diff --git "a/checkpoint-2500/trainer_state.json" "b/checkpoint-2500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2500/trainer_state.json"
@@ -0,0 +1,17521 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.2219753360737697,
+  "eval_steps": 500,
+  "global_step": 2500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008887901344295079,
+      "grad_norm": 0.8810775279998779,
+      "learning_rate": 0.0,
+      "loss": 1.9172,
+      "step": 1
+    },
+    {
+      "epoch": 0.0017775802688590157,
+      "grad_norm": 0.7469067573547363,
+      "learning_rate": 2.1533827903669654e-05,
+      "loss": 1.8244,
+      "step": 2
+    },
+    {
+      "epoch": 0.0026663704032885236,
+      "grad_norm": 0.6811410784721375,
+      "learning_rate": 3.413030972429927e-05,
+      "loss": 1.926,
+      "step": 3
+    },
+    {
+      "epoch": 0.0035551605377180315,
+      "grad_norm": 0.5220848321914673,
+      "learning_rate": 4.306765580733931e-05,
+      "loss": 1.8387,
+      "step": 4
+    },
+    {
+      "epoch": 0.004443950672147539,
+      "grad_norm": 0.5415656566619873,
+      "learning_rate": 5e-05,
+      "loss": 1.8832,
+      "step": 5
+    },
+    {
+      "epoch": 0.005332740806577047,
+      "grad_norm": 0.3646649718284607,
+      "learning_rate": 5.5664137627968925e-05,
+      "loss": 1.8791,
+      "step": 6
+    },
+    {
+      "epoch": 0.006221530941006555,
+      "grad_norm": 0.42480552196502686,
+      "learning_rate": 6.0453097756108376e-05,
+      "loss": 1.7676,
+      "step": 7
+    },
+    {
+      "epoch": 0.007110321075436063,
+      "grad_norm": 0.38259413838386536,
+      "learning_rate": 6.460148371100896e-05,
+      "loss": 1.7888,
+      "step": 8
+    },
+    {
+      "epoch": 0.00799911120986557,
+      "grad_norm": 0.3880155086517334,
+      "learning_rate": 6.826061944859854e-05,
+      "loss": 1.7697,
+      "step": 9
+    },
+    {
+      "epoch": 0.008887901344295079,
+      "grad_norm": 0.5344939231872559,
+      "learning_rate": 7.153382790366967e-05,
+      "loss": 1.6064,
+      "step": 10
+    },
+    {
+      "epoch": 0.009776691478724587,
+      "grad_norm": 0.42493271827697754,
+      "learning_rate": 7.449480512024892e-05,
+      "loss": 1.7261,
+      "step": 11
+    },
+    {
+      "epoch": 0.010665481613154094,
+      "grad_norm": 0.49020832777023315,
+      "learning_rate": 7.719796553163858e-05,
+      "loss": 1.7062,
+      "step": 12
+    },
+    {
+      "epoch": 0.011554271747583602,
+      "grad_norm": 0.5008671283721924,
+      "learning_rate": 7.968463205835412e-05,
+      "loss": 1.7161,
+      "step": 13
+    },
+    {
+      "epoch": 0.01244306188201311,
+      "grad_norm": 0.6378142833709717,
+      "learning_rate": 8.198692565977803e-05,
+      "loss": 1.5226,
+      "step": 14
+    },
+    {
+      "epoch": 0.013331852016442618,
+      "grad_norm": 0.6101044416427612,
+      "learning_rate": 8.413030972429928e-05,
+      "loss": 1.5718,
+      "step": 15
+    },
+    {
+      "epoch": 0.014220642150872126,
+      "grad_norm": 0.4299439787864685,
+      "learning_rate": 8.613531161467861e-05,
+      "loss": 1.7133,
+      "step": 16
+    },
+    {
+      "epoch": 0.015109432285301634,
+      "grad_norm": 0.48655468225479126,
+      "learning_rate": 8.80187213861294e-05,
+      "loss": 1.5746,
+      "step": 17
+    },
+    {
+      "epoch": 0.01599822241973114,
+      "grad_norm": 0.577294647693634,
+      "learning_rate": 8.979444735226819e-05,
+      "loss": 1.4959,
+      "step": 18
+    },
+    {
+      "epoch": 0.016887012554160648,
+      "grad_norm": 0.5080670714378357,
+      "learning_rate": 9.147414002175752e-05,
+      "loss": 1.4947,
+      "step": 19
+    },
+    {
+      "epoch": 0.017775802688590157,
+      "grad_norm": 0.6289512515068054,
+      "learning_rate": 9.306765580733931e-05,
+      "loss": 1.5576,
+      "step": 20
+    },
+    {
+      "epoch": 0.018664592823019664,
+      "grad_norm": 0.5262102484703064,
+      "learning_rate": 9.458340748040766e-05,
+      "loss": 1.5228,
+      "step": 21
+    },
+    {
+      "epoch": 0.019553382957449173,
+      "grad_norm": 0.4739653170108795,
+      "learning_rate": 9.602863302391859e-05,
+      "loss": 1.5751,
+      "step": 22
+    },
+    {
+      "epoch": 0.02044217309187868,
+      "grad_norm": 0.4563744068145752,
+      "learning_rate": 9.740960467331899e-05,
+      "loss": 1.4833,
+      "step": 23
+    },
+    {
+      "epoch": 0.02133096322630819,
+      "grad_norm": 0.5045209527015686,
+      "learning_rate": 9.873179343530825e-05,
+      "loss": 1.5092,
+      "step": 24
+    },
+    {
+      "epoch": 0.022219753360737695,
+      "grad_norm": 0.3677147328853607,
+      "learning_rate": 0.0001,
+      "loss": 1.5253,
+      "step": 25
+    },
+    {
+      "epoch": 0.023108543495167205,
+      "grad_norm": 0.5037134289741516,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 26
+    },
+    {
+      "epoch": 0.02399733362959671,
+      "grad_norm": 0.5134053230285645,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 27
+    },
+    {
+      "epoch": 0.02488612376402622,
+      "grad_norm": 0.4475038945674896,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 28
+    },
+    {
+      "epoch": 0.025774913898455726,
+      "grad_norm": 0.4869753122329712,
+      "learning_rate": 0.0001,
+      "loss": 1.4236,
+      "step": 29
+    },
+    {
+      "epoch": 0.026663704032885236,
+      "grad_norm": 0.3995779752731323,
+      "learning_rate": 0.0001,
+      "loss": 1.4746,
+      "step": 30
+    },
+    {
+      "epoch": 0.027552494167314742,
+      "grad_norm": 0.43385857343673706,
+      "learning_rate": 0.0001,
+      "loss": 1.4692,
+      "step": 31
+    },
+    {
+      "epoch": 0.028441284301744252,
+      "grad_norm": 0.46782711148262024,
+      "learning_rate": 0.0001,
+      "loss": 1.414,
+      "step": 32
+    },
+    {
+      "epoch": 0.029330074436173758,
+      "grad_norm": 0.5026598572731018,
+      "learning_rate": 0.0001,
+      "loss": 1.348,
+      "step": 33
+    },
+    {
+      "epoch": 0.030218864570603268,
+      "grad_norm": 0.3947531580924988,
+      "learning_rate": 0.0001,
+      "loss": 1.325,
+      "step": 34
+    },
+    {
+      "epoch": 0.031107654705032774,
+      "grad_norm": 0.4685068428516388,
+      "learning_rate": 0.0001,
+      "loss": 1.3592,
+      "step": 35
+    },
+    {
+      "epoch": 0.03199644483946228,
+      "grad_norm": 0.5283563733100891,
+      "learning_rate": 0.0001,
+      "loss": 1.4594,
+      "step": 36
+    },
+    {
+      "epoch": 0.03288523497389179,
+      "grad_norm": 0.48610830307006836,
+      "learning_rate": 0.0001,
+      "loss": 1.4441,
+      "step": 37
+    },
+    {
+      "epoch": 0.033774025108321296,
+      "grad_norm": 0.46293529868125916,
+      "learning_rate": 0.0001,
+      "loss": 1.3483,
+      "step": 38
+    },
+    {
+      "epoch": 0.034662815242750805,
+      "grad_norm": 0.5796862244606018,
+      "learning_rate": 0.0001,
+      "loss": 1.3338,
+      "step": 39
+    },
+    {
+      "epoch": 0.035551605377180315,
+      "grad_norm": 0.5200364589691162,
+      "learning_rate": 0.0001,
+      "loss": 1.3805,
+      "step": 40
+    },
+    {
+      "epoch": 0.036440395511609824,
+      "grad_norm": 0.4546230137348175,
+      "learning_rate": 0.0001,
+      "loss": 1.3318,
+      "step": 41
+    },
+    {
+      "epoch": 0.03732918564603933,
+      "grad_norm": 0.5082724094390869,
+      "learning_rate": 0.0001,
+      "loss": 1.2617,
+      "step": 42
+    },
+    {
+      "epoch": 0.03821797578046884,
+      "grad_norm": 0.4639153480529785,
+      "learning_rate": 0.0001,
+      "loss": 1.2998,
+      "step": 43
+    },
+    {
+      "epoch": 0.039106765914898346,
+      "grad_norm": 0.6027429699897766,
+      "learning_rate": 0.0001,
+      "loss": 1.3551,
+      "step": 44
+    },
+    {
+      "epoch": 0.039995556049327856,
+      "grad_norm": 0.4588245451450348,
+      "learning_rate": 0.0001,
+      "loss": 1.297,
+      "step": 45
+    },
+    {
+      "epoch": 0.04088434618375736,
+      "grad_norm": 0.49096405506134033,
+      "learning_rate": 0.0001,
+      "loss": 1.3503,
+      "step": 46
+    },
+    {
+      "epoch": 0.04177313631818687,
+      "grad_norm": 0.5192633867263794,
+      "learning_rate": 0.0001,
+      "loss": 1.3139,
+      "step": 47
+    },
+    {
+      "epoch": 0.04266192645261638,
+      "grad_norm": 0.5095855593681335,
+      "learning_rate": 0.0001,
+      "loss": 1.2607,
+      "step": 48
+    },
+    {
+      "epoch": 0.04355071658704588,
+      "grad_norm": 0.5086390376091003,
+      "learning_rate": 0.0001,
+      "loss": 1.294,
+      "step": 49
+    },
+    {
+      "epoch": 0.04443950672147539,
+      "grad_norm": 0.5870110988616943,
+      "learning_rate": 0.0001,
+      "loss": 1.3086,
+      "step": 50
+    },
+    {
+      "epoch": 0.0453282968559049,
+      "grad_norm": 0.5581070184707642,
+      "learning_rate": 0.0001,
+      "loss": 1.2506,
+      "step": 51
+    },
+    {
+      "epoch": 0.04621708699033441,
+      "grad_norm": 0.5818192362785339,
+      "learning_rate": 0.0001,
+      "loss": 1.3169,
+      "step": 52
+    },
+    {
+      "epoch": 0.04710587712476391,
+      "grad_norm": 0.612249493598938,
+      "learning_rate": 0.0001,
+      "loss": 1.2275,
+      "step": 53
+    },
+    {
+      "epoch": 0.04799466725919342,
+      "grad_norm": 0.712792158126831,
+      "learning_rate": 0.0001,
+      "loss": 1.2525,
+      "step": 54
+    },
+    {
+      "epoch": 0.04888345739362293,
+      "grad_norm": 0.6237956285476685,
+      "learning_rate": 0.0001,
+      "loss": 1.21,
+      "step": 55
+    },
+    {
+      "epoch": 0.04977224752805244,
+      "grad_norm": 0.5647529363632202,
+      "learning_rate": 0.0001,
+      "loss": 1.2615,
+      "step": 56
+    },
+    {
+      "epoch": 0.05066103766248194,
+      "grad_norm": 0.5913922786712646,
+      "learning_rate": 0.0001,
+      "loss": 1.227,
+      "step": 57
+    },
+    {
+      "epoch": 0.05154982779691145,
+      "grad_norm": 0.5439374446868896,
+      "learning_rate": 0.0001,
+      "loss": 1.2046,
+      "step": 58
+    },
+    {
+      "epoch": 0.05243861793134096,
+      "grad_norm": 0.4831898510456085,
+      "learning_rate": 0.0001,
+      "loss": 1.2106,
+      "step": 59
+    },
+    {
+      "epoch": 0.05332740806577047,
+      "grad_norm": 0.47741156816482544,
+      "learning_rate": 0.0001,
+      "loss": 1.2144,
+      "step": 60
+    },
+    {
+      "epoch": 0.054216198200199975,
+      "grad_norm": 0.4893709123134613,
+      "learning_rate": 0.0001,
+      "loss": 1.1871,
+      "step": 61
+    },
+    {
+      "epoch": 0.055104988334629484,
+      "grad_norm": 0.5424261689186096,
+      "learning_rate": 0.0001,
+      "loss": 1.238,
+      "step": 62
+    },
+    {
+      "epoch": 0.055993778469058994,
+      "grad_norm": 0.47771304845809937,
+      "learning_rate": 0.0001,
+      "loss": 1.2568,
+      "step": 63
+    },
+    {
+      "epoch": 0.056882568603488504,
+      "grad_norm": 0.6287993788719177,
+      "learning_rate": 0.0001,
+      "loss": 1.2713,
+      "step": 64
+    },
+    {
+      "epoch": 0.057771358737918006,
+      "grad_norm": 0.5192376971244812,
+      "learning_rate": 0.0001,
+      "loss": 1.1294,
+      "step": 65
+    },
+    {
+      "epoch": 0.058660148872347516,
+      "grad_norm": 0.5453503131866455,
+      "learning_rate": 0.0001,
+      "loss": 1.1706,
+      "step": 66
+    },
+    {
+      "epoch": 0.059548939006777026,
+      "grad_norm": 0.44728267192840576,
+      "learning_rate": 0.0001,
+      "loss": 1.2275,
+      "step": 67
+    },
+    {
+      "epoch": 0.060437729141206535,
+      "grad_norm": 0.561209499835968,
+      "learning_rate": 0.0001,
+      "loss": 1.1971,
+      "step": 68
+    },
+    {
+      "epoch": 0.06132651927563604,
+      "grad_norm": 0.518390417098999,
+      "learning_rate": 0.0001,
+      "loss": 1.2264,
+      "step": 69
+    },
+    {
+      "epoch": 0.06221530941006555,
+      "grad_norm": 0.5499495267868042,
+      "learning_rate": 0.0001,
+      "loss": 1.1627,
+      "step": 70
+    },
+    {
+      "epoch": 0.06310409954449506,
+      "grad_norm": 0.46714553236961365,
+      "learning_rate": 0.0001,
+      "loss": 1.1767,
+      "step": 71
+    },
+    {
+      "epoch": 0.06399288967892457,
+      "grad_norm": 0.5339136123657227,
+      "learning_rate": 0.0001,
+      "loss": 1.129,
+      "step": 72
+    },
+    {
+      "epoch": 0.06488167981335408,
+      "grad_norm": 0.6434009671211243,
+      "learning_rate": 0.0001,
+      "loss": 1.213,
+      "step": 73
+    },
+    {
+      "epoch": 0.06577046994778359,
+      "grad_norm": 0.6940732002258301,
+      "learning_rate": 0.0001,
+      "loss": 1.1562,
+      "step": 74
+    },
+    {
+      "epoch": 0.06665926008221308,
+      "grad_norm": 0.5630552768707275,
+      "learning_rate": 0.0001,
+      "loss": 1.208,
+      "step": 75
+    },
+    {
+      "epoch": 0.06754805021664259,
+      "grad_norm": 0.561046838760376,
+      "learning_rate": 0.0001,
+      "loss": 1.168,
+      "step": 76
+    },
+    {
+      "epoch": 0.0684368403510721,
+      "grad_norm": 0.4617985486984253,
+      "learning_rate": 0.0001,
+      "loss": 1.1431,
+      "step": 77
+    },
+    {
+      "epoch": 0.06932563048550161,
+      "grad_norm": 0.5184434652328491,
+      "learning_rate": 0.0001,
+      "loss": 1.1934,
+      "step": 78
+    },
+    {
+      "epoch": 0.07021442061993112,
+      "grad_norm": 0.6089707016944885,
+      "learning_rate": 0.0001,
+      "loss": 1.2368,
+      "step": 79
+    },
+    {
+      "epoch": 0.07110321075436063,
+      "grad_norm": 0.5018100738525391,
+      "learning_rate": 0.0001,
+      "loss": 1.1722,
+      "step": 80
+    },
+    {
+      "epoch": 0.07199200088879014,
+      "grad_norm": 0.49356648325920105,
+      "learning_rate": 0.0001,
+      "loss": 1.1999,
+      "step": 81
+    },
+    {
+      "epoch": 0.07288079102321965,
+      "grad_norm": 0.544902503490448,
+      "learning_rate": 0.0001,
+      "loss": 1.1496,
+      "step": 82
+    },
+    {
+      "epoch": 0.07376958115764914,
+      "grad_norm": 0.534565806388855,
+      "learning_rate": 0.0001,
+      "loss": 1.0753,
+      "step": 83
+    },
+    {
+      "epoch": 0.07465837129207865,
+      "grad_norm": 0.6407784223556519,
+      "learning_rate": 0.0001,
+      "loss": 1.1728,
+      "step": 84
+    },
+    {
+      "epoch": 0.07554716142650816,
+      "grad_norm": 0.619192361831665,
+      "learning_rate": 0.0001,
+      "loss": 1.179,
+      "step": 85
+    },
+    {
+      "epoch": 0.07643595156093767,
+      "grad_norm": 0.5248496532440186,
+      "learning_rate": 0.0001,
+      "loss": 1.1668,
+      "step": 86
+    },
+    {
+      "epoch": 0.07732474169536718,
+      "grad_norm": 0.5901935696601868,
+      "learning_rate": 0.0001,
+      "loss": 1.1634,
+      "step": 87
+    },
+    {
+      "epoch": 0.07821353182979669,
+      "grad_norm": 0.5717945694923401,
+      "learning_rate": 0.0001,
+      "loss": 1.046,
+      "step": 88
+    },
+    {
+      "epoch": 0.0791023219642262,
+      "grad_norm": 0.5313799381256104,
+      "learning_rate": 0.0001,
+      "loss": 1.1541,
+      "step": 89
+    },
+    {
+      "epoch": 0.07999111209865571,
+      "grad_norm": 0.5816580057144165,
+      "learning_rate": 0.0001,
+      "loss": 1.1505,
+      "step": 90
+    },
+    {
+      "epoch": 0.08087990223308521,
+      "grad_norm": 0.6545852422714233,
+      "learning_rate": 0.0001,
+      "loss": 1.1687,
+      "step": 91
+    },
+    {
+      "epoch": 0.08176869236751472,
+      "grad_norm": 0.5649644136428833,
+      "learning_rate": 0.0001,
+      "loss": 1.2657,
+      "step": 92
+    },
+    {
+      "epoch": 0.08265748250194423,
+      "grad_norm": 0.5403311848640442,
+      "learning_rate": 0.0001,
+      "loss": 1.1041,
+      "step": 93
+    },
+    {
+      "epoch": 0.08354627263637374,
+      "grad_norm": 0.6377498507499695,
+      "learning_rate": 0.0001,
+      "loss": 1.1621,
+      "step": 94
+    },
+    {
+      "epoch": 0.08443506277080325,
+      "grad_norm": 0.5470308065414429,
+      "learning_rate": 0.0001,
+      "loss": 1.0845,
+      "step": 95
+    },
+    {
+      "epoch": 0.08532385290523276,
+      "grad_norm": 0.4540124833583832,
+      "learning_rate": 0.0001,
+      "loss": 1.1723,
+      "step": 96
+    },
+    {
+      "epoch": 0.08621264303966227,
+      "grad_norm": 0.49754512310028076,
+      "learning_rate": 0.0001,
+      "loss": 1.165,
+      "step": 97
+    },
+    {
+      "epoch": 0.08710143317409176,
+      "grad_norm": 0.5652042031288147,
+      "learning_rate": 0.0001,
+      "loss": 1.0729,
+      "step": 98
+    },
+    {
+      "epoch": 0.08799022330852127,
+      "grad_norm": 0.5990537405014038,
+      "learning_rate": 0.0001,
+      "loss": 1.1192,
+      "step": 99
+    },
+    {
+      "epoch": 0.08887901344295078,
+      "grad_norm": 0.49589407444000244,
+      "learning_rate": 0.0001,
+      "loss": 1.0975,
+      "step": 100
+    },
+    {
+      "epoch": 0.08976780357738029,
+      "grad_norm": 0.4846036732196808,
+      "learning_rate": 0.0001,
+      "loss": 1.1206,
+      "step": 101
+    },
+    {
+      "epoch": 0.0906565937118098,
+      "grad_norm": 0.5197208523750305,
+      "learning_rate": 0.0001,
+      "loss": 1.1676,
+      "step": 102
+    },
+    {
+      "epoch": 0.09154538384623931,
+      "grad_norm": 0.500135600566864,
+      "learning_rate": 0.0001,
+      "loss": 1.0938,
+      "step": 103
+    },
+    {
+      "epoch": 0.09243417398066882,
+      "grad_norm": 0.48206326365470886,
+      "learning_rate": 0.0001,
+      "loss": 1.1561,
+      "step": 104
+    },
+    {
+      "epoch": 0.09332296411509833,
+      "grad_norm": 0.5363126397132874,
+      "learning_rate": 0.0001,
+      "loss": 1.1397,
+      "step": 105
+    },
+    {
+      "epoch": 0.09421175424952782,
+      "grad_norm": 0.5342210531234741,
+      "learning_rate": 0.0001,
+      "loss": 1.1213,
+      "step": 106
+    },
+    {
+      "epoch": 0.09510054438395733,
+      "grad_norm": 0.47415241599082947,
+      "learning_rate": 0.0001,
+      "loss": 1.1009,
+      "step": 107
+    },
+    {
+      "epoch": 0.09598933451838684,
+      "grad_norm": 0.5735252499580383,
+      "learning_rate": 0.0001,
+      "loss": 1.0854,
+      "step": 108
+    },
+    {
+      "epoch": 0.09687812465281635,
+      "grad_norm": 0.6142700910568237,
+      "learning_rate": 0.0001,
+      "loss": 1.2542,
+      "step": 109
+    },
+    {
+      "epoch": 0.09776691478724586,
+      "grad_norm": 0.5673418045043945,
+      "learning_rate": 0.0001,
+      "loss": 1.0388,
+      "step": 110
+    },
+    {
+      "epoch": 0.09865570492167537,
+      "grad_norm": 0.5416164398193359,
+      "learning_rate": 0.0001,
+      "loss": 1.0302,
+      "step": 111
+    },
+    {
+      "epoch": 0.09954449505610488,
+      "grad_norm": 0.48338207602500916,
+      "learning_rate": 0.0001,
+      "loss": 1.1465,
+      "step": 112
+    },
+    {
+      "epoch": 0.10043328519053439,
+      "grad_norm": 0.49852466583251953,
+      "learning_rate": 0.0001,
+      "loss": 1.1088,
+      "step": 113
+    },
+    {
+      "epoch": 0.10132207532496389,
+      "grad_norm": 0.5490244626998901,
+      "learning_rate": 0.0001,
+      "loss": 1.0804,
+      "step": 114
+    },
+    {
+      "epoch": 0.1022108654593934,
+      "grad_norm": 7.056009769439697,
+      "learning_rate": 0.0001,
+      "loss": 1.094,
+      "step": 115
+    },
+    {
+      "epoch": 0.1030996555938229,
+      "grad_norm": 0.5697766542434692,
+      "learning_rate": 0.0001,
+      "loss": 1.0271,
+      "step": 116
+    },
+    {
+      "epoch": 0.10398844572825242,
+      "grad_norm": 0.5138140320777893,
+      "learning_rate": 0.0001,
+      "loss": 1.0057,
+      "step": 117
+    },
+    {
+      "epoch": 0.10487723586268193,
+      "grad_norm": 0.4911749064922333,
+      "learning_rate": 0.0001,
+      "loss": 1.1375,
+      "step": 118
+    },
+    {
+      "epoch": 0.10576602599711143,
+      "grad_norm": 0.45733195543289185,
+      "learning_rate": 0.0001,
+      "loss": 1.0889,
+      "step": 119
+    },
+    {
+      "epoch": 0.10665481613154094,
+      "grad_norm": 0.4950946569442749,
+      "learning_rate": 0.0001,
+      "loss": 1.1143,
+      "step": 120
+    },
+    {
+      "epoch": 0.10754360626597045,
+      "grad_norm": 0.5536718964576721,
+      "learning_rate": 0.0001,
+      "loss": 1.1188,
+      "step": 121
+    },
+    {
+      "epoch": 0.10843239640039995,
+      "grad_norm": 0.5105839967727661,
+      "learning_rate": 0.0001,
+      "loss": 1.0501,
+      "step": 122
+    },
+    {
+      "epoch": 0.10932118653482946,
+      "grad_norm": 0.5183722972869873,
+      "learning_rate": 0.0001,
+      "loss": 1.0766,
+      "step": 123
+    },
+    {
+      "epoch": 0.11020997666925897,
+      "grad_norm": 0.5385981798171997,
+      "learning_rate": 0.0001,
+      "loss": 1.0419,
+      "step": 124
+    },
+    {
+      "epoch": 0.11109876680368848,
+      "grad_norm": 0.6438923478126526,
+      "learning_rate": 0.0001,
+      "loss": 1.1333,
+      "step": 125
+    },
+    {
+      "epoch": 0.11198755693811799,
+      "grad_norm": 0.4993588328361511,
+      "learning_rate": 0.0001,
+      "loss": 1.0504,
+      "step": 126
+    },
+    {
+      "epoch": 0.1128763470725475,
+      "grad_norm": 0.5645109415054321,
+      "learning_rate": 0.0001,
+      "loss": 1.0817,
+      "step": 127
+    },
+    {
+      "epoch": 0.11376513720697701,
+      "grad_norm": 0.5064727067947388,
+      "learning_rate": 0.0001,
+      "loss": 1.1588,
+      "step": 128
+    },
+    {
+      "epoch": 0.11465392734140652,
+      "grad_norm": 0.4977114498615265,
+      "learning_rate": 0.0001,
+      "loss": 1.0903,
+      "step": 129
+    },
+    {
+      "epoch": 0.11554271747583601,
+      "grad_norm": 0.5678680539131165,
+      "learning_rate": 0.0001,
+      "loss": 1.1107,
+      "step": 130
+    },
+    {
+      "epoch": 0.11643150761026552,
+      "grad_norm": 0.5574091076850891,
+      "learning_rate": 0.0001,
+      "loss": 1.1081,
+      "step": 131
+    },
+    {
+      "epoch": 0.11732029774469503,
+      "grad_norm": 0.5073116421699524,
+      "learning_rate": 0.0001,
+      "loss": 1.0487,
+      "step": 132
+    },
+    {
+      "epoch": 0.11820908787912454,
+      "grad_norm": 0.4510885179042816,
+      "learning_rate": 0.0001,
+      "loss": 1.0262,
+      "step": 133
+    },
+    {
+      "epoch": 0.11909787801355405,
+      "grad_norm": 0.5130484104156494,
+      "learning_rate": 0.0001,
+      "loss": 1.0612,
+      "step": 134
+    },
+    {
+      "epoch": 0.11998666814798356,
+      "grad_norm": 0.55837082862854,
+      "learning_rate": 0.0001,
+      "loss": 1.1003,
+      "step": 135
+    },
+    {
+      "epoch": 0.12087545828241307,
+      "grad_norm": 0.563966691493988,
+      "learning_rate": 0.0001,
+      "loss": 1.1106,
+      "step": 136
+    },
+    {
+      "epoch": 0.12176424841684258,
+      "grad_norm": 0.495710164308548,
+      "learning_rate": 0.0001,
+      "loss": 1.077,
+      "step": 137
+    },
+    {
+      "epoch": 0.12265303855127208,
+      "grad_norm": 0.4716346561908722,
+      "learning_rate": 0.0001,
+      "loss": 1.0224,
+      "step": 138
+    },
+    {
+      "epoch": 0.12354182868570159,
+      "grad_norm": 0.5098469257354736,
+      "learning_rate": 0.0001,
+      "loss": 1.0821,
+      "step": 139
+    },
+    {
+      "epoch": 0.1244306188201311,
+      "grad_norm": 0.4919681251049042,
+      "learning_rate": 0.0001,
+      "loss": 1.0588,
+      "step": 140
+    },
+    {
+      "epoch": 0.12531940895456062,
+      "grad_norm": 0.4506363570690155,
+      "learning_rate": 0.0001,
+      "loss": 1.0579,
+      "step": 141
+    },
+    {
+      "epoch": 0.12620819908899011,
+      "grad_norm": 0.6020563244819641,
+      "learning_rate": 0.0001,
+      "loss": 1.0176,
+      "step": 142
+    },
+    {
+      "epoch": 0.1270969892234196,
+      "grad_norm": 0.6274687051773071,
+      "learning_rate": 0.0001,
+      "loss": 1.2257,
+      "step": 143
+    },
+    {
+      "epoch": 0.12798577935784913,
+      "grad_norm": 0.49272266030311584,
+      "learning_rate": 0.0001,
+      "loss": 1.0729,
+      "step": 144
+    },
+    {
+      "epoch": 0.12887456949227863,
+      "grad_norm": 0.5412996411323547,
+      "learning_rate": 0.0001,
+      "loss": 1.1025,
+      "step": 145
+    },
+    {
+      "epoch": 0.12976335962670815,
+      "grad_norm": 0.5278550386428833,
+      "learning_rate": 0.0001,
+      "loss": 1.136,
+      "step": 146
+    },
+    {
+      "epoch": 0.13065214976113765,
+      "grad_norm": 0.5578097701072693,
+      "learning_rate": 0.0001,
+      "loss": 1.0407,
+      "step": 147
+    },
+    {
+      "epoch": 0.13154093989556717,
+      "grad_norm": 0.5495776534080505,
+      "learning_rate": 0.0001,
+      "loss": 1.084,
+      "step": 148
+    },
+    {
+      "epoch": 0.13242973002999667,
+      "grad_norm": 0.6431498527526855,
+      "learning_rate": 0.0001,
+      "loss": 1.0953,
+      "step": 149
+    },
+    {
+      "epoch": 0.13331852016442616,
+      "grad_norm": 0.5592057108879089,
+      "learning_rate": 0.0001,
+      "loss": 1.0847,
+      "step": 150
+    },
+    {
+      "epoch": 0.1342073102988557,
+      "grad_norm": 0.6379444599151611,
+      "learning_rate": 0.0001,
+      "loss": 1.0631,
+      "step": 151
+    },
+    {
+      "epoch": 0.13509610043328518,
+      "grad_norm": 0.5859350562095642,
+      "learning_rate": 0.0001,
+      "loss": 1.069,
+      "step": 152
+    },
+    {
+      "epoch": 0.1359848905677147,
+      "grad_norm": 0.5171725153923035,
+      "learning_rate": 0.0001,
+      "loss": 1.0135,
+      "step": 153
+    },
+    {
+      "epoch": 0.1368736807021442,
+      "grad_norm": 0.4534023702144623,
+      "learning_rate": 0.0001,
+      "loss": 1.0122,
+      "step": 154
+    },
+    {
+      "epoch": 0.13776247083657372,
+      "grad_norm": 0.49811944365501404,
+      "learning_rate": 0.0001,
+      "loss": 1.0196,
+      "step": 155
+    },
+    {
+      "epoch": 0.13865126097100322,
+      "grad_norm": 0.46456894278526306,
+      "learning_rate": 0.0001,
+      "loss": 1.0723,
+      "step": 156
+    },
+    {
+      "epoch": 0.13954005110543272,
+      "grad_norm": 0.5349675416946411,
+      "learning_rate": 0.0001,
+      "loss": 1.1131,
+      "step": 157
+    },
+    {
+      "epoch": 0.14042884123986224,
+      "grad_norm": 0.466975599527359,
+      "learning_rate": 0.0001,
+      "loss": 1.0752,
+      "step": 158
+    },
+    {
+      "epoch": 0.14131763137429174,
+      "grad_norm": 0.5469770431518555,
+      "learning_rate": 0.0001,
+      "loss": 1.0607,
+      "step": 159
+    },
+    {
+      "epoch": 0.14220642150872126,
+      "grad_norm": 0.5208465456962585,
+      "learning_rate": 0.0001,
+      "loss": 0.9892,
+      "step": 160
+    },
+    {
+      "epoch": 0.14309521164315075,
+      "grad_norm": 0.543846845626831,
+      "learning_rate": 0.0001,
+      "loss": 1.0861,
+      "step": 161
+    },
+    {
+      "epoch": 0.14398400177758028,
+      "grad_norm": 0.5700931549072266,
+      "learning_rate": 0.0001,
+      "loss": 0.9992,
+      "step": 162
+    },
+    {
+      "epoch": 0.14487279191200977,
+      "grad_norm": 0.5593817830085754,
+      "learning_rate": 0.0001,
+      "loss": 1.0803,
+      "step": 163
+    },
+    {
+      "epoch": 0.1457615820464393,
+      "grad_norm": 0.5960149765014648,
+      "learning_rate": 0.0001,
+      "loss": 1.1054,
+      "step": 164
+    },
+    {
+      "epoch": 0.1466503721808688,
+      "grad_norm": 0.5227658152580261,
+      "learning_rate": 0.0001,
+      "loss": 1.1041,
+      "step": 165
+    },
+    {
+      "epoch": 0.1475391623152983,
+      "grad_norm": 0.5318516492843628,
+      "learning_rate": 0.0001,
+      "loss": 1.1705,
+      "step": 166
+    },
+    {
+      "epoch": 0.1484279524497278,
+      "grad_norm": 0.49774935841560364,
+      "learning_rate": 0.0001,
+      "loss": 1.0942,
+      "step": 167
+    },
+    {
+      "epoch": 0.1493167425841573,
+      "grad_norm": 0.5588480234146118,
+      "learning_rate": 0.0001,
+      "loss": 1.1274,
+      "step": 168
+    },
+    {
+      "epoch": 0.15020553271858683,
+      "grad_norm": 0.5155318379402161,
+      "learning_rate": 0.0001,
+      "loss": 1.0727,
+      "step": 169
+    },
+    {
+      "epoch": 0.15109432285301633,
+      "grad_norm": 0.5063637495040894,
+      "learning_rate": 0.0001,
+      "loss": 1.1389,
+      "step": 170
+    },
+    {
+      "epoch": 0.15198311298744585,
+      "grad_norm": 0.5275964736938477,
+      "learning_rate": 0.0001,
+      "loss": 1.0594,
+      "step": 171
+    },
+    {
+      "epoch": 0.15287190312187535,
+      "grad_norm": 0.48888012766838074,
+      "learning_rate": 0.0001,
+      "loss": 1.0238,
+      "step": 172
+    },
+    {
+      "epoch": 0.15376069325630484,
+      "grad_norm": 0.6187731623649597,
+      "learning_rate": 0.0001,
+      "loss": 1.0535,
+      "step": 173
+    },
+    {
+      "epoch": 0.15464948339073437,
+      "grad_norm": 0.53126460313797,
+      "learning_rate": 0.0001,
+      "loss": 1.026,
+      "step": 174
+    },
+    {
+      "epoch": 0.15553827352516386,
+      "grad_norm": 0.5398485064506531,
+      "learning_rate": 0.0001,
+      "loss": 1.069,
+      "step": 175
+    },
+    {
+      "epoch": 0.15642706365959338,
+      "grad_norm": 0.5406534075737,
+      "learning_rate": 0.0001,
+      "loss": 1.0529,
+      "step": 176
+    },
+    {
+      "epoch": 0.15731585379402288,
+      "grad_norm": 0.48404720425605774,
+      "learning_rate": 0.0001,
+      "loss": 1.0038,
+      "step": 177
+    },
+    {
+      "epoch": 0.1582046439284524,
+      "grad_norm": 0.5885335206985474,
+      "learning_rate": 0.0001,
+      "loss": 1.0029,
+      "step": 178
+    },
+    {
+      "epoch": 0.1590934340628819,
+      "grad_norm": 0.5530521869659424,
+      "learning_rate": 0.0001,
+      "loss": 1.1171,
+      "step": 179
+    },
+    {
+      "epoch": 0.15998222419731142,
+      "grad_norm": 0.5618231892585754,
+      "learning_rate": 0.0001,
+      "loss": 1.1232,
+      "step": 180
+    },
+    {
+      "epoch": 0.16087101433174092,
+      "grad_norm": 0.5092931389808655,
+      "learning_rate": 0.0001,
+      "loss": 1.0364,
+      "step": 181
+    },
+    {
+      "epoch": 0.16175980446617041,
+      "grad_norm": 0.44638335704803467,
+      "learning_rate": 0.0001,
+      "loss": 1.0221,
+      "step": 182
+    },
+    {
+      "epoch": 0.16264859460059994,
+      "grad_norm": 0.4906376004219055,
+      "learning_rate": 0.0001,
+      "loss": 1.023,
+      "step": 183
+    },
+    {
+      "epoch": 0.16353738473502943,
+      "grad_norm": 0.5148441195487976,
+      "learning_rate": 0.0001,
+      "loss": 1.0451,
+      "step": 184
+    },
+    {
+      "epoch": 0.16442617486945896,
+      "grad_norm": 0.5504122972488403,
+      "learning_rate": 0.0001,
+      "loss": 1.1301,
+      "step": 185
+    },
+    {
+      "epoch": 0.16531496500388845,
+      "grad_norm": 0.5042161345481873,
+      "learning_rate": 0.0001,
+      "loss": 1.0545,
+      "step": 186
+    },
+    {
+      "epoch": 0.16620375513831798,
+      "grad_norm": 0.4888613820075989,
+      "learning_rate": 0.0001,
+      "loss": 1.0766,
+      "step": 187
+    },
+    {
+      "epoch": 0.16709254527274747,
+      "grad_norm": 0.46498018503189087,
+      "learning_rate": 0.0001,
+      "loss": 1.0935,
+      "step": 188
+    },
+    {
+      "epoch": 0.16798133540717697,
+      "grad_norm": 0.4573257863521576,
+      "learning_rate": 0.0001,
+      "loss": 1.0511,
+      "step": 189
+    },
+    {
+      "epoch": 0.1688701255416065,
+      "grad_norm": 0.4621419608592987,
+      "learning_rate": 0.0001,
+      "loss": 1.026,
+      "step": 190
+    },
+    {
+      "epoch": 0.169758915676036,
+      "grad_norm": 0.5240814685821533,
+      "learning_rate": 0.0001,
+      "loss": 0.9897,
+      "step": 191
+    },
+    {
+      "epoch": 0.1706477058104655,
+      "grad_norm": 0.4366868734359741,
+      "learning_rate": 0.0001,
+      "loss": 1.0274,
+      "step": 192
+    },
+    {
+      "epoch": 0.171536495944895,
+      "grad_norm": 0.574884831905365,
+      "learning_rate": 0.0001,
+      "loss": 1.0192,
+      "step": 193
+    },
+    {
+      "epoch": 0.17242528607932453,
+      "grad_norm": 0.5254986882209778,
+      "learning_rate": 0.0001,
+      "loss": 1.0043,
+      "step": 194
+    },
+    {
+      "epoch": 0.17331407621375403,
+      "grad_norm": 0.6957873106002808,
+      "learning_rate": 0.0001,
+      "loss": 1.0281,
+      "step": 195
+    },
+    {
+      "epoch": 0.17420286634818352,
+      "grad_norm": 0.5131996273994446,
+      "learning_rate": 0.0001,
+      "loss": 1.0363,
+      "step": 196
+    },
+    {
+      "epoch": 0.17509165648261305,
+      "grad_norm": 0.5054270029067993,
+      "learning_rate": 0.0001,
+      "loss": 1.0708,
+      "step": 197
+    },
+    {
+      "epoch": 0.17598044661704254,
+      "grad_norm": 0.5029573440551758,
+      "learning_rate": 0.0001,
+      "loss": 1.0562,
+      "step": 198
+    },
+    {
+      "epoch": 0.17686923675147206,
+      "grad_norm": 0.5139032602310181,
+      "learning_rate": 0.0001,
+      "loss": 0.995,
+      "step": 199
+    },
+    {
+      "epoch": 0.17775802688590156,
+      "grad_norm": 0.4500792324542999,
+      "learning_rate": 0.0001,
+      "loss": 1.1012,
+      "step": 200
+    },
+    {
+      "epoch": 0.17864681702033108,
+      "grad_norm": 0.5789426565170288,
+      "learning_rate": 0.0001,
+      "loss": 1.1203,
+      "step": 201
+    },
+    {
+      "epoch": 0.17953560715476058,
+      "grad_norm": 0.48242732882499695,
+      "learning_rate": 0.0001,
+      "loss": 1.0759,
+      "step": 202
+    },
+    {
+      "epoch": 0.1804243972891901,
+      "grad_norm": 0.5667662024497986,
+      "learning_rate": 0.0001,
+      "loss": 1.0595,
+      "step": 203
+    },
+    {
+      "epoch": 0.1813131874236196,
+      "grad_norm": 0.6725609302520752,
+      "learning_rate": 0.0001,
+      "loss": 1.0049,
+      "step": 204
+    },
+    {
+      "epoch": 0.1822019775580491,
+      "grad_norm": 0.5460247993469238,
+      "learning_rate": 0.0001,
+      "loss": 1.0646,
+      "step": 205
+    },
+    {
+      "epoch": 0.18309076769247862,
+      "grad_norm": 0.49915874004364014,
+      "learning_rate": 0.0001,
+      "loss": 1.1236,
+      "step": 206
+    },
+    {
+      "epoch": 0.1839795578269081,
+      "grad_norm": 0.522266149520874,
+      "learning_rate": 0.0001,
+      "loss": 1.0688,
+      "step": 207
+    },
+    {
+      "epoch": 0.18486834796133764,
+      "grad_norm": 0.4875168800354004,
+      "learning_rate": 0.0001,
+      "loss": 1.0596,
+      "step": 208
+    },
+    {
+      "epoch": 0.18575713809576713,
+      "grad_norm": 0.5118414759635925,
+      "learning_rate": 0.0001,
+      "loss": 1.0139,
+      "step": 209
+    },
+    {
+      "epoch": 0.18664592823019666,
+      "grad_norm": 0.5071278214454651,
+      "learning_rate": 0.0001,
+      "loss": 1.1415,
+      "step": 210
+    },
+    {
+      "epoch": 0.18753471836462615,
+      "grad_norm": 0.5847178101539612,
+      "learning_rate": 0.0001,
+      "loss": 1.1495,
+      "step": 211
+    },
+    {
+      "epoch": 0.18842350849905565,
+      "grad_norm": 0.45916759967803955,
+      "learning_rate": 0.0001,
+      "loss": 1.0898,
+      "step": 212
+    },
+    {
+      "epoch": 0.18931229863348517,
+      "grad_norm": 0.4967547655105591,
+      "learning_rate": 0.0001,
+      "loss": 0.946,
+      "step": 213
+    },
+    {
+      "epoch": 0.19020108876791467,
+      "grad_norm": 0.48724737763404846,
+      "learning_rate": 0.0001,
+      "loss": 1.0353,
+      "step": 214
+    },
+    {
+      "epoch": 0.1910898789023442,
+      "grad_norm": 0.4389554262161255,
+      "learning_rate": 0.0001,
+      "loss": 1.0129,
+      "step": 215
+    },
+    {
+      "epoch": 0.19197866903677369,
+      "grad_norm": 0.4678506553173065,
+      "learning_rate": 0.0001,
+      "loss": 1.0903,
+      "step": 216
+    },
+    {
+      "epoch": 0.1928674591712032,
+      "grad_norm": 0.5293782353401184,
+      "learning_rate": 0.0001,
+      "loss": 1.0739,
+      "step": 217
+    },
+    {
+      "epoch": 0.1937562493056327,
+      "grad_norm": 0.45177680253982544,
+      "learning_rate": 0.0001,
+      "loss": 1.1057,
+      "step": 218
+    },
+    {
+      "epoch": 0.19464503944006223,
+      "grad_norm": 0.4647086560726166,
+      "learning_rate": 0.0001,
+      "loss": 1.0894,
+      "step": 219
+    },
+    {
+      "epoch": 0.19553382957449172,
+      "grad_norm": 0.5464223027229309,
+      "learning_rate": 0.0001,
+      "loss": 1.0606,
+      "step": 220
+    },
+    {
+      "epoch": 0.19642261970892122,
+      "grad_norm": 0.5713245868682861,
+      "learning_rate": 0.0001,
+      "loss": 1.0621,
+      "step": 221
+    },
+    {
+      "epoch": 0.19731140984335074,
+      "grad_norm": 0.558101236820221,
+      "learning_rate": 0.0001,
+      "loss": 1.0829,
+      "step": 222
+    },
+    {
+      "epoch": 0.19820019997778024,
+      "grad_norm": 0.775971531867981,
+      "learning_rate": 0.0001,
+      "loss": 0.9982,
+      "step": 223
+    },
+    {
+      "epoch": 0.19908899011220976,
+      "grad_norm": 0.4644327163696289,
+      "learning_rate": 0.0001,
+      "loss": 1.0883,
+      "step": 224
+    },
+    {
+      "epoch": 0.19997778024663926,
+      "grad_norm": 0.4939100444316864,
+      "learning_rate": 0.0001,
+      "loss": 1.0367,
+      "step": 225
+    },
+    {
+      "epoch": 0.20086657038106878,
+      "grad_norm": 0.43425729870796204,
+      "learning_rate": 0.0001,
+      "loss": 1.0038,
+      "step": 226
+    },
+    {
+      "epoch": 0.20175536051549828,
+      "grad_norm": 0.508567750453949,
+      "learning_rate": 0.0001,
+      "loss": 1.1134,
+      "step": 227
+    },
+    {
+      "epoch": 0.20264415064992777,
+      "grad_norm": 0.397948294878006,
+      "learning_rate": 0.0001,
+      "loss": 0.9837,
+      "step": 228
+    },
+    {
+      "epoch": 0.2035329407843573,
+      "grad_norm": 0.5290727019309998,
+      "learning_rate": 0.0001,
+      "loss": 1.0965,
+      "step": 229
+    },
+    {
+      "epoch": 0.2044217309187868,
+      "grad_norm": 0.5163105130195618,
+      "learning_rate": 0.0001,
+      "loss": 0.9927,
+      "step": 230
+    },
+    {
+      "epoch": 0.20531052105321632,
+      "grad_norm": 0.4882635474205017,
+      "learning_rate": 0.0001,
+      "loss": 1.0316,
+      "step": 231
+    },
+    {
+      "epoch": 0.2061993111876458,
+      "grad_norm": 0.471646785736084,
+      "learning_rate": 0.0001,
+      "loss": 1.0645,
+      "step": 232
+    },
+    {
+      "epoch": 0.20708810132207534,
+      "grad_norm": 0.5159053206443787,
+      "learning_rate": 0.0001,
+      "loss": 1.0898,
+      "step": 233
+    },
+    {
+      "epoch": 0.20797689145650483,
+      "grad_norm": 0.5597699880599976,
+      "learning_rate": 0.0001,
+      "loss": 1.0726,
+      "step": 234
+    },
+    {
+      "epoch": 0.20886568159093435,
+      "grad_norm": 0.463168203830719,
+      "learning_rate": 0.0001,
+      "loss": 1.0349,
+      "step": 235
+    },
+    {
+      "epoch": 0.20975447172536385,
+      "grad_norm": 0.47213372588157654,
+      "learning_rate": 0.0001,
+      "loss": 1.08,
+      "step": 236
+    },
+    {
+      "epoch": 0.21064326185979335,
+      "grad_norm": 0.46547219157218933,
+      "learning_rate": 0.0001,
+      "loss": 1.084,
+      "step": 237
+    },
+    {
+      "epoch": 0.21153205199422287,
+      "grad_norm": 0.4608129858970642,
+      "learning_rate": 0.0001,
+      "loss": 1.0441,
+      "step": 238
+    },
+    {
+      "epoch": 0.21242084212865237,
+      "grad_norm": 0.5103969573974609,
+      "learning_rate": 0.0001,
+      "loss": 1.0845,
+      "step": 239
+    },
+    {
+      "epoch": 0.2133096322630819,
+      "grad_norm": 0.48750364780426025,
+      "learning_rate": 0.0001,
+      "loss": 1.01,
+      "step": 240
+    },
+    {
+      "epoch": 0.21419842239751138,
+      "grad_norm": 0.4955364465713501,
+      "learning_rate": 0.0001,
+      "loss": 1.0824,
+      "step": 241
+    },
+    {
+      "epoch": 0.2150872125319409,
+      "grad_norm": 0.4669419527053833,
+      "learning_rate": 0.0001,
+      "loss": 1.0061,
+      "step": 242
+    },
+    {
+      "epoch": 0.2159760026663704,
+      "grad_norm": 0.4938861131668091,
+      "learning_rate": 0.0001,
+      "loss": 0.9739,
+      "step": 243
+    },
+    {
+      "epoch": 0.2168647928007999,
+      "grad_norm": 0.5369840860366821,
+      "learning_rate": 0.0001,
+      "loss": 1.1085,
+      "step": 244
+    },
+    {
+      "epoch": 0.21775358293522942,
+      "grad_norm": 0.5239328742027283,
+      "learning_rate": 0.0001,
+      "loss": 0.9869,
+      "step": 245
+    },
+    {
+      "epoch": 0.21864237306965892,
+      "grad_norm": 0.518187403678894,
+      "learning_rate": 0.0001,
+      "loss": 1.046,
+      "step": 246
+    },
+    {
+      "epoch": 0.21953116320408844,
+      "grad_norm": 0.44194117188453674,
+      "learning_rate": 0.0001,
+      "loss": 1.0234,
+      "step": 247
+    },
+    {
+      "epoch": 0.22041995333851794,
