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": 0.2167196273803711,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.8259,
+      "step": 1
+    },
+    {
+      "epoch": 0.004437049362174155,
+      "grad_norm": 0.380119651556015,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.7536,
+      "step": 2
+    },
+    {
+      "epoch": 0.0066555740432612314,
+      "grad_norm": 0.28782230615615845,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.8631,
+      "step": 3
+    },
+    {
+      "epoch": 0.00887409872434831,
+      "grad_norm": 0.23643195629119873,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.7027,
+      "step": 4
+    },
+    {
+      "epoch": 0.011092623405435386,
+      "grad_norm": 0.177652508020401,
+      "learning_rate": 1.111111111111111e-05,
+      "loss": 0.7434,
+      "step": 5
+    },
+    {
+      "epoch": 0.013311148086522463,
+      "grad_norm": 0.2520431876182556,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7035,
+      "step": 6
+    },
+    {
+      "epoch": 0.01552967276760954,
+      "grad_norm": 0.27725693583488464,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.7387,
+      "step": 7
+    },
+    {
+      "epoch": 0.01774819744869662,
+      "grad_norm": 0.21377894282341003,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.9072,
+      "step": 8
+    },
+    {
+      "epoch": 0.019966722129783693,
+      "grad_norm": 0.5910317897796631,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.7977,
+      "step": 9
+    },
+    {
+      "epoch": 0.022185246810870772,
+      "grad_norm": 0.1639687865972519,
+      "learning_rate": 2.222222222222222e-05,
+      "loss": 0.9154,
+      "step": 10
+    },
+    {
+      "epoch": 0.024403771491957847,
+      "grad_norm": 0.5810109972953796,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 1.1596,
+      "step": 11
+    },
+    {
+      "epoch": 0.026622296173044926,
+      "grad_norm": 0.22166898846626282,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6996,
+      "step": 12
+    },
+    {
+      "epoch": 0.028840820854132,
+      "grad_norm": 0.25997990369796753,
+      "learning_rate": 2.8888888888888885e-05,
+      "loss": 0.863,
+      "step": 13
+    },
+    {
+      "epoch": 0.03105934553521908,
+      "grad_norm": 0.19173578917980194,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 1.0604,
+      "step": 14
+    },
+    {
+      "epoch": 0.033277870216306155,
+      "grad_norm": 0.3815133273601532,
+      "learning_rate": 3.333333333333333e-05,
+      "loss": 0.8523,
+      "step": 15
+    },
+    {
+      "epoch": 0.03549639489739324,
+      "grad_norm": 0.2598717212677002,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 0.88,
+      "step": 16
+    },
+    {
+      "epoch": 0.03771491957848031,
+      "grad_norm": 0.28661081194877625,
+      "learning_rate": 3.777777777777777e-05,
+      "loss": 0.8146,
+      "step": 17
+    },
+    {
+      "epoch": 0.03993344425956739,
+      "grad_norm": 0.22358466684818268,
+      "learning_rate": 3.9999999999999996e-05,
+      "loss": 0.7388,
+      "step": 18
+    },
+    {
+      "epoch": 0.04215196894065446,
+      "grad_norm": 0.19268451631069183,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.7281,
+      "step": 19
+    },
+    {
+      "epoch": 0.044370493621741544,
+      "grad_norm": 0.27424678206443787,
+      "learning_rate": 4.444444444444444e-05,
+      "loss": 0.7824,
+      "step": 20
+    },
+    {
+      "epoch": 0.04658901830282862,
+      "grad_norm": 0.252247154712677,
+      "learning_rate": 4.6666666666666665e-05,
+      "loss": 0.8764,
+      "step": 21
+    },
+    {
+      "epoch": 0.048807542983915694,
+      "grad_norm": 0.20753055810928345,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.8175,
+      "step": 22
+    },
+    {
+      "epoch": 0.051026067665002776,
+      "grad_norm": 0.26856139302253723,
+      "learning_rate": 5.111111111111111e-05,
+      "loss": 0.9426,
+      "step": 23
+    },
+    {
+      "epoch": 0.05324459234608985,
+      "grad_norm": 0.1751042753458023,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.7042,
+      "step": 24
+    },
+    {
+      "epoch": 0.05546311702717693,
+      "grad_norm": 0.3058775067329407,
+      "learning_rate": 5.5555555555555545e-05,
+      "loss": 0.9191,
+      "step": 25
+    },
+    {
+      "epoch": 0.057681641708264,
+      "grad_norm": 0.32840612530708313,
+      "learning_rate": 5.777777777777777e-05,
+      "loss": 0.7746,
+      "step": 26
+    },
+    {
+      "epoch": 0.059900166389351084,
+      "grad_norm": 0.23080751299858093,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 0.9714,
+      "step": 27
+    },
+    {
+      "epoch": 0.06211869107043816,
+      "grad_norm": 0.25512954592704773,
+      "learning_rate": 6.222222222222222e-05,
+      "loss": 0.7108,
+      "step": 28
+    },
+    {
+      "epoch": 0.06433721575152523,
+      "grad_norm": 0.24804095923900604,
+      "learning_rate": 6.444444444444444e-05,
+      "loss": 0.8479,
+      "step": 29
+    },
+    {
+      "epoch": 0.06655574043261231,
+      "grad_norm": 0.25404223799705505,
+      "learning_rate": 6.666666666666666e-05,
+      "loss": 0.7454,
+      "step": 30
+    },
+    {
+      "epoch": 0.06877426511369938,
+      "grad_norm": 0.21752962470054626,
+      "learning_rate": 6.888888888888888e-05,
+      "loss": 0.7813,
+      "step": 31
+    },
+    {
+      "epoch": 0.07099278979478647,
+      "grad_norm": 0.261349081993103,
+      "learning_rate": 7.11111111111111e-05,
+      "loss": 0.9202,
+      "step": 32
+    },
+    {
+      "epoch": 0.07321131447587355,
+      "grad_norm": 0.28289204835891724,
+      "learning_rate": 7.333333333333332e-05,
+      "loss": 0.8155,
+      "step": 33
+    },
+    {
+      "epoch": 0.07542983915696062,
+      "grad_norm": 0.3035896122455597,
+      "learning_rate": 7.555555555555554e-05,
+      "loss": 0.8549,
+      "step": 34
+    },
+    {
+      "epoch": 0.0776483638380477,
+      "grad_norm": 0.32377755641937256,
+      "learning_rate": 7.777777777777777e-05,
+      "loss": 0.709,
+      "step": 35
+    },
+    {
+      "epoch": 0.07986688851913477,
+      "grad_norm": 0.272444486618042,
+      "learning_rate": 7.999999999999999e-05,
+      "loss": 0.7805,
+      "step": 36
+    },
+    {
+      "epoch": 0.08208541320022185,
+      "grad_norm": 0.275277704000473,
+      "learning_rate": 8.222222222222222e-05,
+      "loss": 0.9101,
+      "step": 37
+    },
+    {
+      "epoch": 0.08430393788130892,
+      "grad_norm": 0.32304221391677856,
+      "learning_rate": 8.444444444444444e-05,
+      "loss": 0.8556,
+      "step": 38
+    },
+    {
+      "epoch": 0.08652246256239601,
+      "grad_norm": 0.9152739644050598,
+      "learning_rate": 8.666666666666665e-05,
+      "loss": 0.7329,
+      "step": 39
+    },
+    {
+      "epoch": 0.08874098724348309,
+      "grad_norm": 0.29951849579811096,
+      "learning_rate": 8.888888888888888e-05,
+      "loss": 0.8463,
+      "step": 40
+    },
+    {
+      "epoch": 0.09095951192457016,
+      "grad_norm": 0.2968096137046814,
+      "learning_rate": 9.11111111111111e-05,
+      "loss": 0.8174,
+      "step": 41
+    },
+    {
+      "epoch": 0.09317803660565724,
+      "grad_norm": 0.32746949791908264,
+      "learning_rate": 9.333333333333333e-05,
+      "loss": 0.7557,
+      "step": 42
+    },
+    {
+      "epoch": 0.09539656128674431,
+      "grad_norm": 0.32505884766578674,
+      "learning_rate": 9.555555555555555e-05,
+      "loss": 0.8296,
+      "step": 43
+    },
+    {
+      "epoch": 0.09761508596783139,
+      "grad_norm": 0.29400014877319336,
+      "learning_rate": 9.777777777777778e-05,
+      "loss": 0.8364,
+      "step": 44
+    },
+    {
+      "epoch": 0.09983361064891846,
+      "grad_norm": 0.5968295931816101,
+      "learning_rate": 9.999999999999999e-05,
+      "loss": 0.8631,
+      "step": 45
+    },
+    {
+      "epoch": 0.10205213533000555,
+      "grad_norm": 0.3700101375579834,
+      "learning_rate": 0.00010222222222222222,
+      "loss": 0.8306,
+      "step": 46
+    },
+    {
+      "epoch": 0.10427066001109263,
+      "grad_norm": 0.3246069550514221,
+      "learning_rate": 0.00010444444444444444,
+      "loss": 0.8499,
+      "step": 47
+    },
+    {
+      "epoch": 0.1064891846921797,
+      "grad_norm": 0.43050330877304077,
+      "learning_rate": 0.00010666666666666667,
+      "loss": 0.7187,
+      "step": 48
+    },
+    {
+      "epoch": 0.10870770937326678,
+      "grad_norm": 0.3699756860733032,
+      "learning_rate": 0.00010888888888888889,
+      "loss": 0.9064,
+      "step": 49
+    },
+    {
+      "epoch": 0.11092623405435385,
+      "grad_norm": 0.4039502739906311,
+      "learning_rate": 0.00011111111111111109,
+      "loss": 0.8239,
+      "step": 50
+    },
+    {
+      "epoch": 0.11314475873544093,
+      "grad_norm": 0.37197792530059814,
+      "learning_rate": 0.00011333333333333331,
+      "loss": 0.8107,
+      "step": 51
+    },
+    {
+      "epoch": 0.115363283416528,
+      "grad_norm": 0.3142877519130707,
+      "learning_rate": 0.00011555555555555554,
+      "loss": 0.8532,
+      "step": 52
+    },
+    {
+      "epoch": 0.11758180809761509,
+      "grad_norm": 0.40597307682037354,
+      "learning_rate": 0.00011777777777777776,
+      "loss": 0.928,
+      "step": 53
+    },
+    {
+      "epoch": 0.11980033277870217,
+      "grad_norm": 0.40041929483413696,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 0.9558,
+      "step": 54
+    },
+    {
+      "epoch": 0.12201885745978924,
+      "grad_norm": 0.29422467947006226,
+      "learning_rate": 0.00012222222222222221,
+      "loss": 0.6271,
+      "step": 55
+    },
+    {
+      "epoch": 0.12423738214087632,
+      "grad_norm": 0.4203299582004547,
+      "learning_rate": 0.00012444444444444444,
+      "loss": 0.725,
+      "step": 56
+    },
+    {
+      "epoch": 0.1264559068219634,
+      "grad_norm": 0.40294623374938965,
+      "learning_rate": 0.00012666666666666666,
+      "loss": 0.7086,
+      "step": 57
+    },
+    {
+      "epoch": 0.12867443150305047,
+      "grad_norm": 0.49792084097862244,
+      "learning_rate": 0.0001288888888888889,
+      "loss": 0.8292,
+      "step": 58
+    },
+    {
+      "epoch": 0.13089295618413754,
+      "grad_norm": 0.4290768504142761,
+      "learning_rate": 0.00013111111111111111,
+      "loss": 0.7413,
+      "step": 59
+    },
+    {
+      "epoch": 0.13311148086522462,
+      "grad_norm": 0.3451862633228302,
+      "learning_rate": 0.0001333333333333333,
+      "loss": 0.5849,
+      "step": 60
+    },
+    {
+      "epoch": 0.1353300055463117,
+      "grad_norm": 0.3402823805809021,
+      "learning_rate": 0.00013555555555555554,
+      "loss": 0.8508,
+      "step": 61
+    },
+    {
+      "epoch": 0.13754853022739877,
+      "grad_norm": 0.3545222282409668,
+      "learning_rate": 0.00013777777777777776,
+      "loss": 0.7668,
+      "step": 62
+    },
+    {
+      "epoch": 0.13976705490848584,
+      "grad_norm": 0.3091377913951874,
+      "learning_rate": 0.00014,
+      "loss": 0.7597,
+      "step": 63
+    },
+    {
+      "epoch": 0.14198557958957295,
+      "grad_norm": 0.3612399101257324,
+      "learning_rate": 0.0001422222222222222,
+      "loss": 0.7385,
+      "step": 64
+    },
+    {
+      "epoch": 0.14420410427066002,
+      "grad_norm": 0.3823396861553192,
+      "learning_rate": 0.0001444444444444444,
+      "loss": 0.8345,
+      "step": 65
+    },
+    {
+      "epoch": 0.1464226289517471,
+      "grad_norm": 0.2895652949810028,
+      "learning_rate": 0.00014666666666666664,
+      "loss": 0.7398,
+      "step": 66
+    },
+    {
+      "epoch": 0.14864115363283417,
+      "grad_norm": 0.3475222885608673,
+      "learning_rate": 0.00014888888888888886,
+      "loss": 0.8014,
+      "step": 67
+    },
+    {
+      "epoch": 0.15085967831392125,
+      "grad_norm": 0.2549246847629547,
+      "learning_rate": 0.00015111111111111109,
+      "loss": 0.5429,
+      "step": 68
+    },
+    {
+      "epoch": 0.15307820299500832,
+      "grad_norm": 0.3540898561477661,
+      "learning_rate": 0.0001533333333333333,
+      "loss": 0.7581,
+      "step": 69
+    },
+    {
+      "epoch": 0.1552967276760954,
+      "grad_norm": 0.4790879487991333,
+      "learning_rate": 0.00015555555555555554,
+      "loss": 0.8237,
+      "step": 70
+    },
+    {
+      "epoch": 0.15751525235718247,
+      "grad_norm": 0.307131826877594,
+      "learning_rate": 0.00015777777777777776,
+      "loss": 0.6967,
+      "step": 71
+    },
+    {
+      "epoch": 0.15973377703826955,
+      "grad_norm": 0.40200021862983704,
+      "learning_rate": 0.00015999999999999999,
+      "loss": 0.7343,
+      "step": 72
+    },
+    {
+      "epoch": 0.16195230171935662,
+      "grad_norm": 0.3616339862346649,
+      "learning_rate": 0.0001622222222222222,
+      "loss": 0.7051,
+      "step": 73
+    },
+    {
+      "epoch": 0.1641708264004437,
+      "grad_norm": 0.28662922978401184,
+      "learning_rate": 0.00016444444444444444,
+      "loss": 0.7334,
+      "step": 74
+    },
+    {
+      "epoch": 0.16638935108153077,
+      "grad_norm": 0.5068575739860535,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 0.7403,
+      "step": 75
+    },
+    {
+      "epoch": 0.16860787576261785,
+      "grad_norm": 0.3047877252101898,
+      "learning_rate": 0.00016888888888888889,
+      "loss": 1.0307,
+      "step": 76
+    },
+    {
+      "epoch": 0.17082640044370492,
+      "grad_norm": 0.35696130990982056,
+      "learning_rate": 0.0001711111111111111,
+      "loss": 0.8209,
+      "step": 77
+    },
+    {
+      "epoch": 0.17304492512479203,
+      "grad_norm": 0.23237773776054382,
+      "learning_rate": 0.0001733333333333333,
+      "loss": 0.6893,
+      "step": 78
+    },
+    {
+      "epoch": 0.1752634498058791,
+      "grad_norm": 0.33648040890693665,
+      "learning_rate": 0.00017555555555555553,
+      "loss": 0.8784,
+      "step": 79
+    },
+    {
+      "epoch": 0.17748197448696618,
+      "grad_norm": 0.3384227752685547,
+      "learning_rate": 0.00017777777777777776,
+      "loss": 0.7823,
+      "step": 80
+    },
+    {
+      "epoch": 0.17970049916805325,
+      "grad_norm": 0.36117425560951233,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.8377,
+      "step": 81
+    },
+    {
+      "epoch": 0.18191902384914033,
+      "grad_norm": 0.24751441180706024,
+      "learning_rate": 0.0001822222222222222,
+      "loss": 0.7002,
+      "step": 82
+    },
+    {
+      "epoch": 0.1841375485302274,
+      "grad_norm": 0.43273913860321045,
+      "learning_rate": 0.00018444444444444443,
+      "loss": 0.8799,
+      "step": 83
+    },
+    {
+      "epoch": 0.18635607321131448,
+      "grad_norm": 0.26664185523986816,
+      "learning_rate": 0.00018666666666666666,
+      "loss": 0.6694,
+      "step": 84
+    },
+    {
+      "epoch": 0.18857459789240155,
+      "grad_norm": 0.7373461127281189,
+      "learning_rate": 0.00018888888888888888,
+      "loss": 0.7527,
+      "step": 85
+    },
+    {
+      "epoch": 0.19079312257348863,
+      "grad_norm": 0.31846243143081665,
+      "learning_rate": 0.0001911111111111111,
+      "loss": 0.6985,
+      "step": 86
+    },
+    {
+      "epoch": 0.1930116472545757,
+      "grad_norm": 0.2627131938934326,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 0.7603,
+      "step": 87
+    },
+    {
+      "epoch": 0.19523017193566278,
+      "grad_norm": 0.36998966336250305,
+      "learning_rate": 0.00019555555555555556,
+      "loss": 0.7769,
+      "step": 88
+    },
+    {
+      "epoch": 0.19744869661674985,
+      "grad_norm": 0.480451375246048,
+      "learning_rate": 0.00019777777777777776,
+      "loss": 0.8031,
+      "step": 89
+    },
+    {
+      "epoch": 0.19966722129783693,
+      "grad_norm": 0.3571526110172272,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 0.8562,
+      "step": 90
+    },
+    {
+      "epoch": 0.201885745978924,
+      "grad_norm": 0.42584526538848877,
+      "learning_rate": 0.0002022222222222222,
+      "loss": 1.0129,
+      "step": 91
+    },
+    {
+      "epoch": 0.2041042706600111,
+      "grad_norm": 0.30249080061912537,
+      "learning_rate": 0.00020444444444444443,
+      "loss": 0.8191,
+      "step": 92
+    },
+    {
+      "epoch": 0.20632279534109818,
+      "grad_norm": 0.2973940372467041,
+      "learning_rate": 0.00020666666666666666,
+      "loss": 0.8643,
+      "step": 93
+    },
+    {
+      "epoch": 0.20854132002218526,
+      "grad_norm": 0.29512155055999756,
+      "learning_rate": 0.00020888888888888888,
+      "loss": 0.9127,
+      "step": 94
+    },
+    {
+      "epoch": 0.21075984470327233,
+      "grad_norm": 0.3898771405220032,
+      "learning_rate": 0.0002111111111111111,
+      "loss": 0.8398,
+      "step": 95
+    },
+    {
+      "epoch": 0.2129783693843594,
+      "grad_norm": 0.38701146841049194,
+      "learning_rate": 0.00021333333333333333,
+      "loss": 0.972,
+      "step": 96
+    },
+    {
+      "epoch": 0.21519689406544648,
+      "grad_norm": 0.3297521770000458,
+      "learning_rate": 0.00021555555555555556,
+      "loss": 0.7079,
+      "step": 97
+    },
+    {
+      "epoch": 0.21741541874653356,
+      "grad_norm": 0.2672453820705414,
+      "learning_rate": 0.00021777777777777778,
+      "loss": 0.7283,
+      "step": 98
+    },
+    {
+      "epoch": 0.21963394342762063,
+      "grad_norm": 0.2872799336910248,
+      "learning_rate": 0.00021999999999999995,
+      "loss": 0.8921,
+      "step": 99
+    },
+    {
+      "epoch": 0.2218524681087077,
+      "grad_norm": 0.31917476654052734,
+      "learning_rate": 0.00022222222222222218,
+      "loss": 0.8869,
+      "step": 100
+    },
+    {
+      "epoch": 0.22407099278979478,
+      "grad_norm": 0.3244423568248749,
+      "learning_rate": 0.0002244444444444444,
+      "loss": 0.6388,
+      "step": 101
+    },
+    {
+      "epoch": 0.22628951747088186,
+      "grad_norm": 0.2692210078239441,
+      "learning_rate": 0.00022666666666666663,
+      "loss": 0.7595,
+      "step": 102
+    },
+    {
+      "epoch": 0.22850804215196893,
+      "grad_norm": 0.34151414036750793,
+      "learning_rate": 0.00022888888888888885,
+      "loss": 0.7737,
+      "step": 103
+    },
+    {
+      "epoch": 0.230726566833056,
+      "grad_norm": 0.4966946244239807,
+      "learning_rate": 0.00023111111111111108,
+      "loss": 0.7525,
+      "step": 104
+    },
+    {
+      "epoch": 0.23294509151414308,
+      "grad_norm": 0.26454317569732666,
+      "learning_rate": 0.0002333333333333333,
+      "loss": 0.7358,
+      "step": 105
+    },
+    {
+      "epoch": 0.23516361619523019,
+      "grad_norm": 0.28132063150405884,
+      "learning_rate": 0.00023555555555555553,
+      "loss": 0.6623,
+      "step": 106
+    },
+    {
+      "epoch": 0.23738214087631726,
+      "grad_norm": 0.29847851395606995,
+      "learning_rate": 0.00023777777777777775,
+      "loss": 0.7148,
+      "step": 107
+    },
+    {
+      "epoch": 0.23960066555740434,
+      "grad_norm": 0.3481614589691162,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 0.8049,
+      "step": 108
+    },
+    {
+      "epoch": 0.2418191902384914,
+      "grad_norm": 0.22973711788654327,
+      "learning_rate": 0.0002422222222222222,
+      "loss": 0.8939,
+      "step": 109
+    },
+    {
+      "epoch": 0.24403771491957849,
+      "grad_norm": 0.274651437997818,
+      "learning_rate": 0.00024444444444444443,
+      "loss": 0.8453,
+      "step": 110
+    },
+    {
+      "epoch": 0.24625623960066556,
+      "grad_norm": 0.2910013496875763,
+      "learning_rate": 0.0002466666666666666,
+      "loss": 0.6375,
+      "step": 111
+    },
+    {
+      "epoch": 0.24847476428175264,
+      "grad_norm": 0.40011391043663025,
+      "learning_rate": 0.0002488888888888889,
+      "loss": 0.7083,
+      "step": 112
+    },
+    {
+      "epoch": 0.2506932889628397,
+      "grad_norm": 0.36214378476142883,
+      "learning_rate": 0.0002511111111111111,
+      "loss": 0.7957,
+      "step": 113
+    },
+    {
+      "epoch": 0.2529118136439268,
+      "grad_norm": 0.2949477434158325,
+      "learning_rate": 0.00025333333333333333,
+      "loss": 0.8853,
+      "step": 114
+    },
+    {
+      "epoch": 0.25513033832501386,
+      "grad_norm": 0.24429762363433838,
+      "learning_rate": 0.00025555555555555553,
+      "loss": 0.7252,
+      "step": 115
+    },
+    {
+      "epoch": 0.25734886300610094,
+      "grad_norm": 0.37790197134017944,
+      "learning_rate": 0.0002577777777777778,
+      "loss": 0.8496,
+      "step": 116
+    },
+    {
+      "epoch": 0.259567387687188,
+      "grad_norm": 0.440597802400589,
+      "learning_rate": 0.00026,
+      "loss": 0.7722,
+      "step": 117
+    },
+    {
+      "epoch": 0.2617859123682751,
+      "grad_norm": 0.3294551372528076,
+      "learning_rate": 0.00026222222222222223,
+      "loss": 0.7405,
+      "step": 118
+    },
+    {
+      "epoch": 0.26400443704936216,
+      "grad_norm": 0.3052005469799042,
+      "learning_rate": 0.00026444444444444443,
+      "loss": 0.8652,
+      "step": 119
+    },
+    {
+      "epoch": 0.26622296173044924,
+      "grad_norm": 0.3795583248138428,
+      "learning_rate": 0.0002666666666666666,
+      "loss": 1.0319,
+      "step": 120
+    },
+    {
+      "epoch": 0.2684414864115363,
+      "grad_norm": 0.33274900913238525,
+      "learning_rate": 0.0002688888888888889,
+      "loss": 0.8046,
+      "step": 121
+    },
+    {
+      "epoch": 0.2706600110926234,
+      "grad_norm": 0.27135029435157776,
+      "learning_rate": 0.0002711111111111111,
+      "loss": 0.9111,
+      "step": 122
+    },
+    {
+      "epoch": 0.27287853577371046,
+      "grad_norm": 0.26050445437431335,
+      "learning_rate": 0.00027333333333333333,
+      "loss": 0.6475,
+      "step": 123
+    },
+    {
+      "epoch": 0.27509706045479754,
+      "grad_norm": 0.27241647243499756,
+      "learning_rate": 0.0002755555555555555,
+      "loss": 0.8557,
+      "step": 124
+    },
+    {
+      "epoch": 0.2773155851358846,
+      "grad_norm": 0.2890984117984772,
+      "learning_rate": 0.0002777777777777778,
+      "loss": 0.8526,
+      "step": 125
+    },
+    {
+      "epoch": 0.2795341098169717,
+      "grad_norm": 0.2605191171169281,
+      "learning_rate": 0.00028,
+      "loss": 0.6241,
+      "step": 126
+    },
+    {
+      "epoch": 0.2817526344980588,
+      "grad_norm": 0.28794997930526733,
+      "learning_rate": 0.00028222222222222223,
+      "loss": 0.6858,
+      "step": 127
+    },
+    {
+      "epoch": 0.2839711591791459,
+      "grad_norm": 0.2549107074737549,
+      "learning_rate": 0.0002844444444444444,
+      "loss": 0.7434,
+      "step": 128
+    },
+    {
+      "epoch": 0.28618968386023297,
+      "grad_norm": 0.24853132665157318,
+      "learning_rate": 0.0002866666666666667,
+      "loss": 0.7978,
+      "step": 129
+    },
+    {
+      "epoch": 0.28840820854132004,
+      "grad_norm": 0.2531156539916992,
+      "learning_rate": 0.0002888888888888888,
+      "loss": 0.7638,
+      "step": 130
+    },
+    {
+      "epoch": 0.2906267332224071,
+      "grad_norm": 0.33086395263671875,
+      "learning_rate": 0.0002911111111111111,
+      "loss": 0.8865,
+      "step": 131
+    },
+    {
+      "epoch": 0.2928452579034942,
+      "grad_norm": 0.44577595591545105,
+      "learning_rate": 0.00029333333333333327,
+      "loss": 0.6153,
+      "step": 132
+    },
+    {
+      "epoch": 0.29506378258458127,
+      "grad_norm": 0.3071459233760834,
+      "learning_rate": 0.0002955555555555555,
+      "loss": 0.9093,
+      "step": 133
+    },
+    {
+      "epoch": 0.29728230726566834,
+      "grad_norm": 0.2769240438938141,
+      "learning_rate": 0.0002977777777777777,
+      "loss": 0.7318,
+      "step": 134
+    },
+    {
+      "epoch": 0.2995008319467554,
+      "grad_norm": 0.6804690361022949,
+      "learning_rate": 0.0003,
+      "loss": 0.7025,
+      "step": 135
+    },
+    {
+      "epoch": 0.3017193566278425,
+      "grad_norm": 0.3138647675514221,
+      "learning_rate": 0.00029975308641975307,
+      "loss": 0.6561,
+      "step": 136
+    },
+    {
+      "epoch": 0.30393788130892957,
+      "grad_norm": 0.29210060834884644,
+      "learning_rate": 0.00029950617283950616,
+      "loss": 0.9353,
+      "step": 137
+    },
+    {
+      "epoch": 0.30615640599001664,
+      "grad_norm": 0.28114059567451477,
+      "learning_rate": 0.00029925925925925926,
+      "loss": 0.7752,
+      "step": 138
+    },
+    {
+      "epoch": 0.3083749306711037,
+      "grad_norm": 0.2572714388370514,
+      "learning_rate": 0.0002990123456790123,
+      "loss": 0.6928,
+      "step": 139
+    },
+    {
+      "epoch": 0.3105934553521908,
+      "grad_norm": 0.28904077410697937,
+      "learning_rate": 0.0002987654320987654,
+      "loss": 0.7915,
+      "step": 140
+    },
+    {
+      "epoch": 0.31281198003327787,
+      "grad_norm": 0.3937627673149109,
+      "learning_rate": 0.0002985185185185185,
+      "loss": 0.6543,
+      "step": 141
+    },
+    {
+      "epoch": 0.31503050471436495,
+      "grad_norm": 0.30910035967826843,
+      "learning_rate": 0.0002982716049382716,
+      "loss": 0.7556,
+      "step": 142
+    },
+    {
+      "epoch": 0.317249029395452,
+      "grad_norm": 0.29061636328697205,
+      "learning_rate": 0.0002980246913580247,
+      "loss": 0.8013,
+      "step": 143
+    },
+    {
+      "epoch": 0.3194675540765391,
+      "grad_norm": 0.37458187341690063,
+      "learning_rate": 0.0002977777777777777,
+      "loss": 0.7123,
+      "step": 144
+    },
+    {
+      "epoch": 0.32168607875762617,
+      "grad_norm": 0.3350818455219269,
+      "learning_rate": 0.00029753086419753087,
+      "loss": 0.7308,
+      "step": 145
+    },
+    {
+      "epoch": 0.32390460343871325,
+      "grad_norm": 0.2910625636577606,
+      "learning_rate": 0.0002972839506172839,
+      "loss": 0.7201,
+      "step": 146
+    },
+    {
+      "epoch": 0.3261231281198003,
+      "grad_norm": 0.24531260132789612,
+      "learning_rate": 0.000297037037037037,
+      "loss": 0.576,
+      "step": 147
+    },
+    {
+      "epoch": 0.3283416528008874,
+      "grad_norm": 0.28293201327323914,
+      "learning_rate": 0.0002967901234567901,
+      "loss": 0.7154,
+      "step": 148
+    },
+    {
+      "epoch": 0.33056017748197447,
+      "grad_norm": 0.3246794641017914,
+      "learning_rate": 0.0002965432098765432,
+      "loss": 0.8077,
+      "step": 149
+    },
+    {
+      "epoch": 0.33277870216306155,
+      "grad_norm": 0.3558945655822754,
+      "learning_rate": 0.00029629629629629624,
+      "loss": 0.8003,
+      "step": 150
+    },
+    {
+      "epoch": 0.3349972268441486,
+      "grad_norm": 0.2538098394870758,
+      "learning_rate": 0.00029604938271604933,
+      "loss": 0.758,
+      "step": 151
+    },
+    {
+      "epoch": 0.3372157515252357,
+      "grad_norm": 0.4592326283454895,
+      "learning_rate": 0.0002958024691358025,
+      "loss": 0.8848,
+      "step": 152
+    },
+    {
+      "epoch": 0.33943427620632277,
+      "grad_norm": 0.2907700538635254,
+      "learning_rate": 0.0002955555555555555,
+      "loss": 0.8591,
+      "step": 153
+    },
+    {
+      "epoch": 0.34165280088740985,
+      "grad_norm": 0.2621861398220062,
+      "learning_rate": 0.0002953086419753086,
+      "loss": 0.8402,
+      "step": 154
+    },
+    {
+      "epoch": 0.343871325568497,
+      "grad_norm": 0.3475196659564972,
+      "learning_rate": 0.0002950617283950617,
+      "loss": 0.8697,
+      "step": 155
+    },
+    {
+      "epoch": 0.34608985024958405,
+      "grad_norm": 0.25438445806503296,
+      "learning_rate": 0.0002948148148148148,
+      "loss": 0.8944,
+      "step": 156
+    },
+    {
+      "epoch": 0.3483083749306711,
+      "grad_norm": 0.42704662680625916,
+      "learning_rate": 0.00029456790123456785,
+      "loss": 0.7629,
+      "step": 157
+    },
+    {
+      "epoch": 0.3505268996117582,
+      "grad_norm": 0.23807071149349213,
+      "learning_rate": 0.00029432098765432095,
+      "loss": 0.8115,
+      "step": 158
+    },
+    {
+      "epoch": 0.3527454242928453,
+      "grad_norm": 0.2498120218515396,
+      "learning_rate": 0.00029407407407407404,
+      "loss": 0.6282,
+      "step": 159
+    },
+    {
+      "epoch": 0.35496394897393235,
+      "grad_norm": 0.2693820595741272,
+      "learning_rate": 0.00029382716049382714,
+      "loss": 0.6451,
+      "step": 160
+    },
+    {
+      "epoch": 0.35718247365501943,
+      "grad_norm": 0.24963755905628204,
+      "learning_rate": 0.00029358024691358023,
+      "loss": 0.7675,
+      "step": 161
+    },
+    {
+      "epoch": 0.3594009983361065,
+      "grad_norm": 0.2541040778160095,
+      "learning_rate": 0.00029333333333333327,
+      "loss": 0.8628,
+      "step": 162
+    },
+    {
+      "epoch": 0.3616195230171936,
+      "grad_norm": 0.2562277317047119,
+      "learning_rate": 0.0002930864197530864,
+      "loss": 0.7495,
+      "step": 163
+    },
+    {
+      "epoch": 0.36383804769828065,
+      "grad_norm": 0.2182522863149643,
+      "learning_rate": 0.00029283950617283946,
+      "loss": 0.659,
+      "step": 164
+    },
+    {
+      "epoch": 0.36605657237936773,
+      "grad_norm": 0.2577868103981018,
+      "learning_rate": 0.00029259259259259256,
+      "loss": 0.6716,
+      "step": 165
+    },
+    {
+      "epoch": 0.3682750970604548,
+      "grad_norm": 0.26183345913887024,
+      "learning_rate": 0.00029234567901234565,
+      "loss": 0.8737,
+      "step": 166
+    },
+    {
+      "epoch": 0.3704936217415419,
+      "grad_norm": 0.2564447224140167,
+      "learning_rate": 0.00029209876543209875,
+      "loss": 0.7827,
+      "step": 167
+    },
+    {
+      "epoch": 0.37271214642262895,
+      "grad_norm": 0.21394142508506775,
+      "learning_rate": 0.00029185185185185184,
+      "loss": 0.7118,
+      "step": 168
+    },
+    {
+      "epoch": 0.37493067110371603,
+      "grad_norm": 0.3187384307384491,
+      "learning_rate": 0.0002916049382716049,
+      "loss": 0.6139,
+      "step": 169
+    },
+    {
+      "epoch": 0.3771491957848031,
+      "grad_norm": 0.2963434159755707,
+      "learning_rate": 0.000291358024691358,
+      "loss": 0.7997,
+      "step": 170
+    },
+    {
