diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13755 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9994236311239195,
+  "eval_steps": 500,
+  "global_step": 3903,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001536983669548511,
+      "grad_norm": 25.51269023102587,
+      "learning_rate": 3.5805626598465474e-08,
+      "loss": 1.1796,
+      "step": 2
+    },
+    {
+      "epoch": 0.003073967339097022,
+      "grad_norm": 35.249053162120475,
+      "learning_rate": 7.161125319693095e-08,
+      "loss": 1.176,
+      "step": 4
+    },
+    {
+      "epoch": 0.004610951008645533,
+      "grad_norm": 40.978012392510365,
+      "learning_rate": 1.0741687979539642e-07,
+      "loss": 1.1898,
+      "step": 6
+    },
+    {
+      "epoch": 0.006147934678194044,
+      "grad_norm": 37.791685318747554,
+      "learning_rate": 1.432225063938619e-07,
+      "loss": 1.1456,
+      "step": 8
+    },
+    {
+      "epoch": 0.007684918347742555,
+      "grad_norm": 42.35883366195493,
+      "learning_rate": 1.7902813299232735e-07,
+      "loss": 1.2175,
+      "step": 10
+    },
+    {
+      "epoch": 0.009221902017291067,
+      "grad_norm": 48.99379222065678,
+      "learning_rate": 2.1483375959079283e-07,
+      "loss": 1.192,
+      "step": 12
+    },
+    {
+      "epoch": 0.010758885686839578,
+      "grad_norm": 44.18044117170443,
+      "learning_rate": 2.506393861892583e-07,
+      "loss": 1.1534,
+      "step": 14
+    },
+    {
+      "epoch": 0.012295869356388088,
+      "grad_norm": 53.20894364313595,
+      "learning_rate": 2.864450127877238e-07,
+      "loss": 1.1731,
+      "step": 16
+    },
+    {
+      "epoch": 0.0138328530259366,
+      "grad_norm": 24.63444222442823,
+      "learning_rate": 3.2225063938618927e-07,
+      "loss": 1.1526,
+      "step": 18
+    },
+    {
+      "epoch": 0.01536983669548511,
+      "grad_norm": 26.218955749561335,
+      "learning_rate": 3.580562659846547e-07,
+      "loss": 1.1429,
+      "step": 20
+    },
+    {
+      "epoch": 0.01690682036503362,
+      "grad_norm": 24.488025247790667,
+      "learning_rate": 3.938618925831202e-07,
+      "loss": 1.103,
+      "step": 22
+    },
+    {
+      "epoch": 0.018443804034582133,
+      "grad_norm": 36.781877179954705,
+      "learning_rate": 4.2966751918158566e-07,
+      "loss": 1.119,
+      "step": 24
+    },
+    {
+      "epoch": 0.019980787704130643,
+      "grad_norm": 20.069997876411783,
+      "learning_rate": 4.654731457800512e-07,
+      "loss": 1.0916,
+      "step": 26
+    },
+    {
+      "epoch": 0.021517771373679155,
+      "grad_norm": 18.663749580298237,
+      "learning_rate": 5.012787723785166e-07,
+      "loss": 1.0403,
+      "step": 28
+    },
+    {
+      "epoch": 0.023054755043227664,
+      "grad_norm": 30.587706942463424,
+      "learning_rate": 5.37084398976982e-07,
+      "loss": 0.9996,
+      "step": 30
+    },
+    {
+      "epoch": 0.024591738712776177,
+      "grad_norm": 25.127131893569143,
+      "learning_rate": 5.728900255754476e-07,
+      "loss": 0.9013,
+      "step": 32
+    },
+    {
+      "epoch": 0.02612872238232469,
+      "grad_norm": 24.505390983983318,
+      "learning_rate": 6.08695652173913e-07,
+      "loss": 0.8338,
+      "step": 34
+    },
+    {
+      "epoch": 0.0276657060518732,
+      "grad_norm": 28.03624180091526,
+      "learning_rate": 6.445012787723785e-07,
+      "loss": 0.7863,
+      "step": 36
+    },
+    {
+      "epoch": 0.02920268972142171,
+      "grad_norm": 18.04703186938572,
+      "learning_rate": 6.80306905370844e-07,
+      "loss": 0.8265,
+      "step": 38
+    },
+    {
+      "epoch": 0.03073967339097022,
+      "grad_norm": 19.463425683379242,
+      "learning_rate": 7.161125319693094e-07,
+      "loss": 0.775,
+      "step": 40
+    },
+    {
+      "epoch": 0.03227665706051873,
+      "grad_norm": 17.677748750715033,
+      "learning_rate": 7.519181585677749e-07,
+      "loss": 0.6703,
+      "step": 42
+    },
+    {
+      "epoch": 0.03381364073006724,
+      "grad_norm": 10.700820532554927,
+      "learning_rate": 7.877237851662404e-07,
+      "loss": 0.5609,
+      "step": 44
+    },
+    {
+      "epoch": 0.035350624399615754,
+      "grad_norm": 8.475129109313396,
+      "learning_rate": 8.235294117647059e-07,
+      "loss": 0.4812,
+      "step": 46
+    },
+    {
+      "epoch": 0.03688760806916427,
+      "grad_norm": 8.038325889019223,
+      "learning_rate": 8.593350383631713e-07,
+      "loss": 0.5001,
+      "step": 48
+    },
+    {
+      "epoch": 0.03842459173871278,
+      "grad_norm": 5.218235069976644,
+      "learning_rate": 8.951406649616369e-07,
+      "loss": 0.4479,
+      "step": 50
+    },
+    {
+      "epoch": 0.039961575408261285,
+      "grad_norm": 4.835373589811063,
+      "learning_rate": 9.309462915601024e-07,
+      "loss": 0.4323,
+      "step": 52
+    },
+    {
+      "epoch": 0.0414985590778098,
+      "grad_norm": 6.306114301931914,
+      "learning_rate": 9.667519181585676e-07,
+      "loss": 0.4035,
+      "step": 54
+    },
+    {
+      "epoch": 0.04303554274735831,
+      "grad_norm": 4.0054619844059,
+      "learning_rate": 1.0025575447570332e-06,
+      "loss": 0.3893,
+      "step": 56
+    },
+    {
+      "epoch": 0.04457252641690682,
+      "grad_norm": 6.071833866129103,
+      "learning_rate": 1.0383631713554987e-06,
+      "loss": 0.3929,
+      "step": 58
+    },
+    {
+      "epoch": 0.04610951008645533,
+      "grad_norm": 6.484184007577541,
+      "learning_rate": 1.074168797953964e-06,
+      "loss": 0.3561,
+      "step": 60
+    },
+    {
+      "epoch": 0.04764649375600384,
+      "grad_norm": 3.3031873501656173,
+      "learning_rate": 1.1099744245524297e-06,
+      "loss": 0.3598,
+      "step": 62
+    },
+    {
+      "epoch": 0.049183477425552354,
+      "grad_norm": 3.1979226953240896,
+      "learning_rate": 1.1457800511508952e-06,
+      "loss": 0.3505,
+      "step": 64
+    },
+    {
+      "epoch": 0.050720461095100866,
+      "grad_norm": 3.2686119354900667,
+      "learning_rate": 1.1815856777493606e-06,
+      "loss": 0.3402,
+      "step": 66
+    },
+    {
+      "epoch": 0.05225744476464938,
+      "grad_norm": 3.4755981271030425,
+      "learning_rate": 1.217391304347826e-06,
+      "loss": 0.3354,
+      "step": 68
+    },
+    {
+      "epoch": 0.053794428434197884,
+      "grad_norm": 3.7786309539549294,
+      "learning_rate": 1.2531969309462915e-06,
+      "loss": 0.3222,
+      "step": 70
+    },
+    {
+      "epoch": 0.0553314121037464,
+      "grad_norm": 3.189215314687861,
+      "learning_rate": 1.289002557544757e-06,
+      "loss": 0.3269,
+      "step": 72
+    },
+    {
+      "epoch": 0.05686839577329491,
+      "grad_norm": 3.231319202267947,
+      "learning_rate": 1.3248081841432225e-06,
+      "loss": 0.317,
+      "step": 74
+    },
+    {
+      "epoch": 0.05840537944284342,
+      "grad_norm": 3.4059003373513526,
+      "learning_rate": 1.360613810741688e-06,
+      "loss": 0.3116,
+      "step": 76
+    },
+    {
+      "epoch": 0.05994236311239193,
+      "grad_norm": 3.118126356582086,
+      "learning_rate": 1.3964194373401534e-06,
+      "loss": 0.3138,
+      "step": 78
+    },
+    {
+      "epoch": 0.06147934678194044,
+      "grad_norm": 3.4693555336883417,
+      "learning_rate": 1.4322250639386188e-06,
+      "loss": 0.3019,
+      "step": 80
+    },
+    {
+      "epoch": 0.06301633045148895,
+      "grad_norm": 2.970297866234449,
+      "learning_rate": 1.4680306905370844e-06,
+      "loss": 0.293,
+      "step": 82
+    },
+    {
+      "epoch": 0.06455331412103746,
+      "grad_norm": 2.9935089318553216,
+      "learning_rate": 1.5038363171355499e-06,
+      "loss": 0.2807,
+      "step": 84
+    },
+    {
+      "epoch": 0.06609029779058598,
+      "grad_norm": 2.9320835508202747,
+      "learning_rate": 1.5396419437340153e-06,
+      "loss": 0.2981,
+      "step": 86
+    },
+    {
+      "epoch": 0.06762728146013448,
+      "grad_norm": 2.7902350790145958,
+      "learning_rate": 1.5754475703324807e-06,
+      "loss": 0.2882,
+      "step": 88
+    },
+    {
+      "epoch": 0.069164265129683,
+      "grad_norm": 2.913110266507068,
+      "learning_rate": 1.6112531969309462e-06,
+      "loss": 0.2963,
+      "step": 90
+    },
+    {
+      "epoch": 0.07070124879923151,
+      "grad_norm": 3.0624887135891905,
+      "learning_rate": 1.6470588235294118e-06,
+      "loss": 0.2841,
+      "step": 92
+    },
+    {
+      "epoch": 0.07223823246878001,
+      "grad_norm": 2.9694480736928193,
+      "learning_rate": 1.6828644501278772e-06,
+      "loss": 0.2745,
+      "step": 94
+    },
+    {
+      "epoch": 0.07377521613832853,
+      "grad_norm": 2.9922295547415168,
+      "learning_rate": 1.7186700767263426e-06,
+      "loss": 0.2702,
+      "step": 96
+    },
+    {
+      "epoch": 0.07531219980787704,
+      "grad_norm": 2.909619857546067,
+      "learning_rate": 1.754475703324808e-06,
+      "loss": 0.2731,
+      "step": 98
+    },
+    {
+      "epoch": 0.07684918347742556,
+      "grad_norm": 2.894052835979119,
+      "learning_rate": 1.7902813299232737e-06,
+      "loss": 0.28,
+      "step": 100
+    },
+    {
+      "epoch": 0.07838616714697406,
+      "grad_norm": 3.0298785356363256,
+      "learning_rate": 1.8260869565217391e-06,
+      "loss": 0.2604,
+      "step": 102
+    },
+    {
+      "epoch": 0.07992315081652257,
+      "grad_norm": 2.7510733379017367,
+      "learning_rate": 1.8618925831202048e-06,
+      "loss": 0.2745,
+      "step": 104
+    },
+    {
+      "epoch": 0.08146013448607109,
+      "grad_norm": 2.8755196777892267,
+      "learning_rate": 1.89769820971867e-06,
+      "loss": 0.2649,
+      "step": 106
+    },
+    {
+      "epoch": 0.0829971181556196,
+      "grad_norm": 2.7823559727995715,
+      "learning_rate": 1.933503836317135e-06,
+      "loss": 0.2617,
+      "step": 108
+    },
+    {
+      "epoch": 0.0845341018251681,
+      "grad_norm": 2.777608095721294,
+      "learning_rate": 1.9693094629156013e-06,
+      "loss": 0.2421,
+      "step": 110
+    },
+    {
+      "epoch": 0.08607108549471662,
+      "grad_norm": 2.8340487973855075,
+      "learning_rate": 2.0051150895140665e-06,
+      "loss": 0.2725,
+      "step": 112
+    },
+    {
+      "epoch": 0.08760806916426513,
+      "grad_norm": 2.7460631775841913,
+      "learning_rate": 2.040920716112532e-06,
+      "loss": 0.2553,
+      "step": 114
+    },
+    {
+      "epoch": 0.08914505283381365,
+      "grad_norm": 2.930465300765411,
+      "learning_rate": 2.0767263427109973e-06,
+      "loss": 0.2563,
+      "step": 116
+    },
+    {
+      "epoch": 0.09068203650336215,
+      "grad_norm": 2.9408506186054297,
+      "learning_rate": 2.1125319693094626e-06,
+      "loss": 0.2522,
+      "step": 118
+    },
+    {
+      "epoch": 0.09221902017291066,
+      "grad_norm": 2.9075494103912516,
+      "learning_rate": 2.148337595907928e-06,
+      "loss": 0.252,
+      "step": 120
+    },
+    {
+      "epoch": 0.09375600384245918,
+      "grad_norm": 2.772712298254247,
+      "learning_rate": 2.184143222506394e-06,
+      "loss": 0.2443,
+      "step": 122
+    },
+    {
+      "epoch": 0.09529298751200768,
+      "grad_norm": 2.9572827571923654,
+      "learning_rate": 2.2199488491048595e-06,
+      "loss": 0.2371,
+      "step": 124
+    },
+    {
+      "epoch": 0.0968299711815562,
+      "grad_norm": 2.939257511565131,
+      "learning_rate": 2.2557544757033247e-06,
+      "loss": 0.2553,
+      "step": 126
+    },
+    {
+      "epoch": 0.09836695485110471,
+      "grad_norm": 2.8729850658382055,
+      "learning_rate": 2.2915601023017903e-06,
+      "loss": 0.2468,
+      "step": 128
+    },
+    {
+      "epoch": 0.09990393852065321,
+      "grad_norm": 2.7047070980502474,
+      "learning_rate": 2.3273657289002556e-06,
+      "loss": 0.2314,
+      "step": 130
+    },
+    {
+      "epoch": 0.10144092219020173,
+      "grad_norm": 2.7576396912344876,
+      "learning_rate": 2.363171355498721e-06,
+      "loss": 0.231,
+      "step": 132
+    },
+    {
+      "epoch": 0.10297790585975024,
+      "grad_norm": 2.9300704934262054,
+      "learning_rate": 2.398976982097187e-06,
+      "loss": 0.2386,
+      "step": 134
+    },
+    {
+      "epoch": 0.10451488952929876,
+      "grad_norm": 2.630373862177306,
+      "learning_rate": 2.434782608695652e-06,
+      "loss": 0.2259,
+      "step": 136
+    },
+    {
+      "epoch": 0.10605187319884726,
+      "grad_norm": 2.7320707506261286,
+      "learning_rate": 2.4705882352941177e-06,
+      "loss": 0.2393,
+      "step": 138
+    },
+    {
+      "epoch": 0.10758885686839577,
+      "grad_norm": 2.7386829967371384,
+      "learning_rate": 2.506393861892583e-06,
+      "loss": 0.2391,
+      "step": 140
+    },
+    {
+      "epoch": 0.10912584053794429,
+      "grad_norm": 2.5434808295139653,
+      "learning_rate": 2.5421994884910485e-06,
+      "loss": 0.2254,
+      "step": 142
+    },
+    {
+      "epoch": 0.1106628242074928,
+      "grad_norm": 2.59627848497483,
+      "learning_rate": 2.578005115089514e-06,
+      "loss": 0.2323,
+      "step": 144
+    },
+    {
+      "epoch": 0.11219980787704131,
+      "grad_norm": 2.6102361814153836,
+      "learning_rate": 2.6138107416879794e-06,
+      "loss": 0.2212,
+      "step": 146
+    },
+    {
+      "epoch": 0.11373679154658982,
+      "grad_norm": 2.4837198822431157,
+      "learning_rate": 2.649616368286445e-06,
+      "loss": 0.2174,
+      "step": 148
+    },
+    {
+      "epoch": 0.11527377521613832,
+      "grad_norm": 2.487984443326589,
+      "learning_rate": 2.6854219948849103e-06,
+      "loss": 0.2171,
+      "step": 150
+    },
+    {
+      "epoch": 0.11681075888568684,
+      "grad_norm": 2.491267047584632,
+      "learning_rate": 2.721227621483376e-06,
+      "loss": 0.2303,
+      "step": 152
+    },
+    {
+      "epoch": 0.11834774255523535,
+      "grad_norm": 2.2658939217150302,
+      "learning_rate": 2.7570332480818415e-06,
+      "loss": 0.211,
+      "step": 154
+    },
+    {
+      "epoch": 0.11988472622478386,
+      "grad_norm": 2.2884352286474727,
+      "learning_rate": 2.7928388746803067e-06,
+      "loss": 0.2155,
+      "step": 156
+    },
+    {
+      "epoch": 0.12142170989433237,
+      "grad_norm": 2.175477310058195,
+      "learning_rate": 2.8286445012787724e-06,
+      "loss": 0.2101,
+      "step": 158
+    },
+    {
+      "epoch": 0.12295869356388088,
+      "grad_norm": 2.04586024150522,
+      "learning_rate": 2.8644501278772376e-06,
+      "loss": 0.2072,
+      "step": 160
+    },
+    {
+      "epoch": 0.1244956772334294,
+      "grad_norm": 1.937221714698271,
+      "learning_rate": 2.9002557544757037e-06,
+      "loss": 0.2128,
+      "step": 162
+    },
+    {
+      "epoch": 0.1260326609029779,
+      "grad_norm": 1.9127369194130952,
+      "learning_rate": 2.936061381074169e-06,
+      "loss": 0.2164,
+      "step": 164
+    },
+    {
+      "epoch": 0.12756964457252642,
+      "grad_norm": 1.9467054527515646,
+      "learning_rate": 2.971867007672634e-06,
+      "loss": 0.216,
+      "step": 166
+    },
+    {
+      "epoch": 0.12910662824207492,
+      "grad_norm": 1.7444615717696514,
+      "learning_rate": 3.0076726342710997e-06,
+      "loss": 0.185,
+      "step": 168
+    },
+    {
+      "epoch": 0.13064361191162344,
+      "grad_norm": 1.721455823281585,
+      "learning_rate": 3.043478260869565e-06,
+      "loss": 0.2041,
+      "step": 170
+    },
+    {
+      "epoch": 0.13218059558117196,
+      "grad_norm": 1.6944968403589584,
+      "learning_rate": 3.0792838874680306e-06,
+      "loss": 0.2001,
+      "step": 172
+    },
+    {
+      "epoch": 0.13371757925072045,
+      "grad_norm": 1.5902843900468329,
+      "learning_rate": 3.1150895140664962e-06,
+      "loss": 0.1969,
+      "step": 174
+    },
+    {
+      "epoch": 0.13525456292026897,
+      "grad_norm": 1.564694037006993,
+      "learning_rate": 3.1508951406649614e-06,
+      "loss": 0.1945,
+      "step": 176
+    },
+    {
+      "epoch": 0.1367915465898175,
+      "grad_norm": 1.5404270105425055,
+      "learning_rate": 3.186700767263427e-06,
+      "loss": 0.1919,
+      "step": 178
+    },
+    {
+      "epoch": 0.138328530259366,
+      "grad_norm": 1.4600068175795704,
+      "learning_rate": 3.2225063938618923e-06,
+      "loss": 0.2144,
+      "step": 180
+    },
+    {
+      "epoch": 0.1398655139289145,
+      "grad_norm": 1.4972009484455417,
+      "learning_rate": 3.258312020460358e-06,
+      "loss": 0.2029,
+      "step": 182
+    },
+    {
+      "epoch": 0.14140249759846302,
+      "grad_norm": 1.5216950664348157,
+      "learning_rate": 3.2941176470588236e-06,
+      "loss": 0.1997,
+      "step": 184
+    },
+    {
+      "epoch": 0.14293948126801154,
+      "grad_norm": 1.4985918972641317,
+      "learning_rate": 3.3299232736572892e-06,
+      "loss": 0.2036,
+      "step": 186
+    },
+    {
+      "epoch": 0.14447646493756003,
+      "grad_norm": 1.4780907494983677,
+      "learning_rate": 3.3657289002557544e-06,
+      "loss": 0.1981,
+      "step": 188
+    },
+    {
+      "epoch": 0.14601344860710855,
+      "grad_norm": 1.3886567101627987,
+      "learning_rate": 3.4015345268542197e-06,
+      "loss": 0.185,
+      "step": 190
+    },
+    {
+      "epoch": 0.14755043227665707,
+      "grad_norm": 1.398832954652927,
+      "learning_rate": 3.4373401534526853e-06,
+      "loss": 0.2029,
+      "step": 192
+    },
+    {
+      "epoch": 0.14908741594620556,
+      "grad_norm": 1.4597381544671573,
+      "learning_rate": 3.473145780051151e-06,
+      "loss": 0.1881,
+      "step": 194
+    },
+    {
+      "epoch": 0.15062439961575408,
+      "grad_norm": 1.3840318868108508,
+      "learning_rate": 3.508951406649616e-06,
+      "loss": 0.2078,
+      "step": 196
+    },
+    {
+      "epoch": 0.1521613832853026,
+      "grad_norm": 1.4070697917261799,
+      "learning_rate": 3.544757033248082e-06,
+      "loss": 0.1974,
+      "step": 198
+    },
+    {
+      "epoch": 0.15369836695485112,
+      "grad_norm": 1.361393353227654,
+      "learning_rate": 3.5805626598465474e-06,
+      "loss": 0.197,
+      "step": 200
+    },
+    {
+      "epoch": 0.1552353506243996,
+      "grad_norm": 1.434775426937055,
+      "learning_rate": 3.6163682864450126e-06,
+      "loss": 0.2038,
+      "step": 202
+    },
+    {
+      "epoch": 0.15677233429394813,
+      "grad_norm": 1.5065947786249383,
+      "learning_rate": 3.6521739130434783e-06,
+      "loss": 0.2037,
+      "step": 204
+    },
+    {
+      "epoch": 0.15830931796349665,
+      "grad_norm": 1.3442194224119404,
+      "learning_rate": 3.6879795396419435e-06,
+      "loss": 0.1964,
+      "step": 206
+    },
+    {
+      "epoch": 0.15984630163304514,
+      "grad_norm": 1.3670203145371058,
+      "learning_rate": 3.7237851662404096e-06,
+      "loss": 0.195,
+      "step": 208
+    },
+    {
+      "epoch": 0.16138328530259366,
+      "grad_norm": 1.4035980721723869,
+      "learning_rate": 3.7595907928388748e-06,
+      "loss": 0.1961,
+      "step": 210
+    },
+    {
+      "epoch": 0.16292026897214218,
+      "grad_norm": 1.30606746520172,
+      "learning_rate": 3.79539641943734e-06,
+      "loss": 0.1877,
+      "step": 212
+    },
+    {
+      "epoch": 0.16445725264169067,
+      "grad_norm": 1.4103433650854988,
+      "learning_rate": 3.831202046035806e-06,
+      "loss": 0.1941,
+      "step": 214
+    },
+    {
+      "epoch": 0.1659942363112392,
+      "grad_norm": 1.394026552172561,
+      "learning_rate": 3.86700767263427e-06,
+      "loss": 0.1922,
+      "step": 216
+    },
+    {
+      "epoch": 0.1675312199807877,
+      "grad_norm": 1.3829946302657223,
+      "learning_rate": 3.902813299232737e-06,
+      "loss": 0.2048,
+      "step": 218
+    },
+    {
+      "epoch": 0.1690682036503362,
+      "grad_norm": 1.3276260104745685,
+      "learning_rate": 3.9386189258312025e-06,
+      "loss": 0.1892,
+      "step": 220
+    },
+    {
+      "epoch": 0.17060518731988472,
+      "grad_norm": 1.3432125610119714,
+      "learning_rate": 3.974424552429667e-06,
+      "loss": 0.1892,
+      "step": 222
+    },
+    {
+      "epoch": 0.17214217098943324,
+      "grad_norm": 1.3597738268768664,
+      "learning_rate": 4.010230179028133e-06,
+      "loss": 0.1933,
+      "step": 224
+    },
+    {
+      "epoch": 0.17367915465898176,
+      "grad_norm": 1.3784514898304991,
+      "learning_rate": 4.046035805626598e-06,
+      "loss": 0.187,
+      "step": 226
+    },
+    {
+      "epoch": 0.17521613832853025,
+      "grad_norm": 1.283311780182546,
+      "learning_rate": 4.081841432225064e-06,
+      "loss": 0.1889,
+      "step": 228
+    },
+    {
+      "epoch": 0.17675312199807877,
+      "grad_norm": 1.3048663711388646,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.1958,
+      "step": 230
+    },
+    {
+      "epoch": 0.1782901056676273,
+      "grad_norm": 1.3644400089186173,
+      "learning_rate": 4.153452685421995e-06,
+      "loss": 0.1939,
+      "step": 232
+    },
+    {
+      "epoch": 0.17982708933717578,
+      "grad_norm": 1.2592600425837408,
+      "learning_rate": 4.18925831202046e-06,
+      "loss": 0.2031,
+      "step": 234
+    },
+    {
+      "epoch": 0.1813640730067243,
+      "grad_norm": 30.200615105516786,
+      "learning_rate": 4.225063938618925e-06,
+      "loss": 0.1891,
+      "step": 236
+    },
+    {
+      "epoch": 0.18290105667627282,
+      "grad_norm": 1.4965751336406534,
+      "learning_rate": 4.260869565217392e-06,
+      "loss": 0.1878,
+      "step": 238
+    },
+    {
+      "epoch": 0.1844380403458213,
+      "grad_norm": 5.563286529626895,
+      "learning_rate": 4.296675191815856e-06,
+      "loss": 0.1982,
+      "step": 240
+    },
+    {
+      "epoch": 0.18597502401536983,
+      "grad_norm": 1.4805149402757245,
+      "learning_rate": 4.332480818414322e-06,
+      "loss": 0.1942,
+      "step": 242
+    },
+    {
+      "epoch": 0.18751200768491835,
+      "grad_norm": 1.6629859442097377,
+      "learning_rate": 4.368286445012788e-06,
+      "loss": 0.2067,
+      "step": 244
+    },
+    {
+      "epoch": 0.18904899135446687,
+      "grad_norm": 9.155421397230691,
+      "learning_rate": 4.4040920716112525e-06,
+      "loss": 0.1872,
+      "step": 246
+    },
+    {
+      "epoch": 0.19058597502401536,
+      "grad_norm": 1.32114090472292,
+      "learning_rate": 4.439897698209719e-06,
+      "loss": 0.1844,
+      "step": 248
+    },
+    {
+      "epoch": 0.19212295869356388,
+      "grad_norm": 1.2320314387569729,
+      "learning_rate": 4.475703324808184e-06,
+      "loss": 0.1791,
+      "step": 250
+    },
+    {
+      "epoch": 0.1936599423631124,
+      "grad_norm": 1.2291776990139922,
+      "learning_rate": 4.511508951406649e-06,
+      "loss": 0.1799,
+      "step": 252
+    },
+    {
+      "epoch": 0.1951969260326609,
+      "grad_norm": 1.270813260985118,
+      "learning_rate": 4.547314578005115e-06,
+      "loss": 0.1877,
+      "step": 254
+    },
+    {
+      "epoch": 0.19673390970220941,
+      "grad_norm": 1.3202011345679392,
+      "learning_rate": 4.583120204603581e-06,
+      "loss": 0.2013,
+      "step": 256
+    },
+    {
+      "epoch": 0.19827089337175793,
+      "grad_norm": 1.3290195976171792,
+      "learning_rate": 4.618925831202046e-06,
+      "loss": 0.1915,
+      "step": 258
+    },
+    {
+      "epoch": 0.19980787704130643,
+      "grad_norm": 1.2091766956810812,
+      "learning_rate": 4.654731457800511e-06,
+      "loss": 0.1847,
+      "step": 260
+    },
+    {
+      "epoch": 0.20134486071085494,
+      "grad_norm": 1.3107091320741244,
+      "learning_rate": 4.690537084398977e-06,
+      "loss": 0.1844,
+      "step": 262
+    },
+    {
+      "epoch": 0.20288184438040346,
+      "grad_norm": 1.448237078882646,
+      "learning_rate": 4.726342710997442e-06,
+      "loss": 0.1948,
+      "step": 264
+    },
+    {
+      "epoch": 0.20441882804995196,
+      "grad_norm": 1.2336655516702122,
+      "learning_rate": 4.762148337595908e-06,
+      "loss": 0.1949,
+      "step": 266
+    },
+    {
+      "epoch": 0.20595581171950048,
+      "grad_norm": 1.3658099392493064,
+      "learning_rate": 4.797953964194374e-06,
+      "loss": 0.1804,
+      "step": 268
+    },
+    {
+      "epoch": 0.207492795389049,
+      "grad_norm": 1.373403969669977,
+      "learning_rate": 4.8337595907928385e-06,
+      "loss": 0.1905,
+      "step": 270
+    },
+    {
+      "epoch": 0.20902977905859751,
+      "grad_norm": 1.277467955296089,
+      "learning_rate": 4.869565217391304e-06,
+      "loss": 0.1987,
+      "step": 272
+    },
+    {
+      "epoch": 0.210566762728146,
+      "grad_norm": 1.3159132642953701,
+      "learning_rate": 4.90537084398977e-06,
+      "loss": 0.1957,
+      "step": 274
+    },
+    {
+      "epoch": 0.21210374639769453,
+      "grad_norm": 1.278799705973987,
+      "learning_rate": 4.941176470588235e-06,
+      "loss": 0.1903,
+      "step": 276
+    },
+    {
+      "epoch": 0.21364073006724305,
+      "grad_norm": 1.363154755717981,
+      "learning_rate": 4.976982097186701e-06,
+      "loss": 0.181,
+      "step": 278
+    },
+    {
+      "epoch": 0.21517771373679154,
+      "grad_norm": 1.226652497702008,
+      "learning_rate": 5.012787723785166e-06,
+      "loss": 0.185,
+      "step": 280
+    },
+    {
+      "epoch": 0.21671469740634006,
+      "grad_norm": 1.2660471639800883,
+      "learning_rate": 5.0485933503836314e-06,
+      "loss": 0.184,
+      "step": 282
+    },
+    {
+      "epoch": 0.21825168107588858,
+      "grad_norm": 1.204935258599295,
+      "learning_rate": 5.084398976982097e-06,
+      "loss": 0.1816,
+      "step": 284
+    },
+    {
+      "epoch": 0.21978866474543707,
+      "grad_norm": 1.2808963980429877,
+      "learning_rate": 5.120204603580563e-06,
+      "loss": 0.1911,
+      "step": 286
+    },
+    {
+      "epoch": 0.2213256484149856,
+      "grad_norm": 1.194046158945937,
+      "learning_rate": 5.156010230179028e-06,
+      "loss": 0.1838,
+      "step": 288
+    },
+    {
+      "epoch": 0.2228626320845341,
+      "grad_norm": 1.3438720779184878,
+      "learning_rate": 5.191815856777493e-06,
+      "loss": 0.1925,
+      "step": 290
+    },
+    {
+      "epoch": 0.22439961575408263,
+      "grad_norm": 1.1388013545101945,
+      "learning_rate": 5.227621483375959e-06,
+      "loss": 0.1732,
+      "step": 292
+    },
+    {
+      "epoch": 0.22593659942363112,
+      "grad_norm": 1.151418121106973,
+      "learning_rate": 5.2634271099744244e-06,
+      "loss": 0.1936,
+      "step": 294
+    },
+    {
+      "epoch": 0.22747358309317964,
+      "grad_norm": 1.205495348613184,
+      "learning_rate": 5.29923273657289e-06,
+      "loss": 0.1752,
+      "step": 296
+    },
+    {
+      "epoch": 0.22901056676272816,
+      "grad_norm": 1.2040139295608776,
+      "learning_rate": 5.335038363171356e-06,
+      "loss": 0.1873,
+      "step": 298
+    },
+    {
+      "epoch": 0.23054755043227665,
+      "grad_norm": 1.2348832685626474,
+      "learning_rate": 5.3708439897698205e-06,
+      "loss": 0.1898,
+      "step": 300
+    },
+    {
+      "epoch": 0.23208453410182517,
+      "grad_norm": 1.1372824585247818,
+      "learning_rate": 5.406649616368286e-06,
+      "loss": 0.1757,
+      "step": 302
+    },
+    {
+      "epoch": 0.2336215177713737,
+      "grad_norm": 1.231905059549878,
+      "learning_rate": 5.442455242966752e-06,
+      "loss": 0.1952,
+      "step": 304
+    },
+    {
+      "epoch": 0.23515850144092218,
+      "grad_norm": 1.1803349744715428,
+      "learning_rate": 5.478260869565217e-06,
+      "loss": 0.2001,
+      "step": 306
+    },
+    {
+      "epoch": 0.2366954851104707,
+      "grad_norm": 1.277265628143273,
+      "learning_rate": 5.514066496163683e-06,
+      "loss": 0.2008,
+      "step": 308
+    },
+    {
+      "epoch": 0.23823246878001922,
+      "grad_norm": 1.3927848001575907,
+      "learning_rate": 5.549872122762148e-06,
+      "loss": 0.1963,
+      "step": 310
+    },
+    {
+      "epoch": 0.2397694524495677,
+      "grad_norm": 1.2403730899787904,
+      "learning_rate": 5.5856777493606135e-06,
+      "loss": 0.1961,
+      "step": 312
+    },
+    {
+      "epoch": 0.24130643611911623,
+      "grad_norm": 1.163059501211908,
+      "learning_rate": 5.62148337595908e-06,
+      "loss": 0.175,
+      "step": 314
+    },
+    {
+      "epoch": 0.24284341978866475,
+      "grad_norm": 1.1639642316766559,
+      "learning_rate": 5.657289002557545e-06,
+      "loss": 0.1886,
+      "step": 316
+    },
+    {
+      "epoch": 0.24438040345821327,
+      "grad_norm": 1.2382118437766167,
+      "learning_rate": 5.69309462915601e-06,
+      "loss": 0.1973,
+      "step": 318
+    },
+    {
+      "epoch": 0.24591738712776176,
+      "grad_norm": 1.2074114040937993,
+      "learning_rate": 5.728900255754475e-06,
+      "loss": 0.1837,
+      "step": 320
+    },
+    {
+      "epoch": 0.24745437079731028,
+      "grad_norm": 1.295214887261411,
+      "learning_rate": 5.764705882352941e-06,
+      "loss": 0.1979,
+      "step": 322
+    },
+    {
+      "epoch": 0.2489913544668588,
+      "grad_norm": 1.2581192996807238,
+      "learning_rate": 5.800511508951407e-06,
+      "loss": 0.1789,
+      "step": 324
+    },
+    {
+      "epoch": 0.2505283381364073,
+      "grad_norm": 1.254789252666287,
+      "learning_rate": 5.836317135549872e-06,
+      "loss": 0.1941,
+      "step": 326
+    },
+    {
+      "epoch": 0.2520653218059558,
+      "grad_norm": 1.251310029743626,
+      "learning_rate": 5.872122762148338e-06,
+      "loss": 0.189,
+      "step": 328
+    },
+    {
+      "epoch": 0.25360230547550433,
+      "grad_norm": 1.1550514720366614,
+      "learning_rate": 5.9079283887468026e-06,
+      "loss": 0.1968,
+      "step": 330
+    },
+    {
+      "epoch": 0.25513928914505285,
+      "grad_norm": 1.1535871144055385,
+      "learning_rate": 5.943734015345268e-06,
+      "loss": 0.1844,
+      "step": 332
+    },
+    {
+      "epoch": 0.25667627281460137,
+      "grad_norm": 1.2208131864363314,
+      "learning_rate": 5.979539641943734e-06,
+      "loss": 0.1919,
+      "step": 334
+    },
+    {
+      "epoch": 0.25821325648414983,
+      "grad_norm": 1.1683405652224148,
+      "learning_rate": 6.0153452685421995e-06,
+      "loss": 0.1759,
+      "step": 336
+    },
+    {
+      "epoch": 0.25975024015369835,
+      "grad_norm": 1.1952164336575553,
+      "learning_rate": 6.051150895140665e-06,
+      "loss": 0.1795,
+      "step": 338
+    },
+    {
+      "epoch": 0.2612872238232469,
+      "grad_norm": 1.1820897323143373,
+      "learning_rate": 6.08695652173913e-06,
+      "loss": 0.19,
+      "step": 340
+    },
+    {
+      "epoch": 0.2628242074927954,
+      "grad_norm": 1.1800564956613375,
+      "learning_rate": 6.1227621483375955e-06,
+      "loss": 0.1978,
+      "step": 342
+    },
+    {
+      "epoch": 0.2643611911623439,
+      "grad_norm": 1.1467821141624814,
+      "learning_rate": 6.158567774936061e-06,
+      "loss": 0.1911,
+      "step": 344
+    },
+    {
+      "epoch": 0.26589817483189243,
+      "grad_norm": 1.1627264267097417,
+      "learning_rate": 6.194373401534527e-06,
+      "loss": 0.1921,
+      "step": 346
+    },
+    {
+      "epoch": 0.2674351585014409,
+      "grad_norm": 1.195438253587923,
+      "learning_rate": 6.2301790281329925e-06,
+      "loss": 0.1898,
+      "step": 348
+    },
+    {
+      "epoch": 0.2689721421709894,
+      "grad_norm": 1.1780466341960676,
+      "learning_rate": 6.265984654731457e-06,
+      "loss": 0.1792,
+      "step": 350
+    },
+    {
+      "epoch": 0.27050912584053793,
+      "grad_norm": 1.1731330647661304,
+      "learning_rate": 6.301790281329923e-06,
+      "loss": 0.1955,
+      "step": 352
+    },
+    {
+      "epoch": 0.27204610951008645,
+      "grad_norm": 1.1745568091917384,
+      "learning_rate": 6.3375959079283885e-06,
+      "loss": 0.1852,
+      "step": 354
+    },
+    {
+      "epoch": 0.273583093179635,
+      "grad_norm": 1.1783143021627482,
+      "learning_rate": 6.373401534526854e-06,
+      "loss": 0.1845,
+      "step": 356
+    },
+    {
+      "epoch": 0.2751200768491835,
+      "grad_norm": 1.1827170581568967,
+      "learning_rate": 6.40920716112532e-06,
+      "loss": 0.201,
+      "step": 358
+    },
+    {
+      "epoch": 0.276657060518732,
+      "grad_norm": 1.1115484126704642,
+      "learning_rate": 6.445012787723785e-06,
+      "loss": 0.17,
+      "step": 360
+    },
+    {
+      "epoch": 0.2781940441882805,
+      "grad_norm": 1.0326502834598414,
+      "learning_rate": 6.48081841432225e-06,
+      "loss": 0.186,
+      "step": 362
+    },
+    {
+      "epoch": 0.279731027857829,
+      "grad_norm": 1.259244016304707,
+      "learning_rate": 6.516624040920716e-06,
+      "loss": 0.1885,
+      "step": 364
+    },
+    {
+      "epoch": 0.2812680115273775,
+      "grad_norm": 1.1678776079796356,
+      "learning_rate": 6.5524296675191815e-06,
+      "loss": 0.1921,
+      "step": 366
+    },
+    {
+      "epoch": 0.28280499519692603,
+      "grad_norm": 1.0738638308920176,
+      "learning_rate": 6.588235294117647e-06,
+      "loss": 0.1707,
+      "step": 368
+    },
+    {
+      "epoch": 0.28434197886647455,
+      "grad_norm": 1.1594023002547749,
+      "learning_rate": 6.624040920716112e-06,
+      "loss": 0.1903,
+      "step": 370
+    },
+    {
+      "epoch": 0.2858789625360231,
+      "grad_norm": 1.1471065906349844,
+      "learning_rate": 6.6598465473145784e-06,
+      "loss": 0.194,
+      "step": 372
+    },
+    {
+      "epoch": 0.2874159462055716,
+      "grad_norm": 1.1901453575596241,
+      "learning_rate": 6.695652173913043e-06,
+      "loss": 0.1854,
+      "step": 374
+    },
+    {
+      "epoch": 0.28895292987512006,
+      "grad_norm": 1.1928859018793583,
+      "learning_rate": 6.731457800511509e-06,
+      "loss": 0.1906,
+      "step": 376
+    },
+    {
+      "epoch": 0.2904899135446686,
+      "grad_norm": 1.2778835898977514,
+      "learning_rate": 6.7672634271099745e-06,
+      "loss": 0.1871,
+      "step": 378
+    },
+    {
+      "epoch": 0.2920268972142171,
+      "grad_norm": 1.2992357301881277,
+      "learning_rate": 6.803069053708439e-06,
+      "loss": 0.2,
+      "step": 380
+    },
+    {
+      "epoch": 0.2935638808837656,
+      "grad_norm": 1.0621001143169835,
+      "learning_rate": 6.838874680306906e-06,
+      "loss": 0.1735,
+      "step": 382
+    },
+    {
+      "epoch": 0.29510086455331414,
+      "grad_norm": 1.1201679749220157,
+      "learning_rate": 6.874680306905371e-06,
+      "loss": 0.1736,
+      "step": 384
+    },
+    {
+      "epoch": 0.29663784822286265,
+      "grad_norm": 1.1884134035644247,
+      "learning_rate": 6.910485933503836e-06,
+      "loss": 0.1765,
+      "step": 386
+    },
+    {
+      "epoch": 0.2981748318924111,
+      "grad_norm": 1.0331590850791008,
+      "learning_rate": 6.946291560102302e-06,
+      "loss": 0.1751,
+      "step": 388
+    },
+    {
+      "epoch": 0.29971181556195964,
+      "grad_norm": 1.0474474582479285,
+      "learning_rate": 6.982097186700767e-06,
+      "loss": 0.1899,
+      "step": 390
+    },
+    {
+      "epoch": 0.30124879923150816,
+      "grad_norm": 1.1045197145968615,
+      "learning_rate": 6.999998599675296e-06,
+      "loss": 0.1873,
+      "step": 392
+    },
+    {
+      "epoch": 0.3027857829010567,
+      "grad_norm": 1.0461247560001137,
+      "learning_rate": 6.999987397084384e-06,
+      "loss": 0.1844,
+      "step": 394
+    },
+    {
+      "epoch": 0.3043227665706052,
+      "grad_norm": 1.074110109904979,
+      "learning_rate": 6.999964991938417e-06,
+      "loss": 0.1867,
+      "step": 396
+    },
+    {
+      "epoch": 0.3058597502401537,
+      "grad_norm": 1.1795081212239336,
+      "learning_rate": 6.999931384309108e-06,
+      "loss": 0.1927,
+      "step": 398
+    },
+    {
+      "epoch": 0.30739673390970224,
+      "grad_norm": 1.0767670064958226,
+      "learning_rate": 6.999886574304027e-06,
+      "loss": 0.1756,
+      "step": 400
+    },
+    {
+      "epoch": 0.3089337175792507,
+      "grad_norm": 1.1180599613087716,
+      "learning_rate": 6.999830562066599e-06,
+      "loss": 0.1777,
+      "step": 402
+    },
+    {
+      "epoch": 0.3104707012487992,
+      "grad_norm": 1.1283551526882591,
+      "learning_rate": 6.999763347776102e-06,
+      "loss": 0.1806,
+      "step": 404
+    },
+    {
+      "epoch": 0.31200768491834774,
+      "grad_norm": 1.2128389787316596,
+      "learning_rate": 6.999684931647677e-06,
+      "loss": 0.1989,
+      "step": 406
+    },
+    {
+      "epoch": 0.31354466858789626,
+      "grad_norm": 1.0231317030587057,
+      "learning_rate": 6.999595313932308e-06,
+      "loss": 0.1737,
+      "step": 408
+    },
+    {
+      "epoch": 0.3150816522574448,
+      "grad_norm": 1.2500344807446935,
+      "learning_rate": 6.999494494916842e-06,
+      "loss": 0.1906,
+      "step": 410
+    },
+    {
+      "epoch": 0.3166186359269933,
+      "grad_norm": 1.1142795967528607,
+      "learning_rate": 6.999382474923973e-06,
+      "loss": 0.1754,
+      "step": 412
+    },
+    {
+      "epoch": 0.31815561959654176,
+      "grad_norm": 1.0688175628422532,
+      "learning_rate": 6.999259254312248e-06,
+      "loss": 0.1738,
+      "step": 414
+    },
+    {
+      "epoch": 0.3196926032660903,
+      "grad_norm": 1.075624317236362,
+      "learning_rate": 6.999124833476066e-06,
+      "loss": 0.1785,
+      "step": 416
+    },
+    {
+      "epoch": 0.3212295869356388,
+      "grad_norm": 1.0898379845370674,
+      "learning_rate": 6.9989792128456716e-06,
+      "loss": 0.1992,
+      "step": 418
+    },
+    {
+      "epoch": 0.3227665706051873,
+      "grad_norm": 1.027575992474251,
+      "learning_rate": 6.998822392887159e-06,
+      "loss": 0.1773,
+      "step": 420
+    },
+    {
+      "epoch": 0.32430355427473584,
+      "grad_norm": 1.1417631076293755,
+      "learning_rate": 6.9986543741024684e-06,
+      "loss": 0.1905,
+      "step": 422
+    },
+    {
+      "epoch": 0.32584053794428436,
+      "grad_norm": 1.0666179275230614,
+      "learning_rate": 6.998475157029385e-06,
+      "loss": 0.1771,
+      "step": 424
+    },
+    {
+      "epoch": 0.3273775216138329,
+      "grad_norm": 1.178184842738573,
+      "learning_rate": 6.998284742241536e-06,
+      "loss": 0.2027,
+      "step": 426
+    },
+    {
+      "epoch": 0.32891450528338134,
+      "grad_norm": 1.1389042968070178,
+      "learning_rate": 6.99808313034839e-06,
+      "loss": 0.1761,
+      "step": 428
+    },
+    {
+      "epoch": 0.33045148895292986,
+      "grad_norm": 1.1267197030584097,
+      "learning_rate": 6.997870321995255e-06,
+      "loss": 0.1907,
+      "step": 430
+    },
+    {
+      "epoch": 0.3319884726224784,
+      "grad_norm": 1.07793801111443,
+      "learning_rate": 6.9976463178632756e-06,
+      "loss": 0.1826,
+      "step": 432
+    },
+    {
+      "epoch": 0.3335254562920269,
+      "grad_norm": 1.0872443642223155,
+      "learning_rate": 6.99741111866943e-06,
+      "loss": 0.1913,
+      "step": 434
+    },
+    {
+      "epoch": 0.3350624399615754,
+      "grad_norm": 1.1636846292387626,
+      "learning_rate": 6.997164725166531e-06,
+      "loss": 0.1891,
+      "step": 436
+    },
+    {
+      "epoch": 0.33659942363112394,
+      "grad_norm": 1.1189017028456263,
+      "learning_rate": 6.996907138143219e-06,
+      "loss": 0.178,
+      "step": 438
+    },
+    {
+      "epoch": 0.3381364073006724,
+      "grad_norm": 1.0292145655593512,
+      "learning_rate": 6.996638358423965e-06,
+      "loss": 0.1803,
+      "step": 440
+    },
+    {
+      "epoch": 0.3396733909702209,
+      "grad_norm": 1.0390558930376024,
+      "learning_rate": 6.996358386869064e-06,
+      "loss": 0.1881,
+      "step": 442
+    },
+    {
+      "epoch": 0.34121037463976944,
+      "grad_norm": 1.1559870900026559,
+      "learning_rate": 6.996067224374631e-06,
+      "loss": 0.1825,
+      "step": 444
+    },
+    {
+      "epoch": 0.34274735830931796,
+      "grad_norm": 1.1284033103165994,
+      "learning_rate": 6.995764871872603e-06,
+      "loss": 0.1985,
+      "step": 446
+    },
+    {
+      "epoch": 0.3442843419788665,
+      "grad_norm": 1.0182449603206558,
+      "learning_rate": 6.995451330330732e-06,
+      "loss": 0.1963,
+      "step": 448
+    },
+    {
+      "epoch": 0.345821325648415,
+      "grad_norm": 1.0080327819510069,
+      "learning_rate": 6.995126600752583e-06,
+      "loss": 0.189,
+      "step": 450
+    },
+    {
+      "epoch": 0.3473583093179635,
+      "grad_norm": 1.0876359081678704,
+      "learning_rate": 6.994790684177531e-06,
+      "loss": 0.1672,
+      "step": 452
+    },
+    {
+      "epoch": 0.348895292987512,
+      "grad_norm": 1.0950900793628788,
+      "learning_rate": 6.99444358168076e-06,
+      "loss": 0.196,
+      "step": 454
+    },
+    {
+      "epoch": 0.3504322766570605,
+      "grad_norm": 1.0087724624388277,
+      "learning_rate": 6.9940852943732534e-06,
+      "loss": 0.1779,
+      "step": 456
+    },
+    {
+      "epoch": 0.351969260326609,
+      "grad_norm": 1.0777984809693466,
+      "learning_rate": 6.993715823401798e-06,
+      "loss": 0.1867,
+      "step": 458
+    },
+    {
+      "epoch": 0.35350624399615754,
+      "grad_norm": 1.0121221411630168,
+      "learning_rate": 6.993335169948972e-06,
+      "loss": 0.1715,
+      "step": 460
+    },
+    {
+      "epoch": 0.35504322766570606,
+      "grad_norm": 1.011168047129122,
+      "learning_rate": 6.992943335233152e-06,
+      "loss": 0.1707,
+      "step": 462
+    },
+    {
+      "epoch": 0.3565802113352546,
+      "grad_norm": 1.0120465143177686,
+      "learning_rate": 6.992540320508498e-06,
+      "loss": 0.1875,
+      "step": 464
+    },
+    {
+      "epoch": 0.3581171950048031,
+      "grad_norm": 0.9550620727028469,
+      "learning_rate": 6.992126127064956e-06,
+      "loss": 0.1635,
+      "step": 466
+    },
+    {
+      "epoch": 0.35965417867435157,
+      "grad_norm": 1.1350686459542934,
+      "learning_rate": 6.9917007562282535e-06,
+      "loss": 0.1751,
+      "step": 468
+    },
+    {
+      "epoch": 0.3611911623439001,
+      "grad_norm": 1.1572256160083796,
+      "learning_rate": 6.991264209359891e-06,
+      "loss": 0.1883,
+      "step": 470
+    },
+    {
+      "epoch": 0.3627281460134486,
+      "grad_norm": 1.021485377857385,
+      "learning_rate": 6.9908164878571425e-06,
+      "loss": 0.1798,
+      "step": 472
+    },
+    {
+      "epoch": 0.3642651296829971,
+      "grad_norm": 0.9632560183914399,
+      "learning_rate": 6.99035759315305e-06,
+      "loss": 0.1655,
+      "step": 474
+    },
+    {
+      "epoch": 0.36580211335254564,
+      "grad_norm": 1.0013270641998204,
+      "learning_rate": 6.989887526716415e-06,
+      "loss": 0.1927,
+      "step": 476
+    },
+    {
+      "epoch": 0.36733909702209416,
+      "grad_norm": 0.9745686321374288,
+      "learning_rate": 6.9894062900517996e-06,
+      "loss": 0.1674,
+      "step": 478
+    },
+    {
+      "epoch": 0.3688760806916426,
+      "grad_norm": 1.0407042923473835,
+      "learning_rate": 6.988913884699518e-06,
+      "loss": 0.1925,
+      "step": 480
+    },
+    {
+      "epoch": 0.37041306436119115,
+      "grad_norm": 1.0062661091753247,
+      "learning_rate": 6.988410312235632e-06,
+      "loss": 0.1965,
+      "step": 482
+    },
+    {
+      "epoch": 0.37195004803073967,
+      "grad_norm": 1.1013102934828658,
+      "learning_rate": 6.987895574271948e-06,
+      "loss": 0.1998,
+      "step": 484
+    },
+    {
+      "epoch": 0.3734870317002882,
+      "grad_norm": 1.0597643936100356,
+      "learning_rate": 6.987369672456009e-06,
+      "loss": 0.1772,
+      "step": 486
+    },
+    {
+      "epoch": 0.3750240153698367,
+      "grad_norm": 0.9866253310752731,
+      "learning_rate": 6.986832608471089e-06,
+      "loss": 0.1673,
+      "step": 488
+    },
+    {
+      "epoch": 0.3765609990393852,
+      "grad_norm": 1.0602452001077722,
