+    {
+      "epoch": 2.509151414309484,
+      "grad_norm": 0.27579501271247864,
+      "learning_rate": 5.407407407407407e-05,
+      "loss": 0.4234,
+      "step": 1131
+    },
+    {
+      "epoch": 2.5113699389905713,
+      "grad_norm": 0.29730308055877686,
+      "learning_rate": 5.3827160493827156e-05,
+      "loss": 0.5592,
+      "step": 1132
+    },
+    {
+      "epoch": 2.5135884636716583,
+      "grad_norm": 0.25888770818710327,
+      "learning_rate": 5.358024691358024e-05,
+      "loss": 0.4177,
+      "step": 1133
+    },
+    {
+      "epoch": 2.5158069883527454,
+      "grad_norm": 0.2795535922050476,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.586,
+      "step": 1134
+    },
+    {
+      "epoch": 2.5180255130338325,
+      "grad_norm": 0.3462004065513611,
+      "learning_rate": 5.3086419753086415e-05,
+      "loss": 0.312,
+      "step": 1135
+    },
+    {
+      "epoch": 2.5202440377149196,
+      "grad_norm": 0.4033753573894501,
+      "learning_rate": 5.2839506172839496e-05,
+      "loss": 0.5013,
+      "step": 1136
+    },
+    {
+      "epoch": 2.5224625623960066,
+      "grad_norm": 0.3046059310436249,
+      "learning_rate": 5.259259259259259e-05,
+      "loss": 0.4787,
+      "step": 1137
+    },
+    {
+      "epoch": 2.5246810870770937,
+      "grad_norm": 0.242131307721138,
+      "learning_rate": 5.234567901234567e-05,
+      "loss": 0.433,
+      "step": 1138
+    },
+    {
+      "epoch": 2.526899611758181,
+      "grad_norm": 0.38816171884536743,
+      "learning_rate": 5.209876543209876e-05,
+      "loss": 0.6686,
+      "step": 1139
+    },
+    {
+      "epoch": 2.529118136439268,
+      "grad_norm": 0.2915274202823639,
+      "learning_rate": 5.185185185185184e-05,
+      "loss": 0.42,
+      "step": 1140
+    },
+    {
+      "epoch": 2.531336661120355,
+      "grad_norm": 0.3044760525226593,
+      "learning_rate": 5.160493827160494e-05,
+      "loss": 0.3537,
+      "step": 1141
+    },
+    {
+      "epoch": 2.533555185801442,
+      "grad_norm": 0.24314135313034058,
+      "learning_rate": 5.135802469135802e-05,
+      "loss": 0.4703,
+      "step": 1142
+    },
+    {
+      "epoch": 2.535773710482529,
+      "grad_norm": 0.2895704209804535,
+      "learning_rate": 5.111111111111111e-05,
+      "loss": 0.5316,
+      "step": 1143
+    },
+    {
+      "epoch": 2.537992235163616,
+      "grad_norm": 0.44123467803001404,
+      "learning_rate": 5.086419753086419e-05,
+      "loss": 0.7442,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5402107598447032,
+      "grad_norm": 0.23908613622188568,
+      "learning_rate": 5.0617283950617285e-05,
+      "loss": 0.3984,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5424292845257903,
+      "grad_norm": 0.29049402475357056,
+      "learning_rate": 5.0370370370370366e-05,
+      "loss": 0.4492,
+      "step": 1146
+    },
+    {
+      "epoch": 2.5446478092068774,
+      "grad_norm": 0.3551279902458191,
+      "learning_rate": 5.012345679012345e-05,
+      "loss": 0.7222,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5468663338879645,
+      "grad_norm": 0.294590026140213,
+      "learning_rate": 4.9876543209876536e-05,
+      "loss": 0.5301,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5490848585690515,
+      "grad_norm": 0.31074008345603943,
+      "learning_rate": 4.9629629629629625e-05,
+      "loss": 0.4763,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5513033832501386,
+      "grad_norm": 0.318962037563324,
+      "learning_rate": 4.938271604938271e-05,
+      "loss": 0.4458,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5535219079312257,
+      "grad_norm": 0.2321411669254303,
+      "learning_rate": 4.9135802469135795e-05,
+      "loss": 0.497,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5557404326123128,
+      "grad_norm": 0.3084535300731659,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.4997,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5579589572934,
+      "grad_norm": 0.22906942665576935,
+      "learning_rate": 4.864197530864197e-05,
+      "loss": 0.38,
+      "step": 1153
+    },
+    {
+      "epoch": 2.560177481974487,
+      "grad_norm": 0.26602914929389954,
+      "learning_rate": 4.839506172839506e-05,
+      "loss": 0.4671,
+      "step": 1154
+    },
+    {
+      "epoch": 2.562396006655574,
+      "grad_norm": 0.29763728380203247,
+      "learning_rate": 4.814814814814814e-05,
+      "loss": 0.5487,
+      "step": 1155
+    },
+    {
+      "epoch": 2.564614531336661,
+      "grad_norm": 0.3016784191131592,
+      "learning_rate": 4.7901234567901237e-05,
+      "loss": 0.4661,
+      "step": 1156
+    },
+    {
+      "epoch": 2.566833056017748,
+      "grad_norm": 0.2467387467622757,
+      "learning_rate": 4.765432098765432e-05,
+      "loss": 0.3377,
+      "step": 1157
+    },
+    {
+      "epoch": 2.569051580698835,
+      "grad_norm": 0.2624308466911316,
+      "learning_rate": 4.74074074074074e-05,
+      "loss": 0.5323,