+      "epoch": 0.3793677204658902,
+      "grad_norm": 0.3670627772808075,
+      "learning_rate": 0.0002911111111111111,
+      "loss": 0.8605,
+      "step": 171
+    },
+    {
+      "epoch": 0.38158624514697725,
+      "grad_norm": 0.29136520624160767,
+      "learning_rate": 0.00029086419753086417,
+      "loss": 0.6809,
+      "step": 172
+    },
+    {
+      "epoch": 0.38380476982806433,
+      "grad_norm": 0.23281855881214142,
+      "learning_rate": 0.00029061728395061726,
+      "loss": 0.7813,
+      "step": 173
+    },
+    {
+      "epoch": 0.3860232945091514,
+      "grad_norm": 0.2910059988498688,
+      "learning_rate": 0.00029037037037037036,
+      "loss": 0.8936,
+      "step": 174
+    },
+    {
+      "epoch": 0.3882418191902385,
+      "grad_norm": 0.26949936151504517,
+      "learning_rate": 0.00029012345679012345,
+      "loss": 0.7772,
+      "step": 175
+    },
+    {
+      "epoch": 0.39046034387132555,
+      "grad_norm": 0.24920910596847534,
+      "learning_rate": 0.0002898765432098765,
+      "loss": 0.7606,
+      "step": 176
+    },
+    {
+      "epoch": 0.39267886855241263,
+      "grad_norm": 0.3323945701122284,
+      "learning_rate": 0.0002896296296296296,
+      "loss": 0.8299,
+      "step": 177
+    },
+    {
+      "epoch": 0.3948973932334997,
+      "grad_norm": 0.23166917264461517,
+      "learning_rate": 0.0002893827160493827,
+      "loss": 0.7133,
+      "step": 178
+    },
+    {
+      "epoch": 0.3971159179145868,
+      "grad_norm": 0.23577314615249634,
+      "learning_rate": 0.0002891358024691358,
+      "loss": 0.8837,
+      "step": 179
+    },
+    {
+      "epoch": 0.39933444259567386,
+      "grad_norm": 0.2735862135887146,
+      "learning_rate": 0.0002888888888888888,
+      "loss": 0.7391,
+      "step": 180
+    },
+    {
+      "epoch": 0.40155296727676093,
+      "grad_norm": 0.36028414964675903,
+      "learning_rate": 0.00028864197530864197,
+      "loss": 0.861,
+      "step": 181
+    },
+    {
+      "epoch": 0.403771491957848,
+      "grad_norm": 0.23426638543605804,
+      "learning_rate": 0.00028839506172839507,
+      "loss": 0.6658,
+      "step": 182
+    },
+    {
+      "epoch": 0.4059900166389351,
+      "grad_norm": 0.3255574107170105,
+      "learning_rate": 0.0002881481481481481,
+      "loss": 0.7012,
+      "step": 183
+    },
+    {
+      "epoch": 0.4082085413200222,
+      "grad_norm": 0.24860937893390656,
+      "learning_rate": 0.0002879012345679012,
+      "loss": 0.8203,
+      "step": 184
+    },
+    {
+      "epoch": 0.4104270660011093,
+      "grad_norm": 0.2390868067741394,
+      "learning_rate": 0.0002876543209876543,
+      "loss": 0.6305,
+      "step": 185
+    },
+    {
+      "epoch": 0.41264559068219636,
+      "grad_norm": 0.37576478719711304,
+      "learning_rate": 0.0002874074074074074,
+      "loss": 0.8431,
+      "step": 186
+    },
+    {
+      "epoch": 0.41486411536328344,
+      "grad_norm": 0.3355439007282257,
+      "learning_rate": 0.00028716049382716043,
+      "loss": 0.7795,
+      "step": 187
+    },
+    {
+      "epoch": 0.4170826400443705,
+      "grad_norm": 0.2477898746728897,
+      "learning_rate": 0.00028691358024691353,
+      "loss": 0.7379,
+      "step": 188
+    },
+    {
+      "epoch": 0.4193011647254576,
+      "grad_norm": 0.3028287887573242,
+      "learning_rate": 0.0002866666666666667,
+      "loss": 1.0218,
+      "step": 189
+    },
+    {
+      "epoch": 0.42151968940654466,
+      "grad_norm": 0.23490504920482635,
+      "learning_rate": 0.0002864197530864197,
+      "loss": 0.6997,
+      "step": 190
+    },
+    {
+      "epoch": 0.42373821408763174,
+      "grad_norm": 0.21243245899677277,
+      "learning_rate": 0.0002861728395061728,
+      "loss": 0.6163,
+      "step": 191
+    },
+    {
+      "epoch": 0.4259567387687188,
+      "grad_norm": 0.2679629921913147,
+      "learning_rate": 0.0002859259259259259,
+      "loss": 0.8084,
+      "step": 192
+    },
+    {
+      "epoch": 0.4281752634498059,
+      "grad_norm": 0.2221756875514984,
+      "learning_rate": 0.000285679012345679,
+      "loss": 0.7283,
+      "step": 193
+    },
+    {
+      "epoch": 0.43039378813089296,
+      "grad_norm": 0.26024895906448364,
+      "learning_rate": 0.00028543209876543204,
+      "loss": 0.6153,
+      "step": 194
+    },
+    {
+      "epoch": 0.43261231281198004,
+      "grad_norm": 0.31024396419525146,
+      "learning_rate": 0.00028518518518518514,
+      "loss": 0.6931,
+      "step": 195
+    },
+    {
+      "epoch": 0.4348308374930671,
+      "grad_norm": 0.29099681973457336,
+      "learning_rate": 0.00028493827160493824,
+      "loss": 0.9842,
+      "step": 196
+    },
+    {
+      "epoch": 0.4370493621741542,
+      "grad_norm": 0.2998812198638916,
+      "learning_rate": 0.00028469135802469133,
+      "loss": 0.8047,
+      "step": 197
+    },
+    {
+      "epoch": 0.43926788685524126,
+      "grad_norm": 0.23200857639312744,
+      "learning_rate": 0.0002844444444444444,
+      "loss": 0.752,
+      "step": 198
+    },
+    {
+      "epoch": 0.44148641153632834,
+      "grad_norm": 0.32312023639678955,
+      "learning_rate": 0.0002841975308641975,
+      "loss": 0.8416,
+      "step": 199
+    },
+    {
+      "epoch": 0.4437049362174154,
+      "grad_norm": 0.21699897944927216,
+      "learning_rate": 0.0002839506172839506,
+      "loss": 0.7578,
+      "step": 200
+    },
+    {
+      "epoch": 0.4459234608985025,
+      "grad_norm": 0.24872416257858276,
+      "learning_rate": 0.00028370370370370366,
+      "loss": 0.7496,
+      "step": 201
+    },
+    {
+      "epoch": 0.44814198557958956,
+      "grad_norm": 0.2464827001094818,
+      "learning_rate": 0.00028345679012345675,
+      "loss": 0.6443,
+      "step": 202
+    },
+    {
+      "epoch": 0.45036051026067664,
+      "grad_norm": 0.2419295608997345,
+      "learning_rate": 0.00028320987654320985,
+      "loss": 0.7848,
+      "step": 203
+    },
+    {
+      "epoch": 0.4525790349417637,
+      "grad_norm": 0.3549902141094208,
+      "learning_rate": 0.00028296296296296294,
+      "loss": 1.0614,
+      "step": 204
+    },
+    {
+      "epoch": 0.4547975596228508,
+      "grad_norm": 0.31317585706710815,
+      "learning_rate": 0.00028271604938271604,
+      "loss": 0.9018,
+      "step": 205
+    },
+    {
+      "epoch": 0.45701608430393786,
+      "grad_norm": 0.2557738721370697,
+      "learning_rate": 0.0002824691358024691,
+      "loss": 0.7167,
+      "step": 206
+    },
+    {
+      "epoch": 0.45923460898502494,
+      "grad_norm": 0.23882776498794556,
+      "learning_rate": 0.00028222222222222223,
+      "loss": 0.8479,
+      "step": 207
+    },
+    {
+      "epoch": 0.461453133666112,
+      "grad_norm": 0.28240546584129333,
+      "learning_rate": 0.00028197530864197527,
+      "loss": 0.7118,
+      "step": 208
+    },
+    {
+      "epoch": 0.4636716583471991,
+      "grad_norm": 0.3159230053424835,
+      "learning_rate": 0.00028172839506172836,
+      "loss": 0.6441,
+      "step": 209
+    },
+    {
+      "epoch": 0.46589018302828616,
+      "grad_norm": 0.28373125195503235,
+      "learning_rate": 0.00028148148148148146,
+      "loss": 0.7895,
+      "step": 210
+    },
+    {
+      "epoch": 0.46810870770937324,
+      "grad_norm": 0.19972574710845947,
+      "learning_rate": 0.00028123456790123455,
+      "loss": 0.6675,
+      "step": 211
+    },
+    {
+      "epoch": 0.47032723239046037,
+      "grad_norm": 0.3083205223083496,
+      "learning_rate": 0.00028098765432098765,
+      "loss": 0.7718,
+      "step": 212
+    },
+    {
+      "epoch": 0.47254575707154745,
+      "grad_norm": 0.2694496810436249,
+      "learning_rate": 0.0002807407407407407,
+      "loss": 0.7788,
+      "step": 213
+    },
+    {
+      "epoch": 0.4747642817526345,
+      "grad_norm": 0.299120157957077,
+      "learning_rate": 0.0002804938271604938,
+      "loss": 0.7681,
+      "step": 214
+    },
+    {
+      "epoch": 0.4769828064337216,
+      "grad_norm": 0.3270832598209381,
+      "learning_rate": 0.0002802469135802469,
+      "loss": 0.7503,
+      "step": 215
+    },
+    {
+      "epoch": 0.47920133111480867,
+      "grad_norm": 0.296480655670166,
+      "learning_rate": 0.00028,
+      "loss": 0.7776,
+      "step": 216
+    },
+    {
+      "epoch": 0.48141985579589575,
+      "grad_norm": 0.29667648673057556,
+      "learning_rate": 0.00027975308641975307,
+      "loss": 0.7291,
+      "step": 217
+    },
+    {
+      "epoch": 0.4836383804769828,
+      "grad_norm": 0.25731468200683594,
+      "learning_rate": 0.00027950617283950617,
+      "loss": 0.6892,
+      "step": 218
+    },
+    {
+      "epoch": 0.4858569051580699,
+      "grad_norm": 0.30635109543800354,
+      "learning_rate": 0.00027925925925925926,
+      "loss": 0.6766,
+      "step": 219
+    },
+    {
+      "epoch": 0.48807542983915697,
+      "grad_norm": 0.25647830963134766,
+      "learning_rate": 0.0002790123456790123,
+      "loss": 0.7129,
+      "step": 220
+    },
+    {
+      "epoch": 0.49029395452024405,
+      "grad_norm": 0.26678094267845154,
+      "learning_rate": 0.0002787654320987654,
+      "loss": 0.8358,
+      "step": 221
+    },
+    {
+      "epoch": 0.4925124792013311,
+      "grad_norm": 0.2602359354496002,
+      "learning_rate": 0.0002785185185185185,
+      "loss": 0.7974,
+      "step": 222
+    },
+    {
+      "epoch": 0.4947310038824182,
+      "grad_norm": 0.19614528119564056,
+      "learning_rate": 0.0002782716049382716,
+      "loss": 0.8367,
+      "step": 223
+    },
+    {
+      "epoch": 0.49694952856350527,
+      "grad_norm": 0.270206093788147,
+      "learning_rate": 0.00027802469135802463,
+      "loss": 0.8392,
+      "step": 224
+    },
+    {
+      "epoch": 0.49916805324459235,
+      "grad_norm": 0.27967438101768494,
+      "learning_rate": 0.0002777777777777778,
+      "loss": 0.6989,
+      "step": 225
+    },
+    {
+      "epoch": 0.5013865779256794,
+      "grad_norm": 0.3304283618927002,
+      "learning_rate": 0.00027753086419753087,
+      "loss": 0.8841,
+      "step": 226
+    },
+    {
+      "epoch": 0.5036051026067665,
+      "grad_norm": 0.26649901270866394,
+      "learning_rate": 0.0002772839506172839,
+      "loss": 0.9695,
+      "step": 227
+    },
+    {
+      "epoch": 0.5058236272878536,
+      "grad_norm": 0.23750439286231995,
+      "learning_rate": 0.000277037037037037,
+      "loss": 0.7164,
+      "step": 228
+    },
+    {
+      "epoch": 0.5080421519689406,
+      "grad_norm": 0.2571529746055603,
+      "learning_rate": 0.0002767901234567901,
+      "loss": 0.8295,
+      "step": 229
+    },
+    {
+      "epoch": 0.5102606766500277,
+      "grad_norm": 0.23797377943992615,
+      "learning_rate": 0.0002765432098765432,
+      "loss": 0.7802,
+      "step": 230
+    },
+    {
+      "epoch": 0.5124792013311148,
+      "grad_norm": 0.3578810691833496,
+      "learning_rate": 0.00027629629629629624,
+      "loss": 0.6303,
+      "step": 231
+    },
+    {
+      "epoch": 0.5146977260122019,
+      "grad_norm": 0.26352182030677795,
+      "learning_rate": 0.00027604938271604934,
+      "loss": 0.8027,
+      "step": 232
+    },
+    {
+      "epoch": 0.516916250693289,
+      "grad_norm": 0.2451726794242859,
+      "learning_rate": 0.00027580246913580243,
+      "loss": 0.8162,
+      "step": 233
+    },
+    {
+      "epoch": 0.519134775374376,
+      "grad_norm": 0.2787698805332184,
+      "learning_rate": 0.0002755555555555555,
+      "loss": 0.6924,
+      "step": 234
+    },
+    {
+      "epoch": 0.5213533000554631,
+      "grad_norm": 0.36408740282058716,
+      "learning_rate": 0.0002753086419753086,
+      "loss": 0.6189,
+      "step": 235
+    },
+    {
+      "epoch": 0.5235718247365502,
+      "grad_norm": 0.24031764268875122,
+      "learning_rate": 0.0002750617283950617,
+      "loss": 0.7822,
+      "step": 236
+    },
+    {
+      "epoch": 0.5257903494176372,
+      "grad_norm": 0.26255810260772705,
+      "learning_rate": 0.0002748148148148148,
+      "loss": 0.7804,
+      "step": 237
+    },
+    {
+      "epoch": 0.5280088740987243,
+      "grad_norm": 0.3046834170818329,
+      "learning_rate": 0.00027456790123456785,
+      "loss": 0.8223,
+      "step": 238
+    },
+    {
+      "epoch": 0.5302273987798114,
+      "grad_norm": 0.29656538367271423,
+      "learning_rate": 0.00027432098765432095,
+      "loss": 0.6727,
+      "step": 239
+    },
+    {
+      "epoch": 0.5324459234608985,
+      "grad_norm": 0.24222400784492493,
+      "learning_rate": 0.00027407407407407404,
+      "loss": 0.6442,
+      "step": 240
+    },
+    {
+      "epoch": 0.5346644481419855,
+      "grad_norm": 0.29241377115249634,
+      "learning_rate": 0.00027382716049382714,
+      "loss": 0.8001,
+      "step": 241
+    },
+    {
+      "epoch": 0.5368829728230726,
+      "grad_norm": 0.2967418432235718,
+      "learning_rate": 0.00027358024691358023,
+      "loss": 0.701,
+      "step": 242
+    },
+    {
+      "epoch": 0.5391014975041597,
+      "grad_norm": 0.2316524237394333,
+      "learning_rate": 0.00027333333333333333,
+      "loss": 0.656,
+      "step": 243
+    },
+    {
+      "epoch": 0.5413200221852468,
+      "grad_norm": 0.2715781629085541,
+      "learning_rate": 0.0002730864197530864,
+      "loss": 0.6923,
+      "step": 244
+    },
+    {
+      "epoch": 0.5435385468663338,
+      "grad_norm": 0.3755452632904053,
+      "learning_rate": 0.00027283950617283946,
+      "loss": 0.7678,
+      "step": 245
+    },
+    {
+      "epoch": 0.5457570715474209,
+      "grad_norm": 0.32851141691207886,
+      "learning_rate": 0.00027259259259259256,
+      "loss": 0.64,
+      "step": 246
+    },
+    {
+      "epoch": 0.547975596228508,
+      "grad_norm": 0.22852662205696106,
+      "learning_rate": 0.00027234567901234565,
+      "loss": 0.7362,
+      "step": 247
+    },
+    {
+      "epoch": 0.5501941209095951,
+      "grad_norm": 0.29149600863456726,
+      "learning_rate": 0.00027209876543209875,
+      "loss": 0.5983,
+      "step": 248
+    },
+    {
+      "epoch": 0.5524126455906821,
+      "grad_norm": 0.20533788204193115,
+      "learning_rate": 0.00027185185185185184,
+      "loss": 0.6574,
+      "step": 249
+    },
+    {
+      "epoch": 0.5546311702717692,
+      "grad_norm": 0.29093581438064575,
+      "learning_rate": 0.0002716049382716049,
+      "loss": 0.6375,
+      "step": 250
+    },
+    {
+      "epoch": 0.5568496949528563,
+      "grad_norm": 0.3360643684864044,
+      "learning_rate": 0.00027135802469135803,
+      "loss": 0.8229,
+      "step": 251
+    },
+    {
+      "epoch": 0.5590682196339434,
+      "grad_norm": 0.27565768361091614,
+      "learning_rate": 0.0002711111111111111,
+      "loss": 0.6725,
+      "step": 252
+    },
+    {
+      "epoch": 0.5612867443150305,
+      "grad_norm": 0.21658623218536377,
+      "learning_rate": 0.00027086419753086417,
+      "loss": 0.7591,
+      "step": 253
+    },
+    {
+      "epoch": 0.5635052689961176,
+      "grad_norm": 0.28960350155830383,
+      "learning_rate": 0.00027061728395061727,
+      "loss": 0.8615,
+      "step": 254
+    },
+    {
+      "epoch": 0.5657237936772047,
+      "grad_norm": 0.2378702014684677,
+      "learning_rate": 0.00027037037037037036,
+      "loss": 0.6773,
+      "step": 255
+    },
+    {
+      "epoch": 0.5679423183582918,
+      "grad_norm": 0.2553529143333435,
+      "learning_rate": 0.00027012345679012346,
+      "loss": 0.6563,
+      "step": 256
+    },
+    {
+      "epoch": 0.5701608430393789,
+      "grad_norm": 0.2637943923473358,
+      "learning_rate": 0.0002698765432098765,
+      "loss": 0.7121,
+      "step": 257
+    },
+    {
+      "epoch": 0.5723793677204659,
+      "grad_norm": 0.35551363229751587,
+      "learning_rate": 0.0002696296296296296,
+      "loss": 0.8343,
+      "step": 258
+    },
+    {
+      "epoch": 0.574597892401553,
+      "grad_norm": 0.21420539915561676,
+      "learning_rate": 0.0002693827160493827,
+      "loss": 0.75,
+      "step": 259
+    },
+    {
+      "epoch": 0.5768164170826401,
+      "grad_norm": 0.22028757631778717,
+      "learning_rate": 0.0002691358024691358,
+      "loss": 0.7573,
+      "step": 260
+    },
+    {
+      "epoch": 0.5790349417637272,
+      "grad_norm": 0.36668267846107483,
+      "learning_rate": 0.0002688888888888889,
+      "loss": 0.9701,
+      "step": 261
+    },
+    {
+      "epoch": 0.5812534664448142,
+      "grad_norm": 0.28284168243408203,
+      "learning_rate": 0.00026864197530864197,
+      "loss": 0.5681,
+      "step": 262
+    },
+    {
+      "epoch": 0.5834719911259013,
+      "grad_norm": 0.27233126759529114,
+      "learning_rate": 0.000268395061728395,
+      "loss": 0.6894,
+      "step": 263
+    },
+    {
+      "epoch": 0.5856905158069884,
+      "grad_norm": 0.4742943346500397,
+      "learning_rate": 0.0002681481481481481,
+      "loss": 0.669,
+      "step": 264
+    },
+    {
+      "epoch": 0.5879090404880755,
+      "grad_norm": 0.25099506974220276,
+      "learning_rate": 0.0002679012345679012,
+      "loss": 0.7191,
+      "step": 265
+    },
+    {
+      "epoch": 0.5901275651691625,
+      "grad_norm": 0.2981354594230652,
+      "learning_rate": 0.0002676543209876543,
+      "loss": 0.8376,
+      "step": 266
+    },
+    {
+      "epoch": 0.5923460898502496,
+      "grad_norm": 0.28779613971710205,
+      "learning_rate": 0.0002674074074074074,
+      "loss": 0.7136,
+      "step": 267
+    },
+    {
+      "epoch": 0.5945646145313367,
+      "grad_norm": 0.2971796691417694,
+      "learning_rate": 0.00026716049382716043,
+      "loss": 0.8759,
+      "step": 268
+    },
+    {
+      "epoch": 0.5967831392124238,
+      "grad_norm": 0.24150897562503815,
+      "learning_rate": 0.0002669135802469136,
+      "loss": 0.8713,
+      "step": 269
+    },
+    {
+      "epoch": 0.5990016638935108,
+      "grad_norm": 0.31380656361579895,
+      "learning_rate": 0.0002666666666666666,
+      "loss": 0.774,
+      "step": 270
+    },
+    {
+      "epoch": 0.6012201885745979,
+      "grad_norm": 0.39140966534614563,
+      "learning_rate": 0.0002664197530864197,
+      "loss": 0.858,
+      "step": 271
+    },
+    {
+      "epoch": 0.603438713255685,
+      "grad_norm": 0.3054139316082001,
+      "learning_rate": 0.0002661728395061728,
+      "loss": 0.5633,
+      "step": 272
+    },
+    {
+      "epoch": 0.6056572379367721,
+      "grad_norm": 0.29849758744239807,
+      "learning_rate": 0.0002659259259259259,
+      "loss": 0.8113,
+      "step": 273
+    },
+    {
+      "epoch": 0.6078757626178591,
+      "grad_norm": 0.26137489080429077,
+      "learning_rate": 0.000265679012345679,
+      "loss": 0.8754,
+      "step": 274
+    },
+    {
+      "epoch": 0.6100942872989462,
+      "grad_norm": 0.2011139839887619,
+      "learning_rate": 0.00026543209876543205,
+      "loss": 0.5651,
+      "step": 275
+    },
+    {
+      "epoch": 0.6123128119800333,
+      "grad_norm": 0.23980490863323212,
+      "learning_rate": 0.00026518518518518514,
+      "loss": 0.7808,
+      "step": 276
+    },
+    {
+      "epoch": 0.6145313366611204,
+      "grad_norm": 0.2380177229642868,
+      "learning_rate": 0.00026493827160493824,
+      "loss": 0.7763,
+      "step": 277
+    },
+    {
+      "epoch": 0.6167498613422074,
+      "grad_norm": 0.2957155406475067,
+      "learning_rate": 0.00026469135802469133,
+      "loss": 0.6841,
+      "step": 278
+    },
+    {
+      "epoch": 0.6189683860232945,
+      "grad_norm": 0.21348552405834198,
+      "learning_rate": 0.00026444444444444443,
+      "loss": 0.6847,
+      "step": 279
+    },
+    {
+      "epoch": 0.6211869107043816,
+      "grad_norm": 0.23422791063785553,
+      "learning_rate": 0.0002641975308641975,
+      "loss": 0.6404,
+      "step": 280
+    },
+    {
+      "epoch": 0.6234054353854687,
+      "grad_norm": 0.24418871104717255,
+      "learning_rate": 0.0002639506172839506,
+      "loss": 0.6901,
+      "step": 281
+    },
+    {
+      "epoch": 0.6256239600665557,
+      "grad_norm": 0.3436889350414276,
+      "learning_rate": 0.00026370370370370366,
+      "loss": 0.6504,
+      "step": 282
+    },
+    {
+      "epoch": 0.6278424847476428,
+      "grad_norm": 0.1974991410970688,
+      "learning_rate": 0.00026345679012345675,
+      "loss": 0.5366,
+      "step": 283
+    },
+    {
+      "epoch": 0.6300610094287299,
+      "grad_norm": 0.31095439195632935,
+      "learning_rate": 0.00026320987654320985,
+      "loss": 0.6722,
+      "step": 284
+    },
+    {
+      "epoch": 0.632279534109817,
+      "grad_norm": 0.25184550881385803,
+      "learning_rate": 0.00026296296296296294,
+      "loss": 0.7093,
+      "step": 285
+    },
+    {
+      "epoch": 0.634498058790904,
+      "grad_norm": 0.30081167817115784,
+      "learning_rate": 0.00026271604938271604,
+      "loss": 0.7161,
+      "step": 286
+    },
+    {
+      "epoch": 0.6367165834719911,
+      "grad_norm": 0.2533697783946991,
+      "learning_rate": 0.00026246913580246913,
+      "loss": 0.7682,
+      "step": 287
+    },
+    {
+      "epoch": 0.6389351081530782,
+      "grad_norm": 0.5759953856468201,
+      "learning_rate": 0.00026222222222222223,
+      "loss": 0.7091,
+      "step": 288
+    },
+    {
+      "epoch": 0.6411536328341653,
+      "grad_norm": 0.20291733741760254,
+      "learning_rate": 0.00026197530864197527,
+      "loss": 0.807,
+      "step": 289
+    },
+    {
+      "epoch": 0.6433721575152523,
+      "grad_norm": 0.5243172645568848,
+      "learning_rate": 0.00026172839506172837,
+      "loss": 0.7839,
+      "step": 290
+    },
+    {
+      "epoch": 0.6455906821963394,
+      "grad_norm": 0.3201848566532135,
+      "learning_rate": 0.00026148148148148146,
+      "loss": 0.8361,
+      "step": 291
+    },
+    {
+      "epoch": 0.6478092068774265,
+      "grad_norm": 0.23983563482761383,
+      "learning_rate": 0.00026123456790123456,
+      "loss": 0.8274,
+      "step": 292
+    },
+    {
+      "epoch": 0.6500277315585136,
+      "grad_norm": 0.3576997220516205,
+      "learning_rate": 0.0002609876543209876,
+      "loss": 0.7646,
+      "step": 293
+    },
+    {
+      "epoch": 0.6522462562396006,
+      "grad_norm": 0.24339282512664795,
+      "learning_rate": 0.0002607407407407407,
+      "loss": 0.6886,
+      "step": 294
+    },
+    {
+      "epoch": 0.6544647809206877,
+      "grad_norm": 0.24387285113334656,
+      "learning_rate": 0.00026049382716049384,
+      "loss": 0.5715,
+      "step": 295
+    },
+    {
+      "epoch": 0.6566833056017748,
+      "grad_norm": 0.302621454000473,
+      "learning_rate": 0.0002602469135802469,
+      "loss": 0.8409,
+      "step": 296
+    },
+    {
+      "epoch": 0.6589018302828619,
+      "grad_norm": 0.392571359872818,
+      "learning_rate": 0.00026,
+      "loss": 0.6579,
+      "step": 297
+    },
+    {
+      "epoch": 0.6611203549639489,
+      "grad_norm": 0.24242691695690155,
+      "learning_rate": 0.00025975308641975307,
+      "loss": 0.6888,
+      "step": 298
+    },
+    {
+      "epoch": 0.663338879645036,
+      "grad_norm": 0.26514554023742676,
+      "learning_rate": 0.00025950617283950617,
+      "loss": 0.8936,
+      "step": 299
+    },
+    {
+      "epoch": 0.6655574043261231,
+      "grad_norm": 0.29383355379104614,
+      "learning_rate": 0.0002592592592592592,
+      "loss": 0.8674,
+      "step": 300
+    },
+    {
+      "epoch": 0.6677759290072102,
+      "grad_norm": 0.3661712110042572,
+      "learning_rate": 0.0002590123456790123,
+      "loss": 0.6426,
+      "step": 301
+    },
+    {
+      "epoch": 0.6699944536882972,
+      "grad_norm": 0.3070962727069855,
+      "learning_rate": 0.0002587654320987654,
+      "loss": 0.8028,
+      "step": 302
+    },
+    {
+      "epoch": 0.6722129783693843,
+      "grad_norm": 0.2132289707660675,
+      "learning_rate": 0.0002585185185185185,
+      "loss": 0.6673,
+      "step": 303
+    },
+    {
+      "epoch": 0.6744315030504714,
+      "grad_norm": 0.2550564110279083,
+      "learning_rate": 0.0002582716049382716,
+      "loss": 0.5746,
+      "step": 304
+    },
+    {
+      "epoch": 0.6766500277315585,
+      "grad_norm": 0.22844505310058594,
+      "learning_rate": 0.0002580246913580247,
+      "loss": 0.7652,
+      "step": 305
+    },
+    {
+      "epoch": 0.6788685524126455,
+      "grad_norm": 0.30164358019828796,
+      "learning_rate": 0.0002577777777777778,
+      "loss": 0.6536,
+      "step": 306
+    },
+    {
+      "epoch": 0.6810870770937326,
+      "grad_norm": 0.2830958366394043,
+      "learning_rate": 0.0002575308641975308,
+      "loss": 0.7847,
+      "step": 307
+    },
+    {
+      "epoch": 0.6833056017748197,
+      "grad_norm": 0.2820102572441101,
+      "learning_rate": 0.0002572839506172839,
+      "loss": 0.8129,
+      "step": 308
+    },
+    {
+      "epoch": 0.6855241264559068,
+      "grad_norm": 0.24437691271305084,
+      "learning_rate": 0.000257037037037037,
+      "loss": 0.6815,
+      "step": 309
+    },
+    {
+      "epoch": 0.687742651136994,
+      "grad_norm": 0.35614338517189026,
+      "learning_rate": 0.0002567901234567901,
+      "loss": 0.6362,
+      "step": 310
+    },
+    {
+      "epoch": 0.689961175818081,
+      "grad_norm": 0.27009281516075134,
+      "learning_rate": 0.0002565432098765432,
+      "loss": 0.8224,
+      "step": 311
+    },
+    {
+      "epoch": 0.6921797004991681,
+      "grad_norm": 0.25577667355537415,
+      "learning_rate": 0.00025629629629629624,
+      "loss": 0.7625,
+      "step": 312
+    },
+    {
+      "epoch": 0.6943982251802552,
+      "grad_norm": 0.26589664816856384,
+      "learning_rate": 0.0002560493827160494,
+      "loss": 0.6848,
+      "step": 313
+    },
+    {
+      "epoch": 0.6966167498613423,
+      "grad_norm": 0.27902746200561523,
+      "learning_rate": 0.00025580246913580243,
+      "loss": 0.7608,
+      "step": 314
+    },
+    {
+      "epoch": 0.6988352745424293,
+      "grad_norm": 0.23629344999790192,
+      "learning_rate": 0.00025555555555555553,
+      "loss": 0.7078,
+      "step": 315
+    },
+    {
+      "epoch": 0.7010537992235164,
+      "grad_norm": 0.25023365020751953,
+      "learning_rate": 0.0002553086419753086,
+      "loss": 0.6596,
+      "step": 316
+    },
+    {
+      "epoch": 0.7032723239046035,
+      "grad_norm": 0.22880077362060547,
+      "learning_rate": 0.0002550617283950617,
+      "loss": 0.5601,
+      "step": 317
+    },
+    {
+      "epoch": 0.7054908485856906,
+      "grad_norm": 0.27988433837890625,
+      "learning_rate": 0.0002548148148148148,
+      "loss": 0.5444,
+      "step": 318
+    },
+    {
+      "epoch": 0.7077093732667776,
+      "grad_norm": 0.2692090570926666,
+      "learning_rate": 0.00025456790123456785,
+      "loss": 0.7884,
+      "step": 319
+    },
+    {
+      "epoch": 0.7099278979478647,
+      "grad_norm": 0.3106870651245117,
+      "learning_rate": 0.00025432098765432095,
+      "loss": 0.8523,
+      "step": 320
+    },
+    {
+      "epoch": 0.7121464226289518,
+      "grad_norm": 0.23829683661460876,
+      "learning_rate": 0.00025407407407407404,
+      "loss": 0.6541,
+      "step": 321
+    },
+    {
+      "epoch": 0.7143649473100389,
+      "grad_norm": 0.3049006462097168,
+      "learning_rate": 0.00025382716049382714,
+      "loss": 0.8058,
+      "step": 322
+    },
+    {
+      "epoch": 0.7165834719911259,
+      "grad_norm": 0.2554822862148285,
+      "learning_rate": 0.00025358024691358023,
+      "loss": 0.7225,
+      "step": 323
+    },
+    {
+      "epoch": 0.718801996672213,
+      "grad_norm": 0.3805801272392273,
+      "learning_rate": 0.00025333333333333333,
+      "loss": 0.799,
+      "step": 324
+    },
+    {
+      "epoch": 0.7210205213533001,
+      "grad_norm": 0.2639758586883545,
+      "learning_rate": 0.0002530864197530864,
+      "loss": 0.5851,
+      "step": 325
+    },
+    {
+      "epoch": 0.7232390460343872,
+      "grad_norm": 0.26287147402763367,
+      "learning_rate": 0.00025283950617283947,
+      "loss": 0.7474,
+      "step": 326
+    },
+    {
+      "epoch": 0.7254575707154742,
+      "grad_norm": 0.2739676237106323,
+      "learning_rate": 0.00025259259259259256,
+      "loss": 0.8407,
+      "step": 327
+    },
+    {
+      "epoch": 0.7276760953965613,
+      "grad_norm": 0.2047872394323349,
+      "learning_rate": 0.00025234567901234566,
+      "loss": 0.7804,
+      "step": 328
+    },
+    {
+      "epoch": 0.7298946200776484,
+      "grad_norm": 0.2946568727493286,
+      "learning_rate": 0.00025209876543209875,
+      "loss": 0.7719,
+      "step": 329
+    },
+    {
+      "epoch": 0.7321131447587355,
+      "grad_norm": 0.27364087104797363,
+      "learning_rate": 0.0002518518518518518,
+      "loss": 0.7617,
+      "step": 330
+    },
+    {
+      "epoch": 0.7343316694398225,
+      "grad_norm": 0.25805965065956116,
+      "learning_rate": 0.00025160493827160494,
+      "loss": 0.8251,
+      "step": 331
+    },
+    {
+      "epoch": 0.7365501941209096,
+      "grad_norm": 0.23653222620487213,
+      "learning_rate": 0.00025135802469135804,
+      "loss": 0.7494,
+      "step": 332
+    },
+    {
+      "epoch": 0.7387687188019967,
+      "grad_norm": 0.1868569701910019,
+      "learning_rate": 0.0002511111111111111,
+      "loss": 0.5644,
+      "step": 333
+    },
+    {
+      "epoch": 0.7409872434830838,
+      "grad_norm": 0.2814759314060211,
+      "learning_rate": 0.00025086419753086417,
+      "loss": 0.7631,
+      "step": 334
+    },
+    {
+      "epoch": 0.7432057681641708,
+      "grad_norm": 0.31236061453819275,
+      "learning_rate": 0.00025061728395061727,
+      "loss": 0.7689,
+      "step": 335
+    },
+    {
+      "epoch": 0.7454242928452579,
+      "grad_norm": 0.2865675985813141,
+      "learning_rate": 0.00025037037037037036,
+      "loss": 0.7975,
+      "step": 336
+    },
+    {
+      "epoch": 0.747642817526345,
+      "grad_norm": 0.2942241430282593,
+      "learning_rate": 0.0002501234567901234,
+      "loss": 0.7987,
+      "step": 337
+    },
+    {
+      "epoch": 0.7498613422074321,
+      "grad_norm": 0.25397658348083496,
+      "learning_rate": 0.0002498765432098765,
+      "loss": 0.7936,
+      "step": 338
+    },
+    {
+      "epoch": 0.7520798668885191,
+      "grad_norm": 0.33798545598983765,