+      "learning_rate": 6.986284384036193e-06,
+      "loss": 0.1739,
+      "step": 490
+    },
+    {
+      "epoch": 0.37809798270893374,
+      "grad_norm": 1.041644897151479,
+      "learning_rate": 6.985725000906045e-06,
+      "loss": 0.1695,
+      "step": 492
+    },
+    {
+      "epoch": 0.3796349663784822,
+      "grad_norm": 0.952300652923444,
+      "learning_rate": 6.985154460871086e-06,
+      "loss": 0.1801,
+      "step": 494
+    },
+    {
+      "epoch": 0.3811719500480307,
+      "grad_norm": 0.93795302994956,
+      "learning_rate": 6.984572765757467e-06,
+      "loss": 0.1691,
+      "step": 496
+    },
+    {
+      "epoch": 0.38270893371757925,
+      "grad_norm": 0.9322450383167038,
+      "learning_rate": 6.983979917427043e-06,
+      "loss": 0.1789,
+      "step": 498
+    },
+    {
+      "epoch": 0.38424591738712777,
+      "grad_norm": 1.0051761272259416,
+      "learning_rate": 6.98337591777737e-06,
+      "loss": 0.1713,
+      "step": 500
+    },
+    {
+      "epoch": 0.38424591738712777,
+      "eval_loss": 0.16142325103282928,
+      "eval_runtime": 365.2501,
+      "eval_samples_per_second": 50.664,
+      "eval_steps_per_second": 6.335,
+      "step": 500
+    },
+    {
+      "epoch": 0.3857829010566763,
+      "grad_norm": 1.008256178397373,
+      "learning_rate": 6.982760768741694e-06,
+      "loss": 0.178,
+      "step": 502
+    },
+    {
+      "epoch": 0.3873198847262248,
+      "grad_norm": 0.9324814287025068,
+      "learning_rate": 6.982134472288947e-06,
+      "loss": 0.1633,
+      "step": 504
+    },
+    {
+      "epoch": 0.38885686839577327,
+      "grad_norm": 1.0289506080987343,
+      "learning_rate": 6.981497030423744e-06,
+      "loss": 0.1805,
+      "step": 506
+    },
+    {
+      "epoch": 0.3903938520653218,
+      "grad_norm": 0.9347810671889286,
+      "learning_rate": 6.980848445186369e-06,
+      "loss": 0.1727,
+      "step": 508
+    },
+    {
+      "epoch": 0.3919308357348703,
+      "grad_norm": 1.052017862885602,
+      "learning_rate": 6.980188718652778e-06,
+      "loss": 0.1797,
+      "step": 510
+    },
+    {
+      "epoch": 0.39346781940441883,
+      "grad_norm": 1.056395915175445,
+      "learning_rate": 6.9795178529345855e-06,
+      "loss": 0.186,
+      "step": 512
+    },
+    {
+      "epoch": 0.39500480307396735,
+      "grad_norm": 1.0283503364288544,
+      "learning_rate": 6.978835850179057e-06,
+      "loss": 0.1844,
+      "step": 514
+    },
+    {
+      "epoch": 0.39654178674351587,
+      "grad_norm": 0.9433928178661478,
+      "learning_rate": 6.978142712569109e-06,
+      "loss": 0.1719,
+      "step": 516
+    },
+    {
+      "epoch": 0.3980787704130644,
+      "grad_norm": 0.9623302440193999,
+      "learning_rate": 6.9774384423232945e-06,
+      "loss": 0.1613,
+      "step": 518
+    },
+    {
+      "epoch": 0.39961575408261285,
+      "grad_norm": 1.0414121844166375,
+      "learning_rate": 6.976723041695802e-06,
+      "loss": 0.1841,
+      "step": 520
+    },
+    {
+      "epoch": 0.40115273775216137,
+      "grad_norm": 1.0155405045344752,
+      "learning_rate": 6.9759965129764425e-06,
+      "loss": 0.1706,
+      "step": 522
+    },
+    {
+      "epoch": 0.4026897214217099,
+      "grad_norm": 1.0347436462429636,
+      "learning_rate": 6.975258858490648e-06,
+      "loss": 0.1898,
+      "step": 524
+    },
+    {
+      "epoch": 0.4042267050912584,
+      "grad_norm": 0.9905491545072838,
+      "learning_rate": 6.974510080599458e-06,
+      "loss": 0.1701,
+      "step": 526
+    },
+    {
+      "epoch": 0.40576368876080693,
+      "grad_norm": 1.0065784769120174,
+      "learning_rate": 6.973750181699518e-06,
+      "loss": 0.173,
+      "step": 528
+    },
+    {
+      "epoch": 0.40730067243035545,
+      "grad_norm": 1.0028920218026682,
+      "learning_rate": 6.972979164223069e-06,
+      "loss": 0.1815,
+      "step": 530
+    },
+    {
+      "epoch": 0.4088376560999039,
+      "grad_norm": 0.9926648764778461,
+      "learning_rate": 6.972197030637938e-06,
+      "loss": 0.1701,
+      "step": 532
+    },
+    {
+      "epoch": 0.41037463976945243,
+      "grad_norm": 0.9362580434331931,
+      "learning_rate": 6.971403783447532e-06,
+      "loss": 0.162,
+      "step": 534
+    },
+    {
+      "epoch": 0.41191162343900095,
+      "grad_norm": 1.0313345974784995,
+      "learning_rate": 6.97059942519083e-06,
+      "loss": 0.1793,
+      "step": 536
+    },
+    {
+      "epoch": 0.41344860710854947,
+      "grad_norm": 1.053038519408416,
+      "learning_rate": 6.969783958442376e-06,
+      "loss": 0.1744,
+      "step": 538
+    },
+    {
+      "epoch": 0.414985590778098,
+      "grad_norm": 1.09503010697989,
+      "learning_rate": 6.968957385812268e-06,
+      "loss": 0.1837,
+      "step": 540
+    },
+    {
+      "epoch": 0.4165225744476465,
+      "grad_norm": 0.9963387156020179,
+      "learning_rate": 6.968119709946151e-06,
+      "loss": 0.1824,
+      "step": 542
+    },
+    {
+      "epoch": 0.41805955811719503,
+      "grad_norm": 0.9881607706879252,
+      "learning_rate": 6.9672709335252075e-06,
+      "loss": 0.1864,
+      "step": 544
+    },
+    {
+      "epoch": 0.4195965417867435,
+      "grad_norm": 0.9849697365361947,
+      "learning_rate": 6.966411059266153e-06,
+      "loss": 0.1676,
+      "step": 546
+    },
+    {
+      "epoch": 0.421133525456292,
+      "grad_norm": 0.99285568761607,
+      "learning_rate": 6.965540089921224e-06,
+      "loss": 0.1788,
+      "step": 548
+    },
+    {
+      "epoch": 0.42267050912584053,
+      "grad_norm": 1.0020197625846368,
+      "learning_rate": 6.964658028278167e-06,
+      "loss": 0.1717,
+      "step": 550
+    },
+    {
+      "epoch": 0.42420749279538905,
+      "grad_norm": 1.0446372613390171,
+      "learning_rate": 6.963764877160232e-06,
+      "loss": 0.1724,
+      "step": 552
+    },
+    {
+      "epoch": 0.42574447646493757,
+      "grad_norm": 1.0489345870285778,
+      "learning_rate": 6.962860639426168e-06,
+      "loss": 0.1751,
+      "step": 554
+    },
+    {
+      "epoch": 0.4272814601344861,
+      "grad_norm": 0.9913965985343988,
+      "learning_rate": 6.9619453179702036e-06,
+      "loss": 0.1753,
+      "step": 556
+    },
+    {
+      "epoch": 0.42881844380403455,
+      "grad_norm": 0.9746372697415868,
+      "learning_rate": 6.9610189157220465e-06,
+      "loss": 0.1849,
+      "step": 558
+    },
+    {
+      "epoch": 0.4303554274735831,
+      "grad_norm": 1.0035507982106044,
+      "learning_rate": 6.960081435646872e-06,
+      "loss": 0.1887,
+      "step": 560
+    },
+    {
+      "epoch": 0.4318924111431316,
+      "grad_norm": 1.0189535936717342,
+      "learning_rate": 6.95913288074531e-06,
+      "loss": 0.1761,
+      "step": 562
+    },
+    {
+      "epoch": 0.4334293948126801,
+      "grad_norm": 1.0604859650996308,
+      "learning_rate": 6.958173254053442e-06,
+      "loss": 0.1698,
+      "step": 564
+    },
+    {
+      "epoch": 0.43496637848222863,
+      "grad_norm": 1.04513958829485,
+      "learning_rate": 6.957202558642782e-06,
+      "loss": 0.1842,
+      "step": 566
+    },
+    {
+      "epoch": 0.43650336215177715,
+      "grad_norm": 0.9420876517823673,
+      "learning_rate": 6.9562207976202775e-06,
+      "loss": 0.1614,
+      "step": 568
+    },
+    {
+      "epoch": 0.43804034582132567,
+      "grad_norm": 1.0001567101709503,
+      "learning_rate": 6.9552279741282916e-06,
+      "loss": 0.1692,
+      "step": 570
+    },
+    {
+      "epoch": 0.43957732949087414,
+      "grad_norm": 1.0163931461175468,
+      "learning_rate": 6.954224091344593e-06,
+      "loss": 0.1687,
+      "step": 572
+    },
+    {
+      "epoch": 0.44111431316042266,
+      "grad_norm": 0.98643009928601,
+      "learning_rate": 6.953209152482355e-06,
+      "loss": 0.1853,
+      "step": 574
+    },
+    {
+      "epoch": 0.4426512968299712,
+      "grad_norm": 0.9353513380661882,
+      "learning_rate": 6.952183160790133e-06,
+      "loss": 0.1675,
+      "step": 576
+    },
+    {
+      "epoch": 0.4441882804995197,
+      "grad_norm": 0.9020207899109111,
+      "learning_rate": 6.951146119551859e-06,
+      "loss": 0.1679,
+      "step": 578
+    },
+    {
+      "epoch": 0.4457252641690682,
+      "grad_norm": 0.9841310128546708,
+      "learning_rate": 6.950098032086837e-06,
+      "loss": 0.1694,
+      "step": 580
+    },
+    {
+      "epoch": 0.44726224783861673,
+      "grad_norm": 1.0255232655982072,
+      "learning_rate": 6.949038901749723e-06,
+      "loss": 0.1736,
+      "step": 582
+    },
+    {
+      "epoch": 0.44879923150816525,
+      "grad_norm": 0.9032841104351299,
+      "learning_rate": 6.947968731930519e-06,
+      "loss": 0.163,
+      "step": 584
+    },
+    {
+      "epoch": 0.4503362151777137,
+      "grad_norm": 0.9005602498337584,
+      "learning_rate": 6.946887526054563e-06,
+      "loss": 0.1593,
+      "step": 586
+    },
+    {
+      "epoch": 0.45187319884726224,
+      "grad_norm": 0.9439866081688895,
+      "learning_rate": 6.945795287582514e-06,
+      "loss": 0.172,
+      "step": 588
+    },
+    {
+      "epoch": 0.45341018251681076,
+      "grad_norm": 1.015175377754709,
+      "learning_rate": 6.9446920200103465e-06,
+      "loss": 0.1658,
+      "step": 590
+    },
+    {
+      "epoch": 0.4549471661863593,
+      "grad_norm": 0.9730430866278983,
+      "learning_rate": 6.943577726869334e-06,
+      "loss": 0.1607,
+      "step": 592
+    },
+    {
+      "epoch": 0.4564841498559078,
+      "grad_norm": 0.9619897655090097,
+      "learning_rate": 6.942452411726042e-06,
+      "loss": 0.1799,
+      "step": 594
+    },
+    {
+      "epoch": 0.4580211335254563,
+      "grad_norm": 0.9934784984855407,
+      "learning_rate": 6.941316078182312e-06,
+      "loss": 0.1692,
+      "step": 596
+    },
+    {
+      "epoch": 0.4595581171950048,
+      "grad_norm": 0.9337603705095543,
+      "learning_rate": 6.940168729875255e-06,
+      "loss": 0.1787,
+      "step": 598
+    },
+    {
+      "epoch": 0.4610951008645533,
+      "grad_norm": 0.9057632316451069,
+      "learning_rate": 6.939010370477235e-06,
+      "loss": 0.1713,
+      "step": 600
+    },
+    {
+      "epoch": 0.4626320845341018,
+      "grad_norm": 0.989678630378062,
+      "learning_rate": 6.9378410036958635e-06,
+      "loss": 0.1761,
+      "step": 602
+    },
+    {
+      "epoch": 0.46416906820365034,
+      "grad_norm": 0.9385046490524968,
+      "learning_rate": 6.936660633273979e-06,
+      "loss": 0.1745,
+      "step": 604
+    },
+    {
+      "epoch": 0.46570605187319886,
+      "grad_norm": 0.955462629075025,
+      "learning_rate": 6.935469262989644e-06,
+      "loss": 0.1711,
+      "step": 606
+    },
+    {
+      "epoch": 0.4672430355427474,
+      "grad_norm": 0.8795030512447493,
+      "learning_rate": 6.9342668966561245e-06,
+      "loss": 0.169,
+      "step": 608
+    },
+    {
+      "epoch": 0.4687800192122959,
+      "grad_norm": 0.9495870025972418,
+      "learning_rate": 6.933053538121886e-06,
+      "loss": 0.1682,
+      "step": 610
+    },
+    {
+      "epoch": 0.47031700288184436,
+      "grad_norm": 0.9330347326137807,
+      "learning_rate": 6.931829191270576e-06,
+      "loss": 0.1662,
+      "step": 612
+    },
+    {
+      "epoch": 0.4718539865513929,
+      "grad_norm": 0.9017675551372555,
+      "learning_rate": 6.930593860021012e-06,
+      "loss": 0.1628,
+      "step": 614
+    },
+    {
+      "epoch": 0.4733909702209414,
+      "grad_norm": 0.9466561162964194,
+      "learning_rate": 6.929347548327168e-06,
+      "loss": 0.1807,
+      "step": 616
+    },
+    {
+      "epoch": 0.4749279538904899,
+      "grad_norm": 0.9881577337000146,
+      "learning_rate": 6.928090260178169e-06,
+      "loss": 0.1872,
+      "step": 618
+    },
+    {
+      "epoch": 0.47646493756003844,
+      "grad_norm": 0.9646614031070989,
+      "learning_rate": 6.926821999598266e-06,
+      "loss": 0.1709,
+      "step": 620
+    },
+    {
+      "epoch": 0.47800192122958696,
+      "grad_norm": 0.9609197980971407,
+      "learning_rate": 6.9255427706468375e-06,
+      "loss": 0.1774,
+      "step": 622
+    },
+    {
+      "epoch": 0.4795389048991354,
+      "grad_norm": 0.9541380042708364,
+      "learning_rate": 6.92425257741836e-06,
+      "loss": 0.1777,
+      "step": 624
+    },
+    {
+      "epoch": 0.48107588856868394,
+      "grad_norm": 0.898383778954144,
+      "learning_rate": 6.922951424042412e-06,
+      "loss": 0.1691,
+      "step": 626
+    },
+    {
+      "epoch": 0.48261287223823246,
+      "grad_norm": 0.960885030497367,
+      "learning_rate": 6.921639314683648e-06,
+      "loss": 0.1576,
+      "step": 628
+    },
+    {
+      "epoch": 0.484149855907781,
+      "grad_norm": 0.9454433168177289,
+      "learning_rate": 6.92031625354179e-06,
+      "loss": 0.1796,
+      "step": 630
+    },
+    {
+      "epoch": 0.4856868395773295,
+      "grad_norm": 0.9852240424660407,
+      "learning_rate": 6.918982244851616e-06,
+      "loss": 0.171,
+      "step": 632
+    },
+    {
+      "epoch": 0.487223823246878,
+      "grad_norm": 1.0414345737302746,
+      "learning_rate": 6.917637292882944e-06,
+      "loss": 0.1776,
+      "step": 634
+    },
+    {
+      "epoch": 0.48876080691642654,
+      "grad_norm": 0.9675949339406987,
+      "learning_rate": 6.916281401940615e-06,
+      "loss": 0.1806,
+      "step": 636
+    },
+    {
+      "epoch": 0.490297790585975,
+      "grad_norm": 0.982314844842542,
+      "learning_rate": 6.914914576364487e-06,
+      "loss": 0.1785,
+      "step": 638
+    },
+    {
+      "epoch": 0.4918347742555235,
+      "grad_norm": 0.9775652590949767,
+      "learning_rate": 6.913536820529416e-06,
+      "loss": 0.1751,
+      "step": 640
+    },
+    {
+      "epoch": 0.49337175792507204,
+      "grad_norm": 0.8855730733866225,
+      "learning_rate": 6.912148138845241e-06,
+      "loss": 0.1659,
+      "step": 642
+    },
+    {
+      "epoch": 0.49490874159462056,
+      "grad_norm": 1.0573307642737217,
+      "learning_rate": 6.910748535756774e-06,
+      "loss": 0.1781,
+      "step": 644
+    },
+    {
+      "epoch": 0.4964457252641691,
+      "grad_norm": 0.948011615943263,
+      "learning_rate": 6.909338015743782e-06,
+      "loss": 0.1521,
+      "step": 646
+    },
+    {
+      "epoch": 0.4979827089337176,
+      "grad_norm": 1.006209266616953,
+      "learning_rate": 6.907916583320976e-06,
+      "loss": 0.1671,
+      "step": 648
+    },
+    {
+      "epoch": 0.49951969260326606,
+      "grad_norm": 0.9744730979666755,
+      "learning_rate": 6.906484243037992e-06,
+      "loss": 0.1777,
+      "step": 650
+    },
+    {
+      "epoch": 0.5010566762728146,
+      "grad_norm": 1.0140793456518158,
+      "learning_rate": 6.9050409994793835e-06,
+      "loss": 0.1724,
+      "step": 652
+    },
+    {
+      "epoch": 0.5025936599423632,
+      "grad_norm": 0.9953513391992594,
+      "learning_rate": 6.903586857264598e-06,
+      "loss": 0.1612,
+      "step": 654
+    },
+    {
+      "epoch": 0.5041306436119116,
+      "grad_norm": 0.9120318862748673,
+      "learning_rate": 6.9021218210479715e-06,
+      "loss": 0.1565,
+      "step": 656
+    },
+    {
+      "epoch": 0.5056676272814601,
+      "grad_norm": 0.8703090747149029,
+      "learning_rate": 6.900645895518703e-06,
+      "loss": 0.1616,
+      "step": 658
+    },
+    {
+      "epoch": 0.5072046109510087,
+      "grad_norm": 0.9188361211761223,
+      "learning_rate": 6.899159085400851e-06,
+      "loss": 0.1702,
+      "step": 660
+    },
+    {
+      "epoch": 0.5087415946205571,
+      "grad_norm": 0.9052706219772494,
+      "learning_rate": 6.897661395453309e-06,
+      "loss": 0.1671,
+      "step": 662
+    },
+    {
+      "epoch": 0.5102785782901057,
+      "grad_norm": 0.9790524922779839,
+      "learning_rate": 6.896152830469797e-06,
+      "loss": 0.1707,
+      "step": 664
+    },
+    {
+      "epoch": 0.5118155619596542,
+      "grad_norm": 0.9737405664828704,
+      "learning_rate": 6.894633395278839e-06,
+      "loss": 0.1749,
+      "step": 666
+    },
+    {
+      "epoch": 0.5133525456292027,
+      "grad_norm": 0.8786179004815331,
+      "learning_rate": 6.893103094743758e-06,
+      "loss": 0.1662,
+      "step": 668
+    },
+    {
+      "epoch": 0.5148895292987512,
+      "grad_norm": 0.9648594871932485,
+      "learning_rate": 6.891561933762648e-06,
+      "loss": 0.1777,
+      "step": 670
+    },
+    {
+      "epoch": 0.5164265129682997,
+      "grad_norm": 0.9273934239395568,
+      "learning_rate": 6.8900099172683675e-06,
+      "loss": 0.1762,
+      "step": 672
+    },
+    {
+      "epoch": 0.5179634966378482,
+      "grad_norm": 0.9132188955471342,
+      "learning_rate": 6.8884470502285195e-06,
+      "loss": 0.1638,
+      "step": 674
+    },
+    {
+      "epoch": 0.5195004803073967,
+      "grad_norm": 1.0216457284197233,
+      "learning_rate": 6.886873337645439e-06,
+      "loss": 0.1663,
+      "step": 676
+    },
+    {
+      "epoch": 0.5210374639769453,
+      "grad_norm": 0.9342404527623865,
+      "learning_rate": 6.885288784556172e-06,
+      "loss": 0.1692,
+      "step": 678
+    },
+    {
+      "epoch": 0.5225744476464937,
+      "grad_norm": 0.8975547579280767,
+      "learning_rate": 6.883693396032463e-06,
+      "loss": 0.1669,
+      "step": 680
+    },
+    {
+      "epoch": 0.5241114313160423,
+      "grad_norm": 0.8921639188810697,
+      "learning_rate": 6.88208717718074e-06,
+      "loss": 0.1665,
+      "step": 682
+    },
+    {
+      "epoch": 0.5256484149855908,
+      "grad_norm": 0.9687694730815803,
+      "learning_rate": 6.880470133142094e-06,
+      "loss": 0.1674,
+      "step": 684
+    },
+    {
+      "epoch": 0.5271853986551392,
+      "grad_norm": 0.9456241155960425,
+      "learning_rate": 6.878842269092263e-06,
+      "loss": 0.1685,
+      "step": 686
+    },
+    {
+      "epoch": 0.5287223823246878,
+      "grad_norm": 1.0077089550967064,
+      "learning_rate": 6.877203590241621e-06,
+      "loss": 0.1755,
+      "step": 688
+    },
+    {
+      "epoch": 0.5302593659942363,
+      "grad_norm": 0.9499550767483984,
+      "learning_rate": 6.875554101835156e-06,
+      "loss": 0.1727,
+      "step": 690
+    },
+    {
+      "epoch": 0.5317963496637849,
+      "grad_norm": 0.9677648466898067,
+      "learning_rate": 6.873893809152453e-06,
+      "loss": 0.1768,
+      "step": 692
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.9430584650028841,
+      "learning_rate": 6.872222717507679e-06,
+      "loss": 0.1736,
+      "step": 694
+    },
+    {
+      "epoch": 0.5348703170028818,
+      "grad_norm": 0.8724957576172335,
+      "learning_rate": 6.870540832249567e-06,
+      "loss": 0.1586,
+      "step": 696
+    },
+    {
+      "epoch": 0.5364073006724304,
+      "grad_norm": 0.8628795594957057,
+      "learning_rate": 6.868848158761398e-06,
+      "loss": 0.1617,
+      "step": 698
+    },
+    {
+      "epoch": 0.5379442843419788,
+      "grad_norm": 0.9609714958221978,
+      "learning_rate": 6.867144702460982e-06,
+      "loss": 0.1831,
+      "step": 700
+    },
+    {
+      "epoch": 0.5394812680115274,
+      "grad_norm": 0.9942806006570933,
+      "learning_rate": 6.865430468800642e-06,
+      "loss": 0.181,
+      "step": 702
+    },
+    {
+      "epoch": 0.5410182516810759,
+      "grad_norm": 0.9153357695300892,
+      "learning_rate": 6.863705463267197e-06,
+      "loss": 0.1616,
+      "step": 704
+    },
+    {
+      "epoch": 0.5425552353506244,
+      "grad_norm": 0.903669480913219,
+      "learning_rate": 6.861969691381943e-06,
+      "loss": 0.176,
+      "step": 706
+    },
+    {
+      "epoch": 0.5440922190201729,
+      "grad_norm": 0.88720714763806,
+      "learning_rate": 6.860223158700639e-06,
+      "loss": 0.17,
+      "step": 708
+    },
+    {
+      "epoch": 0.5456292026897214,
+      "grad_norm": 0.9051200845778055,
+      "learning_rate": 6.8584658708134825e-06,
+      "loss": 0.1686,
+      "step": 710
+    },
+    {
+      "epoch": 0.54716618635927,
+      "grad_norm": 0.9513875655430767,
+      "learning_rate": 6.856697833345101e-06,
+      "loss": 0.1723,
+      "step": 712
+    },
+    {
+      "epoch": 0.5487031700288184,
+      "grad_norm": 0.9099315275605531,
+      "learning_rate": 6.8549190519545206e-06,
+      "loss": 0.1641,
+      "step": 714
+    },
+    {
+      "epoch": 0.550240153698367,
+      "grad_norm": 0.8702057226720046,
+      "learning_rate": 6.8531295323351655e-06,
+      "loss": 0.1583,
+      "step": 716
+    },
+    {
+      "epoch": 0.5517771373679154,
+      "grad_norm": 0.9086133226107327,
+      "learning_rate": 6.851329280214823e-06,
+      "loss": 0.1637,
+      "step": 718
+    },
+    {
+      "epoch": 0.553314121037464,
+      "grad_norm": 0.908182677978134,
+      "learning_rate": 6.8495183013556365e-06,
+      "loss": 0.1596,
+      "step": 720
+    },
+    {
+      "epoch": 0.5548511047070125,
+      "grad_norm": 0.9856738002008828,
+      "learning_rate": 6.847696601554079e-06,
+      "loss": 0.1701,
+      "step": 722
+    },
+    {
+      "epoch": 0.556388088376561,
+      "grad_norm": 0.9119819563123349,
+      "learning_rate": 6.845864186640944e-06,
+      "loss": 0.1563,
+      "step": 724
+    },
+    {
+      "epoch": 0.5579250720461095,
+      "grad_norm": 0.9080236111588762,
+      "learning_rate": 6.844021062481314e-06,
+      "loss": 0.1753,
+      "step": 726
+    },
+    {
+      "epoch": 0.559462055715658,
+      "grad_norm": 0.8990390737752408,
+      "learning_rate": 6.842167234974556e-06,
+      "loss": 0.1647,
+      "step": 728
+    },
+    {
+      "epoch": 0.5609990393852066,
+      "grad_norm": 0.9761008697066089,
+      "learning_rate": 6.840302710054292e-06,
+      "loss": 0.1708,
+      "step": 730
+    },
+    {
+      "epoch": 0.562536023054755,
+      "grad_norm": 0.9552176939231856,
+      "learning_rate": 6.838427493688384e-06,
+      "loss": 0.1675,
+      "step": 732
+    },
+    {
+      "epoch": 0.5640730067243036,
+      "grad_norm": 0.8757204503224848,
+      "learning_rate": 6.836541591878915e-06,
+      "loss": 0.1624,
+      "step": 734
+    },
+    {
+      "epoch": 0.5656099903938521,
+      "grad_norm": 0.8887299929134649,
+      "learning_rate": 6.834645010662169e-06,
+      "loss": 0.1579,
+      "step": 736
+    },
+    {
+      "epoch": 0.5671469740634005,
+      "grad_norm": 0.9268495074535574,
+      "learning_rate": 6.832737756108613e-06,
+      "loss": 0.1817,
+      "step": 738
+    },
+    {
+      "epoch": 0.5686839577329491,
+      "grad_norm": 0.8930331129406648,
+      "learning_rate": 6.830819834322875e-06,
+      "loss": 0.1722,
+      "step": 740
+    },
+    {
+      "epoch": 0.5702209414024976,
+      "grad_norm": 0.9181728216635983,
+      "learning_rate": 6.828891251443729e-06,
+      "loss": 0.1751,
+      "step": 742
+    },
+    {
+      "epoch": 0.5717579250720461,
+      "grad_norm": 0.9429482896947388,
+      "learning_rate": 6.826952013644067e-06,
+      "loss": 0.1715,
+      "step": 744
+    },
+    {
+      "epoch": 0.5732949087415946,
+      "grad_norm": 0.9118821908902819,
+      "learning_rate": 6.825002127130891e-06,
+      "loss": 0.1696,
+      "step": 746
+    },
+    {
+      "epoch": 0.5748318924111432,
+      "grad_norm": 0.9132457028236339,
+      "learning_rate": 6.823041598145282e-06,
+      "loss": 0.1699,
+      "step": 748
+    },
+    {
+      "epoch": 0.5763688760806917,
+      "grad_norm": 0.8756070623659923,
+      "learning_rate": 6.821070432962387e-06,
+      "loss": 0.1635,
+      "step": 750
+    },
+    {
+      "epoch": 0.5779058597502401,
+      "grad_norm": 0.8779903149650536,
+      "learning_rate": 6.819088637891397e-06,
+      "loss": 0.1571,
+      "step": 752
+    },
+    {
+      "epoch": 0.5794428434197887,
+      "grad_norm": 0.951553518645419,
+      "learning_rate": 6.817096219275525e-06,
+      "loss": 0.177,
+      "step": 754
+    },
+    {
+      "epoch": 0.5809798270893372,
+      "grad_norm": 0.8815920502748197,
+      "learning_rate": 6.815093183491988e-06,
+      "loss": 0.1723,
+      "step": 756
+    },
+    {
+      "epoch": 0.5825168107588857,
+      "grad_norm": 0.8943776455140009,
+      "learning_rate": 6.813079536951986e-06,
+      "loss": 0.1675,
+      "step": 758
+    },
+    {
+      "epoch": 0.5840537944284342,
+      "grad_norm": 0.9042515471791829,
+      "learning_rate": 6.811055286100681e-06,
+      "loss": 0.1668,
+      "step": 760
+    },
+    {
+      "epoch": 0.5855907780979827,
+      "grad_norm": 0.8985203223108896,
+      "learning_rate": 6.809020437417178e-06,
+      "loss": 0.1683,
+      "step": 762
+    },
+    {
+      "epoch": 0.5871277617675312,
+      "grad_norm": 0.8854626460275703,
+      "learning_rate": 6.8069749974145e-06,
+      "loss": 0.1664,
+      "step": 764
+    },
+    {
+      "epoch": 0.5886647454370797,
+      "grad_norm": 0.9849706761614369,
+      "learning_rate": 6.804918972639572e-06,
+      "loss": 0.1603,
+      "step": 766
+    },
+    {
+      "epoch": 0.5902017291066283,
+      "grad_norm": 0.9492268475374204,
+      "learning_rate": 6.802852369673199e-06,
+      "loss": 0.1712,
+      "step": 768
+    },
+    {
+      "epoch": 0.5917387127761767,
+      "grad_norm": 0.9656447681787407,
+      "learning_rate": 6.8007751951300425e-06,
+      "loss": 0.1571,
+      "step": 770
+    },
+    {
+      "epoch": 0.5932756964457253,
+      "grad_norm": 0.8950523002371793,
+      "learning_rate": 6.798687455658602e-06,
+      "loss": 0.166,
+      "step": 772
+    },
+    {
+      "epoch": 0.5948126801152738,
+      "grad_norm": 0.9273600651692718,
+      "learning_rate": 6.79658915794119e-06,
+      "loss": 0.1617,
+      "step": 774
+    },
+    {
+      "epoch": 0.5963496637848222,
+      "grad_norm": 0.9160428728026682,
+      "learning_rate": 6.794480308693916e-06,
+      "loss": 0.1718,
+      "step": 776
+    },
+    {
+      "epoch": 0.5978866474543708,
+      "grad_norm": 0.9675151521212334,
+      "learning_rate": 6.792360914666662e-06,
+      "loss": 0.1618,
+      "step": 778
+    },
+    {
+      "epoch": 0.5994236311239193,
+      "grad_norm": 0.9613967630774912,
+      "learning_rate": 6.79023098264306e-06,
+      "loss": 0.1707,
+      "step": 780
+    },
+    {
+      "epoch": 0.6009606147934679,
+      "grad_norm": 0.8548759973613992,
+      "learning_rate": 6.7880905194404735e-06,
+      "loss": 0.1514,
+      "step": 782
+    },
+    {
+      "epoch": 0.6024975984630163,
+      "grad_norm": 0.890219782689619,
+      "learning_rate": 6.78593953190997e-06,
+      "loss": 0.1715,
+      "step": 784
+    },
+    {
+      "epoch": 0.6040345821325649,
+      "grad_norm": 0.9086155715910085,
+      "learning_rate": 6.783778026936305e-06,
+      "loss": 0.1663,
+      "step": 786
+    },
+    {
+      "epoch": 0.6055715658021134,
+      "grad_norm": 0.8095718809326027,
+      "learning_rate": 6.781606011437898e-06,
+      "loss": 0.1506,
+      "step": 788
+    },
+    {
+      "epoch": 0.6071085494716618,
+      "grad_norm": 0.8993402244627612,
+      "learning_rate": 6.779423492366808e-06,
+      "loss": 0.1778,
+      "step": 790
+    },
+    {
+      "epoch": 0.6086455331412104,
+      "grad_norm": 0.9179176055571641,
+      "learning_rate": 6.777230476708715e-06,
+      "loss": 0.1683,
+      "step": 792
+    },
+    {
+      "epoch": 0.6101825168107589,
+      "grad_norm": 0.880916421665006,
+      "learning_rate": 6.775026971482896e-06,
+      "loss": 0.1679,
+      "step": 794
+    },
+    {
+      "epoch": 0.6117195004803074,
+      "grad_norm": 0.9118630511257958,
+      "learning_rate": 6.7728129837422016e-06,
+      "loss": 0.1647,
+      "step": 796
+    },
+    {
+      "epoch": 0.6132564841498559,
+      "grad_norm": 0.9503223570669749,
+      "learning_rate": 6.770588520573034e-06,
+      "loss": 0.1653,
+      "step": 798
+    },
+    {
+      "epoch": 0.6147934678194045,
+      "grad_norm": 0.914092375436958,
+      "learning_rate": 6.768353589095325e-06,
+      "loss": 0.1706,
+      "step": 800
+    },
+    {
+      "epoch": 0.6163304514889529,
+      "grad_norm": 0.8780710561728171,
+      "learning_rate": 6.766108196462512e-06,
+      "loss": 0.1581,
+      "step": 802
+    },
+    {
+      "epoch": 0.6178674351585014,
+      "grad_norm": 0.9908640319756162,
+      "learning_rate": 6.763852349861517e-06,
+      "loss": 0.1746,
+      "step": 804
+    },
+    {
+      "epoch": 0.61940441882805,
+      "grad_norm": 0.9382036222959125,
+      "learning_rate": 6.761586056512721e-06,
+      "loss": 0.1467,
+      "step": 806
+    },
+    {
+      "epoch": 0.6209414024975984,
+      "grad_norm": 0.8796176529284295,
+      "learning_rate": 6.759309323669945e-06,
+      "loss": 0.1638,
+      "step": 808
+    },
+    {
+      "epoch": 0.622478386167147,
+      "grad_norm": 0.9402322649840099,
+      "learning_rate": 6.757022158620422e-06,
+      "loss": 0.1644,
+      "step": 810
+    },
+    {
+      "epoch": 0.6240153698366955,
+      "grad_norm": 0.9036305217983724,
+      "learning_rate": 6.754724568684775e-06,
+      "loss": 0.1625,
+      "step": 812
+    },
+    {
+      "epoch": 0.6255523535062439,
+      "grad_norm": 0.8200742046612647,
+      "learning_rate": 6.752416561216997e-06,
+      "loss": 0.1617,
+      "step": 814
+    },
+    {
+      "epoch": 0.6270893371757925,
+      "grad_norm": 0.9107960160473824,
+      "learning_rate": 6.750098143604423e-06,
+      "loss": 0.1687,
+      "step": 816
+    },
+    {
+      "epoch": 0.628626320845341,
+      "grad_norm": 0.8972272932978446,
+      "learning_rate": 6.747769323267706e-06,
+      "loss": 0.1746,
+      "step": 818
+    },
+    {
+      "epoch": 0.6301633045148896,
+      "grad_norm": 0.9499481842269516,
+      "learning_rate": 6.7454301076608025e-06,
+      "loss": 0.1696,
+      "step": 820
+    },
+    {
+      "epoch": 0.631700288184438,
+      "grad_norm": 0.8409255790993163,
+      "learning_rate": 6.743080504270933e-06,
+      "loss": 0.1598,
+      "step": 822
+    },
+    {
+      "epoch": 0.6332372718539866,
+      "grad_norm": 0.9027842885644961,
+      "learning_rate": 6.740720520618574e-06,
+      "loss": 0.1585,
+      "step": 824
+    },
+    {
+      "epoch": 0.6347742555235351,
+      "grad_norm": 0.8719606536944999,
+      "learning_rate": 6.738350164257421e-06,
+      "loss": 0.1547,
+      "step": 826
+    },
+    {
+      "epoch": 0.6363112391930835,
+      "grad_norm": 0.9247004952412972,
+      "learning_rate": 6.735969442774372e-06,
+      "loss": 0.173,
+      "step": 828
+    },
+    {
+      "epoch": 0.6378482228626321,
+      "grad_norm": 0.8506134884934429,
+      "learning_rate": 6.733578363789503e-06,
+      "loss": 0.154,
+      "step": 830
+    },
+    {
+      "epoch": 0.6393852065321806,
+      "grad_norm": 0.9171064482180226,
+      "learning_rate": 6.731176934956039e-06,
+      "loss": 0.1652,
+      "step": 832
+    },
+    {
+      "epoch": 0.6409221902017291,
+      "grad_norm": 0.9248200687434835,
+      "learning_rate": 6.728765163960333e-06,
+      "loss": 0.1704,
+      "step": 834
+    },
+    {
+      "epoch": 0.6424591738712776,
+      "grad_norm": 0.8545348365058903,
+      "learning_rate": 6.726343058521839e-06,
+      "loss": 0.1569,
+      "step": 836
+    },
+    {
+      "epoch": 0.6439961575408262,
+      "grad_norm": 0.9043067784462817,
+      "learning_rate": 6.723910626393091e-06,
+      "loss": 0.1715,
+      "step": 838
+    },
+    {
+      "epoch": 0.6455331412103746,
+      "grad_norm": 1.0091403589112276,
+      "learning_rate": 6.721467875359678e-06,
+      "loss": 0.1743,
+      "step": 840
+    },
+    {
+      "epoch": 0.6470701248799231,
+      "grad_norm": 0.8452475270597983,
+      "learning_rate": 6.719014813240213e-06,
+      "loss": 0.1637,
+      "step": 842
+    },
+    {
+      "epoch": 0.6486071085494717,
+      "grad_norm": 0.9000907310330241,
+      "learning_rate": 6.716551447886314e-06,
+      "loss": 0.1664,
+      "step": 844
+    },
+    {
+      "epoch": 0.6501440922190201,
+      "grad_norm": 0.9367454591962908,
+      "learning_rate": 6.714077787182576e-06,
+      "loss": 0.1609,
+      "step": 846
+    },
+    {
+      "epoch": 0.6516810758885687,
+      "grad_norm": 0.8639855347461337,
+      "learning_rate": 6.71159383904655e-06,
+      "loss": 0.1619,
+      "step": 848
+    },
+    {
+      "epoch": 0.6532180595581172,
+      "grad_norm": 0.8444570709766536,
+      "learning_rate": 6.709099611428709e-06,
+      "loss": 0.1569,
+      "step": 850
+    },
+    {
+      "epoch": 0.6547550432276658,
+      "grad_norm": 0.8660206440875692,
+      "learning_rate": 6.706595112312432e-06,
+      "loss": 0.1638,
+      "step": 852
+    },
+    {
+      "epoch": 0.6562920268972142,
+      "grad_norm": 0.9309827842770139,
+      "learning_rate": 6.704080349713974e-06,
+      "loss": 0.1605,
+      "step": 854
+    },
+    {
+      "epoch": 0.6578290105667627,
+      "grad_norm": 0.8714646054346427,
+      "learning_rate": 6.70155533168244e-06,
+      "loss": 0.1595,
+      "step": 856
+    },
+    {
+      "epoch": 0.6593659942363113,
+      "grad_norm": 0.8712554159426241,
+      "learning_rate": 6.699020066299759e-06,
+      "loss": 0.1678,
+      "step": 858
+    },
+    {
+      "epoch": 0.6609029779058597,
+      "grad_norm": 0.9250738676547778,
+      "learning_rate": 6.696474561680663e-06,
+      "loss": 0.1789,
+      "step": 860
+    },
+    {
+      "epoch": 0.6624399615754083,
+      "grad_norm": 0.9111129955962746,
+      "learning_rate": 6.693918825972651e-06,
+      "loss": 0.168,
+      "step": 862
+    },
+    {
+      "epoch": 0.6639769452449568,
+      "grad_norm": 0.9141522182958596,
+      "learning_rate": 6.691352867355973e-06,
+      "loss": 0.16,
+      "step": 864
+    },
+    {
+      "epoch": 0.6655139289145053,
+      "grad_norm": 0.881012224572559,
+      "learning_rate": 6.688776694043602e-06,
+      "loss": 0.1653,
+      "step": 866
+    },
+    {
+      "epoch": 0.6670509125840538,
+      "grad_norm": 0.8632782839785611,
+      "learning_rate": 6.6861903142812e-06,
+      "loss": 0.1564,
+      "step": 868
+    },
+    {
+      "epoch": 0.6685878962536023,
+      "grad_norm": 0.8720327769692922,
+      "learning_rate": 6.683593736347102e-06,
+      "loss": 0.1534,
+      "step": 870
+    },
+    {
+      "epoch": 0.6701248799231508,
+      "grad_norm": 0.876964430748599,
+      "learning_rate": 6.680986968552282e-06,
+      "loss": 0.1657,
+      "step": 872
+    },
+    {
+      "epoch": 0.6716618635926993,
+      "grad_norm": 0.9631837226264072,
+      "learning_rate": 6.6783700192403296e-06,
+      "loss": 0.1683,
+      "step": 874
+    },
+    {
+      "epoch": 0.6731988472622479,
+      "grad_norm": 0.846575004663177,
+      "learning_rate": 6.675742896787425e-06,
+      "loss": 0.1476,
+      "step": 876
+    },
+    {
+      "epoch": 0.6747358309317963,
+      "grad_norm": 0.9308686661082001,
+      "learning_rate": 6.6731056096023065e-06,
+      "loss": 0.1595,
+      "step": 878
+    },
+    {
+      "epoch": 0.6762728146013448,
+      "grad_norm": 0.8915035477659844,
+      "learning_rate": 6.6704581661262486e-06,
+      "loss": 0.1575,
+      "step": 880
+    },
+    {
+      "epoch": 0.6778097982708934,
+      "grad_norm": 0.8362332131345471,
+      "learning_rate": 6.6678005748330346e-06,
+      "loss": 0.1601,
+      "step": 882
+    },
+    {
+      "epoch": 0.6793467819404418,
+      "grad_norm": 0.9249603517376113,
+      "learning_rate": 6.665132844228926e-06,
+      "loss": 0.1654,
+      "step": 884
+    },
+    {
+      "epoch": 0.6808837656099904,
+      "grad_norm": 0.8814575750008693,
+      "learning_rate": 6.662454982852641e-06,
+      "loss": 0.1749,
+      "step": 886
+    },
+    {
+      "epoch": 0.6824207492795389,
+      "grad_norm": 0.899221716183914,
+      "learning_rate": 6.65976699927532e-06,
+      "loss": 0.1682,
+      "step": 888
+    },
+    {
+      "epoch": 0.6839577329490875,
+      "grad_norm": 0.9139541897412309,
+      "learning_rate": 6.657068902100504e-06,
+      "loss": 0.1662,
+      "step": 890
+    },
+    {
+      "epoch": 0.6854947166186359,
+      "grad_norm": 0.9291186479929943,
+      "learning_rate": 6.6543606999641065e-06,
+      "loss": 0.1677,
+      "step": 892
+    },
+    {
+      "epoch": 0.6870317002881844,
+      "grad_norm": 0.8803357017385457,
+      "learning_rate": 6.6516424015343795e-06,
+      "loss": 0.1528,
+      "step": 894
+    },
+    {
+      "epoch": 0.688568683957733,
+      "grad_norm": 0.9230606940030847,
+      "learning_rate": 6.6489140155118964e-06,
+      "loss": 0.1676,
+      "step": 896
+    },
+    {
+      "epoch": 0.6901056676272814,
+      "grad_norm": 0.8703699992143266,
+      "learning_rate": 6.6461755506295145e-06,
+      "loss": 0.1554,
+      "step": 898
+    },
+    {
+      "epoch": 0.69164265129683,
+      "grad_norm": 0.9524256265392508,
+      "learning_rate": 6.643427015652351e-06,
+      "loss": 0.1704,
+      "step": 900
+    },
+    {
+      "epoch": 0.6931796349663785,
+      "grad_norm": 0.9415591239679129,
+      "learning_rate": 6.640668419377758e-06,
+      "loss": 0.1592,
+      "step": 902
+    },
+    {
+      "epoch": 0.694716618635927,
+      "grad_norm": 0.8738330937457875,
+      "learning_rate": 6.6378997706352885e-06,
+      "loss": 0.1692,
+      "step": 904
+    },
+    {
+      "epoch": 0.6962536023054755,
+      "grad_norm": 0.9005686509467422,
+      "learning_rate": 6.635121078286671e-06,
+      "loss": 0.1684,
+      "step": 906
+    },
+    {
+      "epoch": 0.697790585975024,
+      "grad_norm": 0.8296366500080525,
+      "learning_rate": 6.632332351225783e-06,
+      "loss": 0.1593,
+      "step": 908
+    },
+    {
+      "epoch": 0.6993275696445725,
+      "grad_norm": 0.8623498006286683,
+      "learning_rate": 6.629533598378617e-06,
+      "loss": 0.1683,
+      "step": 910
+    },
+    {
+      "epoch": 0.700864553314121,
+      "grad_norm": 0.8462969581032185,
+      "learning_rate": 6.626724828703259e-06,
+      "loss": 0.1534,
+      "step": 912
+    },
+    {
+      "epoch": 0.7024015369836696,
+      "grad_norm": 0.8628311131127484,
+      "learning_rate": 6.623906051189854e-06,
+      "loss": 0.1604,
+      "step": 914
+    },
+    {
+      "epoch": 0.703938520653218,
+      "grad_norm": 0.9145692129791568,
+      "learning_rate": 6.621077274860581e-06,
+      "loss": 0.1594,
+      "step": 916
+    },
+    {
+      "epoch": 0.7054755043227666,
+      "grad_norm": 0.9195013649008027,
+      "learning_rate": 6.618238508769621e-06,
+      "loss": 0.1716,
+      "step": 918
+    },
+    {
+      "epoch": 0.7070124879923151,
+      "grad_norm": 0.9005634587923742,
+      "learning_rate": 6.615389762003131e-06,
+      "loss": 0.169,
+      "step": 920
+    },
+    {
+      "epoch": 0.7085494716618636,
+      "grad_norm": 0.9087616214061071,
+      "learning_rate": 6.612531043679213e-06,
+      "loss": 0.1584,
+      "step": 922
+    },
+    {
+      "epoch": 0.7100864553314121,
+      "grad_norm": 0.8929943346561857,
+      "learning_rate": 6.609662362947886e-06,
+      "loss": 0.1543,
+      "step": 924
+    },
+    {
+      "epoch": 0.7116234390009606,
+      "grad_norm": 0.8923387750870015,
+      "learning_rate": 6.606783728991054e-06,
+      "loss": 0.1576,
+      "step": 926
+    },
+    {
+      "epoch": 0.7131604226705092,
+      "grad_norm": 0.8545883642264139,
+      "learning_rate": 6.603895151022483e-06,
+      "loss": 0.1593,
+      "step": 928
+    },
+    {
+      "epoch": 0.7146974063400576,
+      "grad_norm": 0.8581885831090912,
+      "learning_rate": 6.600996638287762e-06,
+      "loss": 0.1603,
+      "step": 930
+    },
+    {
+      "epoch": 0.7162343900096062,
+      "grad_norm": 0.9326356897470247,
+      "learning_rate": 6.598088200064284e-06,
+      "loss": 0.1731,
+      "step": 932
+    },
+    {
+      "epoch": 0.7177713736791547,
+      "grad_norm": 0.8874229434905837,
+      "learning_rate": 6.595169845661204e-06,
+      "loss": 0.1561,
+      "step": 934
+    },
+    {
+      "epoch": 0.7193083573487031,
+      "grad_norm": 0.880472017082445,
+      "learning_rate": 6.592241584419424e-06,
+      "loss": 0.1589,
+      "step": 936
+    },
+    {
+      "epoch": 0.7208453410182517,
+      "grad_norm": 0.8802659879981632,
+      "learning_rate": 6.589303425711548e-06,
+      "loss": 0.1699,
+      "step": 938
+    },
+    {
+      "epoch": 0.7223823246878002,
+      "grad_norm": 0.8605735232186383,
+      "learning_rate": 6.586355378941866e-06,
+      "loss": 0.1508,
+      "step": 940
+    },
+    {
+      "epoch": 0.7239193083573487,
+      "grad_norm": 0.9218559892018848,
+      "learning_rate": 6.58339745354631e-06,
+      "loss": 0.1536,
+      "step": 942
+    },
+    {
+      "epoch": 0.7254562920268972,
+      "grad_norm": 0.8848127640286009,
+      "learning_rate": 6.580429658992438e-06,
+      "loss": 0.1595,
+      "step": 944
+    },
+    {
+      "epoch": 0.7269932756964457,
+      "grad_norm": 0.9210041621285134,
+      "learning_rate": 6.577452004779393e-06,
+      "loss": 0.1493,
+      "step": 946
+    },
+    {
+      "epoch": 0.7285302593659942,
+      "grad_norm": 0.9094329668162617,
+      "learning_rate": 6.574464500437875e-06,
+      "loss": 0.1656,
+      "step": 948
+    },
+    {
+      "epoch": 0.7300672430355427,
+      "grad_norm": 0.8599101226794522,
+      "learning_rate": 6.571467155530114e-06,
+      "loss": 0.1621,
+      "step": 950
+    },
+    {
+      "epoch": 0.7316042267050913,
+      "grad_norm": 0.8800370037884307,
+      "learning_rate": 6.568459979649836e-06,
+      "loss": 0.1613,
+      "step": 952
+    },
+    {
+      "epoch": 0.7331412103746398,
+      "grad_norm": 0.8133202498369466,
+      "learning_rate": 6.565442982422233e-06,
+      "loss": 0.1498,
+      "step": 954
+    },
+    {
+      "epoch": 0.7346781940441883,
+      "grad_norm": 0.8854492713891652,
+      "learning_rate": 6.5624161735039365e-06,
+      "loss": 0.1576,
+      "step": 956
+    },
+    {
+      "epoch": 0.7362151777137368,
+      "grad_norm": 0.8860066937387789,
+      "learning_rate": 6.559379562582976e-06,
+      "loss": 0.1567,
+      "step": 958
+    },
+    {
+      "epoch": 0.7377521613832853,
+      "grad_norm": 0.801524450434012,
+      "learning_rate": 6.556333159378761e-06,
+      "loss": 0.1507,
+      "step": 960
+    },
+    {
+      "epoch": 0.7392891450528338,
+      "grad_norm": 0.826356057888294,
+      "learning_rate": 6.553276973642037e-06,
+      "loss": 0.1693,
+      "step": 962
+    },
+    {
+      "epoch": 0.7408261287223823,
+      "grad_norm": 0.8294938538531003,
+      "learning_rate": 6.550211015154869e-06,
+      "loss": 0.1558,
+      "step": 964
+    },
+    {
+      "epoch": 0.7423631123919309,
+      "grad_norm": 0.8307597065545927,
+      "learning_rate": 6.547135293730595e-06,
+      "loss": 0.1533,
+      "step": 966
+    },
+    {
+      "epoch": 0.7439000960614793,
+      "grad_norm": 0.8014991418925953,
+      "learning_rate": 6.544049819213806e-06,
+      "loss": 0.1634,
+      "step": 968
+    },
+    {
+      "epoch": 0.7454370797310279,
+      "grad_norm": 0.8998803118527309,
+      "learning_rate": 6.540954601480307e-06,
+      "loss": 0.1678,
+      "step": 970
+    },
+    {
+      "epoch": 0.7469740634005764,
+      "grad_norm": 0.8815150629442536,
+      "learning_rate": 6.537849650437091e-06,
+      "loss": 0.1549,
+      "step": 972
+    },
+    {
+      "epoch": 0.7485110470701248,
+      "grad_norm": 0.825376179394833,
+      "learning_rate": 6.534734976022302e-06,
+      "loss": 0.153,
+      "step": 974
+    },
+    {
+      "epoch": 0.7500480307396734,
+      "grad_norm": 0.8179612985435794,
+      "learning_rate": 6.53161058820521e-06,
+      "loss": 0.1444,
+      "step": 976
+    },
+    {
+      "epoch": 0.7515850144092219,
+      "grad_norm": 0.8727593413193312,
+      "learning_rate": 6.528476496986172e-06,
+      "loss": 0.1517,
+      "step": 978
+    },
+    {
+      "epoch": 0.7531219980787704,
+      "grad_norm": 0.8512575927431834,
+      "learning_rate": 6.525332712396604e-06,