+      "step": 1158
+    },
+    {
+      "epoch": 2.5712701053799223,
+      "grad_norm": 0.3167112171649933,
+      "learning_rate": 4.716049382716049e-05,
+      "loss": 0.4772,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5734886300610094,
+      "grad_norm": 0.3810291290283203,
+      "learning_rate": 4.6913580246913576e-05,
+      "loss": 0.4645,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5757071547420964,
+      "grad_norm": 0.26205170154571533,
+      "learning_rate": 4.6666666666666665e-05,
+      "loss": 0.5172,
+      "step": 1161
+    },
+    {
+      "epoch": 2.5779256794231835,
+      "grad_norm": 0.2953357994556427,
+      "learning_rate": 4.6419753086419746e-05,
+      "loss": 0.4755,
+      "step": 1162
+    },
+    {
+      "epoch": 2.5801442041042706,
+      "grad_norm": 0.2811224162578583,
+      "learning_rate": 4.6172839506172835e-05,
+      "loss": 0.5993,
+      "step": 1163
+    },
+    {
+      "epoch": 2.5823627287853577,
+      "grad_norm": 0.31280505657196045,
+      "learning_rate": 4.592592592592592e-05,
+      "loss": 0.4846,
+      "step": 1164
+    },
+    {
+      "epoch": 2.5845812534664447,
+      "grad_norm": 0.25243163108825684,
+      "learning_rate": 4.567901234567901e-05,
+      "loss": 0.3582,
+      "step": 1165
+    },
+    {
+      "epoch": 2.586799778147532,
+      "grad_norm": 0.23925447463989258,
+      "learning_rate": 4.543209876543209e-05,
+      "loss": 0.3704,
+      "step": 1166
+    },
+    {
+      "epoch": 2.589018302828619,
+      "grad_norm": 0.36358100175857544,
+      "learning_rate": 4.518518518518519e-05,
+      "loss": 0.5795,
+      "step": 1167
+    },
+    {
+      "epoch": 2.591236827509706,
+      "grad_norm": 0.30875033140182495,
+      "learning_rate": 4.493827160493827e-05,
+      "loss": 0.4595,
+      "step": 1168
+    },
+    {
+      "epoch": 2.593455352190793,
+      "grad_norm": 0.23085197806358337,
+      "learning_rate": 4.469135802469135e-05,
+      "loss": 0.4103,
+      "step": 1169
+    },
+    {
+      "epoch": 2.59567387687188,
+      "grad_norm": 0.23379531502723694,
+      "learning_rate": 4.444444444444444e-05,
+      "loss": 0.4347,
+      "step": 1170
+    },
+    {
+      "epoch": 2.597892401552967,
+      "grad_norm": 0.277154803276062,
+      "learning_rate": 4.419753086419753e-05,
+      "loss": 0.4834,
+      "step": 1171
+    },
+    {
+      "epoch": 2.6001109262340543,
+      "grad_norm": 0.2907721698284149,
+      "learning_rate": 4.3950617283950617e-05,
+      "loss": 0.5494,
+      "step": 1172
+    },
+    {
+      "epoch": 2.6023294509151413,
+      "grad_norm": 0.339613676071167,
+      "learning_rate": 4.37037037037037e-05,
+      "loss": 0.4863,
+      "step": 1173
+    },
+    {
+      "epoch": 2.6045479755962284,
+      "grad_norm": 0.2214052975177765,
+      "learning_rate": 4.3456790123456786e-05,
+      "loss": 0.3991,
+      "step": 1174
+    },
+    {
+      "epoch": 2.6067665002773155,
+      "grad_norm": 0.29512521624565125,
+      "learning_rate": 4.3209876543209875e-05,
+      "loss": 0.4203,
+      "step": 1175
+    },
+    {
+      "epoch": 2.6089850249584026,
+      "grad_norm": 0.24430285394191742,
+      "learning_rate": 4.296296296296296e-05,
+      "loss": 0.4429,
+      "step": 1176
+    },
+    {
+      "epoch": 2.6112035496394896,
+      "grad_norm": 0.31122326850891113,
+      "learning_rate": 4.2716049382716045e-05,
+      "loss": 0.4974,
+      "step": 1177
+    },
+    {
+      "epoch": 2.6134220743205767,
+      "grad_norm": 0.3519100844860077,
+      "learning_rate": 4.2469135802469126e-05,
+      "loss": 0.5251,
+      "step": 1178
+    },
+    {
+      "epoch": 2.615640599001664,
+      "grad_norm": 0.28825023770332336,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.5257,
+      "step": 1179
+    },
+    {
+      "epoch": 2.617859123682751,
+      "grad_norm": 0.3672938942909241,
+      "learning_rate": 4.19753086419753e-05,
+      "loss": 0.4635,
+      "step": 1180
+    },
+    {
+      "epoch": 2.620077648363838,
+      "grad_norm": 0.28490591049194336,
+      "learning_rate": 4.172839506172839e-05,
+      "loss": 0.46,
+      "step": 1181
+    },
+    {
+      "epoch": 2.622296173044925,
+      "grad_norm": 0.27138692140579224,
+      "learning_rate": 4.148148148148147e-05,
+      "loss": 0.5118,
+      "step": 1182
+    },
+    {
+      "epoch": 2.624514697726012,
+      "grad_norm": 0.25375887751579285,
+      "learning_rate": 4.123456790123457e-05,
+      "loss": 0.2886,
+      "step": 1183
+    },
+    {
+      "epoch": 2.626733222407099,
+      "grad_norm": 0.2877887189388275,
+      "learning_rate": 4.098765432098765e-05,
+      "loss": 0.4806,
+      "step": 1184
+    },
+    {
+      "epoch": 2.6289517470881862,
+      "grad_norm": 0.31586790084838867,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.5596,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6311702717692733,
+      "grad_norm": 0.2568565309047699,