+      "grad_norm": 0.44847285747528076,
+      "learning_rate": 0.0001,
+      "loss": 1.0692,
+      "step": 248
+    },
+    {
+      "epoch": 0.22130874347294746,
+      "grad_norm": 0.4702758491039276,
+      "learning_rate": 0.0001,
+      "loss": 1.1231,
+      "step": 249
+    },
+    {
+      "epoch": 0.22219753360737696,
+      "grad_norm": 0.43607068061828613,
+      "learning_rate": 0.0001,
+      "loss": 1.0197,
+      "step": 250
+    },
+    {
+      "epoch": 0.22308632374180645,
+      "grad_norm": 0.5397356748580933,
+      "learning_rate": 0.0001,
+      "loss": 1.0527,
+      "step": 251
+    },
+    {
+      "epoch": 0.22397511387623598,
+      "grad_norm": 0.5345639586448669,
+      "learning_rate": 0.0001,
+      "loss": 1.0119,
+      "step": 252
+    },
+    {
+      "epoch": 0.22486390401066547,
+      "grad_norm": 0.45763713121414185,
+      "learning_rate": 0.0001,
+      "loss": 1.0296,
+      "step": 253
+    },
+    {
+      "epoch": 0.225752694145095,
+      "grad_norm": 0.49265143275260925,
+      "learning_rate": 0.0001,
+      "loss": 1.0171,
+      "step": 254
+    },
+    {
+      "epoch": 0.2266414842795245,
+      "grad_norm": 0.5668004751205444,
+      "learning_rate": 0.0001,
+      "loss": 1.0085,
+      "step": 255
+    },
+    {
+      "epoch": 0.22753027441395401,
+      "grad_norm": 0.5550284385681152,
+      "learning_rate": 0.0001,
+      "loss": 1.0006,
+      "step": 256
+    },
+    {
+      "epoch": 0.2284190645483835,
+      "grad_norm": 0.45947736501693726,
+      "learning_rate": 0.0001,
+      "loss": 1.0089,
+      "step": 257
+    },
+    {
+      "epoch": 0.22930785468281303,
+      "grad_norm": 0.5964300632476807,
+      "learning_rate": 0.0001,
+      "loss": 1.0854,
+      "step": 258
+    },
+    {
+      "epoch": 0.23019664481724253,
+      "grad_norm": 0.5322341322898865,
+      "learning_rate": 0.0001,
+      "loss": 1.0521,
+      "step": 259
+    },
+    {
+      "epoch": 0.23108543495167203,
+      "grad_norm": 0.48153069615364075,
+      "learning_rate": 0.0001,
+      "loss": 0.9444,
+      "step": 260
+    },
+    {
+      "epoch": 0.23197422508610155,
+      "grad_norm": 0.45373499393463135,
+      "learning_rate": 0.0001,
+      "loss": 1.047,
+      "step": 261
+    },
+    {
+      "epoch": 0.23286301522053104,
+      "grad_norm": 0.43222519755363464,
+      "learning_rate": 0.0001,
+      "loss": 0.973,
+      "step": 262
+    },
+    {
+      "epoch": 0.23375180535496057,
+      "grad_norm": 0.5555695295333862,
+      "learning_rate": 0.0001,
+      "loss": 1.0127,
+      "step": 263
+    },
+    {
+      "epoch": 0.23464059548939006,
+      "grad_norm": 0.47663983702659607,
+      "learning_rate": 0.0001,
+      "loss": 1.0688,
+      "step": 264
+    },
+    {
+      "epoch": 0.2355293856238196,
+      "grad_norm": 0.5549951791763306,
+      "learning_rate": 0.0001,
+      "loss": 1.0433,
+      "step": 265
+    },
+    {
+      "epoch": 0.23641817575824908,
+      "grad_norm": 0.4606119990348816,
+      "learning_rate": 0.0001,
+      "loss": 0.9985,
+      "step": 266
+    },
+    {
+      "epoch": 0.23730696589267858,
+      "grad_norm": 0.5273025631904602,
+      "learning_rate": 0.0001,
+      "loss": 1.018,
+      "step": 267
+    },
+    {
+      "epoch": 0.2381957560271081,
+      "grad_norm": 0.49982166290283203,
+      "learning_rate": 0.0001,
+      "loss": 0.9592,
+      "step": 268
+    },
+    {
+      "epoch": 0.2390845461615376,
+      "grad_norm": 0.5359534621238708,
+      "learning_rate": 0.0001,
+      "loss": 1.0568,
+      "step": 269
+    },
+    {
+      "epoch": 0.23997333629596712,
+      "grad_norm": 0.48276546597480774,
+      "learning_rate": 0.0001,
+      "loss": 1.0315,
+      "step": 270
+    },
+    {
+      "epoch": 0.24086212643039662,
+      "grad_norm": 0.48728588223457336,
+      "learning_rate": 0.0001,
+      "loss": 1.0827,
+      "step": 271
+    },
+    {
+      "epoch": 0.24175091656482614,
+      "grad_norm": 2.621769428253174,
+      "learning_rate": 0.0001,
+      "loss": 0.997,
+      "step": 272
+    },
+    {
+      "epoch": 0.24263970669925564,
+      "grad_norm": 0.5768371820449829,
+      "learning_rate": 0.0001,
+      "loss": 0.9872,
+      "step": 273
+    },
+    {
+      "epoch": 0.24352849683368516,
+      "grad_norm": 0.5989674925804138,
+      "learning_rate": 0.0001,
+      "loss": 0.9232,
+      "step": 274
+    },
+    {
+      "epoch": 0.24441728696811466,
+      "grad_norm": 0.5455823540687561,
+      "learning_rate": 0.0001,
+      "loss": 1.0062,
+      "step": 275
+    },
+    {
+      "epoch": 0.24530607710254415,
+      "grad_norm": 0.5664008855819702,
+      "learning_rate": 0.0001,
+      "loss": 1.012,
+      "step": 276
+    },
+    {
+      "epoch": 0.24619486723697367,
+      "grad_norm": 0.5294925570487976,
+      "learning_rate": 0.0001,
+      "loss": 1.0285,
+      "step": 277
+    },
+    {
+      "epoch": 0.24708365737140317,
+      "grad_norm": 0.5203514099121094,
+      "learning_rate": 0.0001,
+      "loss": 1.0388,
+      "step": 278
+    },
+    {
+      "epoch": 0.2479724475058327,
+      "grad_norm": 0.5114679336547852,
+      "learning_rate": 0.0001,
+      "loss": 1.0388,
+      "step": 279
+    },
+    {
+      "epoch": 0.2488612376402622,
+      "grad_norm": 0.48817357420921326,
+      "learning_rate": 0.0001,
+      "loss": 0.9768,
+      "step": 280
+    },
+    {
+      "epoch": 0.2497500277746917,
+      "grad_norm": 0.4958358108997345,
+      "learning_rate": 0.0001,
+      "loss": 1.0239,
+      "step": 281
+    },
+    {
+      "epoch": 0.25063881790912124,
+      "grad_norm": 0.4603129029273987,
+      "learning_rate": 0.0001,
+      "loss": 1.0371,
+      "step": 282
+    },
+    {
+      "epoch": 0.25152760804355073,
+      "grad_norm": 0.48228880763053894,
+      "learning_rate": 0.0001,
+      "loss": 1.04,
+      "step": 283
+    },
+    {
+      "epoch": 0.25241639817798023,
+      "grad_norm": 0.5364854335784912,
+      "learning_rate": 0.0001,
+      "loss": 1.0049,
+      "step": 284
+    },
+    {
+      "epoch": 0.2533051883124097,
+      "grad_norm": 0.46868863701820374,
+      "learning_rate": 0.0001,
+      "loss": 0.9842,
+      "step": 285
+    },
+    {
+      "epoch": 0.2541939784468392,
+      "grad_norm": 0.4731464385986328,
+      "learning_rate": 0.0001,
+      "loss": 1.0195,
+      "step": 286
+    },
+    {
+      "epoch": 0.25508276858126877,
+      "grad_norm": 0.5144749879837036,
+      "learning_rate": 0.0001,
+      "loss": 1.0673,
+      "step": 287
+    },
+    {
+      "epoch": 0.25597155871569827,
+      "grad_norm": 0.44826894998550415,
+      "learning_rate": 0.0001,
+      "loss": 0.9981,
+      "step": 288
+    },
+    {
+      "epoch": 0.25686034885012776,
+      "grad_norm": 0.4612467885017395,
+      "learning_rate": 0.0001,
+      "loss": 1.0318,
+      "step": 289
+    },
+    {
+      "epoch": 0.25774913898455726,
+      "grad_norm": 0.4774060845375061,
+      "learning_rate": 0.0001,
+      "loss": 0.9806,
+      "step": 290
+    },
+    {
+      "epoch": 0.25863792911898675,
+      "grad_norm": 0.4614105820655823,
+      "learning_rate": 0.0001,
+      "loss": 1.0404,
+      "step": 291
+    },
+    {
+      "epoch": 0.2595267192534163,
+      "grad_norm": 0.47314900159835815,
+      "learning_rate": 0.0001,
+      "loss": 0.9932,
+      "step": 292
+    },
+    {
+      "epoch": 0.2604155093878458,
+      "grad_norm": 0.5009798407554626,
+      "learning_rate": 0.0001,
+      "loss": 1.1416,
+      "step": 293
+    },
+    {
+      "epoch": 0.2613042995222753,
+      "grad_norm": 0.4358547329902649,
+      "learning_rate": 0.0001,
+      "loss": 1.0379,
+      "step": 294
+    },
+    {
+      "epoch": 0.2621930896567048,
+      "grad_norm": 0.5527064204216003,
+      "learning_rate": 0.0001,
+      "loss": 0.9843,
+      "step": 295
+    },
+    {
+      "epoch": 0.26308187979113434,
+      "grad_norm": 0.47958534955978394,
+      "learning_rate": 0.0001,
+      "loss": 1.0235,
+      "step": 296
+    },
+    {
+      "epoch": 0.26397066992556384,
+      "grad_norm": 0.4394091069698334,
+      "learning_rate": 0.0001,
+      "loss": 0.9711,
+      "step": 297
+    },
+    {
+      "epoch": 0.26485946005999333,
+      "grad_norm": 0.5663338899612427,
+      "learning_rate": 0.0001,
+      "loss": 1.029,
+      "step": 298
+    },
+    {
+      "epoch": 0.26574825019442283,
+      "grad_norm": 0.5038536787033081,
+      "learning_rate": 0.0001,
+      "loss": 1.0097,
+      "step": 299
+    },
+    {
+      "epoch": 0.2666370403288523,
+      "grad_norm": 0.5386257767677307,
+      "learning_rate": 0.0001,
+      "loss": 1.0539,
+      "step": 300
+    },
+    {
+      "epoch": 0.2675258304632819,
+      "grad_norm": 0.4892950654029846,
+      "learning_rate": 0.0001,
+      "loss": 0.99,
+      "step": 301
+    },
+    {
+      "epoch": 0.2684146205977114,
+      "grad_norm": 0.5311320424079895,
+      "learning_rate": 0.0001,
+      "loss": 1.0922,
+      "step": 302
+    },
+    {
+      "epoch": 0.26930341073214087,
+      "grad_norm": 0.4981628656387329,
+      "learning_rate": 0.0001,
+      "loss": 1.0332,
+      "step": 303
+    },
+    {
+      "epoch": 0.27019220086657036,
+      "grad_norm": 0.6175600290298462,
+      "learning_rate": 0.0001,
+      "loss": 1.0552,
+      "step": 304
+    },
+    {
+      "epoch": 0.2710809910009999,
+      "grad_norm": 0.5665920376777649,
+      "learning_rate": 0.0001,
+      "loss": 0.9807,
+      "step": 305
+    },
+    {
+      "epoch": 0.2719697811354294,
+      "grad_norm": 0.4487343430519104,
+      "learning_rate": 0.0001,
+      "loss": 0.9961,
+      "step": 306
+    },
+    {
+      "epoch": 0.2728585712698589,
+      "grad_norm": 0.4994884133338928,
+      "learning_rate": 0.0001,
+      "loss": 0.9809,
+      "step": 307
+    },
+    {
+      "epoch": 0.2737473614042884,
+      "grad_norm": 0.5361630916595459,
+      "learning_rate": 0.0001,
+      "loss": 1.0396,
+      "step": 308
+    },
+    {
+      "epoch": 0.2746361515387179,
+      "grad_norm": 0.49830833077430725,
+      "learning_rate": 0.0001,
+      "loss": 0.966,
+      "step": 309
+    },
+    {
+      "epoch": 0.27552494167314745,
+      "grad_norm": 0.5319890975952148,
+      "learning_rate": 0.0001,
+      "loss": 1.0121,
+      "step": 310
+    },
+    {
+      "epoch": 0.27641373180757695,
+      "grad_norm": 0.5802849531173706,
+      "learning_rate": 0.0001,
+      "loss": 1.0198,
+      "step": 311
+    },
+    {
+      "epoch": 0.27730252194200644,
+      "grad_norm": 0.5048151016235352,
+      "learning_rate": 0.0001,
+      "loss": 0.9996,
+      "step": 312
+    },
+    {
+      "epoch": 0.27819131207643594,
+      "grad_norm": 0.44712725281715393,
+      "learning_rate": 0.0001,
+      "loss": 0.9757,
+      "step": 313
+    },
+    {
+      "epoch": 0.27908010221086543,
+      "grad_norm": 0.48854947090148926,
+      "learning_rate": 0.0001,
+      "loss": 1.0514,
+      "step": 314
+    },
+    {
+      "epoch": 0.279968892345295,
+      "grad_norm": 0.479911208152771,
+      "learning_rate": 0.0001,
+      "loss": 0.9588,
+      "step": 315
+    },
+    {
+      "epoch": 0.2808576824797245,
+      "grad_norm": 0.46088626980781555,
+      "learning_rate": 0.0001,
+      "loss": 1.0577,
+      "step": 316
+    },
+    {
+      "epoch": 0.281746472614154,
+      "grad_norm": 0.5798197388648987,
+      "learning_rate": 0.0001,
+      "loss": 1.0011,
+      "step": 317
+    },
+    {
+      "epoch": 0.28263526274858347,
+      "grad_norm": 0.45959633588790894,
+      "learning_rate": 0.0001,
+      "loss": 0.9358,
+      "step": 318
+    },
+    {
+      "epoch": 0.283524052883013,
+      "grad_norm": 0.4906099736690521,
+      "learning_rate": 0.0001,
+      "loss": 0.9432,
+      "step": 319
+    },
+    {
+      "epoch": 0.2844128430174425,
+      "grad_norm": 0.4779829680919647,
+      "learning_rate": 0.0001,
+      "loss": 0.9471,
+      "step": 320
+    },
+    {
+      "epoch": 0.285301633151872,
+      "grad_norm": 0.5950011014938354,
+      "learning_rate": 0.0001,
+      "loss": 1.0712,
+      "step": 321
+    },
+    {
+      "epoch": 0.2861904232863015,
+      "grad_norm": 0.5539191365242004,
+      "learning_rate": 0.0001,
+      "loss": 1.0329,
+      "step": 322
+    },
+    {
+      "epoch": 0.287079213420731,
+      "grad_norm": 0.44738510251045227,
+      "learning_rate": 0.0001,
+      "loss": 0.9935,
+      "step": 323
+    },
+    {
+      "epoch": 0.28796800355516056,
+      "grad_norm": 0.4675084054470062,
+      "learning_rate": 0.0001,
+      "loss": 0.9979,
+      "step": 324
+    },
+    {
+      "epoch": 0.28885679368959005,
+      "grad_norm": 0.38852187991142273,
+      "learning_rate": 0.0001,
+      "loss": 0.9497,
+      "step": 325
+    },
+    {
+      "epoch": 0.28974558382401955,
+      "grad_norm": 0.5008799433708191,
+      "learning_rate": 0.0001,
+      "loss": 0.9904,
+      "step": 326
+    },
+    {
+      "epoch": 0.29063437395844904,
+      "grad_norm": 0.4492400288581848,
+      "learning_rate": 0.0001,
+      "loss": 1.0489,
+      "step": 327
+    },
+    {
+      "epoch": 0.2915231640928786,
+      "grad_norm": 0.48956772685050964,
+      "learning_rate": 0.0001,
+      "loss": 1.0061,
+      "step": 328
+    },
+    {
+      "epoch": 0.2924119542273081,
+      "grad_norm": 0.5612773299217224,
+      "learning_rate": 0.0001,
+      "loss": 1.0947,
+      "step": 329
+    },
+    {
+      "epoch": 0.2933007443617376,
+      "grad_norm": 0.5352462530136108,
+      "learning_rate": 0.0001,
+      "loss": 1.1113,
+      "step": 330
+    },
+    {
+      "epoch": 0.2941895344961671,
+      "grad_norm": 0.43017029762268066,
+      "learning_rate": 0.0001,
+      "loss": 1.0291,
+      "step": 331
+    },
+    {
+      "epoch": 0.2950783246305966,
+      "grad_norm": 0.5087767839431763,
+      "learning_rate": 0.0001,
+      "loss": 1.0897,
+      "step": 332
+    },
+    {
+      "epoch": 0.29596711476502613,
+      "grad_norm": 0.38609907031059265,
+      "learning_rate": 0.0001,
+      "loss": 0.9629,
+      "step": 333
+    },
+    {
+      "epoch": 0.2968559048994556,
+      "grad_norm": 0.4797438979148865,
+      "learning_rate": 0.0001,
+      "loss": 0.9808,
+      "step": 334
+    },
+    {
+      "epoch": 0.2977446950338851,
+      "grad_norm": 0.4882568418979645,
+      "learning_rate": 0.0001,
+      "loss": 1.0812,
+      "step": 335
+    },
+    {
+      "epoch": 0.2986334851683146,
+      "grad_norm": 0.4409843385219574,
+      "learning_rate": 0.0001,
+      "loss": 1.0034,
+      "step": 336
+    },
+    {
+      "epoch": 0.2995222753027441,
+      "grad_norm": 0.5104953646659851,
+      "learning_rate": 0.0001,
+      "loss": 0.9186,
+      "step": 337
+    },
+    {
+      "epoch": 0.30041106543717366,
+      "grad_norm": 0.45493751764297485,
+      "learning_rate": 0.0001,
+      "loss": 0.9059,
+      "step": 338
+    },
+    {
+      "epoch": 0.30129985557160316,
+      "grad_norm": 0.4311971068382263,
+      "learning_rate": 0.0001,
+      "loss": 1.042,
+      "step": 339
+    },
+    {
+      "epoch": 0.30218864570603265,
+      "grad_norm": 0.43054771423339844,
+      "learning_rate": 0.0001,
+      "loss": 1.0336,
+      "step": 340
+    },
+    {
+      "epoch": 0.30307743584046215,
+      "grad_norm": 0.4950178861618042,
+      "learning_rate": 0.0001,
+      "loss": 1.0515,
+      "step": 341
+    },
+    {
+      "epoch": 0.3039662259748917,
+      "grad_norm": 0.4074246287345886,
+      "learning_rate": 0.0001,
+      "loss": 0.9163,
+      "step": 342
+    },
+    {
+      "epoch": 0.3048550161093212,
+      "grad_norm": 0.45394015312194824,
+      "learning_rate": 0.0001,
+      "loss": 1.0499,
+      "step": 343
+    },
+    {
+      "epoch": 0.3057438062437507,
+      "grad_norm": 0.453685998916626,
+      "learning_rate": 0.0001,
+      "loss": 1.021,
+      "step": 344
+    },
+    {
+      "epoch": 0.3066325963781802,
+      "grad_norm": 0.4687478542327881,
+      "learning_rate": 0.0001,
+      "loss": 1.0305,
+      "step": 345
+    },
+    {
+      "epoch": 0.3075213865126097,
+      "grad_norm": 0.4402327239513397,
+      "learning_rate": 0.0001,
+      "loss": 0.9547,
+      "step": 346
+    },
+    {
+      "epoch": 0.30841017664703924,
+      "grad_norm": 0.41369926929473877,
+      "learning_rate": 0.0001,
+      "loss": 0.981,
+      "step": 347
+    },
+    {
+      "epoch": 0.30929896678146873,
+      "grad_norm": 0.4821121096611023,
+      "learning_rate": 0.0001,
+      "loss": 0.9427,
+      "step": 348
+    },
+    {
+      "epoch": 0.3101877569158982,
+      "grad_norm": 0.4524393379688263,
+      "learning_rate": 0.0001,
+      "loss": 1.0277,
+      "step": 349
+    },
+    {
+      "epoch": 0.3110765470503277,
+      "grad_norm": 0.49870559573173523,
+      "learning_rate": 0.0001,
+      "loss": 1.0075,
+      "step": 350
+    },
+    {
+      "epoch": 0.3119653371847573,
+      "grad_norm": 0.5591082572937012,
+      "learning_rate": 0.0001,
+      "loss": 0.9901,
+      "step": 351
+    },
+    {
+      "epoch": 0.31285412731918677,
+      "grad_norm": 0.4724654257297516,
+      "learning_rate": 0.0001,
+      "loss": 0.9909,
+      "step": 352
+    },
+    {
+      "epoch": 0.31374291745361627,
+      "grad_norm": 0.4514271020889282,
+      "learning_rate": 0.0001,
+      "loss": 1.0259,
+      "step": 353
+    },
+    {
+      "epoch": 0.31463170758804576,
+      "grad_norm": 0.5684020519256592,
+      "learning_rate": 0.0001,
+      "loss": 0.9933,
+      "step": 354
+    },
+    {
+      "epoch": 0.31552049772247526,
+      "grad_norm": 0.442047119140625,
+      "learning_rate": 0.0001,
+      "loss": 1.0394,
+      "step": 355
+    },
+    {
+      "epoch": 0.3164092878569048,
+      "grad_norm": 0.47112616896629333,
+      "learning_rate": 0.0001,
+      "loss": 0.9287,
+      "step": 356
+    },
+    {
+      "epoch": 0.3172980779913343,
+      "grad_norm": 0.5187697410583496,
+      "learning_rate": 0.0001,
+      "loss": 0.9476,
+      "step": 357
+    },
+    {
+      "epoch": 0.3181868681257638,
+      "grad_norm": 0.43961942195892334,
+      "learning_rate": 0.0001,
+      "loss": 0.9474,
+      "step": 358
+    },
+    {
+      "epoch": 0.3190756582601933,
+      "grad_norm": 0.47375670075416565,
+      "learning_rate": 0.0001,
+      "loss": 1.0088,
+      "step": 359
+    },
+    {
+      "epoch": 0.31996444839462285,
+      "grad_norm": 0.4810321629047394,
+      "learning_rate": 0.0001,
+      "loss": 0.9628,
+      "step": 360
+    },
+    {
+      "epoch": 0.32085323852905234,
+      "grad_norm": 0.5187336802482605,
+      "learning_rate": 0.0001,
+      "loss": 1.0055,
+      "step": 361
+    },
+    {
+      "epoch": 0.32174202866348184,
+      "grad_norm": 0.47308310866355896,
+      "learning_rate": 0.0001,
+      "loss": 1.0005,
+      "step": 362
+    },
+    {
+      "epoch": 0.32263081879791133,
+      "grad_norm": 1.5481101274490356,
+      "learning_rate": 0.0001,
+      "loss": 1.0754,
+      "step": 363
+    },
+    {
+      "epoch": 0.32351960893234083,
+      "grad_norm": 0.4808347523212433,
+      "learning_rate": 0.0001,
+      "loss": 1.0567,
+      "step": 364
+    },
+    {
+      "epoch": 0.3244083990667704,
+      "grad_norm": 0.40874922275543213,
+      "learning_rate": 0.0001,
+      "loss": 1.0153,
+      "step": 365
+    },
+    {
+      "epoch": 0.3252971892011999,
+      "grad_norm": 0.5171230435371399,
+      "learning_rate": 0.0001,
+      "loss": 0.9808,
+      "step": 366
+    },
+    {
+      "epoch": 0.3261859793356294,
+      "grad_norm": 0.48159992694854736,
+      "learning_rate": 0.0001,
+      "loss": 0.9873,
+      "step": 367
+    },
+    {
+      "epoch": 0.32707476947005887,
+      "grad_norm": 0.44044238328933716,
+      "learning_rate": 0.0001,
+      "loss": 0.9608,
+      "step": 368
+    },
+    {
+      "epoch": 0.32796355960448836,
+      "grad_norm": 0.4674980342388153,
+      "learning_rate": 0.0001,
+      "loss": 0.9263,
+      "step": 369
+    },
+    {
+      "epoch": 0.3288523497389179,
+      "grad_norm": 0.5395987033843994,
+      "learning_rate": 0.0001,
+      "loss": 0.9548,
+      "step": 370
+    },
+    {
+      "epoch": 0.3297411398733474,
+      "grad_norm": 0.49539071321487427,
+      "learning_rate": 0.0001,
+      "loss": 1.0202,
+      "step": 371
+    },
+    {
+      "epoch": 0.3306299300077769,
+      "grad_norm": 0.4859803020954132,
+      "learning_rate": 0.0001,
+      "loss": 0.9388,
+      "step": 372
+    },
+    {
+      "epoch": 0.3315187201422064,
+      "grad_norm": 0.4504952132701874,
+      "learning_rate": 0.0001,
+      "loss": 1.0092,
+      "step": 373
+    },
+    {
+      "epoch": 0.33240751027663595,
+      "grad_norm": 0.5386714339256287,
+      "learning_rate": 0.0001,
+      "loss": 1.0461,
+      "step": 374
+    },
+    {
+      "epoch": 0.33329630041106545,
+      "grad_norm": 0.5611424446105957,
+      "learning_rate": 0.0001,
+      "loss": 1.0967,
+      "step": 375
+    },
+    {
+      "epoch": 0.33418509054549494,
+      "grad_norm": 0.44047975540161133,
+      "learning_rate": 0.0001,
+      "loss": 0.8973,
+      "step": 376
+    },
+    {
+      "epoch": 0.33507388067992444,
+      "grad_norm": 0.5137032866477966,
+      "learning_rate": 0.0001,
+      "loss": 1.0303,
+      "step": 377
+    },
+    {
+      "epoch": 0.33596267081435394,
+      "grad_norm": 0.47674351930618286,
+      "learning_rate": 0.0001,
+      "loss": 1.0713,
+      "step": 378
+    },
+    {
+      "epoch": 0.3368514609487835,
+      "grad_norm": 0.4222189486026764,
+      "learning_rate": 0.0001,
+      "loss": 0.9894,
+      "step": 379
+    },
+    {
+      "epoch": 0.337740251083213,
+      "grad_norm": 0.41975629329681396,
+      "learning_rate": 0.0001,
+      "loss": 0.9526,
+      "step": 380
+    },
+    {
+      "epoch": 0.3386290412176425,
+      "grad_norm": 0.4654853045940399,
+      "learning_rate": 0.0001,
+      "loss": 0.8748,
+      "step": 381
+    },
+    {
+      "epoch": 0.339517831352072,
+      "grad_norm": 0.39208483695983887,
+      "learning_rate": 0.0001,
+      "loss": 0.9799,
+      "step": 382
+    },
+    {
+      "epoch": 0.3404066214865015,
+      "grad_norm": 0.5432955026626587,
+      "learning_rate": 0.0001,
+      "loss": 1.061,
+      "step": 383
+    },
+    {
+      "epoch": 0.341295411620931,
+      "grad_norm": 0.44643473625183105,
+      "learning_rate": 0.0001,
+      "loss": 0.9823,
+      "step": 384
+    },
+    {
+      "epoch": 0.3421842017553605,
+      "grad_norm": 0.5307298302650452,
+      "learning_rate": 0.0001,
+      "loss": 1.0406,
+      "step": 385
+    },
+    {
+      "epoch": 0.34307299188979,
+      "grad_norm": 0.5106935501098633,
+      "learning_rate": 0.0001,
+      "loss": 1.0473,
+      "step": 386
+    },
+    {
+      "epoch": 0.3439617820242195,
+      "grad_norm": 0.4799475371837616,
+      "learning_rate": 0.0001,
+      "loss": 1.0546,
+      "step": 387
+    },
+    {
+      "epoch": 0.34485057215864906,
+      "grad_norm": 0.37563350796699524,
+      "learning_rate": 0.0001,
+      "loss": 1.0025,
+      "step": 388
+    },
+    {
+      "epoch": 0.34573936229307856,
+      "grad_norm": 0.43214109539985657,
+      "learning_rate": 0.0001,
+      "loss": 0.9771,
+      "step": 389
+    },
+    {
+      "epoch": 0.34662815242750805,
+      "grad_norm": 0.41184356808662415,
+      "learning_rate": 0.0001,
+      "loss": 0.9997,
+      "step": 390
+    },
+    {
+      "epoch": 0.34751694256193755,
+      "grad_norm": 0.49145662784576416,
+      "learning_rate": 0.0001,
+      "loss": 1.0838,
+      "step": 391
+    },
+    {
+      "epoch": 0.34840573269636704,
+      "grad_norm": 0.39192360639572144,
+      "learning_rate": 0.0001,
+      "loss": 0.9671,
+      "step": 392
+    },
+    {
+      "epoch": 0.3492945228307966,
+      "grad_norm": 0.49615415930747986,
+      "learning_rate": 0.0001,
+      "loss": 0.9974,
+      "step": 393
+    },
+    {
+      "epoch": 0.3501833129652261,
+      "grad_norm": 0.48595911264419556,
+      "learning_rate": 0.0001,
+      "loss": 0.9552,
+      "step": 394
+    },
+    {
+      "epoch": 0.3510721030996556,
+      "grad_norm": 0.4774535894393921,
+      "learning_rate": 0.0001,
+      "loss": 0.9908,
+      "step": 395
+    },
+    {
+      "epoch": 0.3519608932340851,
+      "grad_norm": 0.42019304633140564,
+      "learning_rate": 0.0001,
+      "loss": 0.9919,
+      "step": 396
+    },
+    {
+      "epoch": 0.35284968336851463,
+      "grad_norm": 0.4863130450248718,
+      "learning_rate": 0.0001,
+      "loss": 0.9888,
+      "step": 397
+    },
+    {
+      "epoch": 0.35373847350294413,
+      "grad_norm": 0.5684654712677002,
+      "learning_rate": 0.0001,
+      "loss": 1.0284,
+      "step": 398
+    },
+    {
+      "epoch": 0.3546272636373736,
+      "grad_norm": 0.466160386800766,
+      "learning_rate": 0.0001,
+      "loss": 0.999,
+      "step": 399
+    },
+    {
+      "epoch": 0.3555160537718031,
+      "grad_norm": 0.4259321689605713,
+      "learning_rate": 0.0001,
+      "loss": 0.9475,
+      "step": 400
+    },
+    {
+      "epoch": 0.3564048439062326,
+      "grad_norm": 0.4329473376274109,
+      "learning_rate": 0.0001,
+      "loss": 0.971,
+      "step": 401
+    },
+    {
+      "epoch": 0.35729363404066217,
+      "grad_norm": 0.44069400429725647,
+      "learning_rate": 0.0001,
+      "loss": 1.0213,
+      "step": 402
+    },
+    {
+      "epoch": 0.35818242417509166,
+      "grad_norm": 0.46225759387016296,
+      "learning_rate": 0.0001,
+      "loss": 0.9891,
+      "step": 403
+    },
+    {
+      "epoch": 0.35907121430952116,
+      "grad_norm": 0.4280588924884796,
+      "learning_rate": 0.0001,
+      "loss": 1.1054,
+      "step": 404
+    },
+    {
+      "epoch": 0.35996000444395065,
+      "grad_norm": 0.4039415419101715,
+      "learning_rate": 0.0001,
+      "loss": 0.9852,
+      "step": 405
+    },
+    {
+      "epoch": 0.3608487945783802,
+      "grad_norm": 0.45364150404930115,
+      "learning_rate": 0.0001,
+      "loss": 1.0471,
+      "step": 406
+    },
+    {
+      "epoch": 0.3617375847128097,
+      "grad_norm": 0.5891258716583252,
+      "learning_rate": 0.0001,
+      "loss": 1.103,
+      "step": 407
+    },
+    {
+      "epoch": 0.3626263748472392,
+      "grad_norm": 0.47017595171928406,
+      "learning_rate": 0.0001,
+      "loss": 0.8947,
+      "step": 408
+    },
+    {
+      "epoch": 0.3635151649816687,
+      "grad_norm": 0.43023166060447693,
+      "learning_rate": 0.0001,
+      "loss": 1.0324,
+      "step": 409
+    },
+    {
+      "epoch": 0.3644039551160982,
+      "grad_norm": 0.45753541588783264,
+      "learning_rate": 0.0001,
+      "loss": 0.9737,
+      "step": 410
+    },
+    {
+      "epoch": 0.36529274525052774,
+      "grad_norm": 0.48378103971481323,
+      "learning_rate": 0.0001,
+      "loss": 1.059,
+      "step": 411
+    },
+    {
+      "epoch": 0.36618153538495724,
+      "grad_norm": 0.4665151834487915,
+      "learning_rate": 0.0001,
+      "loss": 0.9722,
+      "step": 412
+    },
+    {
+      "epoch": 0.36707032551938673,
+      "grad_norm": 5.9081315994262695,
+      "learning_rate": 0.0001,
+      "loss": 0.9235,
+      "step": 413
+    },
+    {
+      "epoch": 0.3679591156538162,
+      "grad_norm": 0.42533057928085327,
+      "learning_rate": 0.0001,
+      "loss": 0.9788,
+      "step": 414
+    },
+    {
+      "epoch": 0.3688479057882458,
+      "grad_norm": 0.5004814267158508,
+      "learning_rate": 0.0001,
+      "loss": 1.003,
+      "step": 415
+    },
+    {
+      "epoch": 0.3697366959226753,
+      "grad_norm": 0.610554039478302,
+      "learning_rate": 0.0001,
+      "loss": 1.0865,
+      "step": 416
+    },
+    {
+      "epoch": 0.37062548605710477,
+      "grad_norm": 0.49171337485313416,
+      "learning_rate": 0.0001,
+      "loss": 1.0182,
+      "step": 417
+    },
+    {
+      "epoch": 0.37151427619153427,
+      "grad_norm": 0.47732532024383545,
+      "learning_rate": 0.0001,
+      "loss": 0.9888,
+      "step": 418
+    },
+    {
+      "epoch": 0.37240306632596376,
+      "grad_norm": 0.418824702501297,
+      "learning_rate": 0.0001,
+      "loss": 1.0082,
+      "step": 419
+    },
+    {
+      "epoch": 0.3732918564603933,
+      "grad_norm": 0.4232107698917389,
+      "learning_rate": 0.0001,
+      "loss": 0.9977,
+      "step": 420
+    },
+    {
+      "epoch": 0.3741806465948228,
+      "grad_norm": 0.45239681005477905,
+      "learning_rate": 0.0001,
+      "loss": 0.9741,
+      "step": 421
+    },
+    {
+      "epoch": 0.3750694367292523,
+      "grad_norm": 0.4842644929885864,
+      "learning_rate": 0.0001,
+      "loss": 0.9677,
+      "step": 422
+    },
+    {
+      "epoch": 0.3759582268636818,
+      "grad_norm": 0.478947252035141,
+      "learning_rate": 0.0001,
+      "loss": 0.9718,
+      "step": 423
+    },
+    {
+      "epoch": 0.3768470169981113,
+      "grad_norm": 0.5387521386146545,
+      "learning_rate": 0.0001,
+      "loss": 1.0737,
+      "step": 424
+    },
+    {
+      "epoch": 0.37773580713254085,
+      "grad_norm": 0.5804430246353149,
+      "learning_rate": 0.0001,
+      "loss": 1.0273,
+      "step": 425
+    },
+    {
+      "epoch": 0.37862459726697034,
+      "grad_norm": 0.4965243935585022,
+      "learning_rate": 0.0001,
+      "loss": 1.1163,
+      "step": 426
+    },
+    {
+      "epoch": 0.37951338740139984,
+      "grad_norm": 0.5330107808113098,
+      "learning_rate": 0.0001,
+      "loss": 1.0256,
+      "step": 427
+    },
+    {
+      "epoch": 0.38040217753582933,
+      "grad_norm": 0.4444815218448639,
+      "learning_rate": 0.0001,
+      "loss": 0.9643,
+      "step": 428
+    },
+    {
+      "epoch": 0.3812909676702589,
+      "grad_norm": 0.5239233374595642,
+      "learning_rate": 0.0001,
+      "loss": 0.9893,
+      "step": 429
+    },
+    {
+      "epoch": 0.3821797578046884,
+      "grad_norm": 0.4272307753562927,
+      "learning_rate": 0.0001,
+      "loss": 0.9896,
+      "step": 430
+    },
+    {
+      "epoch": 0.3830685479391179,
+      "grad_norm": 0.422820121049881,
+      "learning_rate": 0.0001,
+      "loss": 0.9722,
+      "step": 431
+    },
+    {
+      "epoch": 0.38395733807354737,
+      "grad_norm": 0.45066556334495544,
+      "learning_rate": 0.0001,
+      "loss": 0.9266,
+      "step": 432
+    },
+    {
+      "epoch": 0.38484612820797687,
+      "grad_norm": 0.40709954500198364,
+      "learning_rate": 0.0001,
+      "loss": 0.9006,
+      "step": 433
+    },
+    {
+      "epoch": 0.3857349183424064,
+      "grad_norm": 0.45301303267478943,
+      "learning_rate": 0.0001,
+      "loss": 0.9844,
+      "step": 434
+    },
+    {
+      "epoch": 0.3866237084768359,
+      "grad_norm": 0.4441263675689697,
+      "learning_rate": 0.0001,
+      "loss": 1.0039,
+      "step": 435
+    },
+    {
+      "epoch": 0.3875124986112654,
+      "grad_norm": 0.4564574062824249,
+      "learning_rate": 0.0001,
+      "loss": 1.0397,
+      "step": 436
+    },
+    {
+      "epoch": 0.3884012887456949,
+      "grad_norm": 0.5104243755340576,
+      "learning_rate": 0.0001,
+      "loss": 1.06,
+      "step": 437
+    },
+    {
+      "epoch": 0.38929007888012446,
+      "grad_norm": 0.4639466106891632,
+      "learning_rate": 0.0001,
+      "loss": 1.028,
+      "step": 438
+    },
+    {
+      "epoch": 0.39017886901455395,
+      "grad_norm": 0.4268662929534912,
+      "learning_rate": 0.0001,
+      "loss": 0.9971,
+      "step": 439
+    },
+    {
+      "epoch": 0.39106765914898345,
+      "grad_norm": 0.4981948435306549,
+      "learning_rate": 0.0001,
+      "loss": 0.9946,
+      "step": 440
+    },
+    {
+      "epoch": 0.39195644928341294,
+      "grad_norm": 0.4488162696361542,
+      "learning_rate": 0.0001,
+      "loss": 0.9536,
+      "step": 441
+    },
+    {
+      "epoch": 0.39284523941784244,
+      "grad_norm": 0.4636482298374176,
+      "learning_rate": 0.0001,
+      "loss": 1.1277,
+      "step": 442
+    },
+    {
+      "epoch": 0.393734029552272,
+      "grad_norm": 0.46133679151535034,
+      "learning_rate": 0.0001,
+      "loss": 0.9451,
+      "step": 443
+    },
+    {
+      "epoch": 0.3946228196867015,
+      "grad_norm": 0.4021439850330353,
+      "learning_rate": 0.0001,
+      "loss": 1.0127,
+      "step": 444
+    },
+    {
+      "epoch": 0.395511609821131,
+      "grad_norm": 0.46771371364593506,
+      "learning_rate": 0.0001,
+      "loss": 1.0103,
+      "step": 445
+    },
+    {
+      "epoch": 0.3964003999555605,
+      "grad_norm": 0.5152266025543213,
+      "learning_rate": 0.0001,
+      "loss": 1.0373,
+      "step": 446
+    },
+    {
+      "epoch": 0.39728919008999,
+      "grad_norm": 0.42209741473197937,
+      "learning_rate": 0.0001,
+      "loss": 0.9229,
+      "step": 447
+    },
+    {
+      "epoch": 0.3981779802244195,
+      "grad_norm": 0.46712151169776917,
+      "learning_rate": 0.0001,
+      "loss": 0.9512,
+      "step": 448
+    },
+    {
+      "epoch": 0.399066770358849,
+      "grad_norm": 4.181483268737793,
+      "learning_rate": 0.0001,
+      "loss": 0.9367,
+      "step": 449
+    },
+    {
+      "epoch": 0.3999555604932785,
+      "grad_norm": 0.3845648467540741,
+      "learning_rate": 0.0001,
+      "loss": 0.9741,
+      "step": 450
+    },
+    {
+      "epoch": 0.400844350627708,
+      "grad_norm": 2.03074049949646,
+      "learning_rate": 0.0001,
+      "loss": 0.9775,
+      "step": 451
+    },
+    {
+      "epoch": 0.40173314076213756,
+      "grad_norm": 0.5015456080436707,
+      "learning_rate": 0.0001,
+      "loss": 1.0234,
+      "step": 452
+    },
+    {
+      "epoch": 0.40262193089656706,
+      "grad_norm": 0.4838273525238037,
+      "learning_rate": 0.0001,
+      "loss": 0.9653,
+      "step": 453
+    },
+    {
+      "epoch": 0.40351072103099656,
+      "grad_norm": 0.5604076981544495,
+      "learning_rate": 0.0001,
+      "loss": 0.9732,
+      "step": 454
+    },
+    {
+      "epoch": 0.40439951116542605,
+      "grad_norm": 0.5440881252288818,
+      "learning_rate": 0.0001,
+      "loss": 1.0301,
+      "step": 455
+    },
+    {
+      "epoch": 0.40528830129985555,
+      "grad_norm": 0.9089745879173279,
+      "learning_rate": 0.0001,
+      "loss": 1.0371,
+      "step": 456
+    },
+    {
+      "epoch": 0.4061770914342851,
+      "grad_norm": 0.4472905993461609,
+      "learning_rate": 0.0001,
+      "loss": 1.0375,
+      "step": 457
+    },
+    {
+      "epoch": 0.4070658815687146,
+      "grad_norm": 0.4706718325614929,
+      "learning_rate": 0.0001,
+      "loss": 0.9966,
+      "step": 458
+    },
+    {
+      "epoch": 0.4079546717031441,
+      "grad_norm": 0.48551270365715027,
+      "learning_rate": 0.0001,
+      "loss": 0.9577,
+      "step": 459
+    },
+    {
+      "epoch": 0.4088434618375736,
+      "grad_norm": 0.45593079924583435,
+      "learning_rate": 0.0001,
+      "loss": 0.99,
+      "step": 460
+    },
+    {
+      "epoch": 0.40973225197200314,
+      "grad_norm": 0.4240357577800751,
+      "learning_rate": 0.0001,
+      "loss": 1.0105,
+      "step": 461
+    },
+    {
+      "epoch": 0.41062104210643263,
+      "grad_norm": 0.44513946771621704,
+      "learning_rate": 0.0001,
+      "loss": 0.9697,
+      "step": 462
+    },
+    {
+      "epoch": 0.41150983224086213,
+      "grad_norm": 0.43077144026756287,
+      "learning_rate": 0.0001,
+      "loss": 1.0168,
+      "step": 463
+    },
+    {
+      "epoch": 0.4123986223752916,
+      "grad_norm": 0.47029492259025574,
+      "learning_rate": 0.0001,
+      "loss": 0.94,
+      "step": 464
+    },
+    {
+      "epoch": 0.4132874125097211,
+      "grad_norm": 0.44067198038101196,
+      "learning_rate": 0.0001,
+      "loss": 0.9443,
+      "step": 465
+    },
+    {
+      "epoch": 0.41417620264415067,
+      "grad_norm": 0.6291742324829102,
+      "learning_rate": 0.0001,
+      "loss": 0.9927,
+      "step": 466
+    },
+    {
+      "epoch": 0.41506499277858017,
+      "grad_norm": 0.49682438373565674,
+      "learning_rate": 0.0001,
+      "loss": 0.9411,
+      "step": 467
+    },
+    {
+      "epoch": 0.41595378291300966,
+      "grad_norm": 0.4587923586368561,
+      "learning_rate": 0.0001,
+      "loss": 1.0423,
+      "step": 468
+    },
+    {
+      "epoch": 0.41684257304743916,
+      "grad_norm": 0.5601244568824768,
+      "learning_rate": 0.0001,
+      "loss": 1.0682,
+      "step": 469
+    },
+    {
+      "epoch": 0.4177313631818687,
+      "grad_norm": 0.4534125328063965,
+      "learning_rate": 0.0001,
+      "loss": 1.0591,
+      "step": 470
+    },
+    {
+      "epoch": 0.4186201533162982,
+      "grad_norm": 0.469804972410202,
+      "learning_rate": 0.0001,
+      "loss": 1.0893,
+      "step": 471
+    },
+    {
+      "epoch": 0.4195089434507277,
+      "grad_norm": 0.4769747257232666,
+      "learning_rate": 0.0001,
+      "loss": 0.9974,
+      "step": 472
+    },
+    {
+      "epoch": 0.4203977335851572,
+      "grad_norm": 0.4488048553466797,
+      "learning_rate": 0.0001,
+      "loss": 0.9681,
+      "step": 473
+    },
+    {
+      "epoch": 0.4212865237195867,
+      "grad_norm": 0.4423130750656128,
+      "learning_rate": 0.0001,
+      "loss": 1.0732,
+      "step": 474
+    },
+    {
+      "epoch": 0.42217531385401624,
+      "grad_norm": 0.40248462557792664,
+      "learning_rate": 0.0001,
+      "loss": 0.9631,
+      "step": 475
+    },
+    {
+      "epoch": 0.42306410398844574,
+      "grad_norm": 0.4668256640434265,
+      "learning_rate": 0.0001,
+      "loss": 0.9771,
+      "step": 476
+    },
+    {
+      "epoch": 0.42395289412287523,
+      "grad_norm": 0.46433717012405396,
+      "learning_rate": 0.0001,
+      "loss": 1.02,
+      "step": 477
+    },
+    {
+      "epoch": 0.42484168425730473,
+      "grad_norm": 0.8196445107460022,
+      "learning_rate": 0.0001,
+      "loss": 0.9596,
+      "step": 478
+    },
+    {
+      "epoch": 0.4257304743917342,
+      "grad_norm": 0.5228718519210815,
+      "learning_rate": 0.0001,
+      "loss": 1.0023,
+      "step": 479
+    },
+    {
+      "epoch": 0.4266192645261638,
+      "grad_norm": 0.4933236241340637,
+      "learning_rate": 0.0001,
+      "loss": 0.936,
+      "step": 480
+    },
+    {
+      "epoch": 0.4275080546605933,
+      "grad_norm": 0.4738243818283081,
+      "learning_rate": 0.0001,
+      "loss": 0.9954,
+      "step": 481
+    },
+    {
+      "epoch": 0.42839684479502277,
+      "grad_norm": 0.4303346872329712,
+      "learning_rate": 0.0001,
+      "loss": 1.0311,
+      "step": 482
+    },
+    {
+      "epoch": 0.42928563492945226,
+      "grad_norm": 0.3946980834007263,
+      "learning_rate": 0.0001,
+      "loss": 0.9655,
+      "step": 483
+    },
+    {
+      "epoch": 0.4301744250638818,
+      "grad_norm": 0.4781205952167511,
+      "learning_rate": 0.0001,
+      "loss": 1.0161,
+      "step": 484
+    },
+    {
+      "epoch": 0.4310632151983113,
+      "grad_norm": 0.41279280185699463,
+      "learning_rate": 0.0001,
+      "loss": 0.984,
+      "step": 485
+    },
+    {
+      "epoch": 0.4319520053327408,
+      "grad_norm": 0.42448437213897705,
+      "learning_rate": 0.0001,
+      "loss": 1.0044,
+      "step": 486
+    },
+    {
+      "epoch": 0.4328407954671703,
+      "grad_norm": 0.42571067810058594,
+      "learning_rate": 0.0001,
+      "loss": 0.9963,
+      "step": 487
+    },
+    {
+      "epoch": 0.4337295856015998,
+      "grad_norm": 0.43724721670150757,
+      "learning_rate": 0.0001,
+      "loss": 0.9868,
+      "step": 488
+    },
+    {
+      "epoch": 0.43461837573602935,
+      "grad_norm": 0.506829023361206,
+      "learning_rate": 0.0001,
+      "loss": 1.045,
+      "step": 489
+    },
+    {
+      "epoch": 0.43550716587045885,
+      "grad_norm": 0.3961758315563202,
+      "learning_rate": 0.0001,
+      "loss": 0.9235,
+      "step": 490
+    },
+    {
+      "epoch": 0.43639595600488834,
+      "grad_norm": 0.41079434752464294,
+      "learning_rate": 0.0001,
+      "loss": 0.9944,
+      "step": 491
+    },
+    {
+      "epoch": 0.43728474613931784,
+      "grad_norm": 0.4370504915714264,
+      "learning_rate": 0.0001,
+      "loss": 0.9925,
+      "step": 492
+    },
+    {
+      "epoch": 0.4381735362737474,
+      "grad_norm": 0.4316272735595703,
+      "learning_rate": 0.0001,
+      "loss": 1.0229,
+      "step": 493
+    },
+    {
+      "epoch": 0.4390623264081769,
+      "grad_norm": 0.4933658540248871,
+      "learning_rate": 0.0001,
+      "loss": 1.0974,
+      "step": 494
+    },
+    {
+      "epoch": 0.4399511165426064,
+      "grad_norm": 0.42228391766548157,
+      "learning_rate": 0.0001,
+      "loss": 1.0277,
+      "step": 495
+    },
+    {
+      "epoch": 0.4408399066770359,
+      "grad_norm": 0.45421552658081055,
+      "learning_rate": 0.0001,
+      "loss": 0.8842,
+      "step": 496
+    },
+    {
+      "epoch": 0.44172869681146537,
+      "grad_norm": 0.4748377501964569,
+      "learning_rate": 0.0001,
+      "loss": 0.9488,
+      "step": 497
+    },
+    {
+      "epoch": 0.4426174869458949,
+      "grad_norm": 0.4040847718715668,
+      "learning_rate": 0.0001,
+      "loss": 0.9696,