+      "loss": 0.1573,
+      "step": 980
+    },
+    {
+      "epoch": 0.7546589817483189,
+      "grad_norm": 0.8582941888601493,
+      "learning_rate": 6.522179244498948e-06,
+      "loss": 0.1541,
+      "step": 982
+    },
+    {
+      "epoch": 0.7561959654178675,
+      "grad_norm": 0.8767811552173622,
+      "learning_rate": 6.519016103386639e-06,
+      "loss": 0.1546,
+      "step": 984
+    },
+    {
+      "epoch": 0.757732949087416,
+      "grad_norm": 0.8210784117243145,
+      "learning_rate": 6.5158432991840755e-06,
+      "loss": 0.145,
+      "step": 986
+    },
+    {
+      "epoch": 0.7592699327569644,
+      "grad_norm": 0.8326696940302717,
+      "learning_rate": 6.512660842046582e-06,
+      "loss": 0.1597,
+      "step": 988
+    },
+    {
+      "epoch": 0.760806916426513,
+      "grad_norm": 0.953434296685044,
+      "learning_rate": 6.509468742160382e-06,
+      "loss": 0.153,
+      "step": 990
+    },
+    {
+      "epoch": 0.7623439000960615,
+      "grad_norm": 0.8362202034275996,
+      "learning_rate": 6.506267009742564e-06,
+      "loss": 0.1422,
+      "step": 992
+    },
+    {
+      "epoch": 0.76388088376561,
+      "grad_norm": 0.857361945741671,
+      "learning_rate": 6.503055655041042e-06,
+      "loss": 0.1735,
+      "step": 994
+    },
+    {
+      "epoch": 0.7654178674351585,
+      "grad_norm": 0.891793041703015,
+      "learning_rate": 6.499834688334537e-06,
+      "loss": 0.1555,
+      "step": 996
+    },
+    {
+      "epoch": 0.766954851104707,
+      "grad_norm": 0.9034407482465361,
+      "learning_rate": 6.496604119932528e-06,
+      "loss": 0.1692,
+      "step": 998
+    },
+    {
+      "epoch": 0.7684918347742555,
+      "grad_norm": 0.9404179935204194,
+      "learning_rate": 6.493363960175231e-06,
+      "loss": 0.1549,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7684918347742555,
+      "eval_loss": 0.1432075798511505,
+      "eval_runtime": 362.8754,
+      "eval_samples_per_second": 50.995,
+      "eval_steps_per_second": 6.377,
+      "step": 1000
+    },
+    {
+      "epoch": 0.770028818443804,
+      "grad_norm": 0.8281134972466698,
+      "learning_rate": 6.490114219433558e-06,
+      "loss": 0.1476,
+      "step": 1002
+    },
+    {
+      "epoch": 0.7715658021133526,
+      "grad_norm": 0.8720452925619442,
+      "learning_rate": 6.486854908109089e-06,
+      "loss": 0.1433,
+      "step": 1004
+    },
+    {
+      "epoch": 0.773102785782901,
+      "grad_norm": 0.8720479928054387,
+      "learning_rate": 6.483586036634041e-06,
+      "loss": 0.1553,
+      "step": 1006
+    },
+    {
+      "epoch": 0.7746397694524496,
+      "grad_norm": 0.8412137015802439,
+      "learning_rate": 6.480307615471223e-06,
+      "loss": 0.157,
+      "step": 1008
+    },
+    {
+      "epoch": 0.7761767531219981,
+      "grad_norm": 0.8527177593745339,
+      "learning_rate": 6.4770196551140155e-06,
+      "loss": 0.1614,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7777137367915465,
+      "grad_norm": 0.8821801330028585,
+      "learning_rate": 6.473722166086329e-06,
+      "loss": 0.1525,
+      "step": 1012
+    },
+    {
+      "epoch": 0.7792507204610951,
+      "grad_norm": 0.9018994294254266,
+      "learning_rate": 6.470415158942574e-06,
+      "loss": 0.16,
+      "step": 1014
+    },
+    {
+      "epoch": 0.7807877041306436,
+      "grad_norm": 0.8703842227810923,
+      "learning_rate": 6.467098644267625e-06,
+      "loss": 0.1585,
+      "step": 1016
+    },
+    {
+      "epoch": 0.7823246878001922,
+      "grad_norm": 0.8051081010409243,
+      "learning_rate": 6.46377263267679e-06,
+      "loss": 0.1527,
+      "step": 1018
+    },
+    {
+      "epoch": 0.7838616714697406,
+      "grad_norm": 0.8888259620906063,
+      "learning_rate": 6.460437134815771e-06,
+      "loss": 0.167,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7853986551392892,
+      "grad_norm": 0.9029947228741374,
+      "learning_rate": 6.457092161360633e-06,
+      "loss": 0.1668,
+      "step": 1022
+    },
+    {
+      "epoch": 0.7869356388088377,
+      "grad_norm": 0.8981361575693058,
+      "learning_rate": 6.453737723017775e-06,
+      "loss": 0.1612,
+      "step": 1024
+    },
+    {
+      "epoch": 0.7884726224783861,
+      "grad_norm": 0.9352139489971404,
+      "learning_rate": 6.450373830523886e-06,
+      "loss": 0.1553,
+      "step": 1026
+    },
+    {
+      "epoch": 0.7900096061479347,
+      "grad_norm": 0.8227972574038364,
+      "learning_rate": 6.447000494645916e-06,
+      "loss": 0.1578,
+      "step": 1028
+    },
+    {
+      "epoch": 0.7915465898174832,
+      "grad_norm": 0.8530604144128302,
+      "learning_rate": 6.4436177261810395e-06,
+      "loss": 0.1541,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7930835734870317,
+      "grad_norm": 0.8897585617894813,
+      "learning_rate": 6.440225535956627e-06,
+      "loss": 0.1552,
+      "step": 1032
+    },
+    {
+      "epoch": 0.7946205571565802,
+      "grad_norm": 0.8139899114750058,
+      "learning_rate": 6.436823934830201e-06,
+      "loss": 0.1397,
+      "step": 1034
+    },
+    {
+      "epoch": 0.7961575408261288,
+      "grad_norm": 0.9155078950589843,
+      "learning_rate": 6.433412933689408e-06,
+      "loss": 0.1638,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7976945244956772,
+      "grad_norm": 0.7986982196116793,
+      "learning_rate": 6.429992543451982e-06,
+      "loss": 0.1492,
+      "step": 1038
+    },
+    {
+      "epoch": 0.7992315081652257,
+      "grad_norm": 0.9067813768385491,
+      "learning_rate": 6.426562775065706e-06,
+      "loss": 0.166,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8007684918347743,
+      "grad_norm": 0.8546237308090272,
+      "learning_rate": 6.4231236395083835e-06,
+      "loss": 0.1528,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8023054755043227,
+      "grad_norm": 0.8301245146373359,
+      "learning_rate": 6.419675147787799e-06,
+      "loss": 0.169,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8038424591738713,
+      "grad_norm": 0.9357811284934665,
+      "learning_rate": 6.416217310941682e-06,
+      "loss": 0.1567,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8053794428434198,
+      "grad_norm": 0.8763494765723041,
+      "learning_rate": 6.412750140037675e-06,
+      "loss": 0.1578,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8069164265129684,
+      "grad_norm": 0.8853921013266428,
+      "learning_rate": 6.409273646173296e-06,
+      "loss": 0.1753,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8084534101825168,
+      "grad_norm": 0.8836427743865589,
+      "learning_rate": 6.405787840475904e-06,
+      "loss": 0.1599,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8099903938520653,
+      "grad_norm": 0.8292826766651644,
+      "learning_rate": 6.402292734102661e-06,
+      "loss": 0.1603,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8115273775216139,
+      "grad_norm": 0.7981150884405339,
+      "learning_rate": 6.3987883382405e-06,
+      "loss": 0.152,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8130643611911623,
+      "grad_norm": 0.8941084452594006,
+      "learning_rate": 6.395274664106086e-06,
+      "loss": 0.1675,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8146013448607109,
+      "grad_norm": 0.8843029348683886,
+      "learning_rate": 6.39175172294578e-06,
+      "loss": 0.1647,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8161383285302594,
+      "grad_norm": 0.8607622983535753,
+      "learning_rate": 6.38821952603561e-06,
+      "loss": 0.1642,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8176753121998078,
+      "grad_norm": 0.8712768415639528,
+      "learning_rate": 6.3846780846812234e-06,
+      "loss": 0.165,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8192122958693564,
+      "grad_norm": 0.839138399301467,
+      "learning_rate": 6.381127410217858e-06,
+      "loss": 0.1679,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8207492795389049,
+      "grad_norm": 0.8561911550355878,
+      "learning_rate": 6.377567514010304e-06,
+      "loss": 0.1582,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8222862632084534,
+      "grad_norm": 0.8290678486131811,
+      "learning_rate": 6.373998407452873e-06,
+      "loss": 0.1545,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8238232468780019,
+      "grad_norm": 0.847760798773364,
+      "learning_rate": 6.370420101969349e-06,
+      "loss": 0.1459,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8253602305475505,
+      "grad_norm": 0.8321417732926941,
+      "learning_rate": 6.3668326090129645e-06,
+      "loss": 0.1524,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8268972142170989,
+      "grad_norm": 0.8458379024110787,
+      "learning_rate": 6.363235940066358e-06,
+      "loss": 0.1485,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8284341978866474,
+      "grad_norm": 0.9374727516067862,
+      "learning_rate": 6.359630106641535e-06,
+      "loss": 0.1641,
+      "step": 1078
+    },
+    {
+      "epoch": 0.829971181556196,
+      "grad_norm": 0.8470538370988849,
+      "learning_rate": 6.356015120279837e-06,
+      "loss": 0.1571,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8315081652257444,
+      "grad_norm": 0.8292449900966895,
+      "learning_rate": 6.352390992551903e-06,
+      "loss": 0.1668,
+      "step": 1082
+    },
+    {
+      "epoch": 0.833045148895293,
+      "grad_norm": 0.8873775326690251,
+      "learning_rate": 6.348757735057628e-06,
+      "loss": 0.1573,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8345821325648415,
+      "grad_norm": 0.8792119220842931,
+      "learning_rate": 6.345115359426129e-06,
+      "loss": 0.1578,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8361191162343901,
+      "grad_norm": 0.9098125786437564,
+      "learning_rate": 6.341463877315711e-06,
+      "loss": 0.1631,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8376560999039385,
+      "grad_norm": 0.801489722135179,
+      "learning_rate": 6.337803300413822e-06,
+      "loss": 0.1479,
+      "step": 1090
+    },
+    {
+      "epoch": 0.839193083573487,
+      "grad_norm": 0.9244844513428183,
+      "learning_rate": 6.334133640437025e-06,
+      "loss": 0.1632,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8407300672430356,
+      "grad_norm": 0.881534980036438,
+      "learning_rate": 6.330454909130952e-06,
+      "loss": 0.1595,
+      "step": 1094
+    },
+    {
+      "epoch": 0.842267050912584,
+      "grad_norm": 0.8598183603447899,
+      "learning_rate": 6.326767118270271e-06,
+      "loss": 0.1616,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8438040345821326,
+      "grad_norm": 0.889022724838577,
+      "learning_rate": 6.323070279658648e-06,
+      "loss": 0.1601,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8453410182516811,
+      "grad_norm": 0.8912527206906096,
+      "learning_rate": 6.319364405128706e-06,
+      "loss": 0.1605,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8468780019212296,
+      "grad_norm": 0.8035411272312927,
+      "learning_rate": 6.315649506541995e-06,
+      "loss": 0.1514,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8484149855907781,
+      "grad_norm": 0.866003931969349,
+      "learning_rate": 6.311925595788942e-06,
+      "loss": 0.1535,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8499519692603266,
+      "grad_norm": 0.8292860317473079,
+      "learning_rate": 6.308192684788825e-06,
+      "loss": 0.1517,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8514889529298751,
+      "grad_norm": 0.8869266953820677,
+      "learning_rate": 6.3044507854897265e-06,
+      "loss": 0.1603,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8530259365994236,
+      "grad_norm": 0.8577246546870172,
+      "learning_rate": 6.3006999098684985e-06,
+      "loss": 0.157,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8545629202689722,
+      "grad_norm": 0.8403234008489573,
+      "learning_rate": 6.296940069930725e-06,
+      "loss": 0.1592,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8560999039385206,
+      "grad_norm": 0.8702914831113494,
+      "learning_rate": 6.293171277710682e-06,
+      "loss": 0.1575,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8576368876080691,
+      "grad_norm": 0.8870545228720214,
+      "learning_rate": 6.289393545271299e-06,
+      "loss": 0.163,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8591738712776177,
+      "grad_norm": 0.8835871253233961,
+      "learning_rate": 6.285606884704122e-06,
+      "loss": 0.1505,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8607108549471661,
+      "grad_norm": 0.9222189898206439,
+      "learning_rate": 6.281811308129271e-06,
+      "loss": 0.1565,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8622478386167147,
+      "grad_norm": 0.865733104634757,
+      "learning_rate": 6.278006827695407e-06,
+      "loss": 0.1657,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8637848222862632,
+      "grad_norm": 0.7873625454059449,
+      "learning_rate": 6.274193455579688e-06,
+      "loss": 0.152,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8653218059558118,
+      "grad_norm": 0.9036534746354465,
+      "learning_rate": 6.270371203987733e-06,
+      "loss": 0.1524,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8668587896253602,
+      "grad_norm": 0.8874043561419231,
+      "learning_rate": 6.266540085153581e-06,
+      "loss": 0.1623,
+      "step": 1128
+    },
+    {
+      "epoch": 0.8683957732949087,
+      "grad_norm": 0.8520395043888209,
+      "learning_rate": 6.262700111339654e-06,
+      "loss": 0.1511,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8699327569644573,
+      "grad_norm": 0.814438124524513,
+      "learning_rate": 6.2588512948367144e-06,
+      "loss": 0.143,
+      "step": 1132
+    },
+    {
+      "epoch": 0.8714697406340057,
+      "grad_norm": 0.8620767282571578,
+      "learning_rate": 6.254993647963831e-06,
+      "loss": 0.1616,
+      "step": 1134
+    },
+    {
+      "epoch": 0.8730067243035543,
+      "grad_norm": 0.892703951825046,
+      "learning_rate": 6.251127183068331e-06,
+      "loss": 0.1579,
+      "step": 1136
+    },
+    {
+      "epoch": 0.8745437079731028,
+      "grad_norm": 0.8623333160707259,
+      "learning_rate": 6.247251912525773e-06,
+      "loss": 0.1576,
+      "step": 1138
+    },
+    {
+      "epoch": 0.8760806916426513,
+      "grad_norm": 0.8199322821651265,
+      "learning_rate": 6.243367848739894e-06,
+      "loss": 0.153,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8776176753121998,
+      "grad_norm": 0.8392074331450308,
+      "learning_rate": 6.23947500414258e-06,
+      "loss": 0.1606,
+      "step": 1142
+    },
+    {
+      "epoch": 0.8791546589817483,
+      "grad_norm": 0.8281996951649421,
+      "learning_rate": 6.235573391193819e-06,
+      "loss": 0.1469,
+      "step": 1144
+    },
+    {
+      "epoch": 0.8806916426512968,
+      "grad_norm": 0.8544291264871913,
+      "learning_rate": 6.231663022381666e-06,
+      "loss": 0.17,
+      "step": 1146
+    },
+    {
+      "epoch": 0.8822286263208453,
+      "grad_norm": 0.8699715620381219,
+      "learning_rate": 6.227743910222202e-06,
+      "loss": 0.1543,
+      "step": 1148
+    },
+    {
+      "epoch": 0.8837656099903939,
+      "grad_norm": 0.8096799920711654,
+      "learning_rate": 6.22381606725949e-06,
+      "loss": 0.1562,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8853025936599423,
+      "grad_norm": 0.8090700874638417,
+      "learning_rate": 6.219879506065542e-06,
+      "loss": 0.1499,
+      "step": 1152
+    },
+    {
+      "epoch": 0.8868395773294909,
+      "grad_norm": 0.7867446095253069,
+      "learning_rate": 6.215934239240272e-06,
+      "loss": 0.1558,
+      "step": 1154
+    },
+    {
+      "epoch": 0.8883765609990394,
+      "grad_norm": 0.788584493474745,
+      "learning_rate": 6.211980279411459e-06,
+      "loss": 0.1433,
+      "step": 1156
+    },
+    {
+      "epoch": 0.8899135446685879,
+      "grad_norm": 0.8362819878588067,
+      "learning_rate": 6.208017639234708e-06,
+      "loss": 0.1552,
+      "step": 1158
+    },
+    {
+      "epoch": 0.8914505283381364,
+      "grad_norm": 0.8817509955093219,
+      "learning_rate": 6.204046331393405e-06,
+      "loss": 0.1499,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8929875120076849,
+      "grad_norm": 0.8313021295240178,
+      "learning_rate": 6.20006636859868e-06,
+      "loss": 0.1671,
+      "step": 1162
+    },
+    {
+      "epoch": 0.8945244956772335,
+      "grad_norm": 0.8485741130804667,
+      "learning_rate": 6.196077763589365e-06,
+      "loss": 0.1609,
+      "step": 1164
+    },
+    {
+      "epoch": 0.8960614793467819,
+      "grad_norm": 0.861196387787182,
+      "learning_rate": 6.192080529131955e-06,
+      "loss": 0.1609,
+      "step": 1166
+    },
+    {
+      "epoch": 0.8975984630163305,
+      "grad_norm": 0.8527772834985116,
+      "learning_rate": 6.188074678020563e-06,
+      "loss": 0.1571,
+      "step": 1168
+    },
+    {
+      "epoch": 0.899135446685879,
+      "grad_norm": 0.820498232150602,
+      "learning_rate": 6.184060223076884e-06,
+      "loss": 0.1599,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9006724303554274,
+      "grad_norm": 0.8519749575925564,
+      "learning_rate": 6.180037177150149e-06,
+      "loss": 0.1548,
+      "step": 1172
+    },
+    {
+      "epoch": 0.902209414024976,
+      "grad_norm": 0.8442399564211222,
+      "learning_rate": 6.176005553117091e-06,
+      "loss": 0.151,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9037463976945245,
+      "grad_norm": 0.8937597645248366,
+      "learning_rate": 6.171965363881894e-06,
+      "loss": 0.1682,
+      "step": 1176
+    },
+    {
+      "epoch": 0.905283381364073,
+      "grad_norm": 0.9004779698121571,
+      "learning_rate": 6.167916622376161e-06,
+      "loss": 0.1648,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9068203650336215,
+      "grad_norm": 0.8739876161966016,
+      "learning_rate": 6.163859341558867e-06,
+      "loss": 0.1466,
+      "step": 1180
+    },
+    {
+      "epoch": 0.90835734870317,
+      "grad_norm": 0.8359328154212177,
+      "learning_rate": 6.159793534416318e-06,
+      "loss": 0.1609,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9098943323727186,
+      "grad_norm": 0.7772963511251748,
+      "learning_rate": 6.155719213962113e-06,
+      "loss": 0.1396,
+      "step": 1184
+    },
+    {
+      "epoch": 0.911431316042267,
+      "grad_norm": 0.8267806690490346,
+      "learning_rate": 6.151636393237099e-06,
+      "loss": 0.1557,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9129682997118156,
+      "grad_norm": 0.8299194507951076,
+      "learning_rate": 6.147545085309329e-06,
+      "loss": 0.1535,
+      "step": 1188
+    },
+    {
+      "epoch": 0.914505283381364,
+      "grad_norm": 0.8107731050092446,
+      "learning_rate": 6.143445303274022e-06,
+      "loss": 0.1533,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9160422670509126,
+      "grad_norm": 0.8664903101940159,
+      "learning_rate": 6.139337060253521e-06,
+      "loss": 0.1509,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9175792507204611,
+      "grad_norm": 0.9096007236323014,
+      "learning_rate": 6.135220369397252e-06,
+      "loss": 0.1576,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9191162343900096,
+      "grad_norm": 0.7794208493232124,
+      "learning_rate": 6.131095243881675e-06,
+      "loss": 0.1353,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9206532180595581,
+      "grad_norm": 0.8177662161991649,
+      "learning_rate": 6.1269616969102546e-06,
+      "loss": 0.1409,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9221902017291066,
+      "grad_norm": 0.7668271695157493,
+      "learning_rate": 6.122819741713402e-06,
+      "loss": 0.1442,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9237271853986552,
+      "grad_norm": 0.7607151071645437,
+      "learning_rate": 6.118669391548449e-06,
+      "loss": 0.1497,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9252641690682036,
+      "grad_norm": 0.8748866630574149,
+      "learning_rate": 6.114510659699591e-06,
+      "loss": 0.1611,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9268011527377522,
+      "grad_norm": 0.8598578438295764,
+      "learning_rate": 6.110343559477855e-06,
+      "loss": 0.1542,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9283381364073007,
+      "grad_norm": 0.8347211442345176,
+      "learning_rate": 6.106168104221052e-06,
+      "loss": 0.1488,
+      "step": 1208
+    },
+    {
+      "epoch": 0.9298751200768491,
+      "grad_norm": 0.8798272771765616,
+      "learning_rate": 6.101984307293735e-06,
+      "loss": 0.1528,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9314121037463977,
+      "grad_norm": 0.809406948238894,
+      "learning_rate": 6.097792182087156e-06,
+      "loss": 0.1605,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9329490874159462,
+      "grad_norm": 0.8465513093584821,
+      "learning_rate": 6.093591742019225e-06,
+      "loss": 0.1515,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9344860710854948,
+      "grad_norm": 0.8717872500128261,
+      "learning_rate": 6.089383000534465e-06,
+      "loss": 0.1537,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9360230547550432,
+      "grad_norm": 0.7761882043003053,
+      "learning_rate": 6.085165971103969e-06,
+      "loss": 0.137,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9375600384245918,
+      "grad_norm": 0.8779867026896587,
+      "learning_rate": 6.080940667225359e-06,
+      "loss": 0.1477,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9390970220941403,
+      "grad_norm": 0.8339829528733452,
+      "learning_rate": 6.076707102422741e-06,
+      "loss": 0.1511,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9406340057636887,
+      "grad_norm": 0.8459206215885985,
+      "learning_rate": 6.072465290246659e-06,
+      "loss": 0.1429,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9421709894332373,
+      "grad_norm": 0.8156133417127609,
+      "learning_rate": 6.068215244274061e-06,
+      "loss": 0.1545,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9437079731027858,
+      "grad_norm": 0.8160966643134648,
+      "learning_rate": 6.063956978108245e-06,
+      "loss": 0.1427,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9452449567723343,
+      "grad_norm": 0.8239874532076136,
+      "learning_rate": 6.059690505378819e-06,
+      "loss": 0.1534,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9467819404418828,
+      "grad_norm": 0.8475441460391161,
+      "learning_rate": 6.0554158397416596e-06,
+      "loss": 0.1507,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9483189241114313,
+      "grad_norm": 0.839569089678522,
+      "learning_rate": 6.051132994878868e-06,
+      "loss": 0.1527,
+      "step": 1234
+    },
+    {
+      "epoch": 0.9498559077809798,
+      "grad_norm": 0.7802196203437861,
+      "learning_rate": 6.046841984498722e-06,
+      "loss": 0.1376,
+      "step": 1236
+    },
+    {
+      "epoch": 0.9513928914505283,
+      "grad_norm": 0.8574149706930279,
+      "learning_rate": 6.042542822335638e-06,
+      "loss": 0.1637,
+      "step": 1238
+    },
+    {
+      "epoch": 0.9529298751200769,
+      "grad_norm": 0.8161418800215421,
+      "learning_rate": 6.038235522150124e-06,
+      "loss": 0.1479,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9544668587896253,
+      "grad_norm": 0.882673614104903,
+      "learning_rate": 6.0339200977287316e-06,
+      "loss": 0.1414,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9560038424591739,
+      "grad_norm": 0.8368270452723762,
+      "learning_rate": 6.029596562884021e-06,
+      "loss": 0.158,
+      "step": 1244
+    },
+    {
+      "epoch": 0.9575408261287224,
+      "grad_norm": 0.8428259430927917,
+      "learning_rate": 6.025264931454509e-06,
+      "loss": 0.1474,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9590778097982708,
+      "grad_norm": 0.8478092566057043,
+      "learning_rate": 6.020925217304629e-06,
+      "loss": 0.1526,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9606147934678194,
+      "grad_norm": 0.8333839896657882,
+      "learning_rate": 6.016577434324684e-06,
+      "loss": 0.1419,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9621517771373679,
+      "grad_norm": 0.8445286054177732,
+      "learning_rate": 6.012221596430804e-06,
+      "loss": 0.1596,
+      "step": 1252
+    },
+    {
+      "epoch": 0.9636887608069165,
+      "grad_norm": 0.851762036838009,
+      "learning_rate": 6.0078577175649e-06,
+      "loss": 0.1546,
+      "step": 1254
+    },
+    {
+      "epoch": 0.9652257444764649,
+      "grad_norm": 0.8437848350814667,
+      "learning_rate": 6.00348581169462e-06,
+      "loss": 0.1508,
+      "step": 1256
+    },
+    {
+      "epoch": 0.9667627281460135,
+      "grad_norm": 0.8945339085483407,
+      "learning_rate": 5.9991058928133054e-06,
+      "loss": 0.1576,
+      "step": 1258
+    },
+    {
+      "epoch": 0.968299711815562,
+      "grad_norm": 0.7944819259983578,
+      "learning_rate": 5.994717974939944e-06,
+      "loss": 0.1476,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9698366954851104,
+      "grad_norm": 0.8729176506312757,
+      "learning_rate": 5.990322072119126e-06,
+      "loss": 0.1596,
+      "step": 1262
+    },
+    {
+      "epoch": 0.971373679154659,
+      "grad_norm": 0.8725237140356162,
+      "learning_rate": 5.985918198421002e-06,
+      "loss": 0.1571,
+      "step": 1264
+    },
+    {
+      "epoch": 0.9729106628242075,
+      "grad_norm": 0.7991775531475158,
+      "learning_rate": 5.981506367941233e-06,
+      "loss": 0.1434,
+      "step": 1266
+    },
+    {
+      "epoch": 0.974447646493756,
+      "grad_norm": 0.8382327345169207,
+      "learning_rate": 5.977086594800948e-06,
+      "loss": 0.1596,
+      "step": 1268
+    },
+    {
+      "epoch": 0.9759846301633045,
+      "grad_norm": 0.9529464464579128,
+      "learning_rate": 5.972658893146697e-06,
+      "loss": 0.1514,
+      "step": 1270
+    },
+    {
+      "epoch": 0.9775216138328531,
+      "grad_norm": 0.8484469791654419,
+      "learning_rate": 5.96822327715041e-06,
+      "loss": 0.152,
+      "step": 1272
+    },
+    {
+      "epoch": 0.9790585975024015,
+      "grad_norm": 0.81835031937336,
+      "learning_rate": 5.963779761009348e-06,
+      "loss": 0.1528,
+      "step": 1274
+    },
+    {
+      "epoch": 0.98059558117195,
+      "grad_norm": 0.8310249942970027,
+      "learning_rate": 5.959328358946056e-06,
+      "loss": 0.1473,
+      "step": 1276
+    },
+    {
+      "epoch": 0.9821325648414986,
+      "grad_norm": 0.8856986735480396,
+      "learning_rate": 5.954869085208323e-06,
+      "loss": 0.1609,
+      "step": 1278
+    },
+    {
+      "epoch": 0.983669548511047,
+      "grad_norm": 0.8166362545867224,
+      "learning_rate": 5.9504019540691305e-06,
+      "loss": 0.1465,
+      "step": 1280
+    },
+    {
+      "epoch": 0.9852065321805956,
+      "grad_norm": 0.8533786504203714,
+      "learning_rate": 5.945926979826612e-06,
+      "loss": 0.1555,
+      "step": 1282
+    },
+    {
+      "epoch": 0.9867435158501441,
+      "grad_norm": 0.8994585946450427,
+      "learning_rate": 5.941444176804002e-06,
+      "loss": 0.1495,
+      "step": 1284
+    },
+    {
+      "epoch": 0.9882804995196927,
+      "grad_norm": 0.8108578942364258,
+      "learning_rate": 5.936953559349596e-06,
+      "loss": 0.1505,
+      "step": 1286
+    },
+    {
+      "epoch": 0.9898174831892411,
+      "grad_norm": 0.8282612807656491,
+      "learning_rate": 5.932455141836697e-06,
+      "loss": 0.1561,
+      "step": 1288
+    },
+    {
+      "epoch": 0.9913544668587896,
+      "grad_norm": 0.8532043323116557,
+      "learning_rate": 5.927948938663581e-06,
+      "loss": 0.1559,
+      "step": 1290
+    },
+    {
+      "epoch": 0.9928914505283382,
+      "grad_norm": 0.8582728286463194,
+      "learning_rate": 5.923434964253437e-06,
+      "loss": 0.1628,
+      "step": 1292
+    },
+    {
+      "epoch": 0.9944284341978866,
+      "grad_norm": 0.8145713992914184,
+      "learning_rate": 5.91891323305433e-06,
+      "loss": 0.1503,
+      "step": 1294
+    },
+    {
+      "epoch": 0.9959654178674352,
+      "grad_norm": 0.8011507540671976,
+      "learning_rate": 5.914383759539153e-06,
+      "loss": 0.1439,
+      "step": 1296
+    },
+    {
+      "epoch": 0.9975024015369837,
+      "grad_norm": 0.9045409623476682,
+      "learning_rate": 5.909846558205582e-06,
+      "loss": 0.1533,
+      "step": 1298
+    },
+    {
+      "epoch": 0.9990393852065321,
+      "grad_norm": 0.8994142078460239,
+      "learning_rate": 5.905301643576025e-06,
+      "loss": 0.1635,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0005763688760807,
+      "grad_norm": 1.9268010896112178,
+      "learning_rate": 5.900749030197578e-06,
+      "loss": 0.2109,
+      "step": 1302
+    },
+    {
+      "epoch": 1.0021133525456292,
+      "grad_norm": 0.7330614000908158,
+      "learning_rate": 5.8961887326419804e-06,
+      "loss": 0.0988,
+      "step": 1304
+    },
+    {
+      "epoch": 1.0036503362151776,
+      "grad_norm": 0.7392181978150959,
+      "learning_rate": 5.891620765505566e-06,
+      "loss": 0.1003,
+      "step": 1306
+    },
+    {
+      "epoch": 1.0051873198847263,
+      "grad_norm": 0.7663128310167143,
+      "learning_rate": 5.887045143409216e-06,
+      "loss": 0.1016,
+      "step": 1308
+    },
+    {
+      "epoch": 1.0067243035542748,
+      "grad_norm": 0.7607110644159815,
+      "learning_rate": 5.882461880998317e-06,
+      "loss": 0.0916,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0082612872238232,
+      "grad_norm": 0.8238052235863006,
+      "learning_rate": 5.877870992942704e-06,
+      "loss": 0.0931,
+      "step": 1312
+    },
+    {
+      "epoch": 1.0097982708933717,
+      "grad_norm": 0.8892662986360846,
+      "learning_rate": 5.873272493936625e-06,
+      "loss": 0.1033,
+      "step": 1314
+    },
+    {
+      "epoch": 1.0113352545629202,
+      "grad_norm": 0.7704764998566338,
+      "learning_rate": 5.868666398698687e-06,
+      "loss": 0.1002,
+      "step": 1316
+    },
+    {
+      "epoch": 1.0128722382324689,
+      "grad_norm": 0.7709377348356039,
+      "learning_rate": 5.864052721971809e-06,
+      "loss": 0.0967,
+      "step": 1318
+    },
+    {
+      "epoch": 1.0144092219020173,
+      "grad_norm": 0.7969803446335071,
+      "learning_rate": 5.859431478523179e-06,
+      "loss": 0.1048,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0159462055715658,
+      "grad_norm": 0.6999689392816087,
+      "learning_rate": 5.854802683144201e-06,
+      "loss": 0.088,
+      "step": 1322
+    },
+    {
+      "epoch": 1.0174831892411142,
+      "grad_norm": 0.7760114509424976,
+      "learning_rate": 5.850166350650456e-06,
+      "loss": 0.0991,
+      "step": 1324
+    },
+    {
+      "epoch": 1.019020172910663,
+      "grad_norm": 0.7880085543010041,
+      "learning_rate": 5.845522495881642e-06,
+      "loss": 0.0971,
+      "step": 1326
+    },
+    {
+      "epoch": 1.0205571565802114,
+      "grad_norm": 0.8289396285343062,
+      "learning_rate": 5.840871133701542e-06,
+      "loss": 0.102,
+      "step": 1328
+    },
+    {
+      "epoch": 1.0220941402497599,
+      "grad_norm": 0.7142434734819356,
+      "learning_rate": 5.836212278997961e-06,
+      "loss": 0.0883,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0236311239193083,
+      "grad_norm": 0.7250033576245166,
+      "learning_rate": 5.8315459466826895e-06,
+      "loss": 0.0915,
+      "step": 1332
+    },
+    {
+      "epoch": 1.0251681075888568,
+      "grad_norm": 0.7570476788775075,
+      "learning_rate": 5.826872151691452e-06,
+      "loss": 0.0868,
+      "step": 1334
+    },
+    {
+      "epoch": 1.0267050912584055,
+      "grad_norm": 0.7965255668625996,
+      "learning_rate": 5.822190908983859e-06,
+      "loss": 0.0883,
+      "step": 1336
+    },
+    {
+      "epoch": 1.028242074927954,
+      "grad_norm": 0.8376697395117676,
+      "learning_rate": 5.817502233543355e-06,
+      "loss": 0.0991,
+      "step": 1338
+    },
+    {
+      "epoch": 1.0297790585975024,
+      "grad_norm": 0.7468897147308834,
+      "learning_rate": 5.8128061403771815e-06,
+      "loss": 0.0885,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0313160422670509,
+      "grad_norm": 0.8645299006954267,
+      "learning_rate": 5.8081026445163184e-06,
+      "loss": 0.1036,
+      "step": 1342
+    },
+    {
+      "epoch": 1.0328530259365993,
+      "grad_norm": 0.8029697731003538,
+      "learning_rate": 5.80339176101544e-06,
+      "loss": 0.1027,
+      "step": 1344
+    },
+    {
+      "epoch": 1.034390009606148,
+      "grad_norm": 0.7986403468821157,
+      "learning_rate": 5.798673504952866e-06,
+      "loss": 0.0982,
+      "step": 1346
+    },
+    {
+      "epoch": 1.0359269932756965,
+      "grad_norm": 0.8210887006805161,
+      "learning_rate": 5.793947891430516e-06,
+      "loss": 0.0973,
+      "step": 1348
+    },
+    {
+      "epoch": 1.037463976945245,
+      "grad_norm": 0.7710919739110469,
+      "learning_rate": 5.789214935573857e-06,
+      "loss": 0.097,
+      "step": 1350
+    },
+    {
+      "epoch": 1.0390009606147934,
+      "grad_norm": 0.7754287784899241,
+      "learning_rate": 5.784474652531857e-06,
+      "loss": 0.0962,
+      "step": 1352
+    },
+    {
+      "epoch": 1.0405379442843419,
+      "grad_norm": 0.7811613718003362,
+      "learning_rate": 5.779727057476938e-06,
+      "loss": 0.101,
+      "step": 1354
+    },
+    {
+      "epoch": 1.0420749279538906,
+      "grad_norm": 0.7683352035150061,
+      "learning_rate": 5.774972165604923e-06,
+      "loss": 0.0926,
+      "step": 1356
+    },
+    {
+      "epoch": 1.043611911623439,
+      "grad_norm": 0.8121646663638018,
+      "learning_rate": 5.770209992134992e-06,
+      "loss": 0.1025,
+      "step": 1358
+    },
+    {
+      "epoch": 1.0451488952929875,
+      "grad_norm": 0.7946845821407413,
+      "learning_rate": 5.765440552309633e-06,
+      "loss": 0.098,
+      "step": 1360
+    },
+    {
+      "epoch": 1.046685878962536,
+      "grad_norm": 0.8107985568570227,
+      "learning_rate": 5.760663861394589e-06,
+      "loss": 0.0939,
+      "step": 1362
+    },
+    {
+      "epoch": 1.0482228626320846,
+      "grad_norm": 0.7738201688286012,
+      "learning_rate": 5.755879934678815e-06,
+      "loss": 0.0977,
+      "step": 1364
+    },
+    {
+      "epoch": 1.049759846301633,
+      "grad_norm": 0.8043266326845157,
+      "learning_rate": 5.751088787474421e-06,
+      "loss": 0.104,
+      "step": 1366
+    },
+    {
+      "epoch": 1.0512968299711816,
+      "grad_norm": 0.8212861671574773,
+      "learning_rate": 5.746290435116633e-06,
+      "loss": 0.1012,
+      "step": 1368
+    },
+    {
+      "epoch": 1.05283381364073,
+      "grad_norm": 0.8217655361068997,
+      "learning_rate": 5.7414848929637344e-06,
+      "loss": 0.0948,
+      "step": 1370
+    },
+    {
+      "epoch": 1.0543707973102785,
+      "grad_norm": 0.7398326798927687,
+      "learning_rate": 5.7366721763970276e-06,
+      "loss": 0.0926,
+      "step": 1372
+    },
+    {
+      "epoch": 1.0559077809798272,
+      "grad_norm": 0.8684068565935559,
+      "learning_rate": 5.73185230082077e-06,
+      "loss": 0.0977,
+      "step": 1374
+    },
+    {
+      "epoch": 1.0574447646493756,
+      "grad_norm": 0.812673696526823,
+      "learning_rate": 5.727025281662141e-06,
+      "loss": 0.0922,
+      "step": 1376
+    },
+    {
+      "epoch": 1.058981748318924,
+      "grad_norm": 0.8179613499881782,
+      "learning_rate": 5.722191134371179e-06,
+      "loss": 0.0944,
+      "step": 1378
+    },
+    {
+      "epoch": 1.0605187319884726,
+      "grad_norm": 0.810924645502839,
+      "learning_rate": 5.717349874420742e-06,
+      "loss": 0.0924,
+      "step": 1380
+    },
+    {
+      "epoch": 1.062055715658021,
+      "grad_norm": 0.763255489678059,
+      "learning_rate": 5.71250151730645e-06,
+      "loss": 0.0896,
+      "step": 1382
+    },
+    {
+      "epoch": 1.0635926993275697,
+      "grad_norm": 0.8239338886221861,
+      "learning_rate": 5.707646078546642e-06,
+      "loss": 0.1043,
+      "step": 1384
+    },
+    {
+      "epoch": 1.0651296829971182,
+      "grad_norm": 0.7862418625776587,
+      "learning_rate": 5.702783573682323e-06,
+      "loss": 0.0927,
+      "step": 1386
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 0.7694175148908957,
+      "learning_rate": 5.697914018277113e-06,
+      "loss": 0.0905,
+      "step": 1388
+    },
+    {
+      "epoch": 1.0682036503362151,
+      "grad_norm": 0.8056119084783939,
+      "learning_rate": 5.693037427917201e-06,
+      "loss": 0.0964,
+      "step": 1390
+    },
+    {
+      "epoch": 1.0697406340057638,
+      "grad_norm": 0.827006265409657,
+      "learning_rate": 5.688153818211293e-06,
+      "loss": 0.0905,
+      "step": 1392
+    },
+    {
+      "epoch": 1.0712776176753123,
+      "grad_norm": 0.8282583664130263,
+      "learning_rate": 5.683263204790561e-06,
+      "loss": 0.0915,
+      "step": 1394
+    },
+    {
+      "epoch": 1.0728146013448607,
+      "grad_norm": 0.7937251940987193,
+      "learning_rate": 5.678365603308593e-06,
+      "loss": 0.0969,
+      "step": 1396
+    },
+    {
+      "epoch": 1.0743515850144092,
+      "grad_norm": 0.7954663404413561,
+      "learning_rate": 5.673461029441347e-06,
+      "loss": 0.0996,
+      "step": 1398
+    },
+    {
+      "epoch": 1.0758885686839577,
+      "grad_norm": 0.8270927064110352,
+      "learning_rate": 5.668549498887098e-06,
+      "loss": 0.1034,
+      "step": 1400
+    },
+    {
+      "epoch": 1.0774255523535063,
+      "grad_norm": 0.7670514687913251,
+      "learning_rate": 5.663631027366382e-06,
+      "loss": 0.0913,
+      "step": 1402
+    },
+    {
+      "epoch": 1.0789625360230548,
+      "grad_norm": 0.8715536438358069,
+      "learning_rate": 5.658705630621959e-06,
+      "loss": 0.1004,
+      "step": 1404
+    },
+    {
+      "epoch": 1.0804995196926033,
+      "grad_norm": 0.8161604730434415,
+      "learning_rate": 5.653773324418748e-06,
+      "loss": 0.0943,
+      "step": 1406
+    },
+    {
+      "epoch": 1.0820365033621517,
+      "grad_norm": 0.8667760108772686,
+      "learning_rate": 5.648834124543787e-06,
+      "loss": 0.0954,
+      "step": 1408
+    },
+    {
+      "epoch": 1.0835734870317002,
+      "grad_norm": 0.8231577798724082,
+      "learning_rate": 5.643888046806179e-06,
+      "loss": 0.1,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0851104707012489,
+      "grad_norm": 0.7413599439846686,
+      "learning_rate": 5.638935107037041e-06,
+      "loss": 0.0866,
+      "step": 1412
+    },
+    {
+      "epoch": 1.0866474543707973,
+      "grad_norm": 0.7754362451670733,
+      "learning_rate": 5.633975321089452e-06,
+      "loss": 0.0964,
+      "step": 1414
+    },
+    {
+      "epoch": 1.0881844380403458,
+      "grad_norm": 0.7854818133378504,
+      "learning_rate": 5.629008704838404e-06,
+      "loss": 0.0978,
+      "step": 1416
+    },
+    {
+      "epoch": 1.0897214217098943,
+      "grad_norm": 0.8230491957163081,
+      "learning_rate": 5.624035274180755e-06,
+      "loss": 0.0904,
+      "step": 1418
+    },
+    {
+      "epoch": 1.0912584053794427,
+      "grad_norm": 0.8415692370246974,
+      "learning_rate": 5.619055045035168e-06,
+      "loss": 0.0949,
+      "step": 1420
+    },
+    {
+      "epoch": 1.0927953890489914,
+      "grad_norm": 0.7752546061977003,
+      "learning_rate": 5.6140680333420714e-06,
+      "loss": 0.0977,
+      "step": 1422
+    },
+    {
+      "epoch": 1.09433237271854,
+      "grad_norm": 0.796060251145471,
+      "learning_rate": 5.609074255063598e-06,
+      "loss": 0.0986,
+      "step": 1424
+    },
+    {
+      "epoch": 1.0958693563880884,
+      "grad_norm": 0.8014273381014567,
+      "learning_rate": 5.604073726183545e-06,
+      "loss": 0.0996,
+      "step": 1426
+    },
+    {
+      "epoch": 1.0974063400576368,
+      "grad_norm": 0.8574802894140412,
+      "learning_rate": 5.599066462707311e-06,
+      "loss": 0.0915,
+      "step": 1428
+    },
+    {
+      "epoch": 1.0989433237271853,
+      "grad_norm": 0.7973180847732659,
+      "learning_rate": 5.594052480661852e-06,
+      "loss": 0.098,
+      "step": 1430
+    },
+    {
+      "epoch": 1.100480307396734,
+      "grad_norm": 0.7978978724248726,
+      "learning_rate": 5.589031796095629e-06,
+      "loss": 0.095,
+      "step": 1432
+    },
+    {
+      "epoch": 1.1020172910662824,
+      "grad_norm": 0.7735364885990075,
+      "learning_rate": 5.584004425078556e-06,
+      "loss": 0.0977,
+      "step": 1434
+    },
+    {
+      "epoch": 1.103554274735831,
+      "grad_norm": 0.824645677844322,
+      "learning_rate": 5.5789703837019465e-06,
+      "loss": 0.101,
+      "step": 1436
+    },
+    {
+      "epoch": 1.1050912584053794,
+      "grad_norm": 0.8045096798761333,
+      "learning_rate": 5.5739296880784685e-06,
+      "loss": 0.0995,
+      "step": 1438
+    },
+    {
+      "epoch": 1.106628242074928,
+      "grad_norm": 0.7785417170752582,
+      "learning_rate": 5.568882354342084e-06,
+      "loss": 0.0944,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1081652257444765,
+      "grad_norm": 0.7865456166072924,
+      "learning_rate": 5.5638283986480055e-06,
+      "loss": 0.0993,
+      "step": 1442
+    },
+    {
+      "epoch": 1.109702209414025,
+      "grad_norm": 0.7482822898483027,
+      "learning_rate": 5.5587678371726365e-06,
+      "loss": 0.0923,
+      "step": 1444
+    },
+    {
+      "epoch": 1.1112391930835734,
+      "grad_norm": 0.847672498751733,
+      "learning_rate": 5.553700686113528e-06,
+      "loss": 0.0924,
+      "step": 1446
+    },
+    {
+      "epoch": 1.112776176753122,
+      "grad_norm": 0.8088910111550272,
+      "learning_rate": 5.54862696168932e-06,
+      "loss": 0.0969,
+      "step": 1448
+    },
+    {
+      "epoch": 1.1143131604226706,
+      "grad_norm": 0.7477679679317595,
+      "learning_rate": 5.543546680139695e-06,
+      "loss": 0.089,
+      "step": 1450
+    },
+    {
+      "epoch": 1.115850144092219,
+      "grad_norm": 0.8008986821135872,
+      "learning_rate": 5.5384598577253185e-06,
+      "loss": 0.0979,
+      "step": 1452
+    },
+    {
+      "epoch": 1.1173871277617675,
+      "grad_norm": 0.8322762593719337,
+      "learning_rate": 5.533366510727797e-06,
+      "loss": 0.0959,
+      "step": 1454
+    },
+    {
+      "epoch": 1.118924111431316,
+      "grad_norm": 0.7810694389767996,
+      "learning_rate": 5.528266655449615e-06,
+      "loss": 0.1004,
+      "step": 1456
+    },
+    {
+      "epoch": 1.1204610951008647,
+      "grad_norm": 0.8168615959182653,
+      "learning_rate": 5.523160308214093e-06,
+      "loss": 0.0934,
+      "step": 1458
+    },
+    {
+      "epoch": 1.1219980787704131,
+      "grad_norm": 0.8202613520761898,
+      "learning_rate": 5.518047485365328e-06,
+      "loss": 0.0923,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1235350624399616,
+      "grad_norm": 0.7506685980798529,
+      "learning_rate": 5.512928203268145e-06,
+      "loss": 0.0983,
+      "step": 1462
+    },
+    {
+      "epoch": 1.12507204610951,
+      "grad_norm": 0.7964645064908876,
+      "learning_rate": 5.507802478308043e-06,
+      "loss": 0.1018,
+      "step": 1464
+    },
+    {
+      "epoch": 1.1266090297790585,
+      "grad_norm": 0.7856943241580147,
+      "learning_rate": 5.502670326891141e-06,
+      "loss": 0.0897,
+      "step": 1466
+    },
+    {
+      "epoch": 1.1281460134486072,
+      "grad_norm": 0.9369819821403769,
+      "learning_rate": 5.497531765444132e-06,
+      "loss": 0.0997,
+      "step": 1468
+    },
+    {
+      "epoch": 1.1296829971181557,