+      "learning_rate": 0.00024962962962962965,
+      "loss": 0.8619,
+      "step": 339
+    },
+    {
+      "epoch": 0.7542983915696062,
+      "grad_norm": 0.22395390272140503,
+      "learning_rate": 0.0002493827160493827,
+      "loss": 0.8449,
+      "step": 340
+    },
+    {
+      "epoch": 0.7565169162506933,
+      "grad_norm": 0.4437278211116791,
+      "learning_rate": 0.0002491358024691358,
+      "loss": 0.7646,
+      "step": 341
+    },
+    {
+      "epoch": 0.7587354409317804,
+      "grad_norm": 0.32571929693222046,
+      "learning_rate": 0.0002488888888888889,
+      "loss": 0.5934,
+      "step": 342
+    },
+    {
+      "epoch": 0.7609539656128674,
+      "grad_norm": 0.2665906846523285,
+      "learning_rate": 0.000248641975308642,
+      "loss": 0.5806,
+      "step": 343
+    },
+    {
+      "epoch": 0.7631724902939545,
+      "grad_norm": 0.33049866557121277,
+      "learning_rate": 0.000248395061728395,
+      "loss": 0.7851,
+      "step": 344
+    },
+    {
+      "epoch": 0.7653910149750416,
+      "grad_norm": 0.29507264494895935,
+      "learning_rate": 0.0002481481481481481,
+      "loss": 0.8642,
+      "step": 345
+    },
+    {
+      "epoch": 0.7676095396561287,
+      "grad_norm": 0.2410031110048294,
+      "learning_rate": 0.0002479012345679012,
+      "loss": 0.7291,
+      "step": 346
+    },
+    {
+      "epoch": 0.7698280643372157,
+      "grad_norm": 0.31430765986442566,
+      "learning_rate": 0.0002476543209876543,
+      "loss": 0.5847,
+      "step": 347
+    },
+    {
+      "epoch": 0.7720465890183028,
+      "grad_norm": 0.30985310673713684,
+      "learning_rate": 0.0002474074074074074,
+      "loss": 0.728,
+      "step": 348
+    },
+    {
+      "epoch": 0.7742651136993899,
+      "grad_norm": 0.3210037350654602,
+      "learning_rate": 0.0002471604938271605,
+      "loss": 0.6406,
+      "step": 349
+    },
+    {
+      "epoch": 0.776483638380477,
+      "grad_norm": 0.34360232949256897,
+      "learning_rate": 0.0002469135802469136,
+      "loss": 0.9196,
+      "step": 350
+    },
+    {
+      "epoch": 0.778702163061564,
+      "grad_norm": 0.22999270260334015,
+      "learning_rate": 0.0002466666666666666,
+      "loss": 0.5882,
+      "step": 351
+    },
+    {
+      "epoch": 0.7809206877426511,
+      "grad_norm": 0.29254069924354553,
+      "learning_rate": 0.0002464197530864197,
+      "loss": 0.7582,
+      "step": 352
+    },
+    {
+      "epoch": 0.7831392124237382,
+      "grad_norm": 0.3017931282520294,
+      "learning_rate": 0.0002461728395061728,
+      "loss": 0.7712,
+      "step": 353
+    },
+    {
+      "epoch": 0.7853577371048253,
+      "grad_norm": 0.27810150384902954,
+      "learning_rate": 0.0002459259259259259,
+      "loss": 0.7609,
+      "step": 354
+    },
+    {
+      "epoch": 0.7875762617859123,
+      "grad_norm": 0.35130828619003296,
+      "learning_rate": 0.000245679012345679,
+      "loss": 0.9311,
+      "step": 355
+    },
+    {
+      "epoch": 0.7897947864669994,
+      "grad_norm": 0.25007250905036926,
+      "learning_rate": 0.00024543209876543205,
+      "loss": 0.8541,
+      "step": 356
+    },
+    {
+      "epoch": 0.7920133111480865,
+      "grad_norm": 0.31247836351394653,
+      "learning_rate": 0.0002451851851851852,
+      "loss": 0.7736,
+      "step": 357
+    },
+    {
+      "epoch": 0.7942318358291736,
+      "grad_norm": 0.26175275444984436,
+      "learning_rate": 0.00024493827160493824,
+      "loss": 0.717,
+      "step": 358
+    },
+    {
+      "epoch": 0.7964503605102606,
+      "grad_norm": 0.2646247446537018,
+      "learning_rate": 0.00024469135802469133,
+      "loss": 0.6971,
+      "step": 359
+    },
+    {
+      "epoch": 0.7986688851913477,
+      "grad_norm": 0.3151836693286896,
+      "learning_rate": 0.00024444444444444443,
+      "loss": 0.8856,
+      "step": 360
+    },
+    {
+      "epoch": 0.8008874098724348,
+      "grad_norm": 0.26973995566368103,
+      "learning_rate": 0.0002441975308641975,
+      "loss": 0.7077,
+      "step": 361
+    },
+    {
+      "epoch": 0.8031059345535219,
+      "grad_norm": 0.32068637013435364,
+      "learning_rate": 0.00024395061728395062,
+      "loss": 0.6945,
+      "step": 362
+    },
+    {
+      "epoch": 0.8053244592346089,
+      "grad_norm": 0.231130450963974,
+      "learning_rate": 0.0002437037037037037,
+      "loss": 0.7201,
+      "step": 363
+    },
+    {
+      "epoch": 0.807542983915696,
+      "grad_norm": 0.287585973739624,
+      "learning_rate": 0.00024345679012345678,
+      "loss": 0.939,
+      "step": 364
+    },
+    {
+      "epoch": 0.8097615085967831,
+      "grad_norm": 0.29356417059898376,
+      "learning_rate": 0.00024320987654320985,
+      "loss": 0.9075,
+      "step": 365
+    },
+    {
+      "epoch": 0.8119800332778702,
+      "grad_norm": 0.26167550683021545,
+      "learning_rate": 0.00024296296296296295,
+      "loss": 0.6157,
+      "step": 366
+    },
+    {
+      "epoch": 0.8141985579589573,
+      "grad_norm": 0.2740039825439453,
+      "learning_rate": 0.00024271604938271601,
+      "loss": 0.9434,
+      "step": 367
+    },
+    {
+      "epoch": 0.8164170826400444,
+      "grad_norm": 0.26466289162635803,
+      "learning_rate": 0.0002424691358024691,
+      "loss": 0.7147,
+      "step": 368
+    },
+    {
+      "epoch": 0.8186356073211315,
+      "grad_norm": 0.3322036862373352,
+      "learning_rate": 0.0002422222222222222,
+      "loss": 0.8622,
+      "step": 369
+    },
+    {
+      "epoch": 0.8208541320022186,
+      "grad_norm": 0.24934785068035126,
+      "learning_rate": 0.00024197530864197527,
+      "loss": 0.8206,
+      "step": 370
+    },
+    {
+      "epoch": 0.8230726566833056,
+      "grad_norm": 0.2569073438644409,
+      "learning_rate": 0.0002417283950617284,
+      "loss": 0.6548,
+      "step": 371
+    },
+    {
+      "epoch": 0.8252911813643927,
+      "grad_norm": 0.22757619619369507,
+      "learning_rate": 0.00024148148148148146,
+      "loss": 0.7536,
+      "step": 372
+    },
+    {
+      "epoch": 0.8275097060454798,
+      "grad_norm": 0.26505106687545776,
+      "learning_rate": 0.00024123456790123456,
+      "loss": 0.7122,
+      "step": 373
+    },
+    {
+      "epoch": 0.8297282307265669,
+      "grad_norm": 0.3088465929031372,
+      "learning_rate": 0.00024098765432098763,
+      "loss": 0.7017,
+      "step": 374
+    },
+    {
+      "epoch": 0.831946755407654,
+      "grad_norm": 0.20274297893047333,
+      "learning_rate": 0.00024074074074074072,
+      "loss": 0.6539,
+      "step": 375
+    },
+    {
+      "epoch": 0.834165280088741,
+      "grad_norm": 0.22314627468585968,
+      "learning_rate": 0.0002404938271604938,
+      "loss": 0.8664,
+      "step": 376
+    },
+    {
+      "epoch": 0.8363838047698281,
+      "grad_norm": 0.2935009002685547,
+      "learning_rate": 0.00024024691358024688,
+      "loss": 0.8335,
+      "step": 377
+    },
+    {
+      "epoch": 0.8386023294509152,
+      "grad_norm": 0.2647722661495209,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 0.9782,
+      "step": 378
+    },
+    {
+      "epoch": 0.8408208541320022,
+      "grad_norm": 0.34191229939460754,
+      "learning_rate": 0.00023975308641975305,
+      "loss": 0.8014,
+      "step": 379
+    },
+    {
+      "epoch": 0.8430393788130893,
+      "grad_norm": 0.279151052236557,
+      "learning_rate": 0.00023950617283950617,
+      "loss": 0.7822,
+      "step": 380
+    },
+    {
+      "epoch": 0.8452579034941764,
+      "grad_norm": 0.3598542809486389,
+      "learning_rate": 0.00023925925925925924,
+      "loss": 0.7188,
+      "step": 381
+    },
+    {
+      "epoch": 0.8474764281752635,
+      "grad_norm": 0.22090934216976166,
+      "learning_rate": 0.00023901234567901233,
+      "loss": 0.6811,
+      "step": 382
+    },
+    {
+      "epoch": 0.8496949528563505,
+      "grad_norm": 0.23945635557174683,
+      "learning_rate": 0.0002387654320987654,
+      "loss": 0.7511,
+      "step": 383
+    },
+    {
+      "epoch": 0.8519134775374376,
+      "grad_norm": 0.33367040753364563,
+      "learning_rate": 0.0002385185185185185,
+      "loss": 0.6064,
+      "step": 384
+    },
+    {
+      "epoch": 0.8541320022185247,
+      "grad_norm": 0.264151930809021,
+      "learning_rate": 0.0002382716049382716,
+      "loss": 0.6963,
+      "step": 385
+    },
+    {
+      "epoch": 0.8563505268996118,
+      "grad_norm": 0.3609398603439331,
+      "learning_rate": 0.00023802469135802466,
+      "loss": 0.7249,
+      "step": 386
+    },
+    {
+      "epoch": 0.8585690515806988,
+      "grad_norm": 0.23930279910564423,
+      "learning_rate": 0.00023777777777777775,
+      "loss": 0.9704,
+      "step": 387
+    },
+    {
+      "epoch": 0.8607875762617859,
+      "grad_norm": 0.23176974058151245,
+      "learning_rate": 0.00023753086419753082,
+      "loss": 0.7245,
+      "step": 388
+    },
+    {
+      "epoch": 0.863006100942873,
+      "grad_norm": 0.33359494805336,
+      "learning_rate": 0.00023728395061728394,
+      "loss": 0.7653,
+      "step": 389
+    },
+    {
+      "epoch": 0.8652246256239601,
+      "grad_norm": 0.25733014941215515,
+      "learning_rate": 0.000237037037037037,
+      "loss": 0.7986,
+      "step": 390
+    },
+    {
+      "epoch": 0.8674431503050472,
+      "grad_norm": 0.29932883381843567,
+      "learning_rate": 0.0002367901234567901,
+      "loss": 0.789,
+      "step": 391
+    },
+    {
+      "epoch": 0.8696616749861342,
+      "grad_norm": 0.22833006083965302,
+      "learning_rate": 0.0002365432098765432,
+      "loss": 0.6754,
+      "step": 392
+    },
+    {
+      "epoch": 0.8718801996672213,
+      "grad_norm": 0.23935233056545258,
+      "learning_rate": 0.00023629629629629627,
+      "loss": 0.6924,
+      "step": 393
+    },
+    {
+      "epoch": 0.8740987243483084,
+      "grad_norm": 0.22784388065338135,
+      "learning_rate": 0.00023604938271604937,
+      "loss": 0.726,
+      "step": 394
+    },
+    {
+      "epoch": 0.8763172490293955,
+      "grad_norm": 0.22663845121860504,
+      "learning_rate": 0.00023580246913580243,
+      "loss": 0.6604,
+      "step": 395
+    },
+    {
+      "epoch": 0.8785357737104825,
+      "grad_norm": 0.22075754404067993,
+      "learning_rate": 0.00023555555555555553,
+      "loss": 0.7085,
+      "step": 396
+    },
+    {
+      "epoch": 0.8807542983915696,
+      "grad_norm": 0.2764483094215393,
+      "learning_rate": 0.0002353086419753086,
+      "loss": 0.8206,
+      "step": 397
+    },
+    {
+      "epoch": 0.8829728230726567,
+      "grad_norm": 0.27558431029319763,
+      "learning_rate": 0.00023506172839506172,
+      "loss": 0.8561,
+      "step": 398
+    },
+    {
+      "epoch": 0.8851913477537438,
+      "grad_norm": 0.3313826322555542,
+      "learning_rate": 0.0002348148148148148,
+      "loss": 0.9318,
+      "step": 399
+    },
+    {
+      "epoch": 0.8874098724348308,
+      "grad_norm": 0.2684048116207123,
+      "learning_rate": 0.00023456790123456788,
+      "loss": 0.7949,
+      "step": 400
+    },
+    {
+      "epoch": 0.8896283971159179,
+      "grad_norm": 0.4025936424732208,
+      "learning_rate": 0.00023432098765432098,
+      "loss": 0.6463,
+      "step": 401
+    },
+    {
+      "epoch": 0.891846921797005,
+      "grad_norm": 0.3521644175052643,
+      "learning_rate": 0.00023407407407407405,
+      "loss": 0.7334,
+      "step": 402
+    },
+    {
+      "epoch": 0.894065446478092,
+      "grad_norm": 0.2839796245098114,
+      "learning_rate": 0.00023382716049382714,
+      "loss": 0.6996,
+      "step": 403
+    },
+    {
+      "epoch": 0.8962839711591791,
+      "grad_norm": 0.2888348698616028,
+      "learning_rate": 0.0002335802469135802,
+      "loss": 0.7423,
+      "step": 404
+    },
+    {
+      "epoch": 0.8985024958402662,
+      "grad_norm": 0.28560760617256165,
+      "learning_rate": 0.0002333333333333333,
+      "loss": 0.7123,
+      "step": 405
+    },
+    {
+      "epoch": 0.9007210205213533,
+      "grad_norm": 0.2975189685821533,
+      "learning_rate": 0.00023308641975308637,
+      "loss": 0.5041,
+      "step": 406
+    },
+    {
+      "epoch": 0.9029395452024404,
+      "grad_norm": 0.3158819079399109,
+      "learning_rate": 0.0002328395061728395,
+      "loss": 0.7351,
+      "step": 407
+    },
+    {
+      "epoch": 0.9051580698835274,
+      "grad_norm": 0.3503880500793457,
+      "learning_rate": 0.0002325925925925926,
+      "loss": 0.973,
+      "step": 408
+    },
+    {
+      "epoch": 0.9073765945646145,
+      "grad_norm": 0.22832582890987396,
+      "learning_rate": 0.00023234567901234566,
+      "loss": 0.849,
+      "step": 409
+    },
+    {
+      "epoch": 0.9095951192457016,
+      "grad_norm": 0.2980108857154846,
+      "learning_rate": 0.00023209876543209875,
+      "loss": 0.8164,
+      "step": 410
+    },
+    {
+      "epoch": 0.9118136439267887,
+      "grad_norm": 0.28835344314575195,
+      "learning_rate": 0.00023185185185185182,
+      "loss": 0.7151,
+      "step": 411
+    },
+    {
+      "epoch": 0.9140321686078757,
+      "grad_norm": 0.2831559479236603,
+      "learning_rate": 0.00023160493827160492,
+      "loss": 0.7392,
+      "step": 412
+    },
+    {
+      "epoch": 0.9162506932889628,
+      "grad_norm": 0.2980341613292694,
+      "learning_rate": 0.00023135802469135798,
+      "loss": 0.6243,
+      "step": 413
+    },
+    {
+      "epoch": 0.9184692179700499,
+      "grad_norm": 0.28722143173217773,
+      "learning_rate": 0.00023111111111111108,
+      "loss": 0.7267,
+      "step": 414
+    },
+    {
+      "epoch": 0.920687742651137,
+      "grad_norm": 0.2971077859401703,
+      "learning_rate": 0.0002308641975308642,
+      "loss": 0.6517,
+      "step": 415
+    },
+    {
+      "epoch": 0.922906267332224,
+      "grad_norm": 0.25104406476020813,
+      "learning_rate": 0.00023061728395061727,
+      "loss": 0.6965,
+      "step": 416
+    },
+    {
+      "epoch": 0.9251247920133111,
+      "grad_norm": 0.3970206379890442,
+      "learning_rate": 0.00023037037037037036,
+      "loss": 0.7446,
+      "step": 417
+    },
+    {
+      "epoch": 0.9273433166943982,
+      "grad_norm": 0.2945402264595032,
+      "learning_rate": 0.00023012345679012343,
+      "loss": 0.8148,
+      "step": 418
+    },
+    {
+      "epoch": 0.9295618413754853,
+      "grad_norm": 0.33522069454193115,
+      "learning_rate": 0.00022987654320987653,
+      "loss": 0.7746,
+      "step": 419
+    },
+    {
+      "epoch": 0.9317803660565723,
+      "grad_norm": 0.29551127552986145,
+      "learning_rate": 0.0002296296296296296,
+      "loss": 0.6618,
+      "step": 420
+    },
+    {
+      "epoch": 0.9339988907376594,
+      "grad_norm": 0.3471081852912903,
+      "learning_rate": 0.0002293827160493827,
+      "loss": 0.7453,
+      "step": 421
+    },
+    {
+      "epoch": 0.9362174154187465,
+      "grad_norm": 0.2280748188495636,
+      "learning_rate": 0.00022913580246913579,
+      "loss": 0.8625,
+      "step": 422
+    },
+    {
+      "epoch": 0.9384359400998337,
+      "grad_norm": 0.2733708322048187,
+      "learning_rate": 0.00022888888888888885,
+      "loss": 0.8806,
+      "step": 423
+    },
+    {
+      "epoch": 0.9406544647809207,
+      "grad_norm": 0.30599093437194824,
+      "learning_rate": 0.00022864197530864198,
+      "loss": 0.7868,
+      "step": 424
+    },
+    {
+      "epoch": 0.9428729894620078,
+      "grad_norm": 0.2738279402256012,
+      "learning_rate": 0.00022839506172839504,
+      "loss": 0.7936,
+      "step": 425
+    },
+    {
+      "epoch": 0.9450915141430949,
+      "grad_norm": 0.3843282163143158,
+      "learning_rate": 0.00022814814814814814,
+      "loss": 0.8529,
+      "step": 426
+    },
+    {
+      "epoch": 0.947310038824182,
+      "grad_norm": 0.2600584328174591,
+      "learning_rate": 0.0002279012345679012,
+      "loss": 0.712,
+      "step": 427
+    },
+    {
+      "epoch": 0.949528563505269,
+      "grad_norm": 0.2883090376853943,
+      "learning_rate": 0.0002276543209876543,
+      "loss": 0.8122,
+      "step": 428
+    },
+    {
+      "epoch": 0.9517470881863561,
+      "grad_norm": 0.2851599454879761,
+      "learning_rate": 0.00022740740740740737,
+      "loss": 0.778,
+      "step": 429
+    },
+    {
+      "epoch": 0.9539656128674432,
+      "grad_norm": 0.24541805684566498,
+      "learning_rate": 0.00022716049382716047,
+      "loss": 0.5467,
+      "step": 430
+    },
+    {
+      "epoch": 0.9561841375485303,
+      "grad_norm": 0.2943759560585022,
+      "learning_rate": 0.00022691358024691356,
+      "loss": 0.7773,
+      "step": 431
+    },
+    {
+      "epoch": 0.9584026622296173,
+      "grad_norm": 0.2780015468597412,
+      "learning_rate": 0.00022666666666666663,
+      "loss": 0.7989,
+      "step": 432
+    },
+    {
+      "epoch": 0.9606211869107044,
+      "grad_norm": 0.2987605929374695,
+      "learning_rate": 0.00022641975308641975,
+      "loss": 0.8306,
+      "step": 433
+    },
+    {
+      "epoch": 0.9628397115917915,
+      "grad_norm": 0.3132157623767853,
+      "learning_rate": 0.00022617283950617282,
+      "loss": 0.9412,
+      "step": 434
+    },
+    {
+      "epoch": 0.9650582362728786,
+      "grad_norm": 0.32859885692596436,
+      "learning_rate": 0.00022592592592592591,
+      "loss": 0.7586,
+      "step": 435
+    },
+    {
+      "epoch": 0.9672767609539656,
+      "grad_norm": 0.22929348051548004,
+      "learning_rate": 0.00022567901234567898,
+      "loss": 0.7207,
+      "step": 436
+    },
+    {
+      "epoch": 0.9694952856350527,
+      "grad_norm": 0.28503215312957764,
+      "learning_rate": 0.00022543209876543208,
+      "loss": 0.6832,
+      "step": 437
+    },
+    {
+      "epoch": 0.9717138103161398,
+      "grad_norm": 0.24173609912395477,
+      "learning_rate": 0.00022518518518518517,
+      "loss": 0.6089,
+      "step": 438
+    },
+    {
+      "epoch": 0.9739323349972269,
+      "grad_norm": 0.23539376258850098,
+      "learning_rate": 0.00022493827160493824,
+      "loss": 0.7174,
+      "step": 439
+    },
+    {
+      "epoch": 0.9761508596783139,
+      "grad_norm": 0.3147657513618469,
+      "learning_rate": 0.00022469135802469134,
+      "loss": 0.9489,
+      "step": 440
+    },
+    {
+      "epoch": 0.978369384359401,
+      "grad_norm": 0.2941228449344635,
+      "learning_rate": 0.0002244444444444444,
+      "loss": 0.7463,
+      "step": 441
+    },
+    {
+      "epoch": 0.9805879090404881,
+      "grad_norm": 0.2961880564689636,
+      "learning_rate": 0.00022419753086419753,
+      "loss": 0.6674,
+      "step": 442
+    },
+    {
+      "epoch": 0.9828064337215752,
+      "grad_norm": 0.30243030190467834,
+      "learning_rate": 0.0002239506172839506,
+      "loss": 0.9345,
+      "step": 443
+    },
+    {
+      "epoch": 0.9850249584026622,
+      "grad_norm": 0.27987581491470337,
+      "learning_rate": 0.0002237037037037037,
+      "loss": 0.693,
+      "step": 444
+    },
+    {
+      "epoch": 0.9872434830837493,
+      "grad_norm": 0.20501844584941864,
+      "learning_rate": 0.00022345679012345678,
+      "loss": 0.6729,
+      "step": 445
+    },
+    {
+      "epoch": 0.9894620077648364,
+      "grad_norm": 0.300935834646225,
+      "learning_rate": 0.00022320987654320985,
+      "loss": 0.8197,
+      "step": 446
+    },
+    {
+      "epoch": 0.9916805324459235,
+      "grad_norm": 0.3053737282752991,
+      "learning_rate": 0.00022296296296296295,
+      "loss": 0.624,
+      "step": 447
+    },
+    {
+      "epoch": 0.9938990571270105,
+      "grad_norm": 0.2512013912200928,
+      "learning_rate": 0.00022271604938271602,
+      "loss": 0.7548,
+      "step": 448
+    },
+    {
+      "epoch": 0.9961175818080976,
+      "grad_norm": 0.33708563446998596,
+      "learning_rate": 0.0002224691358024691,
+      "loss": 0.8454,
+      "step": 449
+    },
+    {
+      "epoch": 0.9983361064891847,
+      "grad_norm": 0.30105510354042053,
+      "learning_rate": 0.00022222222222222218,
+      "loss": 0.7847,
+      "step": 450
+    },
+    {
+      "epoch": 1.0005546311702718,
+      "grad_norm": 0.23681364953517914,
+      "learning_rate": 0.0002219753086419753,
+      "loss": 0.9026,
+      "step": 451
+    },
+    {
+      "epoch": 1.0027731558513588,
+      "grad_norm": 0.26799285411834717,
+      "learning_rate": 0.0002217283950617284,
+      "loss": 0.8412,
+      "step": 452
+    },
+    {
+      "epoch": 1.004991680532446,
+      "grad_norm": 0.3256857991218567,
+      "learning_rate": 0.00022148148148148146,
+      "loss": 0.7372,
+      "step": 453
+    },
+    {
+      "epoch": 1.007210205213533,
+      "grad_norm": 0.2166319042444229,
+      "learning_rate": 0.00022123456790123456,
+      "loss": 0.5987,
+      "step": 454
+    },
+    {
+      "epoch": 1.00942872989462,
+      "grad_norm": 0.25850215554237366,
+      "learning_rate": 0.00022098765432098763,
+      "loss": 0.6166,
+      "step": 455
+    },
+    {
+      "epoch": 1.0116472545757071,
+      "grad_norm": 0.23603804409503937,
+      "learning_rate": 0.00022074074074074072,
+      "loss": 0.7852,
+      "step": 456
+    },
+    {
+      "epoch": 1.0138657792567942,
+      "grad_norm": 0.2977489233016968,
+      "learning_rate": 0.0002204938271604938,
+      "loss": 0.6219,
+      "step": 457
+    },
+    {
+      "epoch": 1.0160843039378813,
+      "grad_norm": 0.25002583861351013,
+      "learning_rate": 0.00022024691358024689,
+      "loss": 0.6019,
+      "step": 458
+    },
+    {
+      "epoch": 1.0183028286189684,
+      "grad_norm": 0.24774114787578583,
+      "learning_rate": 0.00021999999999999995,
+      "loss": 0.6292,
+      "step": 459
+    },
+    {
+      "epoch": 1.0205213533000554,
+      "grad_norm": 0.24106049537658691,
+      "learning_rate": 0.00021975308641975308,
+      "loss": 0.5347,
+      "step": 460
+    },
+    {
+      "epoch": 1.0227398779811425,
+      "grad_norm": 0.3368341326713562,
+      "learning_rate": 0.00021950617283950617,
+      "loss": 0.6476,
+      "step": 461
+    },
+    {
+      "epoch": 1.0249584026622296,
+      "grad_norm": 0.24692381918430328,
+      "learning_rate": 0.00021925925925925924,
+      "loss": 0.7612,
+      "step": 462
+    },
+    {
+      "epoch": 1.0271769273433167,
+      "grad_norm": 0.22617410123348236,
+      "learning_rate": 0.00021901234567901233,
+      "loss": 0.5984,
+      "step": 463
+    },
+    {
+      "epoch": 1.0293954520244037,
+      "grad_norm": 0.26592615246772766,
+      "learning_rate": 0.0002187654320987654,
+      "loss": 0.6399,
+      "step": 464
+    },
+    {
+      "epoch": 1.0316139767054908,
+      "grad_norm": 0.3202682435512543,
+      "learning_rate": 0.0002185185185185185,
+      "loss": 0.7008,
+      "step": 465
+    },
+    {
+      "epoch": 1.033832501386578,
+      "grad_norm": 0.27958106994628906,
+      "learning_rate": 0.00021827160493827157,
+      "loss": 0.7392,
+      "step": 466
+    },
+    {
+      "epoch": 1.036051026067665,
+      "grad_norm": 0.3656889498233795,
+      "learning_rate": 0.00021802469135802466,
+      "loss": 0.7971,
+      "step": 467
+    },
+    {
+      "epoch": 1.038269550748752,
+      "grad_norm": 0.3485015332698822,
+      "learning_rate": 0.00021777777777777778,
+      "loss": 0.8756,
+      "step": 468
+    },
+    {
+      "epoch": 1.0404880754298391,
+      "grad_norm": 0.24719741940498352,
+      "learning_rate": 0.00021753086419753085,
+      "loss": 0.7025,
+      "step": 469
+    },
+    {
+      "epoch": 1.0427066001109262,
+      "grad_norm": 0.3202490508556366,
+      "learning_rate": 0.00021728395061728395,
+      "loss": 0.7247,
+      "step": 470
+    },
+    {
+      "epoch": 1.0449251247920133,
+      "grad_norm": 0.26900777220726013,
+      "learning_rate": 0.00021703703703703701,
+      "loss": 0.6841,
+      "step": 471
+    },
+    {
+      "epoch": 1.0471436494731003,
+      "grad_norm": 0.3140398859977722,
+      "learning_rate": 0.0002167901234567901,
+      "loss": 0.8598,
+      "step": 472
+    },
+    {
+      "epoch": 1.0493621741541874,
+      "grad_norm": 0.26383137702941895,
+      "learning_rate": 0.00021654320987654318,
+      "loss": 0.7143,
+      "step": 473
+    },
+    {
+      "epoch": 1.0515806988352745,
+      "grad_norm": 0.2241767793893814,
+      "learning_rate": 0.00021629629629629627,
+      "loss": 0.7804,
+      "step": 474
+    },
+    {
+      "epoch": 1.0537992235163616,
+      "grad_norm": 0.3187519907951355,
+      "learning_rate": 0.00021604938271604937,
+      "loss": 0.6765,
+      "step": 475
+    },
+    {
+      "epoch": 1.0560177481974486,
+      "grad_norm": 0.3121449053287506,
+      "learning_rate": 0.00021580246913580244,
+      "loss": 0.7452,
+      "step": 476
+    },
+    {
+      "epoch": 1.0582362728785357,
+      "grad_norm": 0.31166744232177734,
+      "learning_rate": 0.00021555555555555556,
+      "loss": 0.4848,
+      "step": 477
+    },
+    {
+      "epoch": 1.0604547975596228,
+      "grad_norm": 0.317636638879776,
+      "learning_rate": 0.00021530864197530863,
+      "loss": 0.5866,
+      "step": 478
+    },
+    {
+      "epoch": 1.0626733222407099,
+      "grad_norm": 0.28568732738494873,
+      "learning_rate": 0.00021506172839506172,
+      "loss": 0.655,
+      "step": 479
+    },
+    {
+      "epoch": 1.064891846921797,
+      "grad_norm": 0.37766191363334656,
+      "learning_rate": 0.0002148148148148148,
+      "loss": 0.9238,
+      "step": 480
+    },
+    {
+      "epoch": 1.067110371602884,
+      "grad_norm": 0.33344554901123047,
+      "learning_rate": 0.00021456790123456788,
+      "loss": 0.6688,
+      "step": 481
+    },
+    {
+      "epoch": 1.069328896283971,
+      "grad_norm": 0.35851937532424927,
+      "learning_rate": 0.00021432098765432095,
+      "loss": 0.7746,
+      "step": 482
+    },
+    {
+      "epoch": 1.0715474209650582,
+      "grad_norm": 0.30358707904815674,
+      "learning_rate": 0.00021407407407407405,
+      "loss": 0.934,
+      "step": 483
+    },
+    {
+      "epoch": 1.0737659456461452,
+      "grad_norm": 0.3073350489139557,
+      "learning_rate": 0.00021382716049382714,
+      "loss": 0.8066,
+      "step": 484
+    },
+    {
+      "epoch": 1.0759844703272323,
+      "grad_norm": 0.2752484679222107,
+      "learning_rate": 0.0002135802469135802,
+      "loss": 0.5699,
+      "step": 485
+    },
+    {
+      "epoch": 1.0782029950083194,
+      "grad_norm": 0.2625827193260193,
+      "learning_rate": 0.00021333333333333333,
+      "loss": 0.7051,
+      "step": 486
+    },
+    {
+      "epoch": 1.0804215196894065,
+      "grad_norm": 0.3486410081386566,
+      "learning_rate": 0.0002130864197530864,
+      "loss": 0.6186,
+      "step": 487
+    },
+    {
+      "epoch": 1.0826400443704935,
+      "grad_norm": 0.2784884572029114,
+      "learning_rate": 0.0002128395061728395,
+      "loss": 0.8161,
+      "step": 488
+    },
+    {
+      "epoch": 1.0848585690515806,
+      "grad_norm": 0.2658778131008148,
+      "learning_rate": 0.00021259259259259256,
+      "loss": 0.9697,
+      "step": 489
+    },
+    {
+      "epoch": 1.0870770937326677,
+      "grad_norm": 0.2540466785430908,
+      "learning_rate": 0.00021234567901234566,
+      "loss": 0.8857,
+      "step": 490
+    },
+    {
+      "epoch": 1.0892956184137548,
+      "grad_norm": 0.3410394489765167,
+      "learning_rate": 0.00021209876543209875,
+      "loss": 0.961,
+      "step": 491
+    },
+    {
+      "epoch": 1.0915141430948418,
+      "grad_norm": 0.2105177640914917,
+      "learning_rate": 0.00021185185185185182,
+      "loss": 0.6747,
+      "step": 492
+    },
+    {
+      "epoch": 1.093732667775929,
+      "grad_norm": 0.2830478549003601,
+      "learning_rate": 0.00021160493827160492,
+      "loss": 0.8379,
+      "step": 493
+    },
+    {
+      "epoch": 1.095951192457016,
+      "grad_norm": 0.3059462606906891,
+      "learning_rate": 0.00021135802469135799,
+      "loss": 0.7522,
+      "step": 494
+    },
+    {
+      "epoch": 1.098169717138103,
+      "grad_norm": 0.26358145475387573,
+      "learning_rate": 0.0002111111111111111,
+      "loss": 0.5583,
+      "step": 495
+    },
+    {
+      "epoch": 1.1003882418191901,
+      "grad_norm": 0.2545821964740753,
+      "learning_rate": 0.00021086419753086418,
+      "loss": 0.8056,
+      "step": 496
+    },
+    {
+      "epoch": 1.1026067665002772,
+      "grad_norm": 0.24549247324466705,
+      "learning_rate": 0.00021061728395061727,
+      "loss": 0.7886,
+      "step": 497
+    },
+    {
+      "epoch": 1.1048252911813643,
+      "grad_norm": 0.2521155774593353,
+      "learning_rate": 0.00021037037037037037,
+      "loss": 0.7036,
+      "step": 498
+    },
+    {
+      "epoch": 1.1070438158624514,
+      "grad_norm": 0.282728910446167,
+      "learning_rate": 0.00021012345679012343,
+      "loss": 0.7285,
+      "step": 499
+    },
+    {
+      "epoch": 1.1092623405435384,
+      "grad_norm": 0.27689075469970703,
+      "learning_rate": 0.00020987654320987653,
+      "loss": 0.7262,
+      "step": 500
+    },
+    {
+      "epoch": 1.1114808652246255,
+      "grad_norm": 0.30944883823394775,
+      "learning_rate": 0.0002096296296296296,
+      "loss": 0.6677,
+      "step": 501
+    },
+    {
+      "epoch": 1.1136993899057126,
+      "grad_norm": 0.2773495316505432,
+      "learning_rate": 0.0002093827160493827,
+      "loss": 0.6741,
+      "step": 502
+    },
+    {
+      "epoch": 1.1159179145867997,
+      "grad_norm": 0.22946980595588684,
+      "learning_rate": 0.00020913580246913576,
+      "loss": 0.7674,
+      "step": 503
+    },
+    {
+      "epoch": 1.1181364392678868,
+      "grad_norm": 0.33020997047424316,
+      "learning_rate": 0.00020888888888888888,
+      "loss": 0.7482,
+      "step": 504
+    },
+    {
+      "epoch": 1.1203549639489738,
+      "grad_norm": 0.24645771086215973,
+      "learning_rate": 0.00020864197530864198,
+      "loss": 0.6933,
+      "step": 505
+    },
+    {
+      "epoch": 1.122573488630061,
+      "grad_norm": 0.28131356835365295,
+      "learning_rate": 0.00020839506172839505,
+      "loss": 0.6656,
+      "step": 506
+    },
+    {
+      "epoch": 1.124792013311148,
+      "grad_norm": 0.2974819242954254,
+      "learning_rate": 0.00020814814814814814,
+      "loss": 0.8097,
+      "step": 507
+    },
+    {