+      "step": 498
+    },
+    {
+      "epoch": 0.4435062770803244,
+      "grad_norm": 0.42285656929016113,
+      "learning_rate": 0.0001,
+      "loss": 0.9637,
+      "step": 499
+    },
+    {
+      "epoch": 0.4443950672147539,
+      "grad_norm": 0.3885044455528259,
+      "learning_rate": 0.0001,
+      "loss": 0.985,
+      "step": 500
+    },
+    {
+      "epoch": 0.4452838573491834,
+      "grad_norm": 0.45078715682029724,
+      "learning_rate": 0.0001,
+      "loss": 0.9538,
+      "step": 501
+    },
+    {
+      "epoch": 0.4461726474836129,
+      "grad_norm": 0.4214499294757843,
+      "learning_rate": 0.0001,
+      "loss": 0.8488,
+      "step": 502
+    },
+    {
+      "epoch": 0.44706143761804246,
+      "grad_norm": 0.6104442477226257,
+      "learning_rate": 0.0001,
+      "loss": 1.0681,
+      "step": 503
+    },
+    {
+      "epoch": 0.44795022775247195,
+      "grad_norm": 0.49673840403556824,
+      "learning_rate": 0.0001,
+      "loss": 0.9875,
+      "step": 504
+    },
+    {
+      "epoch": 0.44883901788690145,
+      "grad_norm": 0.4738848805427551,
+      "learning_rate": 0.0001,
+      "loss": 0.9738,
+      "step": 505
+    },
+    {
+      "epoch": 0.44972780802133094,
+      "grad_norm": 0.4084794223308563,
+      "learning_rate": 0.0001,
+      "loss": 0.9617,
+      "step": 506
+    },
+    {
+      "epoch": 0.4506165981557605,
+      "grad_norm": 0.43893682956695557,
+      "learning_rate": 0.0001,
+      "loss": 0.9553,
+      "step": 507
+    },
+    {
+      "epoch": 0.45150538829019,
+      "grad_norm": 0.4741009771823883,
+      "learning_rate": 0.0001,
+      "loss": 0.9863,
+      "step": 508
+    },
+    {
+      "epoch": 0.4523941784246195,
+      "grad_norm": 0.42300963401794434,
+      "learning_rate": 0.0001,
+      "loss": 0.9669,
+      "step": 509
+    },
+    {
+      "epoch": 0.453282968559049,
+      "grad_norm": 0.3663196265697479,
+      "learning_rate": 0.0001,
+      "loss": 0.9606,
+      "step": 510
+    },
+    {
+      "epoch": 0.4541717586934785,
+      "grad_norm": 0.4289178252220154,
+      "learning_rate": 0.0001,
+      "loss": 0.9093,
+      "step": 511
+    },
+    {
+      "epoch": 0.45506054882790803,
+      "grad_norm": 0.41236793994903564,
+      "learning_rate": 0.0001,
+      "loss": 0.9582,
+      "step": 512
+    },
+    {
+      "epoch": 0.4559493389623375,
+      "grad_norm": 0.38569167256355286,
+      "learning_rate": 0.0001,
+      "loss": 0.9804,
+      "step": 513
+    },
+    {
+      "epoch": 0.456838129096767,
+      "grad_norm": 0.42629268765449524,
+      "learning_rate": 0.0001,
+      "loss": 0.9111,
+      "step": 514
+    },
+    {
+      "epoch": 0.4577269192311965,
+      "grad_norm": 0.4302125573158264,
+      "learning_rate": 0.0001,
+      "loss": 0.9435,
+      "step": 515
+    },
+    {
+      "epoch": 0.45861570936562607,
+      "grad_norm": 0.46809178590774536,
+      "learning_rate": 0.0001,
+      "loss": 1.0059,
+      "step": 516
+    },
+    {
+      "epoch": 0.45950449950005556,
+      "grad_norm": 0.4712200164794922,
+      "learning_rate": 0.0001,
+      "loss": 0.9681,
+      "step": 517
+    },
+    {
+      "epoch": 0.46039328963448506,
+      "grad_norm": 0.4966319501399994,
+      "learning_rate": 0.0001,
+      "loss": 1.0278,
+      "step": 518
+    },
+    {
+      "epoch": 0.46128207976891455,
+      "grad_norm": 0.41810038685798645,
+      "learning_rate": 0.0001,
+      "loss": 1.0128,
+      "step": 519
+    },
+    {
+      "epoch": 0.46217086990334405,
+      "grad_norm": 0.4766371250152588,
+      "learning_rate": 0.0001,
+      "loss": 0.9306,
+      "step": 520
+    },
+    {
+      "epoch": 0.4630596600377736,
+      "grad_norm": 0.47531601786613464,
+      "learning_rate": 0.0001,
+      "loss": 1.0213,
+      "step": 521
+    },
+    {
+      "epoch": 0.4639484501722031,
+      "grad_norm": 0.4246899485588074,
+      "learning_rate": 0.0001,
+      "loss": 0.9415,
+      "step": 522
+    },
+    {
+      "epoch": 0.4648372403066326,
+      "grad_norm": 0.38327568769454956,
+      "learning_rate": 0.0001,
+      "loss": 0.976,
+      "step": 523
+    },
+    {
+      "epoch": 0.4657260304410621,
+      "grad_norm": 0.45601171255111694,
+      "learning_rate": 0.0001,
+      "loss": 0.9481,
+      "step": 524
+    },
+    {
+      "epoch": 0.4666148205754916,
+      "grad_norm": 0.4424237608909607,
+      "learning_rate": 0.0001,
+      "loss": 0.974,
+      "step": 525
+    },
+    {
+      "epoch": 0.46750361070992114,
+      "grad_norm": 0.45187127590179443,
+      "learning_rate": 0.0001,
+      "loss": 1.1636,
+      "step": 526
+    },
+    {
+      "epoch": 0.46839240084435063,
+      "grad_norm": 0.44865912199020386,
+      "learning_rate": 0.0001,
+      "loss": 0.9951,
+      "step": 527
+    },
+    {
+      "epoch": 0.4692811909787801,
+      "grad_norm": 0.40876081585884094,
+      "learning_rate": 0.0001,
+      "loss": 0.9765,
+      "step": 528
+    },
+    {
+      "epoch": 0.4701699811132096,
+      "grad_norm": 0.3936661183834076,
+      "learning_rate": 0.0001,
+      "loss": 0.9935,
+      "step": 529
+    },
+    {
+      "epoch": 0.4710587712476392,
+      "grad_norm": 0.422152578830719,
+      "learning_rate": 0.0001,
+      "loss": 0.9956,
+      "step": 530
+    },
+    {
+      "epoch": 0.47194756138206867,
+      "grad_norm": 0.40520817041397095,
+      "learning_rate": 0.0001,
+      "loss": 0.9609,
+      "step": 531
+    },
+    {
+      "epoch": 0.47283635151649817,
+      "grad_norm": 0.43614640831947327,
+      "learning_rate": 0.0001,
+      "loss": 1.0283,
+      "step": 532
+    },
+    {
+      "epoch": 0.47372514165092766,
+      "grad_norm": 0.4783385694026947,
+      "learning_rate": 0.0001,
+      "loss": 1.0388,
+      "step": 533
+    },
+    {
+      "epoch": 0.47461393178535716,
+      "grad_norm": 0.4622490704059601,
+      "learning_rate": 0.0001,
+      "loss": 1.088,
+      "step": 534
+    },
+    {
+      "epoch": 0.4755027219197867,
+      "grad_norm": 0.43488460779190063,
+      "learning_rate": 0.0001,
+      "loss": 0.9386,
+      "step": 535
+    },
+    {
+      "epoch": 0.4763915120542162,
+      "grad_norm": 0.40164855122566223,
+      "learning_rate": 0.0001,
+      "loss": 0.9476,
+      "step": 536
+    },
+    {
+      "epoch": 0.4772803021886457,
+      "grad_norm": 0.35324132442474365,
+      "learning_rate": 0.0001,
+      "loss": 0.9957,
+      "step": 537
+    },
+    {
+      "epoch": 0.4781690923230752,
+      "grad_norm": 0.39218297600746155,
+      "learning_rate": 0.0001,
+      "loss": 0.9682,
+      "step": 538
+    },
+    {
+      "epoch": 0.47905788245750475,
+      "grad_norm": 0.4563474953174591,
+      "learning_rate": 0.0001,
+      "loss": 0.9479,
+      "step": 539
+    },
+    {
+      "epoch": 0.47994667259193424,
+      "grad_norm": 0.4176347255706787,
+      "learning_rate": 0.0001,
+      "loss": 0.9101,
+      "step": 540
+    },
+    {
+      "epoch": 0.48083546272636374,
+      "grad_norm": 0.38945290446281433,
+      "learning_rate": 0.0001,
+      "loss": 0.9083,
+      "step": 541
+    },
+    {
+      "epoch": 0.48172425286079323,
+      "grad_norm": 0.35722264647483826,
+      "learning_rate": 0.0001,
+      "loss": 0.8814,
+      "step": 542
+    },
+    {
+      "epoch": 0.48261304299522273,
+      "grad_norm": 0.43666157126426697,
+      "learning_rate": 0.0001,
+      "loss": 0.9738,
+      "step": 543
+    },
+    {
+      "epoch": 0.4835018331296523,
+      "grad_norm": 0.462503045797348,
+      "learning_rate": 0.0001,
+      "loss": 0.9315,
+      "step": 544
+    },
+    {
+      "epoch": 0.4843906232640818,
+      "grad_norm": 0.48999452590942383,
+      "learning_rate": 0.0001,
+      "loss": 1.024,
+      "step": 545
+    },
+    {
+      "epoch": 0.4852794133985113,
+      "grad_norm": 0.5173038840293884,
+      "learning_rate": 0.0001,
+      "loss": 0.885,
+      "step": 546
+    },
+    {
+      "epoch": 0.48616820353294077,
+      "grad_norm": 0.442202627658844,
+      "learning_rate": 0.0001,
+      "loss": 0.9667,
+      "step": 547
+    },
+    {
+      "epoch": 0.4870569936673703,
+      "grad_norm": 0.5005183219909668,
+      "learning_rate": 0.0001,
+      "loss": 1.0566,
+      "step": 548
+    },
+    {
+      "epoch": 0.4879457838017998,
+      "grad_norm": 0.3552423417568207,
+      "learning_rate": 0.0001,
+      "loss": 0.9279,
+      "step": 549
+    },
+    {
+      "epoch": 0.4888345739362293,
+      "grad_norm": 0.46777617931365967,
+      "learning_rate": 0.0001,
+      "loss": 0.9534,
+      "step": 550
+    },
+    {
+      "epoch": 0.4897233640706588,
+      "grad_norm": 0.43491941690444946,
+      "learning_rate": 0.0001,
+      "loss": 1.069,
+      "step": 551
+    },
+    {
+      "epoch": 0.4906121542050883,
+      "grad_norm": 0.4217972755432129,
+      "learning_rate": 0.0001,
+      "loss": 1.0236,
+      "step": 552
+    },
+    {
+      "epoch": 0.49150094433951785,
+      "grad_norm": 0.43809306621551514,
+      "learning_rate": 0.0001,
+      "loss": 0.9967,
+      "step": 553
+    },
+    {
+      "epoch": 0.49238973447394735,
+      "grad_norm": 0.41759732365608215,
+      "learning_rate": 0.0001,
+      "loss": 0.9486,
+      "step": 554
+    },
+    {
+      "epoch": 0.49327852460837684,
+      "grad_norm": 0.4331734776496887,
+      "learning_rate": 0.0001,
+      "loss": 0.9531,
+      "step": 555
+    },
+    {
+      "epoch": 0.49416731474280634,
+      "grad_norm": 0.4199782907962799,
+      "learning_rate": 0.0001,
+      "loss": 0.9904,
+      "step": 556
+    },
+    {
+      "epoch": 0.49505610487723584,
+      "grad_norm": 0.4305680990219116,
+      "learning_rate": 0.0001,
+      "loss": 0.909,
+      "step": 557
+    },
+    {
+      "epoch": 0.4959448950116654,
+      "grad_norm": 0.5136706233024597,
+      "learning_rate": 0.0001,
+      "loss": 1.121,
+      "step": 558
+    },
+    {
+      "epoch": 0.4968336851460949,
+      "grad_norm": 0.429557204246521,
+      "learning_rate": 0.0001,
+      "loss": 0.9809,
+      "step": 559
+    },
+    {
+      "epoch": 0.4977224752805244,
+      "grad_norm": 0.41944217681884766,
+      "learning_rate": 0.0001,
+      "loss": 0.9802,
+      "step": 560
+    },
+    {
+      "epoch": 0.4986112654149539,
+      "grad_norm": 0.43246370553970337,
+      "learning_rate": 0.0001,
+      "loss": 0.9973,
+      "step": 561
+    },
+    {
+      "epoch": 0.4995000555493834,
+      "grad_norm": 0.3798202872276306,
+      "learning_rate": 0.0001,
+      "loss": 0.9367,
+      "step": 562
+    },
+    {
+      "epoch": 0.5003888456838129,
+      "grad_norm": 0.43042704463005066,
+      "learning_rate": 0.0001,
+      "loss": 1.075,
+      "step": 563
+    },
+    {
+      "epoch": 0.5012776358182425,
+      "grad_norm": 0.3733251392841339,
+      "learning_rate": 0.0001,
+      "loss": 0.9731,
+      "step": 564
+    },
+    {
+      "epoch": 0.5021664259526719,
+      "grad_norm": 0.423252135515213,
+      "learning_rate": 0.0001,
+      "loss": 0.9788,
+      "step": 565
+    },
+    {
+      "epoch": 0.5030552160871015,
+      "grad_norm": 0.43651118874549866,
+      "learning_rate": 0.0001,
+      "loss": 0.9289,
+      "step": 566
+    },
+    {
+      "epoch": 0.5039440062215309,
+      "grad_norm": 0.44451677799224854,
+      "learning_rate": 0.0001,
+      "loss": 0.938,
+      "step": 567
+    },
+    {
+      "epoch": 0.5048327963559605,
+      "grad_norm": 0.4594007730484009,
+      "learning_rate": 0.0001,
+      "loss": 0.9562,
+      "step": 568
+    },
+    {
+      "epoch": 0.50572158649039,
+      "grad_norm": 0.5364981889724731,
+      "learning_rate": 0.0001,
+      "loss": 0.9239,
+      "step": 569
+    },
+    {
+      "epoch": 0.5066103766248194,
+      "grad_norm": 0.45558059215545654,
+      "learning_rate": 0.0001,
+      "loss": 1.0017,
+      "step": 570
+    },
+    {
+      "epoch": 0.507499166759249,
+      "grad_norm": 0.42612048983573914,
+      "learning_rate": 0.0001,
+      "loss": 0.9401,
+      "step": 571
+    },
+    {
+      "epoch": 0.5083879568936784,
+      "grad_norm": 0.4538538157939911,
+      "learning_rate": 0.0001,
+      "loss": 0.9979,
+      "step": 572
+    },
+    {
+      "epoch": 0.509276747028108,
+      "grad_norm": 1.2720234394073486,
+      "learning_rate": 0.0001,
+      "loss": 1.0537,
+      "step": 573
+    },
+    {
+      "epoch": 0.5101655371625375,
+      "grad_norm": 0.4332892894744873,
+      "learning_rate": 0.0001,
+      "loss": 0.9529,
+      "step": 574
+    },
+    {
+      "epoch": 0.511054327296967,
+      "grad_norm": 0.42090901732444763,
+      "learning_rate": 0.0001,
+      "loss": 0.8948,
+      "step": 575
+    },
+    {
+      "epoch": 0.5119431174313965,
+      "grad_norm": 0.4180974066257477,
+      "learning_rate": 0.0001,
+      "loss": 0.9472,
+      "step": 576
+    },
+    {
+      "epoch": 0.512831907565826,
+      "grad_norm": 0.44674041867256165,
+      "learning_rate": 0.0001,
+      "loss": 1.0032,
+      "step": 577
+    },
+    {
+      "epoch": 0.5137206977002555,
+      "grad_norm": 0.38132181763648987,
+      "learning_rate": 0.0001,
+      "loss": 0.9722,
+      "step": 578
+    },
+    {
+      "epoch": 0.5146094878346851,
+      "grad_norm": 0.425112247467041,
+      "learning_rate": 0.0001,
+      "loss": 1.0104,
+      "step": 579
+    },
+    {
+      "epoch": 0.5154982779691145,
+      "grad_norm": 0.4420212507247925,
+      "learning_rate": 0.0001,
+      "loss": 0.8951,
+      "step": 580
+    },
+    {
+      "epoch": 0.5163870681035441,
+      "grad_norm": 0.4047195613384247,
+      "learning_rate": 0.0001,
+      "loss": 0.9103,
+      "step": 581
+    },
+    {
+      "epoch": 0.5172758582379735,
+      "grad_norm": 0.4206266701221466,
+      "learning_rate": 0.0001,
+      "loss": 0.9872,
+      "step": 582
+    },
+    {
+      "epoch": 0.5181646483724031,
+      "grad_norm": 0.46447429060935974,
+      "learning_rate": 0.0001,
+      "loss": 0.9636,
+      "step": 583
+    },
+    {
+      "epoch": 0.5190534385068326,
+      "grad_norm": 0.467122882604599,
+      "learning_rate": 0.0001,
+      "loss": 0.9556,
+      "step": 584
+    },
+    {
+      "epoch": 0.519942228641262,
+      "grad_norm": 0.4438915252685547,
+      "learning_rate": 0.0001,
+      "loss": 0.9775,
+      "step": 585
+    },
+    {
+      "epoch": 0.5208310187756916,
+      "grad_norm": 0.4225422441959381,
+      "learning_rate": 0.0001,
+      "loss": 0.9315,
+      "step": 586
+    },
+    {
+      "epoch": 0.5217198089101212,
+      "grad_norm": 0.3874359726905823,
+      "learning_rate": 0.0001,
+      "loss": 0.9652,
+      "step": 587
+    },
+    {
+      "epoch": 0.5226085990445506,
+      "grad_norm": 0.4789721369743347,
+      "learning_rate": 0.0001,
+      "loss": 0.9336,
+      "step": 588
+    },
+    {
+      "epoch": 0.5234973891789801,
+      "grad_norm": 0.4012058675289154,
+      "learning_rate": 0.0001,
+      "loss": 0.9792,
+      "step": 589
+    },
+    {
+      "epoch": 0.5243861793134096,
+      "grad_norm": 0.4494520425796509,
+      "learning_rate": 0.0001,
+      "loss": 0.9143,
+      "step": 590
+    },
+    {
+      "epoch": 0.5252749694478391,
+      "grad_norm": 0.44614073634147644,
+      "learning_rate": 0.0001,
+      "loss": 0.9286,
+      "step": 591
+    },
+    {
+      "epoch": 0.5261637595822687,
+      "grad_norm": 0.4147653579711914,
+      "learning_rate": 0.0001,
+      "loss": 0.8982,
+      "step": 592
+    },
+    {
+      "epoch": 0.5270525497166981,
+      "grad_norm": 0.4356852173805237,
+      "learning_rate": 0.0001,
+      "loss": 0.9263,
+      "step": 593
+    },
+    {
+      "epoch": 0.5279413398511277,
+      "grad_norm": 0.4420524537563324,
+      "learning_rate": 0.0001,
+      "loss": 0.969,
+      "step": 594
+    },
+    {
+      "epoch": 0.5288301299855571,
+      "grad_norm": 0.4878003001213074,
+      "learning_rate": 0.0001,
+      "loss": 1.0299,
+      "step": 595
+    },
+    {
+      "epoch": 0.5297189201199867,
+      "grad_norm": 0.4248897433280945,
+      "learning_rate": 0.0001,
+      "loss": 1.0166,
+      "step": 596
+    },
+    {
+      "epoch": 0.5306077102544162,
+      "grad_norm": 0.3686140179634094,
+      "learning_rate": 0.0001,
+      "loss": 0.9452,
+      "step": 597
+    },
+    {
+      "epoch": 0.5314965003888457,
+      "grad_norm": 0.3767171800136566,
+      "learning_rate": 0.0001,
+      "loss": 1.0059,
+      "step": 598
+    },
+    {
+      "epoch": 0.5323852905232752,
+      "grad_norm": 0.41753292083740234,
+      "learning_rate": 0.0001,
+      "loss": 0.9117,
+      "step": 599
+    },
+    {
+      "epoch": 0.5332740806577047,
+      "grad_norm": 0.45836883783340454,
+      "learning_rate": 0.0001,
+      "loss": 1.0302,
+      "step": 600
+    },
+    {
+      "epoch": 0.5341628707921342,
+      "grad_norm": 0.40416955947875977,
+      "learning_rate": 0.0001,
+      "loss": 0.9387,
+      "step": 601
+    },
+    {
+      "epoch": 0.5350516609265638,
+      "grad_norm": 0.4028719365596771,
+      "learning_rate": 0.0001,
+      "loss": 1.0046,
+      "step": 602
+    },
+    {
+      "epoch": 0.5359404510609932,
+      "grad_norm": 0.42004886269569397,
+      "learning_rate": 0.0001,
+      "loss": 1.0475,
+      "step": 603
+    },
+    {
+      "epoch": 0.5368292411954227,
+      "grad_norm": 0.44581007957458496,
+      "learning_rate": 0.0001,
+      "loss": 0.8798,
+      "step": 604
+    },
+    {
+      "epoch": 0.5377180313298522,
+      "grad_norm": 0.4007730782032013,
+      "learning_rate": 0.0001,
+      "loss": 0.9578,
+      "step": 605
+    },
+    {
+      "epoch": 0.5386068214642817,
+      "grad_norm": 0.45472684502601624,
+      "learning_rate": 0.0001,
+      "loss": 0.9609,
+      "step": 606
+    },
+    {
+      "epoch": 0.5394956115987113,
+      "grad_norm": 0.45508822798728943,
+      "learning_rate": 0.0001,
+      "loss": 0.9699,
+      "step": 607
+    },
+    {
+      "epoch": 0.5403844017331407,
+      "grad_norm": 0.40522894263267517,
+      "learning_rate": 0.0001,
+      "loss": 0.9791,
+      "step": 608
+    },
+    {
+      "epoch": 0.5412731918675703,
+      "grad_norm": 0.42480531334877014,
+      "learning_rate": 0.0001,
+      "loss": 1.041,
+      "step": 609
+    },
+    {
+      "epoch": 0.5421619820019998,
+      "grad_norm": 0.3351707458496094,
+      "learning_rate": 0.0001,
+      "loss": 1.0673,
+      "step": 610
+    },
+    {
+      "epoch": 0.5430507721364293,
+      "grad_norm": 0.5073234438896179,
+      "learning_rate": 0.0001,
+      "loss": 0.9928,
+      "step": 611
+    },
+    {
+      "epoch": 0.5439395622708588,
+      "grad_norm": 0.4208507239818573,
+      "learning_rate": 0.0001,
+      "loss": 0.9584,
+      "step": 612
+    },
+    {
+      "epoch": 0.5448283524052883,
+      "grad_norm": 0.4468097686767578,
+      "learning_rate": 0.0001,
+      "loss": 0.9225,
+      "step": 613
+    },
+    {
+      "epoch": 0.5457171425397178,
+      "grad_norm": 0.44044622778892517,
+      "learning_rate": 0.0001,
+      "loss": 0.9906,
+      "step": 614
+    },
+    {
+      "epoch": 0.5466059326741474,
+      "grad_norm": 0.5258206725120544,
+      "learning_rate": 0.0001,
+      "loss": 0.9886,
+      "step": 615
+    },
+    {
+      "epoch": 0.5474947228085768,
+      "grad_norm": 0.4392997622489929,
+      "learning_rate": 0.0001,
+      "loss": 1.0121,
+      "step": 616
+    },
+    {
+      "epoch": 0.5483835129430064,
+      "grad_norm": 0.4318770170211792,
+      "learning_rate": 0.0001,
+      "loss": 1.0073,
+      "step": 617
+    },
+    {
+      "epoch": 0.5492723030774358,
+      "grad_norm": 0.4536917805671692,
+      "learning_rate": 0.0001,
+      "loss": 0.9971,
+      "step": 618
+    },
+    {
+      "epoch": 0.5501610932118653,
+      "grad_norm": 0.39293742179870605,
+      "learning_rate": 0.0001,
+      "loss": 0.9785,
+      "step": 619
+    },
+    {
+      "epoch": 0.5510498833462949,
+      "grad_norm": 0.4010220766067505,
+      "learning_rate": 0.0001,
+      "loss": 0.9656,
+      "step": 620
+    },
+    {
+      "epoch": 0.5519386734807243,
+      "grad_norm": 0.43340232968330383,
+      "learning_rate": 0.0001,
+      "loss": 0.8983,
+      "step": 621
+    },
+    {
+      "epoch": 0.5528274636151539,
+      "grad_norm": 0.3749978840351105,
+      "learning_rate": 0.0001,
+      "loss": 0.8759,
+      "step": 622
+    },
+    {
+      "epoch": 0.5537162537495833,
+      "grad_norm": 0.47160354256629944,
+      "learning_rate": 0.0001,
+      "loss": 1.0252,
+      "step": 623
+    },
+    {
+      "epoch": 0.5546050438840129,
+      "grad_norm": 0.3324509859085083,
+      "learning_rate": 0.0001,
+      "loss": 0.9065,
+      "step": 624
+    },
+    {
+      "epoch": 0.5554938340184424,
+      "grad_norm": 0.3825232982635498,
+      "learning_rate": 0.0001,
+      "loss": 0.9536,
+      "step": 625
+    },
+    {
+      "epoch": 0.5563826241528719,
+      "grad_norm": 0.4211728870868683,
+      "learning_rate": 0.0001,
+      "loss": 0.8948,
+      "step": 626
+    },
+    {
+      "epoch": 0.5572714142873014,
+      "grad_norm": 0.40888741612434387,
+      "learning_rate": 0.0001,
+      "loss": 1.0129,
+      "step": 627
+    },
+    {
+      "epoch": 0.5581602044217309,
+      "grad_norm": 0.3999336063861847,
+      "learning_rate": 0.0001,
+      "loss": 0.9486,
+      "step": 628
+    },
+    {
+      "epoch": 0.5590489945561604,
+      "grad_norm": 0.3953676223754883,
+      "learning_rate": 0.0001,
+      "loss": 0.8835,
+      "step": 629
+    },
+    {
+      "epoch": 0.55993778469059,
+      "grad_norm": 3.312321901321411,
+      "learning_rate": 0.0001,
+      "loss": 1.0447,
+      "step": 630
+    },
+    {
+      "epoch": 0.5608265748250194,
+      "grad_norm": 0.6516892313957214,
+      "learning_rate": 0.0001,
+      "loss": 0.8964,
+      "step": 631
+    },
+    {
+      "epoch": 0.561715364959449,
+      "grad_norm": 0.43332597613334656,
+      "learning_rate": 0.0001,
+      "loss": 0.9386,
+      "step": 632
+    },
+    {
+      "epoch": 0.5626041550938785,
+      "grad_norm": 0.36072951555252075,
+      "learning_rate": 0.0001,
+      "loss": 0.9489,
+      "step": 633
+    },
+    {
+      "epoch": 0.563492945228308,
+      "grad_norm": 0.4039455056190491,
+      "learning_rate": 0.0001,
+      "loss": 0.9512,
+      "step": 634
+    },
+    {
+      "epoch": 0.5643817353627375,
+      "grad_norm": 0.40944796800613403,
+      "learning_rate": 0.0001,
+      "loss": 1.0909,
+      "step": 635
+    },
+    {
+      "epoch": 0.5652705254971669,
+      "grad_norm": 0.4490579068660736,
+      "learning_rate": 0.0001,
+      "loss": 0.955,
+      "step": 636
+    },
+    {
+      "epoch": 0.5661593156315965,
+      "grad_norm": 0.4226183593273163,
+      "learning_rate": 0.0001,
+      "loss": 0.9833,
+      "step": 637
+    },
+    {
+      "epoch": 0.567048105766026,
+      "grad_norm": 0.4099627733230591,
+      "learning_rate": 0.0001,
+      "loss": 0.9909,
+      "step": 638
+    },
+    {
+      "epoch": 0.5679368959004555,
+      "grad_norm": 0.39125704765319824,
+      "learning_rate": 0.0001,
+      "loss": 0.9875,
+      "step": 639
+    },
+    {
+      "epoch": 0.568825686034885,
+      "grad_norm": 0.3506604731082916,
+      "learning_rate": 0.0001,
+      "loss": 0.8747,
+      "step": 640
+    },
+    {
+      "epoch": 0.5697144761693145,
+      "grad_norm": 0.6626682281494141,
+      "learning_rate": 0.0001,
+      "loss": 1.0328,
+      "step": 641
+    },
+    {
+      "epoch": 0.570603266303744,
+      "grad_norm": 0.3994938135147095,
+      "learning_rate": 0.0001,
+      "loss": 0.9255,
+      "step": 642
+    },
+    {
+      "epoch": 0.5714920564381736,
+      "grad_norm": 0.4206639528274536,
+      "learning_rate": 0.0001,
+      "loss": 0.9222,
+      "step": 643
+    },
+    {
+      "epoch": 0.572380846572603,
+      "grad_norm": 0.4152818024158478,
+      "learning_rate": 0.0001,
+      "loss": 0.9515,
+      "step": 644
+    },
+    {
+      "epoch": 0.5732696367070326,
+      "grad_norm": 0.3834103047847748,
+      "learning_rate": 0.0001,
+      "loss": 0.9008,
+      "step": 645
+    },
+    {
+      "epoch": 0.574158426841462,
+      "grad_norm": 0.3906621038913727,
+      "learning_rate": 0.0001,
+      "loss": 1.0196,
+      "step": 646
+    },
+    {
+      "epoch": 0.5750472169758916,
+      "grad_norm": 0.4065912067890167,
+      "learning_rate": 0.0001,
+      "loss": 0.9741,
+      "step": 647
+    },
+    {
+      "epoch": 0.5759360071103211,
+      "grad_norm": 0.3870736360549927,
+      "learning_rate": 0.0001,
+      "loss": 0.9403,
+      "step": 648
+    },
+    {
+      "epoch": 0.5768247972447506,
+      "grad_norm": 0.40144529938697815,
+      "learning_rate": 0.0001,
+      "loss": 0.9931,
+      "step": 649
+    },
+    {
+      "epoch": 0.5777135873791801,
+      "grad_norm": 0.40022167563438416,
+      "learning_rate": 0.0001,
+      "loss": 0.9741,
+      "step": 650
+    },
+    {
+      "epoch": 0.5786023775136095,
+      "grad_norm": 0.4081610143184662,
+      "learning_rate": 0.0001,
+      "loss": 0.9509,
+      "step": 651
+    },
+    {
+      "epoch": 0.5794911676480391,
+      "grad_norm": 0.3786165714263916,
+      "learning_rate": 0.0001,
+      "loss": 0.9435,
+      "step": 652
+    },
+    {
+      "epoch": 0.5803799577824686,
+      "grad_norm": 0.3807113468647003,
+      "learning_rate": 0.0001,
+      "loss": 0.9534,
+      "step": 653
+    },
+    {
+      "epoch": 0.5812687479168981,
+      "grad_norm": 0.4126400053501129,
+      "learning_rate": 0.0001,
+      "loss": 0.8815,
+      "step": 654
+    },
+    {
+      "epoch": 0.5821575380513276,
+      "grad_norm": 0.46695202589035034,
+      "learning_rate": 0.0001,
+      "loss": 1.0453,
+      "step": 655
+    },
+    {
+      "epoch": 0.5830463281857572,
+      "grad_norm": 0.43947726488113403,
+      "learning_rate": 0.0001,
+      "loss": 0.9558,
+      "step": 656
+    },
+    {
+      "epoch": 0.5839351183201866,
+      "grad_norm": 0.4180644154548645,
+      "learning_rate": 0.0001,
+      "loss": 0.9144,
+      "step": 657
+    },
+    {
+      "epoch": 0.5848239084546162,
+      "grad_norm": 0.48162519931793213,
+      "learning_rate": 0.0001,
+      "loss": 1.0027,
+      "step": 658
+    },
+    {
+      "epoch": 0.5857126985890456,
+      "grad_norm": 0.42843928933143616,
+      "learning_rate": 0.0001,
+      "loss": 0.9259,
+      "step": 659
+    },
+    {
+      "epoch": 0.5866014887234752,
+      "grad_norm": 0.42587754130363464,
+      "learning_rate": 0.0001,
+      "loss": 0.9438,
+      "step": 660
+    },
+    {
+      "epoch": 0.5874902788579047,
+      "grad_norm": 0.42820873856544495,
+      "learning_rate": 0.0001,
+      "loss": 0.9788,
+      "step": 661
+    },
+    {
+      "epoch": 0.5883790689923342,
+      "grad_norm": 0.44260185956954956,
+      "learning_rate": 0.0001,
+      "loss": 0.951,
+      "step": 662
+    },
+    {
+      "epoch": 0.5892678591267637,
+      "grad_norm": 0.38898608088493347,
+      "learning_rate": 0.0001,
+      "loss": 0.9226,
+      "step": 663
+    },
+    {
+      "epoch": 0.5901566492611932,
+      "grad_norm": 0.4538962244987488,
+      "learning_rate": 0.0001,
+      "loss": 0.9951,
+      "step": 664
+    },
+    {
+      "epoch": 0.5910454393956227,
+      "grad_norm": 0.4434801936149597,
+      "learning_rate": 0.0001,
+      "loss": 0.9968,
+      "step": 665
+    },
+    {
+      "epoch": 0.5919342295300523,
+      "grad_norm": 2.7702479362487793,
+      "learning_rate": 0.0001,
+      "loss": 0.9657,
+      "step": 666
+    },
+    {
+      "epoch": 0.5928230196644817,
+      "grad_norm": 0.4107935130596161,
+      "learning_rate": 0.0001,
+      "loss": 0.8901,
+      "step": 667
+    },
+    {
+      "epoch": 0.5937118097989112,
+      "grad_norm": 5.212909698486328,
+      "learning_rate": 0.0001,
+      "loss": 0.8686,
+      "step": 668
+    },
+    {
+      "epoch": 0.5946005999333407,
+      "grad_norm": 0.4474343955516815,
+      "learning_rate": 0.0001,
+      "loss": 0.8641,
+      "step": 669
+    },
+    {
+      "epoch": 0.5954893900677702,
+      "grad_norm": 0.39664575457572937,
+      "learning_rate": 0.0001,
+      "loss": 0.9605,
+      "step": 670
+    },
+    {
+      "epoch": 0.5963781802021998,
+      "grad_norm": 0.38908809423446655,
+      "learning_rate": 0.0001,
+      "loss": 0.8851,
+      "step": 671
+    },
+    {
+      "epoch": 0.5972669703366292,
+      "grad_norm": 0.39720067381858826,
+      "learning_rate": 0.0001,
+      "loss": 0.9376,
+      "step": 672
+    },
+    {
+      "epoch": 0.5981557604710588,
+      "grad_norm": 0.444224089384079,
+      "learning_rate": 0.0001,
+      "loss": 0.9879,
+      "step": 673
+    },
+    {
+      "epoch": 0.5990445506054882,
+      "grad_norm": 0.461745023727417,
+      "learning_rate": 0.0001,
+      "loss": 1.0298,
+      "step": 674
+    },
+    {
+      "epoch": 0.5999333407399178,
+      "grad_norm": 0.38060104846954346,
+      "learning_rate": 0.0001,
+      "loss": 0.9133,
+      "step": 675
+    },
+    {
+      "epoch": 0.6008221308743473,
+      "grad_norm": 0.4152204990386963,
+      "learning_rate": 0.0001,
+      "loss": 1.0037,
+      "step": 676
+    },
+    {
+      "epoch": 0.6017109210087768,
+      "grad_norm": 0.4251076281070709,
+      "learning_rate": 0.0001,
+      "loss": 0.9217,
+      "step": 677
+    },
+    {
+      "epoch": 0.6025997111432063,
+      "grad_norm": 0.4050005376338959,
+      "learning_rate": 0.0001,
+      "loss": 1.0411,
+      "step": 678
+    },
+    {
+      "epoch": 0.6034885012776359,
+      "grad_norm": 0.4802723228931427,
+      "learning_rate": 0.0001,
+      "loss": 0.9966,
+      "step": 679
+    },
+    {
+      "epoch": 0.6043772914120653,
+      "grad_norm": 0.4158213436603546,
+      "learning_rate": 0.0001,
+      "loss": 0.9684,
+      "step": 680
+    },
+    {
+      "epoch": 0.6052660815464949,
+      "grad_norm": 0.4008488059043884,
+      "learning_rate": 0.0001,
+      "loss": 0.9684,
+      "step": 681
+    },
+    {
+      "epoch": 0.6061548716809243,
+      "grad_norm": 0.4021485149860382,
+      "learning_rate": 0.0001,
+      "loss": 0.8944,
+      "step": 682
+    },
+    {
+      "epoch": 0.6070436618153539,
+      "grad_norm": 0.36115503311157227,
+      "learning_rate": 0.0001,
+      "loss": 0.9677,
+      "step": 683
+    },
+    {
+      "epoch": 0.6079324519497834,
+      "grad_norm": 0.3866066336631775,
+      "learning_rate": 0.0001,
+      "loss": 0.9938,
+      "step": 684
+    },
+    {
+      "epoch": 0.6088212420842128,
+      "grad_norm": 0.47491082549095154,
+      "learning_rate": 0.0001,
+      "loss": 0.9778,
+      "step": 685
+    },
+    {
+      "epoch": 0.6097100322186424,
+      "grad_norm": 0.44795575737953186,
+      "learning_rate": 0.0001,
+      "loss": 0.9964,
+      "step": 686
+    },
+    {
+      "epoch": 0.6105988223530718,
+      "grad_norm": 0.48861074447631836,
+      "learning_rate": 0.0001,
+      "loss": 0.9276,
+      "step": 687
+    },
+    {
+      "epoch": 0.6114876124875014,
+      "grad_norm": 0.4077586233615875,
+      "learning_rate": 0.0001,
+      "loss": 0.9454,
+      "step": 688
+    },
+    {
+      "epoch": 0.6123764026219309,
+      "grad_norm": 0.3845427930355072,
+      "learning_rate": 0.0001,
+      "loss": 0.9704,
+      "step": 689
+    },
+    {
+      "epoch": 0.6132651927563604,
+      "grad_norm": 0.44110408425331116,
+      "learning_rate": 0.0001,
+      "loss": 1.0017,
+      "step": 690
+    },
+    {
+      "epoch": 0.6141539828907899,
+      "grad_norm": 0.38786497712135315,
+      "learning_rate": 0.0001,
+      "loss": 0.9824,
+      "step": 691
+    },
+    {
+      "epoch": 0.6150427730252194,
+      "grad_norm": 0.40381374955177307,
+      "learning_rate": 0.0001,
+      "loss": 0.9184,
+      "step": 692
+    },
+    {
+      "epoch": 0.6159315631596489,
+      "grad_norm": 0.4073936641216278,
+      "learning_rate": 0.0001,
+      "loss": 0.948,
+      "step": 693
+    },
+    {
+      "epoch": 0.6168203532940785,
+      "grad_norm": 0.39875441789627075,
+      "learning_rate": 0.0001,
+      "loss": 0.9917,
+      "step": 694
+    },
+    {
+      "epoch": 0.6177091434285079,
+      "grad_norm": 0.4240407943725586,
+      "learning_rate": 0.0001,
+      "loss": 0.9696,
+      "step": 695
+    },
+    {
+      "epoch": 0.6185979335629375,
+      "grad_norm": 0.4107338488101959,
+      "learning_rate": 0.0001,
+      "loss": 1.0108,
+      "step": 696
+    },
+    {
+      "epoch": 0.619486723697367,
+      "grad_norm": 0.4649637043476105,
+      "learning_rate": 0.0001,
+      "loss": 1.0222,
+      "step": 697
+    },
+    {
+      "epoch": 0.6203755138317965,
+      "grad_norm": 0.398387610912323,
+      "learning_rate": 0.0001,
+      "loss": 0.9952,
+      "step": 698
+    },
+    {
+      "epoch": 0.621264303966226,
+      "grad_norm": 0.4626375734806061,
+      "learning_rate": 0.0001,
+      "loss": 1.0333,
+      "step": 699
+    },
+    {
+      "epoch": 0.6221530941006554,
+      "grad_norm": 0.39919617772102356,
+      "learning_rate": 0.0001,
+      "loss": 0.8835,
+      "step": 700
+    },
+    {
+      "epoch": 0.623041884235085,
+      "grad_norm": 0.3973129391670227,
+      "learning_rate": 0.0001,
+      "loss": 1.0366,
+      "step": 701
+    },
+    {
+      "epoch": 0.6239306743695145,
+      "grad_norm": 0.4347308576107025,
+      "learning_rate": 0.0001,
+      "loss": 0.98,
+      "step": 702
+    },
+    {
+      "epoch": 0.624819464503944,
+      "grad_norm": 0.42371174693107605,
+      "learning_rate": 0.0001,
+      "loss": 0.9625,
+      "step": 703
+    },
+    {
+      "epoch": 0.6257082546383735,
+      "grad_norm": 0.3887942135334015,
+      "learning_rate": 0.0001,
+      "loss": 0.9995,
+      "step": 704
+    },
+    {
+      "epoch": 0.626597044772803,
+      "grad_norm": 0.41830766201019287,
+      "learning_rate": 0.0001,
+      "loss": 1.0613,
+      "step": 705
+    },
+    {
+      "epoch": 0.6274858349072325,
+      "grad_norm": 0.37216201424598694,
+      "learning_rate": 0.0001,
+      "loss": 0.9003,
+      "step": 706
+    },
+    {
+      "epoch": 0.6283746250416621,
+      "grad_norm": 0.3734080493450165,
+      "learning_rate": 0.0001,
+      "loss": 1.0583,
+      "step": 707
+    },
+    {
+      "epoch": 0.6292634151760915,
+      "grad_norm": 0.4222297966480255,
+      "learning_rate": 0.0001,
+      "loss": 0.8877,
+      "step": 708
+    },
+    {
+      "epoch": 0.6301522053105211,
+      "grad_norm": 0.4466676712036133,
+      "learning_rate": 0.0001,
+      "loss": 0.9868,
+      "step": 709
+    },
+    {
+      "epoch": 0.6310409954449505,
+      "grad_norm": 0.4170881509780884,
+      "learning_rate": 0.0001,
+      "loss": 0.9235,
+      "step": 710
+    },
+    {
+      "epoch": 0.6319297855793801,
+      "grad_norm": 0.42569923400878906,
+      "learning_rate": 0.0001,
+      "loss": 1.0111,
+      "step": 711
+    },
+    {
+      "epoch": 0.6328185757138096,
+      "grad_norm": 0.469657838344574,
+      "learning_rate": 0.0001,
+      "loss": 1.1188,
+      "step": 712
+    },
+    {
+      "epoch": 0.6337073658482391,
+      "grad_norm": 0.4612179696559906,
+      "learning_rate": 0.0001,
+      "loss": 1.0049,
+      "step": 713
+    },
+    {
+      "epoch": 0.6345961559826686,
+      "grad_norm": 0.4614596664905548,
+      "learning_rate": 0.0001,
+      "loss": 0.9497,
+      "step": 714
+    },
+    {
+      "epoch": 0.635484946117098,
+      "grad_norm": 0.4139200747013092,
+      "learning_rate": 0.0001,
+      "loss": 0.9691,
+      "step": 715
+    },
+    {
+      "epoch": 0.6363737362515276,
+      "grad_norm": 0.42002081871032715,
+      "learning_rate": 0.0001,
+      "loss": 0.9838,
+      "step": 716
+    },
+    {
+      "epoch": 0.6372625263859572,
+      "grad_norm": 0.4355138838291168,
+      "learning_rate": 0.0001,
+      "loss": 1.0074,
+      "step": 717
+    },
+    {
+      "epoch": 0.6381513165203866,
+      "grad_norm": 0.42147886753082275,
+      "learning_rate": 0.0001,
+      "loss": 0.9929,
+      "step": 718
+    },
+    {
+      "epoch": 0.6390401066548161,
+      "grad_norm": 0.4365898370742798,
+      "learning_rate": 0.0001,
+      "loss": 1.0325,
+      "step": 719
+    },
+    {
+      "epoch": 0.6399288967892457,
+      "grad_norm": 0.4107733964920044,
+      "learning_rate": 0.0001,
+      "loss": 0.9229,
+      "step": 720
+    },
+    {
+      "epoch": 0.6408176869236751,
+      "grad_norm": 0.4189467430114746,
+      "learning_rate": 0.0001,
+      "loss": 1.0041,
+      "step": 721
+    },
+    {
+      "epoch": 0.6417064770581047,
+      "grad_norm": 0.4112345278263092,
+      "learning_rate": 0.0001,
+      "loss": 1.017,
+      "step": 722
+    },
+    {
+      "epoch": 0.6425952671925341,
+      "grad_norm": 0.39325135946273804,
+      "learning_rate": 0.0001,
+      "loss": 0.9264,
+      "step": 723
+    },
+    {
+      "epoch": 0.6434840573269637,
+      "grad_norm": 0.38210329413414,
+      "learning_rate": 0.0001,
+      "loss": 0.9947,
+      "step": 724
+    },
+    {
+      "epoch": 0.6443728474613932,
+      "grad_norm": 0.4400337040424347,
+      "learning_rate": 0.0001,
+      "loss": 0.9118,
+      "step": 725
+    },
+    {
+      "epoch": 0.6452616375958227,
+      "grad_norm": 0.41924870014190674,
+      "learning_rate": 0.0001,
+      "loss": 1.0176,
+      "step": 726
+    },
+    {
+      "epoch": 0.6461504277302522,
+      "grad_norm": 0.37929868698120117,
+      "learning_rate": 0.0001,
+      "loss": 0.906,
+      "step": 727
+    },
+    {
+      "epoch": 0.6470392178646817,
+      "grad_norm": 0.41478589177131653,
+      "learning_rate": 0.0001,
+      "loss": 0.9984,
+      "step": 728
+    },
+    {
+      "epoch": 0.6479280079991112,
+      "grad_norm": 0.3930301368236542,
+      "learning_rate": 0.0001,
+      "loss": 0.9251,
+      "step": 729
+    },
+    {
+      "epoch": 0.6488167981335408,
+      "grad_norm": 0.3816480338573456,
+      "learning_rate": 0.0001,
+      "loss": 0.9501,
+      "step": 730
+    },
+    {
+      "epoch": 0.6497055882679702,
+      "grad_norm": 0.40965738892555237,
+      "learning_rate": 0.0001,
+      "loss": 0.8565,
+      "step": 731
+    },
+    {
+      "epoch": 0.6505943784023998,
+      "grad_norm": 0.42315876483917236,
+      "learning_rate": 0.0001,
+      "loss": 1.0507,
+      "step": 732
+    },
+    {
+      "epoch": 0.6514831685368292,
+      "grad_norm": 0.3918554484844208,
+      "learning_rate": 0.0001,
+      "loss": 0.9513,
+      "step": 733
+    },
+    {
+      "epoch": 0.6523719586712587,
+      "grad_norm": 0.42279428243637085,
+      "learning_rate": 0.0001,
+      "loss": 0.9699,
+      "step": 734
+    },
+    {
+      "epoch": 0.6532607488056883,
+      "grad_norm": 0.41292804479599,
+      "learning_rate": 0.0001,
+      "loss": 0.9509,
+      "step": 735
+    },
+    {
+      "epoch": 0.6541495389401177,
+      "grad_norm": 0.3694443106651306,
+      "learning_rate": 0.0001,
+      "loss": 0.8812,
+      "step": 736
+    },
+    {
+      "epoch": 0.6550383290745473,
+      "grad_norm": 0.42552173137664795,
+      "learning_rate": 0.0001,
+      "loss": 0.9628,
+      "step": 737
+    },
+    {
+      "epoch": 0.6559271192089767,
+      "grad_norm": 0.4035997986793518,
+      "learning_rate": 0.0001,
+      "loss": 0.9955,
+      "step": 738
+    },
+    {
+      "epoch": 0.6568159093434063,
+      "grad_norm": 0.38716575503349304,
+      "learning_rate": 0.0001,
+      "loss": 0.9339,
+      "step": 739
+    },
+    {
+      "epoch": 0.6577046994778358,
+      "grad_norm": 0.37816059589385986,
+      "learning_rate": 0.0001,
+      "loss": 1.0141,
+      "step": 740
+    },
+    {
+      "epoch": 0.6585934896122653,
+      "grad_norm": 0.39885637164115906,
+      "learning_rate": 0.0001,
+      "loss": 0.9982,
+      "step": 741
+    },
+    {
+      "epoch": 0.6594822797466948,
+      "grad_norm": 0.41604557633399963,
+      "learning_rate": 0.0001,
+      "loss": 0.9505,
+      "step": 742
+    },
+    {
+      "epoch": 0.6603710698811244,
+      "grad_norm": 0.3828152120113373,
+      "learning_rate": 0.0001,
+      "loss": 0.9055,
+      "step": 743
+    },
+    {
+      "epoch": 0.6612598600155538,
+      "grad_norm": 0.39052262902259827,
+      "learning_rate": 0.0001,
+      "loss": 0.9922,
+      "step": 744
+    },
+    {
+      "epoch": 0.6621486501499834,
+      "grad_norm": 0.36966073513031006,
+      "learning_rate": 0.0001,
+      "loss": 1.0097,
+      "step": 745
+    },
+    {
+      "epoch": 0.6630374402844128,
+      "grad_norm": 0.40489739179611206,
+      "learning_rate": 0.0001,
+      "loss": 0.9734,
+      "step": 746
+    },
+    {
+      "epoch": 0.6639262304188424,
+      "grad_norm": 0.4752523899078369,
+      "learning_rate": 0.0001,
+      "loss": 1.0276,
+      "step": 747
+    },
+    {
+      "epoch": 0.6648150205532719,
+      "grad_norm": 0.42752546072006226,
+      "learning_rate": 0.0001,
+      "loss": 0.9565,
+      "step": 748
+    },
+    {
+      "epoch": 0.6657038106877013,
+      "grad_norm": 0.3752210736274719,
+      "learning_rate": 0.0001,
+      "loss": 0.9709,