+      "grad_norm": 0.8411240897070034,
+      "learning_rate": 5.492386810414222e-06,
+      "loss": 0.099,
+      "step": 1470
+    },
+    {
+      "epoch": 1.1312199807877041,
+      "grad_norm": 0.7880140283795076,
+      "learning_rate": 5.48723547826908e-06,
+      "loss": 0.0962,
+      "step": 1472
+    },
+    {
+      "epoch": 1.1327569644572526,
+      "grad_norm": 0.7705964805253543,
+      "learning_rate": 5.482077785496794e-06,
+      "loss": 0.0931,
+      "step": 1474
+    },
+    {
+      "epoch": 1.134293948126801,
+      "grad_norm": 0.8165812210553237,
+      "learning_rate": 5.4769137486058e-06,
+      "loss": 0.1001,
+      "step": 1476
+    },
+    {
+      "epoch": 1.1358309317963498,
+      "grad_norm": 0.8057592991161191,
+      "learning_rate": 5.471743384124848e-06,
+      "loss": 0.0943,
+      "step": 1478
+    },
+    {
+      "epoch": 1.1373679154658982,
+      "grad_norm": 0.7946279160573669,
+      "learning_rate": 5.466566708602939e-06,
+      "loss": 0.0968,
+      "step": 1480
+    },
+    {
+      "epoch": 1.1389048991354467,
+      "grad_norm": 0.7589266561686325,
+      "learning_rate": 5.461383738609272e-06,
+      "loss": 0.0919,
+      "step": 1482
+    },
+    {
+      "epoch": 1.1404418828049951,
+      "grad_norm": 0.7750698355623663,
+      "learning_rate": 5.456194490733194e-06,
+      "loss": 0.0905,
+      "step": 1484
+    },
+    {
+      "epoch": 1.1419788664745436,
+      "grad_norm": 0.7612447169194039,
+      "learning_rate": 5.450998981584148e-06,
+      "loss": 0.0876,
+      "step": 1486
+    },
+    {
+      "epoch": 1.1435158501440923,
+      "grad_norm": 0.7659227044968493,
+      "learning_rate": 5.445797227791616e-06,
+      "loss": 0.0913,
+      "step": 1488
+    },
+    {
+      "epoch": 1.1450528338136408,
+      "grad_norm": 0.8296191241219513,
+      "learning_rate": 5.440589246005064e-06,
+      "loss": 0.1023,
+      "step": 1490
+    },
+    {
+      "epoch": 1.1465898174831892,
+      "grad_norm": 0.8332033205007243,
+      "learning_rate": 5.4353750528938995e-06,
+      "loss": 0.1018,
+      "step": 1492
+    },
+    {
+      "epoch": 1.1481268011527377,
+      "grad_norm": 0.755633743693491,
+      "learning_rate": 5.430154665147406e-06,
+      "loss": 0.0982,
+      "step": 1494
+    },
+    {
+      "epoch": 1.1496637848222861,
+      "grad_norm": 0.8294945642825525,
+      "learning_rate": 5.424928099474694e-06,
+      "loss": 0.101,
+      "step": 1496
+    },
+    {
+      "epoch": 1.1512007684918348,
+      "grad_norm": 0.8705045376216894,
+      "learning_rate": 5.419695372604652e-06,
+      "loss": 0.0966,
+      "step": 1498
+    },
+    {
+      "epoch": 1.1527377521613833,
+      "grad_norm": 0.8386991034461505,
+      "learning_rate": 5.414456501285883e-06,
+      "loss": 0.0948,
+      "step": 1500
+    },
+    {
+      "epoch": 1.1527377521613833,
+      "eval_loss": 0.14147287607192993,
+      "eval_runtime": 362.5666,
+      "eval_samples_per_second": 51.039,
+      "eval_steps_per_second": 6.382,
+      "step": 1500
+    },
+    {
+      "epoch": 1.1542747358309318,
+      "grad_norm": 0.808895295948306,
+      "learning_rate": 5.409211502286663e-06,
+      "loss": 0.1023,
+      "step": 1502
+    },
+    {
+      "epoch": 1.1558117195004802,
+      "grad_norm": 0.8129387791807478,
+      "learning_rate": 5.403960392394877e-06,
+      "loss": 0.0917,
+      "step": 1504
+    },
+    {
+      "epoch": 1.157348703170029,
+      "grad_norm": 0.8709858446003281,
+      "learning_rate": 5.398703188417971e-06,
+      "loss": 0.1076,
+      "step": 1506
+    },
+    {
+      "epoch": 1.1588856868395774,
+      "grad_norm": 0.7731592974500305,
+      "learning_rate": 5.393439907182895e-06,
+      "loss": 0.0932,
+      "step": 1508
+    },
+    {
+      "epoch": 1.1604226705091258,
+      "grad_norm": 0.8193191652163263,
+      "learning_rate": 5.388170565536052e-06,
+      "loss": 0.087,
+      "step": 1510
+    },
+    {
+      "epoch": 1.1619596541786743,
+      "grad_norm": 0.8821413504599133,
+      "learning_rate": 5.382895180343243e-06,
+      "loss": 0.1016,
+      "step": 1512
+    },
+    {
+      "epoch": 1.1634966378482228,
+      "grad_norm": 0.8330757982422947,
+      "learning_rate": 5.377613768489613e-06,
+      "loss": 0.0902,
+      "step": 1514
+    },
+    {
+      "epoch": 1.1650336215177715,
+      "grad_norm": 0.7755643958777008,
+      "learning_rate": 5.372326346879597e-06,
+      "loss": 0.0935,
+      "step": 1516
+    },
+    {
+      "epoch": 1.16657060518732,
+      "grad_norm": 0.7798742092931141,
+      "learning_rate": 5.367032932436863e-06,
+      "loss": 0.0907,
+      "step": 1518
+    },
+    {
+      "epoch": 1.1681075888568684,
+      "grad_norm": 0.7669670923992181,
+      "learning_rate": 5.3617335421042644e-06,
+      "loss": 0.1023,
+      "step": 1520
+    },
+    {
+      "epoch": 1.1696445725264168,
+      "grad_norm": 0.8341959799418742,
+      "learning_rate": 5.3564281928437785e-06,
+      "loss": 0.1016,
+      "step": 1522
+    },
+    {
+      "epoch": 1.1711815561959655,
+      "grad_norm": 0.7444403179993622,
+      "learning_rate": 5.351116901636459e-06,
+      "loss": 0.0895,
+      "step": 1524
+    },
+    {
+      "epoch": 1.172718539865514,
+      "grad_norm": 0.788295757514909,
+      "learning_rate": 5.3457996854823736e-06,
+      "loss": 0.0955,
+      "step": 1526
+    },
+    {
+      "epoch": 1.1742555235350625,
+      "grad_norm": 0.8581606629276641,
+      "learning_rate": 5.340476561400559e-06,
+      "loss": 0.0956,
+      "step": 1528
+    },
+    {
+      "epoch": 1.175792507204611,
+      "grad_norm": 0.7680372748823804,
+      "learning_rate": 5.33514754642896e-06,
+      "loss": 0.0892,
+      "step": 1530
+    },
+    {
+      "epoch": 1.1773294908741594,
+      "grad_norm": 0.8579631291966535,
+      "learning_rate": 5.329812657624374e-06,
+      "loss": 0.0947,
+      "step": 1532
+    },
+    {
+      "epoch": 1.178866474543708,
+      "grad_norm": 0.8028770597268665,
+      "learning_rate": 5.324471912062402e-06,
+      "loss": 0.0919,
+      "step": 1534
+    },
+    {
+      "epoch": 1.1804034582132565,
+      "grad_norm": 0.8274846559591892,
+      "learning_rate": 5.319125326837392e-06,
+      "loss": 0.0931,
+      "step": 1536
+    },
+    {
+      "epoch": 1.181940441882805,
+      "grad_norm": 0.8099150871185563,
+      "learning_rate": 5.3137729190623784e-06,
+      "loss": 0.1059,
+      "step": 1538
+    },
+    {
+      "epoch": 1.1834774255523535,
+      "grad_norm": 0.7888074507502317,
+      "learning_rate": 5.308414705869037e-06,
+      "loss": 0.0948,
+      "step": 1540
+    },
+    {
+      "epoch": 1.185014409221902,
+      "grad_norm": 0.736344371575378,
+      "learning_rate": 5.3030507044076244e-06,
+      "loss": 0.0927,
+      "step": 1542
+    },
+    {
+      "epoch": 1.1865513928914506,
+      "grad_norm": 0.8067822180532194,
+      "learning_rate": 5.29768093184692e-06,
+      "loss": 0.0916,
+      "step": 1544
+    },
+    {
+      "epoch": 1.188088376560999,
+      "grad_norm": 0.8588531827838762,
+      "learning_rate": 5.29230540537418e-06,
+      "loss": 0.0962,
+      "step": 1546
+    },
+    {
+      "epoch": 1.1896253602305475,
+      "grad_norm": 0.8183423934241679,
+      "learning_rate": 5.286924142195075e-06,
+      "loss": 0.0908,
+      "step": 1548
+    },
+    {
+      "epoch": 1.191162343900096,
+      "grad_norm": 0.7893384100991108,
+      "learning_rate": 5.2815371595336375e-06,
+      "loss": 0.0949,
+      "step": 1550
+    },
+    {
+      "epoch": 1.1926993275696445,
+      "grad_norm": 0.8369129068810596,
+      "learning_rate": 5.2761444746322054e-06,
+      "loss": 0.1004,
+      "step": 1552
+    },
+    {
+      "epoch": 1.1942363112391932,
+      "grad_norm": 0.852519706321052,
+      "learning_rate": 5.270746104751371e-06,
+      "loss": 0.1071,
+      "step": 1554
+    },
+    {
+      "epoch": 1.1957732949087416,
+      "grad_norm": 0.7067264280513246,
+      "learning_rate": 5.265342067169921e-06,
+      "loss": 0.0866,
+      "step": 1556
+    },
+    {
+      "epoch": 1.19731027857829,
+      "grad_norm": 0.7807816607071811,
+      "learning_rate": 5.259932379184782e-06,
+      "loss": 0.0947,
+      "step": 1558
+    },
+    {
+      "epoch": 1.1988472622478386,
+      "grad_norm": 0.7595283315156356,
+      "learning_rate": 5.254517058110968e-06,
+      "loss": 0.093,
+      "step": 1560
+    },
+    {
+      "epoch": 1.200384245917387,
+      "grad_norm": 0.7355421808089894,
+      "learning_rate": 5.249096121281521e-06,
+      "loss": 0.0929,
+      "step": 1562
+    },
+    {
+      "epoch": 1.2019212295869357,
+      "grad_norm": 0.8309945335729965,
+      "learning_rate": 5.243669586047459e-06,
+      "loss": 0.1038,
+      "step": 1564
+    },
+    {
+      "epoch": 1.2034582132564842,
+      "grad_norm": 0.7783243017854818,
+      "learning_rate": 5.238237469777719e-06,
+      "loss": 0.0919,
+      "step": 1566
+    },
+    {
+      "epoch": 1.2049951969260326,
+      "grad_norm": 0.7536478224038158,
+      "learning_rate": 5.232799789859102e-06,
+      "loss": 0.0832,
+      "step": 1568
+    },
+    {
+      "epoch": 1.206532180595581,
+      "grad_norm": 0.8375220966676139,
+      "learning_rate": 5.227356563696215e-06,
+      "loss": 0.0933,
+      "step": 1570
+    },
+    {
+      "epoch": 1.2080691642651298,
+      "grad_norm": 0.8168530200077441,
+      "learning_rate": 5.221907808711418e-06,
+      "loss": 0.0962,
+      "step": 1572
+    },
+    {
+      "epoch": 1.2096061479346782,
+      "grad_norm": 0.7669312918946458,
+      "learning_rate": 5.216453542344768e-06,
+      "loss": 0.0876,
+      "step": 1574
+    },
+    {
+      "epoch": 1.2111431316042267,
+      "grad_norm": 0.8257303860560026,
+      "learning_rate": 5.210993782053961e-06,
+      "loss": 0.0923,
+      "step": 1576
+    },
+    {
+      "epoch": 1.2126801152737752,
+      "grad_norm": 0.8359416165543359,
+      "learning_rate": 5.205528545314281e-06,
+      "loss": 0.09,
+      "step": 1578
+    },
+    {
+      "epoch": 1.2142170989433236,
+      "grad_norm": 0.8022199675695576,
+      "learning_rate": 5.200057849618535e-06,
+      "loss": 0.0928,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2157540826128723,
+      "grad_norm": 0.809397949668016,
+      "learning_rate": 5.194581712477007e-06,
+      "loss": 0.1011,
+      "step": 1582
+    },
+    {
+      "epoch": 1.2172910662824208,
+      "grad_norm": 0.7620719694922747,
+      "learning_rate": 5.1891001514173994e-06,
+      "loss": 0.0947,
+      "step": 1584
+    },
+    {
+      "epoch": 1.2188280499519693,
+      "grad_norm": 0.8748786606947648,
+      "learning_rate": 5.183613183984768e-06,
+      "loss": 0.1022,
+      "step": 1586
+    },
+    {
+      "epoch": 1.2203650336215177,
+      "grad_norm": 0.8471054549064284,
+      "learning_rate": 5.178120827741481e-06,
+      "loss": 0.0994,
+      "step": 1588
+    },
+    {
+      "epoch": 1.2219020172910664,
+      "grad_norm": 0.8328172662255555,
+      "learning_rate": 5.172623100267148e-06,
+      "loss": 0.09,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2234390009606149,
+      "grad_norm": 0.7751726026024895,
+      "learning_rate": 5.167120019158578e-06,
+      "loss": 0.0872,
+      "step": 1592
+    },
+    {
+      "epoch": 1.2249759846301633,
+      "grad_norm": 0.8122210518131329,
+      "learning_rate": 5.1616116020297065e-06,
+      "loss": 0.0901,
+      "step": 1594
+    },
+    {
+      "epoch": 1.2265129682997118,
+      "grad_norm": 0.7575239070267672,
+      "learning_rate": 5.1560978665115555e-06,
+      "loss": 0.0911,
+      "step": 1596
+    },
+    {
+      "epoch": 1.2280499519692603,
+      "grad_norm": 0.7802284165444789,
+      "learning_rate": 5.150578830252168e-06,
+      "loss": 0.0943,
+      "step": 1598
+    },
+    {
+      "epoch": 1.229586935638809,
+      "grad_norm": 0.8020724802994549,
+      "learning_rate": 5.145054510916552e-06,
+      "loss": 0.0861,
+      "step": 1600
+    },
+    {
+      "epoch": 1.2311239193083574,
+      "grad_norm": 0.9021178290335383,
+      "learning_rate": 5.139524926186624e-06,
+      "loss": 0.1049,
+      "step": 1602
+    },
+    {
+      "epoch": 1.2326609029779059,
+      "grad_norm": 0.7954364435694474,
+      "learning_rate": 5.133990093761158e-06,
+      "loss": 0.0975,
+      "step": 1604
+    },
+    {
+      "epoch": 1.2341978866474543,
+      "grad_norm": 0.862849089104852,
+      "learning_rate": 5.1284500313557214e-06,
+      "loss": 0.1001,
+      "step": 1606
+    },
+    {
+      "epoch": 1.2357348703170028,
+      "grad_norm": 0.8234510165839098,
+      "learning_rate": 5.122904756702622e-06,
+      "loss": 0.096,
+      "step": 1608
+    },
+    {
+      "epoch": 1.2372718539865515,
+      "grad_norm": 0.8180910858318627,
+      "learning_rate": 5.1173542875508495e-06,
+      "loss": 0.0929,
+      "step": 1610
+    },
+    {
+      "epoch": 1.2388088376561,
+      "grad_norm": 0.8254978234033816,
+      "learning_rate": 5.111798641666022e-06,
+      "loss": 0.0986,
+      "step": 1612
+    },
+    {
+      "epoch": 1.2403458213256484,
+      "grad_norm": 0.7970300036804161,
+      "learning_rate": 5.1062378368303286e-06,
+      "loss": 0.0968,
+      "step": 1614
+    },
+    {
+      "epoch": 1.2418828049951969,
+      "grad_norm": 0.7942368711186357,
+      "learning_rate": 5.100671890842464e-06,
+      "loss": 0.0946,
+      "step": 1616
+    },
+    {
+      "epoch": 1.2434197886647453,
+      "grad_norm": 0.781461470784258,
+      "learning_rate": 5.095100821517586e-06,
+      "loss": 0.0909,
+      "step": 1618
+    },
+    {
+      "epoch": 1.244956772334294,
+      "grad_norm": 0.7637849604119566,
+      "learning_rate": 5.089524646687245e-06,
+      "loss": 0.089,
+      "step": 1620
+    },
+    {
+      "epoch": 1.2464937560038425,
+      "grad_norm": 0.8322634659167238,
+      "learning_rate": 5.083943384199339e-06,
+      "loss": 0.0946,
+      "step": 1622
+    },
+    {
+      "epoch": 1.248030739673391,
+      "grad_norm": 0.8673983066924847,
+      "learning_rate": 5.078357051918042e-06,
+      "loss": 0.0997,
+      "step": 1624
+    },
+    {
+      "epoch": 1.2495677233429394,
+      "grad_norm": 0.8093921155750359,
+      "learning_rate": 5.072765667723763e-06,
+      "loss": 0.0924,
+      "step": 1626
+    },
+    {
+      "epoch": 1.2511047070124879,
+      "grad_norm": 0.7682990135468413,
+      "learning_rate": 5.067169249513078e-06,
+      "loss": 0.0908,
+      "step": 1628
+    },
+    {
+      "epoch": 1.2526416906820366,
+      "grad_norm": 0.9219480609910328,
+      "learning_rate": 5.061567815198674e-06,
+      "loss": 0.0992,
+      "step": 1630
+    },
+    {
+      "epoch": 1.254178674351585,
+      "grad_norm": 0.7864964134564114,
+      "learning_rate": 5.055961382709294e-06,
+      "loss": 0.0856,
+      "step": 1632
+    },
+    {
+      "epoch": 1.2557156580211335,
+      "grad_norm": 0.789369364367871,
+      "learning_rate": 5.05034996998968e-06,
+      "loss": 0.101,
+      "step": 1634
+    },
+    {
+      "epoch": 1.257252641690682,
+      "grad_norm": 0.785618855267339,
+      "learning_rate": 5.044733595000514e-06,
+      "loss": 0.0913,
+      "step": 1636
+    },
+    {
+      "epoch": 1.2587896253602304,
+      "grad_norm": 0.8540857446211212,
+      "learning_rate": 5.0391122757183605e-06,
+      "loss": 0.1008,
+      "step": 1638
+    },
+    {
+      "epoch": 1.260326609029779,
+      "grad_norm": 0.8815323610676825,
+      "learning_rate": 5.03348603013561e-06,
+      "loss": 0.0967,
+      "step": 1640
+    },
+    {
+      "epoch": 1.2618635926993276,
+      "grad_norm": 0.8589693358904777,
+      "learning_rate": 5.02785487626042e-06,
+      "loss": 0.0979,
+      "step": 1642
+    },
+    {
+      "epoch": 1.263400576368876,
+      "grad_norm": 0.8415206231916414,
+      "learning_rate": 5.022218832116659e-06,
+      "loss": 0.1038,
+      "step": 1644
+    },
+    {
+      "epoch": 1.2649375600384247,
+      "grad_norm": 0.8392847365814639,
+      "learning_rate": 5.016577915743848e-06,
+      "loss": 0.0999,
+      "step": 1646
+    },
+    {
+      "epoch": 1.266474543707973,
+      "grad_norm": 0.840680092317177,
+      "learning_rate": 5.010932145197101e-06,
+      "loss": 0.0959,
+      "step": 1648
+    },
+    {
+      "epoch": 1.2680115273775217,
+      "grad_norm": 0.8176605216769884,
+      "learning_rate": 5.005281538547071e-06,
+      "loss": 0.0968,
+      "step": 1650
+    },
+    {
+      "epoch": 1.2695485110470701,
+      "grad_norm": 0.7636164672938492,
+      "learning_rate": 4.999626113879891e-06,
+      "loss": 0.0855,
+      "step": 1652
+    },
+    {
+      "epoch": 1.2710854947166186,
+      "grad_norm": 0.8820138872988845,
+      "learning_rate": 4.9939658892971106e-06,
+      "loss": 0.0963,
+      "step": 1654
+    },
+    {
+      "epoch": 1.2726224783861673,
+      "grad_norm": 0.8230226501954376,
+      "learning_rate": 4.9883008829156475e-06,
+      "loss": 0.0908,
+      "step": 1656
+    },
+    {
+      "epoch": 1.2741594620557157,
+      "grad_norm": 0.7954482370699276,
+      "learning_rate": 4.982631112867724e-06,
+      "loss": 0.0949,
+      "step": 1658
+    },
+    {
+      "epoch": 1.2756964457252642,
+      "grad_norm": 0.7446617995311455,
+      "learning_rate": 4.976956597300806e-06,
+      "loss": 0.0922,
+      "step": 1660
+    },
+    {
+      "epoch": 1.2772334293948127,
+      "grad_norm": 0.8316466452619663,
+      "learning_rate": 4.971277354377554e-06,
+      "loss": 0.0961,
+      "step": 1662
+    },
+    {
+      "epoch": 1.2787704130643611,
+      "grad_norm": 0.8717333357230952,
+      "learning_rate": 4.965593402275754e-06,
+      "loss": 0.0989,
+      "step": 1664
+    },
+    {
+      "epoch": 1.2803073967339098,
+      "grad_norm": 0.8635438600294546,
+      "learning_rate": 4.959904759188271e-06,
+      "loss": 0.0965,
+      "step": 1666
+    },
+    {
+      "epoch": 1.2818443804034583,
+      "grad_norm": 0.8348320154619787,
+      "learning_rate": 4.954211443322978e-06,
+      "loss": 0.0982,
+      "step": 1668
+    },
+    {
+      "epoch": 1.2833813640730067,
+      "grad_norm": 0.7587760801820468,
+      "learning_rate": 4.948513472902709e-06,
+      "loss": 0.0894,
+      "step": 1670
+    },
+    {
+      "epoch": 1.2849183477425552,
+      "grad_norm": 0.823101148418266,
+      "learning_rate": 4.942810866165194e-06,
+      "loss": 0.1011,
+      "step": 1672
+    },
+    {
+      "epoch": 1.2864553314121037,
+      "grad_norm": 0.8615662400758028,
+      "learning_rate": 4.937103641363004e-06,
+      "loss": 0.0965,
+      "step": 1674
+    },
+    {
+      "epoch": 1.2879923150816524,
+      "grad_norm": 0.7508304626813055,
+      "learning_rate": 4.931391816763491e-06,
+      "loss": 0.0902,
+      "step": 1676
+    },
+    {
+      "epoch": 1.2895292987512008,
+      "grad_norm": 0.7605099804332683,
+      "learning_rate": 4.925675410648728e-06,
+      "loss": 0.0957,
+      "step": 1678
+    },
+    {
+      "epoch": 1.2910662824207493,
+      "grad_norm": 0.8276765994390878,
+      "learning_rate": 4.919954441315453e-06,
+      "loss": 0.0928,
+      "step": 1680
+    },
+    {
+      "epoch": 1.2926032660902977,
+      "grad_norm": 0.8394305545603477,
+      "learning_rate": 4.914228927075014e-06,
+      "loss": 0.0933,
+      "step": 1682
+    },
+    {
+      "epoch": 1.2941402497598462,
+      "grad_norm": 0.8228704133694662,
+      "learning_rate": 4.908498886253298e-06,
+      "loss": 0.0954,
+      "step": 1684
+    },
+    {
+      "epoch": 1.295677233429395,
+      "grad_norm": 0.8428554748260062,
+      "learning_rate": 4.902764337190685e-06,
+      "loss": 0.101,
+      "step": 1686
+    },
+    {
+      "epoch": 1.2972142170989434,
+      "grad_norm": 0.8148117938502604,
+      "learning_rate": 4.897025298241987e-06,
+      "loss": 0.0939,
+      "step": 1688
+    },
+    {
+      "epoch": 1.2987512007684918,
+      "grad_norm": 0.7520640222305418,
+      "learning_rate": 4.891281787776383e-06,
+      "loss": 0.0897,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3002881844380403,
+      "grad_norm": 0.7757726426358977,
+      "learning_rate": 4.885533824177365e-06,
+      "loss": 0.0907,
+      "step": 1692
+    },
+    {
+      "epoch": 1.3018251681075887,
+      "grad_norm": 0.8281519655081531,
+      "learning_rate": 4.87978142584268e-06,
+      "loss": 0.102,
+      "step": 1694
+    },
+    {
+      "epoch": 1.3033621517771374,
+      "grad_norm": 0.845824623898633,
+      "learning_rate": 4.874024611184266e-06,
+      "loss": 0.0958,
+      "step": 1696
+    },
+    {
+      "epoch": 1.304899135446686,
+      "grad_norm": 0.8422886597995062,
+      "learning_rate": 4.868263398628203e-06,
+      "loss": 0.0944,
+      "step": 1698
+    },
+    {
+      "epoch": 1.3064361191162344,
+      "grad_norm": 0.8314043674269285,
+      "learning_rate": 4.86249780661464e-06,
+      "loss": 0.1048,
+      "step": 1700
+    },
+    {
+      "epoch": 1.3079731027857828,
+      "grad_norm": 0.8629695831852081,
+      "learning_rate": 4.8567278535977475e-06,
+      "loss": 0.1105,
+      "step": 1702
+    },
+    {
+      "epoch": 1.3095100864553313,
+      "grad_norm": 0.8379948759443979,
+      "learning_rate": 4.850953558045653e-06,
+      "loss": 0.0973,
+      "step": 1704
+    },
+    {
+      "epoch": 1.31104707012488,
+      "grad_norm": 0.7985221017745798,
+      "learning_rate": 4.845174938440386e-06,
+      "loss": 0.0994,
+      "step": 1706
+    },
+    {
+      "epoch": 1.3125840537944284,
+      "grad_norm": 0.8271445315963422,
+      "learning_rate": 4.8393920132778144e-06,
+      "loss": 0.089,
+      "step": 1708
+    },
+    {
+      "epoch": 1.314121037463977,
+      "grad_norm": 0.8199327418927324,
+      "learning_rate": 4.833604801067585e-06,
+      "loss": 0.0979,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3156580211335256,
+      "grad_norm": 0.7802159907934838,
+      "learning_rate": 4.827813320333071e-06,
+      "loss": 0.0891,
+      "step": 1712
+    },
+    {
+      "epoch": 1.3171950048030738,
+      "grad_norm": 0.8828098223970235,
+      "learning_rate": 4.822017589611302e-06,
+      "loss": 0.0968,
+      "step": 1714
+    },
+    {
+      "epoch": 1.3187319884726225,
+      "grad_norm": 0.7743361926366948,
+      "learning_rate": 4.816217627452917e-06,
+      "loss": 0.0901,
+      "step": 1716
+    },
+    {
+      "epoch": 1.320268972142171,
+      "grad_norm": 0.8452670586980338,
+      "learning_rate": 4.810413452422094e-06,
+      "loss": 0.0923,
+      "step": 1718
+    },
+    {
+      "epoch": 1.3218059558117194,
+      "grad_norm": 0.8330276027385624,
+      "learning_rate": 4.804605083096499e-06,
+      "loss": 0.0952,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3233429394812681,
+      "grad_norm": 0.805109828914744,
+      "learning_rate": 4.798792538067218e-06,
+      "loss": 0.0946,
+      "step": 1722
+    },
+    {
+      "epoch": 1.3248799231508166,
+      "grad_norm": 0.8508784027134717,
+      "learning_rate": 4.792975835938709e-06,
+      "loss": 0.0957,
+      "step": 1724
+    },
+    {
+      "epoch": 1.326416906820365,
+      "grad_norm": 0.7937254392396291,
+      "learning_rate": 4.787154995328729e-06,
+      "loss": 0.0948,
+      "step": 1726
+    },
+    {
+      "epoch": 1.3279538904899135,
+      "grad_norm": 0.7723472021275456,
+      "learning_rate": 4.781330034868287e-06,
+      "loss": 0.0954,
+      "step": 1728
+    },
+    {
+      "epoch": 1.329490874159462,
+      "grad_norm": 0.8147194013177234,
+      "learning_rate": 4.775500973201573e-06,
+      "loss": 0.0994,
+      "step": 1730
+    },
+    {
+      "epoch": 1.3310278578290107,
+      "grad_norm": 0.7865073470553021,
+      "learning_rate": 4.76966782898591e-06,
+      "loss": 0.0908,
+      "step": 1732
+    },
+    {
+      "epoch": 1.3325648414985591,
+      "grad_norm": 0.8909496268629883,
+      "learning_rate": 4.763830620891682e-06,
+      "loss": 0.0994,
+      "step": 1734
+    },
+    {
+      "epoch": 1.3341018251681076,
+      "grad_norm": 0.8439047859296126,
+      "learning_rate": 4.757989367602286e-06,
+      "loss": 0.0948,
+      "step": 1736
+    },
+    {
+      "epoch": 1.335638808837656,
+      "grad_norm": 0.8146110101810443,
+      "learning_rate": 4.752144087814062e-06,
+      "loss": 0.0932,
+      "step": 1738
+    },
+    {
+      "epoch": 1.3371757925072045,
+      "grad_norm": 0.8639874055710914,
+      "learning_rate": 4.746294800236241e-06,
+      "loss": 0.103,
+      "step": 1740
+    },
+    {
+      "epoch": 1.3387127761767532,
+      "grad_norm": 0.8343222963762533,
+      "learning_rate": 4.740441523590881e-06,
+      "loss": 0.0933,
+      "step": 1742
+    },
+    {
+      "epoch": 1.3402497598463017,
+      "grad_norm": 0.9191398577485145,
+      "learning_rate": 4.734584276612807e-06,
+      "loss": 0.0983,
+      "step": 1744
+    },
+    {
+      "epoch": 1.3417867435158501,
+      "grad_norm": 0.8075250938783436,
+      "learning_rate": 4.7287230780495525e-06,
+      "loss": 0.0958,
+      "step": 1746
+    },
+    {
+      "epoch": 1.3433237271853986,
+      "grad_norm": 0.8741221440894832,
+      "learning_rate": 4.722857946661299e-06,
+      "loss": 0.0997,
+      "step": 1748
+    },
+    {
+      "epoch": 1.344860710854947,
+      "grad_norm": 0.8738624905786072,
+      "learning_rate": 4.7169889012208174e-06,
+      "loss": 0.1046,
+      "step": 1750
+    },
+    {
+      "epoch": 1.3463976945244958,
+      "grad_norm": 0.8003724652951454,
+      "learning_rate": 4.711115960513405e-06,
+      "loss": 0.099,
+      "step": 1752
+    },
+    {
+      "epoch": 1.3479346781940442,
+      "grad_norm": 0.7908228630871454,
+      "learning_rate": 4.705239143336827e-06,
+      "loss": 0.0991,
+      "step": 1754
+    },
+    {
+      "epoch": 1.3494716618635927,
+      "grad_norm": 0.7736523132228244,
+      "learning_rate": 4.6993584685012554e-06,
+      "loss": 0.1028,
+      "step": 1756
+    },
+    {
+      "epoch": 1.3510086455331412,
+      "grad_norm": 0.8141238639955589,
+      "learning_rate": 4.693473954829211e-06,
+      "loss": 0.0954,
+      "step": 1758
+    },
+    {
+      "epoch": 1.3525456292026896,
+      "grad_norm": 0.7526341018348459,
+      "learning_rate": 4.687585621155502e-06,
+      "loss": 0.0905,
+      "step": 1760
+    },
+    {
+      "epoch": 1.3540826128722383,
+      "grad_norm": 0.776573467548187,
+      "learning_rate": 4.68169348632716e-06,
+      "loss": 0.098,
+      "step": 1762
+    },
+    {
+      "epoch": 1.3556195965417868,
+      "grad_norm": 0.8263354203916096,
+      "learning_rate": 4.675797569203389e-06,
+      "loss": 0.1029,
+      "step": 1764
+    },
+    {
+      "epoch": 1.3571565802113352,
+      "grad_norm": 0.7840394777513143,
+      "learning_rate": 4.669897888655494e-06,
+      "loss": 0.0938,
+      "step": 1766
+    },
+    {
+      "epoch": 1.3586935638808837,
+      "grad_norm": 0.8026477022187467,
+      "learning_rate": 4.663994463566828e-06,
+      "loss": 0.0972,
+      "step": 1768
+    },
+    {
+      "epoch": 1.3602305475504322,
+      "grad_norm": 0.7984909693964071,
+      "learning_rate": 4.658087312832729e-06,
+      "loss": 0.0934,
+      "step": 1770
+    },
+    {
+      "epoch": 1.3617675312199808,
+      "grad_norm": 0.8445059001868617,
+      "learning_rate": 4.652176455360459e-06,
+      "loss": 0.0969,
+      "step": 1772
+    },
+    {
+      "epoch": 1.3633045148895293,
+      "grad_norm": 0.8508994597739256,
+      "learning_rate": 4.646261910069147e-06,
+      "loss": 0.0895,
+      "step": 1774
+    },
+    {
+      "epoch": 1.3648414985590778,
+      "grad_norm": 0.7381165146693729,
+      "learning_rate": 4.640343695889721e-06,
+      "loss": 0.0893,
+      "step": 1776
+    },
+    {
+      "epoch": 1.3663784822286265,
+      "grad_norm": 0.7966384986609926,
+      "learning_rate": 4.634421831764857e-06,
+      "loss": 0.0992,
+      "step": 1778
+    },
+    {
+      "epoch": 1.3679154658981747,
+      "grad_norm": 0.7322872414451935,
+      "learning_rate": 4.628496336648911e-06,
+      "loss": 0.0868,
+      "step": 1780
+    },
+    {
+      "epoch": 1.3694524495677234,
+      "grad_norm": 0.7490827880894491,
+      "learning_rate": 4.6225672295078615e-06,
+      "loss": 0.0898,
+      "step": 1782
+    },
+    {
+      "epoch": 1.3709894332372718,
+      "grad_norm": 0.8010232538915442,
+      "learning_rate": 4.616634529319249e-06,
+      "loss": 0.0917,
+      "step": 1784
+    },
+    {
+      "epoch": 1.3725264169068203,
+      "grad_norm": 0.8095923768198692,
+      "learning_rate": 4.61069825507211e-06,
+      "loss": 0.092,
+      "step": 1786
+    },
+    {
+      "epoch": 1.374063400576369,
+      "grad_norm": 0.7861781833880861,
+      "learning_rate": 4.604758425766928e-06,
+      "loss": 0.0891,
+      "step": 1788
+    },
+    {
+      "epoch": 1.3756003842459175,
+      "grad_norm": 0.8541796594703853,
+      "learning_rate": 4.5988150604155585e-06,
+      "loss": 0.0962,
+      "step": 1790
+    },
+    {
+      "epoch": 1.377137367915466,
+      "grad_norm": 0.8065855822446965,
+      "learning_rate": 4.592868178041181e-06,
+      "loss": 0.0976,
+      "step": 1792
+    },
+    {
+      "epoch": 1.3786743515850144,
+      "grad_norm": 0.8147810473519597,
+      "learning_rate": 4.586917797678225e-06,
+      "loss": 0.0931,
+      "step": 1794
+    },
+    {
+      "epoch": 1.3802113352545629,
+      "grad_norm": 0.7931514880597085,
+      "learning_rate": 4.5809639383723245e-06,
+      "loss": 0.0952,
+      "step": 1796
+    },
+    {
+      "epoch": 1.3817483189241115,
+      "grad_norm": 0.8182579050054116,
+      "learning_rate": 4.57500661918024e-06,
+      "loss": 0.0971,
+      "step": 1798
+    },
+    {
+      "epoch": 1.38328530259366,
+      "grad_norm": 0.7690232401474136,
+      "learning_rate": 4.569045859169814e-06,
+      "loss": 0.083,
+      "step": 1800
+    },
+    {
+      "epoch": 1.3848222862632085,
+      "grad_norm": 0.7899387190866939,
+      "learning_rate": 4.563081677419897e-06,
+      "loss": 0.0908,
+      "step": 1802
+    },
+    {
+      "epoch": 1.386359269932757,
+      "grad_norm": 0.8326942252427081,
+      "learning_rate": 4.557114093020294e-06,
+      "loss": 0.1004,
+      "step": 1804
+    },
+    {
+      "epoch": 1.3878962536023054,
+      "grad_norm": 0.868404572818935,
+      "learning_rate": 4.551143125071698e-06,
+      "loss": 0.0989,
+      "step": 1806
+    },
+    {
+      "epoch": 1.389433237271854,
+      "grad_norm": 0.7829935600541224,
+      "learning_rate": 4.545168792685637e-06,
+      "loss": 0.0904,
+      "step": 1808
+    },
+    {
+      "epoch": 1.3909702209414025,
+      "grad_norm": 0.8030113132084329,
+      "learning_rate": 4.539191114984403e-06,
+      "loss": 0.0948,
+      "step": 1810
+    },
+    {
+      "epoch": 1.392507204610951,
+      "grad_norm": 0.8322013967452164,
+      "learning_rate": 4.533210111101e-06,
+      "loss": 0.0889,
+      "step": 1812
+    },
+    {
+      "epoch": 1.3940441882804995,
+      "grad_norm": 0.854566482621519,
+      "learning_rate": 4.527225800179074e-06,
+      "loss": 0.0926,
+      "step": 1814
+    },
+    {
+      "epoch": 1.395581171950048,
+      "grad_norm": 0.8546853000055383,
+      "learning_rate": 4.521238201372854e-06,
+      "loss": 0.0985,
+      "step": 1816
+    },
+    {
+      "epoch": 1.3971181556195966,
+      "grad_norm": 0.7861004953262382,
+      "learning_rate": 4.5152473338471025e-06,
+      "loss": 0.0842,
+      "step": 1818
+    },
+    {
+      "epoch": 1.398655139289145,
+      "grad_norm": 0.8279303392404915,
+      "learning_rate": 4.509253216777034e-06,
+      "loss": 0.093,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4001921229586936,
+      "grad_norm": 0.7776536059593128,
+      "learning_rate": 4.50325586934827e-06,
+      "loss": 0.0892,
+      "step": 1822
+    },
+    {
+      "epoch": 1.401729106628242,
+      "grad_norm": 0.823356682996185,
+      "learning_rate": 4.497255310756767e-06,
+      "loss": 0.0873,
+      "step": 1824
+    },
+    {
+      "epoch": 1.4032660902977905,
+      "grad_norm": 0.781617011211452,
+      "learning_rate": 4.491251560208766e-06,
+      "loss": 0.0962,
+      "step": 1826
+    },
+    {
+      "epoch": 1.4048030739673392,
+      "grad_norm": 0.8182872551902259,
+      "learning_rate": 4.485244636920716e-06,
+      "loss": 0.09,
+      "step": 1828
+    },
+    {
+      "epoch": 1.4063400576368876,
+      "grad_norm": 0.7732731640678324,
+      "learning_rate": 4.479234560119231e-06,
+      "loss": 0.0909,
+      "step": 1830
+    },
+    {
+      "epoch": 1.407877041306436,
+      "grad_norm": 0.8358121488236012,
+      "learning_rate": 4.473221349041009e-06,
+      "loss": 0.0945,
+      "step": 1832
+    },
+    {
+      "epoch": 1.4094140249759846,
+      "grad_norm": 0.783898477086607,
+      "learning_rate": 4.467205022932788e-06,
+      "loss": 0.0868,
+      "step": 1834
+    },
+    {
+      "epoch": 1.410951008645533,
+      "grad_norm": 0.8347502314619144,
+      "learning_rate": 4.4611856010512696e-06,
+      "loss": 0.0859,
+      "step": 1836
+    },
+    {
+      "epoch": 1.4124879923150817,
+      "grad_norm": 0.8029436387846094,
+      "learning_rate": 4.455163102663071e-06,
+      "loss": 0.0986,
+      "step": 1838
+    },
+    {
+      "epoch": 1.4140249759846302,
+      "grad_norm": 0.780720619644275,
+      "learning_rate": 4.449137547044651e-06,
+      "loss": 0.092,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4155619596541786,
+      "grad_norm": 0.7647611604463465,
+      "learning_rate": 4.443108953482255e-06,
+      "loss": 0.0843,
+      "step": 1842
+    },
+    {
+      "epoch": 1.4170989433237273,
+      "grad_norm": 0.7569931424177674,
+      "learning_rate": 4.437077341271854e-06,
+      "loss": 0.1,
+      "step": 1844
+    },
+    {
+      "epoch": 1.4186359269932756,
+      "grad_norm": 0.780426436469334,
+      "learning_rate": 4.431042729719081e-06,
+      "loss": 0.0922,
+      "step": 1846
+    },
+    {
+      "epoch": 1.4201729106628243,
+      "grad_norm": 0.8266309005493522,
+      "learning_rate": 4.425005138139165e-06,
+      "loss": 0.091,
+      "step": 1848
+    },
+    {
+      "epoch": 1.4217098943323727,
+      "grad_norm": 0.7919591525646804,
+      "learning_rate": 4.418964585856878e-06,
+      "loss": 0.0914,
+      "step": 1850
+    },
+    {
+      "epoch": 1.4232468780019212,
+      "grad_norm": 0.8676995223573973,
+      "learning_rate": 4.4129210922064655e-06,
+      "loss": 0.0933,
+      "step": 1852
+    },
+    {
+      "epoch": 1.4247838616714699,
+      "grad_norm": 0.8245651321649498,
+      "learning_rate": 4.406874676531591e-06,
+      "loss": 0.099,
+      "step": 1854
+    },
+    {
+      "epoch": 1.4263208453410183,
+      "grad_norm": 0.8084380870931988,
+      "learning_rate": 4.400825358185267e-06,
+      "loss": 0.094,
+      "step": 1856
+    },
+    {
+      "epoch": 1.4278578290105668,
+      "grad_norm": 0.8383642555183834,
+      "learning_rate": 4.394773156529796e-06,
+      "loss": 0.1035,
+      "step": 1858
+    },
+    {
+      "epoch": 1.4293948126801153,
+      "grad_norm": 0.8496971372491992,
+      "learning_rate": 4.388718090936714e-06,
+      "loss": 0.0961,
+      "step": 1860
+    },
+    {
+      "epoch": 1.4309317963496637,
+      "grad_norm": 0.778441978283022,
+      "learning_rate": 4.38266018078672e-06,
+      "loss": 0.0903,
+      "step": 1862
+    },
+    {
+      "epoch": 1.4324687800192124,
+      "grad_norm": 0.8399196603924103,
+      "learning_rate": 4.376599445469619e-06,
+      "loss": 0.0951,
+      "step": 1864
+    },
+    {
+      "epoch": 1.4340057636887609,
+      "grad_norm": 0.8277510344863601,
+      "learning_rate": 4.370535904384257e-06,
+      "loss": 0.0947,
+      "step": 1866
+    },
+    {
+      "epoch": 1.4355427473583093,
+      "grad_norm": 0.8546854576792136,
+      "learning_rate": 4.3644695769384634e-06,
+      "loss": 0.0978,
+      "step": 1868
+    },
+    {
+      "epoch": 1.4370797310278578,
+      "grad_norm": 0.8038219716990682,
+      "learning_rate": 4.358400482548984e-06,
+      "loss": 0.0901,
+      "step": 1870
+    },
+    {
+      "epoch": 1.4386167146974063,
+      "grad_norm": 0.8079184472772435,
+      "learning_rate": 4.352328640641422e-06,
+      "loss": 0.0885,
+      "step": 1872
+    },
+    {
+      "epoch": 1.440153698366955,
+      "grad_norm": 0.7459154763602515,
+      "learning_rate": 4.346254070650172e-06,
+      "loss": 0.0886,
+      "step": 1874
+    },
+    {
+      "epoch": 1.4416906820365034,
+      "grad_norm": 0.7909488135828993,
+      "learning_rate": 4.340176792018365e-06,
+      "loss": 0.0928,
+      "step": 1876
+    },
+    {
+      "epoch": 1.4432276657060519,
+      "grad_norm": 0.8321374668536021,
+      "learning_rate": 4.3340968241977975e-06,
+      "loss": 0.0932,
+      "step": 1878
+    },
+    {
+      "epoch": 1.4447646493756003,
+      "grad_norm": 0.7923523546601555,
+      "learning_rate": 4.328014186648875e-06,
+      "loss": 0.0867,
+      "step": 1880
+    },
+    {
+      "epoch": 1.4463016330451488,
+      "grad_norm": 0.8123704722222743,
+      "learning_rate": 4.321928898840549e-06,
+      "loss": 0.0883,
+      "step": 1882
+    },
+    {
+      "epoch": 1.4478386167146975,
+      "grad_norm": 0.8443684423046657,
+      "learning_rate": 4.315840980250253e-06,
+      "loss": 0.0941,
+      "step": 1884
+    },
+    {
+      "epoch": 1.449375600384246,
+      "grad_norm": 0.8218259476991273,
+      "learning_rate": 4.30975045036384e-06,
+      "loss": 0.0934,
+      "step": 1886
+    },
+    {
+      "epoch": 1.4509125840537944,
+      "grad_norm": 0.7796868880134117,
+      "learning_rate": 4.3036573286755225e-06,
+      "loss": 0.0904,
+      "step": 1888
+    },
+    {
+      "epoch": 1.4524495677233429,
+      "grad_norm": 0.8382801166615858,
+      "learning_rate": 4.297561634687809e-06,
+      "loss": 0.0963,
+      "step": 1890
+    },
+    {
+      "epoch": 1.4539865513928913,
+      "grad_norm": 0.7638917693205276,
+      "learning_rate": 4.291463387911439e-06,
+      "loss": 0.0846,
+      "step": 1892
+    },
+    {
+      "epoch": 1.45552353506244,
+      "grad_norm": 0.8232905374055427,
+      "learning_rate": 4.2853626078653255e-06,
+      "loss": 0.0907,
+      "step": 1894
+    },
+    {
+      "epoch": 1.4570605187319885,
+      "grad_norm": 0.8327882309952656,
+      "learning_rate": 4.279259314076488e-06,
+      "loss": 0.0924,
+      "step": 1896
+    },
+    {
+      "epoch": 1.458597502401537,
+      "grad_norm": 0.8362322094432861,
+      "learning_rate": 4.273153526079994e-06,
+      "loss": 0.0961,
+      "step": 1898
+    },
+    {
+      "epoch": 1.4601344860710854,
+      "grad_norm": 0.8371159641695654,
+      "learning_rate": 4.2670452634188895e-06,
+      "loss": 0.0857,
+      "step": 1900
+    },
+    {
+      "epoch": 1.4616714697406339,
+      "grad_norm": 0.7922677007185823,
+      "learning_rate": 4.260934545644148e-06,
+      "loss": 0.0919,
+      "step": 1902
+    },
+    {
+      "epoch": 1.4632084534101826,
+      "grad_norm": 0.7923726734910232,
+      "learning_rate": 4.254821392314595e-06,
+      "loss": 0.0903,
+      "step": 1904
+    },
+    {
+      "epoch": 1.464745437079731,
+      "grad_norm": 0.7722229174216336,
+      "learning_rate": 4.248705822996856e-06,
+      "loss": 0.0879,
+      "step": 1906
+    },
+    {
+      "epoch": 1.4662824207492795,
+      "grad_norm": 0.8218964345538021,
+      "learning_rate": 4.242587857265288e-06,
+      "loss": 0.0993,
+      "step": 1908
+    },
+    {
+      "epoch": 1.4678194044188282,
+      "grad_norm": 0.8010254286113982,
+      "learning_rate": 4.236467514701916e-06,
+      "loss": 0.0897,
+      "step": 1910
+    },
+    {
+      "epoch": 1.4693563880883764,
+      "grad_norm": 0.7744562817644429,
+      "learning_rate": 4.230344814896378e-06,
+      "loss": 0.0913,
+      "step": 1912
+    },
+    {
+      "epoch": 1.4708933717579251,
+      "grad_norm": 0.8343016463549663,
+      "learning_rate": 4.22421977744585e-06,
+      "loss": 0.0993,
+      "step": 1914
+    },
+    {
+      "epoch": 1.4724303554274736,
+      "grad_norm": 0.8688236869752541,
+      "learning_rate": 4.2180924219549964e-06,
+      "loss": 0.0997,
+      "step": 1916
+    },
+    {
+      "epoch": 1.473967339097022,
+      "grad_norm": 0.7734307337086462,
+      "learning_rate": 4.211962768035896e-06,
+      "loss": 0.0984,
+      "step": 1918
+    },
+    {
+      "epoch": 1.4755043227665707,
+      "grad_norm": 0.7679518244992084,
+      "learning_rate": 4.205830835307989e-06,
+      "loss": 0.0885,
+      "step": 1920
+    },
+    {
+      "epoch": 1.4770413064361192,
+      "grad_norm": 0.7926045518052727,
+      "learning_rate": 4.199696643398008e-06,
+      "loss": 0.0914,
+      "step": 1922
+    },
+    {
+      "epoch": 1.4785782901056677,
+      "grad_norm": 0.8283722584130274,
+      "learning_rate": 4.193560211939913e-06,
+      "loss": 0.094,
+      "step": 1924
+    },
+    {
+      "epoch": 1.4801152737752161,
+      "grad_norm": 0.7250571779590265,
+      "learning_rate": 4.1874215605748366e-06,
+      "loss": 0.079,
+      "step": 1926
+    },
+    {
+      "epoch": 1.4816522574447646,
+      "grad_norm": 0.8580770642845996,
+      "learning_rate": 4.181280708951015e-06,
+      "loss": 0.1018,
+      "step": 1928
+    },
+    {
+      "epoch": 1.4831892411143133,
+      "grad_norm": 0.8728686047506982,
+      "learning_rate": 4.175137676723726e-06,
+      "loss": 0.0964,
+      "step": 1930
+    },
+    {
+      "epoch": 1.4847262247838617,
+      "grad_norm": 0.7923177951847387,
+      "learning_rate": 4.168992483555231e-06,
+      "loss": 0.095,
+      "step": 1932
+    },
+    {
+      "epoch": 1.4862632084534102,
+      "grad_norm": 0.8120910936666785,
+      "learning_rate": 4.162845149114702e-06,
+      "loss": 0.0969,
+      "step": 1934
+    },
+    {
+      "epoch": 1.4878001921229587,
+      "grad_norm": 0.788706989115151,
+      "learning_rate": 4.15669569307817e-06,
+      "loss": 0.0885,
+      "step": 1936
+    },
+    {
+      "epoch": 1.4893371757925071,
+      "grad_norm": 0.830625147311433,
+      "learning_rate": 4.1505441351284526e-06,
+      "loss": 0.0972,
+      "step": 1938
+    },
+    {
+      "epoch": 1.4908741594620558,
+      "grad_norm": 0.8072145444813656,
+      "learning_rate": 4.144390494955098e-06,
+      "loss": 0.0887,
+      "step": 1940
+    },
+    {
+      "epoch": 1.4924111431316043,
+      "grad_norm": 0.8139524053166517,
+      "learning_rate": 4.138234792254317e-06,
+      "loss": 0.0939,
+      "step": 1942
+    },
+    {
+      "epoch": 1.4939481268011527,
+      "grad_norm": 0.7807279012822973,
+      "learning_rate": 4.132077046728924e-06,
+      "loss": 0.0804,
+      "step": 1944
+    },
+    {
+      "epoch": 1.4954851104707012,
+      "grad_norm": 0.7814387788672251,
+      "learning_rate": 4.125917278088269e-06,
+      "loss": 0.0924,
+      "step": 1946
+    },
+    {
+      "epoch": 1.4970220941402497,
+      "grad_norm": 0.884806793139695,
+      "learning_rate": 4.1197555060481836e-06,
+      "loss": 0.0977,
+      "step": 1948
+    },
+    {
+      "epoch": 1.4985590778097984,
+      "grad_norm": 0.8315758503149716,
+      "learning_rate": 4.1135917503309026e-06,
+      "loss": 0.0895,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5000960614793468,
+      "grad_norm": 0.8674422252417301,
+      "learning_rate": 4.107426030665016e-06,
+      "loss": 0.0882,
+      "step": 1952
+    },
+    {
+      "epoch": 1.5016330451488953,
+      "grad_norm": 0.8449293445755316,
+      "learning_rate": 4.101258366785402e-06,
+      "loss": 0.0967,
+      "step": 1954
+    },
+    {
+      "epoch": 1.5031700288184437,
+      "grad_norm": 0.8359354740918613,
+      "learning_rate": 4.095088778433156e-06,
+      "loss": 0.0998,
+      "step": 1956
+    },
+    {
+      "epoch": 1.5047070124879922,