+      "epoch": 1.127010537992235,
+      "grad_norm": 0.3323066830635071,
+      "learning_rate": 0.0002079012345679012,
+      "loss": 0.67,
+      "step": 508
+    },
+    {
+      "epoch": 1.1292290626733221,
+      "grad_norm": 0.23582278192043304,
+      "learning_rate": 0.0002076543209876543,
+      "loss": 0.5869,
+      "step": 509
+    },
+    {
+      "epoch": 1.1314475873544092,
+      "grad_norm": 0.2586055397987366,
+      "learning_rate": 0.00020740740740740737,
+      "loss": 0.7565,
+      "step": 510
+    },
+    {
+      "epoch": 1.1336661120354963,
+      "grad_norm": 0.27491024136543274,
+      "learning_rate": 0.00020716049382716047,
+      "loss": 0.8276,
+      "step": 511
+    },
+    {
+      "epoch": 1.1358846367165834,
+      "grad_norm": 0.31372207403182983,
+      "learning_rate": 0.00020691358024691354,
+      "loss": 0.9363,
+      "step": 512
+    },
+    {
+      "epoch": 1.1381031613976704,
+      "grad_norm": 0.3633849024772644,
+      "learning_rate": 0.00020666666666666666,
+      "loss": 0.8422,
+      "step": 513
+    },
+    {
+      "epoch": 1.1403216860787575,
+      "grad_norm": 0.3108188211917877,
+      "learning_rate": 0.00020641975308641975,
+      "loss": 0.5441,
+      "step": 514
+    },
+    {
+      "epoch": 1.1425402107598448,
+      "grad_norm": 0.40801548957824707,
+      "learning_rate": 0.00020617283950617282,
+      "loss": 0.8289,
+      "step": 515
+    },
+    {
+      "epoch": 1.1447587354409319,
+      "grad_norm": 0.30879348516464233,
+      "learning_rate": 0.00020592592592592592,
+      "loss": 0.8584,
+      "step": 516
+    },
+    {
+      "epoch": 1.146977260122019,
+      "grad_norm": 0.3504166901111603,
+      "learning_rate": 0.00020567901234567898,
+      "loss": 0.7767,
+      "step": 517
+    },
+    {
+      "epoch": 1.149195784803106,
+      "grad_norm": 0.39951008558273315,
+      "learning_rate": 0.00020543209876543208,
+      "loss": 0.7347,
+      "step": 518
+    },
+    {
+      "epoch": 1.151414309484193,
+      "grad_norm": 0.36428582668304443,
+      "learning_rate": 0.00020518518518518515,
+      "loss": 0.8812,
+      "step": 519
+    },
+    {
+      "epoch": 1.1536328341652802,
+      "grad_norm": 0.30070987343788147,
+      "learning_rate": 0.00020493827160493824,
+      "loss": 0.6311,
+      "step": 520
+    },
+    {
+      "epoch": 1.1558513588463672,
+      "grad_norm": 0.22796699404716492,
+      "learning_rate": 0.00020469135802469136,
+      "loss": 0.6902,
+      "step": 521
+    },
+    {
+      "epoch": 1.1580698835274543,
+      "grad_norm": 0.25958704948425293,
+      "learning_rate": 0.00020444444444444443,
+      "loss": 0.6902,
+      "step": 522
+    },
+    {
+      "epoch": 1.1602884082085414,
+      "grad_norm": 0.327414333820343,
+      "learning_rate": 0.00020419753086419753,
+      "loss": 0.6927,
+      "step": 523
+    },
+    {
+      "epoch": 1.1625069328896285,
+      "grad_norm": 0.28822222352027893,
+      "learning_rate": 0.0002039506172839506,
+      "loss": 0.6539,
+      "step": 524
+    },
+    {
+      "epoch": 1.1647254575707155,
+      "grad_norm": 0.35145804286003113,
+      "learning_rate": 0.0002037037037037037,
+      "loss": 0.8647,
+      "step": 525
+    },
+    {
+      "epoch": 1.1669439822518026,
+      "grad_norm": 0.31477612257003784,
+      "learning_rate": 0.00020345679012345676,
+      "loss": 0.7824,
+      "step": 526
+    },
+    {
+      "epoch": 1.1691625069328897,
+      "grad_norm": 0.27636730670928955,
+      "learning_rate": 0.00020320987654320985,
+      "loss": 0.5986,
+      "step": 527
+    },
+    {
+      "epoch": 1.1713810316139768,
+      "grad_norm": 0.29833418130874634,
+      "learning_rate": 0.00020296296296296295,
+      "loss": 0.6693,
+      "step": 528
+    },
+    {
+      "epoch": 1.1735995562950639,
+      "grad_norm": 0.3670545816421509,
+      "learning_rate": 0.00020271604938271602,
+      "loss": 0.8214,
+      "step": 529
+    },
+    {
+      "epoch": 1.175818080976151,
+      "grad_norm": 0.31870633363723755,
+      "learning_rate": 0.00020246913580246914,
+      "loss": 0.6723,
+      "step": 530
+    },
+    {
+      "epoch": 1.178036605657238,
+      "grad_norm": 0.3955514430999756,
+      "learning_rate": 0.0002022222222222222,
+      "loss": 0.7785,
+      "step": 531
+    },
+    {
+      "epoch": 1.180255130338325,
+      "grad_norm": 0.3128824532032013,
+      "learning_rate": 0.0002019753086419753,
+      "loss": 0.7371,
+      "step": 532
+    },
+    {
+      "epoch": 1.1824736550194122,
+      "grad_norm": 0.35826531052589417,
+      "learning_rate": 0.00020172839506172837,
+      "loss": 0.7284,
+      "step": 533
+    },
+    {
+      "epoch": 1.1846921797004992,
+      "grad_norm": 0.30986204743385315,
+      "learning_rate": 0.00020148148148148147,
+      "loss": 0.5608,
+      "step": 534
+    },
+    {
+      "epoch": 1.1869107043815863,
+      "grad_norm": 0.28916820883750916,
+      "learning_rate": 0.00020123456790123456,
+      "loss": 0.713,
+      "step": 535
+    },
+    {
+      "epoch": 1.1891292290626734,
+      "grad_norm": 0.3706223666667938,
+      "learning_rate": 0.00020098765432098763,
+      "loss": 0.6822,
+      "step": 536
+    },
+    {
+      "epoch": 1.1913477537437605,
+      "grad_norm": 0.3180272579193115,
+      "learning_rate": 0.00020074074074074072,
+      "loss": 0.6445,
+      "step": 537
+    },
+    {
+      "epoch": 1.1935662784248475,
+      "grad_norm": 0.3044776916503906,
+      "learning_rate": 0.0002004938271604938,
+      "loss": 0.7295,
+      "step": 538
+    },
+    {
+      "epoch": 1.1957848031059346,
+      "grad_norm": 0.2969304919242859,
+      "learning_rate": 0.00020024691358024691,
+      "loss": 0.6338,
+      "step": 539
+    },
+    {
+      "epoch": 1.1980033277870217,
+      "grad_norm": 0.38770735263824463,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 0.7256,
+      "step": 540
+    },
+    {
+      "epoch": 1.2002218524681088,
+      "grad_norm": 0.2664363980293274,
+      "learning_rate": 0.00019975308641975308,
+      "loss": 0.6751,
+      "step": 541
+    },
+    {
+      "epoch": 1.2024403771491958,
+      "grad_norm": 0.3109024465084076,
+      "learning_rate": 0.00019950617283950615,
+      "loss": 0.6517,
+      "step": 542
+    },
+    {
+      "epoch": 1.204658901830283,
+      "grad_norm": 0.29338330030441284,
+      "learning_rate": 0.00019925925925925924,
+      "loss": 0.7409,
+      "step": 543
+    },
+    {
+      "epoch": 1.20687742651137,
+      "grad_norm": 0.2865051031112671,
+      "learning_rate": 0.00019901234567901234,
+      "loss": 0.6228,
+      "step": 544
+    },
+    {
+      "epoch": 1.209095951192457,
+      "grad_norm": 0.32752272486686707,
+      "learning_rate": 0.0001987654320987654,
+      "loss": 0.6044,
+      "step": 545
+    },
+    {
+      "epoch": 1.2113144758735441,
+      "grad_norm": 0.289858877658844,
+      "learning_rate": 0.0001985185185185185,
+      "loss": 0.5811,
+      "step": 546
+    },
+    {
+      "epoch": 1.2135330005546312,
+      "grad_norm": 0.3049665093421936,
+      "learning_rate": 0.00019827160493827157,
+      "loss": 0.7133,
+      "step": 547
+    },
+    {
+      "epoch": 1.2157515252357183,
+      "grad_norm": 0.4008253216743469,
+      "learning_rate": 0.0001980246913580247,
+      "loss": 0.6523,
+      "step": 548
+    },
+    {
+      "epoch": 1.2179700499168054,
+      "grad_norm": 0.4490613043308258,
+      "learning_rate": 0.00019777777777777776,
+      "loss": 0.856,
+      "step": 549
+    },
+    {
+      "epoch": 1.2201885745978924,
+      "grad_norm": 0.329410582780838,
+      "learning_rate": 0.00019753086419753085,
+      "loss": 0.8333,
+      "step": 550
+    },
+    {
+      "epoch": 1.2224070992789795,
+      "grad_norm": 0.3436029255390167,
+      "learning_rate": 0.00019728395061728395,
+      "loss": 0.783,
+      "step": 551
+    },
+    {
+      "epoch": 1.2246256239600666,
+      "grad_norm": 0.5023097991943359,
+      "learning_rate": 0.00019703703703703702,
+      "loss": 0.8018,
+      "step": 552
+    },
+    {
+      "epoch": 1.2268441486411537,
+      "grad_norm": 0.27854523062705994,
+      "learning_rate": 0.0001967901234567901,
+      "loss": 0.6922,
+      "step": 553
+    },
+    {
+      "epoch": 1.2290626733222407,
+      "grad_norm": 0.27940770983695984,
+      "learning_rate": 0.00019654320987654318,
+      "loss": 0.7299,
+      "step": 554
+    },
+    {
+      "epoch": 1.2312811980033278,
+      "grad_norm": 0.32032856345176697,
+      "learning_rate": 0.00019629629629629627,
+      "loss": 0.8016,
+      "step": 555
+    },
+    {
+      "epoch": 1.2334997226844149,
+      "grad_norm": 0.2848789095878601,
+      "learning_rate": 0.00019604938271604934,
+      "loss": 0.7605,
+      "step": 556
+    },
+    {
+      "epoch": 1.235718247365502,
+      "grad_norm": 0.29719090461730957,
+      "learning_rate": 0.00019580246913580246,
+      "loss": 0.6823,
+      "step": 557
+    },
+    {
+      "epoch": 1.237936772046589,
+      "grad_norm": 0.3122578263282776,
+      "learning_rate": 0.00019555555555555556,
+      "loss": 0.6085,
+      "step": 558
+    },
+    {
+      "epoch": 1.240155296727676,
+      "grad_norm": 0.25741904973983765,
+      "learning_rate": 0.00019530864197530863,
+      "loss": 0.6915,
+      "step": 559
+    },
+    {
+      "epoch": 1.2423738214087632,
+      "grad_norm": 0.28290796279907227,
+      "learning_rate": 0.00019506172839506172,
+      "loss": 0.6973,
+      "step": 560
+    },
+    {
+      "epoch": 1.2445923460898503,
+      "grad_norm": 0.33513230085372925,
+      "learning_rate": 0.0001948148148148148,
+      "loss": 0.6019,
+      "step": 561
+    },
+    {
+      "epoch": 1.2468108707709373,
+      "grad_norm": 0.2983083724975586,
+      "learning_rate": 0.00019456790123456789,
+      "loss": 0.7758,
+      "step": 562
+    },
+    {
+      "epoch": 1.2490293954520244,
+      "grad_norm": 0.29120442271232605,
+      "learning_rate": 0.00019432098765432095,
+      "loss": 0.8466,
+      "step": 563
+    },
+    {
+      "epoch": 1.2512479201331115,
+      "grad_norm": 0.32034432888031006,
+      "learning_rate": 0.00019407407407407405,
+      "loss": 0.7918,
+      "step": 564
+    },
+    {
+      "epoch": 1.2534664448141986,
+      "grad_norm": 0.2632724940776825,
+      "learning_rate": 0.00019382716049382712,
+      "loss": 0.6672,
+      "step": 565
+    },
+    {
+      "epoch": 1.2556849694952856,
+      "grad_norm": 0.30440056324005127,
+      "learning_rate": 0.00019358024691358024,
+      "loss": 0.8573,
+      "step": 566
+    },
+    {
+      "epoch": 1.2579034941763727,
+      "grad_norm": 0.2771689295768738,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 0.7773,
+      "step": 567
+    },
+    {
+      "epoch": 1.2601220188574598,
+      "grad_norm": 0.28331294655799866,
+      "learning_rate": 0.0001930864197530864,
+      "loss": 0.7412,
+      "step": 568
+    },
+    {
+      "epoch": 1.2623405435385469,
+      "grad_norm": 0.23313188552856445,
+      "learning_rate": 0.0001928395061728395,
+      "loss": 0.6447,
+      "step": 569
+    },
+    {
+      "epoch": 1.264559068219634,
+      "grad_norm": 0.271938681602478,
+      "learning_rate": 0.00019259259259259257,
+      "loss": 0.6021,
+      "step": 570
+    },
+    {
+      "epoch": 1.266777592900721,
+      "grad_norm": 0.2709504961967468,
+      "learning_rate": 0.00019234567901234566,
+      "loss": 0.7105,
+      "step": 571
+    },
+    {
+      "epoch": 1.268996117581808,
+      "grad_norm": 0.3593730032444,
+      "learning_rate": 0.00019209876543209873,
+      "loss": 0.8197,
+      "step": 572
+    },
+    {
+      "epoch": 1.2712146422628952,
+      "grad_norm": 0.29406172037124634,
+      "learning_rate": 0.00019185185185185182,
+      "loss": 0.7803,
+      "step": 573
+    },
+    {
+      "epoch": 1.2734331669439822,
+      "grad_norm": 0.2759162485599518,
+      "learning_rate": 0.00019160493827160495,
+      "loss": 0.7159,
+      "step": 574
+    },
+    {
+      "epoch": 1.2756516916250693,
+      "grad_norm": 0.3199984133243561,
+      "learning_rate": 0.00019135802469135801,
+      "loss": 0.8771,
+      "step": 575
+    },
+    {
+      "epoch": 1.2778702163061564,
+      "grad_norm": 0.30731117725372314,
+      "learning_rate": 0.0001911111111111111,
+      "loss": 0.6451,
+      "step": 576
+    },
+    {
+      "epoch": 1.2800887409872435,
+      "grad_norm": 0.3222436308860779,
+      "learning_rate": 0.00019086419753086418,
+      "loss": 0.7161,
+      "step": 577
+    },
+    {
+      "epoch": 1.2823072656683305,
+      "grad_norm": 0.24637678265571594,
+      "learning_rate": 0.00019061728395061727,
+      "loss": 0.7334,
+      "step": 578
+    },
+    {
+      "epoch": 1.2845257903494176,
+      "grad_norm": 0.24518436193466187,
+      "learning_rate": 0.00019037037037037034,
+      "loss": 0.6141,
+      "step": 579
+    },
+    {
+      "epoch": 1.2867443150305047,
+      "grad_norm": 0.42989033460617065,
+      "learning_rate": 0.00019012345679012344,
+      "loss": 0.9633,
+      "step": 580
+    },
+    {
+      "epoch": 1.2889628397115918,
+      "grad_norm": 0.2623230516910553,
+      "learning_rate": 0.00018987654320987653,
+      "loss": 0.872,
+      "step": 581
+    },
+    {
+      "epoch": 1.2911813643926788,
+      "grad_norm": 0.28632521629333496,
+      "learning_rate": 0.0001896296296296296,
+      "loss": 0.6583,
+      "step": 582
+    },
+    {
+      "epoch": 1.293399889073766,
+      "grad_norm": 0.37467634677886963,
+      "learning_rate": 0.00018938271604938272,
+      "loss": 0.5662,
+      "step": 583
+    },
+    {
+      "epoch": 1.295618413754853,
+      "grad_norm": 0.35922643542289734,
+      "learning_rate": 0.0001891358024691358,
+      "loss": 0.7317,
+      "step": 584
+    },
+    {
+      "epoch": 1.29783693843594,
+      "grad_norm": 0.3045429289340973,
+      "learning_rate": 0.00018888888888888888,
+      "loss": 0.8296,
+      "step": 585
+    },
+    {
+      "epoch": 1.3000554631170271,
+      "grad_norm": 0.4242572784423828,
+      "learning_rate": 0.00018864197530864195,
+      "loss": 0.5689,
+      "step": 586
+    },
+    {
+      "epoch": 1.3022739877981142,
+      "grad_norm": 0.3354453146457672,
+      "learning_rate": 0.00018839506172839505,
+      "loss": 0.6179,
+      "step": 587
+    },
+    {
+      "epoch": 1.3044925124792013,
+      "grad_norm": 0.2874317467212677,
+      "learning_rate": 0.00018814814814814814,
+      "loss": 0.7886,
+      "step": 588
+    },
+    {
+      "epoch": 1.3067110371602884,
+      "grad_norm": 0.27691271901130676,
+      "learning_rate": 0.0001879012345679012,
+      "loss": 0.7029,
+      "step": 589
+    },
+    {
+      "epoch": 1.3089295618413754,
+      "grad_norm": 0.34755873680114746,
+      "learning_rate": 0.0001876543209876543,
+      "loss": 0.712,
+      "step": 590
+    },
+    {
+      "epoch": 1.3111480865224625,
+      "grad_norm": 0.23533223569393158,
+      "learning_rate": 0.00018740740740740737,
+      "loss": 0.5523,
+      "step": 591
+    },
+    {
+      "epoch": 1.3133666112035496,
+      "grad_norm": 0.4206790328025818,
+      "learning_rate": 0.0001871604938271605,
+      "loss": 0.8282,
+      "step": 592
+    },
+    {
+      "epoch": 1.3155851358846367,
+      "grad_norm": 0.2466091811656952,
+      "learning_rate": 0.00018691358024691356,
+      "loss": 0.6705,
+      "step": 593
+    },
+    {
+      "epoch": 1.3178036605657237,
+      "grad_norm": 0.26316744089126587,
+      "learning_rate": 0.00018666666666666666,
+      "loss": 0.8808,
+      "step": 594
+    },
+    {
+      "epoch": 1.3200221852468108,
+      "grad_norm": 0.27022528648376465,
+      "learning_rate": 0.00018641975308641973,
+      "loss": 0.9754,
+      "step": 595
+    },
+    {
+      "epoch": 1.3222407099278979,
+      "grad_norm": 0.31514444947242737,
+      "learning_rate": 0.00018617283950617282,
+      "loss": 0.634,
+      "step": 596
+    },
+    {
+      "epoch": 1.324459234608985,
+      "grad_norm": 0.3509426414966583,
+      "learning_rate": 0.00018592592592592592,
+      "loss": 0.5974,
+      "step": 597
+    },
+    {
+      "epoch": 1.326677759290072,
+      "grad_norm": 0.362454891204834,
+      "learning_rate": 0.00018567901234567899,
+      "loss": 0.7271,
+      "step": 598
+    },
+    {
+      "epoch": 1.328896283971159,
+      "grad_norm": 0.31080344319343567,
+      "learning_rate": 0.00018543209876543208,
+      "loss": 0.5027,
+      "step": 599
+    },
+    {
+      "epoch": 1.3311148086522462,
+      "grad_norm": 0.30572381615638733,
+      "learning_rate": 0.00018518518518518515,
+      "loss": 0.7375,
+      "step": 600
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.3284349739551544,
+      "learning_rate": 0.00018493827160493827,
+      "loss": 0.834,
+      "step": 601
+    },
+    {
+      "epoch": 1.3355518580144203,
+      "grad_norm": 0.36081576347351074,
+      "learning_rate": 0.00018469135802469134,
+      "loss": 0.8055,
+      "step": 602
+    },
+    {
+      "epoch": 1.3377703826955074,
+      "grad_norm": 0.2621290981769562,
+      "learning_rate": 0.00018444444444444443,
+      "loss": 0.6568,
+      "step": 603
+    },
+    {
+      "epoch": 1.3399889073765945,
+      "grad_norm": 0.282374769449234,
+      "learning_rate": 0.00018419753086419753,
+      "loss": 0.7634,
+      "step": 604
+    },
+    {
+      "epoch": 1.3422074320576816,
+      "grad_norm": 0.3122667670249939,
+      "learning_rate": 0.0001839506172839506,
+      "loss": 0.7303,
+      "step": 605
+    },
+    {
+      "epoch": 1.3444259567387689,
+      "grad_norm": 0.286293089389801,
+      "learning_rate": 0.0001837037037037037,
+      "loss": 0.8522,
+      "step": 606
+    },
+    {
+      "epoch": 1.346644481419856,
+      "grad_norm": 0.29735255241394043,
+      "learning_rate": 0.00018345679012345676,
+      "loss": 0.8418,
+      "step": 607
+    },
+    {
+      "epoch": 1.348863006100943,
+      "grad_norm": 0.28353914618492126,
+      "learning_rate": 0.00018320987654320986,
+      "loss": 0.7854,
+      "step": 608
+    },
+    {
+      "epoch": 1.35108153078203,
+      "grad_norm": 0.2539277672767639,
+      "learning_rate": 0.00018296296296296292,
+      "loss": 0.8739,
+      "step": 609
+    },
+    {
+      "epoch": 1.3533000554631172,
+      "grad_norm": 0.3356074094772339,
+      "learning_rate": 0.00018271604938271605,
+      "loss": 0.7112,
+      "step": 610
+    },
+    {
+      "epoch": 1.3555185801442042,
+      "grad_norm": 0.281205952167511,
+      "learning_rate": 0.00018246913580246914,
+      "loss": 0.5676,
+      "step": 611
+    },
+    {
+      "epoch": 1.3577371048252913,
+      "grad_norm": 0.30555519461631775,
+      "learning_rate": 0.0001822222222222222,
+      "loss": 0.7007,
+      "step": 612
+    },
+    {
+      "epoch": 1.3599556295063784,
+      "grad_norm": 0.325361043214798,
+      "learning_rate": 0.0001819753086419753,
+      "loss": 0.7068,
+      "step": 613
+    },
+    {
+      "epoch": 1.3621741541874655,
+      "grad_norm": 0.2869901955127716,
+      "learning_rate": 0.00018172839506172837,
+      "loss": 0.6634,
+      "step": 614
+    },
+    {
+      "epoch": 1.3643926788685525,
+      "grad_norm": 0.28965604305267334,
+      "learning_rate": 0.00018148148148148147,
+      "loss": 0.6614,
+      "step": 615
+    },
+    {
+      "epoch": 1.3666112035496396,
+      "grad_norm": 0.25581103563308716,
+      "learning_rate": 0.00018123456790123454,
+      "loss": 0.6221,
+      "step": 616
+    },
+    {
+      "epoch": 1.3688297282307267,
+      "grad_norm": 0.36997684836387634,
+      "learning_rate": 0.00018098765432098763,
+      "loss": 0.8548,
+      "step": 617
+    },
+    {
+      "epoch": 1.3710482529118138,
+      "grad_norm": 0.2562940716743469,
+      "learning_rate": 0.00018074074074074075,
+      "loss": 0.7298,
+      "step": 618
+    },
+    {
+      "epoch": 1.3732667775929008,
+      "grad_norm": 0.341651052236557,
+      "learning_rate": 0.00018049382716049382,
+      "loss": 0.6944,
+      "step": 619
+    },
+    {
+      "epoch": 1.375485302273988,
+      "grad_norm": 0.3417910635471344,
+      "learning_rate": 0.00018024691358024692,
+      "loss": 0.6838,
+      "step": 620
+    },
+    {
+      "epoch": 1.377703826955075,
+      "grad_norm": 0.2916937470436096,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.8368,
+      "step": 621
+    },
+    {
+      "epoch": 1.379922351636162,
+      "grad_norm": 0.26102912425994873,
+      "learning_rate": 0.00017975308641975308,
+      "loss": 0.5643,
+      "step": 622
+    },
+    {
+      "epoch": 1.3821408763172491,
+      "grad_norm": 0.36503463983535767,
+      "learning_rate": 0.00017950617283950615,
+      "loss": 0.8027,
+      "step": 623
+    },
+    {
+      "epoch": 1.3843594009983362,
+      "grad_norm": 0.3135330080986023,
+      "learning_rate": 0.00017925925925925924,
+      "loss": 0.7236,
+      "step": 624
+    },
+    {
+      "epoch": 1.3865779256794233,
+      "grad_norm": 0.3273005485534668,
+      "learning_rate": 0.0001790123456790123,
+      "loss": 0.6809,
+      "step": 625
+    },
+    {
+      "epoch": 1.3887964503605104,
+      "grad_norm": 0.2877342402935028,
+      "learning_rate": 0.0001787654320987654,
+      "loss": 0.523,
+      "step": 626
+    },
+    {
+      "epoch": 1.3910149750415974,
+      "grad_norm": 0.26204854249954224,
+      "learning_rate": 0.00017851851851851853,
+      "loss": 0.6713,
+      "step": 627
+    },
+    {
+      "epoch": 1.3932334997226845,
+      "grad_norm": 0.30797597765922546,
+      "learning_rate": 0.0001782716049382716,
+      "loss": 0.5729,
+      "step": 628
+    },
+    {
+      "epoch": 1.3954520244037716,
+      "grad_norm": 0.3564557135105133,
+      "learning_rate": 0.0001780246913580247,
+      "loss": 0.7331,
+      "step": 629
+    },
+    {
+      "epoch": 1.3976705490848587,
+      "grad_norm": 0.3141915500164032,
+      "learning_rate": 0.00017777777777777776,
+      "loss": 0.6586,
+      "step": 630
+    },
+    {
+      "epoch": 1.3998890737659457,
+      "grad_norm": 0.37379834055900574,
+      "learning_rate": 0.00017753086419753085,
+      "loss": 0.6656,
+      "step": 631
+    },
+    {
+      "epoch": 1.4021075984470328,
+      "grad_norm": 0.26058465242385864,
+      "learning_rate": 0.00017728395061728392,
+      "loss": 0.5826,
+      "step": 632
+    },
+    {
+      "epoch": 1.4043261231281199,
+      "grad_norm": 0.2606399953365326,
+      "learning_rate": 0.00017703703703703702,
+      "loss": 0.7191,
+      "step": 633
+    },
+    {
+      "epoch": 1.406544647809207,
+      "grad_norm": 0.40708959102630615,
+      "learning_rate": 0.0001767901234567901,
+      "loss": 0.6843,
+      "step": 634
+    },
+    {
+      "epoch": 1.408763172490294,
+      "grad_norm": 0.40571117401123047,
+      "learning_rate": 0.00017654320987654318,
+      "loss": 0.6929,
+      "step": 635
+    },
+    {
+      "epoch": 1.410981697171381,
+      "grad_norm": 0.3446694016456604,
+      "learning_rate": 0.0001762962962962963,
+      "loss": 0.7104,
+      "step": 636
+    },
+    {
+      "epoch": 1.4132002218524682,
+      "grad_norm": 0.2863425016403198,
+      "learning_rate": 0.00017604938271604937,
+      "loss": 0.6228,
+      "step": 637
+    },
+    {
+      "epoch": 1.4154187465335553,
+      "grad_norm": 0.30225810408592224,
+      "learning_rate": 0.00017580246913580247,
+      "loss": 0.7814,
+      "step": 638
+    },
+    {
+      "epoch": 1.4176372712146423,
+      "grad_norm": 0.3592014014720917,
+      "learning_rate": 0.00017555555555555553,
+      "loss": 0.7714,
+      "step": 639
+    },
+    {
+      "epoch": 1.4198557958957294,
+      "grad_norm": 0.35787147283554077,
+      "learning_rate": 0.00017530864197530863,
+      "loss": 0.6857,
+      "step": 640
+    },
+    {
+      "epoch": 1.4220743205768165,
+      "grad_norm": 0.3450486361980438,
+      "learning_rate": 0.00017506172839506172,
+      "loss": 0.7215,
+      "step": 641
+    },
+    {
+      "epoch": 1.4242928452579036,
+      "grad_norm": 0.31643423438072205,
+      "learning_rate": 0.0001748148148148148,
+      "loss": 0.6498,
+      "step": 642
+    },
+    {
+      "epoch": 1.4265113699389906,
+      "grad_norm": 0.353584885597229,
+      "learning_rate": 0.0001745679012345679,
+      "loss": 0.5313,
+      "step": 643
+    },
+    {
+      "epoch": 1.4287298946200777,
+      "grad_norm": 0.3707257807254791,
+      "learning_rate": 0.00017432098765432096,
+      "loss": 0.8282,
+      "step": 644
+    },
+    {
+      "epoch": 1.4309484193011648,
+      "grad_norm": 0.2765117287635803,
+      "learning_rate": 0.00017407407407407408,
+      "loss": 0.6606,
+      "step": 645
+    },
+    {
+      "epoch": 1.4331669439822519,
+      "grad_norm": 0.3104737401008606,
+      "learning_rate": 0.00017382716049382715,
+      "loss": 0.6666,
+      "step": 646
+    },
+    {
+      "epoch": 1.435385468663339,
+      "grad_norm": 0.30970507860183716,
+      "learning_rate": 0.00017358024691358024,
+      "loss": 0.6879,
+      "step": 647
+    },
+    {
+      "epoch": 1.437603993344426,
+      "grad_norm": 0.3165368139743805,
+      "learning_rate": 0.0001733333333333333,
+      "loss": 0.5879,
+      "step": 648
+    },
+    {
+      "epoch": 1.439822518025513,
+      "grad_norm": 0.23878324031829834,
+      "learning_rate": 0.0001730864197530864,
+      "loss": 0.7203,
+      "step": 649
+    },
+    {
+      "epoch": 1.4420410427066002,
+      "grad_norm": 0.37136614322662354,
+      "learning_rate": 0.0001728395061728395,
+      "loss": 0.7047,
+      "step": 650
+    },
+    {
+      "epoch": 1.4442595673876872,
+      "grad_norm": 0.36999979615211487,
+      "learning_rate": 0.00017259259259259257,
+      "loss": 0.747,
+      "step": 651
+    },
+    {
+      "epoch": 1.4464780920687743,
+      "grad_norm": 0.43379828333854675,
+      "learning_rate": 0.00017234567901234566,
+      "loss": 0.733,
+      "step": 652
+    },
+    {
+      "epoch": 1.4486966167498614,
+      "grad_norm": 0.3629532754421234,
+      "learning_rate": 0.00017209876543209873,
+      "loss": 0.5475,
+      "step": 653
+    },
+    {
+      "epoch": 1.4509151414309485,
+      "grad_norm": 0.3684217035770416,
+      "learning_rate": 0.00017185185185185185,
+      "loss": 0.6431,
+      "step": 654
+    },
+    {
+      "epoch": 1.4531336661120355,
+      "grad_norm": 0.29095783829689026,
+      "learning_rate": 0.00017160493827160492,
+      "loss": 0.7291,
+      "step": 655
+    },
+    {
+      "epoch": 1.4553521907931226,
+      "grad_norm": 0.3772960603237152,
+      "learning_rate": 0.00017135802469135802,
+      "loss": 0.6869,
+      "step": 656
+    },
+    {
+      "epoch": 1.4575707154742097,
+      "grad_norm": 0.3238474130630493,
+      "learning_rate": 0.0001711111111111111,
+      "loss": 0.6033,
+      "step": 657
+    },
+    {
+      "epoch": 1.4597892401552968,
+      "grad_norm": 0.2952176630496979,
+      "learning_rate": 0.00017086419753086418,
+      "loss": 0.7152,
+      "step": 658
+    },
+    {
+      "epoch": 1.4620077648363838,
+      "grad_norm": 0.3451940417289734,
+      "learning_rate": 0.00017061728395061727,
+      "loss": 0.94,
+      "step": 659
+    },
+    {
+      "epoch": 1.464226289517471,
+      "grad_norm": 0.36361977458000183,
+      "learning_rate": 0.00017037037037037034,
+      "loss": 0.5669,
+      "step": 660
+    },
+    {
+      "epoch": 1.466444814198558,
+      "grad_norm": 0.3316241204738617,
+      "learning_rate": 0.00017012345679012344,
+      "loss": 0.9412,
+      "step": 661
+    },
+    {
+      "epoch": 1.468663338879645,
+      "grad_norm": 0.3977113962173462,
+      "learning_rate": 0.0001698765432098765,
+      "loss": 0.8137,
+      "step": 662
+    },
+    {
+      "epoch": 1.4708818635607321,
+      "grad_norm": 0.2916874289512634,
+      "learning_rate": 0.00016962962962962963,
+      "loss": 0.6707,
+      "step": 663
+    },
+    {
+      "epoch": 1.4731003882418192,
+      "grad_norm": 0.39612865447998047,
+      "learning_rate": 0.00016938271604938272,
+      "loss": 0.789,
+      "step": 664
+    },
+    {
+      "epoch": 1.4753189129229063,
+      "grad_norm": 0.28594380617141724,
+      "learning_rate": 0.0001691358024691358,
+      "loss": 0.7968,
+      "step": 665
+    },
+    {
+      "epoch": 1.4775374376039934,
+      "grad_norm": 0.2873164117336273,
+      "learning_rate": 0.00016888888888888889,
+      "loss": 0.9177,
+      "step": 666
+    },
+    {
+      "epoch": 1.4797559622850804,
+      "grad_norm": 0.2845945954322815,
+      "learning_rate": 0.00016864197530864195,
+      "loss": 0.7042,
+      "step": 667
+    },
+    {
+      "epoch": 1.4819744869661675,
+      "grad_norm": 0.38443177938461304,
+      "learning_rate": 0.00016839506172839505,
+      "loss": 0.625,
+      "step": 668
+    },
+    {
+      "epoch": 1.4841930116472546,
+      "grad_norm": 0.32710710167884827,
+      "learning_rate": 0.00016814814814814812,
+      "loss": 0.7697,
+      "step": 669
+    },
+    {
+      "epoch": 1.4864115363283417,
+      "grad_norm": 0.31466788053512573,
+      "learning_rate": 0.0001679012345679012,
+      "loss": 0.7095,
+      "step": 670
+    },
+    {
+      "epoch": 1.4886300610094287,
+      "grad_norm": 0.2961171865463257,
+      "learning_rate": 0.00016765432098765433,
+      "loss": 0.6731,
+      "step": 671
+    },
+    {
+      "epoch": 1.4908485856905158,
+      "grad_norm": 0.3103848695755005,
+      "learning_rate": 0.0001674074074074074,
+      "loss": 0.6868,
+      "step": 672
+    },
+    {
+      "epoch": 1.493067110371603,
+      "grad_norm": 0.2996843457221985,
+      "learning_rate": 0.0001671604938271605,
+      "loss": 0.6617,
+      "step": 673
+    },
+    {
+      "epoch": 1.49528563505269,
+      "grad_norm": 0.3195325434207916,
+      "learning_rate": 0.00016691358024691357,
+      "loss": 0.7582,
+      "step": 674
+    },
+    {
+      "epoch": 1.497504159733777,
+      "grad_norm": 0.2709215581417084,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 0.5862,
+      "step": 675
+    },
+    {
+      "epoch": 1.4997226844148641,
+      "grad_norm": 0.30099624395370483,
+      "learning_rate": 0.00016641975308641973,
+      "loss": 0.6575,