+      "step": 749
+    },
+    {
+      "epoch": 0.6665926008221309,
+      "grad_norm": 0.4873553514480591,
+      "learning_rate": 0.0001,
+      "loss": 1.0433,
+      "step": 750
+    },
+    {
+      "epoch": 0.6674813909565603,
+      "grad_norm": 0.4026888906955719,
+      "learning_rate": 0.0001,
+      "loss": 1.0882,
+      "step": 751
+    },
+    {
+      "epoch": 0.6683701810909899,
+      "grad_norm": 0.42212575674057007,
+      "learning_rate": 0.0001,
+      "loss": 0.9615,
+      "step": 752
+    },
+    {
+      "epoch": 0.6692589712254194,
+      "grad_norm": 0.40002021193504333,
+      "learning_rate": 0.0001,
+      "loss": 1.0106,
+      "step": 753
+    },
+    {
+      "epoch": 0.6701477613598489,
+      "grad_norm": 0.4148339331150055,
+      "learning_rate": 0.0001,
+      "loss": 0.9292,
+      "step": 754
+    },
+    {
+      "epoch": 0.6710365514942784,
+      "grad_norm": 0.37192678451538086,
+      "learning_rate": 0.0001,
+      "loss": 0.9844,
+      "step": 755
+    },
+    {
+      "epoch": 0.6719253416287079,
+      "grad_norm": 0.3779515326023102,
+      "learning_rate": 0.0001,
+      "loss": 0.8999,
+      "step": 756
+    },
+    {
+      "epoch": 0.6728141317631374,
+      "grad_norm": 0.38838016986846924,
+      "learning_rate": 0.0001,
+      "loss": 0.9855,
+      "step": 757
+    },
+    {
+      "epoch": 0.673702921897567,
+      "grad_norm": 0.3978157937526703,
+      "learning_rate": 0.0001,
+      "loss": 0.9253,
+      "step": 758
+    },
+    {
+      "epoch": 0.6745917120319964,
+      "grad_norm": 0.38334494829177856,
+      "learning_rate": 0.0001,
+      "loss": 0.8678,
+      "step": 759
+    },
+    {
+      "epoch": 0.675480502166426,
+      "grad_norm": 0.3856591582298279,
+      "learning_rate": 0.0001,
+      "loss": 0.9614,
+      "step": 760
+    },
+    {
+      "epoch": 0.6763692923008554,
+      "grad_norm": 0.6821391582489014,
+      "learning_rate": 0.0001,
+      "loss": 0.9884,
+      "step": 761
+    },
+    {
+      "epoch": 0.677258082435285,
+      "grad_norm": 0.4022596478462219,
+      "learning_rate": 0.0001,
+      "loss": 0.9512,
+      "step": 762
+    },
+    {
+      "epoch": 0.6781468725697145,
+      "grad_norm": 0.3981231153011322,
+      "learning_rate": 0.0001,
+      "loss": 0.9457,
+      "step": 763
+    },
+    {
+      "epoch": 0.679035662704144,
+      "grad_norm": 0.466782808303833,
+      "learning_rate": 0.0001,
+      "loss": 1.0009,
+      "step": 764
+    },
+    {
+      "epoch": 0.6799244528385735,
+      "grad_norm": 0.39416226744651794,
+      "learning_rate": 0.0001,
+      "loss": 0.9204,
+      "step": 765
+    },
+    {
+      "epoch": 0.680813242973003,
+      "grad_norm": 0.3830525279045105,
+      "learning_rate": 0.0001,
+      "loss": 0.9206,
+      "step": 766
+    },
+    {
+      "epoch": 0.6817020331074325,
+      "grad_norm": 0.3653806447982788,
+      "learning_rate": 0.0001,
+      "loss": 0.9601,
+      "step": 767
+    },
+    {
+      "epoch": 0.682590823241862,
+      "grad_norm": 0.4308638274669647,
+      "learning_rate": 0.0001,
+      "loss": 0.9666,
+      "step": 768
+    },
+    {
+      "epoch": 0.6834796133762915,
+      "grad_norm": 0.3821423649787903,
+      "learning_rate": 0.0001,
+      "loss": 1.0252,
+      "step": 769
+    },
+    {
+      "epoch": 0.684368403510721,
+      "grad_norm": 0.459150493144989,
+      "learning_rate": 0.0001,
+      "loss": 0.8923,
+      "step": 770
+    },
+    {
+      "epoch": 0.6852571936451506,
+      "grad_norm": 0.4109600782394409,
+      "learning_rate": 0.0001,
+      "loss": 0.9337,
+      "step": 771
+    },
+    {
+      "epoch": 0.68614598377958,
+      "grad_norm": 0.4196016788482666,
+      "learning_rate": 0.0001,
+      "loss": 1.0004,
+      "step": 772
+    },
+    {
+      "epoch": 0.6870347739140096,
+      "grad_norm": 0.39674344658851624,
+      "learning_rate": 0.0001,
+      "loss": 0.9676,
+      "step": 773
+    },
+    {
+      "epoch": 0.687923564048439,
+      "grad_norm": 0.3917883634567261,
+      "learning_rate": 0.0001,
+      "loss": 0.9255,
+      "step": 774
+    },
+    {
+      "epoch": 0.6888123541828686,
+      "grad_norm": 0.4198206663131714,
+      "learning_rate": 0.0001,
+      "loss": 0.9159,
+      "step": 775
+    },
+    {
+      "epoch": 0.6897011443172981,
+      "grad_norm": 0.4153868556022644,
+      "learning_rate": 0.0001,
+      "loss": 1.0006,
+      "step": 776
+    },
+    {
+      "epoch": 0.6905899344517276,
+      "grad_norm": 0.425462007522583,
+      "learning_rate": 0.0001,
+      "loss": 0.843,
+      "step": 777
+    },
+    {
+      "epoch": 0.6914787245861571,
+      "grad_norm": 0.40152454376220703,
+      "learning_rate": 0.0001,
+      "loss": 0.8695,
+      "step": 778
+    },
+    {
+      "epoch": 0.6923675147205866,
+      "grad_norm": 0.39553624391555786,
+      "learning_rate": 0.0001,
+      "loss": 0.8722,
+      "step": 779
+    },
+    {
+      "epoch": 0.6932563048550161,
+      "grad_norm": 0.4519752860069275,
+      "learning_rate": 0.0001,
+      "loss": 1.0222,
+      "step": 780
+    },
+    {
+      "epoch": 0.6941450949894457,
+      "grad_norm": 0.3791448473930359,
+      "learning_rate": 0.0001,
+      "loss": 0.8787,
+      "step": 781
+    },
+    {
+      "epoch": 0.6950338851238751,
+      "grad_norm": 0.3689649701118469,
+      "learning_rate": 0.0001,
+      "loss": 0.973,
+      "step": 782
+    },
+    {
+      "epoch": 0.6959226752583046,
+      "grad_norm": 0.36836689710617065,
+      "learning_rate": 0.0001,
+      "loss": 0.924,
+      "step": 783
+    },
+    {
+      "epoch": 0.6968114653927341,
+      "grad_norm": 0.4386933147907257,
+      "learning_rate": 0.0001,
+      "loss": 1.0039,
+      "step": 784
+    },
+    {
+      "epoch": 0.6977002555271636,
+      "grad_norm": 0.39001014828681946,
+      "learning_rate": 0.0001,
+      "loss": 0.9655,
+      "step": 785
+    },
+    {
+      "epoch": 0.6985890456615932,
+      "grad_norm": 0.36727163195610046,
+      "learning_rate": 0.0001,
+      "loss": 0.9681,
+      "step": 786
+    },
+    {
+      "epoch": 0.6994778357960226,
+      "grad_norm": 0.3847578465938568,
+      "learning_rate": 0.0001,
+      "loss": 0.9809,
+      "step": 787
+    },
+    {
+      "epoch": 0.7003666259304522,
+      "grad_norm": 0.3912128806114197,
+      "learning_rate": 0.0001,
+      "loss": 0.9828,
+      "step": 788
+    },
+    {
+      "epoch": 0.7012554160648817,
+      "grad_norm": 0.3891061842441559,
+      "learning_rate": 0.0001,
+      "loss": 0.9816,
+      "step": 789
+    },
+    {
+      "epoch": 0.7021442061993112,
+      "grad_norm": 0.40884703397750854,
+      "learning_rate": 0.0001,
+      "loss": 0.9406,
+      "step": 790
+    },
+    {
+      "epoch": 0.7030329963337407,
+      "grad_norm": 0.40863409638404846,
+      "learning_rate": 0.0001,
+      "loss": 0.9473,
+      "step": 791
+    },
+    {
+      "epoch": 0.7039217864681702,
+      "grad_norm": 0.4846546947956085,
+      "learning_rate": 0.0001,
+      "loss": 1.0337,
+      "step": 792
+    },
+    {
+      "epoch": 0.7048105766025997,
+      "grad_norm": 0.4108925759792328,
+      "learning_rate": 0.0001,
+      "loss": 0.8871,
+      "step": 793
+    },
+    {
+      "epoch": 0.7056993667370293,
+      "grad_norm": 0.41259604692459106,
+      "learning_rate": 0.0001,
+      "loss": 0.9616,
+      "step": 794
+    },
+    {
+      "epoch": 0.7065881568714587,
+      "grad_norm": 0.39563241600990295,
+      "learning_rate": 0.0001,
+      "loss": 0.917,
+      "step": 795
+    },
+    {
+      "epoch": 0.7074769470058883,
+      "grad_norm": 0.44878554344177246,
+      "learning_rate": 0.0001,
+      "loss": 0.9133,
+      "step": 796
+    },
+    {
+      "epoch": 0.7083657371403177,
+      "grad_norm": 0.4403840899467468,
+      "learning_rate": 0.0001,
+      "loss": 0.961,
+      "step": 797
+    },
+    {
+      "epoch": 0.7092545272747472,
+      "grad_norm": 0.47209668159484863,
+      "learning_rate": 0.0001,
+      "loss": 0.9692,
+      "step": 798
+    },
+    {
+      "epoch": 0.7101433174091768,
+      "grad_norm": 0.3841542601585388,
+      "learning_rate": 0.0001,
+      "loss": 0.9458,
+      "step": 799
+    },
+    {
+      "epoch": 0.7110321075436062,
+      "grad_norm": 0.46148043870925903,
+      "learning_rate": 0.0001,
+      "loss": 0.8355,
+      "step": 800
+    },
+    {
+      "epoch": 0.7119208976780358,
+      "grad_norm": 0.6854037046432495,
+      "learning_rate": 0.0001,
+      "loss": 0.9275,
+      "step": 801
+    },
+    {
+      "epoch": 0.7128096878124652,
+      "grad_norm": 0.38869839906692505,
+      "learning_rate": 0.0001,
+      "loss": 0.9882,
+      "step": 802
+    },
+    {
+      "epoch": 0.7136984779468948,
+      "grad_norm": 0.4120323956012726,
+      "learning_rate": 0.0001,
+      "loss": 0.87,
+      "step": 803
+    },
+    {
+      "epoch": 0.7145872680813243,
+      "grad_norm": 0.43215325474739075,
+      "learning_rate": 0.0001,
+      "loss": 0.9797,
+      "step": 804
+    },
+    {
+      "epoch": 0.7154760582157538,
+      "grad_norm": 0.4285464882850647,
+      "learning_rate": 0.0001,
+      "loss": 0.9303,
+      "step": 805
+    },
+    {
+      "epoch": 0.7163648483501833,
+      "grad_norm": 0.3982923626899719,
+      "learning_rate": 0.0001,
+      "loss": 0.9645,
+      "step": 806
+    },
+    {
+      "epoch": 0.7172536384846128,
+      "grad_norm": 0.3845166563987732,
+      "learning_rate": 0.0001,
+      "loss": 0.8746,
+      "step": 807
+    },
+    {
+      "epoch": 0.7181424286190423,
+      "grad_norm": 0.34493544697761536,
+      "learning_rate": 0.0001,
+      "loss": 0.8451,
+      "step": 808
+    },
+    {
+      "epoch": 0.7190312187534719,
+      "grad_norm": 0.4371100962162018,
+      "learning_rate": 0.0001,
+      "loss": 1.0098,
+      "step": 809
+    },
+    {
+      "epoch": 0.7199200088879013,
+      "grad_norm": 0.4431426525115967,
+      "learning_rate": 0.0001,
+      "loss": 0.9671,
+      "step": 810
+    },
+    {
+      "epoch": 0.7208087990223309,
+      "grad_norm": 0.39591118693351746,
+      "learning_rate": 0.0001,
+      "loss": 0.9669,
+      "step": 811
+    },
+    {
+      "epoch": 0.7216975891567604,
+      "grad_norm": 0.4144839942455292,
+      "learning_rate": 0.0001,
+      "loss": 1.0489,
+      "step": 812
+    },
+    {
+      "epoch": 0.7225863792911899,
+      "grad_norm": 0.36368703842163086,
+      "learning_rate": 0.0001,
+      "loss": 0.9269,
+      "step": 813
+    },
+    {
+      "epoch": 0.7234751694256194,
+      "grad_norm": 0.34978145360946655,
+      "learning_rate": 0.0001,
+      "loss": 0.925,
+      "step": 814
+    },
+    {
+      "epoch": 0.7243639595600488,
+      "grad_norm": 0.38028684258461,
+      "learning_rate": 0.0001,
+      "loss": 1.0324,
+      "step": 815
+    },
+    {
+      "epoch": 0.7252527496944784,
+      "grad_norm": 0.3885200023651123,
+      "learning_rate": 0.0001,
+      "loss": 0.8994,
+      "step": 816
+    },
+    {
+      "epoch": 0.726141539828908,
+      "grad_norm": 0.3828847110271454,
+      "learning_rate": 0.0001,
+      "loss": 0.977,
+      "step": 817
+    },
+    {
+      "epoch": 0.7270303299633374,
+      "grad_norm": 0.41237783432006836,
+      "learning_rate": 0.0001,
+      "loss": 0.9004,
+      "step": 818
+    },
+    {
+      "epoch": 0.7279191200977669,
+      "grad_norm": 0.3976728916168213,
+      "learning_rate": 0.0001,
+      "loss": 0.8497,
+      "step": 819
+    },
+    {
+      "epoch": 0.7288079102321964,
+      "grad_norm": 0.3727687895298004,
+      "learning_rate": 0.0001,
+      "loss": 0.9518,
+      "step": 820
+    },
+    {
+      "epoch": 0.7296967003666259,
+      "grad_norm": 0.36306801438331604,
+      "learning_rate": 0.0001,
+      "loss": 0.9451,
+      "step": 821
+    },
+    {
+      "epoch": 0.7305854905010555,
+      "grad_norm": 0.38123902678489685,
+      "learning_rate": 0.0001,
+      "loss": 0.9649,
+      "step": 822
+    },
+    {
+      "epoch": 0.7314742806354849,
+      "grad_norm": 0.4005994200706482,
+      "learning_rate": 0.0001,
+      "loss": 0.9271,
+      "step": 823
+    },
+    {
+      "epoch": 0.7323630707699145,
+      "grad_norm": 0.36427778005599976,
+      "learning_rate": 0.0001,
+      "loss": 0.8705,
+      "step": 824
+    },
+    {
+      "epoch": 0.7332518609043439,
+      "grad_norm": 0.39238572120666504,
+      "learning_rate": 0.0001,
+      "loss": 1.0382,
+      "step": 825
+    },
+    {
+      "epoch": 0.7341406510387735,
+      "grad_norm": 0.3285076320171356,
+      "learning_rate": 0.0001,
+      "loss": 0.8946,
+      "step": 826
+    },
+    {
+      "epoch": 0.735029441173203,
+      "grad_norm": 0.4472292363643646,
+      "learning_rate": 0.0001,
+      "loss": 1.0144,
+      "step": 827
+    },
+    {
+      "epoch": 0.7359182313076325,
+      "grad_norm": 0.43173158168792725,
+      "learning_rate": 0.0001,
+      "loss": 0.918,
+      "step": 828
+    },
+    {
+      "epoch": 0.736807021442062,
+      "grad_norm": 0.3840146064758301,
+      "learning_rate": 0.0001,
+      "loss": 0.825,
+      "step": 829
+    },
+    {
+      "epoch": 0.7376958115764916,
+      "grad_norm": 0.3785271644592285,
+      "learning_rate": 0.0001,
+      "loss": 0.9523,
+      "step": 830
+    },
+    {
+      "epoch": 0.738584601710921,
+      "grad_norm": 0.4263155460357666,
+      "learning_rate": 0.0001,
+      "loss": 0.9,
+      "step": 831
+    },
+    {
+      "epoch": 0.7394733918453505,
+      "grad_norm": 0.3733021020889282,
+      "learning_rate": 0.0001,
+      "loss": 0.9072,
+      "step": 832
+    },
+    {
+      "epoch": 0.74036218197978,
+      "grad_norm": 0.4950011074542999,
+      "learning_rate": 0.0001,
+      "loss": 1.0596,
+      "step": 833
+    },
+    {
+      "epoch": 0.7412509721142095,
+      "grad_norm": 0.4155547320842743,
+      "learning_rate": 0.0001,
+      "loss": 0.9846,
+      "step": 834
+    },
+    {
+      "epoch": 0.7421397622486391,
+      "grad_norm": 0.39016374945640564,
+      "learning_rate": 0.0001,
+      "loss": 0.9511,
+      "step": 835
+    },
+    {
+      "epoch": 0.7430285523830685,
+      "grad_norm": 0.44248321652412415,
+      "learning_rate": 0.0001,
+      "loss": 0.9117,
+      "step": 836
+    },
+    {
+      "epoch": 0.7439173425174981,
+      "grad_norm": 0.4027865529060364,
+      "learning_rate": 0.0001,
+      "loss": 0.9217,
+      "step": 837
+    },
+    {
+      "epoch": 0.7448061326519275,
+      "grad_norm": 0.4021622836589813,
+      "learning_rate": 0.0001,
+      "loss": 0.9234,
+      "step": 838
+    },
+    {
+      "epoch": 0.7456949227863571,
+      "grad_norm": 0.4510715901851654,
+      "learning_rate": 0.0001,
+      "loss": 1.0242,
+      "step": 839
+    },
+    {
+      "epoch": 0.7465837129207866,
+      "grad_norm": 0.38624054193496704,
+      "learning_rate": 0.0001,
+      "loss": 0.9285,
+      "step": 840
+    },
+    {
+      "epoch": 0.7474725030552161,
+      "grad_norm": 0.48192909359931946,
+      "learning_rate": 0.0001,
+      "loss": 1.0023,
+      "step": 841
+    },
+    {
+      "epoch": 0.7483612931896456,
+      "grad_norm": 0.4182127118110657,
+      "learning_rate": 0.0001,
+      "loss": 1.04,
+      "step": 842
+    },
+    {
+      "epoch": 0.7492500833240751,
+      "grad_norm": 0.3804481029510498,
+      "learning_rate": 0.0001,
+      "loss": 0.9602,
+      "step": 843
+    },
+    {
+      "epoch": 0.7501388734585046,
+      "grad_norm": 0.3628024458885193,
+      "learning_rate": 0.0001,
+      "loss": 0.955,
+      "step": 844
+    },
+    {
+      "epoch": 0.7510276635929342,
+      "grad_norm": 0.3579307496547699,
+      "learning_rate": 0.0001,
+      "loss": 0.9355,
+      "step": 845
+    },
+    {
+      "epoch": 0.7519164537273636,
+      "grad_norm": 0.36626923084259033,
+      "learning_rate": 0.0001,
+      "loss": 0.9324,
+      "step": 846
+    },
+    {
+      "epoch": 0.7528052438617931,
+      "grad_norm": 0.37422046065330505,
+      "learning_rate": 0.0001,
+      "loss": 1.0267,
+      "step": 847
+    },
+    {
+      "epoch": 0.7536940339962226,
+      "grad_norm": 0.3965827226638794,
+      "learning_rate": 0.0001,
+      "loss": 0.9784,
+      "step": 848
+    },
+    {
+      "epoch": 0.7545828241306521,
+      "grad_norm": 0.36925482749938965,
+      "learning_rate": 0.0001,
+      "loss": 0.9801,
+      "step": 849
+    },
+    {
+      "epoch": 0.7554716142650817,
+      "grad_norm": 0.372070848941803,
+      "learning_rate": 0.0001,
+      "loss": 0.9474,
+      "step": 850
+    },
+    {
+      "epoch": 0.7563604043995111,
+      "grad_norm": 0.4122565984725952,
+      "learning_rate": 0.0001,
+      "loss": 0.9738,
+      "step": 851
+    },
+    {
+      "epoch": 0.7572491945339407,
+      "grad_norm": 0.36244523525238037,
+      "learning_rate": 0.0001,
+      "loss": 0.9174,
+      "step": 852
+    },
+    {
+      "epoch": 0.7581379846683702,
+      "grad_norm": 0.3738951086997986,
+      "learning_rate": 0.0001,
+      "loss": 0.9236,
+      "step": 853
+    },
+    {
+      "epoch": 0.7590267748027997,
+      "grad_norm": 0.3978452980518341,
+      "learning_rate": 0.0001,
+      "loss": 1.0203,
+      "step": 854
+    },
+    {
+      "epoch": 0.7599155649372292,
+      "grad_norm": 0.38576042652130127,
+      "learning_rate": 0.0001,
+      "loss": 0.9305,
+      "step": 855
+    },
+    {
+      "epoch": 0.7608043550716587,
+      "grad_norm": 0.39469850063323975,
+      "learning_rate": 0.0001,
+      "loss": 0.9165,
+      "step": 856
+    },
+    {
+      "epoch": 0.7616931452060882,
+      "grad_norm": 0.4054928719997406,
+      "learning_rate": 0.0001,
+      "loss": 0.8531,
+      "step": 857
+    },
+    {
+      "epoch": 0.7625819353405178,
+      "grad_norm": 0.42798909544944763,
+      "learning_rate": 0.0001,
+      "loss": 0.8747,
+      "step": 858
+    },
+    {
+      "epoch": 0.7634707254749472,
+      "grad_norm": 0.42001426219940186,
+      "learning_rate": 0.0001,
+      "loss": 0.9606,
+      "step": 859
+    },
+    {
+      "epoch": 0.7643595156093768,
+      "grad_norm": 0.3773418366909027,
+      "learning_rate": 0.0001,
+      "loss": 0.9564,
+      "step": 860
+    },
+    {
+      "epoch": 0.7652483057438062,
+      "grad_norm": 0.3583545982837677,
+      "learning_rate": 0.0001,
+      "loss": 0.9112,
+      "step": 861
+    },
+    {
+      "epoch": 0.7661370958782358,
+      "grad_norm": 0.4381794035434723,
+      "learning_rate": 0.0001,
+      "loss": 1.0014,
+      "step": 862
+    },
+    {
+      "epoch": 0.7670258860126653,
+      "grad_norm": 0.40912652015686035,
+      "learning_rate": 0.0001,
+      "loss": 0.9742,
+      "step": 863
+    },
+    {
+      "epoch": 0.7679146761470947,
+      "grad_norm": 0.3959810733795166,
+      "learning_rate": 0.0001,
+      "loss": 0.9635,
+      "step": 864
+    },
+    {
+      "epoch": 0.7688034662815243,
+      "grad_norm": 0.3853726387023926,
+      "learning_rate": 0.0001,
+      "loss": 0.9379,
+      "step": 865
+    },
+    {
+      "epoch": 0.7696922564159537,
+      "grad_norm": 0.4313822388648987,
+      "learning_rate": 0.0001,
+      "loss": 1.0387,
+      "step": 866
+    },
+    {
+      "epoch": 0.7705810465503833,
+      "grad_norm": 0.34767740964889526,
+      "learning_rate": 0.0001,
+      "loss": 0.953,
+      "step": 867
+    },
+    {
+      "epoch": 0.7714698366848128,
+      "grad_norm": 0.414546936750412,
+      "learning_rate": 0.0001,
+      "loss": 0.9447,
+      "step": 868
+    },
+    {
+      "epoch": 0.7723586268192423,
+      "grad_norm": 0.3723791539669037,
+      "learning_rate": 0.0001,
+      "loss": 0.9898,
+      "step": 869
+    },
+    {
+      "epoch": 0.7732474169536718,
+      "grad_norm": 0.3587020933628082,
+      "learning_rate": 0.0001,
+      "loss": 0.9793,
+      "step": 870
+    },
+    {
+      "epoch": 0.7741362070881013,
+      "grad_norm": 0.37663257122039795,
+      "learning_rate": 0.0001,
+      "loss": 0.898,
+      "step": 871
+    },
+    {
+      "epoch": 0.7750249972225308,
+      "grad_norm": 0.4134519398212433,
+      "learning_rate": 0.0001,
+      "loss": 0.8667,
+      "step": 872
+    },
+    {
+      "epoch": 0.7759137873569604,
+      "grad_norm": 0.3675340414047241,
+      "learning_rate": 0.0001,
+      "loss": 0.9326,
+      "step": 873
+    },
+    {
+      "epoch": 0.7768025774913898,
+      "grad_norm": 0.355490118265152,
+      "learning_rate": 0.0001,
+      "loss": 1.0111,
+      "step": 874
+    },
+    {
+      "epoch": 0.7776913676258194,
+      "grad_norm": 0.34475648403167725,
+      "learning_rate": 0.0001,
+      "loss": 0.9417,
+      "step": 875
+    },
+    {
+      "epoch": 0.7785801577602489,
+      "grad_norm": 0.3858400285243988,
+      "learning_rate": 0.0001,
+      "loss": 0.9454,
+      "step": 876
+    },
+    {
+      "epoch": 0.7794689478946784,
+      "grad_norm": 0.41093963384628296,
+      "learning_rate": 0.0001,
+      "loss": 0.9135,
+      "step": 877
+    },
+    {
+      "epoch": 0.7803577380291079,
+      "grad_norm": 0.3808945417404175,
+      "learning_rate": 0.0001,
+      "loss": 1.0214,
+      "step": 878
+    },
+    {
+      "epoch": 0.7812465281635373,
+      "grad_norm": 0.3945116400718689,
+      "learning_rate": 0.0001,
+      "loss": 0.8784,
+      "step": 879
+    },
+    {
+      "epoch": 0.7821353182979669,
+      "grad_norm": 0.34549954533576965,
+      "learning_rate": 0.0001,
+      "loss": 0.8422,
+      "step": 880
+    },
+    {
+      "epoch": 0.7830241084323964,
+      "grad_norm": 0.39158105850219727,
+      "learning_rate": 0.0001,
+      "loss": 0.8915,
+      "step": 881
+    },
+    {
+      "epoch": 0.7839128985668259,
+      "grad_norm": 0.41858598589897156,
+      "learning_rate": 0.0001,
+      "loss": 1.0106,
+      "step": 882
+    },
+    {
+      "epoch": 0.7848016887012554,
+      "grad_norm": 0.3741881251335144,
+      "learning_rate": 0.0001,
+      "loss": 0.9062,
+      "step": 883
+    },
+    {
+      "epoch": 0.7856904788356849,
+      "grad_norm": 0.42688676714897156,
+      "learning_rate": 0.0001,
+      "loss": 1.0078,
+      "step": 884
+    },
+    {
+      "epoch": 0.7865792689701144,
+      "grad_norm": 0.4139765799045563,
+      "learning_rate": 0.0001,
+      "loss": 0.9226,
+      "step": 885
+    },
+    {
+      "epoch": 0.787468059104544,
+      "grad_norm": 0.34299251437187195,
+      "learning_rate": 0.0001,
+      "loss": 0.8292,
+      "step": 886
+    },
+    {
+      "epoch": 0.7883568492389734,
+      "grad_norm": 0.37689974904060364,
+      "learning_rate": 0.0001,
+      "loss": 0.9543,
+      "step": 887
+    },
+    {
+      "epoch": 0.789245639373403,
+      "grad_norm": 0.4176059365272522,
+      "learning_rate": 0.0001,
+      "loss": 0.9437,
+      "step": 888
+    },
+    {
+      "epoch": 0.7901344295078324,
+      "grad_norm": 0.3841492235660553,
+      "learning_rate": 0.0001,
+      "loss": 1.0038,
+      "step": 889
+    },
+    {
+      "epoch": 0.791023219642262,
+      "grad_norm": 0.3431592285633087,
+      "learning_rate": 0.0001,
+      "loss": 0.9337,
+      "step": 890
+    },
+    {
+      "epoch": 0.7919120097766915,
+      "grad_norm": 0.4018877148628235,
+      "learning_rate": 0.0001,
+      "loss": 0.9502,
+      "step": 891
+    },
+    {
+      "epoch": 0.792800799911121,
+      "grad_norm": 0.4328785836696625,
+      "learning_rate": 0.0001,
+      "loss": 1.0362,
+      "step": 892
+    },
+    {
+      "epoch": 0.7936895900455505,
+      "grad_norm": 0.41478991508483887,
+      "learning_rate": 0.0001,
+      "loss": 0.9226,
+      "step": 893
+    },
+    {
+      "epoch": 0.79457838017998,
+      "grad_norm": 0.3804970979690552,
+      "learning_rate": 0.0001,
+      "loss": 0.9237,
+      "step": 894
+    },
+    {
+      "epoch": 0.7954671703144095,
+      "grad_norm": 0.41232219338417053,
+      "learning_rate": 0.0001,
+      "loss": 0.8931,
+      "step": 895
+    },
+    {
+      "epoch": 0.796355960448839,
+      "grad_norm": 0.3760983645915985,
+      "learning_rate": 0.0001,
+      "loss": 0.9839,
+      "step": 896
+    },
+    {
+      "epoch": 0.7972447505832685,
+      "grad_norm": 0.3917902410030365,
+      "learning_rate": 0.0001,
+      "loss": 0.8802,
+      "step": 897
+    },
+    {
+      "epoch": 0.798133540717698,
+      "grad_norm": 0.3652724623680115,
+      "learning_rate": 0.0001,
+      "loss": 0.9058,
+      "step": 898
+    },
+    {
+      "epoch": 0.7990223308521276,
+      "grad_norm": 0.3827455937862396,
+      "learning_rate": 0.0001,
+      "loss": 0.9821,
+      "step": 899
+    },
+    {
+      "epoch": 0.799911120986557,
+      "grad_norm": 0.4203377664089203,
+      "learning_rate": 0.0001,
+      "loss": 1.0054,
+      "step": 900
+    },
+    {
+      "epoch": 0.8007999111209866,
+      "grad_norm": 0.3544522821903229,
+      "learning_rate": 0.0001,
+      "loss": 1.0044,
+      "step": 901
+    },
+    {
+      "epoch": 0.801688701255416,
+      "grad_norm": 0.39687561988830566,
+      "learning_rate": 0.0001,
+      "loss": 0.9896,
+      "step": 902
+    },
+    {
+      "epoch": 0.8025774913898456,
+      "grad_norm": 0.3697212040424347,
+      "learning_rate": 0.0001,
+      "loss": 0.9282,
+      "step": 903
+    },
+    {
+      "epoch": 0.8034662815242751,
+      "grad_norm": 0.37844914197921753,
+      "learning_rate": 0.0001,
+      "loss": 0.9032,
+      "step": 904
+    },
+    {
+      "epoch": 0.8043550716587046,
+      "grad_norm": 0.36820754408836365,
+      "learning_rate": 0.0001,
+      "loss": 0.9314,
+      "step": 905
+    },
+    {
+      "epoch": 0.8052438617931341,
+      "grad_norm": 0.4033227264881134,
+      "learning_rate": 0.0001,
+      "loss": 0.9648,
+      "step": 906
+    },
+    {
+      "epoch": 0.8061326519275636,
+      "grad_norm": 0.3827173411846161,
+      "learning_rate": 0.0001,
+      "loss": 0.8304,
+      "step": 907
+    },
+    {
+      "epoch": 0.8070214420619931,
+      "grad_norm": 0.3617156147956848,
+      "learning_rate": 0.0001,
+      "loss": 0.8478,
+      "step": 908
+    },
+    {
+      "epoch": 0.8079102321964227,
+      "grad_norm": 0.38975533843040466,
+      "learning_rate": 0.0001,
+      "loss": 0.8794,
+      "step": 909
+    },
+    {
+      "epoch": 0.8087990223308521,
+      "grad_norm": 0.39395904541015625,
+      "learning_rate": 0.0001,
+      "loss": 0.8944,
+      "step": 910
+    },
+    {
+      "epoch": 0.8096878124652817,
+      "grad_norm": 0.31278157234191895,
+      "learning_rate": 0.0001,
+      "loss": 0.8736,
+      "step": 911
+    },
+    {
+      "epoch": 0.8105766025997111,
+      "grad_norm": 0.3955362141132355,
+      "learning_rate": 0.0001,
+      "loss": 1.012,
+      "step": 912
+    },
+    {
+      "epoch": 0.8114653927341406,
+      "grad_norm": 0.41375401616096497,
+      "learning_rate": 0.0001,
+      "loss": 0.9767,
+      "step": 913
+    },
+    {
+      "epoch": 0.8123541828685702,
+      "grad_norm": 0.38847875595092773,
+      "learning_rate": 0.0001,
+      "loss": 0.9193,
+      "step": 914
+    },
+    {
+      "epoch": 0.8132429730029996,
+      "grad_norm": 0.42238739132881165,
+      "learning_rate": 0.0001,
+      "loss": 1.0099,
+      "step": 915
+    },
+    {
+      "epoch": 0.8141317631374292,
+      "grad_norm": 0.3599422872066498,
+      "learning_rate": 0.0001,
+      "loss": 0.8548,
+      "step": 916
+    },
+    {
+      "epoch": 0.8150205532718586,
+      "grad_norm": 0.369069367647171,
+      "learning_rate": 0.0001,
+      "loss": 0.9032,
+      "step": 917
+    },
+    {
+      "epoch": 0.8159093434062882,
+      "grad_norm": 0.3789387047290802,
+      "learning_rate": 0.0001,
+      "loss": 0.9597,
+      "step": 918
+    },
+    {
+      "epoch": 0.8167981335407177,
+      "grad_norm": 0.37680602073669434,
+      "learning_rate": 0.0001,
+      "loss": 0.9284,
+      "step": 919
+    },
+    {
+      "epoch": 0.8176869236751472,
+      "grad_norm": 0.4254203140735626,
+      "learning_rate": 0.0001,
+      "loss": 0.9765,
+      "step": 920
+    },
+    {
+      "epoch": 0.8185757138095767,
+      "grad_norm": 0.3756076991558075,
+      "learning_rate": 0.0001,
+      "loss": 0.9469,
+      "step": 921
+    },
+    {
+      "epoch": 0.8194645039440063,
+      "grad_norm": 0.37767136096954346,
+      "learning_rate": 0.0001,
+      "loss": 0.926,
+      "step": 922
+    },
+    {
+      "epoch": 0.8203532940784357,
+      "grad_norm": 0.43672624230384827,
+      "learning_rate": 0.0001,
+      "loss": 0.9305,
+      "step": 923
+    },
+    {
+      "epoch": 0.8212420842128653,
+      "grad_norm": 0.36939603090286255,
+      "learning_rate": 0.0001,
+      "loss": 0.8828,
+      "step": 924
+    },
+    {
+      "epoch": 0.8221308743472947,
+      "grad_norm": 0.41821223497390747,
+      "learning_rate": 0.0001,
+      "loss": 0.9346,
+      "step": 925
+    },
+    {
+      "epoch": 0.8230196644817243,
+      "grad_norm": 0.3761398494243622,
+      "learning_rate": 0.0001,
+      "loss": 0.8826,
+      "step": 926
+    },
+    {
+      "epoch": 0.8239084546161538,
+      "grad_norm": 0.4117630124092102,
+      "learning_rate": 0.0001,
+      "loss": 0.9253,
+      "step": 927
+    },
+    {
+      "epoch": 0.8247972447505832,
+      "grad_norm": 0.31008240580558777,
+      "learning_rate": 0.0001,
+      "loss": 0.9083,
+      "step": 928
+    },
+    {
+      "epoch": 0.8256860348850128,
+      "grad_norm": 0.3615438938140869,
+      "learning_rate": 0.0001,
+      "loss": 0.8833,
+      "step": 929
+    },
+    {
+      "epoch": 0.8265748250194422,
+      "grad_norm": 0.3704020082950592,
+      "learning_rate": 0.0001,
+      "loss": 1.0443,
+      "step": 930
+    },
+    {
+      "epoch": 0.8274636151538718,
+      "grad_norm": 0.3424414098262787,
+      "learning_rate": 0.0001,
+      "loss": 0.959,
+      "step": 931
+    },
+    {
+      "epoch": 0.8283524052883013,
+      "grad_norm": 0.41668984293937683,
+      "learning_rate": 0.0001,
+      "loss": 1.0217,
+      "step": 932
+    },
+    {
+      "epoch": 0.8292411954227308,
+      "grad_norm": 0.35622361302375793,
+      "learning_rate": 0.0001,
+      "loss": 0.9546,
+      "step": 933
+    },
+    {
+      "epoch": 0.8301299855571603,
+      "grad_norm": 0.36572763323783875,
+      "learning_rate": 0.0001,
+      "loss": 0.9514,
+      "step": 934
+    },
+    {
+      "epoch": 0.8310187756915898,
+      "grad_norm": 0.46444347500801086,
+      "learning_rate": 0.0001,
+      "loss": 1.0018,
+      "step": 935
+    },
+    {
+      "epoch": 0.8319075658260193,
+      "grad_norm": 0.39167654514312744,
+      "learning_rate": 0.0001,
+      "loss": 0.9338,
+      "step": 936
+    },
+    {
+      "epoch": 0.8327963559604489,
+      "grad_norm": 0.4243658781051636,
+      "learning_rate": 0.0001,
+      "loss": 0.9801,
+      "step": 937
+    },
+    {
+      "epoch": 0.8336851460948783,
+      "grad_norm": 2.1183691024780273,
+      "learning_rate": 0.0001,
+      "loss": 0.9376,
+      "step": 938
+    },
+    {
+      "epoch": 0.8345739362293079,
+      "grad_norm": 0.4078412353992462,
+      "learning_rate": 0.0001,
+      "loss": 0.9533,
+      "step": 939
+    },
+    {
+      "epoch": 0.8354627263637374,
+      "grad_norm": 0.36636313796043396,
+      "learning_rate": 0.0001,
+      "loss": 0.993,
+      "step": 940
+    },
+    {
+      "epoch": 0.8363515164981669,
+      "grad_norm": 0.42190858721733093,
+      "learning_rate": 0.0001,
+      "loss": 0.9789,
+      "step": 941
+    },
+    {
+      "epoch": 0.8372403066325964,
+      "grad_norm": 0.34506893157958984,
+      "learning_rate": 0.0001,
+      "loss": 0.9759,
+      "step": 942
+    },
+    {
+      "epoch": 0.8381290967670258,
+      "grad_norm": 0.43273380398750305,
+      "learning_rate": 0.0001,
+      "loss": 0.9753,
+      "step": 943
+    },
+    {
+      "epoch": 0.8390178869014554,
+      "grad_norm": 0.36178067326545715,
+      "learning_rate": 0.0001,
+      "loss": 0.9842,
+      "step": 944
+    },
+    {
+      "epoch": 0.839906677035885,
+      "grad_norm": 0.3905907869338989,
+      "learning_rate": 0.0001,
+      "loss": 0.9809,
+      "step": 945
+    },
+    {
+      "epoch": 0.8407954671703144,
+      "grad_norm": 0.3678842782974243,
+      "learning_rate": 0.0001,
+      "loss": 0.939,
+      "step": 946
+    },
+    {
+      "epoch": 0.8416842573047439,
+      "grad_norm": 0.3797389268875122,
+      "learning_rate": 0.0001,
+      "loss": 0.8508,
+      "step": 947
+    },
+    {
+      "epoch": 0.8425730474391734,
+      "grad_norm": 0.43192946910858154,
+      "learning_rate": 0.0001,
+      "loss": 0.9417,
+      "step": 948
+    },
+    {
+      "epoch": 0.8434618375736029,
+      "grad_norm": 0.4288097620010376,
+      "learning_rate": 0.0001,
+      "loss": 0.9428,
+      "step": 949
+    },
+    {
+      "epoch": 0.8443506277080325,
+      "grad_norm": 0.3306112587451935,
+      "learning_rate": 0.0001,
+      "loss": 0.9606,
+      "step": 950
+    },
+    {
+      "epoch": 0.8452394178424619,
+      "grad_norm": 0.3389092683792114,
+      "learning_rate": 0.0001,
+      "loss": 0.8848,
+      "step": 951
+    },
+    {
+      "epoch": 0.8461282079768915,
+      "grad_norm": 0.38675200939178467,
+      "learning_rate": 0.0001,
+      "loss": 0.9428,
+      "step": 952
+    },
+    {
+      "epoch": 0.8470169981113209,
+      "grad_norm": 0.3907453417778015,
+      "learning_rate": 0.0001,
+      "loss": 0.9628,
+      "step": 953
+    },
+    {
+      "epoch": 0.8479057882457505,
+      "grad_norm": 0.3519386649131775,
+      "learning_rate": 0.0001,
+      "loss": 0.9262,
+      "step": 954
+    },
+    {
+      "epoch": 0.84879457838018,
+      "grad_norm": 0.38831275701522827,
+      "learning_rate": 0.0001,
+      "loss": 0.904,
+      "step": 955
+    },
+    {
+      "epoch": 0.8496833685146095,
+      "grad_norm": 0.39892178773880005,
+      "learning_rate": 0.0001,
+      "loss": 0.9897,
+      "step": 956
+    },
+    {
+      "epoch": 0.850572158649039,
+      "grad_norm": 0.4237317144870758,
+      "learning_rate": 0.0001,
+      "loss": 0.9679,
+      "step": 957
+    },
+    {
+      "epoch": 0.8514609487834685,
+      "grad_norm": 0.35414162278175354,
+      "learning_rate": 0.0001,
+      "loss": 0.9362,
+      "step": 958
+    },
+    {
+      "epoch": 0.852349738917898,
+      "grad_norm": 0.34299251437187195,
+      "learning_rate": 0.0001,
+      "loss": 0.9223,
+      "step": 959
+    },
+    {
+      "epoch": 0.8532385290523276,
+      "grad_norm": 0.3854987621307373,
+      "learning_rate": 0.0001,
+      "loss": 0.9285,
+      "step": 960
+    },
+    {
+      "epoch": 0.854127319186757,
+      "grad_norm": 0.3863328993320465,
+      "learning_rate": 0.0001,
+      "loss": 0.9039,
+      "step": 961
+    },
+    {
+      "epoch": 0.8550161093211865,
+      "grad_norm": 0.3316647708415985,
+      "learning_rate": 0.0001,
+      "loss": 0.9445,
+      "step": 962
+    },
+    {
+      "epoch": 0.8559048994556161,
+      "grad_norm": 0.39452093839645386,
+      "learning_rate": 0.0001,
+      "loss": 0.8977,
+      "step": 963
+    },
+    {
+      "epoch": 0.8567936895900455,
+      "grad_norm": 0.3937751352787018,
+      "learning_rate": 0.0001,
+      "loss": 0.914,
+      "step": 964
+    },
+    {
+      "epoch": 0.8576824797244751,
+      "grad_norm": 0.31849080324172974,
+      "learning_rate": 0.0001,
+      "loss": 0.8676,
+      "step": 965
+    },
+    {
+      "epoch": 0.8585712698589045,
+      "grad_norm": 0.3330874741077423,
+      "learning_rate": 0.0001,
+      "loss": 0.8927,
+      "step": 966
+    },
+    {
+      "epoch": 0.8594600599933341,
+      "grad_norm": 0.3816487789154053,
+      "learning_rate": 0.0001,
+      "loss": 0.9715,
+      "step": 967
+    },
+    {
+      "epoch": 0.8603488501277636,
+      "grad_norm": 0.4189969301223755,
+      "learning_rate": 0.0001,
+      "loss": 0.9714,
+      "step": 968
+    },
+    {
+      "epoch": 0.8612376402621931,
+      "grad_norm": 0.3965904712677002,
+      "learning_rate": 0.0001,
+      "loss": 0.9532,
+      "step": 969
+    },
+    {
+      "epoch": 0.8621264303966226,
+      "grad_norm": 0.3900628089904785,
+      "learning_rate": 0.0001,
+      "loss": 0.9577,
+      "step": 970
+    },
+    {
+      "epoch": 0.8630152205310521,
+      "grad_norm": 0.43104588985443115,
+      "learning_rate": 0.0001,
+      "loss": 1.0495,
+      "step": 971
+    },
+    {
+      "epoch": 0.8639040106654816,
+      "grad_norm": 0.39500030875205994,
+      "learning_rate": 0.0001,
+      "loss": 0.9637,
+      "step": 972
+    },
+    {
+      "epoch": 0.8647928007999112,
+      "grad_norm": 0.3971693813800812,
+      "learning_rate": 0.0001,
+      "loss": 1.0178,
+      "step": 973
+    },
+    {
+      "epoch": 0.8656815909343406,
+      "grad_norm": 0.3862766921520233,
+      "learning_rate": 0.0001,
+      "loss": 1.0042,
+      "step": 974
+    },
+    {
+      "epoch": 0.8665703810687702,
+      "grad_norm": 0.3328218162059784,
+      "learning_rate": 0.0001,
+      "loss": 0.9846,
+      "step": 975
+    },
+    {
+      "epoch": 0.8674591712031996,
+      "grad_norm": 0.36906227469444275,
+      "learning_rate": 0.0001,
+      "loss": 0.9593,
+      "step": 976
+    },
+    {
+      "epoch": 0.8683479613376291,
+      "grad_norm": 0.40187719464302063,
+      "learning_rate": 0.0001,
+      "loss": 0.97,
+      "step": 977
+    },
+    {
+      "epoch": 0.8692367514720587,
+      "grad_norm": 0.3088963031768799,
+      "learning_rate": 0.0001,
+      "loss": 0.8898,
+      "step": 978
+    },
+    {
+      "epoch": 0.8701255416064881,
+      "grad_norm": 0.4896550178527832,
+      "learning_rate": 0.0001,
+      "loss": 0.9039,
+      "step": 979
+    },
+    {
+      "epoch": 0.8710143317409177,
+      "grad_norm": 0.39287492632865906,
+      "learning_rate": 0.0001,
+      "loss": 0.8972,
+      "step": 980
+    },
+    {
+      "epoch": 0.8719031218753471,
+      "grad_norm": 0.3530255854129791,
+      "learning_rate": 0.0001,
+      "loss": 0.949,
+      "step": 981
+    },
+    {
+      "epoch": 0.8727919120097767,
+      "grad_norm": 0.4530492126941681,
+      "learning_rate": 0.0001,
+      "loss": 1.0146,
+      "step": 982
+    },
+    {
+      "epoch": 0.8736807021442062,
+      "grad_norm": 0.3938184082508087,
+      "learning_rate": 0.0001,
+      "loss": 0.9926,
+      "step": 983
+    },
+    {
+      "epoch": 0.8745694922786357,
+      "grad_norm": 0.41301020979881287,
+      "learning_rate": 0.0001,
+      "loss": 0.9626,
+      "step": 984
+    },
+    {
+      "epoch": 0.8754582824130652,
+      "grad_norm": 0.5657033920288086,
+      "learning_rate": 0.0001,
+      "loss": 0.9557,
+      "step": 985
+    },
+    {
+      "epoch": 0.8763470725474948,
+      "grad_norm": 0.39124661684036255,
+      "learning_rate": 0.0001,
+      "loss": 0.9334,
+      "step": 986
+    },
+    {
+      "epoch": 0.8772358626819242,
+      "grad_norm": 0.367252916097641,
+      "learning_rate": 0.0001,
+      "loss": 0.9568,
+      "step": 987
+    },
+    {
+      "epoch": 0.8781246528163538,
+      "grad_norm": 0.45111268758773804,
+      "learning_rate": 0.0001,
+      "loss": 0.9269,
+      "step": 988
+    },
+    {
+      "epoch": 0.8790134429507832,
+      "grad_norm": 0.3949245810508728,
+      "learning_rate": 0.0001,
+      "loss": 0.9541,
+      "step": 989
+    },
+    {
+      "epoch": 0.8799022330852128,
+      "grad_norm": 0.3697701096534729,
+      "learning_rate": 0.0001,
+      "loss": 0.858,
+      "step": 990
+    },
+    {
+      "epoch": 0.8807910232196423,
+      "grad_norm": 0.3903849720954895,
+      "learning_rate": 0.0001,
+      "loss": 0.8917,
+      "step": 991
+    },
+    {
+      "epoch": 0.8816798133540718,
+      "grad_norm": 0.3852884769439697,
+      "learning_rate": 0.0001,
+      "loss": 0.9524,
+      "step": 992
+    },
+    {
+      "epoch": 0.8825686034885013,
+      "grad_norm": 0.425076425075531,
+      "learning_rate": 0.0001,
+      "loss": 0.9229,
+      "step": 993
+    },
+    {
+      "epoch": 0.8834573936229307,
+      "grad_norm": 0.3908687233924866,
+      "learning_rate": 0.0001,
+      "loss": 0.8766,
+      "step": 994
+    },
+    {
+      "epoch": 0.8843461837573603,
+      "grad_norm": 0.3858028054237366,
+      "learning_rate": 0.0001,
+      "loss": 0.9996,
+      "step": 995
+    },
+    {
+      "epoch": 0.8852349738917898,
+      "grad_norm": 0.4030248820781708,
+      "learning_rate": 0.0001,
+      "loss": 1.0158,
+      "step": 996
+    },
+    {
+      "epoch": 0.8861237640262193,
+      "grad_norm": 0.3604998290538788,
+      "learning_rate": 0.0001,
+      "loss": 0.8926,
+      "step": 997
+    },
+    {
+      "epoch": 0.8870125541606488,
+      "grad_norm": 0.38563668727874756,
+      "learning_rate": 0.0001,
+      "loss": 0.8513,
+      "step": 998
+    },
+    {
+      "epoch": 0.8879013442950783,
+      "grad_norm": 0.40275925397872925,
+      "learning_rate": 0.0001,
+      "loss": 1.0191,
+      "step": 999
+    },
+    {
+      "epoch": 0.8887901344295078,
+      "grad_norm": 0.3867271840572357,
+      "learning_rate": 0.0001,
+      "loss": 0.8962,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8896789245639374,