+      "learning_rate": 4.0493827160493827e-05,
+      "loss": 0.3547,
+      "step": 1186
+    },
+    {
+      "epoch": 2.6333887964503604,
+      "grad_norm": 0.33998599648475647,
+      "learning_rate": 4.0246913580246915e-05,
+      "loss": 0.5886,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6356073211314475,
+      "grad_norm": 0.27125903964042664,
+      "learning_rate": 3.9999999999999996e-05,
+      "loss": 0.3431,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6378258458125345,
+      "grad_norm": 0.2887658178806305,
+      "learning_rate": 3.975308641975308e-05,
+      "loss": 0.4139,
+      "step": 1189
+    },
+    {
+      "epoch": 2.6400443704936216,
+      "grad_norm": 0.21091485023498535,
+      "learning_rate": 3.950617283950617e-05,
+      "loss": 0.3995,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6422628951747087,
+      "grad_norm": 0.2706880569458008,
+      "learning_rate": 3.9259259259259255e-05,
+      "loss": 0.4124,
+      "step": 1191
+    },
+    {
+      "epoch": 2.6444814198557958,
+      "grad_norm": 0.30743926763534546,
+      "learning_rate": 3.901234567901234e-05,
+      "loss": 0.4902,
+      "step": 1192
+    },
+    {
+      "epoch": 2.646699944536883,
+      "grad_norm": 0.30735859274864197,
+      "learning_rate": 3.8765432098765425e-05,
+      "loss": 0.5958,
+      "step": 1193
+    },
+    {
+      "epoch": 2.64891846921797,
+      "grad_norm": 0.3423071801662445,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 0.4286,
+      "step": 1194
+    },
+    {
+      "epoch": 2.651136993899057,
+      "grad_norm": 0.3121039867401123,
+      "learning_rate": 3.82716049382716e-05,
+      "loss": 0.3855,
+      "step": 1195
+    },
+    {
+      "epoch": 2.653355518580144,
+      "grad_norm": 0.3074365556240082,
+      "learning_rate": 3.802469135802468e-05,
+      "loss": 0.3878,
+      "step": 1196
+    },
+    {
+      "epoch": 2.655574043261231,
+      "grad_norm": 0.32958588004112244,
+      "learning_rate": 3.777777777777777e-05,
+      "loss": 0.3788,
+      "step": 1197
+    },
+    {
+      "epoch": 2.657792567942318,
+      "grad_norm": 0.26261627674102783,
+      "learning_rate": 3.7530864197530867e-05,
+      "loss": 0.4137,
+      "step": 1198
+    },
+    {
+      "epoch": 2.6600110926234053,
+      "grad_norm": 0.2910056412220001,
+      "learning_rate": 3.728395061728395e-05,
+      "loss": 0.4672,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6622296173044924,
+      "grad_norm": 0.37444978952407837,
+      "learning_rate": 3.703703703703703e-05,
+      "loss": 0.5146,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6644481419855794,
+      "grad_norm": 0.29455071687698364,
+      "learning_rate": 3.679012345679012e-05,
+      "loss": 0.4292,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.28878548741340637,
+      "learning_rate": 3.6543209876543207e-05,
+      "loss": 0.5118,
+      "step": 1202
+    },
+    {
+      "epoch": 2.6688851913477536,
+      "grad_norm": 0.2594594359397888,
+      "learning_rate": 3.6296296296296295e-05,
+      "loss": 0.6068,
+      "step": 1203
+    },
+    {
+      "epoch": 2.6711037160288407,
+      "grad_norm": 0.33341702818870544,
+      "learning_rate": 3.604938271604938e-05,
+      "loss": 0.7119,
+      "step": 1204
+    },
+    {
+      "epoch": 2.6733222407099277,
+      "grad_norm": 0.2840867340564728,
+      "learning_rate": 3.5802469135802465e-05,
+      "loss": 0.4428,
+      "step": 1205
+    },
+    {
+      "epoch": 2.675540765391015,
+      "grad_norm": 0.24904444813728333,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 0.4767,
+      "step": 1206
+    },
+    {
+      "epoch": 2.677759290072102,
+      "grad_norm": 0.23573899269104004,
+      "learning_rate": 3.5308641975308635e-05,
+      "loss": 0.4254,
+      "step": 1207
+    },
+    {
+      "epoch": 2.679977814753189,
+      "grad_norm": 0.30268239974975586,
+      "learning_rate": 3.506172839506172e-05,
+      "loss": 0.5123,
+      "step": 1208
+    },
+    {
+      "epoch": 2.682196339434276,
+      "grad_norm": 0.3044212758541107,
+      "learning_rate": 3.481481481481481e-05,
+      "loss": 0.4331,
+      "step": 1209
+    },
+    {
+      "epoch": 2.684414864115363,
+      "grad_norm": 0.2583848237991333,
+      "learning_rate": 3.45679012345679e-05,
+      "loss": 0.4305,
+      "step": 1210
+    },
+    {
+      "epoch": 2.68663338879645,
+      "grad_norm": 0.28169476985931396,
+      "learning_rate": 3.432098765432098e-05,
+      "loss": 0.4784,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6888519134775377,
+      "grad_norm": 0.2685352861881256,
+      "learning_rate": 3.407407407407407e-05,
+      "loss": 0.5689,
+      "step": 1212
+    },
+    {
+      "epoch": 2.691070438158625,
+      "grad_norm": 0.257254958152771,
+      "learning_rate": 3.382716049382716e-05,
+      "loss": 0.4715,
+      "step": 1213
+    },
+    {
+      "epoch": 2.693288962839712,
+      "grad_norm": 0.35312673449516296,