+      "step": 676
+    },
+    {
+      "epoch": 1.5019412090959512,
+      "grad_norm": 0.30555078387260437,
+      "learning_rate": 0.00016617283950617282,
+      "loss": 0.8656,
+      "step": 677
+    },
+    {
+      "epoch": 1.5041597337770383,
+      "grad_norm": 0.38314560055732727,
+      "learning_rate": 0.0001659259259259259,
+      "loss": 0.5369,
+      "step": 678
+    },
+    {
+      "epoch": 1.5063782584581253,
+      "grad_norm": 0.2737380266189575,
+      "learning_rate": 0.000165679012345679,
+      "loss": 0.7028,
+      "step": 679
+    },
+    {
+      "epoch": 1.5085967831392124,
+      "grad_norm": 0.30935999751091003,
+      "learning_rate": 0.0001654320987654321,
+      "loss": 0.6856,
+      "step": 680
+    },
+    {
+      "epoch": 1.5108153078202995,
+      "grad_norm": 0.3298351764678955,
+      "learning_rate": 0.00016518518518518518,
+      "loss": 0.8217,
+      "step": 681
+    },
+    {
+      "epoch": 1.5130338325013866,
+      "grad_norm": 0.5368160009384155,
+      "learning_rate": 0.00016493827160493827,
+      "loss": 0.9832,
+      "step": 682
+    },
+    {
+      "epoch": 1.5152523571824736,
+      "grad_norm": 0.31119877099990845,
+      "learning_rate": 0.00016469135802469134,
+      "loss": 0.7196,
+      "step": 683
+    },
+    {
+      "epoch": 1.5174708818635607,
+      "grad_norm": 0.3705015182495117,
+      "learning_rate": 0.00016444444444444444,
+      "loss": 0.7252,
+      "step": 684
+    },
+    {
+      "epoch": 1.5196894065446478,
+      "grad_norm": 0.29311779141426086,
+      "learning_rate": 0.0001641975308641975,
+      "loss": 0.7487,
+      "step": 685
+    },
+    {
+      "epoch": 1.5219079312257349,
+      "grad_norm": 0.32288309931755066,
+      "learning_rate": 0.0001639506172839506,
+      "loss": 0.6046,
+      "step": 686
+    },
+    {
+      "epoch": 1.524126455906822,
+      "grad_norm": 0.4755495488643646,
+      "learning_rate": 0.0001637037037037037,
+      "loss": 0.7962,
+      "step": 687
+    },
+    {
+      "epoch": 1.526344980587909,
+      "grad_norm": 0.23928476870059967,
+      "learning_rate": 0.00016345679012345676,
+      "loss": 0.8337,
+      "step": 688
+    },
+    {
+      "epoch": 1.528563505268996,
+      "grad_norm": 0.2827261686325073,
+      "learning_rate": 0.00016320987654320988,
+      "loss": 0.6495,
+      "step": 689
+    },
+    {
+      "epoch": 1.5307820299500832,
+      "grad_norm": 0.2969893217086792,
+      "learning_rate": 0.00016296296296296295,
+      "loss": 0.6393,
+      "step": 690
+    },
+    {
+      "epoch": 1.5330005546311702,
+      "grad_norm": 0.28303152322769165,
+      "learning_rate": 0.00016271604938271605,
+      "loss": 0.7185,
+      "step": 691
+    },
+    {
+      "epoch": 1.5352190793122573,
+      "grad_norm": 0.33070868253707886,
+      "learning_rate": 0.00016246913580246912,
+      "loss": 0.8409,
+      "step": 692
+    },
+    {
+      "epoch": 1.5374376039933444,
+      "grad_norm": 0.3248465657234192,
+      "learning_rate": 0.0001622222222222222,
+      "loss": 0.7806,
+      "step": 693
+    },
+    {
+      "epoch": 1.5396561286744315,
+      "grad_norm": 0.3294716775417328,
+      "learning_rate": 0.0001619753086419753,
+      "loss": 0.7606,
+      "step": 694
+    },
+    {
+      "epoch": 1.5418746533555185,
+      "grad_norm": 0.3006589412689209,
+      "learning_rate": 0.00016172839506172837,
+      "loss": 0.6797,
+      "step": 695
+    },
+    {
+      "epoch": 1.5440931780366056,
+      "grad_norm": 0.304825097322464,
+      "learning_rate": 0.00016148148148148147,
+      "loss": 0.7598,
+      "step": 696
+    },
+    {
+      "epoch": 1.5463117027176927,
+      "grad_norm": 0.33681073784828186,
+      "learning_rate": 0.00016123456790123454,
+      "loss": 0.6916,
+      "step": 697
+    },
+    {
+      "epoch": 1.5485302273987798,
+      "grad_norm": 0.31246286630630493,
+      "learning_rate": 0.00016098765432098766,
+      "loss": 0.6751,
+      "step": 698
+    },
+    {
+      "epoch": 1.5507487520798668,
+      "grad_norm": 0.35856175422668457,
+      "learning_rate": 0.00016074074074074073,
+      "loss": 0.774,
+      "step": 699
+    },
+    {
+      "epoch": 1.552967276760954,
+      "grad_norm": 0.37167084217071533,
+      "learning_rate": 0.00016049382716049382,
+      "loss": 0.6346,
+      "step": 700
+    },
+    {
+      "epoch": 1.555185801442041,
+      "grad_norm": 0.3114137351512909,
+      "learning_rate": 0.00016024691358024692,
+      "loss": 0.8808,
+      "step": 701
+    },
+    {
+      "epoch": 1.557404326123128,
+      "grad_norm": 0.3571993112564087,
+      "learning_rate": 0.00015999999999999999,
+      "loss": 0.7832,
+      "step": 702
+    },
+    {
+      "epoch": 1.5596228508042151,
+      "grad_norm": 0.3589223623275757,
+      "learning_rate": 0.00015975308641975308,
+      "loss": 1.0599,
+      "step": 703
+    },
+    {
+      "epoch": 1.5618413754853022,
+      "grad_norm": 0.2767994999885559,
+      "learning_rate": 0.00015950617283950615,
+      "loss": 0.7467,
+      "step": 704
+    },
+    {
+      "epoch": 1.5640599001663893,
+      "grad_norm": 0.4535686671733856,
+      "learning_rate": 0.00015925925925925924,
+      "loss": 0.7503,
+      "step": 705
+    },
+    {
+      "epoch": 1.5662784248474764,
+      "grad_norm": 0.3531101942062378,
+      "learning_rate": 0.0001590123456790123,
+      "loss": 0.6746,
+      "step": 706
+    },
+    {
+      "epoch": 1.5684969495285634,
+      "grad_norm": 0.32757627964019775,
+      "learning_rate": 0.00015876543209876543,
+      "loss": 0.6458,
+      "step": 707
+    },
+    {
+      "epoch": 1.5707154742096505,
+      "grad_norm": 0.26822876930236816,
+      "learning_rate": 0.0001585185185185185,
+      "loss": 0.5542,
+      "step": 708
+    },
+    {
+      "epoch": 1.5729339988907376,
+      "grad_norm": 0.28986552357673645,
+      "learning_rate": 0.0001582716049382716,
+      "loss": 0.7384,
+      "step": 709
+    },
+    {
+      "epoch": 1.5751525235718247,
+      "grad_norm": 0.25871625542640686,
+      "learning_rate": 0.0001580246913580247,
+      "loss": 0.6049,
+      "step": 710
+    },
+    {
+      "epoch": 1.5773710482529117,
+      "grad_norm": 0.3513307571411133,
+      "learning_rate": 0.00015777777777777776,
+      "loss": 1.0653,
+      "step": 711
+    },
+    {
+      "epoch": 1.5795895729339988,
+      "grad_norm": 0.3122536242008209,
+      "learning_rate": 0.00015753086419753086,
+      "loss": 0.6652,
+      "step": 712
+    },
+    {
+      "epoch": 1.581808097615086,
+      "grad_norm": 0.31977444887161255,
+      "learning_rate": 0.00015728395061728392,
+      "loss": 0.5124,
+      "step": 713
+    },
+    {
+      "epoch": 1.584026622296173,
+      "grad_norm": 0.33530670404434204,
+      "learning_rate": 0.00015703703703703702,
+      "loss": 0.8315,
+      "step": 714
+    },
+    {
+      "epoch": 1.58624514697726,
+      "grad_norm": 0.28605493903160095,
+      "learning_rate": 0.0001567901234567901,
+      "loss": 0.6409,
+      "step": 715
+    },
+    {
+      "epoch": 1.5884636716583471,
+      "grad_norm": 0.2935214042663574,
+      "learning_rate": 0.0001565432098765432,
+      "loss": 0.8398,
+      "step": 716
+    },
+    {
+      "epoch": 1.5906821963394342,
+      "grad_norm": 0.27571848034858704,
+      "learning_rate": 0.0001562962962962963,
+      "loss": 0.6377,
+      "step": 717
+    },
+    {
+      "epoch": 1.5929007210205213,
+      "grad_norm": 0.4706502854824066,
+      "learning_rate": 0.00015604938271604937,
+      "loss": 0.7129,
+      "step": 718
+    },
+    {
+      "epoch": 1.5951192457016083,
+      "grad_norm": 0.3261232078075409,
+      "learning_rate": 0.00015580246913580247,
+      "loss": 0.7107,
+      "step": 719
+    },
+    {
+      "epoch": 1.5973377703826954,
+      "grad_norm": 0.37362024188041687,
+      "learning_rate": 0.00015555555555555554,
+      "loss": 0.8066,
+      "step": 720
+    },
+    {
+      "epoch": 1.5995562950637825,
+      "grad_norm": 0.28043627738952637,
+      "learning_rate": 0.00015530864197530863,
+      "loss": 0.7658,
+      "step": 721
+    },
+    {
+      "epoch": 1.6017748197448696,
+      "grad_norm": 0.30754584074020386,
+      "learning_rate": 0.0001550617283950617,
+      "loss": 0.7083,
+      "step": 722
+    },
+    {
+      "epoch": 1.6039933444259566,
+      "grad_norm": 0.4023762345314026,
+      "learning_rate": 0.0001548148148148148,
+      "loss": 0.5566,
+      "step": 723
+    },
+    {
+      "epoch": 1.6062118691070437,
+      "grad_norm": 0.2344517558813095,
+      "learning_rate": 0.00015456790123456792,
+      "loss": 0.8716,
+      "step": 724
+    },
+    {
+      "epoch": 1.6084303937881308,
+      "grad_norm": 0.2978951334953308,
+      "learning_rate": 0.00015432098765432098,
+      "loss": 0.6531,
+      "step": 725
+    },
+    {
+      "epoch": 1.6106489184692179,
+      "grad_norm": 0.42656248807907104,
+      "learning_rate": 0.00015407407407407408,
+      "loss": 0.7658,
+      "step": 726
+    },
+    {
+      "epoch": 1.612867443150305,
+      "grad_norm": 0.3671417534351349,
+      "learning_rate": 0.00015382716049382715,
+      "loss": 0.7286,
+      "step": 727
+    },
+    {
+      "epoch": 1.615085967831392,
+      "grad_norm": 0.37693849205970764,
+      "learning_rate": 0.00015358024691358024,
+      "loss": 0.8357,
+      "step": 728
+    },
+    {
+      "epoch": 1.617304492512479,
+      "grad_norm": 0.31157180666923523,
+      "learning_rate": 0.0001533333333333333,
+      "loss": 0.7669,
+      "step": 729
+    },
+    {
+      "epoch": 1.6195230171935662,
+      "grad_norm": 0.30446523427963257,
+      "learning_rate": 0.0001530864197530864,
+      "loss": 0.7207,
+      "step": 730
+    },
+    {
+      "epoch": 1.6217415418746532,
+      "grad_norm": 0.28724828362464905,
+      "learning_rate": 0.00015283950617283947,
+      "loss": 0.679,
+      "step": 731
+    },
+    {
+      "epoch": 1.6239600665557403,
+      "grad_norm": 0.3182579278945923,
+      "learning_rate": 0.00015259259259259257,
+      "loss": 0.79,
+      "step": 732
+    },
+    {
+      "epoch": 1.6261785912368274,
+      "grad_norm": 0.2652699053287506,
+      "learning_rate": 0.0001523456790123457,
+      "loss": 0.5726,
+      "step": 733
+    },
+    {
+      "epoch": 1.6283971159179145,
+      "grad_norm": 0.33146557211875916,
+      "learning_rate": 0.00015209876543209873,
+      "loss": 0.649,
+      "step": 734
+    },
+    {
+      "epoch": 1.6306156405990015,
+      "grad_norm": 0.2710660696029663,
+      "learning_rate": 0.00015185185185185185,
+      "loss": 0.7221,
+      "step": 735
+    },
+    {
+      "epoch": 1.6328341652800886,
+      "grad_norm": 0.2996942102909088,
+      "learning_rate": 0.00015160493827160492,
+      "loss": 0.6604,
+      "step": 736
+    },
+    {
+      "epoch": 1.6350526899611757,
+      "grad_norm": 0.28071293234825134,
+      "learning_rate": 0.00015135802469135802,
+      "loss": 0.6442,
+      "step": 737
+    },
+    {
+      "epoch": 1.6372712146422628,
+      "grad_norm": 0.3119325637817383,
+      "learning_rate": 0.00015111111111111109,
+      "loss": 0.6362,
+      "step": 738
+    },
+    {
+      "epoch": 1.6394897393233498,
+      "grad_norm": 0.26997172832489014,
+      "learning_rate": 0.00015086419753086418,
+      "loss": 0.7388,
+      "step": 739
+    },
+    {
+      "epoch": 1.641708264004437,
+      "grad_norm": 0.47968146204948425,
+      "learning_rate": 0.00015061728395061728,
+      "loss": 0.7521,
+      "step": 740
+    },
+    {
+      "epoch": 1.643926788685524,
+      "grad_norm": 0.39457523822784424,
+      "learning_rate": 0.00015037037037037034,
+      "loss": 0.8524,
+      "step": 741
+    },
+    {
+      "epoch": 1.646145313366611,
+      "grad_norm": 0.3106069564819336,
+      "learning_rate": 0.00015012345679012347,
+      "loss": 0.5946,
+      "step": 742
+    },
+    {
+      "epoch": 1.6483638380476981,
+      "grad_norm": 0.3754357397556305,
+      "learning_rate": 0.00014987654320987653,
+      "loss": 0.8044,
+      "step": 743
+    },
+    {
+      "epoch": 1.6505823627287852,
+      "grad_norm": 0.3485739827156067,
+      "learning_rate": 0.00014962962962962963,
+      "loss": 0.7037,
+      "step": 744
+    },
+    {
+      "epoch": 1.6528008874098723,
+      "grad_norm": 0.30564582347869873,
+      "learning_rate": 0.0001493827160493827,
+      "loss": 0.6754,
+      "step": 745
+    },
+    {
+      "epoch": 1.6550194120909594,
+      "grad_norm": 0.2982652485370636,
+      "learning_rate": 0.0001491358024691358,
+      "loss": 0.78,
+      "step": 746
+    },
+    {
+      "epoch": 1.6572379367720464,
+      "grad_norm": 0.3643259108066559,
+      "learning_rate": 0.00014888888888888886,
+      "loss": 0.7867,
+      "step": 747
+    },
+    {
+      "epoch": 1.6594564614531335,
+      "grad_norm": 0.2643972337245941,
+      "learning_rate": 0.00014864197530864196,
+      "loss": 0.6558,
+      "step": 748
+    },
+    {
+      "epoch": 1.6616749861342206,
+      "grad_norm": 0.3208940029144287,
+      "learning_rate": 0.00014839506172839505,
+      "loss": 0.6383,
+      "step": 749
+    },
+    {
+      "epoch": 1.6638935108153077,
+      "grad_norm": 0.28904375433921814,
+      "learning_rate": 0.00014814814814814812,
+      "loss": 0.4872,
+      "step": 750
+    },
+    {
+      "epoch": 1.6661120354963947,
+      "grad_norm": 0.30051517486572266,
+      "learning_rate": 0.00014790123456790124,
+      "loss": 0.5728,
+      "step": 751
+    },
+    {
+      "epoch": 1.6683305601774818,
+      "grad_norm": 0.32607942819595337,
+      "learning_rate": 0.0001476543209876543,
+      "loss": 0.6054,
+      "step": 752
+    },
+    {
+      "epoch": 1.670549084858569,
+      "grad_norm": 0.3722771406173706,
+      "learning_rate": 0.0001474074074074074,
+      "loss": 0.6631,
+      "step": 753
+    },
+    {
+      "epoch": 1.672767609539656,
+      "grad_norm": 0.3404600918292999,
+      "learning_rate": 0.00014716049382716047,
+      "loss": 0.7449,
+      "step": 754
+    },
+    {
+      "epoch": 1.674986134220743,
+      "grad_norm": 0.2722650170326233,
+      "learning_rate": 0.00014691358024691357,
+      "loss": 0.7568,
+      "step": 755
+    },
+    {
+      "epoch": 1.6772046589018301,
+      "grad_norm": 0.29268306493759155,
+      "learning_rate": 0.00014666666666666664,
+      "loss": 0.8131,
+      "step": 756
+    },
+    {
+      "epoch": 1.6794231835829172,
+      "grad_norm": 0.3259887397289276,
+      "learning_rate": 0.00014641975308641973,
+      "loss": 0.7477,
+      "step": 757
+    },
+    {
+      "epoch": 1.6816417082640043,
+      "grad_norm": 0.278903067111969,
+      "learning_rate": 0.00014617283950617283,
+      "loss": 0.6889,
+      "step": 758
+    },
+    {
+      "epoch": 1.6838602329450914,
+      "grad_norm": 0.38602930307388306,
+      "learning_rate": 0.00014592592592592592,
+      "loss": 0.7262,
+      "step": 759
+    },
+    {
+      "epoch": 1.6860787576261784,
+      "grad_norm": 0.4022078514099121,
+      "learning_rate": 0.000145679012345679,
+      "loss": 0.7653,
+      "step": 760
+    },
+    {
+      "epoch": 1.6882972823072657,
+      "grad_norm": 0.2767799198627472,
+      "learning_rate": 0.00014543209876543208,
+      "loss": 0.742,
+      "step": 761
+    },
+    {
+      "epoch": 1.6905158069883528,
+      "grad_norm": 0.3517753481864929,
+      "learning_rate": 0.00014518518518518518,
+      "loss": 0.6868,
+      "step": 762
+    },
+    {
+      "epoch": 1.6927343316694399,
+      "grad_norm": 0.3629649877548218,
+      "learning_rate": 0.00014493827160493825,
+      "loss": 1.0878,
+      "step": 763
+    },
+    {
+      "epoch": 1.694952856350527,
+      "grad_norm": 0.3809618353843689,
+      "learning_rate": 0.00014469135802469134,
+      "loss": 0.7238,
+      "step": 764
+    },
+    {
+      "epoch": 1.697171381031614,
+      "grad_norm": 0.3370950520038605,
+      "learning_rate": 0.0001444444444444444,
+      "loss": 0.8901,
+      "step": 765
+    },
+    {
+      "epoch": 1.699389905712701,
+      "grad_norm": 0.2843126654624939,
+      "learning_rate": 0.00014419753086419753,
+      "loss": 0.6644,
+      "step": 766
+    },
+    {
+      "epoch": 1.7016084303937882,
+      "grad_norm": 0.34927135705947876,
+      "learning_rate": 0.0001439506172839506,
+      "loss": 0.921,
+      "step": 767
+    },
+    {
+      "epoch": 1.7038269550748752,
+      "grad_norm": 0.25622934103012085,
+      "learning_rate": 0.0001437037037037037,
+      "loss": 0.575,
+      "step": 768
+    },
+    {
+      "epoch": 1.7060454797559623,
+      "grad_norm": 0.40387478470802307,
+      "learning_rate": 0.00014345679012345676,
+      "loss": 0.7299,
+      "step": 769
+    },
+    {
+      "epoch": 1.7082640044370494,
+      "grad_norm": 0.40031298995018005,
+      "learning_rate": 0.00014320987654320986,
+      "loss": 0.8298,
+      "step": 770
+    },
+    {
+      "epoch": 1.7104825291181365,
+      "grad_norm": 0.27454328536987305,
+      "learning_rate": 0.00014296296296296295,
+      "loss": 0.6409,
+      "step": 771
+    },
+    {
+      "epoch": 1.7127010537992235,
+      "grad_norm": 0.38810995221138,
+      "learning_rate": 0.00014271604938271602,
+      "loss": 0.6641,
+      "step": 772
+    },
+    {
+      "epoch": 1.7149195784803106,
+      "grad_norm": 0.32779669761657715,
+      "learning_rate": 0.00014246913580246912,
+      "loss": 0.789,
+      "step": 773
+    },
+    {
+      "epoch": 1.7171381031613977,
+      "grad_norm": 0.3264537453651428,
+      "learning_rate": 0.0001422222222222222,
+      "loss": 0.8356,
+      "step": 774
+    },
+    {
+      "epoch": 1.7193566278424848,
+      "grad_norm": 0.31787949800491333,
+      "learning_rate": 0.0001419753086419753,
+      "loss": 0.6194,
+      "step": 775
+    },
+    {
+      "epoch": 1.7215751525235718,
+      "grad_norm": 0.3091427683830261,
+      "learning_rate": 0.00014172839506172838,
+      "loss": 0.7274,
+      "step": 776
+    },
+    {
+      "epoch": 1.723793677204659,
+      "grad_norm": 0.30796658992767334,
+      "learning_rate": 0.00014148148148148147,
+      "loss": 0.8659,
+      "step": 777
+    },
+    {
+      "epoch": 1.726012201885746,
+      "grad_norm": 0.3192078471183777,
+      "learning_rate": 0.00014123456790123454,
+      "loss": 0.8007,
+      "step": 778
+    },
+    {
+      "epoch": 1.728230726566833,
+      "grad_norm": 0.3503005802631378,
+      "learning_rate": 0.00014098765432098763,
+      "loss": 0.7897,
+      "step": 779
+    },
+    {
+      "epoch": 1.7304492512479202,
+      "grad_norm": 0.28616148233413696,
+      "learning_rate": 0.00014074074074074073,
+      "loss": 0.7586,
+      "step": 780
+    },
+    {
+      "epoch": 1.7326677759290072,
+      "grad_norm": 0.2781727612018585,
+      "learning_rate": 0.00014049382716049382,
+      "loss": 0.6613,
+      "step": 781
+    },
+    {
+      "epoch": 1.7348863006100943,
+      "grad_norm": 0.3216390311717987,
+      "learning_rate": 0.0001402469135802469,
+      "loss": 0.7288,
+      "step": 782
+    },
+    {
+      "epoch": 1.7371048252911814,
+      "grad_norm": 0.28800565004348755,
+      "learning_rate": 0.00014,
+      "loss": 0.7449,
+      "step": 783
+    },
+    {
+      "epoch": 1.7393233499722685,
+      "grad_norm": 0.32171830534935,
+      "learning_rate": 0.00013975308641975308,
+      "loss": 0.7013,
+      "step": 784
+    },
+    {
+      "epoch": 1.7415418746533555,
+      "grad_norm": 0.3286561965942383,
+      "learning_rate": 0.00013950617283950615,
+      "loss": 0.7229,
+      "step": 785
+    },
+    {
+      "epoch": 1.7437603993344426,
+      "grad_norm": 0.31590333580970764,
+      "learning_rate": 0.00013925925925925925,
+      "loss": 0.7833,
+      "step": 786
+    },
+    {
+      "epoch": 1.7459789240155297,
+      "grad_norm": 0.32752737402915955,
+      "learning_rate": 0.00013901234567901231,
+      "loss": 0.6125,
+      "step": 787
+    },
+    {
+      "epoch": 1.7481974486966168,
+      "grad_norm": 0.4051518738269806,
+      "learning_rate": 0.00013876543209876544,
+      "loss": 0.917,
+      "step": 788
+    },
+    {
+      "epoch": 1.7504159733777038,
+      "grad_norm": 0.2849074900150299,
+      "learning_rate": 0.0001385185185185185,
+      "loss": 0.6409,
+      "step": 789
+    },
+    {
+      "epoch": 1.752634498058791,
+      "grad_norm": 0.3306087851524353,
+      "learning_rate": 0.0001382716049382716,
+      "loss": 0.6316,
+      "step": 790
+    },
+    {
+      "epoch": 1.754853022739878,
+      "grad_norm": 0.3363149166107178,
+      "learning_rate": 0.00013802469135802467,
+      "loss": 0.7477,
+      "step": 791
+    },
+    {
+      "epoch": 1.757071547420965,
+      "grad_norm": 0.3176649212837219,
+      "learning_rate": 0.00013777777777777776,
+      "loss": 0.5782,
+      "step": 792
+    },
+    {
+      "epoch": 1.7592900721020521,
+      "grad_norm": 0.32553067803382874,
+      "learning_rate": 0.00013753086419753086,
+      "loss": 0.6106,
+      "step": 793
+    },
+    {
+      "epoch": 1.7615085967831392,
+      "grad_norm": 0.3698481023311615,
+      "learning_rate": 0.00013728395061728393,
+      "loss": 0.7377,
+      "step": 794
+    },
+    {
+      "epoch": 1.7637271214642263,
+      "grad_norm": 0.35148802399635315,
+      "learning_rate": 0.00013703703703703702,
+      "loss": 0.7406,
+      "step": 795
+    },
+    {
+      "epoch": 1.7659456461453134,
+      "grad_norm": 0.3244885802268982,
+      "learning_rate": 0.00013679012345679012,
+      "loss": 0.8208,
+      "step": 796
+    },
+    {
+      "epoch": 1.7681641708264004,
+      "grad_norm": 0.3165164887905121,
+      "learning_rate": 0.0001365432098765432,
+      "loss": 0.7073,
+      "step": 797
+    },
+    {
+      "epoch": 1.7703826955074875,
+      "grad_norm": 0.26543429493904114,
+      "learning_rate": 0.00013629629629629628,
+      "loss": 0.6895,
+      "step": 798
+    },
+    {
+      "epoch": 1.7726012201885746,
+      "grad_norm": 0.3363840878009796,
+      "learning_rate": 0.00013604938271604937,
+      "loss": 0.7348,
+      "step": 799
+    },
+    {
+      "epoch": 1.7748197448696617,
+      "grad_norm": 0.38652539253234863,
+      "learning_rate": 0.00013580246913580244,
+      "loss": 0.651,
+      "step": 800
+    },
+    {
+      "epoch": 1.7770382695507487,
+      "grad_norm": 0.25464650988578796,
+      "learning_rate": 0.00013555555555555554,
+      "loss": 0.6567,
+      "step": 801
+    },
+    {
+      "epoch": 1.7792567942318358,
+      "grad_norm": 0.3350493013858795,
+      "learning_rate": 0.00013530864197530863,
+      "loss": 0.6296,
+      "step": 802
+    },
+    {
+      "epoch": 1.7814753189129229,
+      "grad_norm": 0.3161368668079376,
+      "learning_rate": 0.00013506172839506173,
+      "loss": 0.6373,
+      "step": 803
+    },
+    {
+      "epoch": 1.78369384359401,
+      "grad_norm": 0.32538414001464844,
+      "learning_rate": 0.0001348148148148148,
+      "loss": 0.7226,
+      "step": 804
+    },
+    {
+      "epoch": 1.785912368275097,
+      "grad_norm": 0.27925610542297363,
+      "learning_rate": 0.0001345679012345679,
+      "loss": 0.6046,
+      "step": 805
+    },
+    {
+      "epoch": 1.788130892956184,
+      "grad_norm": 0.27884551882743835,
+      "learning_rate": 0.00013432098765432099,
+      "loss": 0.6542,
+      "step": 806
+    },
+    {
+      "epoch": 1.7903494176372712,
+      "grad_norm": 0.4178118407726288,
+      "learning_rate": 0.00013407407407407405,
+      "loss": 0.696,
+      "step": 807
+    },
+    {
+      "epoch": 1.7925679423183583,
+      "grad_norm": 0.36381810903549194,
+      "learning_rate": 0.00013382716049382715,
+      "loss": 0.7963,
+      "step": 808
+    },
+    {
+      "epoch": 1.7947864669994453,
+      "grad_norm": 0.2824428081512451,
+      "learning_rate": 0.00013358024691358022,
+      "loss": 0.6187,
+      "step": 809
+    },
+    {
+      "epoch": 1.7970049916805324,
+      "grad_norm": 0.2801170349121094,
+      "learning_rate": 0.0001333333333333333,
+      "loss": 0.6317,
+      "step": 810
+    },
+    {
+      "epoch": 1.7992235163616195,
+      "grad_norm": 0.25746533274650574,
+      "learning_rate": 0.0001330864197530864,
+      "loss": 0.6413,
+      "step": 811
+    },
+    {
+      "epoch": 1.8014420410427066,
+      "grad_norm": 0.2891461253166199,
+      "learning_rate": 0.0001328395061728395,
+      "loss": 0.7191,
+      "step": 812
+    },
+    {
+      "epoch": 1.8036605657237936,
+      "grad_norm": 0.24890680611133575,
+      "learning_rate": 0.00013259259259259257,
+      "loss": 0.7302,
+      "step": 813
+    },
+    {
+      "epoch": 1.8058790904048807,
+      "grad_norm": 0.34262436628341675,
+      "learning_rate": 0.00013234567901234567,
+      "loss": 0.7104,
+      "step": 814
+    },
+    {
+      "epoch": 1.8080976150859678,
+      "grad_norm": 0.39875757694244385,
+      "learning_rate": 0.00013209876543209876,
+      "loss": 0.6223,
+      "step": 815
+    },
+    {
+      "epoch": 1.8103161397670549,
+      "grad_norm": 0.34255075454711914,
+      "learning_rate": 0.00013185185185185183,
+      "loss": 0.9488,
+      "step": 816
+    },
+    {
+      "epoch": 1.8125346644481422,
+      "grad_norm": 0.39329656958580017,
+      "learning_rate": 0.00013160493827160492,
+      "loss": 0.6924,
+      "step": 817
+    },
+    {
+      "epoch": 1.8147531891292292,
+      "grad_norm": 0.279289186000824,
+      "learning_rate": 0.00013135802469135802,
+      "loss": 0.7174,
+      "step": 818
+    },
+    {
+      "epoch": 1.8169717138103163,
+      "grad_norm": 0.38620859384536743,
+      "learning_rate": 0.00013111111111111111,
+      "loss": 0.7702,
+      "step": 819
+    },
+    {
+      "epoch": 1.8191902384914034,
+      "grad_norm": 0.3993968963623047,
+      "learning_rate": 0.00013086419753086418,
+      "loss": 0.7741,
+      "step": 820
+    },
+    {
+      "epoch": 1.8214087631724905,
+      "grad_norm": 0.3427724242210388,
+      "learning_rate": 0.00013061728395061728,
+      "loss": 0.7611,
+      "step": 821
+    },
+    {
+      "epoch": 1.8236272878535775,
+      "grad_norm": 0.5088453888893127,
+      "learning_rate": 0.00013037037037037035,
+      "loss": 0.7957,
+      "step": 822
+    },
+    {
+      "epoch": 1.8258458125346646,
+      "grad_norm": 0.300751656293869,
+      "learning_rate": 0.00013012345679012344,
+      "loss": 0.6266,
+      "step": 823
+    },
+    {
+      "epoch": 1.8280643372157517,
+      "grad_norm": 0.36075693368911743,
+      "learning_rate": 0.00012987654320987654,
+      "loss": 0.842,
+      "step": 824
+    },
+    {
+      "epoch": 1.8302828618968388,
+      "grad_norm": 0.35546496510505676,
+      "learning_rate": 0.0001296296296296296,
+      "loss": 0.7928,
+      "step": 825
+    },
+    {
+      "epoch": 1.8325013865779258,
+      "grad_norm": 0.3749972879886627,
+      "learning_rate": 0.0001293827160493827,
+      "loss": 0.8404,
+      "step": 826
+    },
+    {
+      "epoch": 1.834719911259013,
+      "grad_norm": 0.29471275210380554,
+      "learning_rate": 0.0001291358024691358,
+      "loss": 0.8023,
+      "step": 827
+    },
+    {
+      "epoch": 1.8369384359401,
+      "grad_norm": 0.39225876331329346,
+      "learning_rate": 0.0001288888888888889,
+      "loss": 0.641,
+      "step": 828
+    },
+    {
+      "epoch": 1.839156960621187,
+      "grad_norm": 0.2704963982105255,
+      "learning_rate": 0.00012864197530864196,
+      "loss": 0.7532,
+      "step": 829
+    },
+    {
+      "epoch": 1.8413754853022741,
+      "grad_norm": 0.3629330098628998,
+      "learning_rate": 0.00012839506172839505,
+      "loss": 0.7233,
+      "step": 830
+    },
+    {
+      "epoch": 1.8435940099833612,
+      "grad_norm": 0.3808145821094513,
+      "learning_rate": 0.00012814814814814812,
+      "loss": 0.7652,
+      "step": 831
+    },
+    {
+      "epoch": 1.8458125346644483,
+      "grad_norm": 0.2982264757156372,
+      "learning_rate": 0.00012790123456790122,
+      "loss": 0.7834,
+      "step": 832
+    },
+    {
+      "epoch": 1.8480310593455354,
+      "grad_norm": 0.4305157959461212,
+      "learning_rate": 0.0001276543209876543,
+      "loss": 0.8143,
+      "step": 833
+    },
+    {
+      "epoch": 1.8502495840266224,
+      "grad_norm": 0.25910547375679016,
+      "learning_rate": 0.0001274074074074074,
+      "loss": 0.6114,
+      "step": 834
+    },
+    {
+      "epoch": 1.8524681087077095,
+      "grad_norm": 0.40026503801345825,
+      "learning_rate": 0.00012716049382716047,
+      "loss": 0.7478,
+      "step": 835
+    },
+    {
+      "epoch": 1.8546866333887966,
+      "grad_norm": 0.35869815945625305,
+      "learning_rate": 0.00012691358024691357,
+      "loss": 0.8996,
+      "step": 836
+    },
+    {
+      "epoch": 1.8569051580698837,
+      "grad_norm": 0.38456451892852783,
+      "learning_rate": 0.00012666666666666666,
+      "loss": 0.7947,
+      "step": 837
+    },
+    {
+      "epoch": 1.8591236827509707,
+      "grad_norm": 0.389890193939209,
+      "learning_rate": 0.00012641975308641973,
+      "loss": 0.6674,
+      "step": 838
+    },
+    {
+      "epoch": 1.8613422074320578,
+      "grad_norm": 0.2884763777256012,
+      "learning_rate": 0.00012617283950617283,
+      "loss": 0.5826,
+      "step": 839
+    },
+    {
+      "epoch": 1.8635607321131449,
+      "grad_norm": 0.2756790518760681,
+      "learning_rate": 0.0001259259259259259,
+      "loss": 0.6873,
+      "step": 840
+    },
+    {
+      "epoch": 1.865779256794232,
+      "grad_norm": 0.38118836283683777,
+      "learning_rate": 0.00012567901234567902,
+      "loss": 0.8147,
+      "step": 841
+    },
+    {
+      "epoch": 1.867997781475319,
+      "grad_norm": 0.3216645121574402,
+      "learning_rate": 0.00012543209876543209,
+      "loss": 0.8144,
+      "step": 842
+    },
+    {
+      "epoch": 1.870216306156406,
+      "grad_norm": 0.3267296850681305,
+      "learning_rate": 0.00012518518518518518,
+      "loss": 0.6467,
+      "step": 843
+    },
+    {
+      "epoch": 1.8724348308374932,
+      "grad_norm": 0.3718464970588684,
+      "learning_rate": 0.00012493827160493825,
+      "loss": 0.895,
+      "step": 844
+    },
+    {
+      "epoch": 1.8746533555185803,
+      "grad_norm": 0.24207782745361328,