+      "grad_norm": 0.39497697353363037,
+      "learning_rate": 0.0001,
+      "loss": 0.983,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8905677146983668,
+      "grad_norm": 0.3594026267528534,
+      "learning_rate": 0.0001,
+      "loss": 0.9519,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8914565048327964,
+      "grad_norm": 0.3522585332393646,
+      "learning_rate": 0.0001,
+      "loss": 0.9274,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8923452949672258,
+      "grad_norm": 0.41415807604789734,
+      "learning_rate": 0.0001,
+      "loss": 0.9674,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8932340851016554,
+      "grad_norm": 0.46825456619262695,
+      "learning_rate": 0.0001,
+      "loss": 0.9694,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8941228752360849,
+      "grad_norm": 0.3260214328765869,
+      "learning_rate": 0.0001,
+      "loss": 0.9656,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8950116653705144,
+      "grad_norm": 0.36279264092445374,
+      "learning_rate": 0.0001,
+      "loss": 0.9663,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8959004555049439,
+      "grad_norm": 0.4032383859157562,
+      "learning_rate": 0.0001,
+      "loss": 0.9276,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8967892456393735,
+      "grad_norm": 0.4371550679206848,
+      "learning_rate": 0.0001,
+      "loss": 1.0294,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8976780357738029,
+      "grad_norm": 0.36339619755744934,
+      "learning_rate": 0.0001,
+      "loss": 0.9453,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8985668259082324,
+      "grad_norm": 0.38631105422973633,
+      "learning_rate": 0.0001,
+      "loss": 0.9966,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8994556160426619,
+      "grad_norm": 0.34750688076019287,
+      "learning_rate": 0.0001,
+      "loss": 0.8757,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9003444061770914,
+      "grad_norm": 0.3377476632595062,
+      "learning_rate": 0.0001,
+      "loss": 0.973,
+      "step": 1013
+    },
+    {
+      "epoch": 0.901233196311521,
+      "grad_norm": 0.39975231885910034,
+      "learning_rate": 0.0001,
+      "loss": 0.9392,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9021219864459504,
+      "grad_norm": 0.33771857619285583,
+      "learning_rate": 0.0001,
+      "loss": 0.9934,
+      "step": 1015
+    },
+    {
+      "epoch": 0.90301077658038,
+      "grad_norm": 0.33681410551071167,
+      "learning_rate": 0.0001,
+      "loss": 0.9024,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9038995667148094,
+      "grad_norm": 0.39895641803741455,
+      "learning_rate": 0.0001,
+      "loss": 0.9482,
+      "step": 1017
+    },
+    {
+      "epoch": 0.904788356849239,
+      "grad_norm": 0.3800975978374481,
+      "learning_rate": 0.0001,
+      "loss": 0.9504,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9056771469836685,
+      "grad_norm": 0.3816840350627899,
+      "learning_rate": 0.0001,
+      "loss": 0.9613,
+      "step": 1019
+    },
+    {
+      "epoch": 0.906565937118098,
+      "grad_norm": 0.3548993766307831,
+      "learning_rate": 0.0001,
+      "loss": 0.9234,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9074547272525275,
+      "grad_norm": 0.38023990392684937,
+      "learning_rate": 0.0001,
+      "loss": 0.9697,
+      "step": 1021
+    },
+    {
+      "epoch": 0.908343517386957,
+      "grad_norm": 0.35391300916671753,
+      "learning_rate": 0.0001,
+      "loss": 0.8196,
+      "step": 1022
+    },
+    {
+      "epoch": 0.9092323075213865,
+      "grad_norm": 0.4617893397808075,
+      "learning_rate": 0.0001,
+      "loss": 1.0236,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9101210976558161,
+      "grad_norm": 0.3808736503124237,
+      "learning_rate": 0.0001,
+      "loss": 0.9539,
+      "step": 1024
+    },
+    {
+      "epoch": 0.9110098877902455,
+      "grad_norm": 0.41231459379196167,
+      "learning_rate": 0.0001,
+      "loss": 0.9251,
+      "step": 1025
+    },
+    {
+      "epoch": 0.911898677924675,
+      "grad_norm": 0.34969326853752136,
+      "learning_rate": 0.0001,
+      "loss": 0.9151,
+      "step": 1026
+    },
+    {
+      "epoch": 0.9127874680591045,
+      "grad_norm": 0.3669031262397766,
+      "learning_rate": 0.0001,
+      "loss": 0.8832,
+      "step": 1027
+    },
+    {
+      "epoch": 0.913676258193534,
+      "grad_norm": 0.3683227598667145,
+      "learning_rate": 0.0001,
+      "loss": 0.9324,
+      "step": 1028
+    },
+    {
+      "epoch": 0.9145650483279636,
+      "grad_norm": 0.3768469989299774,
+      "learning_rate": 0.0001,
+      "loss": 0.8713,
+      "step": 1029
+    },
+    {
+      "epoch": 0.915453838462393,
+      "grad_norm": 0.40044888854026794,
+      "learning_rate": 0.0001,
+      "loss": 1.0428,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9163426285968226,
+      "grad_norm": 0.3713497519493103,
+      "learning_rate": 0.0001,
+      "loss": 0.8803,
+      "step": 1031
+    },
+    {
+      "epoch": 0.9172314187312521,
+      "grad_norm": 0.33332690596580505,
+      "learning_rate": 0.0001,
+      "loss": 0.9579,
+      "step": 1032
+    },
+    {
+      "epoch": 0.9181202088656816,
+      "grad_norm": 0.3899736702442169,
+      "learning_rate": 0.0001,
+      "loss": 0.9934,
+      "step": 1033
+    },
+    {
+      "epoch": 0.9190089990001111,
+      "grad_norm": 0.3549093008041382,
+      "learning_rate": 0.0001,
+      "loss": 0.9093,
+      "step": 1034
+    },
+    {
+      "epoch": 0.9198977891345406,
+      "grad_norm": 0.37367865443229675,
+      "learning_rate": 0.0001,
+      "loss": 0.9748,
+      "step": 1035
+    },
+    {
+      "epoch": 0.9207865792689701,
+      "grad_norm": 0.3680754601955414,
+      "learning_rate": 0.0001,
+      "loss": 1.0113,
+      "step": 1036
+    },
+    {
+      "epoch": 0.9216753694033997,
+      "grad_norm": 0.36346691846847534,
+      "learning_rate": 0.0001,
+      "loss": 0.8892,
+      "step": 1037
+    },
+    {
+      "epoch": 0.9225641595378291,
+      "grad_norm": 0.3745006322860718,
+      "learning_rate": 0.0001,
+      "loss": 0.9352,
+      "step": 1038
+    },
+    {
+      "epoch": 0.9234529496722587,
+      "grad_norm": 0.3873980641365051,
+      "learning_rate": 0.0001,
+      "loss": 0.8809,
+      "step": 1039
+    },
+    {
+      "epoch": 0.9243417398066881,
+      "grad_norm": 0.42303210496902466,
+      "learning_rate": 0.0001,
+      "loss": 1.0008,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9252305299411177,
+      "grad_norm": 0.38883471488952637,
+      "learning_rate": 0.0001,
+      "loss": 0.9839,
+      "step": 1041
+    },
+    {
+      "epoch": 0.9261193200755472,
+      "grad_norm": 0.39919066429138184,
+      "learning_rate": 0.0001,
+      "loss": 0.9246,
+      "step": 1042
+    },
+    {
+      "epoch": 0.9270081102099766,
+      "grad_norm": 0.3721344769001007,
+      "learning_rate": 0.0001,
+      "loss": 0.8927,
+      "step": 1043
+    },
+    {
+      "epoch": 0.9278969003444062,
+      "grad_norm": 0.4106079638004303,
+      "learning_rate": 0.0001,
+      "loss": 0.9501,
+      "step": 1044
+    },
+    {
+      "epoch": 0.9287856904788356,
+      "grad_norm": 0.35994967818260193,
+      "learning_rate": 0.0001,
+      "loss": 0.9483,
+      "step": 1045
+    },
+    {
+      "epoch": 0.9296744806132652,
+      "grad_norm": 0.3131955862045288,
+      "learning_rate": 0.0001,
+      "loss": 0.9175,
+      "step": 1046
+    },
+    {
+      "epoch": 0.9305632707476947,
+      "grad_norm": 0.37486889958381653,
+      "learning_rate": 0.0001,
+      "loss": 0.9459,
+      "step": 1047
+    },
+    {
+      "epoch": 0.9314520608821242,
+      "grad_norm": 0.36107298731803894,
+      "learning_rate": 0.0001,
+      "loss": 0.9201,
+      "step": 1048
+    },
+    {
+      "epoch": 0.9323408510165537,
+      "grad_norm": 0.3575218915939331,
+      "learning_rate": 0.0001,
+      "loss": 0.9254,
+      "step": 1049
+    },
+    {
+      "epoch": 0.9332296411509832,
+      "grad_norm": 0.36310702562332153,
+      "learning_rate": 0.0001,
+      "loss": 0.9542,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9341184312854127,
+      "grad_norm": 0.3741386830806732,
+      "learning_rate": 0.0001,
+      "loss": 0.8913,
+      "step": 1051
+    },
+    {
+      "epoch": 0.9350072214198423,
+      "grad_norm": 0.3636328876018524,
+      "learning_rate": 0.0001,
+      "loss": 0.9218,
+      "step": 1052
+    },
+    {
+      "epoch": 0.9358960115542717,
+      "grad_norm": 0.3283315598964691,
+      "learning_rate": 0.0001,
+      "loss": 0.9341,
+      "step": 1053
+    },
+    {
+      "epoch": 0.9367848016887013,
+      "grad_norm": 0.4130898118019104,
+      "learning_rate": 0.0001,
+      "loss": 0.9666,
+      "step": 1054
+    },
+    {
+      "epoch": 0.9376735918231308,
+      "grad_norm": 0.3228664696216583,
+      "learning_rate": 0.0001,
+      "loss": 0.8924,
+      "step": 1055
+    },
+    {
+      "epoch": 0.9385623819575603,
+      "grad_norm": 0.42613714933395386,
+      "learning_rate": 0.0001,
+      "loss": 0.9347,
+      "step": 1056
+    },
+    {
+      "epoch": 0.9394511720919898,
+      "grad_norm": 0.4059881567955017,
+      "learning_rate": 0.0001,
+      "loss": 0.906,
+      "step": 1057
+    },
+    {
+      "epoch": 0.9403399622264192,
+      "grad_norm": 0.4185310900211334,
+      "learning_rate": 0.0001,
+      "loss": 1.0162,
+      "step": 1058
+    },
+    {
+      "epoch": 0.9412287523608488,
+      "grad_norm": 0.38086622953414917,
+      "learning_rate": 0.0001,
+      "loss": 1.0297,
+      "step": 1059
+    },
+    {
+      "epoch": 0.9421175424952783,
+      "grad_norm": 0.3949964642524719,
+      "learning_rate": 0.0001,
+      "loss": 0.9184,
+      "step": 1060
+    },
+    {
+      "epoch": 0.9430063326297078,
+      "grad_norm": 0.3886110186576843,
+      "learning_rate": 0.0001,
+      "loss": 0.9339,
+      "step": 1061
+    },
+    {
+      "epoch": 0.9438951227641373,
+      "grad_norm": 0.37949997186660767,
+      "learning_rate": 0.0001,
+      "loss": 0.9141,
+      "step": 1062
+    },
+    {
+      "epoch": 0.9447839128985668,
+      "grad_norm": 0.34117576479911804,
+      "learning_rate": 0.0001,
+      "loss": 0.9153,
+      "step": 1063
+    },
+    {
+      "epoch": 0.9456727030329963,
+      "grad_norm": 0.36747774481773376,
+      "learning_rate": 0.0001,
+      "loss": 0.9275,
+      "step": 1064
+    },
+    {
+      "epoch": 0.9465614931674259,
+      "grad_norm": 0.3943864703178406,
+      "learning_rate": 0.0001,
+      "loss": 0.8985,
+      "step": 1065
+    },
+    {
+      "epoch": 0.9474502833018553,
+      "grad_norm": 0.3849271237850189,
+      "learning_rate": 0.0001,
+      "loss": 0.956,
+      "step": 1066
+    },
+    {
+      "epoch": 0.9483390734362849,
+      "grad_norm": 0.3633384108543396,
+      "learning_rate": 0.0001,
+      "loss": 0.9836,
+      "step": 1067
+    },
+    {
+      "epoch": 0.9492278635707143,
+      "grad_norm": 0.3905209004878998,
+      "learning_rate": 0.0001,
+      "loss": 0.9753,
+      "step": 1068
+    },
+    {
+      "epoch": 0.9501166537051439,
+      "grad_norm": 0.36972206830978394,
+      "learning_rate": 0.0001,
+      "loss": 0.9108,
+      "step": 1069
+    },
+    {
+      "epoch": 0.9510054438395734,
+      "grad_norm": 0.38415467739105225,
+      "learning_rate": 0.0001,
+      "loss": 0.9539,
+      "step": 1070
+    },
+    {
+      "epoch": 0.9518942339740029,
+      "grad_norm": 0.3597318232059479,
+      "learning_rate": 0.0001,
+      "loss": 0.8953,
+      "step": 1071
+    },
+    {
+      "epoch": 0.9527830241084324,
+      "grad_norm": 0.374531090259552,
+      "learning_rate": 0.0001,
+      "loss": 0.981,
+      "step": 1072
+    },
+    {
+      "epoch": 0.953671814242862,
+      "grad_norm": 0.3501724302768707,
+      "learning_rate": 0.0001,
+      "loss": 0.9589,
+      "step": 1073
+    },
+    {
+      "epoch": 0.9545606043772914,
+      "grad_norm": 0.3414580821990967,
+      "learning_rate": 0.0001,
+      "loss": 0.901,
+      "step": 1074
+    },
+    {
+      "epoch": 0.955449394511721,
+      "grad_norm": 0.3523072600364685,
+      "learning_rate": 0.0001,
+      "loss": 0.901,
+      "step": 1075
+    },
+    {
+      "epoch": 0.9563381846461504,
+      "grad_norm": 0.38694173097610474,
+      "learning_rate": 0.0001,
+      "loss": 0.9884,
+      "step": 1076
+    },
+    {
+      "epoch": 0.9572269747805799,
+      "grad_norm": 0.41436851024627686,
+      "learning_rate": 0.0001,
+      "loss": 0.9123,
+      "step": 1077
+    },
+    {
+      "epoch": 0.9581157649150095,
+      "grad_norm": 0.34676891565322876,
+      "learning_rate": 0.0001,
+      "loss": 0.9113,
+      "step": 1078
+    },
+    {
+      "epoch": 0.9590045550494389,
+      "grad_norm": 0.3826766908168793,
+      "learning_rate": 0.0001,
+      "loss": 0.9213,
+      "step": 1079
+    },
+    {
+      "epoch": 0.9598933451838685,
+      "grad_norm": 0.34997478127479553,
+      "learning_rate": 0.0001,
+      "loss": 0.9178,
+      "step": 1080
+    },
+    {
+      "epoch": 0.9607821353182979,
+      "grad_norm": 0.40335193276405334,
+      "learning_rate": 0.0001,
+      "loss": 1.0525,
+      "step": 1081
+    },
+    {
+      "epoch": 0.9616709254527275,
+      "grad_norm": 0.4107684791088104,
+      "learning_rate": 0.0001,
+      "loss": 1.0239,
+      "step": 1082
+    },
+    {
+      "epoch": 0.962559715587157,
+      "grad_norm": 0.3917597830295563,
+      "learning_rate": 0.0001,
+      "loss": 0.8393,
+      "step": 1083
+    },
+    {
+      "epoch": 0.9634485057215865,
+      "grad_norm": 0.3460274040699005,
+      "learning_rate": 0.0001,
+      "loss": 0.9266,
+      "step": 1084
+    },
+    {
+      "epoch": 0.964337295856016,
+      "grad_norm": 0.36212509870529175,
+      "learning_rate": 0.0001,
+      "loss": 0.9578,
+      "step": 1085
+    },
+    {
+      "epoch": 0.9652260859904455,
+      "grad_norm": 0.4016794264316559,
+      "learning_rate": 0.0001,
+      "loss": 0.9461,
+      "step": 1086
+    },
+    {
+      "epoch": 0.966114876124875,
+      "grad_norm": 0.3891691267490387,
+      "learning_rate": 0.0001,
+      "loss": 0.9576,
+      "step": 1087
+    },
+    {
+      "epoch": 0.9670036662593046,
+      "grad_norm": 0.35440635681152344,
+      "learning_rate": 0.0001,
+      "loss": 0.9546,
+      "step": 1088
+    },
+    {
+      "epoch": 0.967892456393734,
+      "grad_norm": 0.3960229158401489,
+      "learning_rate": 0.0001,
+      "loss": 1.0514,
+      "step": 1089
+    },
+    {
+      "epoch": 0.9687812465281636,
+      "grad_norm": 0.38155800104141235,
+      "learning_rate": 0.0001,
+      "loss": 0.8884,
+      "step": 1090
+    },
+    {
+      "epoch": 0.969670036662593,
+      "grad_norm": 0.3513757884502411,
+      "learning_rate": 0.0001,
+      "loss": 0.8801,
+      "step": 1091
+    },
+    {
+      "epoch": 0.9705588267970225,
+      "grad_norm": 0.38408926129341125,
+      "learning_rate": 0.0001,
+      "loss": 0.8309,
+      "step": 1092
+    },
+    {
+      "epoch": 0.9714476169314521,
+      "grad_norm": 0.378099262714386,
+      "learning_rate": 0.0001,
+      "loss": 0.9089,
+      "step": 1093
+    },
+    {
+      "epoch": 0.9723364070658815,
+      "grad_norm": 0.3182670474052429,
+      "learning_rate": 0.0001,
+      "loss": 0.9489,
+      "step": 1094
+    },
+    {
+      "epoch": 0.9732251972003111,
+      "grad_norm": 0.3988575041294098,
+      "learning_rate": 0.0001,
+      "loss": 0.9364,
+      "step": 1095
+    },
+    {
+      "epoch": 0.9741139873347406,
+      "grad_norm": 0.4182455837726593,
+      "learning_rate": 0.0001,
+      "loss": 1.0097,
+      "step": 1096
+    },
+    {
+      "epoch": 0.9750027774691701,
+      "grad_norm": 0.3374396562576294,
+      "learning_rate": 0.0001,
+      "loss": 0.9467,
+      "step": 1097
+    },
+    {
+      "epoch": 0.9758915676035996,
+      "grad_norm": 0.36527591943740845,
+      "learning_rate": 0.0001,
+      "loss": 0.9365,
+      "step": 1098
+    },
+    {
+      "epoch": 0.9767803577380291,
+      "grad_norm": 0.3679245710372925,
+      "learning_rate": 0.0001,
+      "loss": 0.9387,
+      "step": 1099
+    },
+    {
+      "epoch": 0.9776691478724586,
+      "grad_norm": 0.3694630265235901,
+      "learning_rate": 0.0001,
+      "loss": 0.8805,
+      "step": 1100
+    },
+    {
+      "epoch": 0.9785579380068882,
+      "grad_norm": 0.3750841021537781,
+      "learning_rate": 0.0001,
+      "loss": 0.9024,
+      "step": 1101
+    },
+    {
+      "epoch": 0.9794467281413176,
+      "grad_norm": 0.4152679145336151,
+      "learning_rate": 0.0001,
+      "loss": 1.0473,
+      "step": 1102
+    },
+    {
+      "epoch": 0.9803355182757472,
+      "grad_norm": 0.3526078164577484,
+      "learning_rate": 0.0001,
+      "loss": 0.9227,
+      "step": 1103
+    },
+    {
+      "epoch": 0.9812243084101766,
+      "grad_norm": 0.3585357069969177,
+      "learning_rate": 0.0001,
+      "loss": 0.961,
+      "step": 1104
+    },
+    {
+      "epoch": 0.9821130985446062,
+      "grad_norm": 0.38510292768478394,
+      "learning_rate": 0.0001,
+      "loss": 0.9997,
+      "step": 1105
+    },
+    {
+      "epoch": 0.9830018886790357,
+      "grad_norm": 0.34511297941207886,
+      "learning_rate": 0.0001,
+      "loss": 0.895,
+      "step": 1106
+    },
+    {
+      "epoch": 0.9838906788134651,
+      "grad_norm": 0.36190247535705566,
+      "learning_rate": 0.0001,
+      "loss": 0.9789,
+      "step": 1107
+    },
+    {
+      "epoch": 0.9847794689478947,
+      "grad_norm": 0.3606034219264984,
+      "learning_rate": 0.0001,
+      "loss": 0.9512,
+      "step": 1108
+    },
+    {
+      "epoch": 0.9856682590823241,
+      "grad_norm": 0.31674253940582275,
+      "learning_rate": 0.0001,
+      "loss": 0.9767,
+      "step": 1109
+    },
+    {
+      "epoch": 0.9865570492167537,
+      "grad_norm": 0.3847641944885254,
+      "learning_rate": 0.0001,
+      "loss": 0.9811,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9874458393511832,
+      "grad_norm": 0.28332382440567017,
+      "learning_rate": 0.0001,
+      "loss": 0.9064,
+      "step": 1111
+    },
+    {
+      "epoch": 0.9883346294856127,
+      "grad_norm": 0.3163788318634033,
+      "learning_rate": 0.0001,
+      "loss": 1.0134,
+      "step": 1112
+    },
+    {
+      "epoch": 0.9892234196200422,
+      "grad_norm": 0.37489238381385803,
+      "learning_rate": 0.0001,
+      "loss": 0.9254,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9901122097544717,
+      "grad_norm": 0.3543959856033325,
+      "learning_rate": 0.0001,
+      "loss": 0.9415,
+      "step": 1114
+    },
+    {
+      "epoch": 0.9910009998889012,
+      "grad_norm": 0.42274850606918335,
+      "learning_rate": 0.0001,
+      "loss": 0.9142,
+      "step": 1115
+    },
+    {
+      "epoch": 0.9918897900233308,
+      "grad_norm": 0.38619187474250793,
+      "learning_rate": 0.0001,
+      "loss": 0.9387,
+      "step": 1116
+    },
+    {
+      "epoch": 0.9927785801577602,
+      "grad_norm": 0.379362553358078,
+      "learning_rate": 0.0001,
+      "loss": 0.9879,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9936673702921898,
+      "grad_norm": 0.4227985441684723,
+      "learning_rate": 0.0001,
+      "loss": 1.0083,
+      "step": 1118
+    },
+    {
+      "epoch": 0.9945561604266193,
+      "grad_norm": 0.4064882695674896,
+      "learning_rate": 0.0001,
+      "loss": 0.9726,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9954449505610488,
+      "grad_norm": 0.35058096051216125,
+      "learning_rate": 0.0001,
+      "loss": 0.9862,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9963337406954783,
+      "grad_norm": 0.3268047869205475,
+      "learning_rate": 0.0001,
+      "loss": 0.8753,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9972225308299077,
+      "grad_norm": 0.3554556667804718,
+      "learning_rate": 0.0001,
+      "loss": 1.0007,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9981113209643373,
+      "grad_norm": 0.368964284658432,
+      "learning_rate": 0.0001,
+      "loss": 0.9934,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9990001110987669,
+      "grad_norm": 0.39982903003692627,
+      "learning_rate": 0.0001,
+      "loss": 0.8933,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9998889012331963,
+      "grad_norm": 0.3517761528491974,
+      "learning_rate": 0.0001,
+      "loss": 0.9454,
+      "step": 1125
+    },
+    {
+      "epoch": 1.0007776913676258,
+      "grad_norm": 0.3954550623893738,
+      "learning_rate": 0.0001,
+      "loss": 0.9681,
+      "step": 1126
+    },
+    {
+      "epoch": 1.0016664815020553,
+      "grad_norm": 0.377174973487854,
+      "learning_rate": 0.0001,
+      "loss": 0.8917,
+      "step": 1127
+    },
+    {
+      "epoch": 1.002555271636485,
+      "grad_norm": 0.4547029435634613,
+      "learning_rate": 0.0001,
+      "loss": 0.8837,
+      "step": 1128
+    },
+    {
+      "epoch": 1.0034440617709144,
+      "grad_norm": 0.4161812961101532,
+      "learning_rate": 0.0001,
+      "loss": 1.0238,
+      "step": 1129
+    },
+    {
+      "epoch": 1.0043328519053438,
+      "grad_norm": 0.398183673620224,
+      "learning_rate": 0.0001,
+      "loss": 0.9245,
+      "step": 1130
+    },
+    {
+      "epoch": 1.0052216420397733,
+      "grad_norm": 0.41381561756134033,
+      "learning_rate": 0.0001,
+      "loss": 0.909,
+      "step": 1131
+    },
+    {
+      "epoch": 1.006110432174203,
+      "grad_norm": 0.37161487340927124,
+      "learning_rate": 0.0001,
+      "loss": 0.9355,
+      "step": 1132
+    },
+    {
+      "epoch": 1.0069992223086324,
+      "grad_norm": 0.378110408782959,
+      "learning_rate": 0.0001,
+      "loss": 0.973,
+      "step": 1133
+    },
+    {
+      "epoch": 1.0078880124430618,
+      "grad_norm": 0.3747970163822174,
+      "learning_rate": 0.0001,
+      "loss": 0.8684,
+      "step": 1134
+    },
+    {
+      "epoch": 1.0087768025774915,
+      "grad_norm": 0.3934114873409271,
+      "learning_rate": 0.0001,
+      "loss": 0.9361,
+      "step": 1135
+    },
+    {
+      "epoch": 1.009665592711921,
+      "grad_norm": 0.41165485978126526,
+      "learning_rate": 0.0001,
+      "loss": 0.9385,
+      "step": 1136
+    },
+    {
+      "epoch": 1.0105543828463504,
+      "grad_norm": 0.6741541624069214,
+      "learning_rate": 0.0001,
+      "loss": 0.8731,
+      "step": 1137
+    },
+    {
+      "epoch": 1.01144317298078,
+      "grad_norm": 0.3466031551361084,
+      "learning_rate": 0.0001,
+      "loss": 0.8852,
+      "step": 1138
+    },
+    {
+      "epoch": 1.0123319631152095,
+      "grad_norm": 0.3946625590324402,
+      "learning_rate": 0.0001,
+      "loss": 0.9609,
+      "step": 1139
+    },
+    {
+      "epoch": 1.013220753249639,
+      "grad_norm": 0.3739851713180542,
+      "learning_rate": 0.0001,
+      "loss": 0.7588,
+      "step": 1140
+    },
+    {
+      "epoch": 1.0141095433840683,
+      "grad_norm": 0.3292269706726074,
+      "learning_rate": 0.0001,
+      "loss": 0.9439,
+      "step": 1141
+    },
+    {
+      "epoch": 1.014998333518498,
+      "grad_norm": 0.4359779953956604,
+      "learning_rate": 0.0001,
+      "loss": 0.9791,
+      "step": 1142
+    },
+    {
+      "epoch": 1.0158871236529274,
+      "grad_norm": 0.392539381980896,
+      "learning_rate": 0.0001,
+      "loss": 0.8939,
+      "step": 1143
+    },
+    {
+      "epoch": 1.0167759137873569,
+      "grad_norm": 0.36536142230033875,
+      "learning_rate": 0.0001,
+      "loss": 0.9454,
+      "step": 1144
+    },
+    {
+      "epoch": 1.0176647039217865,
+      "grad_norm": 0.4121522903442383,
+      "learning_rate": 0.0001,
+      "loss": 0.951,
+      "step": 1145
+    },
+    {
+      "epoch": 1.018553494056216,
+      "grad_norm": 0.3721693456172943,
+      "learning_rate": 0.0001,
+      "loss": 0.9443,
+      "step": 1146
+    },
+    {
+      "epoch": 1.0194422841906454,
+      "grad_norm": 0.42354851961135864,
+      "learning_rate": 0.0001,
+      "loss": 0.903,
+      "step": 1147
+    },
+    {
+      "epoch": 1.020331074325075,
+      "grad_norm": 0.3700781464576721,
+      "learning_rate": 0.0001,
+      "loss": 0.9729,
+      "step": 1148
+    },
+    {
+      "epoch": 1.0212198644595045,
+      "grad_norm": 0.37510499358177185,
+      "learning_rate": 0.0001,
+      "loss": 0.9149,
+      "step": 1149
+    },
+    {
+      "epoch": 1.022108654593934,
+      "grad_norm": 0.3892500400543213,
+      "learning_rate": 0.0001,
+      "loss": 0.9894,
+      "step": 1150
+    },
+    {
+      "epoch": 1.0229974447283636,
+      "grad_norm": 0.4017266631126404,
+      "learning_rate": 0.0001,
+      "loss": 0.9369,
+      "step": 1151
+    },
+    {
+      "epoch": 1.023886234862793,
+      "grad_norm": 0.37094780802726746,
+      "learning_rate": 0.0001,
+      "loss": 1.0178,
+      "step": 1152
+    },
+    {
+      "epoch": 1.0247750249972225,
+      "grad_norm": 0.3605554401874542,
+      "learning_rate": 0.0001,
+      "loss": 0.8953,
+      "step": 1153
+    },
+    {
+      "epoch": 1.025663815131652,
+      "grad_norm": 0.3322463035583496,
+      "learning_rate": 0.0001,
+      "loss": 0.8531,
+      "step": 1154
+    },
+    {
+      "epoch": 1.0265526052660816,
+      "grad_norm": 0.35462141036987305,
+      "learning_rate": 0.0001,
+      "loss": 0.9225,
+      "step": 1155
+    },
+    {
+      "epoch": 1.027441395400511,
+      "grad_norm": 0.3974057137966156,
+      "learning_rate": 0.0001,
+      "loss": 0.9839,
+      "step": 1156
+    },
+    {
+      "epoch": 1.0283301855349405,
+      "grad_norm": 0.3556059002876282,
+      "learning_rate": 0.0001,
+      "loss": 0.9588,
+      "step": 1157
+    },
+    {
+      "epoch": 1.0292189756693702,
+      "grad_norm": 0.37689536809921265,
+      "learning_rate": 0.0001,
+      "loss": 0.923,
+      "step": 1158
+    },
+    {
+      "epoch": 1.0301077658037996,
+      "grad_norm": 0.36872097849845886,
+      "learning_rate": 0.0001,
+      "loss": 0.9578,
+      "step": 1159
+    },
+    {
+      "epoch": 1.030996555938229,
+      "grad_norm": 0.44219520688056946,
+      "learning_rate": 0.0001,
+      "loss": 1.0076,
+      "step": 1160
+    },
+    {
+      "epoch": 1.0318853460726587,
+      "grad_norm": 0.3993608355522156,
+      "learning_rate": 0.0001,
+      "loss": 0.9316,
+      "step": 1161
+    },
+    {
+      "epoch": 1.0327741362070881,
+      "grad_norm": 0.3941507041454315,
+      "learning_rate": 0.0001,
+      "loss": 0.9119,
+      "step": 1162
+    },
+    {
+      "epoch": 1.0336629263415176,
+      "grad_norm": 0.43561699986457825,
+      "learning_rate": 0.0001,
+      "loss": 0.8983,
+      "step": 1163
+    },
+    {
+      "epoch": 1.034551716475947,
+      "grad_norm": 0.3520337641239166,
+      "learning_rate": 0.0001,
+      "loss": 0.8171,
+      "step": 1164
+    },
+    {
+      "epoch": 1.0354405066103767,
+      "grad_norm": 0.376632958650589,
+      "learning_rate": 0.0001,
+      "loss": 0.8648,
+      "step": 1165
+    },
+    {
+      "epoch": 1.0363292967448061,
+      "grad_norm": 0.6677507162094116,
+      "learning_rate": 0.0001,
+      "loss": 0.9802,
+      "step": 1166
+    },
+    {
+      "epoch": 1.0372180868792356,
+      "grad_norm": 0.3786594271659851,
+      "learning_rate": 0.0001,
+      "loss": 1.0064,
+      "step": 1167
+    },
+    {
+      "epoch": 1.0381068770136652,
+      "grad_norm": 0.3683249056339264,
+      "learning_rate": 0.0001,
+      "loss": 0.8562,
+      "step": 1168
+    },
+    {
+      "epoch": 1.0389956671480947,
+      "grad_norm": 0.38416755199432373,
+      "learning_rate": 0.0001,
+      "loss": 0.9727,
+      "step": 1169
+    },
+    {
+      "epoch": 1.039884457282524,
+      "grad_norm": 0.3452812135219574,
+      "learning_rate": 0.0001,
+      "loss": 0.8639,
+      "step": 1170
+    },
+    {
+      "epoch": 1.0407732474169538,
+      "grad_norm": 0.3969337046146393,
+      "learning_rate": 0.0001,
+      "loss": 0.9049,
+      "step": 1171
+    },
+    {
+      "epoch": 1.0416620375513832,
+      "grad_norm": 0.3651241362094879,
+      "learning_rate": 0.0001,
+      "loss": 0.8987,
+      "step": 1172
+    },
+    {
+      "epoch": 1.0425508276858126,
+      "grad_norm": 0.4599679410457611,
+      "learning_rate": 0.0001,
+      "loss": 1.0484,
+      "step": 1173
+    },
+    {
+      "epoch": 1.0434396178202423,
+      "grad_norm": 0.4104447066783905,
+      "learning_rate": 0.0001,
+      "loss": 0.9513,
+      "step": 1174
+    },
+    {
+      "epoch": 1.0443284079546717,
+      "grad_norm": 0.39843717217445374,
+      "learning_rate": 0.0001,
+      "loss": 1.0321,
+      "step": 1175
+    },
+    {
+      "epoch": 1.0452171980891012,
+      "grad_norm": 0.3388933837413788,
+      "learning_rate": 0.0001,
+      "loss": 0.8599,
+      "step": 1176
+    },
+    {
+      "epoch": 1.0461059882235306,
+      "grad_norm": 0.3476930856704712,
+      "learning_rate": 0.0001,
+      "loss": 0.8383,
+      "step": 1177
+    },
+    {
+      "epoch": 1.0469947783579603,
+      "grad_norm": 0.37614160776138306,
+      "learning_rate": 0.0001,
+      "loss": 0.9211,
+      "step": 1178
+    },
+    {
+      "epoch": 1.0478835684923897,
+      "grad_norm": 0.37879347801208496,
+      "learning_rate": 0.0001,
+      "loss": 0.9369,
+      "step": 1179
+    },
+    {
+      "epoch": 1.0487723586268192,
+      "grad_norm": 0.3917993903160095,
+      "learning_rate": 0.0001,
+      "loss": 0.9324,
+      "step": 1180
+    },
+    {
+      "epoch": 1.0496611487612488,
+      "grad_norm": 0.32946088910102844,
+      "learning_rate": 0.0001,
+      "loss": 0.913,
+      "step": 1181
+    },
+    {
+      "epoch": 1.0505499388956783,
+      "grad_norm": 0.362324595451355,
+      "learning_rate": 0.0001,
+      "loss": 0.9393,
+      "step": 1182
+    },
+    {
+      "epoch": 1.0514387290301077,
+      "grad_norm": 0.411262184381485,
+      "learning_rate": 0.0001,
+      "loss": 0.9152,
+      "step": 1183
+    },
+    {
+      "epoch": 1.0523275191645374,
+      "grad_norm": 0.3655692934989929,
+      "learning_rate": 0.0001,
+      "loss": 0.8978,
+      "step": 1184
+    },
+    {
+      "epoch": 1.0532163092989668,
+      "grad_norm": 0.4027964174747467,
+      "learning_rate": 0.0001,
+      "loss": 0.954,
+      "step": 1185
+    },
+    {
+      "epoch": 1.0541050994333963,
+      "grad_norm": 0.362482488155365,
+      "learning_rate": 0.0001,
+      "loss": 0.8665,
+      "step": 1186
+    },
+    {
+      "epoch": 1.0549938895678257,
+      "grad_norm": 0.43920066952705383,
+      "learning_rate": 0.0001,
+      "loss": 0.9462,
+      "step": 1187
+    },
+    {
+      "epoch": 1.0558826797022554,
+      "grad_norm": 0.4220573604106903,
+      "learning_rate": 0.0001,
+      "loss": 1.0071,
+      "step": 1188
+    },
+    {
+      "epoch": 1.0567714698366848,
+      "grad_norm": 0.3542270064353943,
+      "learning_rate": 0.0001,
+      "loss": 0.8843,
+      "step": 1189
+    },
+    {
+      "epoch": 1.0576602599711142,
+      "grad_norm": 0.3698994815349579,
+      "learning_rate": 0.0001,
+      "loss": 0.883,
+      "step": 1190
+    },
+    {
+      "epoch": 1.058549050105544,
+      "grad_norm": 0.39016515016555786,
+      "learning_rate": 0.0001,
+      "loss": 1.0091,
+      "step": 1191
+    },
+    {
+      "epoch": 1.0594378402399733,
+      "grad_norm": 0.377370685338974,
+      "learning_rate": 0.0001,
+      "loss": 0.9807,
+      "step": 1192
+    },
+    {
+      "epoch": 1.0603266303744028,
+      "grad_norm": 0.34474626183509827,
+      "learning_rate": 0.0001,
+      "loss": 0.9282,
+      "step": 1193
+    },
+    {
+      "epoch": 1.0612154205088324,
+      "grad_norm": 0.35169529914855957,
+      "learning_rate": 0.0001,
+      "loss": 0.9907,
+      "step": 1194
+    },
+    {
+      "epoch": 1.0621042106432619,
+      "grad_norm": 0.3528148829936981,
+      "learning_rate": 0.0001,
+      "loss": 0.8829,
+      "step": 1195
+    },
+    {
+      "epoch": 1.0629930007776913,
+      "grad_norm": 0.382842481136322,
+      "learning_rate": 0.0001,
+      "loss": 1.0002,
+      "step": 1196
+    },
+    {
+      "epoch": 1.063881790912121,
+      "grad_norm": 0.35213178396224976,
+      "learning_rate": 0.0001,
+      "loss": 0.8423,
+      "step": 1197
+    },
+    {
+      "epoch": 1.0647705810465504,
+      "grad_norm": 0.3510130047798157,
+      "learning_rate": 0.0001,
+      "loss": 0.8727,
+      "step": 1198
+    },
+    {
+      "epoch": 1.0656593711809799,
+      "grad_norm": 0.3832014501094818,
+      "learning_rate": 0.0001,
+      "loss": 0.9299,
+      "step": 1199
+    },
+    {
+      "epoch": 1.0665481613154093,
+      "grad_norm": 0.3763968050479889,
+      "learning_rate": 0.0001,
+      "loss": 0.9403,
+      "step": 1200
+    },
+    {
+      "epoch": 1.067436951449839,
+      "grad_norm": 0.4154573678970337,
+      "learning_rate": 0.0001,
+      "loss": 0.9413,
+      "step": 1201
+    },
+    {
+      "epoch": 1.0683257415842684,
+      "grad_norm": 0.4002286195755005,
+      "learning_rate": 0.0001,
+      "loss": 1.0054,
+      "step": 1202
+    },
+    {
+      "epoch": 1.0692145317186978,
+      "grad_norm": 0.40614718198776245,
+      "learning_rate": 0.0001,
+      "loss": 0.9248,
+      "step": 1203
+    },
+    {
+      "epoch": 1.0701033218531275,
+      "grad_norm": 0.3200342357158661,
+      "learning_rate": 0.0001,
+      "loss": 0.9438,
+      "step": 1204
+    },
+    {
+      "epoch": 1.070992111987557,
+      "grad_norm": 0.4085826575756073,
+      "learning_rate": 0.0001,
+      "loss": 0.9417,
+      "step": 1205
+    },
+    {
+      "epoch": 1.0718809021219864,
+      "grad_norm": 0.3780025541782379,
+      "learning_rate": 0.0001,
+      "loss": 0.9076,
+      "step": 1206
+    },
+    {
+      "epoch": 1.072769692256416,
+      "grad_norm": 0.42114147543907166,
+      "learning_rate": 0.0001,
+      "loss": 0.9071,
+      "step": 1207
+    },
+    {
+      "epoch": 1.0736584823908455,
+      "grad_norm": 0.41555988788604736,
+      "learning_rate": 0.0001,
+      "loss": 0.9986,
+      "step": 1208
+    },
+    {
+      "epoch": 1.074547272525275,
+      "grad_norm": 0.4004509747028351,
+      "learning_rate": 0.0001,
+      "loss": 0.9048,
+      "step": 1209
+    },
+    {
+      "epoch": 1.0754360626597044,
+      "grad_norm": 1.0510995388031006,
+      "learning_rate": 0.0001,
+      "loss": 0.9838,
+      "step": 1210
+    },
+    {
+      "epoch": 1.076324852794134,
+      "grad_norm": 0.3881635069847107,
+      "learning_rate": 0.0001,
+      "loss": 0.978,
+      "step": 1211
+    },
+    {
+      "epoch": 1.0772136429285635,
+      "grad_norm": 0.4103749096393585,
+      "learning_rate": 0.0001,
+      "loss": 0.9817,
+      "step": 1212
+    },
+    {
+      "epoch": 1.078102433062993,
+      "grad_norm": 0.4124666750431061,
+      "learning_rate": 0.0001,
+      "loss": 0.9354,
+      "step": 1213
+    },
+    {
+      "epoch": 1.0789912231974226,
+      "grad_norm": 0.433636337518692,
+      "learning_rate": 0.0001,
+      "loss": 0.9371,
+      "step": 1214
+    },
+    {
+      "epoch": 1.079880013331852,
+      "grad_norm": 0.41402560472488403,
+      "learning_rate": 0.0001,
+      "loss": 0.9656,
+      "step": 1215
+    },
+    {
+      "epoch": 1.0807688034662815,
+      "grad_norm": 0.36485034227371216,
+      "learning_rate": 0.0001,
+      "loss": 0.9697,
+      "step": 1216
+    },
+    {
+      "epoch": 1.0816575936007111,
+      "grad_norm": 0.41599103808403015,
+      "learning_rate": 0.0001,
+      "loss": 0.8941,
+      "step": 1217
+    },
+    {
+      "epoch": 1.0825463837351406,
+      "grad_norm": 0.38060900568962097,
+      "learning_rate": 0.0001,
+      "loss": 0.8354,
+      "step": 1218
+    },
+    {
+      "epoch": 1.08343517386957,
+      "grad_norm": 0.40148356556892395,
+      "learning_rate": 0.0001,
+      "loss": 0.9132,
+      "step": 1219
+    },
+    {
+      "epoch": 1.0843239640039997,
+      "grad_norm": 0.3883506655693054,
+      "learning_rate": 0.0001,
+      "loss": 0.9233,
+      "step": 1220
+    },
+    {
+      "epoch": 1.085212754138429,
+      "grad_norm": 0.34674960374832153,
+      "learning_rate": 0.0001,
+      "loss": 0.8928,
+      "step": 1221
+    },
+    {
+      "epoch": 1.0861015442728585,
+      "grad_norm": 0.35441455245018005,
+      "learning_rate": 0.0001,
+      "loss": 0.8653,
+      "step": 1222
+    },
+    {
+      "epoch": 1.086990334407288,
+      "grad_norm": 0.37511923909187317,
+      "learning_rate": 0.0001,
+      "loss": 0.9179,
+      "step": 1223
+    },
+    {
+      "epoch": 1.0878791245417176,
+      "grad_norm": 0.35857459902763367,
+      "learning_rate": 0.0001,
+      "loss": 0.8866,
+      "step": 1224
+    },
+    {
+      "epoch": 1.088767914676147,
+      "grad_norm": 0.3715518116950989,
+      "learning_rate": 0.0001,
+      "loss": 0.9263,
+      "step": 1225
+    },
+    {
+      "epoch": 1.0896567048105765,
+      "grad_norm": 0.3738372325897217,
+      "learning_rate": 0.0001,
+      "loss": 0.9759,
+      "step": 1226
+    },
+    {
+      "epoch": 1.0905454949450062,
+      "grad_norm": 0.41106364130973816,
+      "learning_rate": 0.0001,
+      "loss": 0.9539,
+      "step": 1227
+    },
+    {
+      "epoch": 1.0914342850794356,
+      "grad_norm": 0.36827585101127625,
+      "learning_rate": 0.0001,
+      "loss": 0.949,
+      "step": 1228
+    },
+    {
+      "epoch": 1.092323075213865,
+      "grad_norm": 0.39710530638694763,
+      "learning_rate": 0.0001,
+      "loss": 0.9351,
+      "step": 1229
+    },
+    {
+      "epoch": 1.0932118653482947,
+      "grad_norm": 0.3501657247543335,
+      "learning_rate": 0.0001,
+      "loss": 0.9026,
+      "step": 1230
+    },
+    {
+      "epoch": 1.0941006554827242,
+      "grad_norm": 0.4335416853427887,
+      "learning_rate": 0.0001,
+      "loss": 0.9438,
+      "step": 1231
+    },
+    {
+      "epoch": 1.0949894456171536,
+      "grad_norm": 0.3835698962211609,
+      "learning_rate": 0.0001,
+      "loss": 0.8951,
+      "step": 1232
+    },
+    {
+      "epoch": 1.095878235751583,
+      "grad_norm": 0.3333055078983307,
+      "learning_rate": 0.0001,
+      "loss": 0.8505,
+      "step": 1233
+    },
+    {
+      "epoch": 1.0967670258860127,
+      "grad_norm": 0.4030056595802307,
+      "learning_rate": 0.0001,
+      "loss": 0.9226,
+      "step": 1234
+    },
+    {
+      "epoch": 1.0976558160204422,
+      "grad_norm": 0.4055632948875427,
+      "learning_rate": 0.0001,
+      "loss": 0.9892,
+      "step": 1235
+    },
+    {
+      "epoch": 1.0985446061548716,
+      "grad_norm": 0.3327998220920563,
+      "learning_rate": 0.0001,
+      "loss": 0.8774,
+      "step": 1236
+    },
+    {
+      "epoch": 1.0994333962893013,
+      "grad_norm": 0.4032350182533264,
+      "learning_rate": 0.0001,
+      "loss": 0.8252,
+      "step": 1237
+    },
+    {
+      "epoch": 1.1003221864237307,
+      "grad_norm": 0.40369337797164917,
+      "learning_rate": 0.0001,
+      "loss": 0.9785,
+      "step": 1238
+    },
+    {
+      "epoch": 1.1012109765581601,
+      "grad_norm": 0.339615136384964,
+      "learning_rate": 0.0001,
+      "loss": 0.9099,
+      "step": 1239
+    },
+    {
+      "epoch": 1.1020997666925898,
+      "grad_norm": 0.4322150945663452,
+      "learning_rate": 0.0001,
+      "loss": 0.9437,
+      "step": 1240
+    },
+    {
+      "epoch": 1.1029885568270192,
+      "grad_norm": 0.4031563103199005,
+      "learning_rate": 0.0001,
+      "loss": 0.9856,
+      "step": 1241
+    },
+    {
+      "epoch": 1.1038773469614487,
+      "grad_norm": 0.36764100193977356,
+      "learning_rate": 0.0001,
+      "loss": 0.9202,
+      "step": 1242
+    },
+    {
+      "epoch": 1.1047661370958783,
+      "grad_norm": 0.35683104395866394,
+      "learning_rate": 0.0001,
+      "loss": 0.8976,
+      "step": 1243
+    },
+    {
+      "epoch": 1.1056549272303078,
+      "grad_norm": 0.3336647152900696,
+      "learning_rate": 0.0001,
+      "loss": 0.875,
+      "step": 1244
+    },
+    {
+      "epoch": 1.1065437173647372,
+      "grad_norm": 0.36215344071388245,
+      "learning_rate": 0.0001,
+      "loss": 0.8985,
+      "step": 1245
+    },
+    {
+      "epoch": 1.1074325074991667,
+      "grad_norm": 0.3436198830604553,
+      "learning_rate": 0.0001,
+      "loss": 0.8672,
+      "step": 1246
+    },
+    {
+      "epoch": 1.1083212976335963,
+      "grad_norm": 0.3478955328464508,
+      "learning_rate": 0.0001,
+      "loss": 0.833,
+      "step": 1247
+    },
+    {
+      "epoch": 1.1092100877680258,
+      "grad_norm": 0.3481768071651459,
+      "learning_rate": 0.0001,
+      "loss": 0.891,
+      "step": 1248
+    },
+    {
+      "epoch": 1.1100988779024552,
+      "grad_norm": 0.37578824162483215,
+      "learning_rate": 0.0001,
+      "loss": 0.88,
+      "step": 1249
+    },
+    {
+      "epoch": 1.1109876680368849,
+      "grad_norm": 0.3523572087287903,
+      "learning_rate": 0.0001,
+      "loss": 0.8152,
+      "step": 1250
+    },
+    {
+      "epoch": 1.1118764581713143,