+      "learning_rate": 3.3580246913580247e-05,
+      "loss": 0.5307,
+      "step": 1214
+    },
+    {
+      "epoch": 2.695507487520799,
+      "grad_norm": 0.32276803255081177,
+      "learning_rate": 3.333333333333333e-05,
+      "loss": 0.418,
+      "step": 1215
+    },
+    {
+      "epoch": 2.697726012201886,
+      "grad_norm": 0.3967615067958832,
+      "learning_rate": 3.3086419753086417e-05,
+      "loss": 0.5094,
+      "step": 1216
+    },
+    {
+      "epoch": 2.699944536882973,
+      "grad_norm": 0.37040090560913086,
+      "learning_rate": 3.2839506172839505e-05,
+      "loss": 0.4979,
+      "step": 1217
+    },
+    {
+      "epoch": 2.70216306156406,
+      "grad_norm": 0.32697996497154236,
+      "learning_rate": 3.2592592592592586e-05,
+      "loss": 0.4212,
+      "step": 1218
+    },
+    {
+      "epoch": 2.7043815862451472,
+      "grad_norm": 0.3230026364326477,
+      "learning_rate": 3.2345679012345675e-05,
+      "loss": 0.5291,
+      "step": 1219
+    },
+    {
+      "epoch": 2.7066001109262343,
+      "grad_norm": 0.26306265592575073,
+      "learning_rate": 3.209876543209876e-05,
+      "loss": 0.5633,
+      "step": 1220
+    },
+    {
+      "epoch": 2.7088186356073214,
+      "grad_norm": 0.2718304395675659,
+      "learning_rate": 3.185185185185185e-05,
+      "loss": 0.4396,
+      "step": 1221
+    },
+    {
+      "epoch": 2.7110371602884085,
+      "grad_norm": 0.30475306510925293,
+      "learning_rate": 3.160493827160493e-05,
+      "loss": 0.5442,
+      "step": 1222
+    },
+    {
+      "epoch": 2.7132556849694955,
+      "grad_norm": 0.2732059359550476,
+      "learning_rate": 3.135802469135802e-05,
+      "loss": 0.569,
+      "step": 1223
+    },
+    {
+      "epoch": 2.7154742096505826,
+      "grad_norm": 0.2485705465078354,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 0.6183,
+      "step": 1224
+    },
+    {
+      "epoch": 2.7176927343316697,
+      "grad_norm": 0.31775492429733276,
+      "learning_rate": 3.08641975308642e-05,
+      "loss": 0.509,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7199112590127568,
+      "grad_norm": 0.303199827671051,
+      "learning_rate": 3.061728395061728e-05,
+      "loss": 0.4271,
+      "step": 1226
+    },
+    {
+      "epoch": 2.722129783693844,
+      "grad_norm": 0.3213540017604828,
+      "learning_rate": 3.0370370370370368e-05,
+      "loss": 0.5339,
+      "step": 1227
+    },
+    {
+      "epoch": 2.724348308374931,
+      "grad_norm": 0.27744725346565247,
+      "learning_rate": 3.0123456790123453e-05,
+      "loss": 0.533,
+      "step": 1228
+    },
+    {
+      "epoch": 2.726566833056018,
+      "grad_norm": 0.3503943979740143,
+      "learning_rate": 2.987654320987654e-05,
+      "loss": 0.454,
+      "step": 1229
+    },
+    {
+      "epoch": 2.728785357737105,
+      "grad_norm": 0.27308955788612366,
+      "learning_rate": 2.9629629629629627e-05,
+      "loss": 0.4014,
+      "step": 1230
+    },
+    {
+      "epoch": 2.731003882418192,
+      "grad_norm": 0.2960096299648285,
+      "learning_rate": 2.9382716049382715e-05,
+      "loss": 0.523,
+      "step": 1231
+    },
+    {
+      "epoch": 2.733222407099279,
+      "grad_norm": 0.28652969002723694,
+      "learning_rate": 2.9135802469135796e-05,
+      "loss": 0.3125,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7354409317803663,
+      "grad_norm": 0.27901625633239746,
+      "learning_rate": 2.8888888888888885e-05,
+      "loss": 0.3945,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7376594564614534,
+      "grad_norm": 0.2694418132305145,
+      "learning_rate": 2.8641975308641973e-05,
+      "loss": 0.4413,
+      "step": 1234
+    },
+    {
+      "epoch": 2.7398779811425404,
+      "grad_norm": 0.24394243955612183,
+      "learning_rate": 2.8395061728395058e-05,
+      "loss": 0.381,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7420965058236275,
+      "grad_norm": 0.2828238904476166,
+      "learning_rate": 2.8148148148148147e-05,
+      "loss": 0.4562,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7443150305047146,
+      "grad_norm": 0.2651687264442444,
+      "learning_rate": 2.790123456790123e-05,
+      "loss": 0.5268,
+      "step": 1237
+    },
+    {
+      "epoch": 2.7465335551858017,
+      "grad_norm": 0.33506590127944946,
+      "learning_rate": 2.765432098765432e-05,
+      "loss": 0.4707,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7487520798668887,
+      "grad_norm": 0.3148481249809265,
+      "learning_rate": 2.7407407407407405e-05,
+      "loss": 0.4294,
+      "step": 1239
+    },
+    {
+      "epoch": 2.750970604547976,
+      "grad_norm": 0.29942113161087036,
+      "learning_rate": 2.7160493827160493e-05,
+      "loss": 0.595,
+      "step": 1240
+    },
+    {
+      "epoch": 2.753189129229063,
+      "grad_norm": 0.2789902985095978,
+      "learning_rate": 2.6913580246913578e-05,
+      "loss": 0.5041,
+      "step": 1241
+    },
+    {