+      "grad_norm": 0.8262628504512934,
+      "learning_rate": 4.088917285355536e-06,
+      "loss": 0.0945,
+      "step": 1958
+    },
+    {
+      "epoch": 1.506243996157541,
+      "grad_norm": 0.7479486694048768,
+      "learning_rate": 4.082743907305897e-06,
+      "loss": 0.0943,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5077809798270894,
+      "grad_norm": 0.8436782180522755,
+      "learning_rate": 4.076568664043625e-06,
+      "loss": 0.1004,
+      "step": 1962
+    },
+    {
+      "epoch": 1.5093179634966378,
+      "grad_norm": 0.7748345501977325,
+      "learning_rate": 4.0703915753340804e-06,
+      "loss": 0.0875,
+      "step": 1964
+    },
+    {
+      "epoch": 1.5108549471661865,
+      "grad_norm": 0.8382516980717638,
+      "learning_rate": 4.064212660948524e-06,
+      "loss": 0.094,
+      "step": 1966
+    },
+    {
+      "epoch": 1.5123919308357348,
+      "grad_norm": 0.7852255919020612,
+      "learning_rate": 4.058031940664067e-06,
+      "loss": 0.0912,
+      "step": 1968
+    },
+    {
+      "epoch": 1.5139289145052834,
+      "grad_norm": 0.8541412693322231,
+      "learning_rate": 4.051849434263595e-06,
+      "loss": 0.0968,
+      "step": 1970
+    },
+    {
+      "epoch": 1.515465898174832,
+      "grad_norm": 0.8000408705444431,
+      "learning_rate": 4.0456651615357155e-06,
+      "loss": 0.0874,
+      "step": 1972
+    },
+    {
+      "epoch": 1.5170028818443804,
+      "grad_norm": 0.7870384108772879,
+      "learning_rate": 4.039479142274686e-06,
+      "loss": 0.089,
+      "step": 1974
+    },
+    {
+      "epoch": 1.518539865513929,
+      "grad_norm": 0.820652722071049,
+      "learning_rate": 4.033291396280355e-06,
+      "loss": 0.0926,
+      "step": 1976
+    },
+    {
+      "epoch": 1.5200768491834773,
+      "grad_norm": 0.8154568505994646,
+      "learning_rate": 4.027101943358098e-06,
+      "loss": 0.0969,
+      "step": 1978
+    },
+    {
+      "epoch": 1.521613832853026,
+      "grad_norm": 0.8770625870074841,
+      "learning_rate": 4.020910803318756e-06,
+      "loss": 0.0993,
+      "step": 1980
+    },
+    {
+      "epoch": 1.5231508165225744,
+      "grad_norm": 0.6758733981469482,
+      "learning_rate": 4.014717995978565e-06,
+      "loss": 0.081,
+      "step": 1982
+    },
+    {
+      "epoch": 1.524687800192123,
+      "grad_norm": 0.8441728203990428,
+      "learning_rate": 4.008523541159104e-06,
+      "loss": 0.1009,
+      "step": 1984
+    },
+    {
+      "epoch": 1.5262247838616716,
+      "grad_norm": 0.8110301274459378,
+      "learning_rate": 4.002327458687218e-06,
+      "loss": 0.0872,
+      "step": 1986
+    },
+    {
+      "epoch": 1.5277617675312198,
+      "grad_norm": 0.8833546884887249,
+      "learning_rate": 3.996129768394969e-06,
+      "loss": 0.0982,
+      "step": 1988
+    },
+    {
+      "epoch": 1.5292987512007685,
+      "grad_norm": 0.8265167766685495,
+      "learning_rate": 3.989930490119561e-06,
+      "loss": 0.0967,
+      "step": 1990
+    },
+    {
+      "epoch": 1.530835734870317,
+      "grad_norm": 0.8707509834765963,
+      "learning_rate": 3.98372964370328e-06,
+      "loss": 0.0956,
+      "step": 1992
+    },
+    {
+      "epoch": 1.5323727185398655,
+      "grad_norm": 0.7763758628665975,
+      "learning_rate": 3.977527248993434e-06,
+      "loss": 0.0867,
+      "step": 1994
+    },
+    {
+      "epoch": 1.5339097022094141,
+      "grad_norm": 0.7789258378178834,
+      "learning_rate": 3.9713233258422855e-06,
+      "loss": 0.0878,
+      "step": 1996
+    },
+    {
+      "epoch": 1.5354466858789624,
+      "grad_norm": 0.7521669320980235,
+      "learning_rate": 3.965117894106988e-06,
+      "loss": 0.0888,
+      "step": 1998
+    },
+    {
+      "epoch": 1.536983669548511,
+      "grad_norm": 0.8412076854628533,
+      "learning_rate": 3.958910973649527e-06,
+      "loss": 0.0846,
+      "step": 2000
+    },
+    {
+      "epoch": 1.536983669548511,
+      "eval_loss": 0.1369212120771408,
+      "eval_runtime": 362.0502,
+      "eval_samples_per_second": 51.112,
+      "eval_steps_per_second": 6.391,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5385206532180595,
+      "grad_norm": 0.7902219827328915,
+      "learning_rate": 3.952702584336648e-06,
+      "loss": 0.0984,
+      "step": 2002
+    },
+    {
+      "epoch": 1.540057636887608,
+      "grad_norm": 0.7801431687989913,
+      "learning_rate": 3.9464927460398e-06,
+      "loss": 0.0979,
+      "step": 2004
+    },
+    {
+      "epoch": 1.5415946205571567,
+      "grad_norm": 0.8810086083716847,
+      "learning_rate": 3.940281478635074e-06,
+      "loss": 0.0916,
+      "step": 2006
+    },
+    {
+      "epoch": 1.5431316042267051,
+      "grad_norm": 0.8496438323924863,
+      "learning_rate": 3.934068802003129e-06,
+      "loss": 0.0943,
+      "step": 2008
+    },
+    {
+      "epoch": 1.5446685878962536,
+      "grad_norm": 0.8624052138866306,
+      "learning_rate": 3.927854736029138e-06,
+      "loss": 0.0941,
+      "step": 2010
+    },
+    {
+      "epoch": 1.546205571565802,
+      "grad_norm": 0.803691628281103,
+      "learning_rate": 3.921639300602719e-06,
+      "loss": 0.0964,
+      "step": 2012
+    },
+    {
+      "epoch": 1.5477425552353505,
+      "grad_norm": 0.7838512381219084,
+      "learning_rate": 3.915422515617876e-06,
+      "loss": 0.0877,
+      "step": 2014
+    },
+    {
+      "epoch": 1.5492795389048992,
+      "grad_norm": 0.8154032044869861,
+      "learning_rate": 3.90920440097293e-06,
+      "loss": 0.0898,
+      "step": 2016
+    },
+    {
+      "epoch": 1.5508165225744477,
+      "grad_norm": 0.7774689436181877,
+      "learning_rate": 3.902984976570459e-06,
+      "loss": 0.0942,
+      "step": 2018
+    },
+    {
+      "epoch": 1.5523535062439962,
+      "grad_norm": 0.7714965765671886,
+      "learning_rate": 3.896764262317232e-06,
+      "loss": 0.0882,
+      "step": 2020
+    },
+    {
+      "epoch": 1.5538904899135446,
+      "grad_norm": 0.8482244169935028,
+      "learning_rate": 3.890542278124151e-06,
+      "loss": 0.092,
+      "step": 2022
+    },
+    {
+      "epoch": 1.555427473583093,
+      "grad_norm": 0.8089857501256293,
+      "learning_rate": 3.884319043906175e-06,
+      "loss": 0.0885,
+      "step": 2024
+    },
+    {
+      "epoch": 1.5569644572526418,
+      "grad_norm": 0.7732451008458415,
+      "learning_rate": 3.878094579582271e-06,
+      "loss": 0.0852,
+      "step": 2026
+    },
+    {
+      "epoch": 1.5585014409221902,
+      "grad_norm": 0.8485314140958313,
+      "learning_rate": 3.871868905075339e-06,
+      "loss": 0.097,
+      "step": 2028
+    },
+    {
+      "epoch": 1.5600384245917387,
+      "grad_norm": 0.7295949942508037,
+      "learning_rate": 3.865642040312155e-06,
+      "loss": 0.0765,
+      "step": 2030
+    },
+    {
+      "epoch": 1.5615754082612874,
+      "grad_norm": 0.8109480864302396,
+      "learning_rate": 3.859414005223303e-06,
+      "loss": 0.0879,
+      "step": 2032
+    },
+    {
+      "epoch": 1.5631123919308356,
+      "grad_norm": 0.8308781255627026,
+      "learning_rate": 3.8531848197431155e-06,
+      "loss": 0.0957,
+      "step": 2034
+    },
+    {
+      "epoch": 1.5646493756003843,
+      "grad_norm": 0.8436684145142693,
+      "learning_rate": 3.846954503809602e-06,
+      "loss": 0.0941,
+      "step": 2036
+    },
+    {
+      "epoch": 1.5661863592699328,
+      "grad_norm": 0.7993009937419601,
+      "learning_rate": 3.840723077364396e-06,
+      "loss": 0.0907,
+      "step": 2038
+    },
+    {
+      "epoch": 1.5677233429394812,
+      "grad_norm": 0.7917608535459787,
+      "learning_rate": 3.834490560352682e-06,
+      "loss": 0.0946,
+      "step": 2040
+    },
+    {
+      "epoch": 1.56926032660903,
+      "grad_norm": 0.775435052389592,
+      "learning_rate": 3.828256972723137e-06,
+      "loss": 0.0866,
+      "step": 2042
+    },
+    {
+      "epoch": 1.5707973102785782,
+      "grad_norm": 0.8417542619925643,
+      "learning_rate": 3.822022334427863e-06,
+      "loss": 0.0956,
+      "step": 2044
+    },
+    {
+      "epoch": 1.5723342939481268,
+      "grad_norm": 0.7534098817985344,
+      "learning_rate": 3.815786665422326e-06,
+      "loss": 0.0907,
+      "step": 2046
+    },
+    {
+      "epoch": 1.5738712776176753,
+      "grad_norm": 0.7454391500744776,
+      "learning_rate": 3.8095499856652907e-06,
+      "loss": 0.0942,
+      "step": 2048
+    },
+    {
+      "epoch": 1.5754082612872238,
+      "grad_norm": 0.8059844013874078,
+      "learning_rate": 3.803312315118758e-06,
+      "loss": 0.0833,
+      "step": 2050
+    },
+    {
+      "epoch": 1.5769452449567725,
+      "grad_norm": 0.831173234796314,
+      "learning_rate": 3.7970736737478976e-06,
+      "loss": 0.0937,
+      "step": 2052
+    },
+    {
+      "epoch": 1.5784822286263207,
+      "grad_norm": 0.8319311638041307,
+      "learning_rate": 3.790834081520988e-06,
+      "loss": 0.0941,
+      "step": 2054
+    },
+    {
+      "epoch": 1.5800192122958694,
+      "grad_norm": 0.8017768441792715,
+      "learning_rate": 3.7845935584093535e-06,
+      "loss": 0.088,
+      "step": 2056
+    },
+    {
+      "epoch": 1.5815561959654179,
+      "grad_norm": 0.8146738006411376,
+      "learning_rate": 3.7783521243872945e-06,
+      "loss": 0.0929,
+      "step": 2058
+    },
+    {
+      "epoch": 1.5830931796349663,
+      "grad_norm": 0.7862299852946094,
+      "learning_rate": 3.7721097994320274e-06,
+      "loss": 0.088,
+      "step": 2060
+    },
+    {
+      "epoch": 1.584630163304515,
+      "grad_norm": 0.7644295350251104,
+      "learning_rate": 3.765866603523621e-06,
+      "loss": 0.0971,
+      "step": 2062
+    },
+    {
+      "epoch": 1.5861671469740632,
+      "grad_norm": 0.8145450168514264,
+      "learning_rate": 3.7596225566449326e-06,
+      "loss": 0.0902,
+      "step": 2064
+    },
+    {
+      "epoch": 1.587704130643612,
+      "grad_norm": 0.8269635685870652,
+      "learning_rate": 3.753377678781543e-06,
+      "loss": 0.0923,
+      "step": 2066
+    },
+    {
+      "epoch": 1.5892411143131604,
+      "grad_norm": 0.7490940289331177,
+      "learning_rate": 3.7471319899216904e-06,
+      "loss": 0.0779,
+      "step": 2068
+    },
+    {
+      "epoch": 1.5907780979827089,
+      "grad_norm": 0.7723884968412873,
+      "learning_rate": 3.7408855100562114e-06,
+      "loss": 0.0901,
+      "step": 2070
+    },
+    {
+      "epoch": 1.5923150816522575,
+      "grad_norm": 0.8319441486290872,
+      "learning_rate": 3.7346382591784747e-06,
+      "loss": 0.0956,
+      "step": 2072
+    },
+    {
+      "epoch": 1.593852065321806,
+      "grad_norm": 0.8690168846750308,
+      "learning_rate": 3.728390257284314e-06,
+      "loss": 0.0953,
+      "step": 2074
+    },
+    {
+      "epoch": 1.5953890489913545,
+      "grad_norm": 0.7909069572145911,
+      "learning_rate": 3.722141524371969e-06,
+      "loss": 0.086,
+      "step": 2076
+    },
+    {
+      "epoch": 1.596926032660903,
+      "grad_norm": 0.8062074972797826,
+      "learning_rate": 3.7158920804420203e-06,
+      "loss": 0.0892,
+      "step": 2078
+    },
+    {
+      "epoch": 1.5984630163304514,
+      "grad_norm": 0.7611492040694079,
+      "learning_rate": 3.7096419454973193e-06,
+      "loss": 0.0969,
+      "step": 2080
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8268228564890713,
+      "learning_rate": 3.703391139542937e-06,
+      "loss": 0.0927,
+      "step": 2082
+    },
+    {
+      "epoch": 1.6015369836695486,
+      "grad_norm": 0.7570539959602375,
+      "learning_rate": 3.6971396825860836e-06,
+      "loss": 0.0833,
+      "step": 2084
+    },
+    {
+      "epoch": 1.603073967339097,
+      "grad_norm": 0.797065245492781,
+      "learning_rate": 3.6908875946360597e-06,
+      "loss": 0.0895,
+      "step": 2086
+    },
+    {
+      "epoch": 1.6046109510086455,
+      "grad_norm": 0.8082111005522078,
+      "learning_rate": 3.68463489570418e-06,
+      "loss": 0.0891,
+      "step": 2088
+    },
+    {
+      "epoch": 1.606147934678194,
+      "grad_norm": 0.8233331752704219,
+      "learning_rate": 3.6783816058037215e-06,
+      "loss": 0.0948,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6076849183477426,
+      "grad_norm": 0.8349017726953393,
+      "learning_rate": 3.672127744949847e-06,
+      "loss": 0.0908,
+      "step": 2092
+    },
+    {
+      "epoch": 1.609221902017291,
+      "grad_norm": 0.7751955893816385,
+      "learning_rate": 3.6658733331595493e-06,
+      "loss": 0.0878,
+      "step": 2094
+    },
+    {
+      "epoch": 1.6107588856868396,
+      "grad_norm": 0.806749618964119,
+      "learning_rate": 3.6596183904515817e-06,
+      "loss": 0.0908,
+      "step": 2096
+    },
+    {
+      "epoch": 1.6122958693563882,
+      "grad_norm": 0.798020323683331,
+      "learning_rate": 3.6533629368464026e-06,
+      "loss": 0.085,
+      "step": 2098
+    },
+    {
+      "epoch": 1.6138328530259365,
+      "grad_norm": 0.8845913837287194,
+      "learning_rate": 3.6471069923661e-06,
+      "loss": 0.0923,
+      "step": 2100
+    },
+    {
+      "epoch": 1.6153698366954852,
+      "grad_norm": 0.7975937926027286,
+      "learning_rate": 3.6408505770343365e-06,
+      "loss": 0.0945,
+      "step": 2102
+    },
+    {
+      "epoch": 1.6169068203650336,
+      "grad_norm": 0.7977338905828745,
+      "learning_rate": 3.6345937108762803e-06,
+      "loss": 0.0886,
+      "step": 2104
+    },
+    {
+      "epoch": 1.618443804034582,
+      "grad_norm": 0.887493292191394,
+      "learning_rate": 3.628336413918541e-06,
+      "loss": 0.0909,
+      "step": 2106
+    },
+    {
+      "epoch": 1.6199807877041308,
+      "grad_norm": 0.7939745654491456,
+      "learning_rate": 3.622078706189112e-06,
+      "loss": 0.0929,
+      "step": 2108
+    },
+    {
+      "epoch": 1.621517771373679,
+      "grad_norm": 0.8183822731345303,
+      "learning_rate": 3.615820607717296e-06,
+      "loss": 0.0858,
+      "step": 2110
+    },
+    {
+      "epoch": 1.6230547550432277,
+      "grad_norm": 0.7807975084192331,
+      "learning_rate": 3.6095621385336505e-06,
+      "loss": 0.0957,
+      "step": 2112
+    },
+    {
+      "epoch": 1.6245917387127762,
+      "grad_norm": 0.8063009489440389,
+      "learning_rate": 3.6033033186699152e-06,
+      "loss": 0.0919,
+      "step": 2114
+    },
+    {
+      "epoch": 1.6261287223823246,
+      "grad_norm": 0.8112570440921344,
+      "learning_rate": 3.597044168158958e-06,
+      "loss": 0.0933,
+      "step": 2116
+    },
+    {
+      "epoch": 1.6276657060518733,
+      "grad_norm": 0.8331881631759025,
+      "learning_rate": 3.5907847070346997e-06,
+      "loss": 0.0927,
+      "step": 2118
+    },
+    {
+      "epoch": 1.6292026897214216,
+      "grad_norm": 0.7415764943928016,
+      "learning_rate": 3.5845249553320595e-06,
+      "loss": 0.0889,
+      "step": 2120
+    },
+    {
+      "epoch": 1.6307396733909703,
+      "grad_norm": 0.8003668361045585,
+      "learning_rate": 3.5782649330868817e-06,
+      "loss": 0.0971,
+      "step": 2122
+    },
+    {
+      "epoch": 1.6322766570605187,
+      "grad_norm": 0.7911658768866721,
+      "learning_rate": 3.5720046603358823e-06,
+      "loss": 0.0878,
+      "step": 2124
+    },
+    {
+      "epoch": 1.6338136407300672,
+      "grad_norm": 0.8045681908855453,
+      "learning_rate": 3.5657441571165754e-06,
+      "loss": 0.0874,
+      "step": 2126
+    },
+    {
+      "epoch": 1.6353506243996159,
+      "grad_norm": 0.893368127193528,
+      "learning_rate": 3.5594834434672142e-06,
+      "loss": 0.0904,
+      "step": 2128
+    },
+    {
+      "epoch": 1.6368876080691641,
+      "grad_norm": 0.7865118382233219,
+      "learning_rate": 3.553222539426724e-06,
+      "loss": 0.0841,
+      "step": 2130
+    },
+    {
+      "epoch": 1.6384245917387128,
+      "grad_norm": 0.7830147718405267,
+      "learning_rate": 3.546961465034641e-06,
+      "loss": 0.0936,
+      "step": 2132
+    },
+    {
+      "epoch": 1.6399615754082613,
+      "grad_norm": 0.8530754022949045,
+      "learning_rate": 3.5407002403310453e-06,
+      "loss": 0.094,
+      "step": 2134
+    },
+    {
+      "epoch": 1.6414985590778097,
+      "grad_norm": 0.7937086579061391,
+      "learning_rate": 3.5344388853565013e-06,
+      "loss": 0.0919,
+      "step": 2136
+    },
+    {
+      "epoch": 1.6430355427473584,
+      "grad_norm": 0.8490157093915195,
+      "learning_rate": 3.528177420151984e-06,
+      "loss": 0.0978,
+      "step": 2138
+    },
+    {
+      "epoch": 1.6445725264169067,
+      "grad_norm": 0.8109475329632754,
+      "learning_rate": 3.521915864758829e-06,
+      "loss": 0.0891,
+      "step": 2140
+    },
+    {
+      "epoch": 1.6461095100864553,
+      "grad_norm": 0.7936184951568639,
+      "learning_rate": 3.5156542392186554e-06,
+      "loss": 0.0916,
+      "step": 2142
+    },
+    {
+      "epoch": 1.6476464937560038,
+      "grad_norm": 0.916031078014482,
+      "learning_rate": 3.5093925635733084e-06,
+      "loss": 0.1031,
+      "step": 2144
+    },
+    {
+      "epoch": 1.6491834774255523,
+      "grad_norm": 0.7265572887246778,
+      "learning_rate": 3.503130857864792e-06,
+      "loss": 0.0872,
+      "step": 2146
+    },
+    {
+      "epoch": 1.650720461095101,
+      "grad_norm": 0.7277873762559617,
+      "learning_rate": 3.496869142135209e-06,
+      "loss": 0.086,
+      "step": 2148
+    },
+    {
+      "epoch": 1.6522574447646494,
+      "grad_norm": 0.7780905988859983,
+      "learning_rate": 3.4906074364266932e-06,
+      "loss": 0.0887,
+      "step": 2150
+    },
+    {
+      "epoch": 1.6537944284341979,
+      "grad_norm": 0.7713751406958493,
+      "learning_rate": 3.4843457607813445e-06,
+      "loss": 0.0871,
+      "step": 2152
+    },
+    {
+      "epoch": 1.6553314121037463,
+      "grad_norm": 0.8650115956133952,
+      "learning_rate": 3.478084135241171e-06,
+      "loss": 0.0948,
+      "step": 2154
+    },
+    {
+      "epoch": 1.6568683957732948,
+      "grad_norm": 0.8279315849380673,
+      "learning_rate": 3.4718225798480157e-06,
+      "loss": 0.0931,
+      "step": 2156
+    },
+    {
+      "epoch": 1.6584053794428435,
+      "grad_norm": 0.8057279660058387,
+      "learning_rate": 3.4655611146435003e-06,
+      "loss": 0.0906,
+      "step": 2158
+    },
+    {
+      "epoch": 1.659942363112392,
+      "grad_norm": 0.8002672116393091,
+      "learning_rate": 3.459299759668954e-06,
+      "loss": 0.0889,
+      "step": 2160
+    },
+    {
+      "epoch": 1.6614793467819404,
+      "grad_norm": 0.7562188109651065,
+      "learning_rate": 3.4530385349653597e-06,
+      "loss": 0.0846,
+      "step": 2162
+    },
+    {
+      "epoch": 1.663016330451489,
+      "grad_norm": 0.9071744451213971,
+      "learning_rate": 3.4467774605732763e-06,
+      "loss": 0.0975,
+      "step": 2164
+    },
+    {
+      "epoch": 1.6645533141210374,
+      "grad_norm": 0.7996156080383284,
+      "learning_rate": 3.440516556532787e-06,
+      "loss": 0.0924,
+      "step": 2166
+    },
+    {
+      "epoch": 1.666090297790586,
+      "grad_norm": 0.7950473833323731,
+      "learning_rate": 3.434255842883424e-06,
+      "loss": 0.0879,
+      "step": 2168
+    },
+    {
+      "epoch": 1.6676272814601345,
+      "grad_norm": 0.7899806697862958,
+      "learning_rate": 3.427995339664118e-06,
+      "loss": 0.0937,
+      "step": 2170
+    },
+    {
+      "epoch": 1.669164265129683,
+      "grad_norm": 0.7336053866465901,
+      "learning_rate": 3.421735066913118e-06,
+      "loss": 0.0835,
+      "step": 2172
+    },
+    {
+      "epoch": 1.6707012487992317,
+      "grad_norm": 0.7728149688439113,
+      "learning_rate": 3.415475044667942e-06,
+      "loss": 0.0927,
+      "step": 2174
+    },
+    {
+      "epoch": 1.67223823246878,
+      "grad_norm": 0.8305527642036483,
+      "learning_rate": 3.4092152929653e-06,
+      "loss": 0.0969,
+      "step": 2176
+    },
+    {
+      "epoch": 1.6737752161383286,
+      "grad_norm": 0.773715213471355,
+      "learning_rate": 3.4029558318410426e-06,
+      "loss": 0.0937,
+      "step": 2178
+    },
+    {
+      "epoch": 1.675312199807877,
+      "grad_norm": 0.8059092352832676,
+      "learning_rate": 3.396696681330085e-06,
+      "loss": 0.0863,
+      "step": 2180
+    },
+    {
+      "epoch": 1.6768491834774255,
+      "grad_norm": 0.803662636009022,
+      "learning_rate": 3.3904378614663507e-06,
+      "loss": 0.084,
+      "step": 2182
+    },
+    {
+      "epoch": 1.6783861671469742,
+      "grad_norm": 0.837705389217021,
+      "learning_rate": 3.3841793922827035e-06,
+      "loss": 0.0882,
+      "step": 2184
+    },
+    {
+      "epoch": 1.6799231508165224,
+      "grad_norm": 0.7922764494195353,
+      "learning_rate": 3.3779212938108883e-06,
+      "loss": 0.0827,
+      "step": 2186
+    },
+    {
+      "epoch": 1.6814601344860711,
+      "grad_norm": 0.7860318255345747,
+      "learning_rate": 3.3716635860814593e-06,
+      "loss": 0.0876,
+      "step": 2188
+    },
+    {
+      "epoch": 1.6829971181556196,
+      "grad_norm": 0.8066504756055669,
+      "learning_rate": 3.365406289123721e-06,
+      "loss": 0.0938,
+      "step": 2190
+    },
+    {
+      "epoch": 1.684534101825168,
+      "grad_norm": 0.7921416532988432,
+      "learning_rate": 3.3591494229656634e-06,
+      "loss": 0.0934,
+      "step": 2192
+    },
+    {
+      "epoch": 1.6860710854947167,
+      "grad_norm": 0.745529905487878,
+      "learning_rate": 3.3528930076339002e-06,
+      "loss": 0.0813,
+      "step": 2194
+    },
+    {
+      "epoch": 1.687608069164265,
+      "grad_norm": 0.832234842686226,
+      "learning_rate": 3.346637063153598e-06,
+      "loss": 0.09,
+      "step": 2196
+    },
+    {
+      "epoch": 1.6891450528338137,
+      "grad_norm": 0.787715663482219,
+      "learning_rate": 3.3403816095484177e-06,
+      "loss": 0.0883,
+      "step": 2198
+    },
+    {
+      "epoch": 1.6906820365033621,
+      "grad_norm": 0.8626434473168217,
+      "learning_rate": 3.3341266668404514e-06,
+      "loss": 0.1017,
+      "step": 2200
+    },
+    {
+      "epoch": 1.6922190201729106,
+      "grad_norm": 0.804548679573353,
+      "learning_rate": 3.3278722550501534e-06,
+      "loss": 0.0894,
+      "step": 2202
+    },
+    {
+      "epoch": 1.6937560038424593,
+      "grad_norm": 0.7757347097137762,
+      "learning_rate": 3.3216183941962793e-06,
+      "loss": 0.0841,
+      "step": 2204
+    },
+    {
+      "epoch": 1.6952929875120075,
+      "grad_norm": 0.7833610959261708,
+      "learning_rate": 3.3153651042958196e-06,
+      "loss": 0.096,
+      "step": 2206
+    },
+    {
+      "epoch": 1.6968299711815562,
+      "grad_norm": 0.7848923969699126,
+      "learning_rate": 3.309112405363941e-06,
+      "loss": 0.0952,
+      "step": 2208
+    },
+    {
+      "epoch": 1.6983669548511047,
+      "grad_norm": 0.8052464321733062,
+      "learning_rate": 3.302860317413917e-06,
+      "loss": 0.0915,
+      "step": 2210
+    },
+    {
+      "epoch": 1.6999039385206531,
+      "grad_norm": 0.8512154630818453,
+      "learning_rate": 3.2966088604570648e-06,
+      "loss": 0.0876,
+      "step": 2212
+    },
+    {
+      "epoch": 1.7014409221902018,
+      "grad_norm": 0.7790903398816384,
+      "learning_rate": 3.2903580545026797e-06,
+      "loss": 0.087,
+      "step": 2214
+    },
+    {
+      "epoch": 1.7029779058597503,
+      "grad_norm": 0.8376240131132313,
+      "learning_rate": 3.28410791955798e-06,
+      "loss": 0.0967,
+      "step": 2216
+    },
+    {
+      "epoch": 1.7045148895292987,
+      "grad_norm": 0.7967627035735608,
+      "learning_rate": 3.2778584756280307e-06,
+      "loss": 0.0915,
+      "step": 2218
+    },
+    {
+      "epoch": 1.7060518731988472,
+      "grad_norm": 0.7702278491400715,
+      "learning_rate": 3.271609742715687e-06,
+      "loss": 0.0862,
+      "step": 2220
+    },
+    {
+      "epoch": 1.7075888568683957,
+      "grad_norm": 0.7544840615892947,
+      "learning_rate": 3.265361740821525e-06,
+      "loss": 0.0833,
+      "step": 2222
+    },
+    {
+      "epoch": 1.7091258405379444,
+      "grad_norm": 0.7621768245734015,
+      "learning_rate": 3.2591144899437885e-06,
+      "loss": 0.0871,
+      "step": 2224
+    },
+    {
+      "epoch": 1.7106628242074928,
+      "grad_norm": 0.8831931871572729,
+      "learning_rate": 3.25286801007831e-06,
+      "loss": 0.0882,
+      "step": 2226
+    },
+    {
+      "epoch": 1.7121998078770413,
+      "grad_norm": 0.8197620745142645,
+      "learning_rate": 3.246622321218458e-06,
+      "loss": 0.0928,
+      "step": 2228
+    },
+    {
+      "epoch": 1.71373679154659,
+      "grad_norm": 0.77604756798234,
+      "learning_rate": 3.2403774433550673e-06,
+      "loss": 0.0832,
+      "step": 2230
+    },
+    {
+      "epoch": 1.7152737752161382,
+      "grad_norm": 0.8842522339003976,
+      "learning_rate": 3.2341333964763795e-06,
+      "loss": 0.0988,
+      "step": 2232
+    },
+    {
+      "epoch": 1.716810758885687,
+      "grad_norm": 0.7768312377818398,
+      "learning_rate": 3.2278902005679734e-06,
+      "loss": 0.0882,
+      "step": 2234
+    },
+    {
+      "epoch": 1.7183477425552354,
+      "grad_norm": 0.8327531068216764,
+      "learning_rate": 3.2216478756127067e-06,
+      "loss": 0.085,
+      "step": 2236
+    },
+    {
+      "epoch": 1.7198847262247838,
+      "grad_norm": 0.8186258517055159,
+      "learning_rate": 3.215406441590646e-06,
+      "loss": 0.0885,
+      "step": 2238
+    },
+    {
+      "epoch": 1.7214217098943325,
+      "grad_norm": 0.6874479376398842,
+      "learning_rate": 3.209165918479012e-06,
+      "loss": 0.0781,
+      "step": 2240
+    },
+    {
+      "epoch": 1.7229586935638808,
+      "grad_norm": 0.8115322816676634,
+      "learning_rate": 3.202926326252103e-06,
+      "loss": 0.0969,
+      "step": 2242
+    },
+    {
+      "epoch": 1.7244956772334294,
+      "grad_norm": 0.8146826671168621,
+      "learning_rate": 3.1966876848812434e-06,
+      "loss": 0.089,
+      "step": 2244
+    },
+    {
+      "epoch": 1.726032660902978,
+      "grad_norm": 0.8248049130832095,
+      "learning_rate": 3.1904500143347092e-06,
+      "loss": 0.0882,
+      "step": 2246
+    },
+    {
+      "epoch": 1.7275696445725264,
+      "grad_norm": 0.8015905899655976,
+      "learning_rate": 3.184213334577675e-06,
+      "loss": 0.0914,
+      "step": 2248
+    },
+    {
+      "epoch": 1.729106628242075,
+      "grad_norm": 0.8160931505603131,
+      "learning_rate": 3.1779776655721374e-06,
+      "loss": 0.0898,
+      "step": 2250
+    },
+    {
+      "epoch": 1.7306436119116233,
+      "grad_norm": 0.8170663856878996,
+      "learning_rate": 3.1717430272768637e-06,
+      "loss": 0.0925,
+      "step": 2252
+    },
+    {
+      "epoch": 1.732180595581172,
+      "grad_norm": 0.8051970712819015,
+      "learning_rate": 3.1655094396473175e-06,
+      "loss": 0.0929,
+      "step": 2254
+    },
+    {
+      "epoch": 1.7337175792507205,
+      "grad_norm": 0.8967076051389512,
+      "learning_rate": 3.1592769226356045e-06,
+      "loss": 0.0931,
+      "step": 2256
+    },
+    {
+      "epoch": 1.735254562920269,
+      "grad_norm": 0.8217397870462823,
+      "learning_rate": 3.153045496190398e-06,
+      "loss": 0.0885,
+      "step": 2258
+    },
+    {
+      "epoch": 1.7367915465898176,
+      "grad_norm": 0.7584555916208618,
+      "learning_rate": 3.1468151802568857e-06,
+      "loss": 0.0868,
+      "step": 2260
+    },
+    {
+      "epoch": 1.7383285302593658,
+      "grad_norm": 0.8377345355524379,
+      "learning_rate": 3.1405859947766965e-06,
+      "loss": 0.084,
+      "step": 2262
+    },
+    {
+      "epoch": 1.7398655139289145,
+      "grad_norm": 0.7974250660258231,
+      "learning_rate": 3.1343579596878455e-06,
+      "loss": 0.0896,
+      "step": 2264
+    },
+    {
+      "epoch": 1.741402497598463,
+      "grad_norm": 0.8423863324362034,
+      "learning_rate": 3.128131094924661e-06,
+      "loss": 0.0914,
+      "step": 2266
+    },
+    {
+      "epoch": 1.7429394812680115,
+      "grad_norm": 0.7998587118371564,
+      "learning_rate": 3.12190542041773e-06,
+      "loss": 0.0877,
+      "step": 2268
+    },
+    {
+      "epoch": 1.7444764649375601,
+      "grad_norm": 0.8420892166338229,
+      "learning_rate": 3.1156809560938246e-06,
+      "loss": 0.0892,
+      "step": 2270
+    },
+    {
+      "epoch": 1.7460134486071084,
+      "grad_norm": 0.7811169390028642,
+      "learning_rate": 3.1094577218758497e-06,
+      "loss": 0.0826,
+      "step": 2272
+    },
+    {
+      "epoch": 1.747550432276657,
+      "grad_norm": 0.856105762521616,
+      "learning_rate": 3.103235737682768e-06,
+      "loss": 0.0925,
+      "step": 2274
+    },
+    {
+      "epoch": 1.7490874159462055,
+      "grad_norm": 0.8434878009386567,
+      "learning_rate": 3.0970150234295416e-06,
+      "loss": 0.0895,
+      "step": 2276
+    },
+    {
+      "epoch": 1.750624399615754,
+      "grad_norm": 0.7912672991129802,
+      "learning_rate": 3.09079559902707e-06,
+      "loss": 0.0844,
+      "step": 2278
+    },
+    {
+      "epoch": 1.7521613832853027,
+      "grad_norm": 0.8093805440566595,
+      "learning_rate": 3.0845774843821242e-06,
+      "loss": 0.0857,
+      "step": 2280
+    },
+    {
+      "epoch": 1.7536983669548512,
+      "grad_norm": 0.7841327968506241,
+      "learning_rate": 3.0783606993972816e-06,
+      "loss": 0.0824,
+      "step": 2282
+    },
+    {
+      "epoch": 1.7552353506243996,
+      "grad_norm": 0.8784822733296287,
+      "learning_rate": 3.072145263970863e-06,
+      "loss": 0.0896,
+      "step": 2284
+    },
+    {
+      "epoch": 1.756772334293948,
+      "grad_norm": 0.785311512287812,
+      "learning_rate": 3.0659311979968707e-06,
+      "loss": 0.0787,
+      "step": 2286
+    },
+    {
+      "epoch": 1.7583093179634965,
+      "grad_norm": 0.8489973269201797,
+      "learning_rate": 3.059718521364926e-06,
+      "loss": 0.0965,
+      "step": 2288
+    },
+    {
+      "epoch": 1.7598463016330452,
+      "grad_norm": 0.8687972454705519,
+      "learning_rate": 3.0535072539602004e-06,
+      "loss": 0.0875,
+      "step": 2290
+    },
+    {
+      "epoch": 1.7613832853025937,
+      "grad_norm": 0.8170562404470127,
+      "learning_rate": 3.0472974156633535e-06,
+      "loss": 0.0947,
+      "step": 2292
+    },
+    {
+      "epoch": 1.7629202689721422,
+      "grad_norm": 0.7542361214197597,
+      "learning_rate": 3.0410890263504736e-06,
+      "loss": 0.0838,
+      "step": 2294
+    },
+    {
+      "epoch": 1.7644572526416908,
+      "grad_norm": 0.8228131888155036,
+      "learning_rate": 3.0348821058930117e-06,
+      "loss": 0.0956,
+      "step": 2296
+    },
+    {
+      "epoch": 1.765994236311239,
+      "grad_norm": 0.8281839285188529,
+      "learning_rate": 3.0286766741577156e-06,
+      "loss": 0.0897,
+      "step": 2298
+    },
+    {
+      "epoch": 1.7675312199807878,
+      "grad_norm": 0.7587083990715107,
+      "learning_rate": 3.022472751006566e-06,
+      "loss": 0.0805,
+      "step": 2300
+    },
+    {
+      "epoch": 1.7690682036503362,
+      "grad_norm": 0.8251420202843136,
+      "learning_rate": 3.0162703562967197e-06,
+      "loss": 0.0845,
+      "step": 2302
+    },
+    {
+      "epoch": 1.7706051873198847,
+      "grad_norm": 0.8065903172521137,
+      "learning_rate": 3.01006950988044e-06,
+      "loss": 0.0891,
+      "step": 2304
+    },
+    {
+      "epoch": 1.7721421709894334,
+      "grad_norm": 0.8124974824000051,
+      "learning_rate": 3.0038702316050317e-06,
+      "loss": 0.0896,
+      "step": 2306
+    },
+    {
+      "epoch": 1.7736791546589816,
+      "grad_norm": 0.7529764285672882,
+      "learning_rate": 2.997672541312782e-06,
+      "loss": 0.0776,
+      "step": 2308
+    },
+    {
+      "epoch": 1.7752161383285303,
+      "grad_norm": 0.8091489227633613,
+      "learning_rate": 2.9914764588408966e-06,
+      "loss": 0.0922,
+      "step": 2310
+    },
+    {
+      "epoch": 1.7767531219980788,
+      "grad_norm": 0.8938294981287168,
+      "learning_rate": 2.985282004021435e-06,
+      "loss": 0.0924,
+      "step": 2312
+    },
+    {
+      "epoch": 1.7782901056676272,
+      "grad_norm": 0.9013544600990553,
+      "learning_rate": 2.979089196681245e-06,
+      "loss": 0.091,
+      "step": 2314
+    },
+    {
+      "epoch": 1.779827089337176,
+      "grad_norm": 0.8597516613442064,
+      "learning_rate": 2.9728980566419016e-06,
+      "loss": 0.0847,
+      "step": 2316
+    },
+    {
+      "epoch": 1.7813640730067242,
+      "grad_norm": 0.8090979220931821,
+      "learning_rate": 2.966708603719645e-06,
+      "loss": 0.091,
+      "step": 2318
+    },
+    {
+      "epoch": 1.7829010566762729,
+      "grad_norm": 0.7916631388831536,
+      "learning_rate": 2.960520857725314e-06,
+      "loss": 0.0867,
+      "step": 2320
+    },
+    {
+      "epoch": 1.7844380403458213,
+      "grad_norm": 0.7616856425912224,
+      "learning_rate": 2.954334838464285e-06,
+      "loss": 0.0892,
+      "step": 2322
+    },
+    {
+      "epoch": 1.7859750240153698,
+      "grad_norm": 0.7512880043816689,
+      "learning_rate": 2.948150565736404e-06,
+      "loss": 0.0825,
+      "step": 2324
+    },
+    {
+      "epoch": 1.7875120076849185,
+      "grad_norm": 0.7903862314768638,
+      "learning_rate": 2.9419680593359335e-06,
+      "loss": 0.0815,
+      "step": 2326
+    },
+    {
+      "epoch": 1.7890489913544667,
+      "grad_norm": 0.826491887413533,
+      "learning_rate": 2.9357873390514757e-06,
+      "loss": 0.0868,
+      "step": 2328
+    },
+    {
+      "epoch": 1.7905859750240154,
+      "grad_norm": 0.8078642556595113,
+      "learning_rate": 2.929608424665921e-06,
+      "loss": 0.0945,
+      "step": 2330
+    },
+    {
+      "epoch": 1.7921229586935639,
+      "grad_norm": 0.8097155781369342,
+      "learning_rate": 2.9234313359563744e-06,
+      "loss": 0.0938,
+      "step": 2332
+    },
+    {
+      "epoch": 1.7936599423631123,
+      "grad_norm": 0.8150186954176805,
+      "learning_rate": 2.9172560926941037e-06,
+      "loss": 0.0828,
+      "step": 2334
+    },
+    {
+      "epoch": 1.795196926032661,
+      "grad_norm": 0.7813243625319553,
+      "learning_rate": 2.9110827146444643e-06,
+      "loss": 0.0858,
+      "step": 2336
+    },
+    {
+      "epoch": 1.7967339097022093,
+      "grad_norm": 0.7791024554992239,
+      "learning_rate": 2.904911221566845e-06,
+      "loss": 0.0844,
+      "step": 2338
+    },
+    {
+      "epoch": 1.798270893371758,
+      "grad_norm": 0.8124678993462557,
+      "learning_rate": 2.898741633214598e-06,
+      "loss": 0.0893,
+      "step": 2340
+    },
+    {
+      "epoch": 1.7998078770413064,
+      "grad_norm": 0.829294783575689,
+      "learning_rate": 2.8925739693349833e-06,
+      "loss": 0.0861,
+      "step": 2342
+    },
+    {
+      "epoch": 1.8013448607108549,
+      "grad_norm": 0.7957217580914165,
+      "learning_rate": 2.886408249669098e-06,
+      "loss": 0.0836,
+      "step": 2344
+    },
+    {
+      "epoch": 1.8028818443804036,
+      "grad_norm": 0.8631788141169373,
+      "learning_rate": 2.880244493951818e-06,
+      "loss": 0.1018,
+      "step": 2346
+    },
+    {
+      "epoch": 1.804418828049952,
+      "grad_norm": 0.6989805954411332,
+      "learning_rate": 2.8740827219117302e-06,
+      "loss": 0.0827,
+      "step": 2348
+    },
+    {
+      "epoch": 1.8059558117195005,
+      "grad_norm": 0.8472076650897804,
+      "learning_rate": 2.867922953271077e-06,
+      "loss": 0.0912,
+      "step": 2350
+    },
+    {
+      "epoch": 1.807492795389049,
+      "grad_norm": 0.8663805858998409,
+      "learning_rate": 2.861765207745683e-06,
+      "loss": 0.094,
+      "step": 2352
+    },
+    {
+      "epoch": 1.8090297790585974,
+      "grad_norm": 0.7756793455854972,
+      "learning_rate": 2.8556095050449032e-06,
+      "loss": 0.086,
+      "step": 2354
+    },
+    {
+      "epoch": 1.810566762728146,
+      "grad_norm": 0.8038188721900218,
+      "learning_rate": 2.8494558648715473e-06,
+      "loss": 0.0895,
+      "step": 2356
+    },
+    {
+      "epoch": 1.8121037463976946,
+      "grad_norm": 0.7788550764950732,
+      "learning_rate": 2.8433043069218307e-06,
+      "loss": 0.0912,
+      "step": 2358
+    },
+    {
+      "epoch": 1.813640730067243,
+      "grad_norm": 0.8331357355337144,
+      "learning_rate": 2.8371548508852977e-06,
+      "loss": 0.0982,
+      "step": 2360
+    },
+    {
+      "epoch": 1.8151777137367917,
+      "grad_norm": 0.8118854152234461,
+      "learning_rate": 2.8310075164447696e-06,
+      "loss": 0.0925,
+      "step": 2362
+    },
+    {
+      "epoch": 1.81671469740634,
+      "grad_norm": 0.8017256660340538,
+      "learning_rate": 2.824862323276273e-06,
+      "loss": 0.0855,
+      "step": 2364
+    },
+    {
+      "epoch": 1.8182516810758886,
+      "grad_norm": 0.7800752551508296,
+      "learning_rate": 2.8187192910489856e-06,
+      "loss": 0.0834,
+      "step": 2366
+    },
+    {
+      "epoch": 1.819788664745437,
+      "grad_norm": 0.8380624718020749,
+      "learning_rate": 2.812578439425164e-06,
+      "loss": 0.0945,
+      "step": 2368
+    },
+    {
+      "epoch": 1.8213256484149856,
+      "grad_norm": 0.8276945914691816,
+      "learning_rate": 2.806439788060088e-06,
+      "loss": 0.0889,
+      "step": 2370
+    },
+    {
+      "epoch": 1.8228626320845343,
+      "grad_norm": 0.7985257731425042,
+      "learning_rate": 2.8003033566019922e-06,
+      "loss": 0.0841,
+      "step": 2372
+    },
+    {
+      "epoch": 1.8243996157540825,
+      "grad_norm": 0.7637963888810898,
+      "learning_rate": 2.7941691646920105e-06,
+      "loss": 0.0751,
+      "step": 2374
+    },
+    {
+      "epoch": 1.8259365994236312,
+      "grad_norm": 0.8413312901210165,
+      "learning_rate": 2.7880372319641042e-06,
+      "loss": 0.089,
+      "step": 2376
+    },
+    {
+      "epoch": 1.8274735830931796,
+      "grad_norm": 0.8586650363460536,
+      "learning_rate": 2.781907578045005e-06,
+      "loss": 0.0887,
+      "step": 2378
+    },
+    {
+      "epoch": 1.829010566762728,
+      "grad_norm": 0.8360563017968606,
+      "learning_rate": 2.77578022255415e-06,
+      "loss": 0.0939,
+      "step": 2380
+    },
+    {
+      "epoch": 1.8305475504322768,
+      "grad_norm": 0.877204652341484,
+      "learning_rate": 2.769655185103623e-06,
+      "loss": 0.0893,
+      "step": 2382
+    },
+    {
+      "epoch": 1.832084534101825,
+      "grad_norm": 0.8297837842829701,
+      "learning_rate": 2.7635324852980843e-06,
+      "loss": 0.0907,
+      "step": 2384
+    },
+    {
+      "epoch": 1.8336215177713737,
+      "grad_norm": 0.8431709281215559,
+      "learning_rate": 2.7574121427347133e-06,
+      "loss": 0.0982,
+      "step": 2386
+    },
+    {
+      "epoch": 1.8351585014409222,
+      "grad_norm": 0.7604258311920689,
+      "learning_rate": 2.751294177003143e-06,
+      "loss": 0.0789,
+      "step": 2388
+    },
+    {
+      "epoch": 1.8366954851104706,
+      "grad_norm": 0.7739652641897983,
+      "learning_rate": 2.745178607685405e-06,
+      "loss": 0.0928,
+      "step": 2390
+    },
+    {
+      "epoch": 1.8382324687800193,
+      "grad_norm": 0.8167833653726546,
+      "learning_rate": 2.7390654543558534e-06,
+      "loss": 0.0842,
+      "step": 2392
+    },
+    {
+      "epoch": 1.8397694524495676,
+      "grad_norm": 0.7908210162723227,
+      "learning_rate": 2.7329547365811104e-06,
+      "loss": 0.0843,
+      "step": 2394
+    },
+    {
+      "epoch": 1.8413064361191163,
+      "grad_norm": 0.8201467727248781,
+      "learning_rate": 2.726846473920006e-06,
+      "loss": 0.0919,
+      "step": 2396
+    },
+    {
+      "epoch": 1.8428434197886647,
+      "grad_norm": 0.8484929622858332,
+      "learning_rate": 2.7207406859235117e-06,
+      "loss": 0.0817,
+      "step": 2398
+    },
+    {
+      "epoch": 1.8443804034582132,
+      "grad_norm": 0.7641741013611666,
+      "learning_rate": 2.714637392134675e-06,
+      "loss": 0.0875,
+      "step": 2400
+    },
+    {
+      "epoch": 1.8459173871277619,
+      "grad_norm": 0.8726969374519596,
+      "learning_rate": 2.708536612088561e-06,
+      "loss": 0.0883,
+      "step": 2402
+    },
+    {
+      "epoch": 1.8474543707973101,
+      "grad_norm": 0.7568052808653818,
+      "learning_rate": 2.702438365312191e-06,
+      "loss": 0.0847,
+      "step": 2404
+    },
+    {
+      "epoch": 1.8489913544668588,
+      "grad_norm": 0.774661999809938,
+      "learning_rate": 2.696342671324478e-06,
+      "loss": 0.0912,
+      "step": 2406
+    },
+    {
+      "epoch": 1.8505283381364073,
+      "grad_norm": 0.8750231519779115,
+      "learning_rate": 2.6902495496361613e-06,
+      "loss": 0.094,
+      "step": 2408
+    },
+    {
+      "epoch": 1.8520653218059557,
+      "grad_norm": 0.7904137960913173,
+      "learning_rate": 2.6841590197497476e-06,
+      "loss": 0.0858,
+      "step": 2410
+    },
+    {
+      "epoch": 1.8536023054755044,
+      "grad_norm": 0.7711313526212195,
+      "learning_rate": 2.678071101159451e-06,
+      "loss": 0.0825,
+      "step": 2412
+    },
+    {
+      "epoch": 1.8551392891450529,
+      "grad_norm": 0.8221962336381063,
+      "learning_rate": 2.6719858133511257e-06,
+      "loss": 0.0906,
+      "step": 2414
+    },
+    {
+      "epoch": 1.8566762728146013,
+      "grad_norm": 0.8337437421475409,
+      "learning_rate": 2.665903175802204e-06,
+      "loss": 0.0868,
+      "step": 2416
+    },
+    {
+      "epoch": 1.8582132564841498,
+      "grad_norm": 0.8378451337853939,
+      "learning_rate": 2.6598232079816353e-06,
+      "loss": 0.0933,
+      "step": 2418
+    },
+    {
+      "epoch": 1.8597502401536983,
+      "grad_norm": 0.7639730128714044,
+      "learning_rate": 2.6537459293498277e-06,
+      "loss": 0.0846,
+      "step": 2420
+    },
+    {
+      "epoch": 1.861287223823247,
+      "grad_norm": 0.7830022284803336,
+      "learning_rate": 2.6476713593585783e-06,
+      "loss": 0.0863,
+      "step": 2422
+    },
+    {
+      "epoch": 1.8628242074927954,
+      "grad_norm": 0.816812769548626,
+      "learning_rate": 2.641599517451016e-06,
+      "loss": 0.0937,
+      "step": 2424
+    },
+    {
+      "epoch": 1.864361191162344,
+      "grad_norm": 0.8742317390360426,
+      "learning_rate": 2.6355304230615356e-06,
+      "loss": 0.0975,
+      "step": 2426
+    },
+    {
+      "epoch": 1.8658981748318926,
+      "grad_norm": 0.8026788777916881,
+      "learning_rate": 2.629464095615743e-06,
+      "loss": 0.0848,
+      "step": 2428
+    },
+    {
+      "epoch": 1.8674351585014408,
+      "grad_norm": 0.8437565172328019,
+      "learning_rate": 2.623400554530382e-06,
+      "loss": 0.0947,
+      "step": 2430
+    },
+    {
+      "epoch": 1.8689721421709895,
+      "grad_norm": 0.801208290733419,
+      "learning_rate": 2.6173398192132812e-06,
+      "loss": 0.0899,
+      "step": 2432
+    },
+    {
+      "epoch": 1.870509125840538,
+      "grad_norm": 0.7932315598508702,
+      "learning_rate": 2.6112819090632854e-06,
+      "loss": 0.0895,
+      "step": 2434
+    },
+    {
+      "epoch": 1.8720461095100864,
+      "grad_norm": 0.9042900524043032,
+      "learning_rate": 2.6052268434702042e-06,
+      "loss": 0.0946,
+      "step": 2436
+    },
+    {
+      "epoch": 1.8735830931796351,
+      "grad_norm": 0.7894581411758911,
+      "learning_rate": 2.599174641814734e-06,
+      "loss": 0.0867,
+      "step": 2438
+    },
+    {
+      "epoch": 1.8751200768491834,
+      "grad_norm": 0.7854020543576818,
+      "learning_rate": 2.59312532346841e-06,
+      "loss": 0.0911,
+      "step": 2440
+    },
+    {
+      "epoch": 1.876657060518732,
+      "grad_norm": 0.815693549418664,
+      "learning_rate": 2.5870789077935335e-06,
+      "loss": 0.0895,
+      "step": 2442
+    },
+    {
+      "epoch": 1.8781940441882805,
+      "grad_norm": 0.7887109470073577,
+      "learning_rate": 2.5810354141431226e-06,
+      "loss": 0.0903,