+      "grad_norm": 0.37987327575683594,
+      "learning_rate": 0.0001,
+      "loss": 0.9859,
+      "step": 1251
+    },
+    {
+      "epoch": 1.1127652483057437,
+      "grad_norm": 0.34919604659080505,
+      "learning_rate": 0.0001,
+      "loss": 0.948,
+      "step": 1252
+    },
+    {
+      "epoch": 1.1136540384401734,
+      "grad_norm": 0.401996910572052,
+      "learning_rate": 0.0001,
+      "loss": 0.9123,
+      "step": 1253
+    },
+    {
+      "epoch": 1.1145428285746029,
+      "grad_norm": 0.3942583501338959,
+      "learning_rate": 0.0001,
+      "loss": 0.8524,
+      "step": 1254
+    },
+    {
+      "epoch": 1.1154316187090323,
+      "grad_norm": 0.3758942484855652,
+      "learning_rate": 0.0001,
+      "loss": 0.8813,
+      "step": 1255
+    },
+    {
+      "epoch": 1.1163204088434617,
+      "grad_norm": 0.3327169716358185,
+      "learning_rate": 0.0001,
+      "loss": 0.8493,
+      "step": 1256
+    },
+    {
+      "epoch": 1.1172091989778914,
+      "grad_norm": 0.34999555349349976,
+      "learning_rate": 0.0001,
+      "loss": 0.8986,
+      "step": 1257
+    },
+    {
+      "epoch": 1.1180979891123208,
+      "grad_norm": 0.3059004843235016,
+      "learning_rate": 0.0001,
+      "loss": 0.8299,
+      "step": 1258
+    },
+    {
+      "epoch": 1.1189867792467503,
+      "grad_norm": 0.40918049216270447,
+      "learning_rate": 0.0001,
+      "loss": 1.0235,
+      "step": 1259
+    },
+    {
+      "epoch": 1.11987556938118,
+      "grad_norm": 0.45679351687431335,
+      "learning_rate": 0.0001,
+      "loss": 0.9245,
+      "step": 1260
+    },
+    {
+      "epoch": 1.1207643595156094,
+      "grad_norm": 0.36050447821617126,
+      "learning_rate": 0.0001,
+      "loss": 0.7914,
+      "step": 1261
+    },
+    {
+      "epoch": 1.1216531496500388,
+      "grad_norm": 0.3530547320842743,
+      "learning_rate": 0.0001,
+      "loss": 0.8892,
+      "step": 1262
+    },
+    {
+      "epoch": 1.1225419397844685,
+      "grad_norm": 0.39871805906295776,
+      "learning_rate": 0.0001,
+      "loss": 0.9088,
+      "step": 1263
+    },
+    {
+      "epoch": 1.123430729918898,
+      "grad_norm": 0.37267911434173584,
+      "learning_rate": 0.0001,
+      "loss": 0.9331,
+      "step": 1264
+    },
+    {
+      "epoch": 1.1243195200533274,
+      "grad_norm": 0.3619190752506256,
+      "learning_rate": 0.0001,
+      "loss": 0.8432,
+      "step": 1265
+    },
+    {
+      "epoch": 1.125208310187757,
+      "grad_norm": 0.3668549656867981,
+      "learning_rate": 0.0001,
+      "loss": 0.9503,
+      "step": 1266
+    },
+    {
+      "epoch": 1.1260971003221865,
+      "grad_norm": 0.34030482172966003,
+      "learning_rate": 0.0001,
+      "loss": 0.8852,
+      "step": 1267
+    },
+    {
+      "epoch": 1.126985890456616,
+      "grad_norm": 0.4020557403564453,
+      "learning_rate": 0.0001,
+      "loss": 0.9593,
+      "step": 1268
+    },
+    {
+      "epoch": 1.1278746805910456,
+      "grad_norm": 0.34024256467819214,
+      "learning_rate": 0.0001,
+      "loss": 0.8787,
+      "step": 1269
+    },
+    {
+      "epoch": 1.128763470725475,
+      "grad_norm": 0.4241231083869934,
+      "learning_rate": 0.0001,
+      "loss": 0.9793,
+      "step": 1270
+    },
+    {
+      "epoch": 1.1296522608599044,
+      "grad_norm": 0.4005233645439148,
+      "learning_rate": 0.0001,
+      "loss": 0.9619,
+      "step": 1271
+    },
+    {
+      "epoch": 1.1305410509943339,
+      "grad_norm": 0.38167819380760193,
+      "learning_rate": 0.0001,
+      "loss": 0.9449,
+      "step": 1272
+    },
+    {
+      "epoch": 1.1314298411287635,
+      "grad_norm": 0.34920740127563477,
+      "learning_rate": 0.0001,
+      "loss": 0.9811,
+      "step": 1273
+    },
+    {
+      "epoch": 1.132318631263193,
+      "grad_norm": 0.3310723900794983,
+      "learning_rate": 0.0001,
+      "loss": 0.8865,
+      "step": 1274
+    },
+    {
+      "epoch": 1.1332074213976224,
+      "grad_norm": 0.3370254933834076,
+      "learning_rate": 0.0001,
+      "loss": 0.8304,
+      "step": 1275
+    },
+    {
+      "epoch": 1.134096211532052,
+      "grad_norm": 0.35583510994911194,
+      "learning_rate": 0.0001,
+      "loss": 0.9203,
+      "step": 1276
+    },
+    {
+      "epoch": 1.1349850016664815,
+      "grad_norm": 0.34542956948280334,
+      "learning_rate": 0.0001,
+      "loss": 0.9403,
+      "step": 1277
+    },
+    {
+      "epoch": 1.135873791800911,
+      "grad_norm": 0.3754197657108307,
+      "learning_rate": 0.0001,
+      "loss": 0.9866,
+      "step": 1278
+    },
+    {
+      "epoch": 1.1367625819353404,
+      "grad_norm": 0.3555785119533539,
+      "learning_rate": 0.0001,
+      "loss": 0.8689,
+      "step": 1279
+    },
+    {
+      "epoch": 1.13765137206977,
+      "grad_norm": 0.33429041504859924,
+      "learning_rate": 0.0001,
+      "loss": 0.833,
+      "step": 1280
+    },
+    {
+      "epoch": 1.1385401622041995,
+      "grad_norm": 0.3541424572467804,
+      "learning_rate": 0.0001,
+      "loss": 0.8759,
+      "step": 1281
+    },
+    {
+      "epoch": 1.139428952338629,
+      "grad_norm": 0.36240896582603455,
+      "learning_rate": 0.0001,
+      "loss": 0.9841,
+      "step": 1282
+    },
+    {
+      "epoch": 1.1403177424730586,
+      "grad_norm": 0.3360092043876648,
+      "learning_rate": 0.0001,
+      "loss": 0.9414,
+      "step": 1283
+    },
+    {
+      "epoch": 1.141206532607488,
+      "grad_norm": 0.3737250864505768,
+      "learning_rate": 0.0001,
+      "loss": 0.9729,
+      "step": 1284
+    },
+    {
+      "epoch": 1.1420953227419175,
+      "grad_norm": 0.37672123312950134,
+      "learning_rate": 0.0001,
+      "loss": 0.8692,
+      "step": 1285
+    },
+    {
+      "epoch": 1.1429841128763472,
+      "grad_norm": 0.3979622721672058,
+      "learning_rate": 0.0001,
+      "loss": 0.9644,
+      "step": 1286
+    },
+    {
+      "epoch": 1.1438729030107766,
+      "grad_norm": 0.4209362268447876,
+      "learning_rate": 0.0001,
+      "loss": 0.9097,
+      "step": 1287
+    },
+    {
+      "epoch": 1.144761693145206,
+      "grad_norm": 0.3987758159637451,
+      "learning_rate": 0.0001,
+      "loss": 0.8919,
+      "step": 1288
+    },
+    {
+      "epoch": 1.1456504832796357,
+      "grad_norm": 0.3739655613899231,
+      "learning_rate": 0.0001,
+      "loss": 0.9623,
+      "step": 1289
+    },
+    {
+      "epoch": 1.1465392734140651,
+      "grad_norm": 0.3652169406414032,
+      "learning_rate": 0.0001,
+      "loss": 0.9033,
+      "step": 1290
+    },
+    {
+      "epoch": 1.1474280635484946,
+      "grad_norm": 0.3900451958179474,
+      "learning_rate": 0.0001,
+      "loss": 0.9795,
+      "step": 1291
+    },
+    {
+      "epoch": 1.1483168536829242,
+      "grad_norm": 0.36547282338142395,
+      "learning_rate": 0.0001,
+      "loss": 0.9595,
+      "step": 1292
+    },
+    {
+      "epoch": 1.1492056438173537,
+      "grad_norm": 0.3713114559650421,
+      "learning_rate": 0.0001,
+      "loss": 0.8854,
+      "step": 1293
+    },
+    {
+      "epoch": 1.1500944339517831,
+      "grad_norm": 0.3759624660015106,
+      "learning_rate": 0.0001,
+      "loss": 0.9053,
+      "step": 1294
+    },
+    {
+      "epoch": 1.1509832240862126,
+      "grad_norm": 0.3536144196987152,
+      "learning_rate": 0.0001,
+      "loss": 1.0054,
+      "step": 1295
+    },
+    {
+      "epoch": 1.1518720142206422,
+      "grad_norm": 0.36850425601005554,
+      "learning_rate": 0.0001,
+      "loss": 0.9134,
+      "step": 1296
+    },
+    {
+      "epoch": 1.1527608043550717,
+      "grad_norm": 0.3645191192626953,
+      "learning_rate": 0.0001,
+      "loss": 1.0067,
+      "step": 1297
+    },
+    {
+      "epoch": 1.153649594489501,
+      "grad_norm": 0.3773975670337677,
+      "learning_rate": 0.0001,
+      "loss": 0.9362,
+      "step": 1298
+    },
+    {
+      "epoch": 1.1545383846239308,
+      "grad_norm": 0.35518959164619446,
+      "learning_rate": 0.0001,
+      "loss": 0.9289,
+      "step": 1299
+    },
+    {
+      "epoch": 1.1554271747583602,
+      "grad_norm": 0.4167252779006958,
+      "learning_rate": 0.0001,
+      "loss": 0.8655,
+      "step": 1300
+    },
+    {
+      "epoch": 1.1563159648927896,
+      "grad_norm": 0.35073405504226685,
+      "learning_rate": 0.0001,
+      "loss": 0.8565,
+      "step": 1301
+    },
+    {
+      "epoch": 1.157204755027219,
+      "grad_norm": 0.3294246196746826,
+      "learning_rate": 0.0001,
+      "loss": 0.8501,
+      "step": 1302
+    },
+    {
+      "epoch": 1.1580935451616488,
+      "grad_norm": 0.3594052493572235,
+      "learning_rate": 0.0001,
+      "loss": 0.96,
+      "step": 1303
+    },
+    {
+      "epoch": 1.1589823352960782,
+      "grad_norm": 0.40062353014945984,
+      "learning_rate": 0.0001,
+      "loss": 0.9908,
+      "step": 1304
+    },
+    {
+      "epoch": 1.1598711254305076,
+      "grad_norm": 0.38618043065071106,
+      "learning_rate": 0.0001,
+      "loss": 0.9185,
+      "step": 1305
+    },
+    {
+      "epoch": 1.1607599155649373,
+      "grad_norm": 0.36038386821746826,
+      "learning_rate": 0.0001,
+      "loss": 0.9733,
+      "step": 1306
+    },
+    {
+      "epoch": 1.1616487056993667,
+      "grad_norm": 0.3996288776397705,
+      "learning_rate": 0.0001,
+      "loss": 0.9534,
+      "step": 1307
+    },
+    {
+      "epoch": 1.1625374958337962,
+      "grad_norm": 0.3757215440273285,
+      "learning_rate": 0.0001,
+      "loss": 0.9219,
+      "step": 1308
+    },
+    {
+      "epoch": 1.1634262859682258,
+      "grad_norm": 0.3682938814163208,
+      "learning_rate": 0.0001,
+      "loss": 0.9141,
+      "step": 1309
+    },
+    {
+      "epoch": 1.1643150761026553,
+      "grad_norm": 0.36964985728263855,
+      "learning_rate": 0.0001,
+      "loss": 0.9956,
+      "step": 1310
+    },
+    {
+      "epoch": 1.1652038662370847,
+      "grad_norm": 0.39142096042633057,
+      "learning_rate": 0.0001,
+      "loss": 0.8502,
+      "step": 1311
+    },
+    {
+      "epoch": 1.1660926563715144,
+      "grad_norm": 0.31965819001197815,
+      "learning_rate": 0.0001,
+      "loss": 0.8842,
+      "step": 1312
+    },
+    {
+      "epoch": 1.1669814465059438,
+      "grad_norm": 0.40179508924484253,
+      "learning_rate": 0.0001,
+      "loss": 0.9441,
+      "step": 1313
+    },
+    {
+      "epoch": 1.1678702366403733,
+      "grad_norm": 0.3406355679035187,
+      "learning_rate": 0.0001,
+      "loss": 0.9468,
+      "step": 1314
+    },
+    {
+      "epoch": 1.168759026774803,
+      "grad_norm": 0.3638407289981842,
+      "learning_rate": 0.0001,
+      "loss": 0.9296,
+      "step": 1315
+    },
+    {
+      "epoch": 1.1696478169092324,
+      "grad_norm": 0.34185290336608887,
+      "learning_rate": 0.0001,
+      "loss": 0.8647,
+      "step": 1316
+    },
+    {
+      "epoch": 1.1705366070436618,
+      "grad_norm": 0.3356599807739258,
+      "learning_rate": 0.0001,
+      "loss": 0.9484,
+      "step": 1317
+    },
+    {
+      "epoch": 1.1714253971780912,
+      "grad_norm": 0.3953014314174652,
+      "learning_rate": 0.0001,
+      "loss": 0.947,
+      "step": 1318
+    },
+    {
+      "epoch": 1.172314187312521,
+      "grad_norm": 0.4338318109512329,
+      "learning_rate": 0.0001,
+      "loss": 0.9389,
+      "step": 1319
+    },
+    {
+      "epoch": 1.1732029774469503,
+      "grad_norm": 0.38250255584716797,
+      "learning_rate": 0.0001,
+      "loss": 0.9113,
+      "step": 1320
+    },
+    {
+      "epoch": 1.1740917675813798,
+      "grad_norm": 1.4235469102859497,
+      "learning_rate": 0.0001,
+      "loss": 0.837,
+      "step": 1321
+    },
+    {
+      "epoch": 1.1749805577158094,
+      "grad_norm": 0.4314570724964142,
+      "learning_rate": 0.0001,
+      "loss": 0.955,
+      "step": 1322
+    },
+    {
+      "epoch": 1.1758693478502389,
+      "grad_norm": 0.3919623792171478,
+      "learning_rate": 0.0001,
+      "loss": 0.9662,
+      "step": 1323
+    },
+    {
+      "epoch": 1.1767581379846683,
+      "grad_norm": 0.3859492540359497,
+      "learning_rate": 0.0001,
+      "loss": 0.8974,
+      "step": 1324
+    },
+    {
+      "epoch": 1.1776469281190978,
+      "grad_norm": 0.39512211084365845,
+      "learning_rate": 0.0001,
+      "loss": 0.9592,
+      "step": 1325
+    },
+    {
+      "epoch": 1.1785357182535274,
+      "grad_norm": 0.34336525201797485,
+      "learning_rate": 0.0001,
+      "loss": 0.8636,
+      "step": 1326
+    },
+    {
+      "epoch": 1.1794245083879569,
+      "grad_norm": 0.39709898829460144,
+      "learning_rate": 0.0001,
+      "loss": 0.8655,
+      "step": 1327
+    },
+    {
+      "epoch": 1.1803132985223863,
+      "grad_norm": 0.4077267348766327,
+      "learning_rate": 0.0001,
+      "loss": 0.9687,
+      "step": 1328
+    },
+    {
+      "epoch": 1.181202088656816,
+      "grad_norm": 0.3338770866394043,
+      "learning_rate": 0.0001,
+      "loss": 0.9293,
+      "step": 1329
+    },
+    {
+      "epoch": 1.1820908787912454,
+      "grad_norm": 0.3750726580619812,
+      "learning_rate": 0.0001,
+      "loss": 0.922,
+      "step": 1330
+    },
+    {
+      "epoch": 1.1829796689256749,
+      "grad_norm": 0.35356971621513367,
+      "learning_rate": 0.0001,
+      "loss": 0.8574,
+      "step": 1331
+    },
+    {
+      "epoch": 1.1838684590601045,
+      "grad_norm": 0.41766199469566345,
+      "learning_rate": 0.0001,
+      "loss": 0.9395,
+      "step": 1332
+    },
+    {
+      "epoch": 1.184757249194534,
+      "grad_norm": 0.37516269087791443,
+      "learning_rate": 0.0001,
+      "loss": 0.9011,
+      "step": 1333
+    },
+    {
+      "epoch": 1.1856460393289634,
+      "grad_norm": 0.3945169746875763,
+      "learning_rate": 0.0001,
+      "loss": 0.9384,
+      "step": 1334
+    },
+    {
+      "epoch": 1.186534829463393,
+      "grad_norm": 0.33570146560668945,
+      "learning_rate": 0.0001,
+      "loss": 0.9534,
+      "step": 1335
+    },
+    {
+      "epoch": 1.1874236195978225,
+      "grad_norm": 0.35922086238861084,
+      "learning_rate": 0.0001,
+      "loss": 0.9344,
+      "step": 1336
+    },
+    {
+      "epoch": 1.188312409732252,
+      "grad_norm": 0.33384042978286743,
+      "learning_rate": 0.0001,
+      "loss": 0.9778,
+      "step": 1337
+    },
+    {
+      "epoch": 1.1892011998666816,
+      "grad_norm": 0.3461971879005432,
+      "learning_rate": 0.0001,
+      "loss": 0.9317,
+      "step": 1338
+    },
+    {
+      "epoch": 1.190089990001111,
+      "grad_norm": 0.366767019033432,
+      "learning_rate": 0.0001,
+      "loss": 0.923,
+      "step": 1339
+    },
+    {
+      "epoch": 1.1909787801355405,
+      "grad_norm": 0.35670191049575806,
+      "learning_rate": 0.0001,
+      "loss": 0.8958,
+      "step": 1340
+    },
+    {
+      "epoch": 1.19186757026997,
+      "grad_norm": 0.3564472496509552,
+      "learning_rate": 0.0001,
+      "loss": 0.8879,
+      "step": 1341
+    },
+    {
+      "epoch": 1.1927563604043996,
+      "grad_norm": 0.34850582480430603,
+      "learning_rate": 0.0001,
+      "loss": 0.9472,
+      "step": 1342
+    },
+    {
+      "epoch": 1.193645150538829,
+      "grad_norm": 0.3118721842765808,
+      "learning_rate": 0.0001,
+      "loss": 0.9182,
+      "step": 1343
+    },
+    {
+      "epoch": 1.1945339406732585,
+      "grad_norm": 0.4474131464958191,
+      "learning_rate": 0.0001,
+      "loss": 0.9986,
+      "step": 1344
+    },
+    {
+      "epoch": 1.1954227308076881,
+      "grad_norm": 0.4048672616481781,
+      "learning_rate": 0.0001,
+      "loss": 0.924,
+      "step": 1345
+    },
+    {
+      "epoch": 1.1963115209421176,
+      "grad_norm": 0.3493656814098358,
+      "learning_rate": 0.0001,
+      "loss": 0.8614,
+      "step": 1346
+    },
+    {
+      "epoch": 1.197200311076547,
+      "grad_norm": 0.32247281074523926,
+      "learning_rate": 0.0001,
+      "loss": 0.8659,
+      "step": 1347
+    },
+    {
+      "epoch": 1.1980891012109764,
+      "grad_norm": 0.4119790196418762,
+      "learning_rate": 0.0001,
+      "loss": 0.9521,
+      "step": 1348
+    },
+    {
+      "epoch": 1.198977891345406,
+      "grad_norm": 0.40390363335609436,
+      "learning_rate": 0.0001,
+      "loss": 0.9203,
+      "step": 1349
+    },
+    {
+      "epoch": 1.1998666814798356,
+      "grad_norm": 0.40085315704345703,
+      "learning_rate": 0.0001,
+      "loss": 1.0135,
+      "step": 1350
+    },
+    {
+      "epoch": 1.200755471614265,
+      "grad_norm": 0.3355647325515747,
+      "learning_rate": 0.0001,
+      "loss": 0.9248,
+      "step": 1351
+    },
+    {
+      "epoch": 1.2016442617486947,
+      "grad_norm": 0.3319724500179291,
+      "learning_rate": 0.0001,
+      "loss": 0.8671,
+      "step": 1352
+    },
+    {
+      "epoch": 1.202533051883124,
+      "grad_norm": 0.4024242162704468,
+      "learning_rate": 0.0001,
+      "loss": 0.9497,
+      "step": 1353
+    },
+    {
+      "epoch": 1.2034218420175535,
+      "grad_norm": 0.3751072287559509,
+      "learning_rate": 0.0001,
+      "loss": 0.9692,
+      "step": 1354
+    },
+    {
+      "epoch": 1.2043106321519832,
+      "grad_norm": 0.3578762114048004,
+      "learning_rate": 0.0001,
+      "loss": 0.9306,
+      "step": 1355
+    },
+    {
+      "epoch": 1.2051994222864126,
+      "grad_norm": 0.41604581475257874,
+      "learning_rate": 0.0001,
+      "loss": 0.9746,
+      "step": 1356
+    },
+    {
+      "epoch": 1.206088212420842,
+      "grad_norm": 0.38614901900291443,
+      "learning_rate": 0.0001,
+      "loss": 0.9475,
+      "step": 1357
+    },
+    {
+      "epoch": 1.2069770025552717,
+      "grad_norm": 0.3261745274066925,
+      "learning_rate": 0.0001,
+      "loss": 0.9745,
+      "step": 1358
+    },
+    {
+      "epoch": 1.2078657926897012,
+      "grad_norm": 0.34567147493362427,
+      "learning_rate": 0.0001,
+      "loss": 0.9237,
+      "step": 1359
+    },
+    {
+      "epoch": 1.2087545828241306,
+      "grad_norm": 0.3168199360370636,
+      "learning_rate": 0.0001,
+      "loss": 0.9789,
+      "step": 1360
+    },
+    {
+      "epoch": 1.2096433729585603,
+      "grad_norm": 0.32394152879714966,
+      "learning_rate": 0.0001,
+      "loss": 0.9087,
+      "step": 1361
+    },
+    {
+      "epoch": 1.2105321630929897,
+      "grad_norm": 0.41741707921028137,
+      "learning_rate": 0.0001,
+      "loss": 0.9361,
+      "step": 1362
+    },
+    {
+      "epoch": 1.2114209532274192,
+      "grad_norm": 0.37223905324935913,
+      "learning_rate": 0.0001,
+      "loss": 0.9193,
+      "step": 1363
+    },
+    {
+      "epoch": 1.2123097433618486,
+      "grad_norm": 0.3394053876399994,
+      "learning_rate": 0.0001,
+      "loss": 0.9014,
+      "step": 1364
+    },
+    {
+      "epoch": 1.2131985334962783,
+      "grad_norm": 0.43821749091148376,
+      "learning_rate": 0.0001,
+      "loss": 0.9964,
+      "step": 1365
+    },
+    {
+      "epoch": 1.2140873236307077,
+      "grad_norm": 0.3783372938632965,
+      "learning_rate": 0.0001,
+      "loss": 0.9388,
+      "step": 1366
+    },
+    {
+      "epoch": 1.2149761137651371,
+      "grad_norm": 0.3495936691761017,
+      "learning_rate": 0.0001,
+      "loss": 0.8109,
+      "step": 1367
+    },
+    {
+      "epoch": 1.2158649038995668,
+      "grad_norm": 0.3929634690284729,
+      "learning_rate": 0.0001,
+      "loss": 0.9825,
+      "step": 1368
+    },
+    {
+      "epoch": 1.2167536940339962,
+      "grad_norm": 0.3742099702358246,
+      "learning_rate": 0.0001,
+      "loss": 0.9098,
+      "step": 1369
+    },
+    {
+      "epoch": 1.2176424841684257,
+      "grad_norm": 0.36973753571510315,
+      "learning_rate": 0.0001,
+      "loss": 0.9053,
+      "step": 1370
+    },
+    {
+      "epoch": 1.2185312743028551,
+      "grad_norm": 0.362594872713089,
+      "learning_rate": 0.0001,
+      "loss": 0.9906,
+      "step": 1371
+    },
+    {
+      "epoch": 1.2194200644372848,
+      "grad_norm": 0.442941278219223,
+      "learning_rate": 0.0001,
+      "loss": 1.0028,
+      "step": 1372
+    },
+    {
+      "epoch": 1.2203088545717142,
+      "grad_norm": 0.41126692295074463,
+      "learning_rate": 0.0001,
+      "loss": 0.9192,
+      "step": 1373
+    },
+    {
+      "epoch": 1.2211976447061437,
+      "grad_norm": 0.3734437823295593,
+      "learning_rate": 0.0001,
+      "loss": 0.919,
+      "step": 1374
+    },
+    {
+      "epoch": 1.2220864348405733,
+      "grad_norm": 0.37153056263923645,
+      "learning_rate": 0.0001,
+      "loss": 0.9253,
+      "step": 1375
+    },
+    {
+      "epoch": 1.2229752249750028,
+      "grad_norm": 0.4021115303039551,
+      "learning_rate": 0.0001,
+      "loss": 0.9013,
+      "step": 1376
+    },
+    {
+      "epoch": 1.2238640151094322,
+      "grad_norm": 0.3954538404941559,
+      "learning_rate": 0.0001,
+      "loss": 0.9907,
+      "step": 1377
+    },
+    {
+      "epoch": 1.2247528052438619,
+      "grad_norm": 0.707315981388092,
+      "learning_rate": 0.0001,
+      "loss": 0.9124,
+      "step": 1378
+    },
+    {
+      "epoch": 1.2256415953782913,
+      "grad_norm": 0.2907516658306122,
+      "learning_rate": 0.0001,
+      "loss": 0.8792,
+      "step": 1379
+    },
+    {
+      "epoch": 1.2265303855127208,
+      "grad_norm": 0.32730191946029663,
+      "learning_rate": 0.0001,
+      "loss": 0.8724,
+      "step": 1380
+    },
+    {
+      "epoch": 1.2274191756471504,
+      "grad_norm": 0.3393136262893677,
+      "learning_rate": 0.0001,
+      "loss": 0.9822,
+      "step": 1381
+    },
+    {
+      "epoch": 1.2283079657815799,
+      "grad_norm": 0.3186226189136505,
+      "learning_rate": 0.0001,
+      "loss": 0.9026,
+      "step": 1382
+    },
+    {
+      "epoch": 1.2291967559160093,
+      "grad_norm": 0.3546365797519684,
+      "learning_rate": 0.0001,
+      "loss": 0.9419,
+      "step": 1383
+    },
+    {
+      "epoch": 1.230085546050439,
+      "grad_norm": 0.33756113052368164,
+      "learning_rate": 0.0001,
+      "loss": 0.9292,
+      "step": 1384
+    },
+    {
+      "epoch": 1.2309743361848684,
+      "grad_norm": 0.357030987739563,
+      "learning_rate": 0.0001,
+      "loss": 0.8748,
+      "step": 1385
+    },
+    {
+      "epoch": 1.2318631263192978,
+      "grad_norm": 0.3271646797657013,
+      "learning_rate": 0.0001,
+      "loss": 0.9411,
+      "step": 1386
+    },
+    {
+      "epoch": 1.2327519164537273,
+      "grad_norm": 0.3888196647167206,
+      "learning_rate": 0.0001,
+      "loss": 0.9634,
+      "step": 1387
+    },
+    {
+      "epoch": 1.233640706588157,
+      "grad_norm": 0.3897395730018616,
+      "learning_rate": 0.0001,
+      "loss": 0.9404,
+      "step": 1388
+    },
+    {
+      "epoch": 1.2345294967225864,
+      "grad_norm": 0.36448606848716736,
+      "learning_rate": 0.0001,
+      "loss": 0.9543,
+      "step": 1389
+    },
+    {
+      "epoch": 1.2354182868570158,
+      "grad_norm": 0.35686194896698,
+      "learning_rate": 0.0001,
+      "loss": 0.8772,
+      "step": 1390
+    },
+    {
+      "epoch": 1.2363070769914455,
+      "grad_norm": 0.3667612373828888,
+      "learning_rate": 0.0001,
+      "loss": 0.8449,
+      "step": 1391
+    },
+    {
+      "epoch": 1.237195867125875,
+      "grad_norm": 0.35356754064559937,
+      "learning_rate": 0.0001,
+      "loss": 0.9099,
+      "step": 1392
+    },
+    {
+      "epoch": 1.2380846572603044,
+      "grad_norm": 0.3918624818325043,
+      "learning_rate": 0.0001,
+      "loss": 0.9164,
+      "step": 1393
+    },
+    {
+      "epoch": 1.2389734473947338,
+      "grad_norm": 0.347768634557724,
+      "learning_rate": 0.0001,
+      "loss": 0.9423,
+      "step": 1394
+    },
+    {
+      "epoch": 1.2398622375291635,
+      "grad_norm": 0.3811168372631073,
+      "learning_rate": 0.0001,
+      "loss": 1.0301,
+      "step": 1395
+    },
+    {
+      "epoch": 1.240751027663593,
+      "grad_norm": 0.3333447575569153,
+      "learning_rate": 0.0001,
+      "loss": 0.9424,
+      "step": 1396
+    },
+    {
+      "epoch": 1.2416398177980223,
+      "grad_norm": 0.3668173551559448,
+      "learning_rate": 0.0001,
+      "loss": 0.9513,
+      "step": 1397
+    },
+    {
+      "epoch": 1.242528607932452,
+      "grad_norm": 0.4245815575122833,
+      "learning_rate": 0.0001,
+      "loss": 0.8381,
+      "step": 1398
+    },
+    {
+      "epoch": 1.2434173980668815,
+      "grad_norm": 0.38784492015838623,
+      "learning_rate": 0.0001,
+      "loss": 0.9348,
+      "step": 1399
+    },
+    {
+      "epoch": 1.244306188201311,
+      "grad_norm": 0.3626404404640198,
+      "learning_rate": 0.0001,
+      "loss": 0.891,
+      "step": 1400
+    },
+    {
+      "epoch": 1.2451949783357406,
+      "grad_norm": 0.3952024579048157,
+      "learning_rate": 0.0001,
+      "loss": 0.8988,
+      "step": 1401
+    },
+    {
+      "epoch": 1.24608376847017,
+      "grad_norm": 0.3714921474456787,
+      "learning_rate": 0.0001,
+      "loss": 0.9597,
+      "step": 1402
+    },
+    {
+      "epoch": 1.2469725586045994,
+      "grad_norm": 0.3674130141735077,
+      "learning_rate": 0.0001,
+      "loss": 0.8814,
+      "step": 1403
+    },
+    {
+      "epoch": 1.247861348739029,
+      "grad_norm": 0.332264244556427,
+      "learning_rate": 0.0001,
+      "loss": 0.8723,
+      "step": 1404
+    },
+    {
+      "epoch": 1.2487501388734585,
+      "grad_norm": 0.44990214705467224,
+      "learning_rate": 0.0001,
+      "loss": 0.8562,
+      "step": 1405
+    },
+    {
+      "epoch": 1.249638929007888,
+      "grad_norm": 0.3691662847995758,
+      "learning_rate": 0.0001,
+      "loss": 0.9808,
+      "step": 1406
+    },
+    {
+      "epoch": 1.2505277191423176,
+      "grad_norm": 0.4463985562324524,
+      "learning_rate": 0.0001,
+      "loss": 1.0113,
+      "step": 1407
+    },
+    {
+      "epoch": 1.251416509276747,
+      "grad_norm": 0.40863966941833496,
+      "learning_rate": 0.0001,
+      "loss": 0.8999,
+      "step": 1408
+    },
+    {
+      "epoch": 1.2523052994111765,
+      "grad_norm": 0.39324069023132324,
+      "learning_rate": 0.0001,
+      "loss": 0.9607,
+      "step": 1409
+    },
+    {
+      "epoch": 1.2531940895456062,
+      "grad_norm": 0.4144541919231415,
+      "learning_rate": 0.0001,
+      "loss": 0.9257,
+      "step": 1410
+    },
+    {
+      "epoch": 1.2540828796800356,
+      "grad_norm": 0.3502121567726135,
+      "learning_rate": 0.0001,
+      "loss": 0.9035,
+      "step": 1411
+    },
+    {
+      "epoch": 1.254971669814465,
+      "grad_norm": 0.3595639765262604,
+      "learning_rate": 0.0001,
+      "loss": 0.958,
+      "step": 1412
+    },
+    {
+      "epoch": 1.2558604599488945,
+      "grad_norm": 0.31337282061576843,
+      "learning_rate": 0.0001,
+      "loss": 0.8338,
+      "step": 1413
+    },
+    {
+      "epoch": 1.256749250083324,
+      "grad_norm": 0.38238781690597534,
+      "learning_rate": 0.0001,
+      "loss": 0.9705,
+      "step": 1414
+    },
+    {
+      "epoch": 1.2576380402177536,
+      "grad_norm": 0.34189799427986145,
+      "learning_rate": 0.0001,
+      "loss": 0.9685,
+      "step": 1415
+    },
+    {
+      "epoch": 1.258526830352183,
+      "grad_norm": 0.3783426582813263,
+      "learning_rate": 0.0001,
+      "loss": 0.8579,
+      "step": 1416
+    },
+    {
+      "epoch": 1.2594156204866125,
+      "grad_norm": 0.35170289874076843,
+      "learning_rate": 0.0001,
+      "loss": 1.0055,
+      "step": 1417
+    },
+    {
+      "epoch": 1.2603044106210421,
+      "grad_norm": 0.3678469657897949,
+      "learning_rate": 0.0001,
+      "loss": 0.9932,
+      "step": 1418
+    },
+    {
+      "epoch": 1.2611932007554716,
+      "grad_norm": 0.35460302233695984,
+      "learning_rate": 0.0001,
+      "loss": 0.8314,
+      "step": 1419
+    },
+    {
+      "epoch": 1.262081990889901,
+      "grad_norm": 0.3786843419075012,
+      "learning_rate": 0.0001,
+      "loss": 0.9391,
+      "step": 1420
+    },
+    {
+      "epoch": 1.2629707810243307,
+      "grad_norm": 0.37704455852508545,
+      "learning_rate": 0.0001,
+      "loss": 0.9314,
+      "step": 1421
+    },
+    {
+      "epoch": 1.2638595711587601,
+      "grad_norm": 0.3811683654785156,
+      "learning_rate": 0.0001,
+      "loss": 0.9184,
+      "step": 1422
+    },
+    {
+      "epoch": 1.2647483612931896,
+      "grad_norm": 0.3323057293891907,
+      "learning_rate": 0.0001,
+      "loss": 0.8478,
+      "step": 1423
+    },
+    {
+      "epoch": 1.2656371514276192,
+      "grad_norm": 0.33943992853164673,
+      "learning_rate": 0.0001,
+      "loss": 0.7872,
+      "step": 1424
+    },
+    {
+      "epoch": 1.2665259415620487,
+      "grad_norm": 0.370331346988678,
+      "learning_rate": 0.0001,
+      "loss": 1.0004,
+      "step": 1425
+    },
+    {
+      "epoch": 1.2674147316964781,
+      "grad_norm": 0.3509312868118286,
+      "learning_rate": 0.0001,
+      "loss": 0.8777,
+      "step": 1426
+    },
+    {
+      "epoch": 1.2683035218309078,
+      "grad_norm": 0.35262539982795715,
+      "learning_rate": 0.0001,
+      "loss": 0.8772,
+      "step": 1427
+    },
+    {
+      "epoch": 1.2691923119653372,
+      "grad_norm": 0.3514516055583954,
+      "learning_rate": 0.0001,
+      "loss": 0.9092,
+      "step": 1428
+    },
+    {
+      "epoch": 1.2700811020997667,
+      "grad_norm": 0.33917155861854553,
+      "learning_rate": 0.0001,
+      "loss": 0.9088,
+      "step": 1429
+    },
+    {
+      "epoch": 1.2709698922341963,
+      "grad_norm": 0.37379592657089233,
+      "learning_rate": 0.0001,
+      "loss": 0.8811,
+      "step": 1430
+    },
+    {
+      "epoch": 1.2718586823686258,
+      "grad_norm": 0.37309855222702026,
+      "learning_rate": 0.0001,
+      "loss": 0.9243,
+      "step": 1431
+    },
+    {
+      "epoch": 1.2727474725030552,
+      "grad_norm": 0.3677520751953125,
+      "learning_rate": 0.0001,
+      "loss": 0.9191,
+      "step": 1432
+    },
+    {
+      "epoch": 1.2736362626374849,
+      "grad_norm": 0.3768025040626526,
+      "learning_rate": 0.0001,
+      "loss": 0.9713,
+      "step": 1433
+    },
+    {
+      "epoch": 1.2745250527719143,
+      "grad_norm": 0.3794255256652832,
+      "learning_rate": 0.0001,
+      "loss": 0.8944,
+      "step": 1434
+    },
+    {
+      "epoch": 1.2754138429063437,
+      "grad_norm": 0.3847828805446625,
+      "learning_rate": 0.0001,
+      "loss": 0.9895,
+      "step": 1435
+    },
+    {
+      "epoch": 1.2763026330407732,
+      "grad_norm": 0.3901936411857605,
+      "learning_rate": 0.0001,
+      "loss": 0.9604,
+      "step": 1436
+    },
+    {
+      "epoch": 1.2771914231752026,
+      "grad_norm": 0.36127346754074097,
+      "learning_rate": 0.0001,
+      "loss": 0.9633,
+      "step": 1437
+    },
+    {
+      "epoch": 1.2780802133096323,
+      "grad_norm": 0.3554041385650635,
+      "learning_rate": 0.0001,
+      "loss": 0.9282,
+      "step": 1438
+    },
+    {
+      "epoch": 1.2789690034440617,
+      "grad_norm": 0.3366556465625763,
+      "learning_rate": 0.0001,
+      "loss": 0.9398,
+      "step": 1439
+    },
+    {
+      "epoch": 1.2798577935784912,
+      "grad_norm": 0.38084515929222107,
+      "learning_rate": 0.0001,
+      "loss": 1.0101,
+      "step": 1440
+    },
+    {
+      "epoch": 1.2807465837129208,
+      "grad_norm": 0.35212430357933044,
+      "learning_rate": 0.0001,
+      "loss": 0.921,
+      "step": 1441
+    },
+    {
+      "epoch": 1.2816353738473503,
+      "grad_norm": 0.37851765751838684,
+      "learning_rate": 0.0001,
+      "loss": 0.8989,
+      "step": 1442
+    },
+    {
+      "epoch": 1.2825241639817797,
+      "grad_norm": 0.31411126255989075,
+      "learning_rate": 0.0001,
+      "loss": 0.8476,
+      "step": 1443
+    },
+    {
+      "epoch": 1.2834129541162094,
+      "grad_norm": 0.35545769333839417,
+      "learning_rate": 0.0001,
+      "loss": 0.9857,
+      "step": 1444
+    },
+    {
+      "epoch": 1.2843017442506388,
+      "grad_norm": 0.36752983927726746,
+      "learning_rate": 0.0001,
+      "loss": 0.8876,
+      "step": 1445
+    },
+    {
+      "epoch": 1.2851905343850683,
+      "grad_norm": 0.3653068542480469,
+      "learning_rate": 0.0001,
+      "loss": 0.8651,
+      "step": 1446
+    },
+    {
+      "epoch": 1.286079324519498,
+      "grad_norm": 0.3388923704624176,
+      "learning_rate": 0.0001,
+      "loss": 0.8922,
+      "step": 1447
+    },
+    {
+      "epoch": 1.2869681146539274,
+      "grad_norm": 0.3574695587158203,
+      "learning_rate": 0.0001,
+      "loss": 0.903,
+      "step": 1448
+    },
+    {
+      "epoch": 1.2878569047883568,
+      "grad_norm": 0.36422044038772583,
+      "learning_rate": 0.0001,
+      "loss": 0.9122,
+      "step": 1449
+    },
+    {
+      "epoch": 1.2887456949227865,
+      "grad_norm": 0.3757425844669342,
+      "learning_rate": 0.0001,
+      "loss": 0.9014,
+      "step": 1450
+    },
+    {
+      "epoch": 1.289634485057216,
+      "grad_norm": 0.3608701229095459,
+      "learning_rate": 0.0001,
+      "loss": 0.9089,
+      "step": 1451
+    },
+    {
+      "epoch": 1.2905232751916453,
+      "grad_norm": 0.3886134922504425,
+      "learning_rate": 0.0001,
+      "loss": 0.817,
+      "step": 1452
+    },
+    {
+      "epoch": 1.291412065326075,
+      "grad_norm": 0.33738628029823303,
+      "learning_rate": 0.0001,
+      "loss": 0.8872,
+      "step": 1453
+    },
+    {
+      "epoch": 1.2923008554605044,
+      "grad_norm": 0.3938889801502228,
+      "learning_rate": 0.0001,
+      "loss": 0.8523,
+      "step": 1454
+    },
+    {
+      "epoch": 1.2931896455949339,
+      "grad_norm": 0.37151628732681274,
+      "learning_rate": 0.0001,
+      "loss": 0.9448,
+      "step": 1455
+    },
+    {
+      "epoch": 1.2940784357293635,
+      "grad_norm": 0.36321815848350525,
+      "learning_rate": 0.0001,
+      "loss": 0.9511,
+      "step": 1456
+    },
+    {
+      "epoch": 1.294967225863793,
+      "grad_norm": 0.35245969891548157,
+      "learning_rate": 0.0001,
+      "loss": 1.0141,
+      "step": 1457
+    },
+    {
+      "epoch": 1.2958560159982224,
+      "grad_norm": 0.3494570851325989,
+      "learning_rate": 0.0001,
+      "loss": 0.9411,
+      "step": 1458
+    },
+    {
+      "epoch": 1.2967448061326519,
+      "grad_norm": 0.3875609338283539,
+      "learning_rate": 0.0001,
+      "loss": 0.9318,
+      "step": 1459
+    },
+    {
+      "epoch": 1.2976335962670813,
+      "grad_norm": 0.3761698305606842,
+      "learning_rate": 0.0001,
+      "loss": 0.8563,
+      "step": 1460
+    },
+    {
+      "epoch": 1.298522386401511,
+      "grad_norm": 0.3185148239135742,
+      "learning_rate": 0.0001,
+      "loss": 0.9486,
+      "step": 1461
+    },
+    {
+      "epoch": 1.2994111765359404,
+      "grad_norm": 0.3480708599090576,
+      "learning_rate": 0.0001,
+      "loss": 0.8733,
+      "step": 1462
+    },
+    {
+      "epoch": 1.3002999666703698,
+      "grad_norm": 0.3743402659893036,
+      "learning_rate": 0.0001,
+      "loss": 0.9778,
+      "step": 1463
+    },
+    {
+      "epoch": 1.3011887568047995,
+      "grad_norm": 0.4024488031864166,
+      "learning_rate": 0.0001,
+      "loss": 0.8907,
+      "step": 1464
+    },
+    {
+      "epoch": 1.302077546939229,
+      "grad_norm": 0.366472452878952,
+      "learning_rate": 0.0001,
+      "loss": 0.9898,
+      "step": 1465
+    },
+    {
+      "epoch": 1.3029663370736584,
+      "grad_norm": 0.36721619963645935,
+      "learning_rate": 0.0001,
+      "loss": 0.8577,
+      "step": 1466
+    },
+    {
+      "epoch": 1.303855127208088,
+      "grad_norm": 0.36933907866477966,
+      "learning_rate": 0.0001,
+      "loss": 0.9365,
+      "step": 1467
+    },
+    {
+      "epoch": 1.3047439173425175,
+      "grad_norm": 0.36210325360298157,
+      "learning_rate": 0.0001,
+      "loss": 0.8787,
+      "step": 1468
+    },
+    {
+      "epoch": 1.305632707476947,
+      "grad_norm": 0.3773205876350403,
+      "learning_rate": 0.0001,
+      "loss": 0.9396,
+      "step": 1469
+    },
+    {
+      "epoch": 1.3065214976113766,
+      "grad_norm": 0.3904916048049927,
+      "learning_rate": 0.0001,
+      "loss": 0.9872,
+      "step": 1470
+    },
+    {
+      "epoch": 1.307410287745806,
+      "grad_norm": 0.38779687881469727,
+      "learning_rate": 0.0001,
+      "loss": 0.9682,
+      "step": 1471
+    },
+    {
+      "epoch": 1.3082990778802355,
+      "grad_norm": 0.3612334728240967,
+      "learning_rate": 0.0001,
+      "loss": 0.8867,
+      "step": 1472
+    },
+    {
+      "epoch": 1.3091878680146651,
+      "grad_norm": 0.4181813597679138,
+      "learning_rate": 0.0001,
+      "loss": 0.9666,
+      "step": 1473
+    },
+    {
+      "epoch": 1.3100766581490946,
+      "grad_norm": 0.4051264524459839,
+      "learning_rate": 0.0001,
+      "loss": 0.9088,
+      "step": 1474
+    },
+    {
+      "epoch": 1.310965448283524,
+      "grad_norm": 0.3271094262599945,
+      "learning_rate": 0.0001,
+      "loss": 0.9018,
+      "step": 1475
+    },
+    {
+      "epoch": 1.3118542384179537,
+      "grad_norm": 0.35540738701820374,
+      "learning_rate": 0.0001,
+      "loss": 1.0157,
+      "step": 1476
+    },
+    {
+      "epoch": 1.3127430285523831,
+      "grad_norm": 0.36050257086753845,
+      "learning_rate": 0.0001,
+      "loss": 0.9347,
+      "step": 1477
+    },
+    {
+      "epoch": 1.3136318186868126,
+      "grad_norm": 0.39505940675735474,
+      "learning_rate": 0.0001,
+      "loss": 0.9321,
+      "step": 1478
+    },
+    {
+      "epoch": 1.3145206088212422,
+      "grad_norm": 0.3984764516353607,
+      "learning_rate": 0.0001,
+      "loss": 0.9472,
+      "step": 1479
+    },
+    {
+      "epoch": 1.3154093989556717,
+      "grad_norm": 0.3784257769584656,
+      "learning_rate": 0.0001,
+      "loss": 0.9075,
+      "step": 1480
+    },
+    {
+      "epoch": 1.316298189090101,
+      "grad_norm": 0.35195082426071167,
+      "learning_rate": 0.0001,
+      "loss": 0.9514,
+      "step": 1481
+    },
+    {
+      "epoch": 1.3171869792245305,
+      "grad_norm": 0.33862170577049255,
+      "learning_rate": 0.0001,
+      "loss": 0.9411,
+      "step": 1482
+    },
+    {
+      "epoch": 1.31807576935896,
+      "grad_norm": 0.3316287696361542,
+      "learning_rate": 0.0001,
+      "loss": 0.9368,
+      "step": 1483
+    },
+    {
+      "epoch": 1.3189645594933896,
+      "grad_norm": 0.3842187225818634,
+      "learning_rate": 0.0001,
+      "loss": 0.9396,
+      "step": 1484
+    },
+    {
+      "epoch": 1.319853349627819,
+      "grad_norm": 0.39101001620292664,
+      "learning_rate": 0.0001,
+      "loss": 0.9329,
+      "step": 1485
+    },
+    {
+      "epoch": 1.3207421397622485,
+      "grad_norm": 0.3348684012889862,
+      "learning_rate": 0.0001,
+      "loss": 0.8891,
+      "step": 1486
+    },
+    {
+      "epoch": 1.3216309298966782,
+      "grad_norm": 0.35855552554130554,
+      "learning_rate": 0.0001,
+      "loss": 0.9148,
+      "step": 1487
+    },
+    {
+      "epoch": 1.3225197200311076,
+      "grad_norm": 0.39019593596458435,
+      "learning_rate": 0.0001,
+      "loss": 0.9043,
+      "step": 1488
+    },
+    {
+      "epoch": 1.323408510165537,
+      "grad_norm": 0.3681492805480957,
+      "learning_rate": 0.0001,
+      "loss": 0.9222,
+      "step": 1489
+    },
+    {
+      "epoch": 1.3242973002999667,
+      "grad_norm": 0.3557593524456024,
+      "learning_rate": 0.0001,
+      "loss": 0.8972,
+      "step": 1490
+    },
+    {
+      "epoch": 1.3251860904343962,
+      "grad_norm": 0.3503437638282776,
+      "learning_rate": 0.0001,
+      "loss": 0.9307,
+      "step": 1491
+    },
+    {
+      "epoch": 1.3260748805688256,
+      "grad_norm": 0.31487593054771423,
+      "learning_rate": 0.0001,
+      "loss": 0.899,
+      "step": 1492
+    },
+    {
+      "epoch": 1.3269636707032553,
+      "grad_norm": 0.3774222433567047,
+      "learning_rate": 0.0001,
+      "loss": 0.8246,
+      "step": 1493
+    },
+    {
+      "epoch": 1.3278524608376847,
+      "grad_norm": 0.3760510981082916,
+      "learning_rate": 0.0001,
+      "loss": 1.0001,
+      "step": 1494
+    },
+    {
+      "epoch": 1.3287412509721142,
+      "grad_norm": 0.3963356912136078,
+      "learning_rate": 0.0001,
+      "loss": 0.9093,
+      "step": 1495
+    },
+    {
+      "epoch": 1.3296300411065438,
+      "grad_norm": 0.3808361887931824,
+      "learning_rate": 0.0001,
+      "loss": 0.9258,
+      "step": 1496
+    },
+    {
+      "epoch": 1.3305188312409733,
+      "grad_norm": 0.39797237515449524,
+      "learning_rate": 0.0001,
+      "loss": 0.9168,
+      "step": 1497
+    },
+    {
+      "epoch": 1.3314076213754027,
+      "grad_norm": 0.36626359820365906,
+      "learning_rate": 0.0001,
+      "loss": 0.9289,
+      "step": 1498
+    },
+    {
+      "epoch": 1.3322964115098324,
+      "grad_norm": 0.38699278235435486,
+      "learning_rate": 0.0001,
+      "loss": 0.8827,
+      "step": 1499
+    },
+    {
+      "epoch": 1.3331852016442618,
+      "grad_norm": 0.37499698996543884,
+      "learning_rate": 0.0001,
+      "loss": 0.9254,
+      "step": 1500
+    },
+    {