+      "epoch": 2.75540765391015,
+      "grad_norm": 0.254192054271698,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.4343,
+      "step": 1242
+    },
+    {
+      "epoch": 2.757626178591237,
+      "grad_norm": 0.25477761030197144,
+      "learning_rate": 2.6419753086419748e-05,
+      "loss": 0.4493,
+      "step": 1243
+    },
+    {
+      "epoch": 2.759844703272324,
+      "grad_norm": 0.32381343841552734,
+      "learning_rate": 2.6172839506172837e-05,
+      "loss": 0.5906,
+      "step": 1244
+    },
+    {
+      "epoch": 2.762063227953411,
+      "grad_norm": 0.2790049612522125,
+      "learning_rate": 2.592592592592592e-05,
+      "loss": 0.5995,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7642817526344983,
+      "grad_norm": 0.28967416286468506,
+      "learning_rate": 2.567901234567901e-05,
+      "loss": 0.4483,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7665002773155853,
+      "grad_norm": 0.28956139087677,
+      "learning_rate": 2.5432098765432095e-05,
+      "loss": 0.5242,
+      "step": 1247
+    },
+    {
+      "epoch": 2.7687188019966724,
+      "grad_norm": 0.2880292534828186,
+      "learning_rate": 2.5185185185185183e-05,
+      "loss": 0.4017,
+      "step": 1248
+    },
+    {
+      "epoch": 2.7709373266777595,
+      "grad_norm": 0.3416897654533386,
+      "learning_rate": 2.4938271604938268e-05,
+      "loss": 0.6263,
+      "step": 1249
+    },
+    {
+      "epoch": 2.7731558513588466,
+      "grad_norm": 0.2564065456390381,
+      "learning_rate": 2.4691358024691357e-05,
+      "loss": 0.4989,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7753743760399336,
+      "grad_norm": 0.3221985697746277,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 0.5802,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7775929007210207,
+      "grad_norm": 0.25522080063819885,
+      "learning_rate": 2.419753086419753e-05,
+      "loss": 0.4842,
+      "step": 1252
+    },
+    {
+      "epoch": 2.779811425402108,
+      "grad_norm": 0.3274691700935364,
+      "learning_rate": 2.3950617283950618e-05,
+      "loss": 0.6838,
+      "step": 1253
+    },
+    {
+      "epoch": 2.782029950083195,
+      "grad_norm": 0.3028249442577362,
+      "learning_rate": 2.37037037037037e-05,
+      "loss": 0.3942,
+      "step": 1254
+    },
+    {
+      "epoch": 2.784248474764282,
+      "grad_norm": 0.25039997696876526,
+      "learning_rate": 2.3456790123456788e-05,
+      "loss": 0.4867,
+      "step": 1255
+    },
+    {
+      "epoch": 2.786466999445369,
+      "grad_norm": 0.3262811005115509,
+      "learning_rate": 2.3209876543209873e-05,
+      "loss": 0.5012,
+      "step": 1256
+    },
+    {
+      "epoch": 2.788685524126456,
+      "grad_norm": 0.367648184299469,
+      "learning_rate": 2.296296296296296e-05,
+      "loss": 0.5831,
+      "step": 1257
+    },
+    {
+      "epoch": 2.790904048807543,
+      "grad_norm": 0.27698493003845215,
+      "learning_rate": 2.2716049382716047e-05,
+      "loss": 0.4414,
+      "step": 1258
+    },
+    {
+      "epoch": 2.7931225734886302,
+      "grad_norm": 0.3209448456764221,
+      "learning_rate": 2.2469135802469135e-05,
+      "loss": 0.6594,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7953410981697173,
+      "grad_norm": 0.26347240805625916,
+      "learning_rate": 2.222222222222222e-05,
+      "loss": 0.5156,
+      "step": 1260
+    },
+    {
+      "epoch": 2.7975596228508044,
+      "grad_norm": 0.2997935116291046,
+      "learning_rate": 2.1975308641975308e-05,
+      "loss": 0.3789,
+      "step": 1261
+    },
+    {
+      "epoch": 2.7997781475318915,
+      "grad_norm": 0.37121424078941345,
+      "learning_rate": 2.1728395061728393e-05,
+      "loss": 0.5227,
+      "step": 1262
+    },
+    {
+      "epoch": 2.8019966722129785,
+      "grad_norm": 0.33274465799331665,
+      "learning_rate": 2.148148148148148e-05,
+      "loss": 0.3701,
+      "step": 1263
+    },
+    {
+      "epoch": 2.8042151968940656,
+      "grad_norm": 0.308185875415802,
+      "learning_rate": 2.1234567901234563e-05,
+      "loss": 0.4998,
+      "step": 1264
+    },
+    {
+      "epoch": 2.8064337215751527,
+      "grad_norm": 0.3436746597290039,
+      "learning_rate": 2.098765432098765e-05,
+      "loss": 0.6083,
+      "step": 1265
+    },
+    {
+      "epoch": 2.8086522462562398,
+      "grad_norm": 0.3519856035709381,
+      "learning_rate": 2.0740740740740737e-05,
+      "loss": 0.4131,
+      "step": 1266
+    },
+    {
+      "epoch": 2.810870770937327,
+      "grad_norm": 0.25393161177635193,
+      "learning_rate": 2.0493827160493825e-05,
+      "loss": 0.6249,
+      "step": 1267
+    },
+    {
+      "epoch": 2.813089295618414,
+      "grad_norm": 0.32563331723213196,
+      "learning_rate": 2.0246913580246913e-05,
+      "loss": 0.4273,
+      "step": 1268
+    },
+    {
+      "epoch": 2.815307820299501,
+      "grad_norm": 0.2718135118484497,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.4032,