+      "step": 2444
+    },
+    {
+      "epoch": 1.879731027857829,
+      "grad_norm": 0.7896777569740607,
+      "learning_rate": 2.5749948618608356e-06,
+      "loss": 0.0809,
+      "step": 2446
+    },
+    {
+      "epoch": 1.8812680115273777,
+      "grad_norm": 0.7615729562820233,
+      "learning_rate": 2.5689572702809203e-06,
+      "loss": 0.0838,
+      "step": 2448
+    },
+    {
+      "epoch": 1.882804995196926,
+      "grad_norm": 0.8582942386637096,
+      "learning_rate": 2.562922658728145e-06,
+      "loss": 0.09,
+      "step": 2450
+    },
+    {
+      "epoch": 1.8843419788664746,
+      "grad_norm": 0.7611707316427125,
+      "learning_rate": 2.556891046517745e-06,
+      "loss": 0.0775,
+      "step": 2452
+    },
+    {
+      "epoch": 1.885878962536023,
+      "grad_norm": 0.7455166273658402,
+      "learning_rate": 2.5508624529553496e-06,
+      "loss": 0.0866,
+      "step": 2454
+    },
+    {
+      "epoch": 1.8874159462055715,
+      "grad_norm": 0.8572197727943554,
+      "learning_rate": 2.5448368973369295e-06,
+      "loss": 0.0867,
+      "step": 2456
+    },
+    {
+      "epoch": 1.8889529298751202,
+      "grad_norm": 0.8079812248707947,
+      "learning_rate": 2.5388143989487295e-06,
+      "loss": 0.09,
+      "step": 2458
+    },
+    {
+      "epoch": 1.8904899135446684,
+      "grad_norm": 0.8079920947645508,
+      "learning_rate": 2.5327949770672125e-06,
+      "loss": 0.0872,
+      "step": 2460
+    },
+    {
+      "epoch": 1.8920268972142171,
+      "grad_norm": 0.8329864584858913,
+      "learning_rate": 2.5267786509589907e-06,
+      "loss": 0.0892,
+      "step": 2462
+    },
+    {
+      "epoch": 1.8935638808837656,
+      "grad_norm": 0.8044295715601575,
+      "learning_rate": 2.52076543988077e-06,
+      "loss": 0.0874,
+      "step": 2464
+    },
+    {
+      "epoch": 1.895100864553314,
+      "grad_norm": 0.7566977212719446,
+      "learning_rate": 2.5147553630792827e-06,
+      "loss": 0.079,
+      "step": 2466
+    },
+    {
+      "epoch": 1.8966378482228627,
+      "grad_norm": 0.7703458660050836,
+      "learning_rate": 2.5087484397912354e-06,
+      "loss": 0.0803,
+      "step": 2468
+    },
+    {
+      "epoch": 1.898174831892411,
+      "grad_norm": 0.7472342418402756,
+      "learning_rate": 2.5027446892432335e-06,
+      "loss": 0.0831,
+      "step": 2470
+    },
+    {
+      "epoch": 1.8997118155619597,
+      "grad_norm": 0.7860659793588969,
+      "learning_rate": 2.496744130651731e-06,
+      "loss": 0.0861,
+      "step": 2472
+    },
+    {
+      "epoch": 1.9012487992315081,
+      "grad_norm": 0.7782192620932303,
+      "learning_rate": 2.4907467832229655e-06,
+      "loss": 0.0904,
+      "step": 2474
+    },
+    {
+      "epoch": 1.9027857829010566,
+      "grad_norm": 0.7976874165660952,
+      "learning_rate": 2.4847526661528974e-06,
+      "loss": 0.0868,
+      "step": 2476
+    },
+    {
+      "epoch": 1.9043227665706053,
+      "grad_norm": 0.8130144869116065,
+      "learning_rate": 2.4787617986271457e-06,
+      "loss": 0.0887,
+      "step": 2478
+    },
+    {
+      "epoch": 1.9058597502401537,
+      "grad_norm": 0.8306432379209792,
+      "learning_rate": 2.4727741998209278e-06,
+      "loss": 0.0817,
+      "step": 2480
+    },
+    {
+      "epoch": 1.9073967339097022,
+      "grad_norm": 0.7830733463300852,
+      "learning_rate": 2.4667898888989997e-06,
+      "loss": 0.0894,
+      "step": 2482
+    },
+    {
+      "epoch": 1.9089337175792507,
+      "grad_norm": 0.7962828362447372,
+      "learning_rate": 2.460808885015596e-06,
+      "loss": 0.0833,
+      "step": 2484
+    },
+    {
+      "epoch": 1.9104707012487991,
+      "grad_norm": 0.84538791014193,
+      "learning_rate": 2.454831207314364e-06,
+      "loss": 0.0898,
+      "step": 2486
+    },
+    {
+      "epoch": 1.9120076849183478,
+      "grad_norm": 0.7980052835225272,
+      "learning_rate": 2.4488568749283024e-06,
+      "loss": 0.0867,
+      "step": 2488
+    },
+    {
+      "epoch": 1.9135446685878963,
+      "grad_norm": 0.7823748772923179,
+      "learning_rate": 2.4428859069797065e-06,
+      "loss": 0.086,
+      "step": 2490
+    },
+    {
+      "epoch": 1.9150816522574448,
+      "grad_norm": 0.8497213682534336,
+      "learning_rate": 2.4369183225801037e-06,
+      "loss": 0.0876,
+      "step": 2492
+    },
+    {
+      "epoch": 1.9166186359269934,
+      "grad_norm": 0.729672453173719,
+      "learning_rate": 2.430954140830187e-06,
+      "loss": 0.08,
+      "step": 2494
+    },
+    {
+      "epoch": 1.9181556195965417,
+      "grad_norm": 0.8535703260306475,
+      "learning_rate": 2.424993380819759e-06,
+      "loss": 0.0809,
+      "step": 2496
+    },
+    {
+      "epoch": 1.9196926032660904,
+      "grad_norm": 0.8459889716819726,
+      "learning_rate": 2.419036061627676e-06,
+      "loss": 0.0898,
+      "step": 2498
+    },
+    {
+      "epoch": 1.9212295869356388,
+      "grad_norm": 0.8483259441228239,
+      "learning_rate": 2.4130822023217745e-06,
+      "loss": 0.0841,
+      "step": 2500
+    },
+    {
+      "epoch": 1.9212295869356388,
+      "eval_loss": 0.13174882531166077,
+      "eval_runtime": 362.8867,
+      "eval_samples_per_second": 50.994,
+      "eval_steps_per_second": 6.377,
+      "step": 2500
+    },
+    {
+      "epoch": 1.9227665706051873,
+      "grad_norm": 0.7537082873290889,
+      "learning_rate": 2.40713182195882e-06,
+      "loss": 0.0835,
+      "step": 2502
+    },
+    {
+      "epoch": 1.924303554274736,
+      "grad_norm": 0.8025926008593399,
+      "learning_rate": 2.401184939584441e-06,
+      "loss": 0.091,
+      "step": 2504
+    },
+    {
+      "epoch": 1.9258405379442842,
+      "grad_norm": 0.8587481263834186,
+      "learning_rate": 2.3952415742330715e-06,
+      "loss": 0.0817,
+      "step": 2506
+    },
+    {
+      "epoch": 1.927377521613833,
+      "grad_norm": 0.8496889861192923,
+      "learning_rate": 2.38930174492789e-06,
+      "loss": 0.0871,
+      "step": 2508
+    },
+    {
+      "epoch": 1.9289145052833814,
+      "grad_norm": 0.7449830563434601,
+      "learning_rate": 2.383365470680753e-06,
+      "loss": 0.0889,
+      "step": 2510
+    },
+    {
+      "epoch": 1.9304514889529298,
+      "grad_norm": 0.7842488010652338,
+      "learning_rate": 2.377432770492138e-06,
+      "loss": 0.0896,
+      "step": 2512
+    },
+    {
+      "epoch": 1.9319884726224785,
+      "grad_norm": 0.7604948684079603,
+      "learning_rate": 2.3715036633510887e-06,
+      "loss": 0.0768,
+      "step": 2514
+    },
+    {
+      "epoch": 1.9335254562920268,
+      "grad_norm": 0.8622740265461863,
+      "learning_rate": 2.365578168235143e-06,
+      "loss": 0.0884,
+      "step": 2516
+    },
+    {
+      "epoch": 1.9350624399615755,
+      "grad_norm": 0.8075457215159016,
+      "learning_rate": 2.3596563041102794e-06,
+      "loss": 0.0859,
+      "step": 2518
+    },
+    {
+      "epoch": 1.936599423631124,
+      "grad_norm": 0.7735312713920628,
+      "learning_rate": 2.3537380899308532e-06,
+      "loss": 0.0888,
+      "step": 2520
+    },
+    {
+      "epoch": 1.9381364073006724,
+      "grad_norm": 0.7250205794783319,
+      "learning_rate": 2.347823544639541e-06,
+      "loss": 0.0812,
+      "step": 2522
+    },
+    {
+      "epoch": 1.939673390970221,
+      "grad_norm": 0.8168723067376802,
+      "learning_rate": 2.3419126871672716e-06,
+      "loss": 0.085,
+      "step": 2524
+    },
+    {
+      "epoch": 1.9412103746397693,
+      "grad_norm": 0.7558226843999474,
+      "learning_rate": 2.3360055364331726e-06,
+      "loss": 0.0842,
+      "step": 2526
+    },
+    {
+      "epoch": 1.942747358309318,
+      "grad_norm": 0.8367673308249525,
+      "learning_rate": 2.3301021113445057e-06,
+      "loss": 0.0859,
+      "step": 2528
+    },
+    {
+      "epoch": 1.9442843419788665,
+      "grad_norm": 0.771933781721656,
+      "learning_rate": 2.3242024307966115e-06,
+      "loss": 0.0853,
+      "step": 2530
+    },
+    {
+      "epoch": 1.945821325648415,
+      "grad_norm": 0.8022281457204921,
+      "learning_rate": 2.3183065136728395e-06,
+      "loss": 0.0829,
+      "step": 2532
+    },
+    {
+      "epoch": 1.9473583093179636,
+      "grad_norm": 0.8770111872133347,
+      "learning_rate": 2.3124143788444994e-06,
+      "loss": 0.0871,
+      "step": 2534
+    },
+    {
+      "epoch": 1.9488952929875119,
+      "grad_norm": 0.7289175750768725,
+      "learning_rate": 2.3065260451707887e-06,
+      "loss": 0.0765,
+      "step": 2536
+    },
+    {
+      "epoch": 1.9504322766570605,
+      "grad_norm": 0.8573432100144109,
+      "learning_rate": 2.3006415314987453e-06,
+      "loss": 0.0926,
+      "step": 2538
+    },
+    {
+      "epoch": 1.951969260326609,
+      "grad_norm": 0.8049758874747357,
+      "learning_rate": 2.2947608566631738e-06,
+      "loss": 0.09,
+      "step": 2540
+    },
+    {
+      "epoch": 1.9535062439961575,
+      "grad_norm": 0.8070974588287757,
+      "learning_rate": 2.288884039486595e-06,
+      "loss": 0.0805,
+      "step": 2542
+    },
+    {
+      "epoch": 1.9550432276657062,
+      "grad_norm": 0.7988980263528441,
+      "learning_rate": 2.2830110987791816e-06,
+      "loss": 0.0868,
+      "step": 2544
+    },
+    {
+      "epoch": 1.9565802113352546,
+      "grad_norm": 0.8018076390389216,
+      "learning_rate": 2.277142053338701e-06,
+      "loss": 0.0833,
+      "step": 2546
+    },
+    {
+      "epoch": 1.958117195004803,
+      "grad_norm": 0.8761877037291468,
+      "learning_rate": 2.271276921950448e-06,
+      "loss": 0.0889,
+      "step": 2548
+    },
+    {
+      "epoch": 1.9596541786743515,
+      "grad_norm": 0.8222253983712655,
+      "learning_rate": 2.265415723387194e-06,
+      "loss": 0.0884,
+      "step": 2550
+    },
+    {
+      "epoch": 1.9611911623439,
+      "grad_norm": 0.8067488111444437,
+      "learning_rate": 2.259558476409119e-06,
+      "loss": 0.0878,
+      "step": 2552
+    },
+    {
+      "epoch": 1.9627281460134487,
+      "grad_norm": 0.7772228994300671,
+      "learning_rate": 2.253705199763759e-06,
+      "loss": 0.0827,
+      "step": 2554
+    },
+    {
+      "epoch": 1.9642651296829972,
+      "grad_norm": 0.8171211068617626,
+      "learning_rate": 2.247855912185938e-06,
+      "loss": 0.0853,
+      "step": 2556
+    },
+    {
+      "epoch": 1.9658021133525456,
+      "grad_norm": 0.7146445332349611,
+      "learning_rate": 2.242010632397715e-06,
+      "loss": 0.0782,
+      "step": 2558
+    },
+    {
+      "epoch": 1.967339097022094,
+      "grad_norm": 0.7891447556881424,
+      "learning_rate": 2.2361693791083176e-06,
+      "loss": 0.0887,
+      "step": 2560
+    },
+    {
+      "epoch": 1.9688760806916425,
+      "grad_norm": 0.7846780714035666,
+      "learning_rate": 2.230332171014091e-06,
+      "loss": 0.0851,
+      "step": 2562
+    },
+    {
+      "epoch": 1.9704130643611912,
+      "grad_norm": 0.8098130826581111,
+      "learning_rate": 2.2244990267984265e-06,
+      "loss": 0.0915,
+      "step": 2564
+    },
+    {
+      "epoch": 1.9719500480307397,
+      "grad_norm": 0.777976273941637,
+      "learning_rate": 2.2186699651317143e-06,
+      "loss": 0.0802,
+      "step": 2566
+    },
+    {
+      "epoch": 1.9734870317002882,
+      "grad_norm": 0.8355719530974074,
+      "learning_rate": 2.2128450046712702e-06,
+      "loss": 0.0904,
+      "step": 2568
+    },
+    {
+      "epoch": 1.9750240153698368,
+      "grad_norm": 0.7798614757356649,
+      "learning_rate": 2.2070241640612915e-06,
+      "loss": 0.0838,
+      "step": 2570
+    },
+    {
+      "epoch": 1.976560999039385,
+      "grad_norm": 0.8328604297726137,
+      "learning_rate": 2.2012074619327824e-06,
+      "loss": 0.0865,
+      "step": 2572
+    },
+    {
+      "epoch": 1.9780979827089338,
+      "grad_norm": 0.793696845619225,
+      "learning_rate": 2.195394916903502e-06,
+      "loss": 0.0782,
+      "step": 2574
+    },
+    {
+      "epoch": 1.9796349663784822,
+      "grad_norm": 0.7815626290849819,
+      "learning_rate": 2.1895865475779054e-06,
+      "loss": 0.0788,
+      "step": 2576
+    },
+    {
+      "epoch": 1.9811719500480307,
+      "grad_norm": 0.8439158338940291,
+      "learning_rate": 2.1837823725470835e-06,
+      "loss": 0.0889,
+      "step": 2578
+    },
+    {
+      "epoch": 1.9827089337175794,
+      "grad_norm": 0.8288860856763497,
+      "learning_rate": 2.177982410388699e-06,
+      "loss": 0.0884,
+      "step": 2580
+    },
+    {
+      "epoch": 1.9842459173871276,
+      "grad_norm": 0.8187063512218921,
+      "learning_rate": 2.1721866796669302e-06,
+      "loss": 0.088,
+      "step": 2582
+    },
+    {
+      "epoch": 1.9857829010566763,
+      "grad_norm": 0.7887200768254294,
+      "learning_rate": 2.166395198932414e-06,
+      "loss": 0.0836,
+      "step": 2584
+    },
+    {
+      "epoch": 1.9873198847262248,
+      "grad_norm": 0.8058422750958129,
+      "learning_rate": 2.160607986722186e-06,
+      "loss": 0.0872,
+      "step": 2586
+    },
+    {
+      "epoch": 1.9888568683957732,
+      "grad_norm": 0.8186786966538023,
+      "learning_rate": 2.154825061559614e-06,
+      "loss": 0.0921,
+      "step": 2588
+    },
+    {
+      "epoch": 1.990393852065322,
+      "grad_norm": 0.8648916595771557,
+      "learning_rate": 2.149046441954347e-06,
+      "loss": 0.0859,
+      "step": 2590
+    },
+    {
+      "epoch": 1.9919308357348702,
+      "grad_norm": 0.8612159243706975,
+      "learning_rate": 2.1432721464022532e-06,
+      "loss": 0.0997,
+      "step": 2592
+    },
+    {
+      "epoch": 1.9934678194044189,
+      "grad_norm": 0.8214954020731858,
+      "learning_rate": 2.137502193385361e-06,
+      "loss": 0.0875,
+      "step": 2594
+    },
+    {
+      "epoch": 1.9950048030739673,
+      "grad_norm": 0.7979371383049858,
+      "learning_rate": 2.1317366013717983e-06,
+      "loss": 0.0843,
+      "step": 2596
+    },
+    {
+      "epoch": 1.9965417867435158,
+      "grad_norm": 0.7620709623218802,
+      "learning_rate": 2.125975388815733e-06,
+      "loss": 0.0793,
+      "step": 2598
+    },
+    {
+      "epoch": 1.9980787704130645,
+      "grad_norm": 0.7554666873391905,
+      "learning_rate": 2.1202185741573206e-06,
+      "loss": 0.0871,
+      "step": 2600
+    },
+    {
+      "epoch": 1.9996157540826127,
+      "grad_norm": 0.7689027561104417,
+      "learning_rate": 2.1144661758226355e-06,
+      "loss": 0.0787,
+      "step": 2602
+    },
+    {
+      "epoch": 2.0011527377521614,
+      "grad_norm": 0.6010513132525606,
+      "learning_rate": 2.108718212223618e-06,
+      "loss": 0.0835,
+      "step": 2604
+    },
+    {
+      "epoch": 2.00268972142171,
+      "grad_norm": 0.5735836702570751,
+      "learning_rate": 2.1029747017580132e-06,
+      "loss": 0.0485,
+      "step": 2606
+    },
+    {
+      "epoch": 2.0042267050912583,
+      "grad_norm": 0.5702538041925809,
+      "learning_rate": 2.0972356628093154e-06,
+      "loss": 0.0452,
+      "step": 2608
+    },
+    {
+      "epoch": 2.005763688760807,
+      "grad_norm": 0.5895447128781794,
+      "learning_rate": 2.091501113746703e-06,
+      "loss": 0.0446,
+      "step": 2610
+    },
+    {
+      "epoch": 2.0073006724303553,
+      "grad_norm": 0.6069626452449407,
+      "learning_rate": 2.085771072924988e-06,
+      "loss": 0.0439,
+      "step": 2612
+    },
+    {
+      "epoch": 2.008837656099904,
+      "grad_norm": 0.6625893476183761,
+      "learning_rate": 2.080045558684546e-06,
+      "loss": 0.0426,
+      "step": 2614
+    },
+    {
+      "epoch": 2.0103746397694526,
+      "grad_norm": 0.7034714400173309,
+      "learning_rate": 2.0743245893512725e-06,
+      "loss": 0.0433,
+      "step": 2616
+    },
+    {
+      "epoch": 2.011911623439001,
+      "grad_norm": 0.7707738025199298,
+      "learning_rate": 2.0686081832365095e-06,
+      "loss": 0.045,
+      "step": 2618
+    },
+    {
+      "epoch": 2.0134486071085496,
+      "grad_norm": 0.793812018682344,
+      "learning_rate": 2.0628963586369966e-06,
+      "loss": 0.043,
+      "step": 2620
+    },
+    {
+      "epoch": 2.014985590778098,
+      "grad_norm": 0.7298066829996939,
+      "learning_rate": 2.057189133834806e-06,
+      "loss": 0.0413,
+      "step": 2622
+    },
+    {
+      "epoch": 2.0165225744476465,
+      "grad_norm": 0.7317715663050453,
+      "learning_rate": 2.051486527097292e-06,
+      "loss": 0.0413,
+      "step": 2624
+    },
+    {
+      "epoch": 2.018059558117195,
+      "grad_norm": 0.7006778779832138,
+      "learning_rate": 2.045788556677023e-06,
+      "loss": 0.0425,
+      "step": 2626
+    },
+    {
+      "epoch": 2.0195965417867434,
+      "grad_norm": 0.747566077495069,
+      "learning_rate": 2.04009524081173e-06,
+      "loss": 0.0416,
+      "step": 2628
+    },
+    {
+      "epoch": 2.021133525456292,
+      "grad_norm": 0.6905039170767364,
+      "learning_rate": 2.034406597724246e-06,
+      "loss": 0.0366,
+      "step": 2630
+    },
+    {
+      "epoch": 2.0226705091258403,
+      "grad_norm": 0.6513726296876595,
+      "learning_rate": 2.0287226456224464e-06,
+      "loss": 0.0417,
+      "step": 2632
+    },
+    {
+      "epoch": 2.024207492795389,
+      "grad_norm": 0.6710660115404629,
+      "learning_rate": 2.0230434026991936e-06,
+      "loss": 0.0439,
+      "step": 2634
+    },
+    {
+      "epoch": 2.0257444764649377,
+      "grad_norm": 0.6687092945349834,
+      "learning_rate": 2.0173688871322763e-06,
+      "loss": 0.0396,
+      "step": 2636
+    },
+    {
+      "epoch": 2.027281460134486,
+      "grad_norm": 0.7735624904374961,
+      "learning_rate": 2.011699117084352e-06,
+      "loss": 0.0431,
+      "step": 2638
+    },
+    {
+      "epoch": 2.0288184438040346,
+      "grad_norm": 0.7166923451943271,
+      "learning_rate": 2.0060341107028893e-06,
+      "loss": 0.0439,
+      "step": 2640
+    },
+    {
+      "epoch": 2.030355427473583,
+      "grad_norm": 0.7077961035403634,
+      "learning_rate": 2.0003738861201104e-06,
+      "loss": 0.0499,
+      "step": 2642
+    },
+    {
+      "epoch": 2.0318924111431316,
+      "grad_norm": 0.6883244536280981,
+      "learning_rate": 1.994718461452929e-06,
+      "loss": 0.0443,
+      "step": 2644
+    },
+    {
+      "epoch": 2.0334293948126803,
+      "grad_norm": 0.6214200682417671,
+      "learning_rate": 1.9890678548028994e-06,
+      "loss": 0.0425,
+      "step": 2646
+    },
+    {
+      "epoch": 2.0349663784822285,
+      "grad_norm": 0.6365455258734565,
+      "learning_rate": 1.9834220842561525e-06,
+      "loss": 0.0438,
+      "step": 2648
+    },
+    {
+      "epoch": 2.036503362151777,
+      "grad_norm": 0.6181452869288234,
+      "learning_rate": 1.9777811678833405e-06,
+      "loss": 0.0363,
+      "step": 2650
+    },
+    {
+      "epoch": 2.038040345821326,
+      "grad_norm": 0.6491919777612177,
+      "learning_rate": 1.972145123739581e-06,
+      "loss": 0.0397,
+      "step": 2652
+    },
+    {
+      "epoch": 2.039577329490874,
+      "grad_norm": 0.6585608787162763,
+      "learning_rate": 1.9665139698643894e-06,
+      "loss": 0.0368,
+      "step": 2654
+    },
+    {
+      "epoch": 2.041114313160423,
+      "grad_norm": 0.7727352776818418,
+      "learning_rate": 1.960887724281639e-06,
+      "loss": 0.0454,
+      "step": 2656
+    },
+    {
+      "epoch": 2.042651296829971,
+      "grad_norm": 0.6357364480821968,
+      "learning_rate": 1.955266404999487e-06,
+      "loss": 0.0415,
+      "step": 2658
+    },
+    {
+      "epoch": 2.0441882804995197,
+      "grad_norm": 0.6281847150403672,
+      "learning_rate": 1.9496500300103206e-06,
+      "loss": 0.0371,
+      "step": 2660
+    },
+    {
+      "epoch": 2.0457252641690684,
+      "grad_norm": 0.6912027377771374,
+      "learning_rate": 1.944038617290707e-06,
+      "loss": 0.04,
+      "step": 2662
+    },
+    {
+      "epoch": 2.0472622478386167,
+      "grad_norm": 0.7127284309719673,
+      "learning_rate": 1.938432184801327e-06,
+      "loss": 0.0444,
+      "step": 2664
+    },
+    {
+      "epoch": 2.0487992315081653,
+      "grad_norm": 0.6753527235959302,
+      "learning_rate": 1.9328307504869223e-06,
+      "loss": 0.0406,
+      "step": 2666
+    },
+    {
+      "epoch": 2.0503362151777136,
+      "grad_norm": 0.6882383050291725,
+      "learning_rate": 1.9272343322762377e-06,
+      "loss": 0.0393,
+      "step": 2668
+    },
+    {
+      "epoch": 2.0518731988472623,
+      "grad_norm": 0.7047552202339743,
+      "learning_rate": 1.9216429480819575e-06,
+      "loss": 0.0438,
+      "step": 2670
+    },
+    {
+      "epoch": 2.053410182516811,
+      "grad_norm": 0.7596140785601401,
+      "learning_rate": 1.9160566158006613e-06,
+      "loss": 0.0414,
+      "step": 2672
+    },
+    {
+      "epoch": 2.054947166186359,
+      "grad_norm": 0.6985050704664985,
+      "learning_rate": 1.9104753533127555e-06,
+      "loss": 0.0397,
+      "step": 2674
+    },
+    {
+      "epoch": 2.056484149855908,
+      "grad_norm": 0.6562480639793371,
+      "learning_rate": 1.9048991784824146e-06,
+      "loss": 0.0401,
+      "step": 2676
+    },
+    {
+      "epoch": 2.058021133525456,
+      "grad_norm": 0.69494658065462,
+      "learning_rate": 1.8993281091575362e-06,
+      "loss": 0.0404,
+      "step": 2678
+    },
+    {
+      "epoch": 2.059558117195005,
+      "grad_norm": 0.7358984218115102,
+      "learning_rate": 1.8937621631696722e-06,
+      "loss": 0.0422,
+      "step": 2680
+    },
+    {
+      "epoch": 2.0610951008645535,
+      "grad_norm": 0.6513459605049308,
+      "learning_rate": 1.8882013583339773e-06,
+      "loss": 0.0401,
+      "step": 2682
+    },
+    {
+      "epoch": 2.0626320845341017,
+      "grad_norm": 0.6939024794365927,
+      "learning_rate": 1.8826457124491504e-06,
+      "loss": 0.0445,
+      "step": 2684
+    },
+    {
+      "epoch": 2.0641690682036504,
+      "grad_norm": 0.6668447837064994,
+      "learning_rate": 1.8770952432973784e-06,
+      "loss": 0.0392,
+      "step": 2686
+    },
+    {
+      "epoch": 2.0657060518731987,
+      "grad_norm": 0.6666480416435122,
+      "learning_rate": 1.87154996864428e-06,
+      "loss": 0.0409,
+      "step": 2688
+    },
+    {
+      "epoch": 2.0672430355427474,
+      "grad_norm": 0.6709721661234623,
+      "learning_rate": 1.8660099062388431e-06,
+      "loss": 0.0405,
+      "step": 2690
+    },
+    {
+      "epoch": 2.068780019212296,
+      "grad_norm": 0.685700314026161,
+      "learning_rate": 1.8604750738133756e-06,
+      "loss": 0.0397,
+      "step": 2692
+    },
+    {
+      "epoch": 2.0703170028818443,
+      "grad_norm": 0.6515559887232024,
+      "learning_rate": 1.8549454890834497e-06,
+      "loss": 0.0384,
+      "step": 2694
+    },
+    {
+      "epoch": 2.071853986551393,
+      "grad_norm": 0.709788654208775,
+      "learning_rate": 1.8494211697478323e-06,
+      "loss": 0.0415,
+      "step": 2696
+    },
+    {
+      "epoch": 2.073390970220941,
+      "grad_norm": 0.6775450512874873,
+      "learning_rate": 1.8439021334884444e-06,
+      "loss": 0.0386,
+      "step": 2698
+    },
+    {
+      "epoch": 2.07492795389049,
+      "grad_norm": 0.7750933505013274,
+      "learning_rate": 1.8383883979702934e-06,
+      "loss": 0.0426,
+      "step": 2700
+    },
+    {
+      "epoch": 2.0764649375600386,
+      "grad_norm": 0.6584818113536601,
+      "learning_rate": 1.8328799808414227e-06,
+      "loss": 0.038,
+      "step": 2702
+    },
+    {
+      "epoch": 2.078001921229587,
+      "grad_norm": 0.7829082201798115,
+      "learning_rate": 1.8273768997328525e-06,
+      "loss": 0.04,
+      "step": 2704
+    },
+    {
+      "epoch": 2.0795389048991355,
+      "grad_norm": 0.734651522047483,
+      "learning_rate": 1.8218791722585205e-06,
+      "loss": 0.0415,
+      "step": 2706
+    },
+    {
+      "epoch": 2.0810758885686838,
+      "grad_norm": 0.6479085477298625,
+      "learning_rate": 1.8163868160152308e-06,
+      "loss": 0.0372,
+      "step": 2708
+    },
+    {
+      "epoch": 2.0826128722382324,
+      "grad_norm": 0.7161779560377441,
+      "learning_rate": 1.8108998485826017e-06,
+      "loss": 0.0439,
+      "step": 2710
+    },
+    {
+      "epoch": 2.084149855907781,
+      "grad_norm": 0.6464214115435033,
+      "learning_rate": 1.8054182875229925e-06,
+      "loss": 0.0401,
+      "step": 2712
+    },
+    {
+      "epoch": 2.0856868395773294,
+      "grad_norm": 0.694251707574077,
+      "learning_rate": 1.799942150381465e-06,
+      "loss": 0.0387,
+      "step": 2714
+    },
+    {
+      "epoch": 2.087223823246878,
+      "grad_norm": 0.7296255965243119,
+      "learning_rate": 1.7944714546857195e-06,
+      "loss": 0.0428,
+      "step": 2716
+    },
+    {
+      "epoch": 2.0887608069164267,
+      "grad_norm": 0.693023285089645,
+      "learning_rate": 1.7890062179460383e-06,
+      "loss": 0.0425,
+      "step": 2718
+    },
+    {
+      "epoch": 2.090297790585975,
+      "grad_norm": 0.8091932276327041,
+      "learning_rate": 1.7835464576552334e-06,
+      "loss": 0.0442,
+      "step": 2720
+    },
+    {
+      "epoch": 2.0918347742555237,
+      "grad_norm": 0.7647763579617219,
+      "learning_rate": 1.7780921912885828e-06,
+      "loss": 0.0423,
+      "step": 2722
+    },
+    {
+      "epoch": 2.093371757925072,
+      "grad_norm": 0.7252365506572918,
+      "learning_rate": 1.7726434363037843e-06,
+      "loss": 0.0401,
+      "step": 2724
+    },
+    {
+      "epoch": 2.0949087415946206,
+      "grad_norm": 0.7457710289722953,
+      "learning_rate": 1.7672002101408983e-06,
+      "loss": 0.0441,
+      "step": 2726
+    },
+    {
+      "epoch": 2.0964457252641693,
+      "grad_norm": 0.8268423722666511,
+      "learning_rate": 1.761762530222281e-06,
+      "loss": 0.0428,
+      "step": 2728
+    },
+    {
+      "epoch": 2.0979827089337175,
+      "grad_norm": 0.7059199502684411,
+      "learning_rate": 1.756330413952541e-06,
+      "loss": 0.0415,
+      "step": 2730
+    },
+    {
+      "epoch": 2.099519692603266,
+      "grad_norm": 0.6799281357494354,
+      "learning_rate": 1.7509038787184795e-06,
+      "loss": 0.0433,
+      "step": 2732
+    },
+    {
+      "epoch": 2.1010566762728144,
+      "grad_norm": 0.6560319432868118,
+      "learning_rate": 1.7454829418890321e-06,
+      "loss": 0.0404,
+      "step": 2734
+    },
+    {
+      "epoch": 2.102593659942363,
+      "grad_norm": 0.6785113040948152,
+      "learning_rate": 1.7400676208152185e-06,
+      "loss": 0.0387,
+      "step": 2736
+    },
+    {
+      "epoch": 2.104130643611912,
+      "grad_norm": 0.6843492876225117,
+      "learning_rate": 1.7346579328300795e-06,
+      "loss": 0.0437,
+      "step": 2738
+    },
+    {
+      "epoch": 2.10566762728146,
+      "grad_norm": 0.6887111820251511,
+      "learning_rate": 1.7292538952486288e-06,
+      "loss": 0.0384,
+      "step": 2740
+    },
+    {
+      "epoch": 2.1072046109510087,
+      "grad_norm": 0.7567201231202126,
+      "learning_rate": 1.7238555253677945e-06,
+      "loss": 0.0438,
+      "step": 2742
+    },
+    {
+      "epoch": 2.108741594620557,
+      "grad_norm": 0.6548000735118099,
+      "learning_rate": 1.7184628404663628e-06,
+      "loss": 0.0392,
+      "step": 2744
+    },
+    {
+      "epoch": 2.1102785782901057,
+      "grad_norm": 0.6916814777807151,
+      "learning_rate": 1.713075857804926e-06,
+      "loss": 0.039,
+      "step": 2746
+    },
+    {
+      "epoch": 2.1118155619596544,
+      "grad_norm": 0.675062961353931,
+      "learning_rate": 1.7076945946258195e-06,
+      "loss": 0.0409,
+      "step": 2748
+    },
+    {
+      "epoch": 2.1133525456292026,
+      "grad_norm": 0.7099794758199118,
+      "learning_rate": 1.702319068153079e-06,
+      "loss": 0.0422,
+      "step": 2750
+    },
+    {
+      "epoch": 2.1148895292987513,
+      "grad_norm": 0.8255111866724547,
+      "learning_rate": 1.6969492955923765e-06,
+      "loss": 0.0467,
+      "step": 2752
+    },
+    {
+      "epoch": 2.1164265129682995,
+      "grad_norm": 0.7089306864322071,
+      "learning_rate": 1.6915852941309628e-06,
+      "loss": 0.0372,
+      "step": 2754
+    },
+    {
+      "epoch": 2.117963496637848,
+      "grad_norm": 0.7195267991677364,
+      "learning_rate": 1.6862270809376217e-06,
+      "loss": 0.0441,
+      "step": 2756
+    },
+    {
+      "epoch": 2.119500480307397,
+      "grad_norm": 0.7600552648629298,
+      "learning_rate": 1.6808746731626085e-06,
+      "loss": 0.0498,
+      "step": 2758
+    },
+    {
+      "epoch": 2.121037463976945,
+      "grad_norm": 0.6719618311917168,
+      "learning_rate": 1.6755280879375975e-06,
+      "loss": 0.0365,
+      "step": 2760
+    },
+    {
+      "epoch": 2.122574447646494,
+      "grad_norm": 0.6725979801550628,
+      "learning_rate": 1.6701873423756275e-06,
+      "loss": 0.038,
+      "step": 2762
+    },
+    {
+      "epoch": 2.124111431316042,
+      "grad_norm": 0.7133205530228818,
+      "learning_rate": 1.6648524535710401e-06,
+      "loss": 0.041,
+      "step": 2764
+    },
+    {
+      "epoch": 2.1256484149855908,
+      "grad_norm": 0.6458222389869777,
+      "learning_rate": 1.6595234385994398e-06,
+      "loss": 0.0433,
+      "step": 2766
+    },
+    {
+      "epoch": 2.1271853986551394,
+      "grad_norm": 0.7422996801701388,
+      "learning_rate": 1.6542003145176265e-06,
+      "loss": 0.0425,
+      "step": 2768
+    },
+    {
+      "epoch": 2.1287223823246877,
+      "grad_norm": 0.7076408311621847,
+      "learning_rate": 1.648883098363542e-06,
+      "loss": 0.0416,
+      "step": 2770
+    },
+    {
+      "epoch": 2.1302593659942364,
+      "grad_norm": 0.7798719878560143,
+      "learning_rate": 1.6435718071562212e-06,
+      "loss": 0.0495,
+      "step": 2772
+    },
+    {
+      "epoch": 2.1317963496637846,
+      "grad_norm": 0.7240342459534611,
+      "learning_rate": 1.6382664578957359e-06,
+      "loss": 0.0392,
+      "step": 2774
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 0.7052969420907925,
+      "learning_rate": 1.6329670675631369e-06,
+      "loss": 0.0419,
+      "step": 2776
+    },
+    {
+      "epoch": 2.134870317002882,
+      "grad_norm": 0.7301031702593861,
+      "learning_rate": 1.6276736531204044e-06,
+      "loss": 0.0386,
+      "step": 2778
+    },
+    {
+      "epoch": 2.1364073006724302,
+      "grad_norm": 0.6875032643695771,
+      "learning_rate": 1.6223862315103865e-06,
+      "loss": 0.0393,
+      "step": 2780
+    },
+    {
+      "epoch": 2.137944284341979,
+      "grad_norm": 0.6775172137731008,
+      "learning_rate": 1.617104819656758e-06,
+      "loss": 0.0399,
+      "step": 2782
+    },
+    {
+      "epoch": 2.1394812680115276,
+      "grad_norm": 0.7408298536972419,
+      "learning_rate": 1.6118294344639496e-06,
+      "loss": 0.0431,
+      "step": 2784
+    },
+    {
+      "epoch": 2.141018251681076,
+      "grad_norm": 0.7531648570461281,
+      "learning_rate": 1.6065600928171054e-06,
+      "loss": 0.0401,
+      "step": 2786
+    },
+    {
+      "epoch": 2.1425552353506245,
+      "grad_norm": 0.6778919207003176,
+      "learning_rate": 1.6012968115820306e-06,
+      "loss": 0.0399,
+      "step": 2788
+    },
+    {
+      "epoch": 2.1440922190201728,
+      "grad_norm": 0.7039227096859256,
+      "learning_rate": 1.5960396076051233e-06,
+      "loss": 0.0402,
+      "step": 2790
+    },
+    {
+      "epoch": 2.1456292026897215,
+      "grad_norm": 0.6789632610059323,
+      "learning_rate": 1.5907884977133366e-06,
+      "loss": 0.0373,
+      "step": 2792
+    },
+    {
+      "epoch": 2.1471661863592697,
+      "grad_norm": 0.6884435472545695,
+      "learning_rate": 1.5855434987141163e-06,
+      "loss": 0.0447,
+      "step": 2794
+    },
+    {
+      "epoch": 2.1487031700288184,
+      "grad_norm": 0.6920660658427317,
+      "learning_rate": 1.580304627395348e-06,
+      "loss": 0.0454,
+      "step": 2796
+    },
+    {
+      "epoch": 2.150240153698367,
+      "grad_norm": 0.7745737611033345,
+      "learning_rate": 1.5750719005253062e-06,
+      "loss": 0.0413,
+      "step": 2798
+    },
+    {
+      "epoch": 2.1517771373679153,
+      "grad_norm": 0.6907792630115484,
+      "learning_rate": 1.569845334852595e-06,
+      "loss": 0.041,
+      "step": 2800
+    },
+    {
+      "epoch": 2.153314121037464,
+      "grad_norm": 0.6604941051087728,
+      "learning_rate": 1.5646249471060995e-06,
+      "loss": 0.0431,
+      "step": 2802
+    },
+    {
+      "epoch": 2.1548511047070127,
+      "grad_norm": 0.7080559060460522,
+      "learning_rate": 1.5594107539949365e-06,
+      "loss": 0.0421,
+      "step": 2804
+    },
+    {
+      "epoch": 2.156388088376561,
+      "grad_norm": 0.7149705649090393,
+      "learning_rate": 1.5542027722083853e-06,
+      "loss": 0.0388,
+      "step": 2806
+    },
+    {
+      "epoch": 2.1579250720461096,
+      "grad_norm": 0.7146661785693218,
+      "learning_rate": 1.5490010184158523e-06,
+      "loss": 0.0398,
+      "step": 2808
+    },
+    {
+      "epoch": 2.159462055715658,
+      "grad_norm": 0.7240758501470305,
+      "learning_rate": 1.543805509266806e-06,
+      "loss": 0.045,
+      "step": 2810
+    },
+    {
+      "epoch": 2.1609990393852065,
+      "grad_norm": 0.7001839005699572,
+      "learning_rate": 1.5386162613907287e-06,
+      "loss": 0.0377,
+      "step": 2812
+    },
+    {
+      "epoch": 2.1625360230547552,
+      "grad_norm": 0.7094421429704845,
+      "learning_rate": 1.5334332913970623e-06,
+      "loss": 0.0425,
+      "step": 2814
+    },
+    {
+      "epoch": 2.1640730067243035,
+      "grad_norm": 0.6847460522981713,
+      "learning_rate": 1.5282566158751524e-06,
+      "loss": 0.0402,
+      "step": 2816
+    },
+    {
+      "epoch": 2.165609990393852,
+      "grad_norm": 0.7120209696945222,
+      "learning_rate": 1.5230862513941995e-06,
+      "loss": 0.0399,
+      "step": 2818
+    },
+    {
+      "epoch": 2.1671469740634004,
+      "grad_norm": 0.6907346621733982,
+      "learning_rate": 1.5179222145032073e-06,
+      "loss": 0.0447,
+      "step": 2820
+    },
+    {
+      "epoch": 2.168683957732949,
+      "grad_norm": 0.7773144554098375,
+      "learning_rate": 1.5127645217309192e-06,
+      "loss": 0.0434,
+      "step": 2822
+    },
+    {
+      "epoch": 2.1702209414024978,
+      "grad_norm": 0.7117848076597437,
+      "learning_rate": 1.5076131895857784e-06,
+      "loss": 0.0423,
+      "step": 2824
+    },
+    {
+      "epoch": 2.171757925072046,
+      "grad_norm": 0.7027544209261267,
+      "learning_rate": 1.5024682345558675e-06,
+      "loss": 0.0401,
+      "step": 2826
+    },
+    {
+      "epoch": 2.1732949087415947,
+      "grad_norm": 0.7033554856080803,
+      "learning_rate": 1.4973296731088581e-06,
+      "loss": 0.0388,
+      "step": 2828
+    },
+    {
+      "epoch": 2.174831892411143,
+      "grad_norm": 0.7715815363157628,
+      "learning_rate": 1.4921975216919582e-06,
+      "loss": 0.04,
+      "step": 2830
+    },
+    {
+      "epoch": 2.1763688760806916,
+      "grad_norm": 0.7862111914567127,
+      "learning_rate": 1.4870717967318554e-06,
+      "loss": 0.0392,
+      "step": 2832
+    },
+    {
+      "epoch": 2.1779058597502403,
+      "grad_norm": 0.7221877404149011,
+      "learning_rate": 1.4819525146346723e-06,
+      "loss": 0.0461,
+      "step": 2834
+    },
+    {
+      "epoch": 2.1794428434197886,
+      "grad_norm": 0.6911027142163597,
+      "learning_rate": 1.4768396917859073e-06,
+      "loss": 0.0382,
+      "step": 2836
+    },
+    {
+      "epoch": 2.1809798270893372,
+      "grad_norm": 0.7119349195163982,
+      "learning_rate": 1.4717333445503851e-06,
+      "loss": 0.0384,
+      "step": 2838
+    },
+    {
+      "epoch": 2.1825168107588855,
+      "grad_norm": 0.7025660651802623,
+      "learning_rate": 1.4666334892722047e-06,
+      "loss": 0.0393,
+      "step": 2840
+    },
+    {
+      "epoch": 2.184053794428434,
+      "grad_norm": 0.7083054615445171,
+      "learning_rate": 1.4615401422746807e-06,
+      "loss": 0.0411,
+      "step": 2842
+    },
+    {
+      "epoch": 2.185590778097983,
+      "grad_norm": 0.7283963127887443,
+      "learning_rate": 1.4564533198603048e-06,
+      "loss": 0.0417,
+      "step": 2844
+    },
+    {
+      "epoch": 2.187127761767531,
+      "grad_norm": 0.6929199237523709,
+      "learning_rate": 1.45137303831068e-06,
+      "loss": 0.0378,
+      "step": 2846
+    },
+    {
+      "epoch": 2.18866474543708,
+      "grad_norm": 0.7003853343345394,
+      "learning_rate": 1.4462993138864725e-06,
+      "loss": 0.0389,
+      "step": 2848
+    },
+    {
+      "epoch": 2.1902017291066285,
+      "grad_norm": 0.6983653676006323,
+      "learning_rate": 1.441232162827364e-06,
+      "loss": 0.039,
+      "step": 2850
+    },
+    {
+      "epoch": 2.1917387127761767,
+      "grad_norm": 0.7159975682093909,
+      "learning_rate": 1.4361716013519952e-06,
+      "loss": 0.0409,
+      "step": 2852
+    },
+    {
+      "epoch": 2.1932756964457254,
+      "grad_norm": 0.7754416019144194,
+      "learning_rate": 1.4311176456579158e-06,
+      "loss": 0.0427,
+      "step": 2854
+    },
+    {
+      "epoch": 2.1948126801152736,
+      "grad_norm": 0.7345157001852429,
+      "learning_rate": 1.4260703119215324e-06,
+      "loss": 0.0425,
+      "step": 2856
+    },
+    {
+      "epoch": 2.1963496637848223,
+      "grad_norm": 0.6537911636031369,
+      "learning_rate": 1.4210296162980526e-06,
+      "loss": 0.04,
+      "step": 2858
+    },
+    {
+      "epoch": 2.1978866474543706,
+      "grad_norm": 0.6733339321649235,
+      "learning_rate": 1.415995574921444e-06,
+      "loss": 0.0358,
+      "step": 2860
+    },
+    {
+      "epoch": 2.1994236311239193,
+      "grad_norm": 0.6520397122582371,
+      "learning_rate": 1.4109682039043717e-06,
+      "loss": 0.0367,
+      "step": 2862
+    },
+    {
+      "epoch": 2.200960614793468,
+      "grad_norm": 0.7851918798134458,
+      "learning_rate": 1.4059475193381485e-06,
+      "loss": 0.044,
+      "step": 2864
+    },
+    {
+      "epoch": 2.202497598463016,
+      "grad_norm": 0.6944754890985299,
+      "learning_rate": 1.4009335372926895e-06,
+      "loss": 0.0388,
+      "step": 2866
+    },
+    {
+      "epoch": 2.204034582132565,
+      "grad_norm": 0.7185388294778169,
+      "learning_rate": 1.3959262738164548e-06,
+      "loss": 0.0434,
+      "step": 2868
+    },
+    {
+      "epoch": 2.2055715658021136,
+      "grad_norm": 0.7581514840792837,
+      "learning_rate": 1.3909257449364012e-06,
+      "loss": 0.0419,
+      "step": 2870
+    },
+    {
+      "epoch": 2.207108549471662,
+      "grad_norm": 0.6783363501218831,
+      "learning_rate": 1.38593196665793e-06,
+      "loss": 0.0374,
+      "step": 2872
+    },
+    {
+      "epoch": 2.2086455331412105,
+      "grad_norm": 0.6911512020404058,
+      "learning_rate": 1.3809449549648313e-06,
+      "loss": 0.037,
+      "step": 2874
+    },
+    {
+      "epoch": 2.2101825168107587,
+      "grad_norm": 0.6909020138454802,
+      "learning_rate": 1.3759647258192446e-06,
+      "loss": 0.0388,
+      "step": 2876
+    },
+    {
+      "epoch": 2.2117195004803074,
+      "grad_norm": 0.6798676329168656,
+      "learning_rate": 1.370991295161596e-06,
+      "loss": 0.0371,
+      "step": 2878
+    },
+    {
+      "epoch": 2.213256484149856,
+      "grad_norm": 0.7757828920522275,
+      "learning_rate": 1.3660246789105472e-06,
+      "loss": 0.0383,
+      "step": 2880
+    },
+    {
+      "epoch": 2.2147934678194043,
+      "grad_norm": 0.7289028293318881,
+      "learning_rate": 1.3610648929629593e-06,
+      "loss": 0.0408,
+      "step": 2882
+    },
+    {
+      "epoch": 2.216330451488953,
+      "grad_norm": 0.7420967431557955,
+      "learning_rate": 1.3561119531938205e-06,
+      "loss": 0.043,
+      "step": 2884
+    },
+    {
+      "epoch": 2.2178674351585013,
+      "grad_norm": 0.7163966573368784,
+      "learning_rate": 1.3511658754562126e-06,
+      "loss": 0.0396,
+      "step": 2886
+    },
+    {
+      "epoch": 2.21940441882805,
+      "grad_norm": 0.7419349249047851,
+      "learning_rate": 1.3462266755812522e-06,
+      "loss": 0.0417,
+      "step": 2888
+    },
+    {
+      "epoch": 2.2209414024975986,
+      "grad_norm": 0.7800199446876597,
+      "learning_rate": 1.3412943693780412e-06,
+      "loss": 0.0422,
+      "step": 2890
+    },
+    {
+      "epoch": 2.222478386167147,
+      "grad_norm": 0.6961688291543298,
+      "learning_rate": 1.3363689726336185e-06,
+      "loss": 0.0407,
+      "step": 2892
+    },
+    {
+      "epoch": 2.2240153698366956,
+      "grad_norm": 0.7098183143734048,
+      "learning_rate": 1.3314505011129031e-06,
+      "loss": 0.0441,
+      "step": 2894
+    },
+    {
+      "epoch": 2.225552353506244,
+      "grad_norm": 0.6599491965120293,
+      "learning_rate": 1.3265389705586513e-06,
+      "loss": 0.0391,
+      "step": 2896
+    },
+    {
+      "epoch": 2.2270893371757925,
+      "grad_norm": 0.6447579173261855,
+      "learning_rate": 1.3216343966914074e-06,
+      "loss": 0.0346,
+      "step": 2898
+    },
+    {
+      "epoch": 2.228626320845341,
+      "grad_norm": 0.7146194112938924,
+      "learning_rate": 1.31673679520944e-06,
+      "loss": 0.0414,
+      "step": 2900
+    },
+    {
+      "epoch": 2.2301633045148894,
+      "grad_norm": 0.7599787345630702,
+      "learning_rate": 1.3118461817887071e-06,
+      "loss": 0.0426,
+      "step": 2902
+    },
+    {
+      "epoch": 2.231700288184438,
+      "grad_norm": 0.7045861225059481,
+      "learning_rate": 1.3069625720827984e-06,
+      "loss": 0.0369,
+      "step": 2904
+    },
+    {
+      "epoch": 2.2332372718539863,
+      "grad_norm": 0.6678597621037076,
+      "learning_rate": 1.302085981722887e-06,
+      "loss": 0.0378,
+      "step": 2906
+    },
+    {
+      "epoch": 2.234774255523535,
+      "grad_norm": 0.7318552001241747,
+      "learning_rate": 1.2972164263176784e-06,
+      "loss": 0.043,
+      "step": 2908
+    },
+    {
+      "epoch": 2.2363112391930837,
+      "grad_norm": 0.7347881359720747,
+      "learning_rate": 1.2923539214533586e-06,
+      "loss": 0.0435,
+      "step": 2910
+    },
+    {
+      "epoch": 2.237848222862632,
+      "grad_norm": 0.7833005087501644,
+      "learning_rate": 1.2874984826935493e-06,
+      "loss": 0.0411,
+      "step": 2912
+    },
+    {
+      "epoch": 2.2393852065321806,
+      "grad_norm": 0.7340931719392261,
+      "learning_rate": 1.2826501255792588e-06,
+      "loss": 0.0419,
+      "step": 2914
+    },
+    {
+      "epoch": 2.2409221902017293,
+      "grad_norm": 0.7244781359138952,
+      "learning_rate": 1.277808865628821e-06,
+      "loss": 0.0403,
+      "step": 2916
+    },
+    {
+      "epoch": 2.2424591738712776,
+      "grad_norm": 0.7012819390876723,
+      "learning_rate": 1.2729747183378591e-06,
+      "loss": 0.041,
+      "step": 2918
+    },
+    {
+      "epoch": 2.2439961575408263,
+      "grad_norm": 0.7456233730244353,
+      "learning_rate": 1.2681476991792295e-06,
+      "loss": 0.0421,
+      "step": 2920
+    },
+    {
+      "epoch": 2.2455331412103745,
+      "grad_norm": 0.6802488303834758,
+      "learning_rate": 1.2633278236029728e-06,
+      "loss": 0.0362,
+      "step": 2922
+    },
+    {
+      "epoch": 2.247070124879923,
+      "grad_norm": 0.7518464030635749,
+      "learning_rate": 1.2585151070362655e-06,
+      "loss": 0.0404,
+      "step": 2924
+    },
+    {
+      "epoch": 2.2486071085494714,
+      "grad_norm": 0.6909362534084197,
+      "learning_rate": 1.253709564883368e-06,
+      "loss": 0.0404,
+      "step": 2926
+    },
+    {
+      "epoch": 2.25014409221902,
+      "grad_norm": 0.752505311933665,
+      "learning_rate": 1.2489112125255795e-06,
+      "loss": 0.0387,
+      "step": 2928
+    },
+    {
+      "epoch": 2.251681075888569,
+      "grad_norm": 0.74911032053493,
+      "learning_rate": 1.2441200653211853e-06,
+      "loss": 0.0383,
+      "step": 2930
+    },
+    {
+      "epoch": 2.253218059558117,