+      "epoch": 1.3340739917786912,
+      "grad_norm": 0.34208056330680847,
+      "learning_rate": 0.0001,
+      "loss": 0.8792,
+      "step": 1501
+    },
+    {
+      "epoch": 1.334962781913121,
+      "grad_norm": 0.3909793496131897,
+      "learning_rate": 0.0001,
+      "loss": 0.9986,
+      "step": 1502
+    },
+    {
+      "epoch": 1.3358515720475503,
+      "grad_norm": 0.35674503445625305,
+      "learning_rate": 0.0001,
+      "loss": 0.8902,
+      "step": 1503
+    },
+    {
+      "epoch": 1.3367403621819798,
+      "grad_norm": 0.35735827684402466,
+      "learning_rate": 0.0001,
+      "loss": 0.9898,
+      "step": 1504
+    },
+    {
+      "epoch": 1.3376291523164092,
+      "grad_norm": 0.39872074127197266,
+      "learning_rate": 0.0001,
+      "loss": 1.048,
+      "step": 1505
+    },
+    {
+      "epoch": 1.3385179424508387,
+      "grad_norm": 0.4395754039287567,
+      "learning_rate": 0.0001,
+      "loss": 0.8621,
+      "step": 1506
+    },
+    {
+      "epoch": 1.3394067325852683,
+      "grad_norm": 0.3650192320346832,
+      "learning_rate": 0.0001,
+      "loss": 0.8788,
+      "step": 1507
+    },
+    {
+      "epoch": 1.3402955227196978,
+      "grad_norm": 0.35158872604370117,
+      "learning_rate": 0.0001,
+      "loss": 0.8228,
+      "step": 1508
+    },
+    {
+      "epoch": 1.3411843128541272,
+      "grad_norm": 0.3553421199321747,
+      "learning_rate": 0.0001,
+      "loss": 0.9102,
+      "step": 1509
+    },
+    {
+      "epoch": 1.3420731029885569,
+      "grad_norm": 0.32123276591300964,
+      "learning_rate": 0.0001,
+      "loss": 0.8808,
+      "step": 1510
+    },
+    {
+      "epoch": 1.3429618931229863,
+      "grad_norm": 0.35363566875457764,
+      "learning_rate": 0.0001,
+      "loss": 0.8895,
+      "step": 1511
+    },
+    {
+      "epoch": 1.3438506832574157,
+      "grad_norm": 0.38487884402275085,
+      "learning_rate": 0.0001,
+      "loss": 0.8713,
+      "step": 1512
+    },
+    {
+      "epoch": 1.3447394733918454,
+      "grad_norm": 0.3799397051334381,
+      "learning_rate": 0.0001,
+      "loss": 0.8982,
+      "step": 1513
+    },
+    {
+      "epoch": 1.3456282635262748,
+      "grad_norm": 0.3352080285549164,
+      "learning_rate": 0.0001,
+      "loss": 0.914,
+      "step": 1514
+    },
+    {
+      "epoch": 1.3465170536607043,
+      "grad_norm": 0.6567979454994202,
+      "learning_rate": 0.0001,
+      "loss": 0.978,
+      "step": 1515
+    },
+    {
+      "epoch": 1.347405843795134,
+      "grad_norm": 0.4132377505302429,
+      "learning_rate": 0.0001,
+      "loss": 1.0033,
+      "step": 1516
+    },
+    {
+      "epoch": 1.3482946339295634,
+      "grad_norm": 0.4519813060760498,
+      "learning_rate": 0.0001,
+      "loss": 0.9213,
+      "step": 1517
+    },
+    {
+      "epoch": 1.3491834240639928,
+      "grad_norm": 0.35851332545280457,
+      "learning_rate": 0.0001,
+      "loss": 0.9634,
+      "step": 1518
+    },
+    {
+      "epoch": 1.3500722141984225,
+      "grad_norm": 0.39293810725212097,
+      "learning_rate": 0.0001,
+      "loss": 0.9069,
+      "step": 1519
+    },
+    {
+      "epoch": 1.350961004332852,
+      "grad_norm": 0.36240848898887634,
+      "learning_rate": 0.0001,
+      "loss": 0.8621,
+      "step": 1520
+    },
+    {
+      "epoch": 1.3518497944672814,
+      "grad_norm": 0.3404124975204468,
+      "learning_rate": 0.0001,
+      "loss": 0.8964,
+      "step": 1521
+    },
+    {
+      "epoch": 1.352738584601711,
+      "grad_norm": 0.31596043705940247,
+      "learning_rate": 0.0001,
+      "loss": 0.9091,
+      "step": 1522
+    },
+    {
+      "epoch": 1.3536273747361405,
+      "grad_norm": 0.3596150279045105,
+      "learning_rate": 0.0001,
+      "loss": 0.9275,
+      "step": 1523
+    },
+    {
+      "epoch": 1.35451616487057,
+      "grad_norm": 0.35948145389556885,
+      "learning_rate": 0.0001,
+      "loss": 0.8821,
+      "step": 1524
+    },
+    {
+      "epoch": 1.3554049550049996,
+      "grad_norm": 0.3627282381057739,
+      "learning_rate": 0.0001,
+      "loss": 0.9501,
+      "step": 1525
+    },
+    {
+      "epoch": 1.356293745139429,
+      "grad_norm": 0.38337674736976624,
+      "learning_rate": 0.0001,
+      "loss": 0.8677,
+      "step": 1526
+    },
+    {
+      "epoch": 1.3571825352738585,
+      "grad_norm": 0.38964757323265076,
+      "learning_rate": 0.0001,
+      "loss": 0.9327,
+      "step": 1527
+    },
+    {
+      "epoch": 1.358071325408288,
+      "grad_norm": 0.3563820421695709,
+      "learning_rate": 0.0001,
+      "loss": 0.836,
+      "step": 1528
+    },
+    {
+      "epoch": 1.3589601155427173,
+      "grad_norm": 0.3765336573123932,
+      "learning_rate": 0.0001,
+      "loss": 0.9198,
+      "step": 1529
+    },
+    {
+      "epoch": 1.359848905677147,
+      "grad_norm": 0.3981674611568451,
+      "learning_rate": 0.0001,
+      "loss": 0.9539,
+      "step": 1530
+    },
+    {
+      "epoch": 1.3607376958115764,
+      "grad_norm": 0.4140075147151947,
+      "learning_rate": 0.0001,
+      "loss": 0.9104,
+      "step": 1531
+    },
+    {
+      "epoch": 1.3616264859460059,
+      "grad_norm": 0.4137401282787323,
+      "learning_rate": 0.0001,
+      "loss": 1.0583,
+      "step": 1532
+    },
+    {
+      "epoch": 1.3625152760804355,
+      "grad_norm": 0.3649982511997223,
+      "learning_rate": 0.0001,
+      "loss": 0.8334,
+      "step": 1533
+    },
+    {
+      "epoch": 1.363404066214865,
+      "grad_norm": 0.3175910711288452,
+      "learning_rate": 0.0001,
+      "loss": 0.8834,
+      "step": 1534
+    },
+    {
+      "epoch": 1.3642928563492944,
+      "grad_norm": 0.3742975890636444,
+      "learning_rate": 0.0001,
+      "loss": 0.9664,
+      "step": 1535
+    },
+    {
+      "epoch": 1.365181646483724,
+      "grad_norm": 0.369791179895401,
+      "learning_rate": 0.0001,
+      "loss": 0.8947,
+      "step": 1536
+    },
+    {
+      "epoch": 1.3660704366181535,
+      "grad_norm": 0.33241912722587585,
+      "learning_rate": 0.0001,
+      "loss": 0.9392,
+      "step": 1537
+    },
+    {
+      "epoch": 1.366959226752583,
+      "grad_norm": 0.345084011554718,
+      "learning_rate": 0.0001,
+      "loss": 0.8655,
+      "step": 1538
+    },
+    {
+      "epoch": 1.3678480168870126,
+      "grad_norm": 0.33112481236457825,
+      "learning_rate": 0.0001,
+      "loss": 0.9325,
+      "step": 1539
+    },
+    {
+      "epoch": 1.368736807021442,
+      "grad_norm": 0.3624488413333893,
+      "learning_rate": 0.0001,
+      "loss": 0.9253,
+      "step": 1540
+    },
+    {
+      "epoch": 1.3696255971558715,
+      "grad_norm": 0.3580976724624634,
+      "learning_rate": 0.0001,
+      "loss": 0.8796,
+      "step": 1541
+    },
+    {
+      "epoch": 1.3705143872903012,
+      "grad_norm": 0.30134427547454834,
+      "learning_rate": 0.0001,
+      "loss": 0.8049,
+      "step": 1542
+    },
+    {
+      "epoch": 1.3714031774247306,
+      "grad_norm": 0.3331443965435028,
+      "learning_rate": 0.0001,
+      "loss": 0.9014,
+      "step": 1543
+    },
+    {
+      "epoch": 1.37229196755916,
+      "grad_norm": 0.33350786566734314,
+      "learning_rate": 0.0001,
+      "loss": 0.9826,
+      "step": 1544
+    },
+    {
+      "epoch": 1.3731807576935897,
+      "grad_norm": 0.3651019036769867,
+      "learning_rate": 0.0001,
+      "loss": 0.9215,
+      "step": 1545
+    },
+    {
+      "epoch": 1.3740695478280192,
+      "grad_norm": 0.3742315173149109,
+      "learning_rate": 0.0001,
+      "loss": 0.9774,
+      "step": 1546
+    },
+    {
+      "epoch": 1.3749583379624486,
+      "grad_norm": 0.32286152243614197,
+      "learning_rate": 0.0001,
+      "loss": 0.8499,
+      "step": 1547
+    },
+    {
+      "epoch": 1.3758471280968783,
+      "grad_norm": 0.33028700947761536,
+      "learning_rate": 0.0001,
+      "loss": 0.8511,
+      "step": 1548
+    },
+    {
+      "epoch": 1.3767359182313077,
+      "grad_norm": 0.3529914617538452,
+      "learning_rate": 0.0001,
+      "loss": 0.8478,
+      "step": 1549
+    },
+    {
+      "epoch": 1.3776247083657371,
+      "grad_norm": 0.34889042377471924,
+      "learning_rate": 0.0001,
+      "loss": 0.8968,
+      "step": 1550
+    },
+    {
+      "epoch": 1.3785134985001666,
+      "grad_norm": 0.35926511883735657,
+      "learning_rate": 0.0001,
+      "loss": 0.9326,
+      "step": 1551
+    },
+    {
+      "epoch": 1.379402288634596,
+      "grad_norm": 0.3612864315509796,
+      "learning_rate": 0.0001,
+      "loss": 0.9212,
+      "step": 1552
+    },
+    {
+      "epoch": 1.3802910787690257,
+      "grad_norm": 0.342424601316452,
+      "learning_rate": 0.0001,
+      "loss": 0.8462,
+      "step": 1553
+    },
+    {
+      "epoch": 1.3811798689034551,
+      "grad_norm": 0.36990538239479065,
+      "learning_rate": 0.0001,
+      "loss": 0.8569,
+      "step": 1554
+    },
+    {
+      "epoch": 1.3820686590378846,
+      "grad_norm": 0.3396192491054535,
+      "learning_rate": 0.0001,
+      "loss": 0.8649,
+      "step": 1555
+    },
+    {
+      "epoch": 1.3829574491723142,
+      "grad_norm": 0.3837670087814331,
+      "learning_rate": 0.0001,
+      "loss": 0.8626,
+      "step": 1556
+    },
+    {
+      "epoch": 1.3838462393067437,
+      "grad_norm": 0.3661665618419647,
+      "learning_rate": 0.0001,
+      "loss": 0.9781,
+      "step": 1557
+    },
+    {
+      "epoch": 1.384735029441173,
+      "grad_norm": 0.41726961731910706,
+      "learning_rate": 0.0001,
+      "loss": 0.9522,
+      "step": 1558
+    },
+    {
+      "epoch": 1.3856238195756028,
+      "grad_norm": 0.4061615467071533,
+      "learning_rate": 0.0001,
+      "loss": 0.9127,
+      "step": 1559
+    },
+    {
+      "epoch": 1.3865126097100322,
+      "grad_norm": 0.32869914174079895,
+      "learning_rate": 0.0001,
+      "loss": 0.8947,
+      "step": 1560
+    },
+    {
+      "epoch": 1.3874013998444616,
+      "grad_norm": 0.6287943124771118,
+      "learning_rate": 0.0001,
+      "loss": 0.8759,
+      "step": 1561
+    },
+    {
+      "epoch": 1.3882901899788913,
+      "grad_norm": 0.562345027923584,
+      "learning_rate": 0.0001,
+      "loss": 0.9109,
+      "step": 1562
+    },
+    {
+      "epoch": 1.3891789801133207,
+      "grad_norm": 0.39289891719818115,
+      "learning_rate": 0.0001,
+      "loss": 0.9831,
+      "step": 1563
+    },
+    {
+      "epoch": 1.3900677702477502,
+      "grad_norm": 0.4826609194278717,
+      "learning_rate": 0.0001,
+      "loss": 0.8846,
+      "step": 1564
+    },
+    {
+      "epoch": 1.3909565603821799,
+      "grad_norm": 0.4759630262851715,
+      "learning_rate": 0.0001,
+      "loss": 0.7826,
+      "step": 1565
+    },
+    {
+      "epoch": 1.3918453505166093,
+      "grad_norm": 0.8236848711967468,
+      "learning_rate": 0.0001,
+      "loss": 0.9145,
+      "step": 1566
+    },
+    {
+      "epoch": 1.3927341406510387,
+      "grad_norm": 0.39659222960472107,
+      "learning_rate": 0.0001,
+      "loss": 0.9594,
+      "step": 1567
+    },
+    {
+      "epoch": 1.3936229307854684,
+      "grad_norm": 0.32191386818885803,
+      "learning_rate": 0.0001,
+      "loss": 0.8515,
+      "step": 1568
+    },
+    {
+      "epoch": 1.3945117209198978,
+      "grad_norm": 0.397210955619812,
+      "learning_rate": 0.0001,
+      "loss": 0.9207,
+      "step": 1569
+    },
+    {
+      "epoch": 1.3954005110543273,
+      "grad_norm": 0.36070799827575684,
+      "learning_rate": 0.0001,
+      "loss": 0.932,
+      "step": 1570
+    },
+    {
+      "epoch": 1.396289301188757,
+      "grad_norm": 0.35527974367141724,
+      "learning_rate": 0.0001,
+      "loss": 0.9069,
+      "step": 1571
+    },
+    {
+      "epoch": 1.3971780913231864,
+      "grad_norm": 0.39062851667404175,
+      "learning_rate": 0.0001,
+      "loss": 0.9763,
+      "step": 1572
+    },
+    {
+      "epoch": 1.3980668814576158,
+      "grad_norm": 0.3772708475589752,
+      "learning_rate": 0.0001,
+      "loss": 0.9299,
+      "step": 1573
+    },
+    {
+      "epoch": 1.3989556715920453,
+      "grad_norm": 0.3413572609424591,
+      "learning_rate": 0.0001,
+      "loss": 0.9314,
+      "step": 1574
+    },
+    {
+      "epoch": 1.3998444617264747,
+      "grad_norm": 0.4356358051300049,
+      "learning_rate": 0.0001,
+      "loss": 0.8741,
+      "step": 1575
+    },
+    {
+      "epoch": 1.4007332518609044,
+      "grad_norm": 0.3393174409866333,
+      "learning_rate": 0.0001,
+      "loss": 0.8657,
+      "step": 1576
+    },
+    {
+      "epoch": 1.4016220419953338,
+      "grad_norm": 0.35734716057777405,
+      "learning_rate": 0.0001,
+      "loss": 0.9329,
+      "step": 1577
+    },
+    {
+      "epoch": 1.4025108321297632,
+      "grad_norm": 0.3565090298652649,
+      "learning_rate": 0.0001,
+      "loss": 0.9528,
+      "step": 1578
+    },
+    {
+      "epoch": 1.403399622264193,
+      "grad_norm": 0.39361128211021423,
+      "learning_rate": 0.0001,
+      "loss": 0.8599,
+      "step": 1579
+    },
+    {
+      "epoch": 1.4042884123986223,
+      "grad_norm": 0.4015718102455139,
+      "learning_rate": 0.0001,
+      "loss": 0.8092,
+      "step": 1580
+    },
+    {
+      "epoch": 1.4051772025330518,
+      "grad_norm": 0.35963454842567444,
+      "learning_rate": 0.0001,
+      "loss": 0.9064,
+      "step": 1581
+    },
+    {
+      "epoch": 1.4060659926674814,
+      "grad_norm": 0.3792876601219177,
+      "learning_rate": 0.0001,
+      "loss": 0.9415,
+      "step": 1582
+    },
+    {
+      "epoch": 1.4069547828019109,
+      "grad_norm": 0.45550185441970825,
+      "learning_rate": 0.0001,
+      "loss": 0.9823,
+      "step": 1583
+    },
+    {
+      "epoch": 1.4078435729363403,
+      "grad_norm": 0.3728649318218231,
+      "learning_rate": 0.0001,
+      "loss": 0.9192,
+      "step": 1584
+    },
+    {
+      "epoch": 1.40873236307077,
+      "grad_norm": 0.35824286937713623,
+      "learning_rate": 0.0001,
+      "loss": 0.9029,
+      "step": 1585
+    },
+    {
+      "epoch": 1.4096211532051994,
+      "grad_norm": 0.34785544872283936,
+      "learning_rate": 0.0001,
+      "loss": 0.8923,
+      "step": 1586
+    },
+    {
+      "epoch": 1.4105099433396289,
+      "grad_norm": 0.3750496208667755,
+      "learning_rate": 0.0001,
+      "loss": 0.97,
+      "step": 1587
+    },
+    {
+      "epoch": 1.4113987334740585,
+      "grad_norm": 0.365651398897171,
+      "learning_rate": 0.0001,
+      "loss": 0.9155,
+      "step": 1588
+    },
+    {
+      "epoch": 1.412287523608488,
+      "grad_norm": 0.389883428812027,
+      "learning_rate": 0.0001,
+      "loss": 0.8792,
+      "step": 1589
+    },
+    {
+      "epoch": 1.4131763137429174,
+      "grad_norm": 0.3404034674167633,
+      "learning_rate": 0.0001,
+      "loss": 0.8753,
+      "step": 1590
+    },
+    {
+      "epoch": 1.414065103877347,
+      "grad_norm": 0.3574478328227997,
+      "learning_rate": 0.0001,
+      "loss": 0.884,
+      "step": 1591
+    },
+    {
+      "epoch": 1.4149538940117765,
+      "grad_norm": 0.3543257415294647,
+      "learning_rate": 0.0001,
+      "loss": 0.9347,
+      "step": 1592
+    },
+    {
+      "epoch": 1.415842684146206,
+      "grad_norm": 0.37373191118240356,
+      "learning_rate": 0.0001,
+      "loss": 0.9138,
+      "step": 1593
+    },
+    {
+      "epoch": 1.4167314742806356,
+      "grad_norm": 0.7103110551834106,
+      "learning_rate": 0.0001,
+      "loss": 0.918,
+      "step": 1594
+    },
+    {
+      "epoch": 1.417620264415065,
+      "grad_norm": 0.377218097448349,
+      "learning_rate": 0.0001,
+      "loss": 1.0207,
+      "step": 1595
+    },
+    {
+      "epoch": 1.4185090545494945,
+      "grad_norm": 0.3593274652957916,
+      "learning_rate": 0.0001,
+      "loss": 0.8988,
+      "step": 1596
+    },
+    {
+      "epoch": 1.419397844683924,
+      "grad_norm": 0.3128172755241394,
+      "learning_rate": 0.0001,
+      "loss": 0.9314,
+      "step": 1597
+    },
+    {
+      "epoch": 1.4202866348183536,
+      "grad_norm": 0.39537543058395386,
+      "learning_rate": 0.0001,
+      "loss": 0.9583,
+      "step": 1598
+    },
+    {
+      "epoch": 1.421175424952783,
+      "grad_norm": 0.38807258009910583,
+      "learning_rate": 0.0001,
+      "loss": 0.9978,
+      "step": 1599
+    },
+    {
+      "epoch": 1.4220642150872125,
+      "grad_norm": 0.32216835021972656,
+      "learning_rate": 0.0001,
+      "loss": 0.8949,
+      "step": 1600
+    },
+    {
+      "epoch": 1.422953005221642,
+      "grad_norm": 0.3530212938785553,
+      "learning_rate": 0.0001,
+      "loss": 0.9434,
+      "step": 1601
+    },
+    {
+      "epoch": 1.4238417953560716,
+      "grad_norm": 0.38593819737434387,
+      "learning_rate": 0.0001,
+      "loss": 0.8588,
+      "step": 1602
+    },
+    {
+      "epoch": 1.424730585490501,
+      "grad_norm": 0.319916307926178,
+      "learning_rate": 0.0001,
+      "loss": 0.9491,
+      "step": 1603
+    },
+    {
+      "epoch": 1.4256193756249305,
+      "grad_norm": 0.46837642788887024,
+      "learning_rate": 0.0001,
+      "loss": 0.8415,
+      "step": 1604
+    },
+    {
+      "epoch": 1.4265081657593601,
+      "grad_norm": 0.3597429394721985,
+      "learning_rate": 0.0001,
+      "loss": 0.9981,
+      "step": 1605
+    },
+    {
+      "epoch": 1.4273969558937896,
+      "grad_norm": 0.33240172266960144,
+      "learning_rate": 0.0001,
+      "loss": 0.9084,
+      "step": 1606
+    },
+    {
+      "epoch": 1.428285746028219,
+      "grad_norm": 0.3380354046821594,
+      "learning_rate": 0.0001,
+      "loss": 0.9233,
+      "step": 1607
+    },
+    {
+      "epoch": 1.4291745361626487,
+      "grad_norm": 0.503455400466919,
+      "learning_rate": 0.0001,
+      "loss": 0.9825,
+      "step": 1608
+    },
+    {
+      "epoch": 1.430063326297078,
+      "grad_norm": 0.3610774576663971,
+      "learning_rate": 0.0001,
+      "loss": 1.0047,
+      "step": 1609
+    },
+    {
+      "epoch": 1.4309521164315075,
+      "grad_norm": 0.44202589988708496,
+      "learning_rate": 0.0001,
+      "loss": 0.9203,
+      "step": 1610
+    },
+    {
+      "epoch": 1.4318409065659372,
+      "grad_norm": 0.3535199761390686,
+      "learning_rate": 0.0001,
+      "loss": 0.9925,
+      "step": 1611
+    },
+    {
+      "epoch": 1.4327296967003667,
+      "grad_norm": 0.3379201292991638,
+      "learning_rate": 0.0001,
+      "loss": 0.9715,
+      "step": 1612
+    },
+    {
+      "epoch": 1.433618486834796,
+      "grad_norm": 0.34348997473716736,
+      "learning_rate": 0.0001,
+      "loss": 0.9106,
+      "step": 1613
+    },
+    {
+      "epoch": 1.4345072769692258,
+      "grad_norm": 0.341406911611557,
+      "learning_rate": 0.0001,
+      "loss": 0.9345,
+      "step": 1614
+    },
+    {
+      "epoch": 1.4353960671036552,
+      "grad_norm": 0.2867315411567688,
+      "learning_rate": 0.0001,
+      "loss": 0.8473,
+      "step": 1615
+    },
+    {
+      "epoch": 1.4362848572380846,
+      "grad_norm": 0.3688201308250427,
+      "learning_rate": 0.0001,
+      "loss": 0.887,
+      "step": 1616
+    },
+    {
+      "epoch": 1.4371736473725143,
+      "grad_norm": 0.3555888235569,
+      "learning_rate": 0.0001,
+      "loss": 0.8905,
+      "step": 1617
+    },
+    {
+      "epoch": 1.4380624375069437,
+      "grad_norm": 0.3761601448059082,
+      "learning_rate": 0.0001,
+      "loss": 0.8659,
+      "step": 1618
+    },
+    {
+      "epoch": 1.4389512276413732,
+      "grad_norm": 0.3437979519367218,
+      "learning_rate": 0.0001,
+      "loss": 0.9541,
+      "step": 1619
+    },
+    {
+      "epoch": 1.4398400177758026,
+      "grad_norm": 0.33345940709114075,
+      "learning_rate": 0.0001,
+      "loss": 0.9721,
+      "step": 1620
+    },
+    {
+      "epoch": 1.4407288079102323,
+      "grad_norm": 0.356842577457428,
+      "learning_rate": 0.0001,
+      "loss": 0.9055,
+      "step": 1621
+    },
+    {
+      "epoch": 1.4416175980446617,
+      "grad_norm": 0.35060185194015503,
+      "learning_rate": 0.0001,
+      "loss": 0.8517,
+      "step": 1622
+    },
+    {
+      "epoch": 1.4425063881790912,
+      "grad_norm": 0.3705711364746094,
+      "learning_rate": 0.0001,
+      "loss": 1.0321,
+      "step": 1623
+    },
+    {
+      "epoch": 1.4433951783135206,
+      "grad_norm": 0.38974469900131226,
+      "learning_rate": 0.0001,
+      "loss": 0.8974,
+      "step": 1624
+    },
+    {
+      "epoch": 1.4442839684479503,
+      "grad_norm": 0.40210771560668945,
+      "learning_rate": 0.0001,
+      "loss": 0.9183,
+      "step": 1625
+    },
+    {
+      "epoch": 1.4451727585823797,
+      "grad_norm": 0.3603565990924835,
+      "learning_rate": 0.0001,
+      "loss": 0.8622,
+      "step": 1626
+    },
+    {
+      "epoch": 1.4460615487168091,
+      "grad_norm": 0.3595213294029236,
+      "learning_rate": 0.0001,
+      "loss": 0.931,
+      "step": 1627
+    },
+    {
+      "epoch": 1.4469503388512388,
+      "grad_norm": 0.34834766387939453,
+      "learning_rate": 0.0001,
+      "loss": 0.9076,
+      "step": 1628
+    },
+    {
+      "epoch": 1.4478391289856682,
+      "grad_norm": 0.3451787531375885,
+      "learning_rate": 0.0001,
+      "loss": 0.8818,
+      "step": 1629
+    },
+    {
+      "epoch": 1.4487279191200977,
+      "grad_norm": 0.3447902202606201,
+      "learning_rate": 0.0001,
+      "loss": 0.8725,
+      "step": 1630
+    },
+    {
+      "epoch": 1.4496167092545273,
+      "grad_norm": 0.3713982403278351,
+      "learning_rate": 0.0001,
+      "loss": 0.9179,
+      "step": 1631
+    },
+    {
+      "epoch": 1.4505054993889568,
+      "grad_norm": 0.41470077633857727,
+      "learning_rate": 0.0001,
+      "loss": 0.8618,
+      "step": 1632
+    },
+    {
+      "epoch": 1.4513942895233862,
+      "grad_norm": 0.35781964659690857,
+      "learning_rate": 0.0001,
+      "loss": 0.8232,
+      "step": 1633
+    },
+    {
+      "epoch": 1.452283079657816,
+      "grad_norm": 0.32825222611427307,
+      "learning_rate": 0.0001,
+      "loss": 0.8527,
+      "step": 1634
+    },
+    {
+      "epoch": 1.4531718697922453,
+      "grad_norm": 0.3815636932849884,
+      "learning_rate": 0.0001,
+      "loss": 0.9824,
+      "step": 1635
+    },
+    {
+      "epoch": 1.4540606599266748,
+      "grad_norm": 0.3856026828289032,
+      "learning_rate": 0.0001,
+      "loss": 0.9522,
+      "step": 1636
+    },
+    {
+      "epoch": 1.4549494500611044,
+      "grad_norm": 0.3923175036907196,
+      "learning_rate": 0.0001,
+      "loss": 0.8057,
+      "step": 1637
+    },
+    {
+      "epoch": 1.4558382401955339,
+      "grad_norm": 0.3747645914554596,
+      "learning_rate": 0.0001,
+      "loss": 0.9438,
+      "step": 1638
+    },
+    {
+      "epoch": 1.4567270303299633,
+      "grad_norm": 0.3561338782310486,
+      "learning_rate": 0.0001,
+      "loss": 0.8827,
+      "step": 1639
+    },
+    {
+      "epoch": 1.457615820464393,
+      "grad_norm": 0.32757651805877686,
+      "learning_rate": 0.0001,
+      "loss": 0.8562,
+      "step": 1640
+    },
+    {
+      "epoch": 1.4585046105988224,
+      "grad_norm": 0.378717839717865,
+      "learning_rate": 0.0001,
+      "loss": 0.9642,
+      "step": 1641
+    },
+    {
+      "epoch": 1.4593934007332519,
+      "grad_norm": 0.3806663751602173,
+      "learning_rate": 0.0001,
+      "loss": 0.9111,
+      "step": 1642
+    },
+    {
+      "epoch": 1.4602821908676813,
+      "grad_norm": 0.34055235981941223,
+      "learning_rate": 0.0001,
+      "loss": 0.9138,
+      "step": 1643
+    },
+    {
+      "epoch": 1.461170981002111,
+      "grad_norm": 0.3508027493953705,
+      "learning_rate": 0.0001,
+      "loss": 0.8867,
+      "step": 1644
+    },
+    {
+      "epoch": 1.4620597711365404,
+      "grad_norm": 0.3853304386138916,
+      "learning_rate": 0.0001,
+      "loss": 0.895,
+      "step": 1645
+    },
+    {
+      "epoch": 1.4629485612709698,
+      "grad_norm": 0.39283403754234314,
+      "learning_rate": 0.0001,
+      "loss": 0.8977,
+      "step": 1646
+    },
+    {
+      "epoch": 1.4638373514053993,
+      "grad_norm": 0.3606354892253876,
+      "learning_rate": 0.0001,
+      "loss": 0.8564,
+      "step": 1647
+    },
+    {
+      "epoch": 1.464726141539829,
+      "grad_norm": 0.3763819932937622,
+      "learning_rate": 0.0001,
+      "loss": 0.8714,
+      "step": 1648
+    },
+    {
+      "epoch": 1.4656149316742584,
+      "grad_norm": 0.3701646625995636,
+      "learning_rate": 0.0001,
+      "loss": 0.9512,
+      "step": 1649
+    },
+    {
+      "epoch": 1.4665037218086878,
+      "grad_norm": 0.383543998003006,
+      "learning_rate": 0.0001,
+      "loss": 0.9052,
+      "step": 1650
+    },
+    {
+      "epoch": 1.4673925119431175,
+      "grad_norm": 0.3642030954360962,
+      "learning_rate": 0.0001,
+      "loss": 0.8511,
+      "step": 1651
+    },
+    {
+      "epoch": 1.468281302077547,
+      "grad_norm": 0.38401633501052856,
+      "learning_rate": 0.0001,
+      "loss": 0.9076,
+      "step": 1652
+    },
+    {
+      "epoch": 1.4691700922119764,
+      "grad_norm": 0.3193959891796112,
+      "learning_rate": 0.0001,
+      "loss": 0.8841,
+      "step": 1653
+    },
+    {
+      "epoch": 1.470058882346406,
+      "grad_norm": 0.35109949111938477,
+      "learning_rate": 0.0001,
+      "loss": 0.9746,
+      "step": 1654
+    },
+    {
+      "epoch": 1.4709476724808355,
+      "grad_norm": 0.38246193528175354,
+      "learning_rate": 0.0001,
+      "loss": 0.9658,
+      "step": 1655
+    },
+    {
+      "epoch": 1.471836462615265,
+      "grad_norm": 0.3813583552837372,
+      "learning_rate": 0.0001,
+      "loss": 0.9618,
+      "step": 1656
+    },
+    {
+      "epoch": 1.4727252527496946,
+      "grad_norm": 0.345525860786438,
+      "learning_rate": 0.0001,
+      "loss": 1.0142,
+      "step": 1657
+    },
+    {
+      "epoch": 1.473614042884124,
+      "grad_norm": 0.3886154592037201,
+      "learning_rate": 0.0001,
+      "loss": 0.9028,
+      "step": 1658
+    },
+    {
+      "epoch": 1.4745028330185534,
+      "grad_norm": 0.31922218203544617,
+      "learning_rate": 0.0001,
+      "loss": 0.9339,
+      "step": 1659
+    },
+    {
+      "epoch": 1.4753916231529831,
+      "grad_norm": 0.38563981652259827,
+      "learning_rate": 0.0001,
+      "loss": 1.0242,
+      "step": 1660
+    },
+    {
+      "epoch": 1.4762804132874126,
+      "grad_norm": 0.3879290223121643,
+      "learning_rate": 0.0001,
+      "loss": 0.9703,
+      "step": 1661
+    },
+    {
+      "epoch": 1.477169203421842,
+      "grad_norm": 0.36282819509506226,
+      "learning_rate": 0.0001,
+      "loss": 0.9179,
+      "step": 1662
+    },
+    {
+      "epoch": 1.4780579935562717,
+      "grad_norm": 0.39247390627861023,
+      "learning_rate": 0.0001,
+      "loss": 0.8528,
+      "step": 1663
+    },
+    {
+      "epoch": 1.478946783690701,
+      "grad_norm": 0.37190213799476624,
+      "learning_rate": 0.0001,
+      "loss": 0.8894,
+      "step": 1664
+    },
+    {
+      "epoch": 1.4798355738251305,
+      "grad_norm": 0.35375040769577026,
+      "learning_rate": 0.0001,
+      "loss": 0.862,
+      "step": 1665
+    },
+    {
+      "epoch": 1.48072436395956,
+      "grad_norm": 0.38695651292800903,
+      "learning_rate": 0.0001,
+      "loss": 0.9118,
+      "step": 1666
+    },
+    {
+      "epoch": 1.4816131540939896,
+      "grad_norm": 0.34383878111839294,
+      "learning_rate": 0.0001,
+      "loss": 0.874,
+      "step": 1667
+    },
+    {
+      "epoch": 1.482501944228419,
+      "grad_norm": 0.3391941785812378,
+      "learning_rate": 0.0001,
+      "loss": 0.9457,
+      "step": 1668
+    },
+    {
+      "epoch": 1.4833907343628485,
+      "grad_norm": 0.3345963954925537,
+      "learning_rate": 0.0001,
+      "loss": 0.8521,
+      "step": 1669
+    },
+    {
+      "epoch": 1.484279524497278,
+      "grad_norm": 0.3772295117378235,
+      "learning_rate": 0.0001,
+      "loss": 0.8987,
+      "step": 1670
+    },
+    {
+      "epoch": 1.4851683146317076,
+      "grad_norm": 0.3609481751918793,
+      "learning_rate": 0.0001,
+      "loss": 0.8841,
+      "step": 1671
+    },
+    {
+      "epoch": 1.486057104766137,
+      "grad_norm": 0.3498031795024872,
+      "learning_rate": 0.0001,
+      "loss": 0.9713,
+      "step": 1672
+    },
+    {
+      "epoch": 1.4869458949005665,
+      "grad_norm": 0.33337390422821045,
+      "learning_rate": 0.0001,
+      "loss": 0.9458,
+      "step": 1673
+    },
+    {
+      "epoch": 1.4878346850349962,
+      "grad_norm": 0.3364051282405853,
+      "learning_rate": 0.0001,
+      "loss": 0.8227,
+      "step": 1674
+    },
+    {
+      "epoch": 1.4887234751694256,
+      "grad_norm": 0.3339247405529022,
+      "learning_rate": 0.0001,
+      "loss": 0.9191,
+      "step": 1675
+    },
+    {
+      "epoch": 1.489612265303855,
+      "grad_norm": 0.3879411816596985,
+      "learning_rate": 0.0001,
+      "loss": 0.9857,
+      "step": 1676
+    },
+    {
+      "epoch": 1.4905010554382847,
+      "grad_norm": 0.3417137563228607,
+      "learning_rate": 0.0001,
+      "loss": 0.8781,
+      "step": 1677
+    },
+    {
+      "epoch": 1.4913898455727141,
+      "grad_norm": 0.3149627447128296,
+      "learning_rate": 0.0001,
+      "loss": 0.9536,
+      "step": 1678
+    },
+    {
+      "epoch": 1.4922786357071436,
+      "grad_norm": 0.3448125720024109,
+      "learning_rate": 0.0001,
+      "loss": 0.8991,
+      "step": 1679
+    },
+    {
+      "epoch": 1.4931674258415732,
+      "grad_norm": 0.3406059741973877,
+      "learning_rate": 0.0001,
+      "loss": 0.9704,
+      "step": 1680
+    },
+    {
+      "epoch": 1.4940562159760027,
+      "grad_norm": 0.3938051164150238,
+      "learning_rate": 0.0001,
+      "loss": 0.9602,
+      "step": 1681
+    },
+    {
+      "epoch": 1.4949450061104321,
+      "grad_norm": 0.35217610001564026,
+      "learning_rate": 0.0001,
+      "loss": 0.8743,
+      "step": 1682
+    },
+    {
+      "epoch": 1.4958337962448618,
+      "grad_norm": 0.3443836271762848,
+      "learning_rate": 0.0001,
+      "loss": 0.8809,
+      "step": 1683
+    },
+    {
+      "epoch": 1.4967225863792912,
+      "grad_norm": 0.36990824341773987,
+      "learning_rate": 0.0001,
+      "loss": 0.9191,
+      "step": 1684
+    },
+    {
+      "epoch": 1.4976113765137207,
+      "grad_norm": 0.3850703239440918,
+      "learning_rate": 0.0001,
+      "loss": 0.9439,
+      "step": 1685
+    },
+    {
+      "epoch": 1.4985001666481503,
+      "grad_norm": 0.336357444524765,
+      "learning_rate": 0.0001,
+      "loss": 0.9015,
+      "step": 1686
+    },
+    {
+      "epoch": 1.4993889567825798,
+      "grad_norm": 0.33322861790657043,
+      "learning_rate": 0.0001,
+      "loss": 0.9134,
+      "step": 1687
+    },
+    {
+      "epoch": 1.5002777469170092,
+      "grad_norm": 0.3278595507144928,
+      "learning_rate": 0.0001,
+      "loss": 0.902,
+      "step": 1688
+    },
+    {
+      "epoch": 1.5011665370514389,
+      "grad_norm": 0.36542853713035583,
+      "learning_rate": 0.0001,
+      "loss": 0.9095,
+      "step": 1689
+    },
+    {
+      "epoch": 1.502055327185868,
+      "grad_norm": 0.3316444456577301,
+      "learning_rate": 0.0001,
+      "loss": 0.8739,
+      "step": 1690
+    },
+    {
+      "epoch": 1.5029441173202978,
+      "grad_norm": 0.34426775574684143,
+      "learning_rate": 0.0001,
+      "loss": 0.8909,
+      "step": 1691
+    },
+    {
+      "epoch": 1.5038329074547274,
+      "grad_norm": 0.3481121063232422,
+      "learning_rate": 0.0001,
+      "loss": 0.9043,
+      "step": 1692
+    },
+    {
+      "epoch": 1.5047216975891566,
+      "grad_norm": 0.3318212628364563,
+      "learning_rate": 0.0001,
+      "loss": 0.9219,
+      "step": 1693
+    },
+    {
+      "epoch": 1.5056104877235863,
+      "grad_norm": 0.3894112706184387,
+      "learning_rate": 0.0001,
+      "loss": 0.9349,
+      "step": 1694
+    },
+    {
+      "epoch": 1.5064992778580157,
+      "grad_norm": 0.3638782501220703,
+      "learning_rate": 0.0001,
+      "loss": 0.8549,
+      "step": 1695
+    },
+    {
+      "epoch": 1.5073880679924452,
+      "grad_norm": 0.3686305284500122,
+      "learning_rate": 0.0001,
+      "loss": 0.8293,
+      "step": 1696
+    },
+    {
+      "epoch": 1.5082768581268748,
+      "grad_norm": 0.3255722224712372,
+      "learning_rate": 0.0001,
+      "loss": 0.9002,
+      "step": 1697
+    },
+    {
+      "epoch": 1.5091656482613043,
+      "grad_norm": 0.3039862811565399,
+      "learning_rate": 0.0001,
+      "loss": 0.838,
+      "step": 1698
+    },
+    {
+      "epoch": 1.5100544383957337,
+      "grad_norm": 0.372408926486969,
+      "learning_rate": 0.0001,
+      "loss": 0.9254,
+      "step": 1699
+    },
+    {
+      "epoch": 1.5109432285301634,
+      "grad_norm": 0.3547128438949585,
+      "learning_rate": 0.0001,
+      "loss": 0.9622,
+      "step": 1700
+    },
+    {
+      "epoch": 1.5118320186645928,
+      "grad_norm": 0.3619207441806793,
+      "learning_rate": 0.0001,
+      "loss": 0.9454,
+      "step": 1701
+    },
+    {
+      "epoch": 1.5127208087990223,
+      "grad_norm": 0.347741961479187,
+      "learning_rate": 0.0001,
+      "loss": 0.8835,
+      "step": 1702
+    },
+    {
+      "epoch": 1.513609598933452,
+      "grad_norm": 0.3651529848575592,
+      "learning_rate": 0.0001,
+      "loss": 0.9481,
+      "step": 1703
+    },
+    {
+      "epoch": 1.5144983890678814,
+      "grad_norm": 0.3714975416660309,
+      "learning_rate": 0.0001,
+      "loss": 0.8664,
+      "step": 1704
+    },
+    {
+      "epoch": 1.5153871792023108,
+      "grad_norm": 0.38629162311553955,
+      "learning_rate": 0.0001,
+      "loss": 0.911,
+      "step": 1705
+    },
+    {
+      "epoch": 1.5162759693367405,
+      "grad_norm": 0.31858816742897034,
+      "learning_rate": 0.0001,
+      "loss": 0.8828,
+      "step": 1706
+    },
+    {
+      "epoch": 1.51716475947117,
+      "grad_norm": 0.34497538208961487,
+      "learning_rate": 0.0001,
+      "loss": 0.8967,
+      "step": 1707
+    },
+    {
+      "epoch": 1.5180535496055994,
+      "grad_norm": 0.34334251284599304,
+      "learning_rate": 0.0001,
+      "loss": 0.9161,
+      "step": 1708
+    },
+    {
+      "epoch": 1.518942339740029,
+      "grad_norm": 0.35556355118751526,
+      "learning_rate": 0.0001,
+      "loss": 0.9303,
+      "step": 1709
+    },
+    {
+      "epoch": 1.5198311298744582,
+      "grad_norm": 0.2841368019580841,
+      "learning_rate": 0.0001,
+      "loss": 0.9444,
+      "step": 1710
+    },
+    {
+      "epoch": 1.520719920008888,
+      "grad_norm": 0.36247727274894714,
+      "learning_rate": 0.0001,
+      "loss": 0.9561,
+      "step": 1711
+    },
+    {
+      "epoch": 1.5216087101433176,
+      "grad_norm": 0.3494194447994232,
+      "learning_rate": 0.0001,
+      "loss": 0.9227,
+      "step": 1712
+    },
+    {
+      "epoch": 1.5224975002777468,
+      "grad_norm": 0.34062543511390686,
+      "learning_rate": 0.0001,
+      "loss": 0.9002,
+      "step": 1713
+    },
+    {
+      "epoch": 1.5233862904121764,
+      "grad_norm": 0.32053127884864807,
+      "learning_rate": 0.0001,
+      "loss": 0.8472,
+      "step": 1714
+    },
+    {
+      "epoch": 1.524275080546606,
+      "grad_norm": 0.34881147742271423,
+      "learning_rate": 0.0001,
+      "loss": 0.996,
+      "step": 1715
+    },
+    {
+      "epoch": 1.5251638706810353,
+      "grad_norm": 0.31298136711120605,
+      "learning_rate": 0.0001,
+      "loss": 0.9664,
+      "step": 1716
+    },
+    {
+      "epoch": 1.526052660815465,
+      "grad_norm": 0.35817843675613403,
+      "learning_rate": 0.0001,
+      "loss": 0.8581,
+      "step": 1717
+    },
+    {
+      "epoch": 1.5269414509498944,
+      "grad_norm": 0.36280620098114014,
+      "learning_rate": 0.0001,
+      "loss": 0.9752,
+      "step": 1718
+    },
+    {
+      "epoch": 1.5278302410843239,
+      "grad_norm": 0.3819250464439392,
+      "learning_rate": 0.0001,
+      "loss": 0.8537,
+      "step": 1719
+    },
+    {
+      "epoch": 1.5287190312187535,
+      "grad_norm": 0.319489061832428,
+      "learning_rate": 0.0001,
+      "loss": 0.8939,
+      "step": 1720
+    },
+    {
+      "epoch": 1.529607821353183,
+      "grad_norm": 0.3549860417842865,
+      "learning_rate": 0.0001,
+      "loss": 0.9703,
+      "step": 1721
+    },
+    {
+      "epoch": 1.5304966114876124,
+      "grad_norm": 0.3653222620487213,
+      "learning_rate": 0.0001,
+      "loss": 0.9454,
+      "step": 1722
+    },
+    {
+      "epoch": 1.531385401622042,
+      "grad_norm": 1.377524495124817,
+      "learning_rate": 0.0001,
+      "loss": 0.9496,
+      "step": 1723
+    },
+    {
+      "epoch": 1.5322741917564715,
+      "grad_norm": 0.37237972021102905,
+      "learning_rate": 0.0001,
+      "loss": 0.9882,
+      "step": 1724
+    },
+    {
+      "epoch": 1.533162981890901,
+      "grad_norm": 0.3513152003288269,
+      "learning_rate": 0.0001,
+      "loss": 0.9387,
+      "step": 1725
+    },
+    {
+      "epoch": 1.5340517720253306,
+      "grad_norm": 1.2794930934906006,
+      "learning_rate": 0.0001,
+      "loss": 0.9713,
+      "step": 1726
+    },
+    {
+      "epoch": 1.53494056215976,
+      "grad_norm": 0.37315309047698975,
+      "learning_rate": 0.0001,
+      "loss": 0.9837,
+      "step": 1727
+    },
+    {
+      "epoch": 1.5358293522941895,
+      "grad_norm": 0.34950658679008484,
+      "learning_rate": 0.0001,
+      "loss": 0.9261,
+      "step": 1728
+    },
+    {
+      "epoch": 1.5367181424286191,
+      "grad_norm": 0.3580825924873352,
+      "learning_rate": 0.0001,
+      "loss": 0.9007,
+      "step": 1729
+    },
+    {
+      "epoch": 1.5376069325630486,
+      "grad_norm": 0.6059324145317078,
+      "learning_rate": 0.0001,
+      "loss": 0.9215,
+      "step": 1730
+    },
+    {
+      "epoch": 1.538495722697478,
+      "grad_norm": 2.688345432281494,
+      "learning_rate": 0.0001,
+      "loss": 0.9347,
+      "step": 1731
+    },
+    {
+      "epoch": 1.5393845128319077,
+      "grad_norm": 1.058393955230713,
+      "learning_rate": 0.0001,
+      "loss": 0.8553,
+      "step": 1732
+    },
+    {
+      "epoch": 1.540273302966337,
+      "grad_norm": 0.5399541854858398,
+      "learning_rate": 0.0001,
+      "loss": 0.9347,
+      "step": 1733
+    },
+    {
+      "epoch": 1.5411620931007666,
+      "grad_norm": 0.38701578974723816,
+      "learning_rate": 0.0001,
+      "loss": 0.9532,
+      "step": 1734
+    },
+    {
+      "epoch": 1.5420508832351962,
+      "grad_norm": 0.3606896698474884,
+      "learning_rate": 0.0001,
+      "loss": 0.9679,
+      "step": 1735
+    },
+    {
+      "epoch": 1.5429396733696255,
+      "grad_norm": 0.33948925137519836,
+      "learning_rate": 0.0001,
+      "loss": 0.8631,
+      "step": 1736
+    },
+    {
+      "epoch": 1.5438284635040551,
+      "grad_norm": 0.3091464638710022,
+      "learning_rate": 0.0001,
+      "loss": 0.8451,
+      "step": 1737
+    },
+    {
+      "epoch": 1.5447172536384848,
+      "grad_norm": 0.3580194413661957,
+      "learning_rate": 0.0001,
+      "loss": 0.9166,
+      "step": 1738
+    },
+    {
+      "epoch": 1.545606043772914,
+      "grad_norm": 0.32534483075141907,
+      "learning_rate": 0.0001,
+      "loss": 0.8588,
+      "step": 1739
+    },
+    {
+      "epoch": 1.5464948339073437,
+      "grad_norm": 0.37172192335128784,
+      "learning_rate": 0.0001,
+      "loss": 0.8881,
+      "step": 1740
+    },
+    {
+      "epoch": 1.547383624041773,
+      "grad_norm": 0.3773120045661926,
+      "learning_rate": 0.0001,
+      "loss": 0.9553,
+      "step": 1741
+    },
+    {
+      "epoch": 1.5482724141762025,
+      "grad_norm": 0.37788835167884827,
+      "learning_rate": 0.0001,
+      "loss": 0.9444,
+      "step": 1742
+    },
+    {
+      "epoch": 1.5491612043106322,
+      "grad_norm": 0.33638015389442444,
+      "learning_rate": 0.0001,
+      "loss": 0.8482,
+      "step": 1743
+    },
+    {
+      "epoch": 1.5500499944450616,
+      "grad_norm": 0.38720908761024475,
+      "learning_rate": 0.0001,
+      "loss": 0.8876,
+      "step": 1744
+    },
+    {
+      "epoch": 1.550938784579491,
+      "grad_norm": 0.3406068980693817,
+      "learning_rate": 0.0001,
+      "loss": 0.9519,
+      "step": 1745
+    },
+    {
+      "epoch": 1.5518275747139207,
+      "grad_norm": 0.3615241050720215,
+      "learning_rate": 0.0001,
+      "loss": 0.9574,
+      "step": 1746
+    },
+    {
+      "epoch": 1.5527163648483502,
+      "grad_norm": 0.34514304995536804,
+      "learning_rate": 0.0001,
+      "loss": 0.9059,
+      "step": 1747
+    },
+    {
+      "epoch": 1.5536051549827796,
+      "grad_norm": 0.3243461549282074,
+      "learning_rate": 0.0001,
+      "loss": 0.888,
+      "step": 1748
+    },
+    {
+      "epoch": 1.5544939451172093,
+      "grad_norm": 0.37375468015670776,
+      "learning_rate": 0.0001,
+      "loss": 0.7914,
+      "step": 1749
+    },
+    {
+      "epoch": 1.5553827352516387,
+      "grad_norm": 0.37953221797943115,
+      "learning_rate": 0.0001,
+      "loss": 0.8981,
+      "step": 1750
+    },
+    {