+      "learning_rate": 0.00012469135802469134,
+      "loss": 0.6672,
+      "step": 845
+    },
+    {
+      "epoch": 1.8768718801996673,
+      "grad_norm": 0.35015353560447693,
+      "learning_rate": 0.00012444444444444444,
+      "loss": 0.7884,
+      "step": 846
+    },
+    {
+      "epoch": 1.8790904048807544,
+      "grad_norm": 0.30226242542266846,
+      "learning_rate": 0.0001241975308641975,
+      "loss": 0.6873,
+      "step": 847
+    },
+    {
+      "epoch": 1.8813089295618415,
+      "grad_norm": 0.32037070393562317,
+      "learning_rate": 0.0001239506172839506,
+      "loss": 0.6733,
+      "step": 848
+    },
+    {
+      "epoch": 1.8835274542429286,
+      "grad_norm": 0.3182600736618042,
+      "learning_rate": 0.0001237037037037037,
+      "loss": 0.6955,
+      "step": 849
+    },
+    {
+      "epoch": 1.8857459789240156,
+      "grad_norm": 0.3208216726779938,
+      "learning_rate": 0.0001234567901234568,
+      "loss": 0.7466,
+      "step": 850
+    },
+    {
+      "epoch": 1.8879645036051027,
+      "grad_norm": 0.30803990364074707,
+      "learning_rate": 0.00012320987654320986,
+      "loss": 0.7166,
+      "step": 851
+    },
+    {
+      "epoch": 1.8901830282861898,
+      "grad_norm": 0.30191513895988464,
+      "learning_rate": 0.00012296296296296296,
+      "loss": 0.6748,
+      "step": 852
+    },
+    {
+      "epoch": 1.8924015529672769,
+      "grad_norm": 0.3444351255893707,
+      "learning_rate": 0.00012271604938271602,
+      "loss": 0.5594,
+      "step": 853
+    },
+    {
+      "epoch": 1.894620077648364,
+      "grad_norm": 0.3323015868663788,
+      "learning_rate": 0.00012246913580246912,
+      "loss": 0.6796,
+      "step": 854
+    },
+    {
+      "epoch": 1.896838602329451,
+      "grad_norm": 0.2612377405166626,
+      "learning_rate": 0.00012222222222222221,
+      "loss": 1.0069,
+      "step": 855
+    },
+    {
+      "epoch": 1.899057127010538,
+      "grad_norm": 0.2681643068790436,
+      "learning_rate": 0.00012197530864197531,
+      "loss": 0.5953,
+      "step": 856
+    },
+    {
+      "epoch": 1.9012756516916252,
+      "grad_norm": 0.3278005123138428,
+      "learning_rate": 0.00012172839506172839,
+      "loss": 0.5959,
+      "step": 857
+    },
+    {
+      "epoch": 1.9034941763727122,
+      "grad_norm": 0.38157182931900024,
+      "learning_rate": 0.00012148148148148147,
+      "loss": 0.8317,
+      "step": 858
+    },
+    {
+      "epoch": 1.9057127010537993,
+      "grad_norm": 0.3950137197971344,
+      "learning_rate": 0.00012123456790123455,
+      "loss": 0.8901,
+      "step": 859
+    },
+    {
+      "epoch": 1.9079312257348864,
+      "grad_norm": 0.3997960686683655,
+      "learning_rate": 0.00012098765432098764,
+      "loss": 0.7423,
+      "step": 860
+    },
+    {
+      "epoch": 1.9101497504159735,
+      "grad_norm": 0.27877548336982727,
+      "learning_rate": 0.00012074074074074073,
+      "loss": 0.914,
+      "step": 861
+    },
+    {
+      "epoch": 1.9123682750970605,
+      "grad_norm": 0.35123705863952637,
+      "learning_rate": 0.00012049382716049381,
+      "loss": 0.5634,
+      "step": 862
+    },
+    {
+      "epoch": 1.9145867997781476,
+      "grad_norm": 0.31742504239082336,
+      "learning_rate": 0.0001202469135802469,
+      "loss": 0.5575,
+      "step": 863
+    },
+    {
+      "epoch": 1.9168053244592347,
+      "grad_norm": 0.390474408864975,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 0.7188,
+      "step": 864
+    },
+    {
+      "epoch": 1.9190238491403218,
+      "grad_norm": 0.27736327052116394,
+      "learning_rate": 0.00011975308641975308,
+      "loss": 0.6971,
+      "step": 865
+    },
+    {
+      "epoch": 1.9212423738214088,
+      "grad_norm": 0.40486714243888855,
+      "learning_rate": 0.00011950617283950617,
+      "loss": 0.6838,
+      "step": 866
+    },
+    {
+      "epoch": 1.923460898502496,
+      "grad_norm": 0.3235892653465271,
+      "learning_rate": 0.00011925925925925925,
+      "loss": 0.594,
+      "step": 867
+    },
+    {
+      "epoch": 1.925679423183583,
+      "grad_norm": 0.3409321904182434,
+      "learning_rate": 0.00011901234567901233,
+      "loss": 0.9011,
+      "step": 868
+    },
+    {
+      "epoch": 1.92789794786467,
+      "grad_norm": 0.34094899892807007,
+      "learning_rate": 0.00011876543209876541,
+      "loss": 0.7765,
+      "step": 869
+    },
+    {
+      "epoch": 1.9301164725457571,
+      "grad_norm": 0.4011700451374054,
+      "learning_rate": 0.0001185185185185185,
+      "loss": 0.9597,
+      "step": 870
+    },
+    {
+      "epoch": 1.9323349972268442,
+      "grad_norm": 0.355100154876709,
+      "learning_rate": 0.0001182716049382716,
+      "loss": 0.7505,
+      "step": 871
+    },
+    {
+      "epoch": 1.9345535219079313,
+      "grad_norm": 0.3918919265270233,
+      "learning_rate": 0.00011802469135802468,
+      "loss": 0.744,
+      "step": 872
+    },
+    {
+      "epoch": 1.9367720465890184,
+      "grad_norm": 0.3756035566329956,
+      "learning_rate": 0.00011777777777777776,
+      "loss": 0.625,
+      "step": 873
+    },
+    {
+      "epoch": 1.9389905712701054,
+      "grad_norm": 0.3947852849960327,
+      "learning_rate": 0.00011753086419753086,
+      "loss": 0.7491,
+      "step": 874
+    },
+    {
+      "epoch": 1.9412090959511925,
+      "grad_norm": 0.27899986505508423,
+      "learning_rate": 0.00011728395061728394,
+      "loss": 0.8339,
+      "step": 875
+    },
+    {
+      "epoch": 1.9434276206322796,
+      "grad_norm": 0.40907737612724304,
+      "learning_rate": 0.00011703703703703702,
+      "loss": 0.6557,
+      "step": 876
+    },
+    {
+      "epoch": 1.9456461453133667,
+      "grad_norm": 0.2738918960094452,
+      "learning_rate": 0.0001167901234567901,
+      "loss": 0.5621,
+      "step": 877
+    },
+    {
+      "epoch": 1.9478646699944537,
+      "grad_norm": 0.2920062243938446,
+      "learning_rate": 0.00011654320987654319,
+      "loss": 0.6615,
+      "step": 878
+    },
+    {
+      "epoch": 1.9500831946755408,
+      "grad_norm": 0.3285408318042755,
+      "learning_rate": 0.0001162962962962963,
+      "loss": 0.7438,
+      "step": 879
+    },
+    {
+      "epoch": 1.9523017193566279,
+      "grad_norm": 0.27441486716270447,
+      "learning_rate": 0.00011604938271604938,
+      "loss": 0.6549,
+      "step": 880
+    },
+    {
+      "epoch": 1.954520244037715,
+      "grad_norm": 0.3172101080417633,
+      "learning_rate": 0.00011580246913580246,
+      "loss": 0.7178,
+      "step": 881
+    },
+    {
+      "epoch": 1.956738768718802,
+      "grad_norm": 0.2766132354736328,
+      "learning_rate": 0.00011555555555555554,
+      "loss": 0.7242,
+      "step": 882
+    },
+    {
+      "epoch": 1.958957293399889,
+      "grad_norm": 0.31022900342941284,
+      "learning_rate": 0.00011530864197530863,
+      "loss": 0.8407,
+      "step": 883
+    },
+    {
+      "epoch": 1.9611758180809762,
+      "grad_norm": 0.38427525758743286,
+      "learning_rate": 0.00011506172839506172,
+      "loss": 0.7315,
+      "step": 884
+    },
+    {
+      "epoch": 1.9633943427620633,
+      "grad_norm": 0.3450315296649933,
+      "learning_rate": 0.0001148148148148148,
+      "loss": 0.715,
+      "step": 885
+    },
+    {
+      "epoch": 1.9656128674431503,
+      "grad_norm": 0.3379027545452118,
+      "learning_rate": 0.00011456790123456789,
+      "loss": 0.7842,
+      "step": 886
+    },
+    {
+      "epoch": 1.9678313921242374,
+      "grad_norm": 0.38613322377204895,
+      "learning_rate": 0.00011432098765432099,
+      "loss": 0.81,
+      "step": 887
+    },
+    {
+      "epoch": 1.9700499168053245,
+      "grad_norm": 1.4010131359100342,
+      "learning_rate": 0.00011407407407407407,
+      "loss": 0.6937,
+      "step": 888
+    },
+    {
+      "epoch": 1.9722684414864116,
+      "grad_norm": 0.295520156621933,
+      "learning_rate": 0.00011382716049382715,
+      "loss": 0.6375,
+      "step": 889
+    },
+    {
+      "epoch": 1.9744869661674986,
+      "grad_norm": 0.2881146967411041,
+      "learning_rate": 0.00011358024691358023,
+      "loss": 0.5929,
+      "step": 890
+    },
+    {
+      "epoch": 1.9767054908485857,
+      "grad_norm": 0.45779848098754883,
+      "learning_rate": 0.00011333333333333331,
+      "loss": 0.8378,
+      "step": 891
+    },
+    {
+      "epoch": 1.9789240155296728,
+      "grad_norm": 0.3618588149547577,
+      "learning_rate": 0.00011308641975308641,
+      "loss": 0.5644,
+      "step": 892
+    },
+    {
+      "epoch": 1.9811425402107599,
+      "grad_norm": 0.3937014937400818,
+      "learning_rate": 0.00011283950617283949,
+      "loss": 0.6868,
+      "step": 893
+    },
+    {
+      "epoch": 1.983361064891847,
+      "grad_norm": 0.2787335216999054,
+      "learning_rate": 0.00011259259259259259,
+      "loss": 0.6386,
+      "step": 894
+    },
+    {
+      "epoch": 1.985579589572934,
+      "grad_norm": 0.31302133202552795,
+      "learning_rate": 0.00011234567901234567,
+      "loss": 0.8036,
+      "step": 895
+    },
+    {
+      "epoch": 1.987798114254021,
+      "grad_norm": 0.32790157198905945,
+      "learning_rate": 0.00011209876543209876,
+      "loss": 0.6945,
+      "step": 896
+    },
+    {
+      "epoch": 1.9900166389351082,
+      "grad_norm": 0.30650439858436584,
+      "learning_rate": 0.00011185185185185184,
+      "loss": 0.5674,
+      "step": 897
+    },
+    {
+      "epoch": 1.9922351636161952,
+      "grad_norm": 0.3315053880214691,
+      "learning_rate": 0.00011160493827160493,
+      "loss": 0.5573,
+      "step": 898
+    },
+    {
+      "epoch": 1.9944536882972823,
+      "grad_norm": 0.30042293667793274,
+      "learning_rate": 0.00011135802469135801,
+      "loss": 0.7087,
+      "step": 899
+    },
+    {
+      "epoch": 1.9966722129783694,
+      "grad_norm": 0.44175413250923157,
+      "learning_rate": 0.00011111111111111109,
+      "loss": 0.776,
+      "step": 900
+    },
+    {
+      "epoch": 1.9988907376594565,
+      "grad_norm": 0.32071128487586975,
+      "learning_rate": 0.0001108641975308642,
+      "loss": 0.6358,
+      "step": 901
+    },
+    {
+      "epoch": 2.0011092623405435,
+      "grad_norm": 0.30683010816574097,
+      "learning_rate": 0.00011061728395061728,
+      "loss": 0.7083,
+      "step": 902
+    },
+    {
+      "epoch": 2.0033277870216306,
+      "grad_norm": 0.3082660734653473,
+      "learning_rate": 0.00011037037037037036,
+      "loss": 0.5953,
+      "step": 903
+    },
+    {
+      "epoch": 2.0055463117027177,
+      "grad_norm": 0.39927366375923157,
+      "learning_rate": 0.00011012345679012344,
+      "loss": 0.6913,
+      "step": 904
+    },
+    {
+      "epoch": 2.0077648363838048,
+      "grad_norm": 0.3141803741455078,
+      "learning_rate": 0.00010987654320987654,
+      "loss": 0.77,
+      "step": 905
+    },
+    {
+      "epoch": 2.009983361064892,
+      "grad_norm": 0.4064863920211792,
+      "learning_rate": 0.00010962962962962962,
+      "loss": 0.6212,
+      "step": 906
+    },
+    {
+      "epoch": 2.012201885745979,
+      "grad_norm": 0.2454594075679779,
+      "learning_rate": 0.0001093827160493827,
+      "loss": 0.6472,
+      "step": 907
+    },
+    {
+      "epoch": 2.014420410427066,
+      "grad_norm": 0.2857517898082733,
+      "learning_rate": 0.00010913580246913578,
+      "loss": 0.7127,
+      "step": 908
+    },
+    {
+      "epoch": 2.016638935108153,
+      "grad_norm": 0.3515011668205261,
+      "learning_rate": 0.00010888888888888889,
+      "loss": 0.7819,
+      "step": 909
+    },
+    {
+      "epoch": 2.01885745978924,
+      "grad_norm": 0.32326120138168335,
+      "learning_rate": 0.00010864197530864197,
+      "loss": 0.7064,
+      "step": 910
+    },
+    {
+      "epoch": 2.021075984470327,
+      "grad_norm": 0.28967180848121643,
+      "learning_rate": 0.00010839506172839505,
+      "loss": 0.7131,
+      "step": 911
+    },
+    {
+      "epoch": 2.0232945091514143,
+      "grad_norm": 0.3091174066066742,
+      "learning_rate": 0.00010814814814814814,
+      "loss": 0.6319,
+      "step": 912
+    },
+    {
+      "epoch": 2.0255130338325014,
+      "grad_norm": 0.3107006847858429,
+      "learning_rate": 0.00010790123456790122,
+      "loss": 0.563,
+      "step": 913
+    },
+    {
+      "epoch": 2.0277315585135884,
+      "grad_norm": 0.2923293709754944,
+      "learning_rate": 0.00010765432098765431,
+      "loss": 0.6959,
+      "step": 914
+    },
+    {
+      "epoch": 2.0299500831946755,
+      "grad_norm": 0.39304766058921814,
+      "learning_rate": 0.0001074074074074074,
+      "loss": 0.6565,
+      "step": 915
+    },
+    {
+      "epoch": 2.0321686078757626,
+      "grad_norm": 0.2778719365596771,
+      "learning_rate": 0.00010716049382716048,
+      "loss": 0.5742,
+      "step": 916
+    },
+    {
+      "epoch": 2.0343871325568497,
+      "grad_norm": 0.3342873752117157,
+      "learning_rate": 0.00010691358024691357,
+      "loss": 0.6357,
+      "step": 917
+    },
+    {
+      "epoch": 2.0366056572379367,
+      "grad_norm": 0.3781198561191559,
+      "learning_rate": 0.00010666666666666667,
+      "loss": 0.6057,
+      "step": 918
+    },
+    {
+      "epoch": 2.038824181919024,
+      "grad_norm": 0.32263439893722534,
+      "learning_rate": 0.00010641975308641975,
+      "loss": 0.7998,
+      "step": 919
+    },
+    {
+      "epoch": 2.041042706600111,
+      "grad_norm": 0.2679537832736969,
+      "learning_rate": 0.00010617283950617283,
+      "loss": 0.544,
+      "step": 920
+    },
+    {
+      "epoch": 2.043261231281198,
+      "grad_norm": 0.3113681674003601,
+      "learning_rate": 0.00010592592592592591,
+      "loss": 0.5866,
+      "step": 921
+    },
+    {
+      "epoch": 2.045479755962285,
+      "grad_norm": 0.3256099224090576,
+      "learning_rate": 0.00010567901234567899,
+      "loss": 0.7615,
+      "step": 922
+    },
+    {
+      "epoch": 2.047698280643372,
+      "grad_norm": 0.2568422257900238,
+      "learning_rate": 0.00010543209876543209,
+      "loss": 0.5962,
+      "step": 923
+    },
+    {
+      "epoch": 2.049916805324459,
+      "grad_norm": 0.325097918510437,
+      "learning_rate": 0.00010518518518518518,
+      "loss": 0.5286,
+      "step": 924
+    },
+    {
+      "epoch": 2.0521353300055463,
+      "grad_norm": 0.26789817214012146,
+      "learning_rate": 0.00010493827160493826,
+      "loss": 0.5988,
+      "step": 925
+    },
+    {
+      "epoch": 2.0543538546866333,
+      "grad_norm": 0.3094393014907837,
+      "learning_rate": 0.00010469135802469135,
+      "loss": 0.742,
+      "step": 926
+    },
+    {
+      "epoch": 2.0565723793677204,
+      "grad_norm": 0.3021222651004791,
+      "learning_rate": 0.00010444444444444444,
+      "loss": 0.7331,
+      "step": 927
+    },
+    {
+      "epoch": 2.0587909040488075,
+      "grad_norm": 0.34265440702438354,
+      "learning_rate": 0.00010419753086419752,
+      "loss": 0.539,
+      "step": 928
+    },
+    {
+      "epoch": 2.0610094287298946,
+      "grad_norm": 0.3500079810619354,
+      "learning_rate": 0.0001039506172839506,
+      "loss": 0.7859,
+      "step": 929
+    },
+    {
+      "epoch": 2.0632279534109816,
+      "grad_norm": 0.3058033287525177,
+      "learning_rate": 0.00010370370370370369,
+      "loss": 0.6641,
+      "step": 930
+    },
+    {
+      "epoch": 2.0654464780920687,
+      "grad_norm": 0.29699403047561646,
+      "learning_rate": 0.00010345679012345677,
+      "loss": 0.791,
+      "step": 931
+    },
+    {
+      "epoch": 2.067665002773156,
+      "grad_norm": 0.2828279435634613,
+      "learning_rate": 0.00010320987654320988,
+      "loss": 0.5583,
+      "step": 932
+    },
+    {
+      "epoch": 2.069883527454243,
+      "grad_norm": 0.32840675115585327,
+      "learning_rate": 0.00010296296296296296,
+      "loss": 0.7839,
+      "step": 933
+    },
+    {
+      "epoch": 2.07210205213533,
+      "grad_norm": 0.37871822714805603,
+      "learning_rate": 0.00010271604938271604,
+      "loss": 0.8039,
+      "step": 934
+    },
+    {
+      "epoch": 2.074320576816417,
+      "grad_norm": 0.335502564907074,
+      "learning_rate": 0.00010246913580246912,
+      "loss": 0.6094,
+      "step": 935
+    },
+    {
+      "epoch": 2.076539101497504,
+      "grad_norm": 0.33749639987945557,
+      "learning_rate": 0.00010222222222222222,
+      "loss": 0.9484,
+      "step": 936
+    },
+    {
+      "epoch": 2.078757626178591,
+      "grad_norm": 0.3336297869682312,
+      "learning_rate": 0.0001019753086419753,
+      "loss": 0.7506,
+      "step": 937
+    },
+    {
+      "epoch": 2.0809761508596782,
+      "grad_norm": 0.3587692081928253,
+      "learning_rate": 0.00010172839506172838,
+      "loss": 0.5446,
+      "step": 938
+    },
+    {
+      "epoch": 2.0831946755407653,
+      "grad_norm": 0.39508581161499023,
+      "learning_rate": 0.00010148148148148147,
+      "loss": 0.6371,
+      "step": 939
+    },
+    {
+      "epoch": 2.0854132002218524,
+      "grad_norm": 0.2936362326145172,
+      "learning_rate": 0.00010123456790123457,
+      "loss": 0.6163,
+      "step": 940
+    },
+    {
+      "epoch": 2.0876317249029395,
+      "grad_norm": 0.3153306841850281,
+      "learning_rate": 0.00010098765432098765,
+      "loss": 0.7952,
+      "step": 941
+    },
+    {
+      "epoch": 2.0898502495840265,
+      "grad_norm": 0.35774263739585876,
+      "learning_rate": 0.00010074074074074073,
+      "loss": 0.746,
+      "step": 942
+    },
+    {
+      "epoch": 2.0920687742651136,
+      "grad_norm": 0.355527400970459,
+      "learning_rate": 0.00010049382716049381,
+      "loss": 0.6741,
+      "step": 943
+    },
+    {
+      "epoch": 2.0942872989462007,
+      "grad_norm": 0.2694290280342102,
+      "learning_rate": 0.0001002469135802469,
+      "loss": 0.5772,
+      "step": 944
+    },
+    {
+      "epoch": 2.0965058236272878,
+      "grad_norm": 0.2968207001686096,
+      "learning_rate": 9.999999999999999e-05,
+      "loss": 0.8499,
+      "step": 945
+    },
+    {
+      "epoch": 2.098724348308375,
+      "grad_norm": 0.4148792028427124,
+      "learning_rate": 9.975308641975307e-05,
+      "loss": 0.6845,
+      "step": 946
+    },
+    {
+      "epoch": 2.100942872989462,
+      "grad_norm": 0.36994293332099915,
+      "learning_rate": 9.950617283950617e-05,
+      "loss": 0.6734,
+      "step": 947
+    },
+    {
+      "epoch": 2.103161397670549,
+      "grad_norm": 0.40943992137908936,
+      "learning_rate": 9.925925925925925e-05,
+      "loss": 0.6856,
+      "step": 948
+    },
+    {
+      "epoch": 2.105379922351636,
+      "grad_norm": 0.31637802720069885,
+      "learning_rate": 9.901234567901234e-05,
+      "loss": 0.6327,
+      "step": 949
+    },
+    {
+      "epoch": 2.107598447032723,
+      "grad_norm": 0.32384446263313293,
+      "learning_rate": 9.876543209876543e-05,
+      "loss": 0.7136,
+      "step": 950
+    },
+    {
+      "epoch": 2.10981697171381,
+      "grad_norm": 0.3463316261768341,
+      "learning_rate": 9.851851851851851e-05,
+      "loss": 0.6732,
+      "step": 951
+    },
+    {
+      "epoch": 2.1120354963948973,
+      "grad_norm": 0.3144516944885254,
+      "learning_rate": 9.827160493827159e-05,
+      "loss": 0.5973,
+      "step": 952
+    },
+    {
+      "epoch": 2.1142540210759844,
+      "grad_norm": 0.4218892455101013,
+      "learning_rate": 9.802469135802467e-05,
+      "loss": 0.7083,
+      "step": 953
+    },
+    {
+      "epoch": 2.1164725457570714,
+      "grad_norm": 0.32757800817489624,
+      "learning_rate": 9.777777777777778e-05,
+      "loss": 0.7467,
+      "step": 954
+    },
+    {
+      "epoch": 2.1186910704381585,
+      "grad_norm": 0.3414016366004944,
+      "learning_rate": 9.753086419753086e-05,
+      "loss": 0.6019,
+      "step": 955
+    },
+    {
+      "epoch": 2.1209095951192456,
+      "grad_norm": 0.35521650314331055,
+      "learning_rate": 9.728395061728394e-05,
+      "loss": 0.7681,
+      "step": 956
+    },
+    {
+      "epoch": 2.1231281198003327,
+      "grad_norm": 0.3359363377094269,
+      "learning_rate": 9.703703703703702e-05,
+      "loss": 0.8535,
+      "step": 957
+    },
+    {
+      "epoch": 2.1253466444814197,
+      "grad_norm": 0.2991136610507965,
+      "learning_rate": 9.679012345679012e-05,
+      "loss": 0.6931,
+      "step": 958
+    },
+    {
+      "epoch": 2.127565169162507,
+      "grad_norm": 0.43483608961105347,
+      "learning_rate": 9.65432098765432e-05,
+      "loss": 0.658,
+      "step": 959
+    },
+    {
+      "epoch": 2.129783693843594,
+      "grad_norm": 0.36569491028785706,
+      "learning_rate": 9.629629629629628e-05,
+      "loss": 0.9663,
+      "step": 960
+    },
+    {
+      "epoch": 2.132002218524681,
+      "grad_norm": 0.3341977894306183,
+      "learning_rate": 9.604938271604936e-05,
+      "loss": 0.5969,
+      "step": 961
+    },
+    {
+      "epoch": 2.134220743205768,
+      "grad_norm": 0.4225453734397888,
+      "learning_rate": 9.580246913580247e-05,
+      "loss": 0.9784,
+      "step": 962
+    },
+    {
+      "epoch": 2.136439267886855,
+      "grad_norm": 0.4027768671512604,
+      "learning_rate": 9.555555555555555e-05,
+      "loss": 0.7403,
+      "step": 963
+    },
+    {
+      "epoch": 2.138657792567942,
+      "grad_norm": 0.34804868698120117,
+      "learning_rate": 9.530864197530864e-05,
+      "loss": 0.7218,
+      "step": 964
+    },
+    {
+      "epoch": 2.1408763172490293,
+      "grad_norm": 0.3041076064109802,
+      "learning_rate": 9.506172839506172e-05,
+      "loss": 0.7794,
+      "step": 965
+    },
+    {
+      "epoch": 2.1430948419301163,
+      "grad_norm": 0.3609686493873596,
+      "learning_rate": 9.48148148148148e-05,
+      "loss": 0.7747,
+      "step": 966
+    },
+    {
+      "epoch": 2.1453133666112034,
+      "grad_norm": 0.3663080334663391,
+      "learning_rate": 9.45679012345679e-05,
+      "loss": 0.8099,
+      "step": 967
+    },
+    {
+      "epoch": 2.1475318912922905,
+      "grad_norm": 0.293468713760376,
+      "learning_rate": 9.432098765432098e-05,
+      "loss": 0.5885,
+      "step": 968
+    },
+    {
+      "epoch": 2.1497504159733776,
+      "grad_norm": 0.3324252665042877,
+      "learning_rate": 9.407407407407407e-05,
+      "loss": 0.7715,
+      "step": 969
+    },
+    {
+      "epoch": 2.1519689406544646,
+      "grad_norm": 0.3613545894622803,
+      "learning_rate": 9.382716049382715e-05,
+      "loss": 0.759,
+      "step": 970
+    },
+    {
+      "epoch": 2.1541874653355517,
+      "grad_norm": 0.47570472955703735,
+      "learning_rate": 9.358024691358025e-05,
+      "loss": 0.4884,
+      "step": 971
+    },
+    {
+      "epoch": 2.156405990016639,
+      "grad_norm": 0.3041134476661682,
+      "learning_rate": 9.333333333333333e-05,
+      "loss": 0.5472,
+      "step": 972
+    },
+    {
+      "epoch": 2.158624514697726,
+      "grad_norm": 0.3073522746562958,
+      "learning_rate": 9.308641975308641e-05,
+      "loss": 0.6938,
+      "step": 973
+    },
+    {
+      "epoch": 2.160843039378813,
+      "grad_norm": 0.3302704393863678,
+      "learning_rate": 9.283950617283949e-05,
+      "loss": 0.7356,
+      "step": 974
+    },
+    {
+      "epoch": 2.1630615640599,
+      "grad_norm": 0.2714685797691345,
+      "learning_rate": 9.259259259259257e-05,
+      "loss": 0.673,
+      "step": 975
+    },
+    {
+      "epoch": 2.165280088740987,
+      "grad_norm": 0.32952073216438293,
+      "learning_rate": 9.234567901234567e-05,
+      "loss": 0.6415,
+      "step": 976
+    },
+    {
+      "epoch": 2.167498613422074,
+      "grad_norm": 0.2673444151878357,
+      "learning_rate": 9.209876543209876e-05,
+      "loss": 0.4356,
+      "step": 977
+    },
+    {
+      "epoch": 2.1697171381031612,
+      "grad_norm": 0.3107616603374481,
+      "learning_rate": 9.185185185185185e-05,
+      "loss": 0.6125,
+      "step": 978
+    },
+    {
+      "epoch": 2.1719356627842483,
+      "grad_norm": 0.32444843649864197,
+      "learning_rate": 9.160493827160493e-05,
+      "loss": 0.9993,
+      "step": 979
+    },
+    {
+      "epoch": 2.1741541874653354,
+      "grad_norm": 0.4467172622680664,
+      "learning_rate": 9.135802469135802e-05,
+      "loss": 0.7594,
+      "step": 980
+    },
+    {
+      "epoch": 2.1763727121464225,
+      "grad_norm": 0.3748447597026825,
+      "learning_rate": 9.11111111111111e-05,
+      "loss": 0.5858,
+      "step": 981
+    },
+    {
+      "epoch": 2.1785912368275095,
+      "grad_norm": 0.3629694879055023,
+      "learning_rate": 9.086419753086419e-05,
+      "loss": 0.5798,
+      "step": 982
+    },
+    {
+      "epoch": 2.1808097615085966,
+      "grad_norm": 0.39069244265556335,
+      "learning_rate": 9.061728395061727e-05,
+      "loss": 0.9651,
+      "step": 983
+    },
+    {
+      "epoch": 2.1830282861896837,
+      "grad_norm": 0.2800096869468689,
+      "learning_rate": 9.037037037037038e-05,
+      "loss": 0.5838,
+      "step": 984
+    },
+    {
+      "epoch": 2.1852468108707708,
+      "grad_norm": 0.3680480122566223,
+      "learning_rate": 9.012345679012346e-05,
+      "loss": 0.6811,
+      "step": 985
+    },
+    {
+      "epoch": 2.187465335551858,
+      "grad_norm": 0.3076974153518677,
+      "learning_rate": 8.987654320987654e-05,
+      "loss": 0.7052,
+      "step": 986
+    },
+    {
+      "epoch": 2.189683860232945,
+      "grad_norm": 0.3076222836971283,
+      "learning_rate": 8.962962962962962e-05,
+      "loss": 0.4577,
+      "step": 987
+    },
+    {
+      "epoch": 2.191902384914032,
+      "grad_norm": 0.3724336326122284,
+      "learning_rate": 8.93827160493827e-05,
+      "loss": 0.6319,
+      "step": 988
+    },
+    {
+      "epoch": 2.194120909595119,
+      "grad_norm": 0.3186887800693512,
+      "learning_rate": 8.91358024691358e-05,
+      "loss": 0.5808,
+      "step": 989
+    },
+    {
+      "epoch": 2.196339434276206,
+      "grad_norm": 0.3216995596885681,
+      "learning_rate": 8.888888888888888e-05,
+      "loss": 0.6164,
+      "step": 990
+    },
+    {
+      "epoch": 2.1985579589572932,
+      "grad_norm": 0.3395521640777588,
+      "learning_rate": 8.864197530864196e-05,
+      "loss": 0.629,
+      "step": 991
+    },
+    {
+      "epoch": 2.2007764836383803,
+      "grad_norm": 0.2792010009288788,
+      "learning_rate": 8.839506172839506e-05,
+      "loss": 0.5596,
+      "step": 992
+    },
+    {
+      "epoch": 2.2029950083194674,
+      "grad_norm": 0.34687718749046326,
+      "learning_rate": 8.814814814814815e-05,
+      "loss": 1.1191,
+      "step": 993
+    },
+    {
+      "epoch": 2.2052135330005544,
+      "grad_norm": 0.37574487924575806,
+      "learning_rate": 8.790123456790123e-05,
+      "loss": 0.5286,
+      "step": 994
+    },
+    {
+      "epoch": 2.2074320576816415,
+      "grad_norm": 0.30812448263168335,
+      "learning_rate": 8.765432098765431e-05,
+      "loss": 0.5201,
+      "step": 995
+    },
+    {
+      "epoch": 2.2096505823627286,
+      "grad_norm": 0.40352487564086914,
+      "learning_rate": 8.74074074074074e-05,
+      "loss": 0.8149,
+      "step": 996
+    },
+    {
+      "epoch": 2.2118691070438157,
+      "grad_norm": 0.2878148555755615,
+      "learning_rate": 8.716049382716048e-05,
+      "loss": 0.5824,
+      "step": 997
+    },
+    {
+      "epoch": 2.2140876317249027,
+      "grad_norm": 0.3396584093570709,
+      "learning_rate": 8.691358024691357e-05,
+      "loss": 0.709,
+      "step": 998
+    },
+    {
+      "epoch": 2.21630615640599,
+      "grad_norm": 0.36563000082969666,
+      "learning_rate": 8.666666666666665e-05,
+      "loss": 0.4879,
+      "step": 999
+    },
+    {
+      "epoch": 2.218524681087077,
+      "grad_norm": 0.4133862853050232,
+      "learning_rate": 8.641975308641975e-05,
+      "loss": 0.6586,
+      "step": 1000
+    },
+    {
+      "epoch": 2.220743205768164,
+      "grad_norm": 0.4625363051891327,
+      "learning_rate": 8.617283950617283e-05,
+      "loss": 0.6783,
+      "step": 1001
+    },
+    {
+      "epoch": 2.222961730449251,
+      "grad_norm": 0.3060675263404846,
+      "learning_rate": 8.592592592592593e-05,
+      "loss": 0.7208,
+      "step": 1002
+    },
+    {
+      "epoch": 2.225180255130338,
+      "grad_norm": 0.29190897941589355,
+      "learning_rate": 8.567901234567901e-05,
+      "loss": 0.6016,
+      "step": 1003
+    },
+    {
+      "epoch": 2.227398779811425,
+      "grad_norm": 0.39983999729156494,
+      "learning_rate": 8.543209876543209e-05,
+      "loss": 0.5774,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2296173044925123,
+      "grad_norm": 0.36002904176712036,
+      "learning_rate": 8.518518518518517e-05,
+      "loss": 0.6914,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2318358291735994,
+      "grad_norm": 0.36424368619918823,
+      "learning_rate": 8.493827160493825e-05,
+      "loss": 0.7596,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2340543538546864,
+      "grad_norm": 0.37448421120643616,
+      "learning_rate": 8.469135802469136e-05,
+      "loss": 0.556,
+      "step": 1007
+    },
+    {
+      "epoch": 2.2362728785357735,
+      "grad_norm": 0.34816887974739075,
+      "learning_rate": 8.444444444444444e-05,
+      "loss": 0.7636,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2384914032168606,
+      "grad_norm": 0.3246499300003052,
+      "learning_rate": 8.419753086419752e-05,
+      "loss": 0.7171,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2407099278979477,
+      "grad_norm": 0.3131142854690552,
+      "learning_rate": 8.39506172839506e-05,
+      "loss": 0.5859,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2429284525790347,
+      "grad_norm": 0.3964233696460724,
+      "learning_rate": 8.37037037037037e-05,
+      "loss": 0.5967,
+      "step": 1011
+    },
+    {
+      "epoch": 2.245146977260122,
+      "grad_norm": 0.30504992604255676,
+      "learning_rate": 8.345679012345678e-05,
+      "loss": 0.5538,
+      "step": 1012
+    },
+    {
+      "epoch": 2.247365501941209,
+      "grad_norm": 0.46755149960517883,
+      "learning_rate": 8.320987654320986e-05,
+      "loss": 0.6802,
+      "step": 1013
+    },
+    {
+      "epoch": 2.249584026622296,
+      "grad_norm": 0.3534623086452484,