+      "grad_norm": 0.7048961647841477,
+      "learning_rate": 1.2393361386054102e-06,
+      "loss": 0.0359,
+      "step": 2932
+    },
+    {
+      "epoch": 2.2547550432276657,
+      "grad_norm": 0.7455505928042365,
+      "learning_rate": 1.2345594476903678e-06,
+      "loss": 0.0388,
+      "step": 2934
+    },
+    {
+      "epoch": 2.2562920268972144,
+      "grad_norm": 0.7618852898444295,
+      "learning_rate": 1.2297900078650073e-06,
+      "loss": 0.0406,
+      "step": 2936
+    },
+    {
+      "epoch": 2.2578290105667627,
+      "grad_norm": 0.7877656713372072,
+      "learning_rate": 1.225027834395077e-06,
+      "loss": 0.0418,
+      "step": 2938
+    },
+    {
+      "epoch": 2.2593659942363113,
+      "grad_norm": 0.7487876696738023,
+      "learning_rate": 1.2202729425230628e-06,
+      "loss": 0.0435,
+      "step": 2940
+    },
+    {
+      "epoch": 2.2609029779058596,
+      "grad_norm": 0.7549339920384163,
+      "learning_rate": 1.215525347468143e-06,
+      "loss": 0.039,
+      "step": 2942
+    },
+    {
+      "epoch": 2.2624399615754083,
+      "grad_norm": 0.7381797557455133,
+      "learning_rate": 1.210785064426143e-06,
+      "loss": 0.0401,
+      "step": 2944
+    },
+    {
+      "epoch": 2.263976945244957,
+      "grad_norm": 0.7303121808696214,
+      "learning_rate": 1.2060521085694836e-06,
+      "loss": 0.0393,
+      "step": 2946
+    },
+    {
+      "epoch": 2.265513928914505,
+      "grad_norm": 0.7091244479813347,
+      "learning_rate": 1.2013264950471333e-06,
+      "loss": 0.0426,
+      "step": 2948
+    },
+    {
+      "epoch": 2.267050912584054,
+      "grad_norm": 0.8368399757213874,
+      "learning_rate": 1.196608238984561e-06,
+      "loss": 0.0488,
+      "step": 2950
+    },
+    {
+      "epoch": 2.268587896253602,
+      "grad_norm": 0.7827298872789893,
+      "learning_rate": 1.191897355483681e-06,
+      "loss": 0.0439,
+      "step": 2952
+    },
+    {
+      "epoch": 2.270124879923151,
+      "grad_norm": 0.7171969138664943,
+      "learning_rate": 1.1871938596228177e-06,
+      "loss": 0.0388,
+      "step": 2954
+    },
+    {
+      "epoch": 2.2716618635926995,
+      "grad_norm": 0.7079483043665579,
+      "learning_rate": 1.1824977664566453e-06,
+      "loss": 0.0418,
+      "step": 2956
+    },
+    {
+      "epoch": 2.2731988472622477,
+      "grad_norm": 0.7574665616251981,
+      "learning_rate": 1.1778090910161426e-06,
+      "loss": 0.0416,
+      "step": 2958
+    },
+    {
+      "epoch": 2.2747358309317964,
+      "grad_norm": 0.7111652785046836,
+      "learning_rate": 1.1731278483085481e-06,
+      "loss": 0.0427,
+      "step": 2960
+    },
+    {
+      "epoch": 2.2762728146013447,
+      "grad_norm": 0.7535827297817603,
+      "learning_rate": 1.1684540533173104e-06,
+      "loss": 0.041,
+      "step": 2962
+    },
+    {
+      "epoch": 2.2778097982708934,
+      "grad_norm": 0.7519434123378339,
+      "learning_rate": 1.1637877210020395e-06,
+      "loss": 0.0408,
+      "step": 2964
+    },
+    {
+      "epoch": 2.279346781940442,
+      "grad_norm": 0.7008462349602121,
+      "learning_rate": 1.1591288662984594e-06,
+      "loss": 0.0397,
+      "step": 2966
+    },
+    {
+      "epoch": 2.2808837656099903,
+      "grad_norm": 0.7152887905478794,
+      "learning_rate": 1.154477504118357e-06,
+      "loss": 0.038,
+      "step": 2968
+    },
+    {
+      "epoch": 2.282420749279539,
+      "grad_norm": 0.7322203327399467,
+      "learning_rate": 1.1498336493495444e-06,
+      "loss": 0.0424,
+      "step": 2970
+    },
+    {
+      "epoch": 2.283957732949087,
+      "grad_norm": 0.705019954856083,
+      "learning_rate": 1.1451973168557992e-06,
+      "loss": 0.0393,
+      "step": 2972
+    },
+    {
+      "epoch": 2.285494716618636,
+      "grad_norm": 0.7314831773427871,
+      "learning_rate": 1.1405685214768224e-06,
+      "loss": 0.042,
+      "step": 2974
+    },
+    {
+      "epoch": 2.2870317002881846,
+      "grad_norm": 0.7024944801405867,
+      "learning_rate": 1.1359472780281916e-06,
+      "loss": 0.04,
+      "step": 2976
+    },
+    {
+      "epoch": 2.288568683957733,
+      "grad_norm": 0.7216103123781318,
+      "learning_rate": 1.1313336013013139e-06,
+      "loss": 0.04,
+      "step": 2978
+    },
+    {
+      "epoch": 2.2901056676272815,
+      "grad_norm": 0.6819781243729732,
+      "learning_rate": 1.126727506063375e-06,
+      "loss": 0.0383,
+      "step": 2980
+    },
+    {
+      "epoch": 2.29164265129683,
+      "grad_norm": 0.6997754986778041,
+      "learning_rate": 1.122129007057296e-06,
+      "loss": 0.0402,
+      "step": 2982
+    },
+    {
+      "epoch": 2.2931796349663784,
+      "grad_norm": 0.6626184454931626,
+      "learning_rate": 1.1175381190016833e-06,
+      "loss": 0.038,
+      "step": 2984
+    },
+    {
+      "epoch": 2.294716618635927,
+      "grad_norm": 0.6890948335939371,
+      "learning_rate": 1.1129548565907842e-06,
+      "loss": 0.041,
+      "step": 2986
+    },
+    {
+      "epoch": 2.2962536023054754,
+      "grad_norm": 0.6433359932240779,
+      "learning_rate": 1.108379234494435e-06,
+      "loss": 0.0351,
+      "step": 2988
+    },
+    {
+      "epoch": 2.297790585975024,
+      "grad_norm": 0.7392619695394935,
+      "learning_rate": 1.103811267358019e-06,
+      "loss": 0.0414,
+      "step": 2990
+    },
+    {
+      "epoch": 2.2993275696445723,
+      "grad_norm": 0.7246127735420398,
+      "learning_rate": 1.0992509698024226e-06,
+      "loss": 0.0389,
+      "step": 2992
+    },
+    {
+      "epoch": 2.300864553314121,
+      "grad_norm": 0.7447654773215131,
+      "learning_rate": 1.0946983564239754e-06,
+      "loss": 0.0459,
+      "step": 2994
+    },
+    {
+      "epoch": 2.3024015369836697,
+      "grad_norm": 0.7324507127817105,
+      "learning_rate": 1.0901534417944175e-06,
+      "loss": 0.0377,
+      "step": 2996
+    },
+    {
+      "epoch": 2.303938520653218,
+      "grad_norm": 0.6915535971116652,
+      "learning_rate": 1.0856162404608466e-06,
+      "loss": 0.0361,
+      "step": 2998
+    },
+    {
+      "epoch": 2.3054755043227666,
+      "grad_norm": 0.7490224319565619,
+      "learning_rate": 1.0810867669456705e-06,
+      "loss": 0.0375,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3054755043227666,
+      "eval_loss": 0.1596132069826126,
+      "eval_runtime": 361.7237,
+      "eval_samples_per_second": 51.158,
+      "eval_steps_per_second": 6.397,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3070124879923153,
+      "grad_norm": 0.7500197718141116,
+      "learning_rate": 1.0765650357465648e-06,
+      "loss": 0.0439,
+      "step": 3002
+    },
+    {
+      "epoch": 2.3085494716618635,
+      "grad_norm": 0.7509126020866208,
+      "learning_rate": 1.07205106133642e-06,
+      "loss": 0.0399,
+      "step": 3004
+    },
+    {
+      "epoch": 2.310086455331412,
+      "grad_norm": 0.6985233959996506,
+      "learning_rate": 1.0675448581633016e-06,
+      "loss": 0.0368,
+      "step": 3006
+    },
+    {
+      "epoch": 2.3116234390009605,
+      "grad_norm": 0.7109894469419616,
+      "learning_rate": 1.063046440650405e-06,
+      "loss": 0.0394,
+      "step": 3008
+    },
+    {
+      "epoch": 2.313160422670509,
+      "grad_norm": 0.6685986841427571,
+      "learning_rate": 1.0585558231959986e-06,
+      "loss": 0.0354,
+      "step": 3010
+    },
+    {
+      "epoch": 2.314697406340058,
+      "grad_norm": 0.6537755179648417,
+      "learning_rate": 1.0540730201733887e-06,
+      "loss": 0.0341,
+      "step": 3012
+    },
+    {
+      "epoch": 2.316234390009606,
+      "grad_norm": 0.7680457966259363,
+      "learning_rate": 1.0495980459308696e-06,
+      "loss": 0.0401,
+      "step": 3014
+    },
+    {
+      "epoch": 2.3177713736791548,
+      "grad_norm": 0.7160760279013273,
+      "learning_rate": 1.0451309147916773e-06,
+      "loss": 0.0385,
+      "step": 3016
+    },
+    {
+      "epoch": 2.319308357348703,
+      "grad_norm": 0.6930648625031354,
+      "learning_rate": 1.040671641053945e-06,
+      "loss": 0.0385,
+      "step": 3018
+    },
+    {
+      "epoch": 2.3208453410182517,
+      "grad_norm": 0.7212560558538098,
+      "learning_rate": 1.036220238990653e-06,
+      "loss": 0.038,
+      "step": 3020
+    },
+    {
+      "epoch": 2.3223823246878004,
+      "grad_norm": 0.7246474758799715,
+      "learning_rate": 1.0317767228495906e-06,
+      "loss": 0.0371,
+      "step": 3022
+    },
+    {
+      "epoch": 2.3239193083573486,
+      "grad_norm": 0.7534642781877979,
+      "learning_rate": 1.0273411068533037e-06,
+      "loss": 0.0387,
+      "step": 3024
+    },
+    {
+      "epoch": 2.3254562920268973,
+      "grad_norm": 0.7525530298880164,
+      "learning_rate": 1.0229134051990528e-06,
+      "loss": 0.0406,
+      "step": 3026
+    },
+    {
+      "epoch": 2.3269932756964455,
+      "grad_norm": 0.7939486095316215,
+      "learning_rate": 1.018493632058767e-06,
+      "loss": 0.0445,
+      "step": 3028
+    },
+    {
+      "epoch": 2.3285302593659942,
+      "grad_norm": 0.7218972892190336,
+      "learning_rate": 1.0140818015789975e-06,
+      "loss": 0.0405,
+      "step": 3030
+    },
+    {
+      "epoch": 2.330067243035543,
+      "grad_norm": 0.7983391471167234,
+      "learning_rate": 1.009677927880873e-06,
+      "loss": 0.04,
+      "step": 3032
+    },
+    {
+      "epoch": 2.331604226705091,
+      "grad_norm": 0.7867565276794657,
+      "learning_rate": 1.0052820250600568e-06,
+      "loss": 0.0433,
+      "step": 3034
+    },
+    {
+      "epoch": 2.33314121037464,
+      "grad_norm": 0.7385399550662357,
+      "learning_rate": 1.000894107186695e-06,
+      "loss": 0.0358,
+      "step": 3036
+    },
+    {
+      "epoch": 2.334678194044188,
+      "grad_norm": 0.7023204396957986,
+      "learning_rate": 9.965141883053803e-07,
+      "loss": 0.0409,
+      "step": 3038
+    },
+    {
+      "epoch": 2.3362151777137368,
+      "grad_norm": 0.7239223187447263,
+      "learning_rate": 9.921422824351001e-07,
+      "loss": 0.0401,
+      "step": 3040
+    },
+    {
+      "epoch": 2.3377521613832855,
+      "grad_norm": 0.7201483530659722,
+      "learning_rate": 9.87778403569196e-07,
+      "loss": 0.0374,
+      "step": 3042
+    },
+    {
+      "epoch": 2.3392891450528337,
+      "grad_norm": 0.7067782590935108,
+      "learning_rate": 9.834225656753167e-07,
+      "loss": 0.0364,
+      "step": 3044
+    },
+    {
+      "epoch": 2.3408261287223824,
+      "grad_norm": 0.767002733961986,
+      "learning_rate": 9.790747826953707e-07,
+      "loss": 0.0382,
+      "step": 3046
+    },
+    {
+      "epoch": 2.342363112391931,
+      "grad_norm": 0.7748316117552638,
+      "learning_rate": 9.747350685454906e-07,
+      "loss": 0.0416,
+      "step": 3048
+    },
+    {
+      "epoch": 2.3439000960614793,
+      "grad_norm": 0.6930253730222432,
+      "learning_rate": 9.704034371159801e-07,
+      "loss": 0.0409,
+      "step": 3050
+    },
+    {
+      "epoch": 2.345437079731028,
+      "grad_norm": 0.7552114221352043,
+      "learning_rate": 9.660799022712694e-07,
+      "loss": 0.0412,
+      "step": 3052
+    },
+    {
+      "epoch": 2.3469740634005762,
+      "grad_norm": 0.7612525976410068,
+      "learning_rate": 9.617644778498773e-07,
+      "loss": 0.0383,
+      "step": 3054
+    },
+    {
+      "epoch": 2.348511047070125,
+      "grad_norm": 0.7280586427447336,
+      "learning_rate": 9.574571776643617e-07,
+      "loss": 0.0433,
+      "step": 3056
+    },
+    {
+      "epoch": 2.350048030739673,
+      "grad_norm": 0.6693940571370718,
+      "learning_rate": 9.531580155012778e-07,
+      "loss": 0.04,
+      "step": 3058
+    },
+    {
+      "epoch": 2.351585014409222,
+      "grad_norm": 0.7048821090945985,
+      "learning_rate": 9.488670051211332e-07,
+      "loss": 0.0394,
+      "step": 3060
+    },
+    {
+      "epoch": 2.3531219980787705,
+      "grad_norm": 0.7505448748331673,
+      "learning_rate": 9.4458416025834e-07,
+      "loss": 0.0392,
+      "step": 3062
+    },
+    {
+      "epoch": 2.354658981748319,
+      "grad_norm": 0.6771190590157495,
+      "learning_rate": 9.403094946211808e-07,
+      "loss": 0.0413,
+      "step": 3064
+    },
+    {
+      "epoch": 2.3561959654178675,
+      "grad_norm": 0.7260675282763582,
+      "learning_rate": 9.360430218917558e-07,
+      "loss": 0.0398,
+      "step": 3066
+    },
+    {
+      "epoch": 2.357732949087416,
+      "grad_norm": 0.687104316635614,
+      "learning_rate": 9.317847557259387e-07,
+      "loss": 0.0388,
+      "step": 3068
+    },
+    {
+      "epoch": 2.3592699327569644,
+      "grad_norm": 0.6959356324751542,
+      "learning_rate": 9.275347097533404e-07,
+      "loss": 0.0409,
+      "step": 3070
+    },
+    {
+      "epoch": 2.360806916426513,
+      "grad_norm": 0.7566129208877186,
+      "learning_rate": 9.232928975772597e-07,
+      "loss": 0.0386,
+      "step": 3072
+    },
+    {
+      "epoch": 2.3623439000960613,
+      "grad_norm": 0.7158570072440973,
+      "learning_rate": 9.190593327746406e-07,
+      "loss": 0.0389,
+      "step": 3074
+    },
+    {
+      "epoch": 2.36388088376561,
+      "grad_norm": 0.7207667687469005,
+      "learning_rate": 9.148340288960307e-07,
+      "loss": 0.0409,
+      "step": 3076
+    },
+    {
+      "epoch": 2.3654178674351587,
+      "grad_norm": 0.7445605342380751,
+      "learning_rate": 9.106169994655347e-07,
+      "loss": 0.0407,
+      "step": 3078
+    },
+    {
+      "epoch": 2.366954851104707,
+      "grad_norm": 0.7736558101051751,
+      "learning_rate": 9.064082579807752e-07,
+      "loss": 0.0387,
+      "step": 3080
+    },
+    {
+      "epoch": 2.3684918347742556,
+      "grad_norm": 0.7435778083456558,
+      "learning_rate": 9.022078179128439e-07,
+      "loss": 0.0422,
+      "step": 3082
+    },
+    {
+      "epoch": 2.370028818443804,
+      "grad_norm": 0.680093837395743,
+      "learning_rate": 8.980156927062642e-07,
+      "loss": 0.0341,
+      "step": 3084
+    },
+    {
+      "epoch": 2.3715658021133526,
+      "grad_norm": 0.801244212730542,
+      "learning_rate": 8.93831895778948e-07,
+      "loss": 0.0372,
+      "step": 3086
+    },
+    {
+      "epoch": 2.3731027857829012,
+      "grad_norm": 0.7357307681580836,
+      "learning_rate": 8.896564405221447e-07,
+      "loss": 0.0377,
+      "step": 3088
+    },
+    {
+      "epoch": 2.3746397694524495,
+      "grad_norm": 0.7209769877560075,
+      "learning_rate": 8.85489340300409e-07,
+      "loss": 0.0423,
+      "step": 3090
+    },
+    {
+      "epoch": 2.376176753121998,
+      "grad_norm": 0.7330652449279741,
+      "learning_rate": 8.813306084515513e-07,
+      "loss": 0.0416,
+      "step": 3092
+    },
+    {
+      "epoch": 2.3777137367915464,
+      "grad_norm": 0.7504076654960494,
+      "learning_rate": 8.771802582865972e-07,
+      "loss": 0.0422,
+      "step": 3094
+    },
+    {
+      "epoch": 2.379250720461095,
+      "grad_norm": 0.6852298932331159,
+      "learning_rate": 8.730383030897463e-07,
+      "loss": 0.038,
+      "step": 3096
+    },
+    {
+      "epoch": 2.380787704130644,
+      "grad_norm": 0.6881568732248152,
+      "learning_rate": 8.689047561183245e-07,
+      "loss": 0.0382,
+      "step": 3098
+    },
+    {
+      "epoch": 2.382324687800192,
+      "grad_norm": 0.6634157742716904,
+      "learning_rate": 8.647796306027476e-07,
+      "loss": 0.0379,
+      "step": 3100
+    },
+    {
+      "epoch": 2.3838616714697407,
+      "grad_norm": 0.7328461753734546,
+      "learning_rate": 8.606629397464787e-07,
+      "loss": 0.0408,
+      "step": 3102
+    },
+    {
+      "epoch": 2.385398655139289,
+      "grad_norm": 0.7159969869187992,
+      "learning_rate": 8.565546967259782e-07,
+      "loss": 0.0351,
+      "step": 3104
+    },
+    {
+      "epoch": 2.3869356388088376,
+      "grad_norm": 0.7267286204071172,
+      "learning_rate": 8.524549146906714e-07,
+      "loss": 0.0399,
+      "step": 3106
+    },
+    {
+      "epoch": 2.3884726224783863,
+      "grad_norm": 0.6908926932230512,
+      "learning_rate": 8.483636067629014e-07,
+      "loss": 0.0434,
+      "step": 3108
+    },
+    {
+      "epoch": 2.3900096061479346,
+      "grad_norm": 0.697297173531623,
+      "learning_rate": 8.442807860378868e-07,
+      "loss": 0.0398,
+      "step": 3110
+    },
+    {
+      "epoch": 2.3915465898174832,
+      "grad_norm": 0.704515687419953,
+      "learning_rate": 8.40206465583683e-07,
+      "loss": 0.041,
+      "step": 3112
+    },
+    {
+      "epoch": 2.393083573487032,
+      "grad_norm": 0.717830937968017,
+      "learning_rate": 8.361406584411343e-07,
+      "loss": 0.0382,
+      "step": 3114
+    },
+    {
+      "epoch": 2.39462055715658,
+      "grad_norm": 0.6666435251973266,
+      "learning_rate": 8.320833776238384e-07,
+      "loss": 0.0355,
+      "step": 3116
+    },
+    {
+      "epoch": 2.396157540826129,
+      "grad_norm": 0.7678187211880796,
+      "learning_rate": 8.280346361181063e-07,
+      "loss": 0.0445,
+      "step": 3118
+    },
+    {
+      "epoch": 2.397694524495677,
+      "grad_norm": 0.6358919496522946,
+      "learning_rate": 8.239944468829094e-07,
+      "loss": 0.038,
+      "step": 3120
+    },
+    {
+      "epoch": 2.399231508165226,
+      "grad_norm": 0.6575677433822967,
+      "learning_rate": 8.199628228498507e-07,
+      "loss": 0.0382,
+      "step": 3122
+    },
+    {
+      "epoch": 2.400768491834774,
+      "grad_norm": 0.7346980326551955,
+      "learning_rate": 8.159397769231166e-07,
+      "loss": 0.0406,
+      "step": 3124
+    },
+    {
+      "epoch": 2.4023054755043227,
+      "grad_norm": 0.6780951881965817,
+      "learning_rate": 8.119253219794369e-07,
+      "loss": 0.0376,
+      "step": 3126
+    },
+    {
+      "epoch": 2.4038424591738714,
+      "grad_norm": 0.7518449654269455,
+      "learning_rate": 8.079194708680458e-07,
+      "loss": 0.0432,
+      "step": 3128
+    },
+    {
+      "epoch": 2.4053794428434196,
+      "grad_norm": 0.6735026739540719,
+      "learning_rate": 8.039222364106351e-07,
+      "loss": 0.0353,
+      "step": 3130
+    },
+    {
+      "epoch": 2.4069164265129683,
+      "grad_norm": 0.6342721214297197,
+      "learning_rate": 7.999336314013204e-07,
+      "loss": 0.0402,
+      "step": 3132
+    },
+    {
+      "epoch": 2.408453410182517,
+      "grad_norm": 0.7230053307776508,
+      "learning_rate": 7.959536686065951e-07,
+      "loss": 0.0391,
+      "step": 3134
+    },
+    {
+      "epoch": 2.4099903938520653,
+      "grad_norm": 0.7653416656760699,
+      "learning_rate": 7.919823607652921e-07,
+      "loss": 0.0399,
+      "step": 3136
+    },
+    {
+      "epoch": 2.411527377521614,
+      "grad_norm": 0.8030771738565683,
+      "learning_rate": 7.880197205885418e-07,
+      "loss": 0.0429,
+      "step": 3138
+    },
+    {
+      "epoch": 2.413064361191162,
+      "grad_norm": 0.7931388452091263,
+      "learning_rate": 7.840657607597282e-07,
+      "loss": 0.0428,
+      "step": 3140
+    },
+    {
+      "epoch": 2.414601344860711,
+      "grad_norm": 0.7430780202524496,
+      "learning_rate": 7.80120493934458e-07,
+      "loss": 0.0431,
+      "step": 3142
+    },
+    {
+      "epoch": 2.4161383285302596,
+      "grad_norm": 0.6756429003384381,
+      "learning_rate": 7.761839327405105e-07,
+      "loss": 0.038,
+      "step": 3144
+    },
+    {
+      "epoch": 2.417675312199808,
+      "grad_norm": 0.7445655284954013,
+      "learning_rate": 7.722560897777989e-07,
+      "loss": 0.0378,
+      "step": 3146
+    },
+    {
+      "epoch": 2.4192122958693565,
+      "grad_norm": 0.780771575546592,
+      "learning_rate": 7.683369776183342e-07,
+      "loss": 0.0398,
+      "step": 3148
+    },
+    {
+      "epoch": 2.4207492795389047,
+      "grad_norm": 0.7856619296684908,
+      "learning_rate": 7.644266088061811e-07,
+      "loss": 0.0436,
+      "step": 3150
+    },
+    {
+      "epoch": 2.4222862632084534,
+      "grad_norm": 0.7727263360626014,
+      "learning_rate": 7.605249958574199e-07,
+      "loss": 0.0435,
+      "step": 3152
+    },
+    {
+      "epoch": 2.423823246878002,
+      "grad_norm": 0.7022557826012432,
+      "learning_rate": 7.566321512601064e-07,
+      "loss": 0.039,
+      "step": 3154
+    },
+    {
+      "epoch": 2.4253602305475503,
+      "grad_norm": 0.7135877902071903,
+      "learning_rate": 7.527480874742269e-07,
+      "loss": 0.0404,
+      "step": 3156
+    },
+    {
+      "epoch": 2.426897214217099,
+      "grad_norm": 0.726091607679137,
+      "learning_rate": 7.488728169316684e-07,
+      "loss": 0.039,
+      "step": 3158
+    },
+    {
+      "epoch": 2.4284341978866473,
+      "grad_norm": 0.6803423609158856,
+      "learning_rate": 7.450063520361706e-07,
+      "loss": 0.0382,
+      "step": 3160
+    },
+    {
+      "epoch": 2.429971181556196,
+      "grad_norm": 0.631085298189217,
+      "learning_rate": 7.411487051632861e-07,
+      "loss": 0.0369,
+      "step": 3162
+    },
+    {
+      "epoch": 2.4315081652257446,
+      "grad_norm": 0.68946022968755,
+      "learning_rate": 7.372998886603466e-07,
+      "loss": 0.0383,
+      "step": 3164
+    },
+    {
+      "epoch": 2.433045148895293,
+      "grad_norm": 0.6831510201370967,
+      "learning_rate": 7.33459914846419e-07,
+      "loss": 0.0431,
+      "step": 3166
+    },
+    {
+      "epoch": 2.4345821325648416,
+      "grad_norm": 0.7684038963667288,
+      "learning_rate": 7.296287960122672e-07,
+      "loss": 0.0414,
+      "step": 3168
+    },
+    {
+      "epoch": 2.43611911623439,
+      "grad_norm": 0.7347215978088572,
+      "learning_rate": 7.258065444203128e-07,
+      "loss": 0.0416,
+      "step": 3170
+    },
+    {
+      "epoch": 2.4376560999039385,
+      "grad_norm": 0.7643085623961072,
+      "learning_rate": 7.219931723045929e-07,
+      "loss": 0.0433,
+      "step": 3172
+    },
+    {
+      "epoch": 2.439193083573487,
+      "grad_norm": 0.7071330978188225,
+      "learning_rate": 7.181886918707297e-07,
+      "loss": 0.037,
+      "step": 3174
+    },
+    {
+      "epoch": 2.4407300672430354,
+      "grad_norm": 0.6841434498931844,
+      "learning_rate": 7.143931152958791e-07,
+      "loss": 0.0372,
+      "step": 3176
+    },
+    {
+      "epoch": 2.442267050912584,
+      "grad_norm": 0.692314572344589,
+      "learning_rate": 7.106064547287002e-07,
+      "loss": 0.0345,
+      "step": 3178
+    },
+    {
+      "epoch": 2.443804034582133,
+      "grad_norm": 0.7363574996455537,
+      "learning_rate": 7.068287222893179e-07,
+      "loss": 0.0379,
+      "step": 3180
+    },
+    {
+      "epoch": 2.445341018251681,
+      "grad_norm": 0.6822850019837609,
+      "learning_rate": 7.030599300692748e-07,
+      "loss": 0.0368,
+      "step": 3182
+    },
+    {
+      "epoch": 2.4468780019212297,
+      "grad_norm": 0.7168612788689864,
+      "learning_rate": 6.993000901315013e-07,
+      "loss": 0.0375,
+      "step": 3184
+    },
+    {
+      "epoch": 2.448414985590778,
+      "grad_norm": 0.7882428320516394,
+      "learning_rate": 6.955492145102735e-07,
+      "loss": 0.0411,
+      "step": 3186
+    },
+    {
+      "epoch": 2.4499519692603267,
+      "grad_norm": 0.7300434190827599,
+      "learning_rate": 6.918073152111746e-07,
+      "loss": 0.0411,
+      "step": 3188
+    },
+    {
+      "epoch": 2.451488952929875,
+      "grad_norm": 0.7866316609425517,
+      "learning_rate": 6.88074404211058e-07,
+      "loss": 0.0439,
+      "step": 3190
+    },
+    {
+      "epoch": 2.4530259365994236,
+      "grad_norm": 0.7171956812535517,
+      "learning_rate": 6.843504934580055e-07,
+      "loss": 0.0388,
+      "step": 3192
+    },
+    {
+      "epoch": 2.4545629202689723,
+      "grad_norm": 0.704134115079434,
+      "learning_rate": 6.806355948712931e-07,
+      "loss": 0.0366,
+      "step": 3194
+    },
+    {
+      "epoch": 2.4560999039385205,
+      "grad_norm": 0.7589267365486366,
+      "learning_rate": 6.76929720341353e-07,
+      "loss": 0.0402,
+      "step": 3196
+    },
+    {
+      "epoch": 2.457636887608069,
+      "grad_norm": 0.7438363100690424,
+      "learning_rate": 6.732328817297294e-07,
+      "loss": 0.0383,
+      "step": 3198
+    },
+    {
+      "epoch": 2.459173871277618,
+      "grad_norm": 0.7366824621405825,
+      "learning_rate": 6.695450908690482e-07,
+      "loss": 0.0402,
+      "step": 3200
+    },
+    {
+      "epoch": 2.460710854947166,
+      "grad_norm": 0.7557689558570855,
+      "learning_rate": 6.658663595629751e-07,
+      "loss": 0.0386,
+      "step": 3202
+    },
+    {
+      "epoch": 2.462247838616715,
+      "grad_norm": 0.7592608254472845,
+      "learning_rate": 6.621966995861775e-07,
+      "loss": 0.0383,
+      "step": 3204
+    },
+    {
+      "epoch": 2.463784822286263,
+      "grad_norm": 0.7295386054315369,
+      "learning_rate": 6.585361226842899e-07,
+      "loss": 0.0361,
+      "step": 3206
+    },
+    {
+      "epoch": 2.4653218059558117,
+      "grad_norm": 0.6625997315002057,
+      "learning_rate": 6.548846405738713e-07,
+      "loss": 0.038,
+      "step": 3208
+    },
+    {
+      "epoch": 2.46685878962536,
+      "grad_norm": 0.7014497663989586,
+      "learning_rate": 6.512422649423715e-07,
+      "loss": 0.0412,
+      "step": 3210
+    },
+    {
+      "epoch": 2.4683957732949087,
+      "grad_norm": 0.783558414213728,
+      "learning_rate": 6.476090074480966e-07,
+      "loss": 0.0393,
+      "step": 3212
+    },
+    {
+      "epoch": 2.4699327569644574,
+      "grad_norm": 0.7244776560799225,
+      "learning_rate": 6.439848797201623e-07,
+      "loss": 0.038,
+      "step": 3214
+    },
+    {
+      "epoch": 2.4714697406340056,
+      "grad_norm": 0.7224162285528907,
+      "learning_rate": 6.403698933584653e-07,
+      "loss": 0.0409,
+      "step": 3216
+    },
+    {
+      "epoch": 2.4730067243035543,
+      "grad_norm": 0.761865817385147,
+      "learning_rate": 6.367640599336425e-07,
+      "loss": 0.0397,
+      "step": 3218
+    },
+    {
+      "epoch": 2.474543707973103,
+      "grad_norm": 0.6853505678703133,
+      "learning_rate": 6.331673909870353e-07,
+      "loss": 0.0364,
+      "step": 3220
+    },
+    {
+      "epoch": 2.476080691642651,
+      "grad_norm": 0.716004689687549,
+      "learning_rate": 6.295798980306516e-07,
+      "loss": 0.0384,
+      "step": 3222
+    },
+    {
+      "epoch": 2.4776176753122,
+      "grad_norm": 0.6629622601789238,
+      "learning_rate": 6.260015925471279e-07,
+      "loss": 0.0382,
+      "step": 3224
+    },
+    {
+      "epoch": 2.479154658981748,
+      "grad_norm": 0.6579568031971793,
+      "learning_rate": 6.224324859896957e-07,
+      "loss": 0.0355,
+      "step": 3226
+    },
+    {
+      "epoch": 2.480691642651297,
+      "grad_norm": 0.7068591652816046,
+      "learning_rate": 6.18872589782143e-07,
+      "loss": 0.041,
+      "step": 3228
+    },
+    {
+      "epoch": 2.4822286263208455,
+      "grad_norm": 0.7304546722170435,
+      "learning_rate": 6.153219153187772e-07,
+      "loss": 0.04,
+      "step": 3230
+    },
+    {
+      "epoch": 2.4837656099903938,
+      "grad_norm": 0.7585585312793138,
+      "learning_rate": 6.117804739643907e-07,
+      "loss": 0.0394,
+      "step": 3232
+    },
+    {
+      "epoch": 2.4853025936599424,
+      "grad_norm": 0.7054671882965444,
+      "learning_rate": 6.082482770542192e-07,
+      "loss": 0.0378,
+      "step": 3234
+    },
+    {
+      "epoch": 2.4868395773294907,
+      "grad_norm": 0.703977106591607,
+      "learning_rate": 6.047253358939142e-07,
+      "loss": 0.0383,
+      "step": 3236
+    },
+    {
+      "epoch": 2.4883765609990394,
+      "grad_norm": 0.6400130530644246,
+      "learning_rate": 6.01211661759501e-07,
+      "loss": 0.0397,
+      "step": 3238
+    },
+    {
+      "epoch": 2.489913544668588,
+      "grad_norm": 0.7278600446549451,
+      "learning_rate": 5.977072658973393e-07,
+      "loss": 0.0384,
+      "step": 3240
+    },
+    {
+      "epoch": 2.4914505283381363,
+      "grad_norm": 0.746204962644141,
+      "learning_rate": 5.942121595240963e-07,
+      "loss": 0.0362,
+      "step": 3242
+    },
+    {
+      "epoch": 2.492987512007685,
+      "grad_norm": 0.7292907227238307,
+      "learning_rate": 5.907263538267036e-07,
+      "loss": 0.0386,
+      "step": 3244
+    },
+    {
+      "epoch": 2.4945244956772337,
+      "grad_norm": 0.7163392398345111,
+      "learning_rate": 5.872498599623248e-07,
+      "loss": 0.0401,
+      "step": 3246
+    },
+    {
+      "epoch": 2.496061479346782,
+      "grad_norm": 0.7435508674073057,
+      "learning_rate": 5.837826890583188e-07,
+      "loss": 0.0416,
+      "step": 3248
+    },
+    {
+      "epoch": 2.4975984630163306,
+      "grad_norm": 0.72066787754061,
+      "learning_rate": 5.803248522122008e-07,
+      "loss": 0.0394,
+      "step": 3250
+    },
+    {
+      "epoch": 2.499135446685879,
+      "grad_norm": 0.732919013969617,
+      "learning_rate": 5.768763604916157e-07,
+      "loss": 0.0411,
+      "step": 3252
+    },
+    {
+      "epoch": 2.5006724303554275,
+      "grad_norm": 0.7405159195872731,
+      "learning_rate": 5.734372249342942e-07,
+      "loss": 0.0417,
+      "step": 3254
+    },
+    {
+      "epoch": 2.5022094140249758,
+      "grad_norm": 0.7197000994936994,
+      "learning_rate": 5.700074565480184e-07,
+      "loss": 0.041,
+      "step": 3256
+    },
+    {
+      "epoch": 2.5037463976945245,
+      "grad_norm": 0.7015912360666993,
+      "learning_rate": 5.665870663105918e-07,
+      "loss": 0.0384,
+      "step": 3258
+    },
+    {
+      "epoch": 2.505283381364073,
+      "grad_norm": 0.7221454603303137,
+      "learning_rate": 5.631760651697987e-07,
+      "loss": 0.0406,
+      "step": 3260
+    },
+    {
+      "epoch": 2.5068203650336214,
+      "grad_norm": 0.6898827529762818,
+      "learning_rate": 5.59774464043373e-07,
+      "loss": 0.0347,
+      "step": 3262
+    },
+    {
+      "epoch": 2.50835734870317,
+      "grad_norm": 0.7561654186730838,
+      "learning_rate": 5.56382273818961e-07,
+      "loss": 0.0401,
+      "step": 3264
+    },
+    {
+      "epoch": 2.5098943323727188,
+      "grad_norm": 0.7634858564317603,
+      "learning_rate": 5.529995053540845e-07,
+      "loss": 0.0393,
+      "step": 3266
+    },
+    {
+      "epoch": 2.511431316042267,
+      "grad_norm": 0.7573259002994046,
+      "learning_rate": 5.496261694761138e-07,
+      "loss": 0.0369,
+      "step": 3268
+    },
+    {
+      "epoch": 2.5129682997118157,
+      "grad_norm": 0.712091317631427,
+      "learning_rate": 5.46262276982225e-07,
+      "loss": 0.0364,
+      "step": 3270
+    },
+    {
+      "epoch": 2.514505283381364,
+      "grad_norm": 0.7452058266292231,
+      "learning_rate": 5.429078386393659e-07,
+      "loss": 0.0376,
+      "step": 3272
+    },
+    {
+      "epoch": 2.5160422670509126,
+      "grad_norm": 0.6945382662735007,
+      "learning_rate": 5.3956286518423e-07,
+      "loss": 0.0383,
+      "step": 3274
+    },
+    {
+      "epoch": 2.517579250720461,
+      "grad_norm": 0.7542364397801513,
+      "learning_rate": 5.362273673232104e-07,
+      "loss": 0.0404,
+      "step": 3276
+    },
+    {
+      "epoch": 2.5191162343900095,
+      "grad_norm": 0.6802822757529954,
+      "learning_rate": 5.329013557323747e-07,
+      "loss": 0.0373,
+      "step": 3278
+    },
+    {
+      "epoch": 2.520653218059558,
+      "grad_norm": 0.6897346854529715,
+      "learning_rate": 5.295848410574261e-07,
+      "loss": 0.0371,
+      "step": 3280
+    },
+    {
+      "epoch": 2.5221902017291065,
+      "grad_norm": 0.7751087080675236,
+      "learning_rate": 5.262778339136709e-07,
+      "loss": 0.0395,
+      "step": 3282
+    },
+    {
+      "epoch": 2.523727185398655,
+      "grad_norm": 0.739466897171713,
+      "learning_rate": 5.229803448859851e-07,
+      "loss": 0.0387,
+      "step": 3284
+    },
+    {
+      "epoch": 2.525264169068204,
+      "grad_norm": 0.7326229919534164,
+      "learning_rate": 5.196923845287773e-07,
+      "loss": 0.0399,
+      "step": 3286
+    },
+    {
+      "epoch": 2.526801152737752,
+      "grad_norm": 0.6978471766129466,
+      "learning_rate": 5.164139633659586e-07,
+      "loss": 0.0371,
+      "step": 3288
+    },
+    {
+      "epoch": 2.5283381364073008,
+      "grad_norm": 0.7419075355790281,
+      "learning_rate": 5.1314509189091e-07,
+      "loss": 0.0418,
+      "step": 3290
+    },
+    {
+      "epoch": 2.5298751200768494,
+      "grad_norm": 0.6611449090491568,
+      "learning_rate": 5.098857805664424e-07,
+      "loss": 0.0363,
+      "step": 3292
+    },
+    {
+      "epoch": 2.5314121037463977,
+      "grad_norm": 0.6982486243686108,
+      "learning_rate": 5.066360398247698e-07,
+      "loss": 0.0379,
+      "step": 3294
+    },
+    {
+      "epoch": 2.532949087415946,
+      "grad_norm": 0.6911500588630615,
+      "learning_rate": 5.033958800674717e-07,
+      "loss": 0.0367,
+      "step": 3296
+    },
+    {
+      "epoch": 2.5344860710854946,
+      "grad_norm": 0.669502871148484,
+      "learning_rate": 5.001653116654625e-07,
+      "loss": 0.0345,
+      "step": 3298
+    },
+    {
+      "epoch": 2.5360230547550433,
+      "grad_norm": 0.6572589542045425,
+      "learning_rate": 4.969443449589576e-07,
+      "loss": 0.0366,
+      "step": 3300
+    },
+    {
+      "epoch": 2.5375600384245915,
+      "grad_norm": 0.6553120661704981,
+      "learning_rate": 4.937329902574367e-07,
+      "loss": 0.0382,
+      "step": 3302
+    },
+    {
+      "epoch": 2.5390970220941402,
+      "grad_norm": 0.7504177803505651,
+      "learning_rate": 4.90531257839617e-07,
+      "loss": 0.0376,
+      "step": 3304
+    },
+    {
+      "epoch": 2.540634005763689,
+      "grad_norm": 0.6822045806022221,
+      "learning_rate": 4.873391579534186e-07,
+      "loss": 0.0383,
+      "step": 3306
+    },
+    {
+      "epoch": 2.542170989433237,
+      "grad_norm": 0.6780574733579846,
+      "learning_rate": 4.841567008159255e-07,
+      "loss": 0.0352,
+      "step": 3308
+    },
+    {
+      "epoch": 2.543707973102786,
+      "grad_norm": 0.721214268014702,
+      "learning_rate": 4.809838966133612e-07,
+      "loss": 0.0396,
+      "step": 3310
+    },
+    {
+      "epoch": 2.5452449567723345,
+      "grad_norm": 0.7202960873503238,
+      "learning_rate": 4.778207555010525e-07,
+      "loss": 0.043,
+      "step": 3312
+    },
+    {
+      "epoch": 2.5467819404418828,
+      "grad_norm": 0.6736522099205745,
+      "learning_rate": 4.74667287603396e-07,
+      "loss": 0.0379,
+      "step": 3314
+    },
+    {
+      "epoch": 2.5483189241114315,
+      "grad_norm": 0.7223414892006996,
+      "learning_rate": 4.715235030138284e-07,
+      "loss": 0.036,
+      "step": 3316
+    },
+    {
+      "epoch": 2.5498559077809797,
+      "grad_norm": 0.6901701228966389,
+      "learning_rate": 4.6838941179479045e-07,
+      "loss": 0.0389,
+      "step": 3318
+    },
+    {
+      "epoch": 2.5513928914505284,
+      "grad_norm": 0.7617514775980445,
+      "learning_rate": 4.652650239776981e-07,
+      "loss": 0.0399,
+      "step": 3320
+    },
+    {
+      "epoch": 2.5529298751200766,
+      "grad_norm": 0.7419234804032083,
+      "learning_rate": 4.6215034956290966e-07,
+      "loss": 0.0411,
+      "step": 3322
+    },
+    {
+      "epoch": 2.5544668587896253,
+      "grad_norm": 0.7045434950164664,
+      "learning_rate": 4.590453985196933e-07,
+      "loss": 0.0342,
+      "step": 3324
+    },
+    {
+      "epoch": 2.556003842459174,
+      "grad_norm": 0.6495176322545015,
+      "learning_rate": 4.5595018078619483e-07,
+      "loss": 0.0369,
+      "step": 3326
+    },
+    {
+      "epoch": 2.5575408261287222,
+      "grad_norm": 0.6788801032608669,
+      "learning_rate": 4.5286470626940446e-07,
+      "loss": 0.0351,
+      "step": 3328
+    },
+    {
+      "epoch": 2.559077809798271,
+      "grad_norm": 0.7617522250263151,
+      "learning_rate": 4.497889848451303e-07,
+      "loss": 0.0392,
+      "step": 3330
+    },
+    {
+      "epoch": 2.5606147934678196,
+      "grad_norm": 0.6989723366985957,
+      "learning_rate": 4.4672302635796267e-07,
+      "loss": 0.0385,
+      "step": 3332
+    },
+    {
+      "epoch": 2.562151777137368,
+      "grad_norm": 0.7054197171986639,
+      "learning_rate": 4.436668406212402e-07,
+      "loss": 0.0385,
+      "step": 3334
+    },
+    {
+      "epoch": 2.5636887608069165,
+      "grad_norm": 0.6582212133744816,
+      "learning_rate": 4.4062043741702416e-07,
+      "loss": 0.0357,
+      "step": 3336
+    },
+    {
+      "epoch": 2.565225744476465,
+      "grad_norm": 0.7280116384325578,
+      "learning_rate": 4.37583826496064e-07,
+      "loss": 0.0437,
+      "step": 3338
+    },
+    {
+      "epoch": 2.5667627281460135,
+      "grad_norm": 0.8381876496092886,
+      "learning_rate": 4.3455701757776623e-07,
+      "loss": 0.0388,
+      "step": 3340
+    },
+    {
+      "epoch": 2.5682997118155617,
+      "grad_norm": 0.7357236219117014,
+      "learning_rate": 4.315400203501649e-07,
+      "loss": 0.0389,
+      "step": 3342
+    },
+    {
+      "epoch": 2.5698366954851104,
+      "grad_norm": 0.7154920208066208,
+      "learning_rate": 4.2853284446988595e-07,
+      "loss": 0.0394,
+      "step": 3344
+    },
+    {
+      "epoch": 2.571373679154659,
+      "grad_norm": 0.7461232486368099,
+      "learning_rate": 4.2553549956212485e-07,
+      "loss": 0.0378,
+      "step": 3346
+    },
+    {
+      "epoch": 2.5729106628242073,
+      "grad_norm": 0.6390926012256907,
+      "learning_rate": 4.225479952206074e-07,
+      "loss": 0.0344,
+      "step": 3348
+    },
+    {
+      "epoch": 2.574447646493756,
+      "grad_norm": 0.6265363948197101,
+      "learning_rate": 4.195703410075615e-07,
+      "loss": 0.0358,
+      "step": 3350
+    },
+    {
+      "epoch": 2.5759846301633047,
+      "grad_norm": 0.7056231822693347,
+      "learning_rate": 4.1660254645368936e-07,
+      "loss": 0.0383,
+      "step": 3352
+    },
+    {
+      "epoch": 2.577521613832853,
+      "grad_norm": 0.7197705899541366,
+      "learning_rate": 4.1364462105813486e-07,
+      "loss": 0.0382,
+      "step": 3354
+    },
+    {
+      "epoch": 2.5790585975024016,
+      "grad_norm": 0.7234355293425967,
+      "learning_rate": 4.10696574288452e-07,
+      "loss": 0.0362,
+      "step": 3356
+    },
+    {
+      "epoch": 2.5805955811719503,
+      "grad_norm": 0.6700554042034669,
+      "learning_rate": 4.077584155805774e-07,
+      "loss": 0.0383,
+      "step": 3358
+    },
+    {
+      "epoch": 2.5821325648414986,
+      "grad_norm": 0.6989541320884164,
+      "learning_rate": 4.048301543387956e-07,
+      "loss": 0.0334,
+      "step": 3360
+    },
+    {
+      "epoch": 2.583669548511047,
+      "grad_norm": 0.7142263192429504,
+      "learning_rate": 4.0191179993571623e-07,
+      "loss": 0.0408,
+      "step": 3362
+    },
+    {
+      "epoch": 2.5852065321805955,
+      "grad_norm": 0.665689407268323,
+      "learning_rate": 3.990033617122375e-07,
+      "loss": 0.0343,
+      "step": 3364
+    },
+    {
+      "epoch": 2.586743515850144,
+      "grad_norm": 0.6706336215208009,
+      "learning_rate": 3.961048489775169e-07,
+      "loss": 0.0393,
+      "step": 3366
+    },
+    {
+      "epoch": 2.5882804995196924,
+      "grad_norm": 0.746265905426555,
+      "learning_rate": 3.932162710089454e-07,
+      "loss": 0.0403,
+      "step": 3368
+    },
+    {
+      "epoch": 2.589817483189241,
+      "grad_norm": 0.7118457765158708,
+      "learning_rate": 3.903376370521146e-07,
+      "loss": 0.0381,
+      "step": 3370
+    },
+    {
+      "epoch": 2.59135446685879,
+      "grad_norm": 0.6708203366211161,
+      "learning_rate": 3.874689563207873e-07,
+      "loss": 0.0341,
+      "step": 3372
+    },
+    {
+      "epoch": 2.592891450528338,
+      "grad_norm": 0.6330517123604927,
+      "learning_rate": 3.8461023799686917e-07,
+      "loss": 0.0364,
+      "step": 3374
+    },
+    {
+      "epoch": 2.5944284341978867,
+      "grad_norm": 0.7349779518025492,
+      "learning_rate": 3.817614912303794e-07,
+      "loss": 0.0382,
+      "step": 3376
+    },
+    {
+      "epoch": 2.5959654178674354,
+      "grad_norm": 0.704462327359088,
+      "learning_rate": 3.7892272513942003e-07,
+      "loss": 0.0373,
+      "step": 3378
+    },
+    {
+      "epoch": 2.5975024015369836,
+      "grad_norm": 0.6642196959244105,
+      "learning_rate": 3.760939488101463e-07,
+      "loss": 0.036,
+      "step": 3380
+    },
+    {
+      "epoch": 2.5990393852065323,
+      "grad_norm": 0.6891895315956058,
+      "learning_rate": 3.7327517129674055e-07,
+      "loss": 0.0374,
+      "step": 3382
+    },
+    {
+      "epoch": 2.6005763688760806,
+      "grad_norm": 0.7644451152560637,
+      "learning_rate": 3.70466401621383e-07,
+      "loss": 0.0405,
+      "step": 3384
+    },
+    {
+      "epoch": 2.6021133525456293,
+      "grad_norm": 0.7737376511884962,
+      "learning_rate": 3.676676487742172e-07,
+      "loss": 0.0366,
+      "step": 3386
+    },
+    {
+      "epoch": 2.6036503362151775,
+      "grad_norm": 0.7181650532484439,
+      "learning_rate": 3.648789217133282e-07,
+      "loss": 0.0374,
+      "step": 3388
+    },
+    {
+      "epoch": 2.605187319884726,
+      "grad_norm": 0.6915226559770837,
+      "learning_rate": 3.621002293647111e-07,
+      "loss": 0.0368,
+      "step": 3390
+    },
+    {
+      "epoch": 2.606724303554275,
+      "grad_norm": 0.7570072346163113,
+      "learning_rate": 3.593315806222415e-07,
+      "loss": 0.04,
+      "step": 3392
+    },
+    {
+      "epoch": 2.608261287223823,
+      "grad_norm": 0.8257096784505046,
+      "learning_rate": 3.5657298434764884e-07,
+      "loss": 0.0409,
+      "step": 3394
+    },
+    {
+      "epoch": 2.609798270893372,
+      "grad_norm": 0.7210127478796208,
+      "learning_rate": 3.538244493704862e-07,
+      "loss": 0.0382,
+      "step": 3396
+    },
+    {
+      "epoch": 2.6113352545629205,
+      "grad_norm": 0.7234247835643788,
+      "learning_rate": 3.510859844881033e-07,
+      "loss": 0.0374,
+      "step": 3398
+    },
+    {
+      "epoch": 2.6128722382324687,
+      "grad_norm": 0.7384845763851767,
+      "learning_rate": 3.4835759846562047e-07,
+      "loss": 0.0386,
+      "step": 3400
+    },
+    {
+      "epoch": 2.6144092219020174,
+      "grad_norm": 0.7147474025303922,
+      "learning_rate": 3.45639300035894e-07,
+      "loss": 0.0346,
+      "step": 3402
+    },
+    {
+      "epoch": 2.6159462055715657,
+      "grad_norm": 0.6463222792959664,
+      "learning_rate": 3.429310978994955e-07,
+      "loss": 0.0336,
+      "step": 3404
+    },
+    {
+      "epoch": 2.6174831892411143,
+      "grad_norm": 0.7095670666778837,
+      "learning_rate": 3.402330007246798e-07,
+      "loss": 0.0354,
+      "step": 3406
+    },
+    {
+      "epoch": 2.6190201729106626,
+      "grad_norm": 0.6506286288306146,
+      "learning_rate": 3.3754501714735867e-07,
+      "loss": 0.0373,
+      "step": 3408
+    },
+    {
+      "epoch": 2.6205571565802113,
+      "grad_norm": 0.7120588233101057,
+      "learning_rate": 3.34867155771074e-07,
+      "loss": 0.0387,
+      "step": 3410
+    },
+    {
+      "epoch": 2.62209414024976,
+      "grad_norm": 0.7586108600316385,
+      "learning_rate": 3.321994251669659e-07,
+      "loss": 0.0372,
+      "step": 3412
+    },
+    {
+      "epoch": 2.623631123919308,
+      "grad_norm": 0.8046131513589455,
+      "learning_rate": 3.295418338737517e-07,
+      "loss": 0.0394,
+      "step": 3414
+    },
+    {
+      "epoch": 2.625168107588857,
+      "grad_norm": 0.6335632790717333,
+      "learning_rate": 3.2689439039769407e-07,
+      "loss": 0.0344,
+      "step": 3416
+    },
+    {
+      "epoch": 2.6267050912584056,
+      "grad_norm": 0.7948767979535154,
+      "learning_rate": 3.2425710321257503e-07,
+      "loss": 0.0413,
+      "step": 3418
+    },
+    {
+      "epoch": 2.628242074927954,
+      "grad_norm": 0.6525048541754009,