+      "step": 1269
+    },
+    {
+      "epoch": 2.817526344980588,
+      "grad_norm": 0.2950599789619446,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 0.4207,
+      "step": 1270
+    },
+    {
+      "epoch": 2.819744869661675,
+      "grad_norm": 0.3212021291255951,
+      "learning_rate": 1.950617283950617e-05,
+      "loss": 0.4366,
+      "step": 1271
+    },
+    {
+      "epoch": 2.821963394342762,
+      "grad_norm": 0.3315774202346802,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.4767,
+      "step": 1272
+    },
+    {
+      "epoch": 2.8241819190238493,
+      "grad_norm": 0.2837589681148529,
+      "learning_rate": 1.901234567901234e-05,
+      "loss": 0.4703,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8264004437049364,
+      "grad_norm": 0.29627227783203125,
+      "learning_rate": 1.8765432098765433e-05,
+      "loss": 0.4745,
+      "step": 1274
+    },
+    {
+      "epoch": 2.8286189683860234,
+      "grad_norm": 0.3321774899959564,
+      "learning_rate": 1.8518518518518515e-05,
+      "loss": 0.5968,
+      "step": 1275
+    },
+    {
+      "epoch": 2.8308374930671105,
+      "grad_norm": 0.37471216917037964,
+      "learning_rate": 1.8271604938271603e-05,
+      "loss": 0.4182,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8330560177481976,
+      "grad_norm": 0.33006352186203003,
+      "learning_rate": 1.802469135802469e-05,
+      "loss": 0.5731,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8352745424292847,
+      "grad_norm": 0.2877190113067627,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.3881,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8374930671103717,
+      "grad_norm": 0.2881687879562378,
+      "learning_rate": 1.753086419753086e-05,
+      "loss": 0.4283,
+      "step": 1279
+    },
+    {
+      "epoch": 2.839711591791459,
+      "grad_norm": 0.2627007067203522,
+      "learning_rate": 1.728395061728395e-05,
+      "loss": 0.4638,
+      "step": 1280
+    },
+    {
+      "epoch": 2.841930116472546,
+      "grad_norm": 0.30371254682540894,
+      "learning_rate": 1.7037037037037035e-05,
+      "loss": 0.4026,
+      "step": 1281
+    },
+    {
+      "epoch": 2.844148641153633,
+      "grad_norm": 0.2697737514972687,
+      "learning_rate": 1.6790123456790123e-05,
+      "loss": 0.613,
+      "step": 1282
+    },
+    {
+      "epoch": 2.84636716583472,
+      "grad_norm": 0.2805211544036865,
+      "learning_rate": 1.6543209876543208e-05,
+      "loss": 0.4358,
+      "step": 1283
+    },
+    {
+      "epoch": 2.848585690515807,
+      "grad_norm": 0.2741427421569824,
+      "learning_rate": 1.6296296296296293e-05,
+      "loss": 0.3824,
+      "step": 1284
+    },
+    {
+      "epoch": 2.850804215196894,
+      "grad_norm": 0.32560959458351135,
+      "learning_rate": 1.604938271604938e-05,
+      "loss": 0.4875,
+      "step": 1285
+    },
+    {
+      "epoch": 2.8530227398779813,
+      "grad_norm": 0.25359609723091125,
+      "learning_rate": 1.5802469135802467e-05,
+      "loss": 0.3994,
+      "step": 1286
+    },
+    {
+      "epoch": 2.8552412645590683,
+      "grad_norm": 0.25484853982925415,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.2825,
+      "step": 1287
+    },
+    {
+      "epoch": 2.8574597892401554,
+      "grad_norm": 0.26220130920410156,
+      "learning_rate": 1.530864197530864e-05,
+      "loss": 0.4516,
+      "step": 1288
+    },
+    {
+      "epoch": 2.8596783139212425,
+      "grad_norm": 0.44897976517677307,
+      "learning_rate": 1.5061728395061727e-05,
+      "loss": 0.4281,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8618968386023296,
+      "grad_norm": 0.2895084023475647,
+      "learning_rate": 1.4814814814814813e-05,
+      "loss": 0.3966,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8641153632834166,
+      "grad_norm": 0.2595643401145935,
+      "learning_rate": 1.4567901234567898e-05,
+      "loss": 0.4628,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8663338879645037,
+      "grad_norm": 0.31123635172843933,
+      "learning_rate": 1.4320987654320987e-05,
+      "loss": 0.3891,
+      "step": 1292
+    },
+    {
+      "epoch": 2.868552412645591,
+      "grad_norm": 0.3240855634212494,
+      "learning_rate": 1.4074074074074073e-05,
+      "loss": 0.6483,
+      "step": 1293
+    },
+    {
+      "epoch": 2.870770937326678,
+      "grad_norm": 0.27346086502075195,
+      "learning_rate": 1.382716049382716e-05,
+      "loss": 0.5581,
+      "step": 1294
+    },
+    {
+      "epoch": 2.872989462007765,
+      "grad_norm": 0.3081636428833008,
+      "learning_rate": 1.3580246913580247e-05,
+      "loss": 0.5294,
+      "step": 1295
+    },
+    {
+      "epoch": 2.875207986688852,
+      "grad_norm": 0.28265485167503357,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.4739,
+      "step": 1296
+    },
+    {
+      "epoch": 2.877426511369939,
+      "grad_norm": 0.28777042031288147,
+      "learning_rate": 1.3086419753086418e-05,