+      "learning_rate": 8.296296296296295e-05,
+      "loss": 0.8367,
+      "step": 1014
+    },
+    {
+      "epoch": 2.251802551303383,
+      "grad_norm": 0.330941766500473,
+      "learning_rate": 8.271604938271605e-05,
+      "loss": 0.8372,
+      "step": 1015
+    },
+    {
+      "epoch": 2.25402107598447,
+      "grad_norm": 0.3522794544696808,
+      "learning_rate": 8.246913580246914e-05,
+      "loss": 0.7413,
+      "step": 1016
+    },
+    {
+      "epoch": 2.256239600665557,
+      "grad_norm": 0.35905721783638,
+      "learning_rate": 8.222222222222222e-05,
+      "loss": 0.8538,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2584581253466443,
+      "grad_norm": 0.3598293364048004,
+      "learning_rate": 8.19753086419753e-05,
+      "loss": 0.918,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2606766500277313,
+      "grad_norm": 0.33987298607826233,
+      "learning_rate": 8.172839506172838e-05,
+      "loss": 0.8135,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2628951747088184,
+      "grad_norm": 0.405911922454834,
+      "learning_rate": 8.148148148148148e-05,
+      "loss": 0.7815,
+      "step": 1020
+    },
+    {
+      "epoch": 2.2651136993899055,
+      "grad_norm": 0.36983832716941833,
+      "learning_rate": 8.123456790123456e-05,
+      "loss": 0.7588,
+      "step": 1021
+    },
+    {
+      "epoch": 2.2673322240709926,
+      "grad_norm": 0.42615360021591187,
+      "learning_rate": 8.098765432098765e-05,
+      "loss": 0.6601,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2695507487520796,
+      "grad_norm": 0.32522353529930115,
+      "learning_rate": 8.074074074074073e-05,
+      "loss": 0.7458,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2717692734331667,
+      "grad_norm": 0.39460816979408264,
+      "learning_rate": 8.049382716049383e-05,
+      "loss": 0.7396,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2739877981142538,
+      "grad_norm": 0.26139745116233826,
+      "learning_rate": 8.024691358024691e-05,
+      "loss": 0.6766,
+      "step": 1025
+    },
+    {
+      "epoch": 2.276206322795341,
+      "grad_norm": 0.4569029211997986,
+      "learning_rate": 7.999999999999999e-05,
+      "loss": 0.7269,
+      "step": 1026
+    },
+    {
+      "epoch": 2.278424847476428,
+      "grad_norm": 0.4914579391479492,
+      "learning_rate": 7.975308641975307e-05,
+      "loss": 0.699,
+      "step": 1027
+    },
+    {
+      "epoch": 2.280643372157515,
+      "grad_norm": 0.38146674633026123,
+      "learning_rate": 7.950617283950616e-05,
+      "loss": 0.5836,
+      "step": 1028
+    },
+    {
+      "epoch": 2.2828618968386025,
+      "grad_norm": 0.3164074420928955,
+      "learning_rate": 7.925925925925925e-05,
+      "loss": 0.6519,
+      "step": 1029
+    },
+    {
+      "epoch": 2.2850804215196896,
+      "grad_norm": 0.3870079815387726,
+      "learning_rate": 7.901234567901235e-05,
+      "loss": 0.8724,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2872989462007767,
+      "grad_norm": 0.40606939792633057,
+      "learning_rate": 7.876543209876543e-05,
+      "loss": 0.7741,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2895174708818637,
+      "grad_norm": 0.29621872305870056,
+      "learning_rate": 7.851851851851851e-05,
+      "loss": 0.7245,
+      "step": 1032
+    },
+    {
+      "epoch": 2.291735995562951,
+      "grad_norm": 0.3516596555709839,
+      "learning_rate": 7.82716049382716e-05,
+      "loss": 0.5376,
+      "step": 1033
+    },
+    {
+      "epoch": 2.293954520244038,
+      "grad_norm": 0.3870883285999298,
+      "learning_rate": 7.802469135802469e-05,
+      "loss": 0.8126,
+      "step": 1034
+    },
+    {
+      "epoch": 2.296173044925125,
+      "grad_norm": 0.2679266333580017,
+      "learning_rate": 7.777777777777777e-05,
+      "loss": 0.5435,
+      "step": 1035
+    },
+    {
+      "epoch": 2.298391569606212,
+      "grad_norm": 0.37198925018310547,
+      "learning_rate": 7.753086419753085e-05,
+      "loss": 0.6573,
+      "step": 1036
+    },
+    {
+      "epoch": 2.300610094287299,
+      "grad_norm": 0.30731192231178284,
+      "learning_rate": 7.728395061728396e-05,
+      "loss": 0.5732,
+      "step": 1037
+    },
+    {
+      "epoch": 2.302828618968386,
+      "grad_norm": 0.303986132144928,
+      "learning_rate": 7.703703703703704e-05,
+      "loss": 0.7707,
+      "step": 1038
+    },
+    {
+      "epoch": 2.3050471436494733,
+      "grad_norm": 0.34620535373687744,
+      "learning_rate": 7.679012345679012e-05,
+      "loss": 0.6496,
+      "step": 1039
+    },
+    {
+      "epoch": 2.3072656683305603,
+      "grad_norm": 0.3175440728664398,
+      "learning_rate": 7.65432098765432e-05,
+      "loss": 0.6434,
+      "step": 1040
+    },
+    {
+      "epoch": 2.3094841930116474,
+      "grad_norm": 0.34240737557411194,
+      "learning_rate": 7.629629629629628e-05,
+      "loss": 0.8621,
+      "step": 1041
+    },
+    {
+      "epoch": 2.3117027176927345,
+      "grad_norm": 0.3621320426464081,
+      "learning_rate": 7.604938271604937e-05,
+      "loss": 0.6733,
+      "step": 1042
+    },
+    {
+      "epoch": 2.3139212423738216,
+      "grad_norm": 0.44679510593414307,
+      "learning_rate": 7.580246913580246e-05,
+      "loss": 0.6691,
+      "step": 1043
+    },
+    {
+      "epoch": 2.3161397670549086,
+      "grad_norm": 0.32941627502441406,
+      "learning_rate": 7.555555555555554e-05,
+      "loss": 0.6376,
+      "step": 1044
+    },
+    {
+      "epoch": 2.3183582917359957,
+      "grad_norm": 0.4651550054550171,
+      "learning_rate": 7.530864197530864e-05,
+      "loss": 0.6882,
+      "step": 1045
+    },
+    {
+      "epoch": 2.320576816417083,
+      "grad_norm": 0.3475878834724426,
+      "learning_rate": 7.506172839506173e-05,
+      "loss": 0.8214,
+      "step": 1046
+    },
+    {
+      "epoch": 2.32279534109817,
+      "grad_norm": 0.3238202929496765,
+      "learning_rate": 7.481481481481481e-05,
+      "loss": 0.6053,
+      "step": 1047
+    },
+    {
+      "epoch": 2.325013865779257,
+      "grad_norm": 0.3107715845108032,
+      "learning_rate": 7.45679012345679e-05,
+      "loss": 0.6556,
+      "step": 1048
+    },
+    {
+      "epoch": 2.327232390460344,
+      "grad_norm": 0.49945971369743347,
+      "learning_rate": 7.432098765432098e-05,
+      "loss": 0.7583,
+      "step": 1049
+    },
+    {
+      "epoch": 2.329450915141431,
+      "grad_norm": 0.3475596606731415,
+      "learning_rate": 7.407407407407406e-05,
+      "loss": 0.7105,
+      "step": 1050
+    },
+    {
+      "epoch": 2.331669439822518,
+      "grad_norm": 0.3444492816925049,
+      "learning_rate": 7.382716049382715e-05,
+      "loss": 0.6782,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3338879645036053,
+      "grad_norm": 0.34761983156204224,
+      "learning_rate": 7.358024691358024e-05,
+      "loss": 0.5447,
+      "step": 1052
+    },
+    {
+      "epoch": 2.3361064891846923,
+      "grad_norm": 0.38653799891471863,
+      "learning_rate": 7.333333333333332e-05,
+      "loss": 0.903,
+      "step": 1053
+    },
+    {
+      "epoch": 2.3383250138657794,
+      "grad_norm": 0.36463484168052673,
+      "learning_rate": 7.308641975308641e-05,
+      "loss": 0.702,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3405435385468665,
+      "grad_norm": 0.39046764373779297,
+      "learning_rate": 7.28395061728395e-05,
+      "loss": 0.6875,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3427620632279536,
+      "grad_norm": 0.3187321722507477,
+      "learning_rate": 7.259259259259259e-05,
+      "loss": 0.7419,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3449805879090406,
+      "grad_norm": 0.4015580713748932,
+      "learning_rate": 7.234567901234567e-05,
+      "loss": 0.6435,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3471991125901277,
+      "grad_norm": 0.3675442636013031,
+      "learning_rate": 7.209876543209877e-05,
+      "loss": 0.7006,
+      "step": 1058
+    },
+    {
+      "epoch": 2.3494176372712148,
+      "grad_norm": 0.3764418363571167,
+      "learning_rate": 7.185185185185185e-05,
+      "loss": 0.6821,
+      "step": 1059
+    },
+    {
+      "epoch": 2.351636161952302,
+      "grad_norm": 0.3968859016895294,
+      "learning_rate": 7.160493827160493e-05,
+      "loss": 0.6123,
+      "step": 1060
+    },
+    {
+      "epoch": 2.353854686633389,
+      "grad_norm": 0.3037530183792114,
+      "learning_rate": 7.135802469135801e-05,
+      "loss": 0.5963,
+      "step": 1061
+    },
+    {
+      "epoch": 2.356073211314476,
+      "grad_norm": 0.3054431080818176,
+      "learning_rate": 7.11111111111111e-05,
+      "loss": 0.6499,
+      "step": 1062
+    },
+    {
+      "epoch": 2.358291735995563,
+      "grad_norm": 0.3349511921405792,
+      "learning_rate": 7.086419753086419e-05,
+      "loss": 0.6187,
+      "step": 1063
+    },
+    {
+      "epoch": 2.36051026067665,
+      "grad_norm": 0.4058036506175995,
+      "learning_rate": 7.061728395061727e-05,
+      "loss": 0.7584,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3627287853577372,
+      "grad_norm": 0.39039862155914307,
+      "learning_rate": 7.037037037037036e-05,
+      "loss": 0.5986,
+      "step": 1065
+    },
+    {
+      "epoch": 2.3649473100388243,
+      "grad_norm": 0.2948319613933563,
+      "learning_rate": 7.012345679012345e-05,
+      "loss": 0.6139,
+      "step": 1066
+    },
+    {
+      "epoch": 2.3671658347199114,
+      "grad_norm": 0.4672812819480896,
+      "learning_rate": 6.987654320987654e-05,
+      "loss": 0.825,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3693843594009985,
+      "grad_norm": 0.38048335909843445,
+      "learning_rate": 6.962962962962962e-05,
+      "loss": 0.8687,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3716028840820855,
+      "grad_norm": 0.28383272886276245,
+      "learning_rate": 6.938271604938272e-05,
+      "loss": 0.5008,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3738214087631726,
+      "grad_norm": 0.43459123373031616,
+      "learning_rate": 6.91358024691358e-05,
+      "loss": 0.6115,
+      "step": 1070
+    },
+    {
+      "epoch": 2.3760399334442597,
+      "grad_norm": 0.37523216009140015,
+      "learning_rate": 6.888888888888888e-05,
+      "loss": 0.8478,
+      "step": 1071
+    },
+    {
+      "epoch": 2.3782584581253468,
+      "grad_norm": 0.3041827380657196,
+      "learning_rate": 6.864197530864196e-05,
+      "loss": 0.7552,
+      "step": 1072
+    },
+    {
+      "epoch": 2.380476982806434,
+      "grad_norm": 0.3452499806880951,
+      "learning_rate": 6.839506172839506e-05,
+      "loss": 0.773,
+      "step": 1073
+    },
+    {
+      "epoch": 2.382695507487521,
+      "grad_norm": 0.34864675998687744,
+      "learning_rate": 6.814814814814814e-05,
+      "loss": 0.8391,
+      "step": 1074
+    },
+    {
+      "epoch": 2.384914032168608,
+      "grad_norm": 0.39141833782196045,
+      "learning_rate": 6.790123456790122e-05,
+      "loss": 0.5627,
+      "step": 1075
+    },
+    {
+      "epoch": 2.387132556849695,
+      "grad_norm": 0.4450438618659973,
+      "learning_rate": 6.765432098765432e-05,
+      "loss": 0.6031,
+      "step": 1076
+    },
+    {
+      "epoch": 2.389351081530782,
+      "grad_norm": 0.32962727546691895,
+      "learning_rate": 6.74074074074074e-05,
+      "loss": 0.8934,
+      "step": 1077
+    },
+    {
+      "epoch": 2.391569606211869,
+      "grad_norm": 0.41162562370300293,
+      "learning_rate": 6.716049382716049e-05,
+      "loss": 0.9187,
+      "step": 1078
+    },
+    {
+      "epoch": 2.3937881308929563,
+      "grad_norm": 0.4733867943286896,
+      "learning_rate": 6.691358024691357e-05,
+      "loss": 0.6872,
+      "step": 1079
+    },
+    {
+      "epoch": 2.3960066555740434,
+      "grad_norm": 0.33442676067352295,
+      "learning_rate": 6.666666666666666e-05,
+      "loss": 0.6079,
+      "step": 1080
+    },
+    {
+      "epoch": 2.3982251802551304,
+      "grad_norm": 0.35997340083122253,
+      "learning_rate": 6.641975308641975e-05,
+      "loss": 0.7141,
+      "step": 1081
+    },
+    {
+      "epoch": 2.4004437049362175,
+      "grad_norm": 0.37527135014533997,
+      "learning_rate": 6.617283950617283e-05,
+      "loss": 0.6404,
+      "step": 1082
+    },
+    {
+      "epoch": 2.4026622296173046,
+      "grad_norm": 0.2717944085597992,
+      "learning_rate": 6.592592592592591e-05,
+      "loss": 0.6928,
+      "step": 1083
+    },
+    {
+      "epoch": 2.4048807542983917,
+      "grad_norm": 0.36003822088241577,
+      "learning_rate": 6.567901234567901e-05,
+      "loss": 0.6252,
+      "step": 1084
+    },
+    {
+      "epoch": 2.4070992789794787,
+      "grad_norm": 0.37141597270965576,
+      "learning_rate": 6.543209876543209e-05,
+      "loss": 0.5406,
+      "step": 1085
+    },
+    {
+      "epoch": 2.409317803660566,
+      "grad_norm": 0.3468698263168335,
+      "learning_rate": 6.518518518518517e-05,
+      "loss": 0.5818,
+      "step": 1086
+    },
+    {
+      "epoch": 2.411536328341653,
+      "grad_norm": 0.3632766902446747,
+      "learning_rate": 6.493827160493827e-05,
+      "loss": 0.6013,
+      "step": 1087
+    },
+    {
+      "epoch": 2.41375485302274,
+      "grad_norm": 0.3762660026550293,
+      "learning_rate": 6.469135802469135e-05,
+      "loss": 0.5937,
+      "step": 1088
+    },
+    {
+      "epoch": 2.415973377703827,
+      "grad_norm": 0.3033027946949005,
+      "learning_rate": 6.444444444444444e-05,
+      "loss": 0.5853,
+      "step": 1089
+    },
+    {
+      "epoch": 2.418191902384914,
+      "grad_norm": 0.35676154494285583,
+      "learning_rate": 6.419753086419753e-05,
+      "loss": 0.674,
+      "step": 1090
+    },
+    {
+      "epoch": 2.420410427066001,
+      "grad_norm": 0.30847400426864624,
+      "learning_rate": 6.395061728395061e-05,
+      "loss": 0.6831,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4226289517470883,
+      "grad_norm": 0.3360093832015991,
+      "learning_rate": 6.37037037037037e-05,
+      "loss": 0.5932,
+      "step": 1092
+    },
+    {
+      "epoch": 2.4248474764281753,
+      "grad_norm": 0.3796274960041046,
+      "learning_rate": 6.345679012345678e-05,
+      "loss": 0.653,
+      "step": 1093
+    },
+    {
+      "epoch": 2.4270660011092624,
+      "grad_norm": 0.42600107192993164,
+      "learning_rate": 6.320987654320987e-05,
+      "loss": 0.8228,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4292845257903495,
+      "grad_norm": 0.32950371503829956,
+      "learning_rate": 6.296296296296295e-05,
+      "loss": 0.5308,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4315030504714366,
+      "grad_norm": 0.3656352758407593,
+      "learning_rate": 6.271604938271604e-05,
+      "loss": 0.7284,
+      "step": 1096
+    },
+    {
+      "epoch": 2.4337215751525236,
+      "grad_norm": 0.3264632225036621,
+      "learning_rate": 6.246913580246912e-05,
+      "loss": 0.6587,
+      "step": 1097
+    },
+    {
+      "epoch": 2.4359400998336107,
+      "grad_norm": 0.3639947175979614,
+      "learning_rate": 6.222222222222222e-05,
+      "loss": 0.6732,
+      "step": 1098
+    },
+    {
+      "epoch": 2.438158624514698,
+      "grad_norm": 0.32109498977661133,
+      "learning_rate": 6.19753086419753e-05,
+      "loss": 0.5782,
+      "step": 1099
+    },
+    {
+      "epoch": 2.440377149195785,
+      "grad_norm": 0.2942397892475128,
+      "learning_rate": 6.17283950617284e-05,
+      "loss": 0.5964,
+      "step": 1100
+    },
+    {
+      "epoch": 2.442595673876872,
+      "grad_norm": 0.3522522747516632,
+      "learning_rate": 6.148148148148148e-05,
+      "loss": 0.6497,
+      "step": 1101
+    },
+    {
+      "epoch": 2.444814198557959,
+      "grad_norm": 0.3488999605178833,
+      "learning_rate": 6.123456790123456e-05,
+      "loss": 0.6767,
+      "step": 1102
+    },
+    {
+      "epoch": 2.447032723239046,
+      "grad_norm": 0.3627452254295349,
+      "learning_rate": 6.0987654320987655e-05,
+      "loss": 0.728,
+      "step": 1103
+    },
+    {
+      "epoch": 2.449251247920133,
+      "grad_norm": 0.41831135749816895,
+      "learning_rate": 6.0740740740740736e-05,
+      "loss": 0.4978,
+      "step": 1104
+    },
+    {
+      "epoch": 2.4514697726012202,
+      "grad_norm": 0.31789958477020264,
+      "learning_rate": 6.049382716049382e-05,
+      "loss": 0.5682,
+      "step": 1105
+    },
+    {
+      "epoch": 2.4536882972823073,
+      "grad_norm": 0.35635948181152344,
+      "learning_rate": 6.0246913580246906e-05,
+      "loss": 0.6238,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4559068219633944,
+      "grad_norm": 0.47133147716522217,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 0.6863,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4581253466444815,
+      "grad_norm": 0.31902948021888733,
+      "learning_rate": 5.975308641975308e-05,
+      "loss": 0.5635,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4603438713255685,
+      "grad_norm": 0.433309406042099,
+      "learning_rate": 5.9506172839506165e-05,
+      "loss": 0.4396,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4625623960066556,
+      "grad_norm": 0.4524514079093933,
+      "learning_rate": 5.925925925925925e-05,
+      "loss": 0.6356,
+      "step": 1110
+    },
+    {
+      "epoch": 2.4647809206877427,
+      "grad_norm": 0.37966808676719666,
+      "learning_rate": 5.901234567901234e-05,
+      "loss": 0.6018,
+      "step": 1111
+    },
+    {
+      "epoch": 2.4669994453688298,
+      "grad_norm": 0.3257821500301361,
+      "learning_rate": 5.876543209876543e-05,
+      "loss": 0.5915,
+      "step": 1112
+    },
+    {
+      "epoch": 2.469217970049917,
+      "grad_norm": 0.48562905192375183,
+      "learning_rate": 5.851851851851851e-05,
+      "loss": 0.6302,
+      "step": 1113
+    },
+    {
+      "epoch": 2.471436494731004,
+      "grad_norm": 0.3250422477722168,
+      "learning_rate": 5.827160493827159e-05,
+      "loss": 0.6747,
+      "step": 1114
+    },
+    {
+      "epoch": 2.473655019412091,
+      "grad_norm": 0.36961784958839417,
+      "learning_rate": 5.802469135802469e-05,
+      "loss": 0.5736,
+      "step": 1115
+    },
+    {
+      "epoch": 2.475873544093178,
+      "grad_norm": 0.5119171738624573,
+      "learning_rate": 5.777777777777777e-05,
+      "loss": 0.8694,
+      "step": 1116
+    },
+    {
+      "epoch": 2.478092068774265,
+      "grad_norm": 0.37949636578559875,
+      "learning_rate": 5.753086419753086e-05,
+      "loss": 0.6805,
+      "step": 1117
+    },
+    {
+      "epoch": 2.480310593455352,
+      "grad_norm": 0.397210955619812,
+      "learning_rate": 5.7283950617283946e-05,
+      "loss": 0.5761,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4825291181364393,
+      "grad_norm": 0.3840007185935974,
+      "learning_rate": 5.7037037037037035e-05,
+      "loss": 0.6796,
+      "step": 1119
+    },
+    {
+      "epoch": 2.4847476428175264,
+      "grad_norm": 0.3360983729362488,
+      "learning_rate": 5.6790123456790116e-05,
+      "loss": 0.7156,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4869661674986134,
+      "grad_norm": 0.4211987555027008,
+      "learning_rate": 5.6543209876543205e-05,
+      "loss": 0.651,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4891846921797005,
+      "grad_norm": 0.36395126581192017,
+      "learning_rate": 5.629629629629629e-05,
+      "loss": 0.7818,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4914032168607876,
+      "grad_norm": 0.4043983519077301,
+      "learning_rate": 5.604938271604938e-05,
+      "loss": 0.6381,
+      "step": 1123
+    },
+    {
+      "epoch": 2.4936217415418747,
+      "grad_norm": 0.3778453469276428,
+      "learning_rate": 5.580246913580246e-05,
+      "loss": 0.7598,
+      "step": 1124
+    },
+    {
+      "epoch": 2.4958402662229617,
+      "grad_norm": 0.311799556016922,
+      "learning_rate": 5.5555555555555545e-05,
+      "loss": 0.7919,
+      "step": 1125
+    },
+    {
+      "epoch": 2.498058790904049,
+      "grad_norm": 0.3689611554145813,
+      "learning_rate": 5.530864197530864e-05,
+      "loss": 0.6851,
+      "step": 1126
+    },
+    {
+      "epoch": 2.500277315585136,
+      "grad_norm": 0.32257241010665894,
+      "learning_rate": 5.506172839506172e-05,
+      "loss": 0.5531,
+      "step": 1127
+    },
+    {
+      "epoch": 2.502495840266223,
+      "grad_norm": 0.44063225388526917,
+      "learning_rate": 5.481481481481481e-05,
+      "loss": 0.6448,
+      "step": 1128
+    },
+    {
+      "epoch": 2.50471436494731,
+      "grad_norm": 0.3749827444553375,
+      "learning_rate": 5.456790123456789e-05,
+      "loss": 0.7515,
+      "step": 1129
+    },
+    {
+      "epoch": 2.506932889628397,
+      "grad_norm": 0.4113692343235016,
+      "learning_rate": 5.4320987654320986e-05,
+      "loss": 0.6705,
+      "step": 1130
+    },
+    {
+      "epoch": 2.509151414309484,
+      "grad_norm": 0.4451723098754883,
+      "learning_rate": 5.407407407407407e-05,
+      "loss": 0.6284,
+      "step": 1131
+    },
+    {
+      "epoch": 2.5113699389905713,
+      "grad_norm": 0.32581984996795654,
+      "learning_rate": 5.3827160493827156e-05,
+      "loss": 0.7726,
+      "step": 1132
+    },
+    {
+      "epoch": 2.5135884636716583,
+      "grad_norm": 0.34149596095085144,
+      "learning_rate": 5.358024691358024e-05,
+      "loss": 0.5945,
+      "step": 1133
+    },
+    {
+      "epoch": 2.5158069883527454,
+      "grad_norm": 0.36342647671699524,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.8107,
+      "step": 1134
+    },
+    {
+      "epoch": 2.5180255130338325,
+      "grad_norm": 0.42264389991760254,
+      "learning_rate": 5.3086419753086415e-05,
+      "loss": 0.5343,
+      "step": 1135
+    },
+    {
+      "epoch": 2.5202440377149196,
+      "grad_norm": 0.523398756980896,
+      "learning_rate": 5.2839506172839496e-05,
+      "loss": 0.7595,
+      "step": 1136
+    },
+    {
+      "epoch": 2.5224625623960066,
+      "grad_norm": 0.3850318193435669,
+      "learning_rate": 5.259259259259259e-05,
+      "loss": 0.7257,
+      "step": 1137
+    },
+    {
+      "epoch": 2.5246810870770937,
+      "grad_norm": 0.37276557087898254,
+      "learning_rate": 5.234567901234567e-05,
+      "loss": 0.665,
+      "step": 1138
+    },
+    {
+      "epoch": 2.526899611758181,
+      "grad_norm": 0.4208609163761139,
+      "learning_rate": 5.209876543209876e-05,
+      "loss": 0.9216,
+      "step": 1139
+    },
+    {
+      "epoch": 2.529118136439268,
+      "grad_norm": 0.38741543889045715,
+      "learning_rate": 5.185185185185184e-05,
+      "loss": 0.6637,
+      "step": 1140
+    },
+    {
+      "epoch": 2.531336661120355,
+      "grad_norm": 0.4953962564468384,
+      "learning_rate": 5.160493827160494e-05,
+      "loss": 0.6325,
+      "step": 1141
+    },
+    {
+      "epoch": 2.533555185801442,
+      "grad_norm": 0.3442378640174866,
+      "learning_rate": 5.135802469135802e-05,
+      "loss": 0.658,
+      "step": 1142
+    },
+    {
+      "epoch": 2.535773710482529,
+      "grad_norm": 0.3987411856651306,
+      "learning_rate": 5.111111111111111e-05,
+      "loss": 0.8066,
+      "step": 1143
+    },
+    {
+      "epoch": 2.537992235163616,
+      "grad_norm": 0.49136677384376526,
+      "learning_rate": 5.086419753086419e-05,
+      "loss": 1.0484,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5402107598447032,
+      "grad_norm": 0.3194064497947693,
+      "learning_rate": 5.0617283950617285e-05,
+      "loss": 0.5841,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5424292845257903,
+      "grad_norm": 0.3506169021129608,
+      "learning_rate": 5.0370370370370366e-05,
+      "loss": 0.6848,
+      "step": 1146
+    },
+    {
+      "epoch": 2.5446478092068774,
+      "grad_norm": 0.44297873973846436,
+      "learning_rate": 5.012345679012345e-05,
+      "loss": 1.0249,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5468663338879645,
+      "grad_norm": 0.4493277668952942,
+      "learning_rate": 4.9876543209876536e-05,
+      "loss": 0.7373,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5490848585690515,
+      "grad_norm": 0.4456116259098053,
+      "learning_rate": 4.9629629629629625e-05,
+      "loss": 0.7053,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5513033832501386,
+      "grad_norm": 0.44051024317741394,
+      "learning_rate": 4.938271604938271e-05,
+      "loss": 0.6748,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5535219079312257,
+      "grad_norm": 0.3373473286628723,
+      "learning_rate": 4.9135802469135795e-05,
+      "loss": 0.6563,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5557404326123128,
+      "grad_norm": 0.34927576780319214,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.6583,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5579589572934,
+      "grad_norm": 0.3529128134250641,
+      "learning_rate": 4.864197530864197e-05,
+      "loss": 0.597,
+      "step": 1153
+    },
+    {
+      "epoch": 2.560177481974487,
+      "grad_norm": 0.38114437460899353,
+      "learning_rate": 4.839506172839506e-05,
+      "loss": 0.6222,
+      "step": 1154
+    },
+    {
+      "epoch": 2.562396006655574,
+      "grad_norm": 0.4912923276424408,
+      "learning_rate": 4.814814814814814e-05,
+      "loss": 0.7611,
+      "step": 1155
+    },
+    {
+      "epoch": 2.564614531336661,
+      "grad_norm": 0.37824904918670654,
+      "learning_rate": 4.7901234567901237e-05,
+      "loss": 0.6744,
+      "step": 1156
+    },
+    {
+      "epoch": 2.566833056017748,
+      "grad_norm": 0.35213157534599304,
+      "learning_rate": 4.765432098765432e-05,
+      "loss": 0.4892,
+      "step": 1157
+    },
+    {
+      "epoch": 2.569051580698835,
+      "grad_norm": 0.30457833409309387,
+      "learning_rate": 4.74074074074074e-05,
+      "loss": 0.7113,
+      "step": 1158
+    },
+    {
+      "epoch": 2.5712701053799223,
+      "grad_norm": 0.37088507413864136,
+      "learning_rate": 4.716049382716049e-05,
+      "loss": 0.7059,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5734886300610094,
+      "grad_norm": 0.4210187792778015,
+      "learning_rate": 4.6913580246913576e-05,
+      "loss": 0.6615,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5757071547420964,
+      "grad_norm": 0.33626729249954224,
+      "learning_rate": 4.6666666666666665e-05,
+      "loss": 0.7067,
+      "step": 1161
+    },
+    {
+      "epoch": 2.5779256794231835,
+      "grad_norm": 0.3614247441291809,
+      "learning_rate": 4.6419753086419746e-05,
+      "loss": 0.6298,
+      "step": 1162
+    },
+    {
+      "epoch": 2.5801442041042706,
+      "grad_norm": 0.37730133533477783,
+      "learning_rate": 4.6172839506172835e-05,
+      "loss": 0.8217,
+      "step": 1163
+    },
+    {
+      "epoch": 2.5823627287853577,
+      "grad_norm": 0.3866690695285797,
+      "learning_rate": 4.592592592592592e-05,
+      "loss": 0.7002,
+      "step": 1164
+    },
+    {
+      "epoch": 2.5845812534664447,
+      "grad_norm": 0.3395581543445587,
+      "learning_rate": 4.567901234567901e-05,
+      "loss": 0.5581,
+      "step": 1165
+    },
+    {
+      "epoch": 2.586799778147532,
+      "grad_norm": 0.294520765542984,
+      "learning_rate": 4.543209876543209e-05,
+      "loss": 0.5132,
+      "step": 1166
+    },
+    {
+      "epoch": 2.589018302828619,
+      "grad_norm": 0.4104853868484497,
+      "learning_rate": 4.518518518518519e-05,
+      "loss": 0.8177,
+      "step": 1167
+    },
+    {
+      "epoch": 2.591236827509706,
+      "grad_norm": 0.42259836196899414,
+      "learning_rate": 4.493827160493827e-05,
+      "loss": 0.6694,
+      "step": 1168
+    },
+    {
+      "epoch": 2.593455352190793,
+      "grad_norm": 0.3451855480670929,
+      "learning_rate": 4.469135802469135e-05,
+      "loss": 0.5781,
+      "step": 1169
+    },
+    {
+      "epoch": 2.59567387687188,
+      "grad_norm": 0.37934038043022156,
+      "learning_rate": 4.444444444444444e-05,
+      "loss": 0.6594,
+      "step": 1170
+    },
+    {
+      "epoch": 2.597892401552967,
+      "grad_norm": 0.3825890123844147,
+      "learning_rate": 4.419753086419753e-05,
+      "loss": 0.6771,
+      "step": 1171
+    },
+    {
+      "epoch": 2.6001109262340543,
+      "grad_norm": 0.3449947237968445,
+      "learning_rate": 4.3950617283950617e-05,
+      "loss": 0.7588,
+      "step": 1172
+    },
+    {
+      "epoch": 2.6023294509151413,
+      "grad_norm": 0.43323570489883423,
+      "learning_rate": 4.37037037037037e-05,
+      "loss": 0.7212,
+      "step": 1173
+    },
+    {
+      "epoch": 2.6045479755962284,
+      "grad_norm": 0.2862163186073303,
+      "learning_rate": 4.3456790123456786e-05,
+      "loss": 0.5954,
+      "step": 1174
+    },
+    {
+      "epoch": 2.6067665002773155,
+      "grad_norm": 0.4550541341304779,
+      "learning_rate": 4.3209876543209875e-05,
+      "loss": 0.6373,
+      "step": 1175
+    },
+    {
+      "epoch": 2.6089850249584026,
+      "grad_norm": 0.29838016629219055,
+      "learning_rate": 4.296296296296296e-05,
+      "loss": 0.6363,
+      "step": 1176
+    },
+    {
+      "epoch": 2.6112035496394896,
+      "grad_norm": 0.3695347309112549,
+      "learning_rate": 4.2716049382716045e-05,
+      "loss": 0.7622,
+      "step": 1177
+    },
+    {
+      "epoch": 2.6134220743205767,
+      "grad_norm": 0.3937763273715973,
+      "learning_rate": 4.2469135802469126e-05,
+      "loss": 0.7455,
+      "step": 1178
+    },
+    {
+      "epoch": 2.615640599001664,
+      "grad_norm": 0.3865770101547241,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.7528,
+      "step": 1179
+    },
+    {
+      "epoch": 2.617859123682751,
+      "grad_norm": 0.3427639901638031,
+      "learning_rate": 4.19753086419753e-05,
+      "loss": 0.7431,
+      "step": 1180
+    },
+    {
+      "epoch": 2.620077648363838,
+      "grad_norm": 0.3482280671596527,
+      "learning_rate": 4.172839506172839e-05,
+      "loss": 0.6407,
+      "step": 1181
+    },
+    {
+      "epoch": 2.622296173044925,
+      "grad_norm": 0.33790749311447144,
+      "learning_rate": 4.148148148148147e-05,
+      "loss": 0.7596,
+      "step": 1182
+    },
+    {
+      "epoch": 2.624514697726012,
+      "grad_norm": 0.366729199886322,