+      "learning_rate": 3.216299807596697e-07,
+      "loss": 0.0364,
+      "step": 3420
+    },
+    {
+      "epoch": 2.6297790585975025,
+      "grad_norm": 0.6986942893445027,
+      "learning_rate": 3.190130314477178e-07,
+      "loss": 0.0424,
+      "step": 3422
+    },
+    {
+      "epoch": 2.631316042267051,
+      "grad_norm": 0.6837105014346364,
+      "learning_rate": 3.164062636528975e-07,
+      "loss": 0.036,
+      "step": 3424
+    },
+    {
+      "epoch": 2.6328530259365994,
+      "grad_norm": 0.6890104339174508,
+      "learning_rate": 3.138096857188001e-07,
+      "loss": 0.0343,
+      "step": 3426
+    },
+    {
+      "epoch": 2.6343900096061477,
+      "grad_norm": 0.7400251284066703,
+      "learning_rate": 3.1122330595639864e-07,
+      "loss": 0.0338,
+      "step": 3428
+    },
+    {
+      "epoch": 2.6359269932756964,
+      "grad_norm": 0.7249635097965398,
+      "learning_rate": 3.0864713264402697e-07,
+      "loss": 0.0409,
+      "step": 3430
+    },
+    {
+      "epoch": 2.637463976945245,
+      "grad_norm": 0.6532133989690416,
+      "learning_rate": 3.060811740273496e-07,
+      "loss": 0.0381,
+      "step": 3432
+    },
+    {
+      "epoch": 2.6390009606147933,
+      "grad_norm": 0.7524963132197209,
+      "learning_rate": 3.03525438319338e-07,
+      "loss": 0.0403,
+      "step": 3434
+    },
+    {
+      "epoch": 2.640537944284342,
+      "grad_norm": 0.7653173809713844,
+      "learning_rate": 3.0097993370024114e-07,
+      "loss": 0.039,
+      "step": 3436
+    },
+    {
+      "epoch": 2.6420749279538907,
+      "grad_norm": 0.7480100060268294,
+      "learning_rate": 2.9844466831756e-07,
+      "loss": 0.0414,
+      "step": 3438
+    },
+    {
+      "epoch": 2.643611911623439,
+      "grad_norm": 0.7160584018035305,
+      "learning_rate": 2.9591965028602584e-07,
+      "loss": 0.0355,
+      "step": 3440
+    },
+    {
+      "epoch": 2.6451488952929876,
+      "grad_norm": 0.7617333555835748,
+      "learning_rate": 2.9340488768756837e-07,
+      "loss": 0.0402,
+      "step": 3442
+    },
+    {
+      "epoch": 2.6466858789625363,
+      "grad_norm": 0.6637519028178263,
+      "learning_rate": 2.909003885712919e-07,
+      "loss": 0.0391,
+      "step": 3444
+    },
+    {
+      "epoch": 2.6482228626320845,
+      "grad_norm": 0.6831014037948032,
+      "learning_rate": 2.8840616095345085e-07,
+      "loss": 0.0362,
+      "step": 3446
+    },
+    {
+      "epoch": 2.649759846301633,
+      "grad_norm": 0.6786000704889272,
+      "learning_rate": 2.859222128174235e-07,
+      "loss": 0.0378,
+      "step": 3448
+    },
+    {
+      "epoch": 2.6512968299711814,
+      "grad_norm": 0.8326763730243029,
+      "learning_rate": 2.83448552113686e-07,
+      "loss": 0.0405,
+      "step": 3450
+    },
+    {
+      "epoch": 2.65283381364073,
+      "grad_norm": 0.7015006547909788,
+      "learning_rate": 2.809851867597875e-07,
+      "loss": 0.0367,
+      "step": 3452
+    },
+    {
+      "epoch": 2.6543707973102784,
+      "grad_norm": 0.710615174193323,
+      "learning_rate": 2.7853212464032146e-07,
+      "loss": 0.041,
+      "step": 3454
+    },
+    {
+      "epoch": 2.655907780979827,
+      "grad_norm": 0.6849647259800099,
+      "learning_rate": 2.7608937360690814e-07,
+      "loss": 0.0363,
+      "step": 3456
+    },
+    {
+      "epoch": 2.6574447646493757,
+      "grad_norm": 0.7276694595529292,
+      "learning_rate": 2.736569414781617e-07,
+      "loss": 0.0383,
+      "step": 3458
+    },
+    {
+      "epoch": 2.658981748318924,
+      "grad_norm": 0.774608853311179,
+      "learning_rate": 2.7123483603966824e-07,
+      "loss": 0.0402,
+      "step": 3460
+    },
+    {
+      "epoch": 2.6605187319884727,
+      "grad_norm": 0.7083364735387814,
+      "learning_rate": 2.6882306504396143e-07,
+      "loss": 0.0401,
+      "step": 3462
+    },
+    {
+      "epoch": 2.6620557156580213,
+      "grad_norm": 0.6714099006775984,
+      "learning_rate": 2.664216362104964e-07,
+      "loss": 0.0354,
+      "step": 3464
+    },
+    {
+      "epoch": 2.6635926993275696,
+      "grad_norm": 0.6437816346046543,
+      "learning_rate": 2.64030557225627e-07,
+      "loss": 0.035,
+      "step": 3466
+    },
+    {
+      "epoch": 2.6651296829971183,
+      "grad_norm": 0.7386329454344525,
+      "learning_rate": 2.6164983574257875e-07,
+      "loss": 0.0411,
+      "step": 3468
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.7173981240012108,
+      "learning_rate": 2.592794793814257e-07,
+      "loss": 0.0351,
+      "step": 3470
+    },
+    {
+      "epoch": 2.668203650336215,
+      "grad_norm": 0.7373052648241346,
+      "learning_rate": 2.569194957290666e-07,
+      "loss": 0.0367,
+      "step": 3472
+    },
+    {
+      "epoch": 2.6697406340057634,
+      "grad_norm": 0.7328221093520572,
+      "learning_rate": 2.5456989233919775e-07,
+      "loss": 0.0395,
+      "step": 3474
+    },
+    {
+      "epoch": 2.671277617675312,
+      "grad_norm": 0.6381830501668562,
+      "learning_rate": 2.5223067673229285e-07,
+      "loss": 0.0336,
+      "step": 3476
+    },
+    {
+      "epoch": 2.672814601344861,
+      "grad_norm": 0.7536818768122339,
+      "learning_rate": 2.4990185639557777e-07,
+      "loss": 0.0398,
+      "step": 3478
+    },
+    {
+      "epoch": 2.674351585014409,
+      "grad_norm": 0.7737529199470804,
+      "learning_rate": 2.475834387830031e-07,
+      "loss": 0.0379,
+      "step": 3480
+    },
+    {
+      "epoch": 2.6758885686839577,
+      "grad_norm": 0.7616061727604434,
+      "learning_rate": 2.452754313152246e-07,
+      "loss": 0.0403,
+      "step": 3482
+    },
+    {
+      "epoch": 2.6774255523535064,
+      "grad_norm": 0.7209975392281303,
+      "learning_rate": 2.429778413795775e-07,
+      "loss": 0.0383,
+      "step": 3484
+    },
+    {
+      "epoch": 2.6789625360230547,
+      "grad_norm": 0.6777003415447909,
+      "learning_rate": 2.4069067633005426e-07,
+      "loss": 0.0397,
+      "step": 3486
+    },
+    {
+      "epoch": 2.6804995196926034,
+      "grad_norm": 0.7469038413792495,
+      "learning_rate": 2.3841394348727856e-07,
+      "loss": 0.0392,
+      "step": 3488
+    },
+    {
+      "epoch": 2.682036503362152,
+      "grad_norm": 0.7456099202324231,
+      "learning_rate": 2.3614765013848365e-07,
+      "loss": 0.0389,
+      "step": 3490
+    },
+    {
+      "epoch": 2.6835734870317003,
+      "grad_norm": 0.758937959458986,
+      "learning_rate": 2.338918035374881e-07,
+      "loss": 0.0381,
+      "step": 3492
+    },
+    {
+      "epoch": 2.6851104707012485,
+      "grad_norm": 0.7223222799308419,
+      "learning_rate": 2.3164641090467535e-07,
+      "loss": 0.0399,
+      "step": 3494
+    },
+    {
+      "epoch": 2.686647454370797,
+      "grad_norm": 0.750470365319683,
+      "learning_rate": 2.29411479426966e-07,
+      "loss": 0.0389,
+      "step": 3496
+    },
+    {
+      "epoch": 2.688184438040346,
+      "grad_norm": 0.8119786466865458,
+      "learning_rate": 2.27187016257798e-07,
+      "loss": 0.0394,
+      "step": 3498
+    },
+    {
+      "epoch": 2.689721421709894,
+      "grad_norm": 0.7063154674095581,
+      "learning_rate": 2.2497302851710354e-07,
+      "loss": 0.0366,
+      "step": 3500
+    },
+    {
+      "epoch": 2.689721421709894,
+      "eval_loss": 0.15855057537555695,
+      "eval_runtime": 360.8492,
+      "eval_samples_per_second": 51.282,
+      "eval_steps_per_second": 6.413,
+      "step": 3500
+    },
+    {
+      "epoch": 2.691258405379443,
+      "grad_norm": 0.7006450588051517,
+      "learning_rate": 2.2276952329128462e-07,
+      "loss": 0.0357,
+      "step": 3502
+    },
+    {
+      "epoch": 2.6927953890489915,
+      "grad_norm": 0.6689362757218708,
+      "learning_rate": 2.2057650763319235e-07,
+      "loss": 0.0377,
+      "step": 3504
+    },
+    {
+      "epoch": 2.6943323727185398,
+      "grad_norm": 0.7529959367992736,
+      "learning_rate": 2.1839398856210274e-07,
+      "loss": 0.0381,
+      "step": 3506
+    },
+    {
+      "epoch": 2.6958693563880884,
+      "grad_norm": 0.7460063989500526,
+      "learning_rate": 2.1622197306369529e-07,
+      "loss": 0.0403,
+      "step": 3508
+    },
+    {
+      "epoch": 2.697406340057637,
+      "grad_norm": 0.7548043108085488,
+      "learning_rate": 2.1406046809003016e-07,
+      "loss": 0.0382,
+      "step": 3510
+    },
+    {
+      "epoch": 2.6989433237271854,
+      "grad_norm": 0.7421499181728671,
+      "learning_rate": 2.1190948055952634e-07,
+      "loss": 0.0382,
+      "step": 3512
+    },
+    {
+      "epoch": 2.700480307396734,
+      "grad_norm": 0.712291075185728,
+      "learning_rate": 2.097690173569392e-07,
+      "loss": 0.0374,
+      "step": 3514
+    },
+    {
+      "epoch": 2.7020172910662823,
+      "grad_norm": 0.6764437704157867,
+      "learning_rate": 2.0763908533333742e-07,
+      "loss": 0.0358,
+      "step": 3516
+    },
+    {
+      "epoch": 2.703554274735831,
+      "grad_norm": 0.7710410887906691,
+      "learning_rate": 2.0551969130608366e-07,
+      "loss": 0.0405,
+      "step": 3518
+    },
+    {
+      "epoch": 2.7050912584053792,
+      "grad_norm": 0.6628607572891497,
+      "learning_rate": 2.0341084205881088e-07,
+      "loss": 0.0334,
+      "step": 3520
+    },
+    {
+      "epoch": 2.706628242074928,
+      "grad_norm": 0.7047102030885535,
+      "learning_rate": 2.0131254434139894e-07,
+      "loss": 0.0364,
+      "step": 3522
+    },
+    {
+      "epoch": 2.7081652257444766,
+      "grad_norm": 0.7511681830084664,
+      "learning_rate": 1.992248048699576e-07,
+      "loss": 0.0423,
+      "step": 3524
+    },
+    {
+      "epoch": 2.709702209414025,
+      "grad_norm": 0.6655250328949831,
+      "learning_rate": 1.971476303268007e-07,
+      "loss": 0.0355,
+      "step": 3526
+    },
+    {
+      "epoch": 2.7112391930835735,
+      "grad_norm": 0.7160428780372823,
+      "learning_rate": 1.950810273604274e-07,
+      "loss": 0.0375,
+      "step": 3528
+    },
+    {
+      "epoch": 2.712776176753122,
+      "grad_norm": 0.7004993169642538,
+      "learning_rate": 1.930250025855006e-07,
+      "loss": 0.0381,
+      "step": 3530
+    },
+    {
+      "epoch": 2.7143131604226705,
+      "grad_norm": 0.6778326775797812,
+      "learning_rate": 1.90979562582822e-07,
+      "loss": 0.0359,
+      "step": 3532
+    },
+    {
+      "epoch": 2.715850144092219,
+      "grad_norm": 0.7164337600080629,
+      "learning_rate": 1.8894471389931838e-07,
+      "loss": 0.0373,
+      "step": 3534
+    },
+    {
+      "epoch": 2.7173871277617674,
+      "grad_norm": 0.7141990651264797,
+      "learning_rate": 1.869204630480142e-07,
+      "loss": 0.0375,
+      "step": 3536
+    },
+    {
+      "epoch": 2.718924111431316,
+      "grad_norm": 0.7001303648061724,
+      "learning_rate": 1.8490681650801216e-07,
+      "loss": 0.0355,
+      "step": 3538
+    },
+    {
+      "epoch": 2.7204610951008643,
+      "grad_norm": 0.7165227897430166,
+      "learning_rate": 1.829037807244751e-07,
+      "loss": 0.0361,
+      "step": 3540
+    },
+    {
+      "epoch": 2.721998078770413,
+      "grad_norm": 0.7103603815004305,
+      "learning_rate": 1.8091136210860293e-07,
+      "loss": 0.0394,
+      "step": 3542
+    },
+    {
+      "epoch": 2.7235350624399617,
+      "grad_norm": 0.7231011724413698,
+      "learning_rate": 1.789295670376128e-07,
+      "loss": 0.0408,
+      "step": 3544
+    },
+    {
+      "epoch": 2.72507204610951,
+      "grad_norm": 0.7187891219293795,
+      "learning_rate": 1.7695840185471828e-07,
+      "loss": 0.0353,
+      "step": 3546
+    },
+    {
+      "epoch": 2.7266090297790586,
+      "grad_norm": 0.7249222842840717,
+      "learning_rate": 1.7499787286910896e-07,
+      "loss": 0.0383,
+      "step": 3548
+    },
+    {
+      "epoch": 2.7281460134486073,
+      "grad_norm": 0.6726610987757241,
+      "learning_rate": 1.7304798635593227e-07,
+      "loss": 0.0345,
+      "step": 3550
+    },
+    {
+      "epoch": 2.7296829971181555,
+      "grad_norm": 0.7798682542487884,
+      "learning_rate": 1.711087485562714e-07,
+      "loss": 0.0395,
+      "step": 3552
+    },
+    {
+      "epoch": 2.7312199807877042,
+      "grad_norm": 0.8489020328845095,
+      "learning_rate": 1.6918016567712457e-07,
+      "loss": 0.0388,
+      "step": 3554
+    },
+    {
+      "epoch": 2.732756964457253,
+      "grad_norm": 0.7675735405799808,
+      "learning_rate": 1.672622438913869e-07,
+      "loss": 0.0413,
+      "step": 3556
+    },
+    {
+      "epoch": 2.734293948126801,
+      "grad_norm": 0.7672370602760468,
+      "learning_rate": 1.6535498933783083e-07,
+      "loss": 0.0392,
+      "step": 3558
+    },
+    {
+      "epoch": 2.7358309317963494,
+      "grad_norm": 0.706536918085323,
+      "learning_rate": 1.634584081210853e-07,
+      "loss": 0.0353,
+      "step": 3560
+    },
+    {
+      "epoch": 2.737367915465898,
+      "grad_norm": 0.6514716793107945,
+      "learning_rate": 1.6157250631161624e-07,
+      "loss": 0.029,
+      "step": 3562
+    },
+    {
+      "epoch": 2.7389048991354468,
+      "grad_norm": 0.8151870302671732,
+      "learning_rate": 1.5969728994570786e-07,
+      "loss": 0.0451,
+      "step": 3564
+    },
+    {
+      "epoch": 2.740441882804995,
+      "grad_norm": 0.7201013965367818,
+      "learning_rate": 1.5783276502544413e-07,
+      "loss": 0.0375,
+      "step": 3566
+    },
+    {
+      "epoch": 2.7419788664745437,
+      "grad_norm": 0.8279664659096065,
+      "learning_rate": 1.5597893751868574e-07,
+      "loss": 0.0377,
+      "step": 3568
+    },
+    {
+      "epoch": 2.7435158501440924,
+      "grad_norm": 0.7556263425034218,
+      "learning_rate": 1.541358133590562e-07,
+      "loss": 0.0383,
+      "step": 3570
+    },
+    {
+      "epoch": 2.7450528338136406,
+      "grad_norm": 0.786743386378996,
+      "learning_rate": 1.5230339844592033e-07,
+      "loss": 0.0389,
+      "step": 3572
+    },
+    {
+      "epoch": 2.7465898174831893,
+      "grad_norm": 0.6751334601351806,
+      "learning_rate": 1.504816986443635e-07,
+      "loss": 0.0348,
+      "step": 3574
+    },
+    {
+      "epoch": 2.748126801152738,
+      "grad_norm": 0.7230502301443176,
+      "learning_rate": 1.4867071978517626e-07,
+      "loss": 0.0381,
+      "step": 3576
+    },
+    {
+      "epoch": 2.7496637848222862,
+      "grad_norm": 0.7424159284142868,
+      "learning_rate": 1.4687046766483425e-07,
+      "loss": 0.0377,
+      "step": 3578
+    },
+    {
+      "epoch": 2.751200768491835,
+      "grad_norm": 0.7033065172321098,
+      "learning_rate": 1.450809480454787e-07,
+      "loss": 0.0392,
+      "step": 3580
+    },
+    {
+      "epoch": 2.752737752161383,
+      "grad_norm": 0.7422740281311779,
+      "learning_rate": 1.4330216665490024e-07,
+      "loss": 0.0388,
+      "step": 3582
+    },
+    {
+      "epoch": 2.754274735830932,
+      "grad_norm": 0.6749501870256193,
+      "learning_rate": 1.4153412918651736e-07,
+      "loss": 0.0292,
+      "step": 3584
+    },
+    {
+      "epoch": 2.75581171950048,
+      "grad_norm": 0.7383402984400056,
+      "learning_rate": 1.3977684129936095e-07,
+      "loss": 0.0362,
+      "step": 3586
+    },
+    {
+      "epoch": 2.757348703170029,
+      "grad_norm": 0.7344177105323607,
+      "learning_rate": 1.3803030861805686e-07,
+      "loss": 0.0369,
+      "step": 3588
+    },
+    {
+      "epoch": 2.7588856868395775,
+      "grad_norm": 0.7788733949519481,
+      "learning_rate": 1.3629453673280367e-07,
+      "loss": 0.0386,
+      "step": 3590
+    },
+    {
+      "epoch": 2.7604226705091257,
+      "grad_norm": 0.7523955526390752,
+      "learning_rate": 1.3456953119935832e-07,
+      "loss": 0.0402,
+      "step": 3592
+    },
+    {
+      "epoch": 2.7619596541786744,
+      "grad_norm": 0.6387136592938574,
+      "learning_rate": 1.328552975390179e-07,
+      "loss": 0.0366,
+      "step": 3594
+    },
+    {
+      "epoch": 2.763496637848223,
+      "grad_norm": 0.7765496538466939,
+      "learning_rate": 1.3115184123860141e-07,
+      "loss": 0.038,
+      "step": 3596
+    },
+    {
+      "epoch": 2.7650336215177713,
+      "grad_norm": 0.6197630907677361,
+      "learning_rate": 1.2945916775043285e-07,
+      "loss": 0.0347,
+      "step": 3598
+    },
+    {
+      "epoch": 2.76657060518732,
+      "grad_norm": 0.6507185122236504,
+      "learning_rate": 1.2777728249232174e-07,
+      "loss": 0.0366,
+      "step": 3600
+    },
+    {
+      "epoch": 2.7681075888568683,
+      "grad_norm": 0.7129555271323699,
+      "learning_rate": 1.261061908475476e-07,
+      "loss": 0.0382,
+      "step": 3602
+    },
+    {
+      "epoch": 2.769644572526417,
+      "grad_norm": 0.7063456528141899,
+      "learning_rate": 1.244458981648443e-07,
+      "loss": 0.0357,
+      "step": 3604
+    },
+    {
+      "epoch": 2.771181556195965,
+      "grad_norm": 0.7313820214800656,
+      "learning_rate": 1.2279640975837857e-07,
+      "loss": 0.0361,
+      "step": 3606
+    },
+    {
+      "epoch": 2.772718539865514,
+      "grad_norm": 0.7892493813976871,
+      "learning_rate": 1.2115773090773718e-07,
+      "loss": 0.0383,
+      "step": 3608
+    },
+    {
+      "epoch": 2.7742555235350626,
+      "grad_norm": 0.7011080024188364,
+      "learning_rate": 1.1952986685790674e-07,
+      "loss": 0.0348,
+      "step": 3610
+    },
+    {
+      "epoch": 2.775792507204611,
+      "grad_norm": 0.7212790327473237,
+      "learning_rate": 1.1791282281925968e-07,
+      "loss": 0.0386,
+      "step": 3612
+    },
+    {
+      "epoch": 2.7773294908741595,
+      "grad_norm": 0.6824883473362624,
+      "learning_rate": 1.163066039675369e-07,
+      "loss": 0.0351,
+      "step": 3614
+    },
+    {
+      "epoch": 2.778866474543708,
+      "grad_norm": 0.6850223544852229,
+      "learning_rate": 1.1471121544382851e-07,
+      "loss": 0.0356,
+      "step": 3616
+    },
+    {
+      "epoch": 2.7804034582132564,
+      "grad_norm": 0.7426536823064712,
+      "learning_rate": 1.131266623545612e-07,
+      "loss": 0.0375,
+      "step": 3618
+    },
+    {
+      "epoch": 2.781940441882805,
+      "grad_norm": 0.7258544058495259,
+      "learning_rate": 1.1155294977148028e-07,
+      "loss": 0.0364,
+      "step": 3620
+    },
+    {
+      "epoch": 2.783477425552354,
+      "grad_norm": 0.7068237112208213,
+      "learning_rate": 1.0999008273163257e-07,
+      "loss": 0.0427,
+      "step": 3622
+    },
+    {
+      "epoch": 2.785014409221902,
+      "grad_norm": 0.723771053564773,
+      "learning_rate": 1.0843806623735207e-07,
+      "loss": 0.0378,
+      "step": 3624
+    },
+    {
+      "epoch": 2.7865513928914503,
+      "grad_norm": 0.8519854888198337,
+      "learning_rate": 1.0689690525624167e-07,
+      "loss": 0.04,
+      "step": 3626
+    },
+    {
+      "epoch": 2.788088376560999,
+      "grad_norm": 0.706127807785433,
+      "learning_rate": 1.0536660472115993e-07,
+      "loss": 0.0393,
+      "step": 3628
+    },
+    {
+      "epoch": 2.7896253602305476,
+      "grad_norm": 0.6705497949051035,
+      "learning_rate": 1.0384716953020323e-07,
+      "loss": 0.0333,
+      "step": 3630
+    },
+    {
+      "epoch": 2.791162343900096,
+      "grad_norm": 0.6886374108827851,
+      "learning_rate": 1.0233860454669097e-07,
+      "loss": 0.0354,
+      "step": 3632
+    },
+    {
+      "epoch": 2.7926993275696446,
+      "grad_norm": 0.6453996780069803,
+      "learning_rate": 1.0084091459914929e-07,
+      "loss": 0.0333,
+      "step": 3634
+    },
+    {
+      "epoch": 2.7942363112391932,
+      "grad_norm": 0.7016064494453276,
+      "learning_rate": 9.935410448129705e-08,
+      "loss": 0.0388,
+      "step": 3636
+    },
+    {
+      "epoch": 2.7957732949087415,
+      "grad_norm": 0.7099217898470163,
+      "learning_rate": 9.787817895202876e-08,
+      "loss": 0.0341,
+      "step": 3638
+    },
+    {
+      "epoch": 2.79731027857829,
+      "grad_norm": 0.7392413254776985,
+      "learning_rate": 9.641314273540175e-08,
+      "loss": 0.0412,
+      "step": 3640
+    },
+    {
+      "epoch": 2.798847262247839,
+      "grad_norm": 0.6659018777001534,
+      "learning_rate": 9.495900052061629e-08,
+      "loss": 0.036,
+      "step": 3642
+    },
+    {
+      "epoch": 2.800384245917387,
+      "grad_norm": 0.7498210451825361,
+      "learning_rate": 9.351575696200754e-08,
+      "loss": 0.0379,
+      "step": 3644
+    },
+    {
+      "epoch": 2.801921229586936,
+      "grad_norm": 0.7188638489637743,
+      "learning_rate": 9.208341667902487e-08,
+      "loss": 0.0349,
+      "step": 3646
+    },
+    {
+      "epoch": 2.803458213256484,
+      "grad_norm": 0.6975773584570585,
+      "learning_rate": 9.06619842562183e-08,
+      "loss": 0.0323,
+      "step": 3648
+    },
+    {
+      "epoch": 2.8049951969260327,
+      "grad_norm": 0.6554005252536124,
+      "learning_rate": 8.925146424322644e-08,
+      "loss": 0.0329,
+      "step": 3650
+    },
+    {
+      "epoch": 2.806532180595581,
+      "grad_norm": 0.7073740246874296,
+      "learning_rate": 8.78518611547594e-08,
+      "loss": 0.0386,
+      "step": 3652
+    },
+    {
+      "epoch": 2.8080691642651296,
+      "grad_norm": 0.6691649384368833,
+      "learning_rate": 8.64631794705844e-08,
+      "loss": 0.0399,
+      "step": 3654
+    },
+    {
+      "epoch": 2.8096061479346783,
+      "grad_norm": 0.7533417497153267,
+      "learning_rate": 8.508542363551296e-08,
+      "loss": 0.0376,
+      "step": 3656
+    },
+    {
+      "epoch": 2.8111431316042266,
+      "grad_norm": 0.7324347703672368,
+      "learning_rate": 8.371859805938497e-08,
+      "loss": 0.0364,
+      "step": 3658
+    },
+    {
+      "epoch": 2.8126801152737753,
+      "grad_norm": 0.6931453128827018,
+      "learning_rate": 8.236270711705662e-08,
+      "loss": 0.0326,
+      "step": 3660
+    },
+    {
+      "epoch": 2.814217098943324,
+      "grad_norm": 0.6612068285353259,
+      "learning_rate": 8.101775514838372e-08,
+      "loss": 0.0374,
+      "step": 3662
+    },
+    {
+      "epoch": 2.815754082612872,
+      "grad_norm": 0.6742383780806924,
+      "learning_rate": 7.968374645820964e-08,
+      "loss": 0.0362,
+      "step": 3664
+    },
+    {
+      "epoch": 2.817291066282421,
+      "grad_norm": 0.7414092505129248,
+      "learning_rate": 7.836068531635249e-08,
+      "loss": 0.0403,
+      "step": 3666
+    },
+    {
+      "epoch": 2.818828049951969,
+      "grad_norm": 0.6917889501190945,
+      "learning_rate": 7.704857595758802e-08,
+      "loss": 0.036,
+      "step": 3668
+    },
+    {
+      "epoch": 2.820365033621518,
+      "grad_norm": 0.7254060046131036,
+      "learning_rate": 7.574742258163952e-08,
+      "loss": 0.0403,
+      "step": 3670
+    },
+    {
+      "epoch": 2.821902017291066,
+      "grad_norm": 0.7112553826453433,
+      "learning_rate": 7.445722935316307e-08,
+      "loss": 0.0392,
+      "step": 3672
+    },
+    {
+      "epoch": 2.8234390009606147,
+      "grad_norm": 0.7231907969490486,
+      "learning_rate": 7.317800040173311e-08,
+      "loss": 0.0388,
+      "step": 3674
+    },
+    {
+      "epoch": 2.8249759846301634,
+      "grad_norm": 0.7143348806863665,
+      "learning_rate": 7.190973982183124e-08,
+      "loss": 0.0356,
+      "step": 3676
+    },
+    {
+      "epoch": 2.8265129682997117,
+      "grad_norm": 0.722256824360624,
+      "learning_rate": 7.065245167283179e-08,
+      "loss": 0.0365,
+      "step": 3678
+    },
+    {
+      "epoch": 2.8280499519692603,
+      "grad_norm": 0.749657038879464,
+      "learning_rate": 6.940613997898826e-08,
+      "loss": 0.0377,
+      "step": 3680
+    },
+    {
+      "epoch": 2.829586935638809,
+      "grad_norm": 0.7506478416183923,
+      "learning_rate": 6.817080872942393e-08,
+      "loss": 0.039,
+      "step": 3682
+    },
+    {
+      "epoch": 2.8311239193083573,
+      "grad_norm": 0.7166949562340486,
+      "learning_rate": 6.694646187811371e-08,
+      "loss": 0.0381,
+      "step": 3684
+    },
+    {
+      "epoch": 2.832660902977906,
+      "grad_norm": 0.6919045066011337,
+      "learning_rate": 6.573310334387544e-08,
+      "loss": 0.0359,
+      "step": 3686
+    },
+    {
+      "epoch": 2.8341978866474546,
+      "grad_norm": 0.7652082696602207,
+      "learning_rate": 6.453073701035644e-08,
+      "loss": 0.0387,
+      "step": 3688
+    },
+    {
+      "epoch": 2.835734870317003,
+      "grad_norm": 0.6875494379565337,
+      "learning_rate": 6.333936672602058e-08,
+      "loss": 0.0358,
+      "step": 3690
+    },
+    {
+      "epoch": 2.837271853986551,
+      "grad_norm": 0.704153070629626,
+      "learning_rate": 6.215899630413668e-08,
+      "loss": 0.0399,
+      "step": 3692
+    },
+    {
+      "epoch": 2.8388088376561,
+      "grad_norm": 0.7683845653293229,
+      "learning_rate": 6.098962952276449e-08,
+      "loss": 0.0376,
+      "step": 3694
+    },
+    {
+      "epoch": 2.8403458213256485,
+      "grad_norm": 0.7444818377336334,
+      "learning_rate": 5.983127012474498e-08,
+      "loss": 0.0375,
+      "step": 3696
+    },
+    {
+      "epoch": 2.8418828049951967,
+      "grad_norm": 0.7131478664262563,
+      "learning_rate": 5.8683921817687943e-08,
+      "loss": 0.0377,
+      "step": 3698
+    },
+    {
+      "epoch": 2.8434197886647454,
+      "grad_norm": 0.6844062973629339,
+      "learning_rate": 5.7547588273958336e-08,
+      "loss": 0.0348,
+      "step": 3700
+    },
+    {
+      "epoch": 2.844956772334294,
+      "grad_norm": 0.6917343381332636,
+      "learning_rate": 5.6422273130665835e-08,
+      "loss": 0.0378,
+      "step": 3702
+    },
+    {
+      "epoch": 2.8464937560038424,
+      "grad_norm": 0.7486578311506366,
+      "learning_rate": 5.5307979989653534e-08,
+      "loss": 0.041,
+      "step": 3704
+    },
+    {
+      "epoch": 2.848030739673391,
+      "grad_norm": 0.7065517002670538,
+      "learning_rate": 5.420471241748592e-08,
+      "loss": 0.0344,
+      "step": 3706
+    },
+    {
+      "epoch": 2.8495677233429397,
+      "grad_norm": 0.7218309216210936,
+      "learning_rate": 5.311247394543761e-08,
+      "loss": 0.0396,
+      "step": 3708
+    },
+    {
+      "epoch": 2.851104707012488,
+      "grad_norm": 0.7344762397348806,
+      "learning_rate": 5.203126806948127e-08,
+      "loss": 0.0398,
+      "step": 3710
+    },
+    {
+      "epoch": 2.8526416906820367,
+      "grad_norm": 0.6978785514396233,
+      "learning_rate": 5.0961098250277166e-08,
+      "loss": 0.035,
+      "step": 3712
+    },
+    {
+      "epoch": 2.854178674351585,
+      "grad_norm": 0.7492370044630394,
+      "learning_rate": 4.990196791316304e-08,
+      "loss": 0.0395,
+      "step": 3714
+    },
+    {
+      "epoch": 2.8557156580211336,
+      "grad_norm": 0.7347118419552384,
+      "learning_rate": 4.8853880448140876e-08,
+      "loss": 0.0373,
+      "step": 3716
+    },
+    {
+      "epoch": 2.857252641690682,
+      "grad_norm": 0.6710823443691567,
+      "learning_rate": 4.781683920986801e-08,
+      "loss": 0.0353,
+      "step": 3718
+    },
+    {
+      "epoch": 2.8587896253602305,
+      "grad_norm": 0.7656067717475032,
+      "learning_rate": 4.679084751764467e-08,
+      "loss": 0.0405,
+      "step": 3720
+    },
+    {
+      "epoch": 2.860326609029779,
+      "grad_norm": 0.6658481972086318,
+      "learning_rate": 4.5775908655405814e-08,
+      "loss": 0.0392,
+      "step": 3722
+    },
+    {
+      "epoch": 2.8618635926993274,
+      "grad_norm": 0.7325672821180539,
+      "learning_rate": 4.4772025871709085e-08,
+      "loss": 0.0357,
+      "step": 3724
+    },
+    {
+      "epoch": 2.863400576368876,
+      "grad_norm": 0.7131675994968449,
+      "learning_rate": 4.377920237972238e-08,
+      "loss": 0.0391,
+      "step": 3726
+    },
+    {
+      "epoch": 2.864937560038425,
+      "grad_norm": 0.735115333347456,
+      "learning_rate": 4.279744135721764e-08,
+      "loss": 0.0383,
+      "step": 3728
+    },
+    {
+      "epoch": 2.866474543707973,
+      "grad_norm": 0.7142932075003074,
+      "learning_rate": 4.182674594655839e-08,
+      "loss": 0.0347,
+      "step": 3730
+    },
+    {
+      "epoch": 2.8680115273775217,
+      "grad_norm": 0.7488967955446748,
+      "learning_rate": 4.0867119254689664e-08,
+      "loss": 0.0387,
+      "step": 3732
+    },
+    {
+      "epoch": 2.86954851104707,
+      "grad_norm": 0.7704536931618361,
+      "learning_rate": 3.991856435312868e-08,
+      "loss": 0.0399,
+      "step": 3734
+    },
+    {
+      "epoch": 2.8710854947166187,
+      "grad_norm": 0.7243897987142581,
+      "learning_rate": 3.898108427795355e-08,
+      "loss": 0.0407,
+      "step": 3736
+    },
+    {
+      "epoch": 2.872622478386167,
+      "grad_norm": 0.7696208552815814,
+      "learning_rate": 3.805468202979706e-08,
+      "loss": 0.0377,
+      "step": 3738
+    },
+    {
+      "epoch": 2.8741594620557156,
+      "grad_norm": 0.732607042331649,
+      "learning_rate": 3.7139360573832715e-08,
+      "loss": 0.0398,
+      "step": 3740
+    },
+    {
+      "epoch": 2.8756964457252643,
+      "grad_norm": 0.6798024620244276,
+      "learning_rate": 3.6235122839767707e-08,
+      "loss": 0.0338,
+      "step": 3742
+    },
+    {
+      "epoch": 2.8772334293948125,
+      "grad_norm": 0.7560881846201208,
+      "learning_rate": 3.534197172183323e-08,
+      "loss": 0.0385,
+      "step": 3744
+    },
+    {
+      "epoch": 2.878770413064361,
+      "grad_norm": 0.7216366213167684,
+      "learning_rate": 3.4459910078775914e-08,
+      "loss": 0.0404,
+      "step": 3746
+    },
+    {
+      "epoch": 2.88030739673391,
+      "grad_norm": 0.7186180485084087,
+      "learning_rate": 3.358894073384616e-08,
+      "loss": 0.0388,
+      "step": 3748
+    },
+    {
+      "epoch": 2.881844380403458,
+      "grad_norm": 0.755800097162076,
+      "learning_rate": 3.2729066474792734e-08,
+      "loss": 0.0355,
+      "step": 3750
+    },
+    {
+      "epoch": 2.883381364073007,
+      "grad_norm": 0.7190348005662992,
+      "learning_rate": 3.18802900538499e-08,
+      "loss": 0.0387,
+      "step": 3752
+    },
+    {
+      "epoch": 2.8849183477425555,
+      "grad_norm": 0.6881759771848993,
+      "learning_rate": 3.104261418773241e-08,
+      "loss": 0.0371,
+      "step": 3754
+    },
+    {
+      "epoch": 2.8864553314121038,
+      "grad_norm": 0.7368031850460476,
+      "learning_rate": 3.0216041557624196e-08,
+      "loss": 0.0363,
+      "step": 3756
+    },
+    {
+      "epoch": 2.887992315081652,
+      "grad_norm": 0.6818452373614833,
+      "learning_rate": 2.9400574809169856e-08,
+      "loss": 0.0385,
+      "step": 3758
+    },
+    {
+      "epoch": 2.8895292987512007,
+      "grad_norm": 0.671665343657444,
+      "learning_rate": 2.859621655246841e-08,
+      "loss": 0.0394,
+      "step": 3760
+    },
+    {
+      "epoch": 2.8910662824207494,
+      "grad_norm": 0.8081746540797051,
+      "learning_rate": 2.7802969362062057e-08,
+      "loss": 0.0426,
+      "step": 3762
+    },
+    {
+      "epoch": 2.8926032660902976,
+      "grad_norm": 0.7953691794225282,
+      "learning_rate": 2.702083577693071e-08,
+      "loss": 0.0401,
+      "step": 3764
+    },
+    {
+      "epoch": 2.8941402497598463,
+      "grad_norm": 0.6780256825935983,
+      "learning_rate": 2.624981830048151e-08,
+      "loss": 0.038,
+      "step": 3766
+    },
+    {
+      "epoch": 2.895677233429395,
+      "grad_norm": 0.7601952372611993,
+      "learning_rate": 2.5489919400542236e-08,
+      "loss": 0.0348,
+      "step": 3768
+    },
+    {
+      "epoch": 2.8972142170989432,
+      "grad_norm": 0.7521761082939742,
+      "learning_rate": 2.4741141509353136e-08,
+      "loss": 0.0376,
+      "step": 3770
+    },
+    {
+      "epoch": 2.898751200768492,
+      "grad_norm": 0.7042258551510416,
+      "learning_rate": 2.4003487023557978e-08,
+      "loss": 0.0383,
+      "step": 3772
+    },
+    {
+      "epoch": 2.9002881844380406,
+      "grad_norm": 0.6754060246689739,
+      "learning_rate": 2.3276958304198235e-08,
+      "loss": 0.0356,
+      "step": 3774
+    },
+    {
+      "epoch": 2.901825168107589,
+      "grad_norm": 0.738067712348994,
+      "learning_rate": 2.2561557676705314e-08,
+      "loss": 0.0441,
+      "step": 3776
+    },
+    {
+      "epoch": 2.9033621517771375,
+      "grad_norm": 0.7057955885412858,
+      "learning_rate": 2.1857287430891213e-08,
+      "loss": 0.0378,
+      "step": 3778
+    },
+    {
+      "epoch": 2.9048991354466858,
+      "grad_norm": 0.6886592794650539,
+      "learning_rate": 2.1164149820942722e-08,
+      "loss": 0.0366,
+      "step": 3780
+    },
+    {
+      "epoch": 2.9064361191162345,
+      "grad_norm": 0.7673640335969266,
+      "learning_rate": 2.048214706541479e-08,
+      "loss": 0.0406,
+      "step": 3782
+    },
+    {
+      "epoch": 2.9079731027857827,
+      "grad_norm": 0.6678527359649526,
+      "learning_rate": 1.9811281347221597e-08,
+      "loss": 0.037,
+      "step": 3784
+    },
+    {
+      "epoch": 2.9095100864553314,
+      "grad_norm": 0.7127890261581394,
+      "learning_rate": 1.9151554813630734e-08,
+      "loss": 0.0361,
+      "step": 3786
+    },
+    {
+      "epoch": 2.91104707012488,
+      "grad_norm": 0.7583876230260865,
+      "learning_rate": 1.850296957625658e-08,
+      "loss": 0.0378,
+      "step": 3788
+    },
+    {
+      "epoch": 2.9125840537944283,
+      "grad_norm": 0.6957251087608395,
+      "learning_rate": 1.7865527711052932e-08,
+      "loss": 0.0392,
+      "step": 3790
+    },
+    {
+      "epoch": 2.914121037463977,
+      "grad_norm": 0.7245984947394084,
+      "learning_rate": 1.7239231258306397e-08,
+      "loss": 0.0387,
+      "step": 3792
+    },
+    {
+      "epoch": 2.9156580211335257,
+      "grad_norm": 0.6799380535511241,
+      "learning_rate": 1.662408222262979e-08,
+      "loss": 0.0359,
+      "step": 3794
+    },
+    {
+      "epoch": 2.917195004803074,
+      "grad_norm": 0.6832560309649182,
+      "learning_rate": 1.6020082572956674e-08,
+      "loss": 0.0379,
+      "step": 3796
+    },
+    {
+      "epoch": 2.9187319884726226,
+      "grad_norm": 0.7506399618941308,
+      "learning_rate": 1.542723424253323e-08,
+      "loss": 0.0433,
+      "step": 3798
+    },
+    {
+      "epoch": 2.920268972142171,
+      "grad_norm": 0.7012670552597634,
+      "learning_rate": 1.4845539128913954e-08,
+      "loss": 0.038,
+      "step": 3800
+    },
+    {
+      "epoch": 2.9218059558117195,
+      "grad_norm": 0.8131855947668669,
+      "learning_rate": 1.4274999093955077e-08,
+      "loss": 0.038,
+      "step": 3802
+    },
+    {
+      "epoch": 2.9233429394812678,
+      "grad_norm": 0.72352053716851,
+      "learning_rate": 1.371561596380677e-08,
+      "loss": 0.04,
+      "step": 3804
+    },
+    {
+      "epoch": 2.9248799231508165,
+      "grad_norm": 0.7277652228743811,
+      "learning_rate": 1.3167391528910831e-08,
+      "loss": 0.0404,
+      "step": 3806
+    },
+    {
+      "epoch": 2.926416906820365,
+      "grad_norm": 0.6608930767109022,
+      "learning_rate": 1.2630327543991349e-08,
+      "loss": 0.0333,
+      "step": 3808
+    },
+    {
+      "epoch": 2.9279538904899134,
+      "grad_norm": 0.7251459149646445,
+      "learning_rate": 1.2104425728051981e-08,
+      "loss": 0.035,
+      "step": 3810
+    },
+    {
+      "epoch": 2.929490874159462,
+      "grad_norm": 0.7280772115836247,
+      "learning_rate": 1.1589687764367807e-08,
+      "loss": 0.0366,
+      "step": 3812
+    },
+    {
+      "epoch": 2.9310278578290108,
+      "grad_norm": 0.7583993906119615,
+      "learning_rate": 1.1086115300482203e-08,
+      "loss": 0.0377,
+      "step": 3814
+    },
+    {
+      "epoch": 2.932564841498559,
+      "grad_norm": 0.6946315289320816,
+      "learning_rate": 1.0593709948200635e-08,
+      "loss": 0.0375,
+      "step": 3816
+    },
+    {
+      "epoch": 2.9341018251681077,
+      "grad_norm": 0.7540614936256114,
+      "learning_rate": 1.011247328358561e-08,
+      "loss": 0.0419,
+      "step": 3818
+    },
+    {
+      "epoch": 2.9356388088376564,
+      "grad_norm": 0.7380916570292597,
+      "learning_rate": 9.642406846950446e-09,
+      "loss": 0.0391,
+      "step": 3820
+    },
+    {
+      "epoch": 2.9371757925072046,
+      "grad_norm": 0.7229342418450578,
+      "learning_rate": 9.183512142857342e-09,
+      "loss": 0.0363,
+      "step": 3822
+    },
+    {
+      "epoch": 2.938712776176753,
+      "grad_norm": 0.7727279695915876,
+      "learning_rate": 8.73579064010882e-09,
+      "loss": 0.0395,
+      "step": 3824
+    },
+    {
+      "epoch": 2.9402497598463015,
+      "grad_norm": 0.741689517658393,
+      "learning_rate": 8.299243771746179e-09,
+      "loss": 0.0386,
+      "step": 3826
+    },
+    {
+      "epoch": 2.9417867435158502,
+      "grad_norm": 0.7231470558103498,
+      "learning_rate": 7.873872935043269e-09,
+      "loss": 0.041,
+      "step": 3828
+    },
+    {
+      "epoch": 2.9433237271853985,
+      "grad_norm": 0.6584072678052216,
+      "learning_rate": 7.459679491501448e-09,
+      "loss": 0.0345,
+      "step": 3830
+    },
+    {
+      "epoch": 2.944860710854947,
+      "grad_norm": 0.6835760459984861,
+      "learning_rate": 7.0566647668476315e-09,
+      "loss": 0.0384,
+      "step": 3832
+    },
+    {
+      "epoch": 2.946397694524496,
+      "grad_norm": 0.7191660536723004,
+      "learning_rate": 6.6648300510276925e-09,
+      "loss": 0.0358,
+      "step": 3834
+    },
+    {
+      "epoch": 2.947934678194044,
+      "grad_norm": 0.7051985526337444,
+      "learning_rate": 6.284176598202573e-09,
+      "loss": 0.0369,
+      "step": 3836
+    },
+    {
+      "epoch": 2.9494716618635928,
+      "grad_norm": 0.6889042075604795,
+      "learning_rate": 5.914705626746342e-09,
+      "loss": 0.0346,
+      "step": 3838
+    },
+    {
+      "epoch": 2.9510086455331415,
+      "grad_norm": 0.6797954961166619,
+      "learning_rate": 5.556418319239975e-09,
+      "loss": 0.0358,
+      "step": 3840
+    },
+    {
+      "epoch": 2.9525456292026897,
+      "grad_norm": 0.7081235515177567,
+      "learning_rate": 5.209315822468252e-09,
+      "loss": 0.04,
+      "step": 3842
+    },
+    {
+      "epoch": 2.9540826128722384,
+      "grad_norm": 0.7586899396756654,
+      "learning_rate": 4.873399247417032e-09,
+      "loss": 0.041,
+      "step": 3844
+    },
+    {
+      "epoch": 2.9556195965417866,
+      "grad_norm": 0.6546459502260559,
+      "learning_rate": 4.548669669267813e-09,
+      "loss": 0.0379,
+      "step": 3846
+    },
+    {
+      "epoch": 2.9571565802113353,
+      "grad_norm": 0.7194117012638997,
+      "learning_rate": 4.23512812739657e-09,
+      "loss": 0.0352,
+      "step": 3848
+    },
+    {
+      "epoch": 2.9586935638808836,
+      "grad_norm": 0.6550475400228738,
+      "learning_rate": 3.9327756253683125e-09,
+      "loss": 0.0338,
+      "step": 3850
+    },
+    {
+      "epoch": 2.9602305475504322,
+      "grad_norm": 0.6471117547944658,
+      "learning_rate": 3.6416131309355283e-09,
+      "loss": 0.0327,
+      "step": 3852
+    },
+    {
+      "epoch": 2.961767531219981,
+      "grad_norm": 0.6643076347264795,
+      "learning_rate": 3.361641576034302e-09,
+      "loss": 0.0366,
+      "step": 3854
+    },
+    {
+      "epoch": 2.963304514889529,
+      "grad_norm": 0.7533191809512965,
+      "learning_rate": 3.0928618567808153e-09,
+      "loss": 0.0401,
+      "step": 3856
+    },
+    {
+      "epoch": 2.964841498559078,
+      "grad_norm": 0.7192365559237837,
+      "learning_rate": 2.835274833469403e-09,
+      "loss": 0.0372,
+      "step": 3858
+    },
+    {
+      "epoch": 2.9663784822286265,
+      "grad_norm": 0.720379151963908,
+      "learning_rate": 2.588881330570225e-09,
+      "loss": 0.0362,
+      "step": 3860
+    },
+    {
+      "epoch": 2.967915465898175,
+      "grad_norm": 0.7348332898518517,
+      "learning_rate": 2.3536821367246e-09,
+      "loss": 0.0342,
+      "step": 3862
+    },
+    {
+      "epoch": 2.9694524495677235,
+      "grad_norm": 0.7451802490285939,
+      "learning_rate": 2.1296780047446172e-09,
+      "loss": 0.0393,
+      "step": 3864
+    },
+    {
+      "epoch": 2.9709894332372717,
+      "grad_norm": 0.5897360637433469,
+      "learning_rate": 1.9168696516092544e-09,
+      "loss": 0.035,
+      "step": 3866
+    },
+    {
+      "epoch": 2.9725264169068204,
+      "grad_norm": 0.6721100987536596,
+      "learning_rate": 1.7152577584635952e-09,
+      "loss": 0.0376,
+      "step": 3868
+    },
+    {
+      "epoch": 2.9740634005763686,
+      "grad_norm": 0.7372791169185593,
+      "learning_rate": 1.524842970614948e-09,
+      "loss": 0.0393,
+      "step": 3870
+    },
+    {
+      "epoch": 2.9756003842459173,
+      "grad_norm": 0.7155596616446167,
+      "learning_rate": 1.3456258975312885e-09,
+      "loss": 0.041,
+      "step": 3872
+    },
+    {
+      "epoch": 2.977137367915466,
+      "grad_norm": 0.7243460282148372,
+      "learning_rate": 1.1776071128412614e-09,
+      "loss": 0.0373,
+      "step": 3874
+    },
+    {
+      "epoch": 2.9786743515850143,
+      "grad_norm": 0.8216882893956271,
+      "learning_rate": 1.0207871543287394e-09,
+      "loss": 0.0387,
+      "step": 3876
+    },
+    {
+      "epoch": 2.980211335254563,
+      "grad_norm": 0.7555366446961992,
+      "learning_rate": 8.7516652393399e-10,
+      "loss": 0.0387,
+      "step": 3878
+    },
+    {
+      "epoch": 2.9817483189241116,
+      "grad_norm": 0.7303863102314947,
+      "learning_rate": 7.40745687751343e-10,
+      "loss": 0.0375,
+      "step": 3880
+    },
+    {
+      "epoch": 2.98328530259366,
+      "grad_norm": 0.7352509297982905,
+      "learning_rate": 6.175250760268591e-10,
+      "loss": 0.0398,
+      "step": 3882
+    },
+    {
+      "epoch": 2.9848222862632086,
+      "grad_norm": 0.7396812555733191,
+      "learning_rate": 5.055050831579422e-10,
+      "loss": 0.0454,
+      "step": 3884
+    },
+    {
+      "epoch": 2.986359269932757,
+      "grad_norm": 0.6935348372155691,
+      "learning_rate": 4.0468606769178403e-10,
+      "loss": 0.0353,
+      "step": 3886
+    },
+    {
+      "epoch": 2.9878962536023055,
+      "grad_norm": 0.7901782130270826,
+      "learning_rate": 3.150683523238107e-10,
+      "loss": 0.0376,
+      "step": 3888
+    },
+    {
+      "epoch": 2.9894332372718537,
+      "grad_norm": 0.6621002431203192,
+      "learning_rate": 2.366522238972935e-10,
+      "loss": 0.0356,
+      "step": 3890
+    },
+    {
+      "epoch": 2.9909702209414024,
+      "grad_norm": 0.7167270045432917,
+      "learning_rate": 1.6943793340179482e-10,
+      "loss": 0.0381,
+      "step": 3892
+    },
+    {
+      "epoch": 2.992507204610951,
+      "grad_norm": 0.70796009257099,
+      "learning_rate": 1.1342569597277973e-10,
+      "loss": 0.0376,
+      "step": 3894
+    },
+    {
+      "epoch": 2.9940441882804993,
+      "grad_norm": 0.725988884787138,
+      "learning_rate": 6.861569089161578e-11,
+      "loss": 0.0354,
+      "step": 3896
+    },
+    {
+      "epoch": 2.995581171950048,
+      "grad_norm": 0.6619948029465499,
+      "learning_rate": 3.500806158285297e-11,
+      "loss": 0.0362,
+      "step": 3898
+    },
+    {
+      "epoch": 2.9971181556195967,
+      "grad_norm": 0.7007360064163597,
+      "learning_rate": 1.2602915616166665e-11,
+      "loss": 0.0403,
+      "step": 3900
+    },
+    {
+      "epoch": 2.998655139289145,
+      "grad_norm": 0.6392213863532241,
+      "learning_rate": 1.4003247044147572e-12,
+      "loss": 0.0384,
+      "step": 3902
+    },
+    {
+      "epoch": 2.9994236311239195,
+      "step": 3903,
+      "total_flos": 1225143666540544.0,
+      "train_loss": 0.11350150189982632,
+      "train_runtime": 22267.9689,
+      "train_samples_per_second": 22.437,
+      "train_steps_per_second": 0.175
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 3903,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1225143666540544.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}