+      "loss": 0.4424,
+      "step": 1297
+    },
+    {
+      "epoch": 2.879645036051026,
+      "grad_norm": 0.3454686999320984,
+      "learning_rate": 1.2839506172839505e-05,
+      "loss": 0.5693,
+      "step": 1298
+    },
+    {
+      "epoch": 2.8818635607321132,
+      "grad_norm": 0.31595227122306824,
+      "learning_rate": 1.2592592592592592e-05,
+      "loss": 0.6419,
+      "step": 1299
+    },
+    {
+      "epoch": 2.8840820854132003,
+      "grad_norm": 0.28077563643455505,
+      "learning_rate": 1.2345679012345678e-05,
+      "loss": 0.4338,
+      "step": 1300
+    },
+    {
+      "epoch": 2.8863006100942874,
+      "grad_norm": 0.37803104519844055,
+      "learning_rate": 1.2098765432098765e-05,
+      "loss": 0.4679,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8885191347753745,
+      "grad_norm": 0.3873048722743988,
+      "learning_rate": 1.185185185185185e-05,
+      "loss": 0.3848,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8907376594564616,
+      "grad_norm": 0.2962777614593506,
+      "learning_rate": 1.1604938271604937e-05,
+      "loss": 0.4658,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8929561841375486,
+      "grad_norm": 0.32748886942863464,
+      "learning_rate": 1.1358024691358023e-05,
+      "loss": 0.4866,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8951747088186357,
+      "grad_norm": 0.33596014976501465,
+      "learning_rate": 1.111111111111111e-05,
+      "loss": 0.4003,
+      "step": 1305
+    },
+    {
+      "epoch": 2.8973932334997228,
+      "grad_norm": 0.3027206063270569,
+      "learning_rate": 1.0864197530864197e-05,
+      "loss": 0.4864,
+      "step": 1306
+    },
+    {
+      "epoch": 2.89961175818081,
+      "grad_norm": 0.286678284406662,
+      "learning_rate": 1.0617283950617282e-05,
+      "loss": 0.4611,
+      "step": 1307
+    },
+    {
+      "epoch": 2.901830282861897,
+      "grad_norm": 0.2828173339366913,
+      "learning_rate": 1.0370370370370368e-05,
+      "loss": 0.5277,
+      "step": 1308
+    },
+    {
+      "epoch": 2.904048807542984,
+      "grad_norm": 0.29805052280426025,
+      "learning_rate": 1.0123456790123457e-05,
+      "loss": 0.4285,
+      "step": 1309
+    },
+    {
+      "epoch": 2.906267332224071,
+      "grad_norm": 0.22442780435085297,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 0.2856,
+      "step": 1310
+    },
+    {
+      "epoch": 2.908485856905158,
+      "grad_norm": 0.2582552433013916,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.5215,
+      "step": 1311
+    },
+    {
+      "epoch": 2.9107043815862452,
+      "grad_norm": 0.25813403725624084,
+      "learning_rate": 9.382716049382717e-06,
+      "loss": 0.486,
+      "step": 1312
+    },
+    {
+      "epoch": 2.9129229062673323,
+      "grad_norm": 0.31789693236351013,
+      "learning_rate": 9.135802469135802e-06,
+      "loss": 0.328,
+      "step": 1313
+    },
+    {
+      "epoch": 2.9151414309484194,
+      "grad_norm": 0.24600911140441895,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.4782,
+      "step": 1314
+    },
+    {
+      "epoch": 2.9173599556295065,
+      "grad_norm": 0.2871638536453247,
+      "learning_rate": 8.641975308641975e-06,
+      "loss": 0.4636,
+      "step": 1315
+    },
+    {
+      "epoch": 2.9195784803105935,
+      "grad_norm": 0.2895622253417969,
+      "learning_rate": 8.395061728395062e-06,
+      "loss": 0.7178,
+      "step": 1316
+    },
+    {
+      "epoch": 2.9217970049916806,
+      "grad_norm": 0.2977471947669983,
+      "learning_rate": 8.148148148148147e-06,
+      "loss": 0.5255,
+      "step": 1317
+    },
+    {
+      "epoch": 2.9240155296727677,
+      "grad_norm": 0.29984888434410095,
+      "learning_rate": 7.901234567901233e-06,
+      "loss": 0.516,
+      "step": 1318
+    },
+    {
+      "epoch": 2.9262340543538548,
+      "grad_norm": 0.2472100704908371,
+      "learning_rate": 7.65432098765432e-06,
+      "loss": 0.4578,
+      "step": 1319
+    },
+    {
+      "epoch": 2.928452579034942,
+      "grad_norm": 0.29188069701194763,
+      "learning_rate": 7.407407407407407e-06,
+      "loss": 0.5206,
+      "step": 1320
+    },
+    {
+      "epoch": 2.930671103716029,
+      "grad_norm": 0.28032782673835754,
+      "learning_rate": 7.160493827160493e-06,
+      "loss": 0.6177,
+      "step": 1321
+    },
+    {
+      "epoch": 2.932889628397116,
+      "grad_norm": 0.25077563524246216,
+      "learning_rate": 6.91358024691358e-06,
+      "loss": 0.4534,
+      "step": 1322
+    },
+    {
+      "epoch": 2.935108153078203,
+      "grad_norm": 0.33414068818092346,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.4343,
+      "step": 1323
+    },
+    {
+      "epoch": 2.93732667775929,
+      "grad_norm": 0.28644508123397827,
+      "learning_rate": 6.4197530864197525e-06,
+      "loss": 0.3835,
+      "step": 1324
+    },
+    {
+      "epoch": 2.939545202440377,
+      "grad_norm": 0.3566669821739197,