+      "learning_rate": 4.123456790123457e-05,
+      "loss": 0.4556,
+      "step": 1183
+    },
+    {
+      "epoch": 2.626733222407099,
+      "grad_norm": 0.4306209981441498,
+      "learning_rate": 4.098765432098765e-05,
+      "loss": 0.7033,
+      "step": 1184
+    },
+    {
+      "epoch": 2.6289517470881862,
+      "grad_norm": 0.38049623370170593,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.7882,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6311702717692733,
+      "grad_norm": 0.3654687702655792,
+      "learning_rate": 4.0493827160493827e-05,
+      "loss": 0.5655,
+      "step": 1186
+    },
+    {
+      "epoch": 2.6333887964503604,
+      "grad_norm": 0.43000659346580505,
+      "learning_rate": 4.0246913580246915e-05,
+      "loss": 0.8894,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6356073211314475,
+      "grad_norm": 0.35160139203071594,
+      "learning_rate": 3.9999999999999996e-05,
+      "loss": 0.531,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6378258458125345,
+      "grad_norm": 0.41925349831581116,
+      "learning_rate": 3.975308641975308e-05,
+      "loss": 0.6287,
+      "step": 1189
+    },
+    {
+      "epoch": 2.6400443704936216,
+      "grad_norm": 0.2996477484703064,
+      "learning_rate": 3.950617283950617e-05,
+      "loss": 0.5894,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6422628951747087,
+      "grad_norm": 0.3519132137298584,
+      "learning_rate": 3.9259259259259255e-05,
+      "loss": 0.5787,
+      "step": 1191
+    },
+    {
+      "epoch": 2.6444814198557958,
+      "grad_norm": 0.3597591817378998,
+      "learning_rate": 3.901234567901234e-05,
+      "loss": 0.739,
+      "step": 1192
+    },
+    {
+      "epoch": 2.646699944536883,
+      "grad_norm": 0.3572014272212982,
+      "learning_rate": 3.8765432098765425e-05,
+      "loss": 0.8411,
+      "step": 1193
+    },
+    {
+      "epoch": 2.64891846921797,
+      "grad_norm": 0.49726173281669617,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 0.6023,
+      "step": 1194
+    },
+    {
+      "epoch": 2.651136993899057,
+      "grad_norm": 0.411594420671463,
+      "learning_rate": 3.82716049382716e-05,
+      "loss": 0.6025,
+      "step": 1195
+    },
+    {
+      "epoch": 2.653355518580144,
+      "grad_norm": 0.4449329078197479,
+      "learning_rate": 3.802469135802468e-05,
+      "loss": 0.613,
+      "step": 1196
+    },
+    {
+      "epoch": 2.655574043261231,
+      "grad_norm": 0.4007834792137146,
+      "learning_rate": 3.777777777777777e-05,
+      "loss": 0.5728,
+      "step": 1197
+    },
+    {
+      "epoch": 2.657792567942318,
+      "grad_norm": 0.35383889079093933,
+      "learning_rate": 3.7530864197530867e-05,
+      "loss": 0.59,
+      "step": 1198
+    },
+    {
+      "epoch": 2.6600110926234053,
+      "grad_norm": 0.3726763129234314,
+      "learning_rate": 3.728395061728395e-05,
+      "loss": 0.6817,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6622296173044924,
+      "grad_norm": 0.4918183982372284,
+      "learning_rate": 3.703703703703703e-05,
+      "loss": 0.7734,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6644481419855794,
+      "grad_norm": 0.3921082317829132,
+      "learning_rate": 3.679012345679012e-05,
+      "loss": 0.5855,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.3500710427761078,
+      "learning_rate": 3.6543209876543207e-05,
+      "loss": 0.7169,
+      "step": 1202
+    },
+    {
+      "epoch": 2.6688851913477536,
+      "grad_norm": 0.350633442401886,
+      "learning_rate": 3.6296296296296295e-05,
+      "loss": 0.843,
+      "step": 1203
+    },
+    {
+      "epoch": 2.6711037160288407,
+      "grad_norm": 0.36611026525497437,
+      "learning_rate": 3.604938271604938e-05,
+      "loss": 0.9445,
+      "step": 1204
+    },
+    {
+      "epoch": 2.6733222407099277,
+      "grad_norm": 0.32130178809165955,
+      "learning_rate": 3.5802469135802465e-05,
+      "loss": 0.6251,
+      "step": 1205
+    },
+    {
+      "epoch": 2.675540765391015,
+      "grad_norm": 0.38302144408226013,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 0.6811,
+      "step": 1206
+    },
+    {
+      "epoch": 2.677759290072102,
+      "grad_norm": 0.41110023856163025,
+      "learning_rate": 3.5308641975308635e-05,
+      "loss": 0.6695,
+      "step": 1207
+    },
+    {
+      "epoch": 2.679977814753189,
+      "grad_norm": 0.43156564235687256,
+      "learning_rate": 3.506172839506172e-05,
+      "loss": 0.7629,
+      "step": 1208
+    },
+    {
+      "epoch": 2.682196339434276,
+      "grad_norm": 0.36374178528785706,
+      "learning_rate": 3.481481481481481e-05,
+      "loss": 0.639,
+      "step": 1209
+    },
+    {
+      "epoch": 2.684414864115363,
+      "grad_norm": 0.40637218952178955,
+      "learning_rate": 3.45679012345679e-05,
+      "loss": 0.62,
+      "step": 1210
+    },
+    {
+      "epoch": 2.68663338879645,
+      "grad_norm": 0.32890716195106506,
+      "learning_rate": 3.432098765432098e-05,
+      "loss": 0.6494,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6888519134775377,
+      "grad_norm": 0.34911808371543884,
+      "learning_rate": 3.407407407407407e-05,
+      "loss": 0.798,
+      "step": 1212
+    },
+    {
+      "epoch": 2.691070438158625,
+      "grad_norm": 0.3288569748401642,
+      "learning_rate": 3.382716049382716e-05,
+      "loss": 0.6642,
+      "step": 1213
+    },
+    {
+      "epoch": 2.693288962839712,
+      "grad_norm": 0.36653268337249756,
+      "learning_rate": 3.3580246913580247e-05,
+      "loss": 0.7126,
+      "step": 1214
+    },
+    {
+      "epoch": 2.695507487520799,
+      "grad_norm": 0.3479364812374115,
+      "learning_rate": 3.333333333333333e-05,
+      "loss": 0.6324,
+      "step": 1215
+    },
+    {
+      "epoch": 2.697726012201886,
+      "grad_norm": 0.40026071667671204,
+      "learning_rate": 3.3086419753086417e-05,
+      "loss": 0.8029,
+      "step": 1216
+    },
+    {
+      "epoch": 2.699944536882973,
+      "grad_norm": 0.3726097345352173,
+      "learning_rate": 3.2839506172839505e-05,
+      "loss": 0.7859,
+      "step": 1217
+    },
+    {
+      "epoch": 2.70216306156406,
+      "grad_norm": 0.45412957668304443,
+      "learning_rate": 3.2592592592592586e-05,
+      "loss": 0.6761,
+      "step": 1218
+    },
+    {
+      "epoch": 2.7043815862451472,
+      "grad_norm": 0.3936997354030609,
+      "learning_rate": 3.2345679012345675e-05,
+      "loss": 0.72,
+      "step": 1219
+    },
+    {
+      "epoch": 2.7066001109262343,
+      "grad_norm": 0.3688371777534485,
+      "learning_rate": 3.209876543209876e-05,
+      "loss": 0.7343,
+      "step": 1220
+    },
+    {
+      "epoch": 2.7088186356073214,
+      "grad_norm": 0.37290459871292114,
+      "learning_rate": 3.185185185185185e-05,
+      "loss": 0.7259,
+      "step": 1221
+    },
+    {
+      "epoch": 2.7110371602884085,
+      "grad_norm": 0.351614773273468,
+      "learning_rate": 3.160493827160493e-05,
+      "loss": 0.9476,
+      "step": 1222
+    },
+    {
+      "epoch": 2.7132556849694955,
+      "grad_norm": 0.3665882349014282,
+      "learning_rate": 3.135802469135802e-05,
+      "loss": 0.7916,
+      "step": 1223
+    },
+    {
+      "epoch": 2.7154742096505826,
+      "grad_norm": 0.3111608624458313,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 0.8192,
+      "step": 1224
+    },
+    {
+      "epoch": 2.7176927343316697,
+      "grad_norm": 0.47173646092414856,
+      "learning_rate": 3.08641975308642e-05,
+      "loss": 0.742,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7199112590127568,
+      "grad_norm": 0.3495829701423645,
+      "learning_rate": 3.061728395061728e-05,
+      "loss": 0.5991,
+      "step": 1226
+    },
+    {
+      "epoch": 2.722129783693844,
+      "grad_norm": 0.4144182801246643,
+      "learning_rate": 3.0370370370370368e-05,
+      "loss": 0.7165,
+      "step": 1227
+    },
+    {
+      "epoch": 2.724348308374931,
+      "grad_norm": 0.34134775400161743,
+      "learning_rate": 3.0123456790123453e-05,
+      "loss": 0.745,
+      "step": 1228
+    },
+    {
+      "epoch": 2.726566833056018,
+      "grad_norm": 0.40796345472335815,
+      "learning_rate": 2.987654320987654e-05,
+      "loss": 0.7099,
+      "step": 1229
+    },
+    {
+      "epoch": 2.728785357737105,
+      "grad_norm": 0.338306725025177,
+      "learning_rate": 2.9629629629629627e-05,
+      "loss": 0.6187,
+      "step": 1230
+    },
+    {
+      "epoch": 2.731003882418192,
+      "grad_norm": 0.4072520136833191,
+      "learning_rate": 2.9382716049382715e-05,
+      "loss": 0.7471,
+      "step": 1231
+    },
+    {
+      "epoch": 2.733222407099279,
+      "grad_norm": 0.3976074159145355,
+      "learning_rate": 2.9135802469135796e-05,
+      "loss": 0.4748,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7354409317803663,
+      "grad_norm": 0.40216678380966187,
+      "learning_rate": 2.8888888888888885e-05,
+      "loss": 0.5721,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7376594564614534,
+      "grad_norm": 0.33358097076416016,
+      "learning_rate": 2.8641975308641973e-05,
+      "loss": 0.6315,
+      "step": 1234
+    },
+    {
+      "epoch": 2.7398779811425404,
+      "grad_norm": 0.3056701719760895,
+      "learning_rate": 2.8395061728395058e-05,
+      "loss": 0.6673,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7420965058236275,
+      "grad_norm": 0.3281489908695221,
+      "learning_rate": 2.8148148148148147e-05,
+      "loss": 0.6475,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7443150305047146,
+      "grad_norm": 0.3202396035194397,
+      "learning_rate": 2.790123456790123e-05,
+      "loss": 0.7457,
+      "step": 1237
+    },
+    {
+      "epoch": 2.7465335551858017,
+      "grad_norm": 0.4637419879436493,
+      "learning_rate": 2.765432098765432e-05,
+      "loss": 0.7351,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7487520798668887,
+      "grad_norm": 0.4060627520084381,
+      "learning_rate": 2.7407407407407405e-05,
+      "loss": 0.6025,
+      "step": 1239
+    },
+    {
+      "epoch": 2.750970604547976,
+      "grad_norm": 0.3486776351928711,
+      "learning_rate": 2.7160493827160493e-05,
+      "loss": 0.7604,
+      "step": 1240
+    },
+    {
+      "epoch": 2.753189129229063,
+      "grad_norm": 0.34644681215286255,
+      "learning_rate": 2.6913580246913578e-05,
+      "loss": 0.7352,
+      "step": 1241
+    },
+    {
+      "epoch": 2.75540765391015,
+      "grad_norm": 0.3410528600215912,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6591,
+      "step": 1242
+    },
+    {
+      "epoch": 2.757626178591237,
+      "grad_norm": 0.32501497864723206,
+      "learning_rate": 2.6419753086419748e-05,
+      "loss": 0.6415,
+      "step": 1243
+    },
+    {
+      "epoch": 2.759844703272324,
+      "grad_norm": 0.3995802104473114,
+      "learning_rate": 2.6172839506172837e-05,
+      "loss": 0.8061,
+      "step": 1244
+    },
+    {
+      "epoch": 2.762063227953411,
+      "grad_norm": 0.336206316947937,
+      "learning_rate": 2.592592592592592e-05,
+      "loss": 0.8745,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7642817526344983,
+      "grad_norm": 0.35639190673828125,
+      "learning_rate": 2.567901234567901e-05,
+      "loss": 0.7144,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7665002773155853,
+      "grad_norm": 0.34624427556991577,
+      "learning_rate": 2.5432098765432095e-05,
+      "loss": 0.7128,
+      "step": 1247
+    },
+    {
+      "epoch": 2.7687188019966724,
+      "grad_norm": 0.38196006417274475,
+      "learning_rate": 2.5185185185185183e-05,
+      "loss": 0.6168,
+      "step": 1248
+    },
+    {
+      "epoch": 2.7709373266777595,
+      "grad_norm": 0.32413601875305176,
+      "learning_rate": 2.4938271604938268e-05,
+      "loss": 0.8261,
+      "step": 1249
+    },
+    {
+      "epoch": 2.7731558513588466,
+      "grad_norm": 0.34929120540618896,
+      "learning_rate": 2.4691358024691357e-05,
+      "loss": 0.6727,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7753743760399336,
+      "grad_norm": 0.37564265727996826,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 0.8242,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7775929007210207,
+      "grad_norm": 0.33285558223724365,
+      "learning_rate": 2.419753086419753e-05,
+      "loss": 0.6205,
+      "step": 1252
+    },
+    {
+      "epoch": 2.779811425402108,
+      "grad_norm": 0.4596811532974243,
+      "learning_rate": 2.3950617283950618e-05,
+      "loss": 0.9268,
+      "step": 1253
+    },
+    {
+      "epoch": 2.782029950083195,
+      "grad_norm": 0.42930254340171814,
+      "learning_rate": 2.37037037037037e-05,
+      "loss": 0.5897,
+      "step": 1254
+    },
+    {
+      "epoch": 2.784248474764282,
+      "grad_norm": 0.2988002598285675,
+      "learning_rate": 2.3456790123456788e-05,
+      "loss": 0.6582,
+      "step": 1255
+    },
+    {
+      "epoch": 2.786466999445369,
+      "grad_norm": 0.3756999671459198,
+      "learning_rate": 2.3209876543209873e-05,
+      "loss": 0.7009,
+      "step": 1256
+    },
+    {
+      "epoch": 2.788685524126456,
+      "grad_norm": 0.43179574608802795,
+      "learning_rate": 2.296296296296296e-05,
+      "loss": 0.8466,
+      "step": 1257
+    },
+    {
+      "epoch": 2.790904048807543,
+      "grad_norm": 0.4446340799331665,
+      "learning_rate": 2.2716049382716047e-05,
+      "loss": 0.6367,
+      "step": 1258
+    },
+    {
+      "epoch": 2.7931225734886302,
+      "grad_norm": 0.4565364420413971,
+      "learning_rate": 2.2469135802469135e-05,
+      "loss": 0.8962,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7953410981697173,
+      "grad_norm": 0.3654424846172333,
+      "learning_rate": 2.222222222222222e-05,
+      "loss": 0.7056,
+      "step": 1260
+    },
+    {
+      "epoch": 2.7975596228508044,
+      "grad_norm": 0.3186659514904022,
+      "learning_rate": 2.1975308641975308e-05,
+      "loss": 0.5495,
+      "step": 1261
+    },
+    {
+      "epoch": 2.7997781475318915,
+      "grad_norm": 0.4120534360408783,
+      "learning_rate": 2.1728395061728393e-05,
+      "loss": 0.7143,
+      "step": 1262
+    },
+    {
+      "epoch": 2.8019966722129785,
+      "grad_norm": 0.35582518577575684,
+      "learning_rate": 2.148148148148148e-05,
+      "loss": 0.5925,
+      "step": 1263
+    },
+    {
+      "epoch": 2.8042151968940656,
+      "grad_norm": 0.38848698139190674,
+      "learning_rate": 2.1234567901234563e-05,
+      "loss": 0.7256,
+      "step": 1264
+    },
+    {
+      "epoch": 2.8064337215751527,
+      "grad_norm": 0.4256631135940552,
+      "learning_rate": 2.098765432098765e-05,
+      "loss": 0.8521,
+      "step": 1265
+    },
+    {
+      "epoch": 2.8086522462562398,
+      "grad_norm": 0.4274256229400635,
+      "learning_rate": 2.0740740740740737e-05,
+      "loss": 0.605,
+      "step": 1266
+    },
+    {
+      "epoch": 2.810870770937327,
+      "grad_norm": 0.31895482540130615,
+      "learning_rate": 2.0493827160493825e-05,
+      "loss": 0.8141,
+      "step": 1267
+    },
+    {
+      "epoch": 2.813089295618414,
+      "grad_norm": 0.41098976135253906,
+      "learning_rate": 2.0246913580246913e-05,
+      "loss": 0.5943,
+      "step": 1268
+    },
+    {
+      "epoch": 2.815307820299501,
+      "grad_norm": 0.3214184045791626,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.5767,
+      "step": 1269
+    },
+    {
+      "epoch": 2.817526344980588,
+      "grad_norm": 0.33707132935523987,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 0.5985,
+      "step": 1270
+    },
+    {
+      "epoch": 2.819744869661675,
+      "grad_norm": 0.3730061650276184,
+      "learning_rate": 1.950617283950617e-05,
+      "loss": 0.6366,
+      "step": 1271
+    },
+    {
+      "epoch": 2.821963394342762,
+      "grad_norm": 0.4043295979499817,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.6635,
+      "step": 1272
+    },
+    {
+      "epoch": 2.8241819190238493,
+      "grad_norm": 0.34972065687179565,
+      "learning_rate": 1.901234567901234e-05,
+      "loss": 0.6591,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8264004437049364,
+      "grad_norm": 0.377803772687912,
+      "learning_rate": 1.8765432098765433e-05,
+      "loss": 0.6876,
+      "step": 1274
+    },
+    {
+      "epoch": 2.8286189683860234,
+      "grad_norm": 0.43975675106048584,
+      "learning_rate": 1.8518518518518515e-05,
+      "loss": 0.8433,
+      "step": 1275
+    },
+    {
+      "epoch": 2.8308374930671105,
+      "grad_norm": 0.4129805862903595,
+      "learning_rate": 1.8271604938271603e-05,
+      "loss": 0.6128,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8330560177481976,
+      "grad_norm": 0.3992515504360199,
+      "learning_rate": 1.802469135802469e-05,
+      "loss": 0.8231,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8352745424292847,
+      "grad_norm": 0.3519774079322815,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.5861,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8374930671103717,
+      "grad_norm": 0.4027957618236542,
+      "learning_rate": 1.753086419753086e-05,
+      "loss": 0.6461,
+      "step": 1279
+    },
+    {
+      "epoch": 2.839711591791459,
+      "grad_norm": 0.3712104558944702,
+      "learning_rate": 1.728395061728395e-05,
+      "loss": 0.667,
+      "step": 1280
+    },
+    {
+      "epoch": 2.841930116472546,
+      "grad_norm": 0.27034059166908264,
+      "learning_rate": 1.7037037037037035e-05,
+      "loss": 0.5569,
+      "step": 1281
+    },
+    {
+      "epoch": 2.844148641153633,
+      "grad_norm": 0.40547284483909607,
+      "learning_rate": 1.6790123456790123e-05,
+      "loss": 0.8388,
+      "step": 1282
+    },
+    {
+      "epoch": 2.84636716583472,
+      "grad_norm": 0.6386278867721558,
+      "learning_rate": 1.6543209876543208e-05,
+      "loss": 0.5935,
+      "step": 1283
+    },
+    {
+      "epoch": 2.848585690515807,
+      "grad_norm": 0.3278522789478302,
+      "learning_rate": 1.6296296296296293e-05,
+      "loss": 0.5315,
+      "step": 1284
+    },
+    {
+      "epoch": 2.850804215196894,
+      "grad_norm": 0.3663184344768524,
+      "learning_rate": 1.604938271604938e-05,
+      "loss": 0.6507,
+      "step": 1285
+    },
+    {
+      "epoch": 2.8530227398779813,
+      "grad_norm": 0.3412255048751831,
+      "learning_rate": 1.5802469135802467e-05,
+      "loss": 0.5841,
+      "step": 1286
+    },
+    {
+      "epoch": 2.8552412645590683,
+      "grad_norm": 0.35918501019477844,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.429,
+      "step": 1287
+    },
+    {
+      "epoch": 2.8574597892401554,
+      "grad_norm": 0.3154831826686859,
+      "learning_rate": 1.530864197530864e-05,
+      "loss": 0.6429,
+      "step": 1288
+    },
+    {
+      "epoch": 2.8596783139212425,
+      "grad_norm": 0.6046378016471863,
+      "learning_rate": 1.5061728395061727e-05,
+      "loss": 0.6676,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8618968386023296,
+      "grad_norm": 0.39476388692855835,
+      "learning_rate": 1.4814814814814813e-05,
+      "loss": 0.5814,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8641153632834166,
+      "grad_norm": 0.3529339134693146,
+      "learning_rate": 1.4567901234567898e-05,
+      "loss": 0.6639,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8663338879645037,
+      "grad_norm": 0.42051246762275696,
+      "learning_rate": 1.4320987654320987e-05,
+      "loss": 0.6224,
+      "step": 1292
+    },
+    {
+      "epoch": 2.868552412645591,
+      "grad_norm": 0.3897562325000763,
+      "learning_rate": 1.4074074074074073e-05,
+      "loss": 0.8855,
+      "step": 1293
+    },
+    {
+      "epoch": 2.870770937326678,
+      "grad_norm": 0.34242159128189087,
+      "learning_rate": 1.382716049382716e-05,
+      "loss": 0.806,
+      "step": 1294
+    },
+    {
+      "epoch": 2.872989462007765,
+      "grad_norm": 0.39863669872283936,
+      "learning_rate": 1.3580246913580247e-05,
+      "loss": 0.7796,
+      "step": 1295
+    },
+    {
+      "epoch": 2.875207986688852,
+      "grad_norm": 0.28337791562080383,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6309,
+      "step": 1296
+    },
+    {
+      "epoch": 2.877426511369939,
+      "grad_norm": 0.3512384593486786,
+      "learning_rate": 1.3086419753086418e-05,
+      "loss": 0.6327,
+      "step": 1297
+    },
+    {
+      "epoch": 2.879645036051026,
+      "grad_norm": 0.3786504864692688,
+      "learning_rate": 1.2839506172839505e-05,
+      "loss": 0.7245,
+      "step": 1298
+    },
+    {
+      "epoch": 2.8818635607321132,
+      "grad_norm": 0.34923163056373596,
+      "learning_rate": 1.2592592592592592e-05,
+      "loss": 0.8614,
+      "step": 1299
+    },
+    {
+      "epoch": 2.8840820854132003,
+      "grad_norm": 0.3500804603099823,
+      "learning_rate": 1.2345679012345678e-05,
+      "loss": 0.5442,
+      "step": 1300
+    },
+    {
+      "epoch": 2.8863006100942874,
+      "grad_norm": 0.4108477532863617,
+      "learning_rate": 1.2098765432098765e-05,
+      "loss": 0.725,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8885191347753745,
+      "grad_norm": 0.36196666955947876,
+      "learning_rate": 1.185185185185185e-05,
+      "loss": 0.5503,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8907376594564616,
+      "grad_norm": 0.44769877195358276,
+      "learning_rate": 1.1604938271604937e-05,
+      "loss": 0.7031,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8929561841375486,
+      "grad_norm": 0.4032876789569855,
+      "learning_rate": 1.1358024691358023e-05,
+      "loss": 0.7061,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8951747088186357,
+      "grad_norm": 0.3978791832923889,
+      "learning_rate": 1.111111111111111e-05,
+      "loss": 0.7025,
+      "step": 1305
+    },
+    {
+      "epoch": 2.8973932334997228,
+      "grad_norm": 0.378939688205719,
+      "learning_rate": 1.0864197530864197e-05,
+      "loss": 0.6771,
+      "step": 1306
+    },
+    {
+      "epoch": 2.89961175818081,
+      "grad_norm": 0.46190688014030457,
+      "learning_rate": 1.0617283950617282e-05,
+      "loss": 0.677,
+      "step": 1307
+    },
+    {
+      "epoch": 2.901830282861897,
+      "grad_norm": 0.40030258893966675,
+      "learning_rate": 1.0370370370370368e-05,
+      "loss": 0.7338,
+      "step": 1308
+    },
+    {
+      "epoch": 2.904048807542984,
+      "grad_norm": 0.41620171070098877,
+      "learning_rate": 1.0123456790123457e-05,
+      "loss": 0.6474,
+      "step": 1309
+    },
+    {
+      "epoch": 2.906267332224071,
+      "grad_norm": 0.3200892508029938,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 0.4254,
+      "step": 1310
+    },
+    {
+      "epoch": 2.908485856905158,
+      "grad_norm": 0.30711278319358826,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.7118,
+      "step": 1311
+    },
+    {
+      "epoch": 2.9107043815862452,
+      "grad_norm": 0.37710142135620117,
+      "learning_rate": 9.382716049382717e-06,
+      "loss": 0.6679,
+      "step": 1312
+    },
+    {
+      "epoch": 2.9129229062673323,
+      "grad_norm": 0.4593709409236908,
+      "learning_rate": 9.135802469135802e-06,
+      "loss": 0.5259,
+      "step": 1313
+    },
+    {
+      "epoch": 2.9151414309484194,
+      "grad_norm": 0.3127303123474121,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.679,
+      "step": 1314
+    },
+    {
+      "epoch": 2.9173599556295065,
+      "grad_norm": 0.3648340702056885,
+      "learning_rate": 8.641975308641975e-06,
+      "loss": 0.6737,
+      "step": 1315
+    },
+    {
+      "epoch": 2.9195784803105935,
+      "grad_norm": 0.39914053678512573,
+      "learning_rate": 8.395061728395062e-06,
+      "loss": 0.9316,
+      "step": 1316
+    },
+    {
+      "epoch": 2.9217970049916806,
+      "grad_norm": 0.37575289607048035,
+      "learning_rate": 8.148148148148147e-06,
+      "loss": 0.7505,
+      "step": 1317
+    },
+    {
+      "epoch": 2.9240155296727677,
+      "grad_norm": 0.34847092628479004,
+      "learning_rate": 7.901234567901233e-06,
+      "loss": 0.7238,
+      "step": 1318
+    },
+    {
+      "epoch": 2.9262340543538548,
+      "grad_norm": 0.3633224368095398,
+      "learning_rate": 7.65432098765432e-06,
+      "loss": 0.648,
+      "step": 1319
+    },
+    {
+      "epoch": 2.928452579034942,
+      "grad_norm": 0.3765181303024292,
+      "learning_rate": 7.407407407407407e-06,
+      "loss": 0.7943,
+      "step": 1320
+    },
+    {
+      "epoch": 2.930671103716029,
+      "grad_norm": 0.3698227107524872,
+      "learning_rate": 7.160493827160493e-06,
+      "loss": 0.7919,
+      "step": 1321
+    },
+    {
+      "epoch": 2.932889628397116,
+      "grad_norm": 0.3493306338787079,
+      "learning_rate": 6.91358024691358e-06,
+      "loss": 0.6661,
+      "step": 1322
+    },
+    {
+      "epoch": 2.935108153078203,
+      "grad_norm": 0.4068938195705414,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.6675,
+      "step": 1323
+    },
+    {
+      "epoch": 2.93732667775929,
+      "grad_norm": 0.3794713020324707,
+      "learning_rate": 6.4197530864197525e-06,
+      "loss": 0.6468,
+      "step": 1324
+    },
+    {
+      "epoch": 2.939545202440377,
+      "grad_norm": 0.4686014950275421,
+      "learning_rate": 6.172839506172839e-06,
+      "loss": 0.7808,
+      "step": 1325
+    },
+    {
+      "epoch": 2.9417637271214643,
+      "grad_norm": 0.3734581470489502,
+      "learning_rate": 5.925925925925925e-06,
+      "loss": 0.6765,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9439822518025514,
+      "grad_norm": 0.4299374222755432,
+      "learning_rate": 5.679012345679012e-06,
+      "loss": 0.7129,
+      "step": 1327
+    },
+    {
+      "epoch": 2.9462007764836384,
+      "grad_norm": 0.30758053064346313,
+      "learning_rate": 5.432098765432098e-06,
+      "loss": 0.8138,
+      "step": 1328
+    },
+    {
+      "epoch": 2.9484193011647255,
+      "grad_norm": 0.36681562662124634,
+      "learning_rate": 5.185185185185184e-06,
+      "loss": 0.5586,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9506378258458126,
+      "grad_norm": 0.39631277322769165,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.7247,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9528563505268997,
+      "grad_norm": 0.4128675162792206,
+      "learning_rate": 4.691358024691358e-06,
+      "loss": 0.7964,
+      "step": 1331
+    },
+    {
+      "epoch": 2.9550748752079867,
+      "grad_norm": 0.2892022132873535,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.6945,
+      "step": 1332
+    },
+    {
+      "epoch": 2.957293399889074,
+      "grad_norm": 0.42822718620300293,
+      "learning_rate": 4.197530864197531e-06,
+      "loss": 0.5976,
+      "step": 1333
+    },
+    {
+      "epoch": 2.959511924570161,
+      "grad_norm": 0.3877113163471222,
+      "learning_rate": 3.950617283950617e-06,
+      "loss": 0.6921,
+      "step": 1334
+    },
+    {
+      "epoch": 2.961730449251248,
+      "grad_norm": 0.33067578077316284,
+      "learning_rate": 3.7037037037037033e-06,
+      "loss": 0.7964,
+      "step": 1335
+    },
+    {
+      "epoch": 2.963948973932335,
+      "grad_norm": 0.45144280791282654,
+      "learning_rate": 3.45679012345679e-06,
+      "loss": 0.8757,
+      "step": 1336
+    },
+    {
+      "epoch": 2.966167498613422,
+      "grad_norm": 0.4545195698738098,
+      "learning_rate": 3.2098765432098762e-06,
+      "loss": 0.6213,
+      "step": 1337
+    },
+    {
+      "epoch": 2.968386023294509,
+      "grad_norm": 0.3137918710708618,
+      "learning_rate": 2.9629629629629625e-06,
+      "loss": 0.6447,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9706045479755963,
+      "grad_norm": 0.3208416998386383,
+      "learning_rate": 2.716049382716049e-06,
+      "loss": 0.5429,
+      "step": 1339
+    },
+    {
+      "epoch": 2.9728230726566833,
+      "grad_norm": 0.3757649064064026,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 0.7607,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9750415973377704,
+      "grad_norm": 0.4091593623161316,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.7485,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9772601220188575,
+      "grad_norm": 0.4239092469215393,
+      "learning_rate": 1.9753086419753083e-06,
+      "loss": 0.8213,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9794786466999446,
+      "grad_norm": 0.38245269656181335,
+      "learning_rate": 1.728395061728395e-06,
+      "loss": 0.6481,
+      "step": 1343
+    },
+    {
+      "epoch": 2.9816971713810316,
+      "grad_norm": 0.4324122369289398,
+      "learning_rate": 1.4814814814814812e-06,
+      "loss": 0.7333,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9839156960621187,
+      "grad_norm": 0.3000708818435669,
+      "learning_rate": 1.234567901234568e-06,
+      "loss": 0.8968,
+      "step": 1345
+    },
+    {
+      "epoch": 2.986134220743206,
+      "grad_norm": 0.42410048842430115,
+      "learning_rate": 9.876543209876542e-07,
+      "loss": 0.8807,
+      "step": 1346
+    },
+    {
+      "epoch": 2.988352745424293,
+      "grad_norm": 0.33582401275634766,
+      "learning_rate": 7.407407407407406e-07,
+      "loss": 0.6089,
+      "step": 1347
+    },
+    {
+      "epoch": 2.99057127010538,
+      "grad_norm": 0.4067254662513733,
+      "learning_rate": 4.938271604938271e-07,
+      "loss": 0.666,
+      "step": 1348
+    },
+    {
+      "epoch": 2.992789794786467,
+      "grad_norm": 0.4285987615585327,
+      "learning_rate": 2.4691358024691354e-07,
+      "loss": 0.7118,
+      "step": 1349
+    },
+    {
+      "epoch": 2.995008319467554,
+      "grad_norm": 0.34644412994384766,
+      "learning_rate": 0.0,
+      "loss": 0.5851,
+      "step": 1350
+    },
+    {
+      "epoch": 2.995008319467554,
+      "step": 1350,
+      "total_flos": 2.334453955087565e+16,
+      "train_loss": 0.7255008474985758,
+      "train_runtime": 1758.1957,
+      "train_samples_per_second": 12.304,
+      "train_steps_per_second": 0.768
+    }
+  ],
+  "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": 2.334453955087565e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}