+      "learning_rate": 6.172839506172839e-06,
+      "loss": 0.5482,
+      "step": 1325
+    },
+    {
+      "epoch": 2.9417637271214643,
+      "grad_norm": 0.30680161714553833,
+      "learning_rate": 5.925925925925925e-06,
+      "loss": 0.4796,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9439822518025514,
+      "grad_norm": 0.3375394344329834,
+      "learning_rate": 5.679012345679012e-06,
+      "loss": 0.5337,
+      "step": 1327
+    },
+    {
+      "epoch": 2.9462007764836384,
+      "grad_norm": 0.2626802623271942,
+      "learning_rate": 5.432098765432098e-06,
+      "loss": 0.6186,
+      "step": 1328
+    },
+    {
+      "epoch": 2.9484193011647255,
+      "grad_norm": 0.2848595976829529,
+      "learning_rate": 5.185185185185184e-06,
+      "loss": 0.3795,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9506378258458126,
+      "grad_norm": 0.28433355689048767,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.5122,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9528563505268997,
+      "grad_norm": 0.3099772036075592,
+      "learning_rate": 4.691358024691358e-06,
+      "loss": 0.5397,
+      "step": 1331
+    },
+    {
+      "epoch": 2.9550748752079867,
+      "grad_norm": 0.2349434196949005,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.522,
+      "step": 1332
+    },
+    {
+      "epoch": 2.957293399889074,
+      "grad_norm": 0.3928101658821106,
+      "learning_rate": 4.197530864197531e-06,
+      "loss": 0.4151,
+      "step": 1333
+    },
+    {
+      "epoch": 2.959511924570161,
+      "grad_norm": 0.3285221755504608,
+      "learning_rate": 3.950617283950617e-06,
+      "loss": 0.4894,
+      "step": 1334
+    },
+    {
+      "epoch": 2.961730449251248,
+      "grad_norm": 0.2856891453266144,
+      "learning_rate": 3.7037037037037033e-06,
+      "loss": 0.5939,
+      "step": 1335
+    },
+    {
+      "epoch": 2.963948973932335,
+      "grad_norm": 0.33456167578697205,
+      "learning_rate": 3.45679012345679e-06,
+      "loss": 0.5935,
+      "step": 1336
+    },
+    {
+      "epoch": 2.966167498613422,
+      "grad_norm": 0.2973114252090454,
+      "learning_rate": 3.2098765432098762e-06,
+      "loss": 0.4376,
+      "step": 1337
+    },
+    {
+      "epoch": 2.968386023294509,
+      "grad_norm": 0.2715815007686615,
+      "learning_rate": 2.9629629629629625e-06,
+      "loss": 0.451,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9706045479755963,
+      "grad_norm": 0.26056328415870667,
+      "learning_rate": 2.716049382716049e-06,
+      "loss": 0.3796,
+      "step": 1339
+    },
+    {
+      "epoch": 2.9728230726566833,
+      "grad_norm": 0.4245537221431732,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 0.5778,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9750415973377704,
+      "grad_norm": 0.28082960844039917,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.5011,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9772601220188575,
+      "grad_norm": 0.40667977929115295,
+      "learning_rate": 1.9753086419753083e-06,
+      "loss": 0.553,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9794786466999446,
+      "grad_norm": 0.28392481803894043,
+      "learning_rate": 1.728395061728395e-06,
+      "loss": 0.463,
+      "step": 1343
+    },
+    {
+      "epoch": 2.9816971713810316,
+      "grad_norm": 0.36204710602760315,
+      "learning_rate": 1.4814814814814812e-06,
+      "loss": 0.4518,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9839156960621187,
+      "grad_norm": 0.22228075563907623,
+      "learning_rate": 1.234567901234568e-06,
+      "loss": 0.6984,
+      "step": 1345
+    },
+    {
+      "epoch": 2.986134220743206,
+      "grad_norm": 0.3683073818683624,
+      "learning_rate": 9.876543209876542e-07,
+      "loss": 0.6846,
+      "step": 1346
+    },
+    {
+      "epoch": 2.988352745424293,
+      "grad_norm": 0.22791865468025208,
+      "learning_rate": 7.407407407407406e-07,
+      "loss": 0.4088,
+      "step": 1347
+    },
+    {
+      "epoch": 2.99057127010538,
+      "grad_norm": 0.2904137372970581,
+      "learning_rate": 4.938271604938271e-07,
+      "loss": 0.4281,
+      "step": 1348
+    },
+    {
+      "epoch": 2.992789794786467,
+      "grad_norm": 0.3128870129585266,
+      "learning_rate": 2.4691358024691354e-07,
+      "loss": 0.4765,
+      "step": 1349
+    },
+    {
+      "epoch": 2.995008319467554,
+      "grad_norm": 0.25493577122688293,
+      "learning_rate": 0.0,
+      "loss": 0.4009,
+      "step": 1350
+    },
+    {
+      "epoch": 2.995008319467554,
+      "step": 1350,
+      "total_flos": 5.088784523806638e+17,
+      "train_loss": 0.5411236907817699,
+      "train_runtime": 4356.0395,
+      "train_samples_per_second": 4.966,
+      "train_steps_per_second": 0.31
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1350,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.088784523806638e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}