+      "epoch": 1.5562715253860682,
+      "grad_norm": 0.36913594603538513,
+      "learning_rate": 0.0001,
+      "loss": 0.9042,
+      "step": 1751
+    },
+    {
+      "epoch": 1.5571603155204978,
+      "grad_norm": 0.40167394280433655,
+      "learning_rate": 0.0001,
+      "loss": 0.9929,
+      "step": 1752
+    },
+    {
+      "epoch": 1.5580491056549273,
+      "grad_norm": 0.3700322210788727,
+      "learning_rate": 0.0001,
+      "loss": 0.9464,
+      "step": 1753
+    },
+    {
+      "epoch": 1.5589378957893567,
+      "grad_norm": 0.4139519929885864,
+      "learning_rate": 0.0001,
+      "loss": 0.8966,
+      "step": 1754
+    },
+    {
+      "epoch": 1.5598266859237864,
+      "grad_norm": 1.0801198482513428,
+      "learning_rate": 0.0001,
+      "loss": 0.8788,
+      "step": 1755
+    },
+    {
+      "epoch": 1.5607154760582156,
+      "grad_norm": 0.8275560140609741,
+      "learning_rate": 0.0001,
+      "loss": 0.8335,
+      "step": 1756
+    },
+    {
+      "epoch": 1.5616042661926453,
+      "grad_norm": 0.5179623961448669,
+      "learning_rate": 0.0001,
+      "loss": 0.9282,
+      "step": 1757
+    },
+    {
+      "epoch": 1.562493056327075,
+      "grad_norm": 0.3609626889228821,
+      "learning_rate": 0.0001,
+      "loss": 0.8669,
+      "step": 1758
+    },
+    {
+      "epoch": 1.5633818464615041,
+      "grad_norm": 0.35782667994499207,
+      "learning_rate": 0.0001,
+      "loss": 0.8826,
+      "step": 1759
+    },
+    {
+      "epoch": 1.5642706365959338,
+      "grad_norm": 0.5242655873298645,
+      "learning_rate": 0.0001,
+      "loss": 0.8888,
+      "step": 1760
+    },
+    {
+      "epoch": 1.5651594267303635,
+      "grad_norm": 0.3657304346561432,
+      "learning_rate": 0.0001,
+      "loss": 0.9309,
+      "step": 1761
+    },
+    {
+      "epoch": 1.5660482168647927,
+      "grad_norm": 0.5481138825416565,
+      "learning_rate": 0.0001,
+      "loss": 0.8967,
+      "step": 1762
+    },
+    {
+      "epoch": 1.5669370069992223,
+      "grad_norm": 0.39052364230155945,
+      "learning_rate": 0.0001,
+      "loss": 0.9363,
+      "step": 1763
+    },
+    {
+      "epoch": 1.5678257971336518,
+      "grad_norm": 0.5495015978813171,
+      "learning_rate": 0.0001,
+      "loss": 0.9439,
+      "step": 1764
+    },
+    {
+      "epoch": 1.5687145872680812,
+      "grad_norm": 0.3862490653991699,
+      "learning_rate": 0.0001,
+      "loss": 0.9504,
+      "step": 1765
+    },
+    {
+      "epoch": 1.5696033774025109,
+      "grad_norm": 0.4059329330921173,
+      "learning_rate": 0.0001,
+      "loss": 0.9468,
+      "step": 1766
+    },
+    {
+      "epoch": 1.5704921675369403,
+      "grad_norm": 0.3757980465888977,
+      "learning_rate": 0.0001,
+      "loss": 0.9091,
+      "step": 1767
+    },
+    {
+      "epoch": 1.5713809576713698,
+      "grad_norm": 0.32866010069847107,
+      "learning_rate": 0.0001,
+      "loss": 0.9177,
+      "step": 1768
+    },
+    {
+      "epoch": 1.5722697478057994,
+      "grad_norm": 0.3823925852775574,
+      "learning_rate": 0.0001,
+      "loss": 0.8151,
+      "step": 1769
+    },
+    {
+      "epoch": 1.5731585379402289,
+      "grad_norm": 0.3286689519882202,
+      "learning_rate": 0.0001,
+      "loss": 0.9327,
+      "step": 1770
+    },
+    {
+      "epoch": 1.5740473280746583,
+      "grad_norm": 0.3756827712059021,
+      "learning_rate": 0.0001,
+      "loss": 0.9451,
+      "step": 1771
+    },
+    {
+      "epoch": 1.574936118209088,
+      "grad_norm": 0.34540703892707825,
+      "learning_rate": 0.0001,
+      "loss": 0.906,
+      "step": 1772
+    },
+    {
+      "epoch": 1.5758249083435174,
+      "grad_norm": 0.3327772617340088,
+      "learning_rate": 0.0001,
+      "loss": 0.8418,
+      "step": 1773
+    },
+    {
+      "epoch": 1.5767136984779468,
+      "grad_norm": 0.3816230893135071,
+      "learning_rate": 0.0001,
+      "loss": 0.9186,
+      "step": 1774
+    },
+    {
+      "epoch": 1.5776024886123765,
+      "grad_norm": 0.3620496094226837,
+      "learning_rate": 0.0001,
+      "loss": 0.9283,
+      "step": 1775
+    },
+    {
+      "epoch": 1.578491278746806,
+      "grad_norm": 0.3261551558971405,
+      "learning_rate": 0.0001,
+      "loss": 0.8566,
+      "step": 1776
+    },
+    {
+      "epoch": 1.5793800688812354,
+      "grad_norm": 0.3376888334751129,
+      "learning_rate": 0.0001,
+      "loss": 0.8816,
+      "step": 1777
+    },
+    {
+      "epoch": 1.580268859015665,
+      "grad_norm": 0.34106653928756714,
+      "learning_rate": 0.0001,
+      "loss": 0.8366,
+      "step": 1778
+    },
+    {
+      "epoch": 1.5811576491500943,
+      "grad_norm": 0.39439812302589417,
+      "learning_rate": 0.0001,
+      "loss": 0.9562,
+      "step": 1779
+    },
+    {
+      "epoch": 1.582046439284524,
+      "grad_norm": 0.4464170038700104,
+      "learning_rate": 0.0001,
+      "loss": 0.9191,
+      "step": 1780
+    },
+    {
+      "epoch": 1.5829352294189536,
+      "grad_norm": 0.3573848605155945,
+      "learning_rate": 0.0001,
+      "loss": 0.8999,
+      "step": 1781
+    },
+    {
+      "epoch": 1.5838240195533828,
+      "grad_norm": 0.3367520272731781,
+      "learning_rate": 0.0001,
+      "loss": 0.8198,
+      "step": 1782
+    },
+    {
+      "epoch": 1.5847128096878125,
+      "grad_norm": 0.7480552196502686,
+      "learning_rate": 0.0001,
+      "loss": 0.9561,
+      "step": 1783
+    },
+    {
+      "epoch": 1.5856015998222421,
+      "grad_norm": 0.3286367952823639,
+      "learning_rate": 0.0001,
+      "loss": 0.9986,
+      "step": 1784
+    },
+    {
+      "epoch": 1.5864903899566714,
+      "grad_norm": 0.32625967264175415,
+      "learning_rate": 0.0001,
+      "loss": 0.9228,
+      "step": 1785
+    },
+    {
+      "epoch": 1.587379180091101,
+      "grad_norm": 0.3715958893299103,
+      "learning_rate": 0.0001,
+      "loss": 0.9178,
+      "step": 1786
+    },
+    {
+      "epoch": 1.5882679702255305,
+      "grad_norm": 0.34337377548217773,
+      "learning_rate": 0.0001,
+      "loss": 0.9656,
+      "step": 1787
+    },
+    {
+      "epoch": 1.58915676035996,
+      "grad_norm": 0.36245790123939514,
+      "learning_rate": 0.0001,
+      "loss": 0.9587,
+      "step": 1788
+    },
+    {
+      "epoch": 1.5900455504943896,
+      "grad_norm": 0.4087197780609131,
+      "learning_rate": 0.0001,
+      "loss": 0.9179,
+      "step": 1789
+    },
+    {
+      "epoch": 1.590934340628819,
+      "grad_norm": 0.4008493423461914,
+      "learning_rate": 0.0001,
+      "loss": 0.8691,
+      "step": 1790
+    },
+    {
+      "epoch": 1.5918231307632484,
+      "grad_norm": 0.8266111016273499,
+      "learning_rate": 0.0001,
+      "loss": 1.0221,
+      "step": 1791
+    },
+    {
+      "epoch": 1.592711920897678,
+      "grad_norm": 0.3512238562107086,
+      "learning_rate": 0.0001,
+      "loss": 0.8909,
+      "step": 1792
+    },
+    {
+      "epoch": 1.5936007110321075,
+      "grad_norm": 0.3301697075366974,
+      "learning_rate": 0.0001,
+      "loss": 0.9072,
+      "step": 1793
+    },
+    {
+      "epoch": 1.594489501166537,
+      "grad_norm": 0.3027113974094391,
+      "learning_rate": 0.0001,
+      "loss": 0.8956,
+      "step": 1794
+    },
+    {
+      "epoch": 1.5953782913009666,
+      "grad_norm": 0.34008100628852844,
+      "learning_rate": 0.0001,
+      "loss": 0.922,
+      "step": 1795
+    },
+    {
+      "epoch": 1.596267081435396,
+      "grad_norm": 0.37972933053970337,
+      "learning_rate": 0.0001,
+      "loss": 0.8718,
+      "step": 1796
+    },
+    {
+      "epoch": 1.5971558715698255,
+      "grad_norm": 0.38649454712867737,
+      "learning_rate": 0.0001,
+      "loss": 0.8555,
+      "step": 1797
+    },
+    {
+      "epoch": 1.5980446617042552,
+      "grad_norm": 0.3708219826221466,
+      "learning_rate": 0.0001,
+      "loss": 0.8959,
+      "step": 1798
+    },
+    {
+      "epoch": 1.5989334518386846,
+      "grad_norm": 0.36001938581466675,
+      "learning_rate": 0.0001,
+      "loss": 0.8835,
+      "step": 1799
+    },
+    {
+      "epoch": 1.599822241973114,
+      "grad_norm": 0.36445868015289307,
+      "learning_rate": 0.0001,
+      "loss": 0.9237,
+      "step": 1800
+    },
+    {
+      "epoch": 1.6007110321075437,
+      "grad_norm": 0.41779571771621704,
+      "learning_rate": 0.0001,
+      "loss": 0.9475,
+      "step": 1801
+    },
+    {
+      "epoch": 1.601599822241973,
+      "grad_norm": 0.48116981983184814,
+      "learning_rate": 0.0001,
+      "loss": 0.9844,
+      "step": 1802
+    },
+    {
+      "epoch": 1.6024886123764026,
+      "grad_norm": 0.3373110592365265,
+      "learning_rate": 0.0001,
+      "loss": 0.8992,
+      "step": 1803
+    },
+    {
+      "epoch": 1.6033774025108323,
+      "grad_norm": 0.33223381638526917,
+      "learning_rate": 0.0001,
+      "loss": 0.8746,
+      "step": 1804
+    },
+    {
+      "epoch": 1.6042661926452615,
+      "grad_norm": 0.3680526912212372,
+      "learning_rate": 0.0001,
+      "loss": 0.9382,
+      "step": 1805
+    },
+    {
+      "epoch": 1.6051549827796912,
+      "grad_norm": 0.5222595930099487,
+      "learning_rate": 0.0001,
+      "loss": 0.9586,
+      "step": 1806
+    },
+    {
+      "epoch": 1.6060437729141208,
+      "grad_norm": 0.35774093866348267,
+      "learning_rate": 0.0001,
+      "loss": 0.9331,
+      "step": 1807
+    },
+    {
+      "epoch": 1.60693256304855,
+      "grad_norm": 0.528674840927124,
+      "learning_rate": 0.0001,
+      "loss": 0.9237,
+      "step": 1808
+    },
+    {
+      "epoch": 1.6078213531829797,
+      "grad_norm": 0.41820868849754333,
+      "learning_rate": 0.0001,
+      "loss": 0.9443,
+      "step": 1809
+    },
+    {
+      "epoch": 1.6087101433174091,
+      "grad_norm": 0.3307277262210846,
+      "learning_rate": 0.0001,
+      "loss": 0.8068,
+      "step": 1810
+    },
+    {
+      "epoch": 1.6095989334518386,
+      "grad_norm": 0.4219682812690735,
+      "learning_rate": 0.0001,
+      "loss": 0.9318,
+      "step": 1811
+    },
+    {
+      "epoch": 1.6104877235862682,
+      "grad_norm": 0.42429181933403015,
+      "learning_rate": 0.0001,
+      "loss": 0.8943,
+      "step": 1812
+    },
+    {
+      "epoch": 1.6113765137206977,
+      "grad_norm": 1.4591997861862183,
+      "learning_rate": 0.0001,
+      "loss": 0.8419,
+      "step": 1813
+    },
+    {
+      "epoch": 1.6122653038551271,
+      "grad_norm": 0.49615946412086487,
+      "learning_rate": 0.0001,
+      "loss": 0.9466,
+      "step": 1814
+    },
+    {
+      "epoch": 1.6131540939895568,
+      "grad_norm": 0.5319680571556091,
+      "learning_rate": 0.0001,
+      "loss": 0.8491,
+      "step": 1815
+    },
+    {
+      "epoch": 1.6140428841239862,
+      "grad_norm": 1.1099143028259277,
+      "learning_rate": 0.0001,
+      "loss": 0.9024,
+      "step": 1816
+    },
+    {
+      "epoch": 1.6149316742584157,
+      "grad_norm": 0.3784678280353546,
+      "learning_rate": 0.0001,
+      "loss": 0.9001,
+      "step": 1817
+    },
+    {
+      "epoch": 1.6158204643928453,
+      "grad_norm": 0.42704182863235474,
+      "learning_rate": 0.0001,
+      "loss": 0.9171,
+      "step": 1818
+    },
+    {
+      "epoch": 1.6167092545272748,
+      "grad_norm": 0.3222212791442871,
+      "learning_rate": 0.0001,
+      "loss": 0.8957,
+      "step": 1819
+    },
+    {
+      "epoch": 1.6175980446617042,
+      "grad_norm": 0.3985123038291931,
+      "learning_rate": 0.0001,
+      "loss": 0.803,
+      "step": 1820
+    },
+    {
+      "epoch": 1.6184868347961339,
+      "grad_norm": 0.3731878697872162,
+      "learning_rate": 0.0001,
+      "loss": 0.9434,
+      "step": 1821
+    },
+    {
+      "epoch": 1.6193756249305633,
+      "grad_norm": 0.35805362462997437,
+      "learning_rate": 0.0001,
+      "loss": 0.9385,
+      "step": 1822
+    },
+    {
+      "epoch": 1.6202644150649927,
+      "grad_norm": 0.353607177734375,
+      "learning_rate": 0.0001,
+      "loss": 0.9146,
+      "step": 1823
+    },
+    {
+      "epoch": 1.6211532051994224,
+      "grad_norm": 0.3861144185066223,
+      "learning_rate": 0.0001,
+      "loss": 0.9313,
+      "step": 1824
+    },
+    {
+      "epoch": 1.6220419953338518,
+      "grad_norm": 0.3477698564529419,
+      "learning_rate": 0.0001,
+      "loss": 1.0114,
+      "step": 1825
+    },
+    {
+      "epoch": 1.6229307854682813,
+      "grad_norm": 0.3571338355541229,
+      "learning_rate": 0.0001,
+      "loss": 0.9542,
+      "step": 1826
+    },
+    {
+      "epoch": 1.623819575602711,
+      "grad_norm": 0.3161649703979492,
+      "learning_rate": 0.0001,
+      "loss": 0.8867,
+      "step": 1827
+    },
+    {
+      "epoch": 1.6247083657371402,
+      "grad_norm": 0.34788355231285095,
+      "learning_rate": 0.0001,
+      "loss": 0.9498,
+      "step": 1828
+    },
+    {
+      "epoch": 1.6255971558715698,
+      "grad_norm": 0.3480173647403717,
+      "learning_rate": 0.0001,
+      "loss": 0.9012,
+      "step": 1829
+    },
+    {
+      "epoch": 1.6264859460059995,
+      "grad_norm": 0.3352920413017273,
+      "learning_rate": 0.0001,
+      "loss": 0.9405,
+      "step": 1830
+    },
+    {
+      "epoch": 1.6273747361404287,
+      "grad_norm": 0.3569203019142151,
+      "learning_rate": 0.0001,
+      "loss": 0.8903,
+      "step": 1831
+    },
+    {
+      "epoch": 1.6282635262748584,
+      "grad_norm": 0.36906954646110535,
+      "learning_rate": 0.0001,
+      "loss": 0.9598,
+      "step": 1832
+    },
+    {
+      "epoch": 1.6291523164092878,
+      "grad_norm": 0.3525664508342743,
+      "learning_rate": 0.0001,
+      "loss": 0.9319,
+      "step": 1833
+    },
+    {
+      "epoch": 1.6300411065437173,
+      "grad_norm": 0.3371136784553528,
+      "learning_rate": 0.0001,
+      "loss": 0.9193,
+      "step": 1834
+    },
+    {
+      "epoch": 1.630929896678147,
+      "grad_norm": 0.3484685719013214,
+      "learning_rate": 0.0001,
+      "loss": 0.9035,
+      "step": 1835
+    },
+    {
+      "epoch": 1.6318186868125764,
+      "grad_norm": 0.3327842354774475,
+      "learning_rate": 0.0001,
+      "loss": 0.7831,
+      "step": 1836
+    },
+    {
+      "epoch": 1.6327074769470058,
+      "grad_norm": 0.290935754776001,
+      "learning_rate": 0.0001,
+      "loss": 0.8634,
+      "step": 1837
+    },
+    {
+      "epoch": 1.6335962670814355,
+      "grad_norm": 0.35313087701797485,
+      "learning_rate": 0.0001,
+      "loss": 0.8898,
+      "step": 1838
+    },
+    {
+      "epoch": 1.634485057215865,
+      "grad_norm": 0.35210633277893066,
+      "learning_rate": 0.0001,
+      "loss": 0.9164,
+      "step": 1839
+    },
+    {
+      "epoch": 1.6353738473502943,
+      "grad_norm": 0.36426299810409546,
+      "learning_rate": 0.0001,
+      "loss": 0.8369,
+      "step": 1840
+    },
+    {
+      "epoch": 1.636262637484724,
+      "grad_norm": 0.36496469378471375,
+      "learning_rate": 0.0001,
+      "loss": 0.94,
+      "step": 1841
+    },
+    {
+      "epoch": 1.6371514276191534,
+      "grad_norm": 0.3476478159427643,
+      "learning_rate": 0.0001,
+      "loss": 0.8577,
+      "step": 1842
+    },
+    {
+      "epoch": 1.6380402177535829,
+      "grad_norm": 0.3203316926956177,
+      "learning_rate": 0.0001,
+      "loss": 0.861,
+      "step": 1843
+    },
+    {
+      "epoch": 1.6389290078880125,
+      "grad_norm": 0.3230277895927429,
+      "learning_rate": 0.0001,
+      "loss": 0.8456,
+      "step": 1844
+    },
+    {
+      "epoch": 1.639817798022442,
+      "grad_norm": 0.3015528917312622,
+      "learning_rate": 0.0001,
+      "loss": 0.8744,
+      "step": 1845
+    },
+    {
+      "epoch": 1.6407065881568714,
+      "grad_norm": 0.35518017411231995,
+      "learning_rate": 0.0001,
+      "loss": 0.8698,
+      "step": 1846
+    },
+    {
+      "epoch": 1.641595378291301,
+      "grad_norm": 0.35839059948921204,
+      "learning_rate": 0.0001,
+      "loss": 0.8935,
+      "step": 1847
+    },
+    {
+      "epoch": 1.6424841684257305,
+      "grad_norm": 0.3377850651741028,
+      "learning_rate": 0.0001,
+      "loss": 0.9225,
+      "step": 1848
+    },
+    {
+      "epoch": 1.64337295856016,
+      "grad_norm": 0.3225672245025635,
+      "learning_rate": 0.0001,
+      "loss": 0.9355,
+      "step": 1849
+    },
+    {
+      "epoch": 1.6442617486945896,
+      "grad_norm": 0.3357776999473572,
+      "learning_rate": 0.0001,
+      "loss": 0.8404,
+      "step": 1850
+    },
+    {
+      "epoch": 1.6451505388290188,
+      "grad_norm": 0.3443959653377533,
+      "learning_rate": 0.0001,
+      "loss": 0.9009,
+      "step": 1851
+    },
+    {
+      "epoch": 1.6460393289634485,
+      "grad_norm": 0.3641587495803833,
+      "learning_rate": 0.0001,
+      "loss": 0.9419,
+      "step": 1852
+    },
+    {
+      "epoch": 1.6469281190978782,
+      "grad_norm": 0.3614901304244995,
+      "learning_rate": 0.0001,
+      "loss": 0.8922,
+      "step": 1853
+    },
+    {
+      "epoch": 1.6478169092323074,
+      "grad_norm": 0.3756221532821655,
+      "learning_rate": 0.0001,
+      "loss": 0.8531,
+      "step": 1854
+    },
+    {
+      "epoch": 1.648705699366737,
+      "grad_norm": 0.32705435156822205,
+      "learning_rate": 0.0001,
+      "loss": 0.9559,
+      "step": 1855
+    },
+    {
+      "epoch": 1.6495944895011665,
+      "grad_norm": 0.354168564081192,
+      "learning_rate": 0.0001,
+      "loss": 0.8389,
+      "step": 1856
+    },
+    {
+      "epoch": 1.650483279635596,
+      "grad_norm": 0.3950870633125305,
+      "learning_rate": 0.0001,
+      "loss": 0.8642,
+      "step": 1857
+    },
+    {
+      "epoch": 1.6513720697700256,
+      "grad_norm": 0.3480079174041748,
+      "learning_rate": 0.0001,
+      "loss": 0.9008,
+      "step": 1858
+    },
+    {
+      "epoch": 1.652260859904455,
+      "grad_norm": 0.37953078746795654,
+      "learning_rate": 0.0001,
+      "loss": 0.9134,
+      "step": 1859
+    },
+    {
+      "epoch": 1.6531496500388845,
+      "grad_norm": 0.9250193238258362,
+      "learning_rate": 0.0001,
+      "loss": 0.8301,
+      "step": 1860
+    },
+    {
+      "epoch": 1.6540384401733141,
+      "grad_norm": 0.33579471707344055,
+      "learning_rate": 0.0001,
+      "loss": 0.8996,
+      "step": 1861
+    },
+    {
+      "epoch": 1.6549272303077436,
+      "grad_norm": 0.37209945917129517,
+      "learning_rate": 0.0001,
+      "loss": 0.8403,
+      "step": 1862
+    },
+    {
+      "epoch": 1.655816020442173,
+      "grad_norm": 0.3377666473388672,
+      "learning_rate": 0.0001,
+      "loss": 0.871,
+      "step": 1863
+    },
+    {
+      "epoch": 1.6567048105766027,
+      "grad_norm": 0.37800464034080505,
+      "learning_rate": 0.0001,
+      "loss": 0.9146,
+      "step": 1864
+    },
+    {
+      "epoch": 1.6575936007110321,
+      "grad_norm": 0.4948398172855377,
+      "learning_rate": 0.0001,
+      "loss": 0.9485,
+      "step": 1865
+    },
+    {
+      "epoch": 1.6584823908454616,
+      "grad_norm": 0.4751080274581909,
+      "learning_rate": 0.0001,
+      "loss": 0.8765,
+      "step": 1866
+    },
+    {
+      "epoch": 1.6593711809798912,
+      "grad_norm": 0.375337690114975,
+      "learning_rate": 0.0001,
+      "loss": 0.9074,
+      "step": 1867
+    },
+    {
+      "epoch": 1.6602599711143207,
+      "grad_norm": 1.2624320983886719,
+      "learning_rate": 0.0001,
+      "loss": 0.9032,
+      "step": 1868
+    },
+    {
+      "epoch": 1.66114876124875,
+      "grad_norm": 0.7726836800575256,
+      "learning_rate": 0.0001,
+      "loss": 0.9992,
+      "step": 1869
+    },
+    {
+      "epoch": 1.6620375513831798,
+      "grad_norm": 0.38178691267967224,
+      "learning_rate": 0.0001,
+      "loss": 0.9621,
+      "step": 1870
+    },
+    {
+      "epoch": 1.6629263415176092,
+      "grad_norm": 0.39392024278640747,
+      "learning_rate": 0.0001,
+      "loss": 0.9873,
+      "step": 1871
+    },
+    {
+      "epoch": 1.6638151316520386,
+      "grad_norm": 1.0342029333114624,
+      "learning_rate": 0.0001,
+      "loss": 0.9192,
+      "step": 1872
+    },
+    {
+      "epoch": 1.6647039217864683,
+      "grad_norm": 0.3234097361564636,
+      "learning_rate": 0.0001,
+      "loss": 0.9132,
+      "step": 1873
+    },
+    {
+      "epoch": 1.6655927119208975,
+      "grad_norm": 2.331127405166626,
+      "learning_rate": 0.0001,
+      "loss": 0.9651,
+      "step": 1874
+    },
+    {
+      "epoch": 1.6664815020553272,
+      "grad_norm": 0.3696269989013672,
+      "learning_rate": 0.0001,
+      "loss": 0.8714,
+      "step": 1875
+    },
+    {
+      "epoch": 1.6673702921897569,
+      "grad_norm": 0.3183539807796478,
+      "learning_rate": 0.0001,
+      "loss": 0.8484,
+      "step": 1876
+    },
+    {
+      "epoch": 1.668259082324186,
+      "grad_norm": 0.3290097713470459,
+      "learning_rate": 0.0001,
+      "loss": 0.9776,
+      "step": 1877
+    },
+    {
+      "epoch": 1.6691478724586157,
+      "grad_norm": 0.3337092101573944,
+      "learning_rate": 0.0001,
+      "loss": 0.9359,
+      "step": 1878
+    },
+    {
+      "epoch": 1.6700366625930452,
+      "grad_norm": 0.3510020971298218,
+      "learning_rate": 0.0001,
+      "loss": 0.9407,
+      "step": 1879
+    },
+    {
+      "epoch": 1.6709254527274746,
+      "grad_norm": 0.3333737254142761,
+      "learning_rate": 0.0001,
+      "loss": 0.8822,
+      "step": 1880
+    },
+    {
+      "epoch": 1.6718142428619043,
+      "grad_norm": 0.3105640709400177,
+      "learning_rate": 0.0001,
+      "loss": 0.8645,
+      "step": 1881
+    },
+    {
+      "epoch": 1.6727030329963337,
+      "grad_norm": 0.29750367999076843,
+      "learning_rate": 0.0001,
+      "loss": 0.9338,
+      "step": 1882
+    },
+    {
+      "epoch": 1.6735918231307632,
+      "grad_norm": 0.34713029861450195,
+      "learning_rate": 0.0001,
+      "loss": 0.9299,
+      "step": 1883
+    },
+    {
+      "epoch": 1.6744806132651928,
+      "grad_norm": 0.3725203275680542,
+      "learning_rate": 0.0001,
+      "loss": 0.9744,
+      "step": 1884
+    },
+    {
+      "epoch": 1.6753694033996223,
+      "grad_norm": 0.3104175925254822,
+      "learning_rate": 0.0001,
+      "loss": 0.9057,
+      "step": 1885
+    },
+    {
+      "epoch": 1.6762581935340517,
+      "grad_norm": 0.37686604261398315,
+      "learning_rate": 0.0001,
+      "loss": 1.0123,
+      "step": 1886
+    },
+    {
+      "epoch": 1.6771469836684814,
+      "grad_norm": 0.29763513803482056,
+      "learning_rate": 0.0001,
+      "loss": 0.8857,
+      "step": 1887
+    },
+    {
+      "epoch": 1.6780357738029108,
+      "grad_norm": 0.3398146629333496,
+      "learning_rate": 0.0001,
+      "loss": 0.8808,
+      "step": 1888
+    },
+    {
+      "epoch": 1.6789245639373402,
+      "grad_norm": 0.33802369236946106,
+      "learning_rate": 0.0001,
+      "loss": 0.8505,
+      "step": 1889
+    },
+    {
+      "epoch": 1.67981335407177,
+      "grad_norm": 0.28601887822151184,
+      "learning_rate": 0.0001,
+      "loss": 0.8208,
+      "step": 1890
+    },
+    {
+      "epoch": 1.6807021442061993,
+      "grad_norm": 0.31592151522636414,
+      "learning_rate": 0.0001,
+      "loss": 0.864,
+      "step": 1891
+    },
+    {
+      "epoch": 1.6815909343406288,
+      "grad_norm": 0.3350915014743805,
+      "learning_rate": 0.0001,
+      "loss": 0.8605,
+      "step": 1892
+    },
+    {
+      "epoch": 1.6824797244750584,
+      "grad_norm": 0.37870457768440247,
+      "learning_rate": 0.0001,
+      "loss": 0.9751,
+      "step": 1893
+    },
+    {
+      "epoch": 1.6833685146094879,
+      "grad_norm": 0.3566136956214905,
+      "learning_rate": 0.0001,
+      "loss": 0.9273,
+      "step": 1894
+    },
+    {
+      "epoch": 1.6842573047439173,
+      "grad_norm": 0.3385891318321228,
+      "learning_rate": 0.0001,
+      "loss": 0.8724,
+      "step": 1895
+    },
+    {
+      "epoch": 1.685146094878347,
+      "grad_norm": 0.3746001720428467,
+      "learning_rate": 0.0001,
+      "loss": 0.9185,
+      "step": 1896
+    },
+    {
+      "epoch": 1.6860348850127762,
+      "grad_norm": 0.3757399320602417,
+      "learning_rate": 0.0001,
+      "loss": 0.8831,
+      "step": 1897
+    },
+    {
+      "epoch": 1.6869236751472059,
+      "grad_norm": 0.33278876543045044,
+      "learning_rate": 0.0001,
+      "loss": 0.8558,
+      "step": 1898
+    },
+    {
+      "epoch": 1.6878124652816355,
+      "grad_norm": 0.3170175552368164,
+      "learning_rate": 0.0001,
+      "loss": 0.8932,
+      "step": 1899
+    },
+    {
+      "epoch": 1.6887012554160648,
+      "grad_norm": 0.32355326414108276,
+      "learning_rate": 0.0001,
+      "loss": 0.8787,
+      "step": 1900
+    },
+    {
+      "epoch": 1.6895900455504944,
+      "grad_norm": 0.31958630681037903,
+      "learning_rate": 0.0001,
+      "loss": 0.9199,
+      "step": 1901
+    },
+    {
+      "epoch": 1.6904788356849239,
+      "grad_norm": 0.3621491491794586,
+      "learning_rate": 0.0001,
+      "loss": 0.8973,
+      "step": 1902
+    },
+    {
+      "epoch": 1.6913676258193533,
+      "grad_norm": 0.48310527205467224,
+      "learning_rate": 0.0001,
+      "loss": 0.8819,
+      "step": 1903
+    },
+    {
+      "epoch": 1.692256415953783,
+      "grad_norm": 0.3930363059043884,
+      "learning_rate": 0.0001,
+      "loss": 0.9095,
+      "step": 1904
+    },
+    {
+      "epoch": 1.6931452060882124,
+      "grad_norm": 0.5452322363853455,
+      "learning_rate": 0.0001,
+      "loss": 0.8147,
+      "step": 1905
+    },
+    {
+      "epoch": 1.6940339962226418,
+      "grad_norm": 0.3676657974720001,
+      "learning_rate": 0.0001,
+      "loss": 0.8653,
+      "step": 1906
+    },
+    {
+      "epoch": 1.6949227863570715,
+      "grad_norm": 0.5217211246490479,
+      "learning_rate": 0.0001,
+      "loss": 0.8173,
+      "step": 1907
+    },
+    {
+      "epoch": 1.695811576491501,
+      "grad_norm": 0.4746188223361969,
+      "learning_rate": 0.0001,
+      "loss": 0.9034,
+      "step": 1908
+    },
+    {
+      "epoch": 1.6967003666259304,
+      "grad_norm": 0.6145462989807129,
+      "learning_rate": 0.0001,
+      "loss": 0.9044,
+      "step": 1909
+    },
+    {
+      "epoch": 1.69758915676036,
+      "grad_norm": 0.43200716376304626,
+      "learning_rate": 0.0001,
+      "loss": 0.9742,
+      "step": 1910
+    },
+    {
+      "epoch": 1.6984779468947895,
+      "grad_norm": 0.3674427568912506,
+      "learning_rate": 0.0001,
+      "loss": 0.9574,
+      "step": 1911
+    },
+    {
+      "epoch": 1.699366737029219,
+      "grad_norm": 0.3205876052379608,
+      "learning_rate": 0.0001,
+      "loss": 0.83,
+      "step": 1912
+    },
+    {
+      "epoch": 1.7002555271636486,
+      "grad_norm": 0.41618865728378296,
+      "learning_rate": 0.0001,
+      "loss": 0.9021,
+      "step": 1913
+    },
+    {
+      "epoch": 1.701144317298078,
+      "grad_norm": 0.40217745304107666,
+      "learning_rate": 0.0001,
+      "loss": 0.8975,
+      "step": 1914
+    },
+    {
+      "epoch": 1.7020331074325075,
+      "grad_norm": 0.37823233008384705,
+      "learning_rate": 0.0001,
+      "loss": 0.9554,
+      "step": 1915
+    },
+    {
+      "epoch": 1.7029218975669371,
+      "grad_norm": 0.37665289640426636,
+      "learning_rate": 0.0001,
+      "loss": 0.9769,
+      "step": 1916
+    },
+    {
+      "epoch": 1.7038106877013666,
+      "grad_norm": 0.3486294746398926,
+      "learning_rate": 0.0001,
+      "loss": 0.9428,
+      "step": 1917
+    },
+    {
+      "epoch": 1.704699477835796,
+      "grad_norm": 0.40671101212501526,
+      "learning_rate": 0.0001,
+      "loss": 0.9127,
+      "step": 1918
+    },
+    {
+      "epoch": 1.7055882679702257,
+      "grad_norm": 0.35006022453308105,
+      "learning_rate": 0.0001,
+      "loss": 0.9617,
+      "step": 1919
+    },
+    {
+      "epoch": 1.7064770581046549,
+      "grad_norm": 0.36815375089645386,
+      "learning_rate": 0.0001,
+      "loss": 0.9224,
+      "step": 1920
+    },
+    {
+      "epoch": 1.7073658482390845,
+      "grad_norm": 0.34530341625213623,
+      "learning_rate": 0.0001,
+      "loss": 0.8884,
+      "step": 1921
+    },
+    {
+      "epoch": 1.7082546383735142,
+      "grad_norm": 0.3507446348667145,
+      "learning_rate": 0.0001,
+      "loss": 0.9667,
+      "step": 1922
+    },
+    {
+      "epoch": 1.7091434285079434,
+      "grad_norm": 0.34675371646881104,
+      "learning_rate": 0.0001,
+      "loss": 0.8347,
+      "step": 1923
+    },
+    {
+      "epoch": 1.710032218642373,
+      "grad_norm": 0.34880420565605164,
+      "learning_rate": 0.0001,
+      "loss": 0.8989,
+      "step": 1924
+    },
+    {
+      "epoch": 1.7109210087768025,
+      "grad_norm": 0.35518354177474976,
+      "learning_rate": 0.0001,
+      "loss": 0.8713,
+      "step": 1925
+    },
+    {
+      "epoch": 1.711809798911232,
+      "grad_norm": 0.34916481375694275,
+      "learning_rate": 0.0001,
+      "loss": 0.8733,
+      "step": 1926
+    },
+    {
+      "epoch": 1.7126985890456616,
+      "grad_norm": 0.3095919191837311,
+      "learning_rate": 0.0001,
+      "loss": 0.9293,
+      "step": 1927
+    },
+    {
+      "epoch": 1.713587379180091,
+      "grad_norm": 0.36070945858955383,
+      "learning_rate": 0.0001,
+      "loss": 0.9314,
+      "step": 1928
+    },
+    {
+      "epoch": 1.7144761693145205,
+      "grad_norm": 0.36071017384529114,
+      "learning_rate": 0.0001,
+      "loss": 1.0066,
+      "step": 1929
+    },
+    {
+      "epoch": 1.7153649594489502,
+      "grad_norm": 0.3244760036468506,
+      "learning_rate": 0.0001,
+      "loss": 0.8946,
+      "step": 1930
+    },
+    {
+      "epoch": 1.7162537495833796,
+      "grad_norm": 0.34101998805999756,
+      "learning_rate": 0.0001,
+      "loss": 0.8418,
+      "step": 1931
+    },
+    {
+      "epoch": 1.717142539717809,
+      "grad_norm": 0.38111263513565063,
+      "learning_rate": 0.0001,
+      "loss": 0.855,
+      "step": 1932
+    },
+    {
+      "epoch": 1.7180313298522387,
+      "grad_norm": 0.3378105163574219,
+      "learning_rate": 0.0001,
+      "loss": 0.8144,
+      "step": 1933
+    },
+    {
+      "epoch": 1.7189201199866682,
+      "grad_norm": 0.3542186915874481,
+      "learning_rate": 0.0001,
+      "loss": 0.8838,
+      "step": 1934
+    },
+    {
+      "epoch": 1.7198089101210976,
+      "grad_norm": 0.3762187957763672,
+      "learning_rate": 0.0001,
+      "loss": 0.8893,
+      "step": 1935
+    },
+    {
+      "epoch": 1.7206977002555273,
+      "grad_norm": 0.3594928979873657,
+      "learning_rate": 0.0001,
+      "loss": 0.9449,
+      "step": 1936
+    },
+    {
+      "epoch": 1.7215864903899567,
+      "grad_norm": 0.3138940632343292,
+      "learning_rate": 0.0001,
+      "loss": 0.9581,
+      "step": 1937
+    },
+    {
+      "epoch": 1.7224752805243861,
+      "grad_norm": 0.3370933532714844,
+      "learning_rate": 0.0001,
+      "loss": 0.903,
+      "step": 1938
+    },
+    {
+      "epoch": 1.7233640706588158,
+      "grad_norm": 0.3399626612663269,
+      "learning_rate": 0.0001,
+      "loss": 0.8864,
+      "step": 1939
+    },
+    {
+      "epoch": 1.7242528607932452,
+      "grad_norm": 0.329193115234375,
+      "learning_rate": 0.0001,
+      "loss": 0.9119,
+      "step": 1940
+    },
+    {
+      "epoch": 1.7251416509276747,
+      "grad_norm": 0.3717508316040039,
+      "learning_rate": 0.0001,
+      "loss": 0.8974,
+      "step": 1941
+    },
+    {
+      "epoch": 1.7260304410621043,
+      "grad_norm": 0.3529798090457916,
+      "learning_rate": 0.0001,
+      "loss": 0.9278,
+      "step": 1942
+    },
+    {
+      "epoch": 1.7269192311965336,
+      "grad_norm": 0.38855284452438354,
+      "learning_rate": 0.0001,
+      "loss": 0.8737,
+      "step": 1943
+    },
+    {
+      "epoch": 1.7278080213309632,
+      "grad_norm": 0.3287023901939392,
+      "learning_rate": 0.0001,
+      "loss": 0.9743,
+      "step": 1944
+    },
+    {
+      "epoch": 1.728696811465393,
+      "grad_norm": 0.3897152841091156,
+      "learning_rate": 0.0001,
+      "loss": 0.8451,
+      "step": 1945
+    },
+    {
+      "epoch": 1.729585601599822,
+      "grad_norm": 0.3124273419380188,
+      "learning_rate": 0.0001,
+      "loss": 0.8386,
+      "step": 1946
+    },
+    {
+      "epoch": 1.7304743917342518,
+      "grad_norm": 0.2802036702632904,
+      "learning_rate": 0.0001,
+      "loss": 0.8524,
+      "step": 1947
+    },
+    {
+      "epoch": 1.7313631818686812,
+      "grad_norm": 0.33942630887031555,
+      "learning_rate": 0.0001,
+      "loss": 0.8643,
+      "step": 1948
+    },
+    {
+      "epoch": 1.7322519720031107,
+      "grad_norm": 0.3702329695224762,
+      "learning_rate": 0.0001,
+      "loss": 0.9303,
+      "step": 1949
+    },
+    {
+      "epoch": 1.7331407621375403,
+      "grad_norm": 0.31775155663490295,
+      "learning_rate": 0.0001,
+      "loss": 0.8537,
+      "step": 1950
+    },
+    {
+      "epoch": 1.7340295522719698,
+      "grad_norm": 0.3505903482437134,
+      "learning_rate": 0.0001,
+      "loss": 0.8882,
+      "step": 1951
+    },
+    {
+      "epoch": 1.7349183424063992,
+      "grad_norm": 0.3102082908153534,
+      "learning_rate": 0.0001,
+      "loss": 0.8552,
+      "step": 1952
+    },
+    {
+      "epoch": 1.7358071325408289,
+      "grad_norm": 0.40489935874938965,
+      "learning_rate": 0.0001,
+      "loss": 0.945,
+      "step": 1953
+    },
+    {
+      "epoch": 1.7366959226752583,
+      "grad_norm": 0.35709646344184875,
+      "learning_rate": 0.0001,
+      "loss": 0.8496,
+      "step": 1954
+    },
+    {
+      "epoch": 1.7375847128096877,
+      "grad_norm": 0.33537212014198303,
+      "learning_rate": 0.0001,
+      "loss": 0.9205,
+      "step": 1955
+    },
+    {
+      "epoch": 1.7384735029441174,
+      "grad_norm": 0.3619324564933777,
+      "learning_rate": 0.0001,
+      "loss": 0.9342,
+      "step": 1956
+    },
+    {
+      "epoch": 1.7393622930785468,
+      "grad_norm": 0.31395861506462097,
+      "learning_rate": 0.0001,
+      "loss": 0.8663,
+      "step": 1957
+    },
+    {
+      "epoch": 1.7402510832129763,
+      "grad_norm": 0.3466954827308655,
+      "learning_rate": 0.0001,
+      "loss": 0.8824,
+      "step": 1958
+    },
+    {
+      "epoch": 1.741139873347406,
+      "grad_norm": 0.34001612663269043,
+      "learning_rate": 0.0001,
+      "loss": 0.9166,
+      "step": 1959
+    },
+    {
+      "epoch": 1.7420286634818354,
+      "grad_norm": 0.31668078899383545,
+      "learning_rate": 0.0001,
+      "loss": 0.8834,
+      "step": 1960
+    },
+    {
+      "epoch": 1.7429174536162648,
+      "grad_norm": 0.33864402770996094,
+      "learning_rate": 0.0001,
+      "loss": 0.9826,
+      "step": 1961
+    },
+    {
+      "epoch": 1.7438062437506945,
+      "grad_norm": 0.34717845916748047,
+      "learning_rate": 0.0001,
+      "loss": 0.9625,
+      "step": 1962
+    },
+    {
+      "epoch": 1.744695033885124,
+      "grad_norm": 0.3487595021724701,
+      "learning_rate": 0.0001,
+      "loss": 0.9137,
+      "step": 1963
+    },
+    {
+      "epoch": 1.7455838240195534,
+      "grad_norm": 0.28423944115638733,
+      "learning_rate": 0.0001,
+      "loss": 0.8175,
+      "step": 1964
+    },
+    {
+      "epoch": 1.746472614153983,
+      "grad_norm": 0.4026448428630829,
+      "learning_rate": 0.0001,
+      "loss": 0.9801,
+      "step": 1965
+    },
+    {
+      "epoch": 1.7473614042884122,
+      "grad_norm": 0.34367966651916504,
+      "learning_rate": 0.0001,
+      "loss": 0.908,
+      "step": 1966
+    },
+    {
+      "epoch": 1.748250194422842,
+      "grad_norm": 0.3341876268386841,
+      "learning_rate": 0.0001,
+      "loss": 0.9105,
+      "step": 1967
+    },
+    {
+      "epoch": 1.7491389845572716,
+      "grad_norm": 0.3579274117946625,
+      "learning_rate": 0.0001,
+      "loss": 0.8702,
+      "step": 1968
+    },
+    {
+      "epoch": 1.7500277746917008,
+      "grad_norm": 0.30886539816856384,
+      "learning_rate": 0.0001,
+      "loss": 0.9024,
+      "step": 1969
+    },
+    {
+      "epoch": 1.7509165648261305,
+      "grad_norm": 0.33523109555244446,
+      "learning_rate": 0.0001,
+      "loss": 0.8876,
+      "step": 1970
+    },
+    {
+      "epoch": 1.75180535496056,
+      "grad_norm": 0.32511594891548157,
+      "learning_rate": 0.0001,
+      "loss": 0.9099,
+      "step": 1971
+    },
+    {
+      "epoch": 1.7526941450949893,
+      "grad_norm": 0.3819718658924103,
+      "learning_rate": 0.0001,
+      "loss": 0.9776,
+      "step": 1972
+    },
+    {
+      "epoch": 1.753582935229419,
+      "grad_norm": 0.32984215021133423,
+      "learning_rate": 0.0001,
+      "loss": 0.9106,
+      "step": 1973
+    },
+    {
+      "epoch": 1.7544717253638484,
+      "grad_norm": 0.4052921235561371,
+      "learning_rate": 0.0001,
+      "loss": 0.9176,
+      "step": 1974
+    },
+    {
+      "epoch": 1.7553605154982779,
+      "grad_norm": 0.3490274250507355,
+      "learning_rate": 0.0001,
+      "loss": 0.9902,
+      "step": 1975
+    },
+    {
+      "epoch": 1.7562493056327075,
+      "grad_norm": 0.37016966938972473,
+      "learning_rate": 0.0001,
+      "loss": 0.9497,
+      "step": 1976
+    },
+    {
+      "epoch": 1.757138095767137,
+      "grad_norm": 0.3575468957424164,
+      "learning_rate": 0.0001,
+      "loss": 0.864,
+      "step": 1977
+    },
+    {
+      "epoch": 1.7580268859015664,
+      "grad_norm": 0.34194570779800415,
+      "learning_rate": 0.0001,
+      "loss": 0.9862,
+      "step": 1978
+    },
+    {
+      "epoch": 1.758915676035996,
+      "grad_norm": 0.361395001411438,
+      "learning_rate": 0.0001,
+      "loss": 0.8745,
+      "step": 1979
+    },
+    {
+      "epoch": 1.7598044661704255,
+      "grad_norm": 0.3402308225631714,
+      "learning_rate": 0.0001,
+      "loss": 0.9091,
+      "step": 1980
+    },
+    {
+      "epoch": 1.760693256304855,
+      "grad_norm": 0.3822394013404846,
+      "learning_rate": 0.0001,
+      "loss": 0.9479,
+      "step": 1981
+    },
+    {
+      "epoch": 1.7615820464392846,
+      "grad_norm": 0.3259856700897217,
+      "learning_rate": 0.0001,
+      "loss": 0.912,
+      "step": 1982
+    },
+    {
+      "epoch": 1.762470836573714,
+      "grad_norm": 0.40468284487724304,
+      "learning_rate": 0.0001,
+      "loss": 0.8525,
+      "step": 1983
+    },
+    {
+      "epoch": 1.7633596267081435,
+      "grad_norm": 0.4105701446533203,
+      "learning_rate": 0.0001,
+      "loss": 0.8759,
+      "step": 1984
+    },
+    {
+      "epoch": 1.7642484168425732,
+      "grad_norm": 0.5805673003196716,
+      "learning_rate": 0.0001,
+      "loss": 1.0095,
+      "step": 1985
+    },
+    {
+      "epoch": 1.7651372069770026,
+      "grad_norm": 0.6856684684753418,
+      "learning_rate": 0.0001,
+      "loss": 0.9672,
+      "step": 1986
+    },
+    {
+      "epoch": 1.766025997111432,
+      "grad_norm": 0.3889956474304199,
+      "learning_rate": 0.0001,
+      "loss": 0.8923,
+      "step": 1987
+    },
+    {
+      "epoch": 1.7669147872458617,
+      "grad_norm": 0.37936434149742126,
+      "learning_rate": 0.0001,
+      "loss": 0.9341,
+      "step": 1988
+    },
+    {
+      "epoch": 1.767803577380291,
+      "grad_norm": 0.34393537044525146,
+      "learning_rate": 0.0001,
+      "loss": 0.8836,
+      "step": 1989
+    },
+    {
+      "epoch": 1.7686923675147206,
+      "grad_norm": 0.34456318616867065,
+      "learning_rate": 0.0001,
+      "loss": 0.9741,
+      "step": 1990
+    },
+    {
+      "epoch": 1.7695811576491502,
+      "grad_norm": 0.32842886447906494,
+      "learning_rate": 0.0001,
+      "loss": 0.9137,
+      "step": 1991
+    },
+    {
+      "epoch": 1.7704699477835795,
+      "grad_norm": 0.35680335760116577,
+      "learning_rate": 0.0001,
+      "loss": 0.8332,
+      "step": 1992
+    },
+    {
+      "epoch": 1.7713587379180091,
+      "grad_norm": 0.3687342703342438,
+      "learning_rate": 0.0001,
+      "loss": 0.9065,
+      "step": 1993
+    },
+    {
+      "epoch": 1.7722475280524386,
+      "grad_norm": 0.39165356755256653,
+      "learning_rate": 0.0001,
+      "loss": 0.933,
+      "step": 1994
+    },
+    {
+      "epoch": 1.773136318186868,
+      "grad_norm": 0.36600562930107117,
+      "learning_rate": 0.0001,
+      "loss": 0.8694,
+      "step": 1995
+    },
+    {
+      "epoch": 1.7740251083212977,
+      "grad_norm": 0.3371865749359131,
+      "learning_rate": 0.0001,
+      "loss": 0.9622,
+      "step": 1996
+    },
+    {
+      "epoch": 1.7749138984557271,
+      "grad_norm": 0.3732564449310303,
+      "learning_rate": 0.0001,
+      "loss": 1.0046,
+      "step": 1997
+    },
+    {
+      "epoch": 1.7758026885901566,
+      "grad_norm": 0.33094069361686707,
+      "learning_rate": 0.0001,
+      "loss": 0.8466,
+      "step": 1998
+    },
+    {
+      "epoch": 1.7766914787245862,
+      "grad_norm": 0.3384708762168884,
+      "learning_rate": 0.0001,
+      "loss": 0.855,
+      "step": 1999
+    },
+    {
+      "epoch": 1.7775802688590157,
+      "grad_norm": 0.3246244788169861,
+      "learning_rate": 0.0001,
+      "loss": 0.9129,
+      "step": 2000
+    },
+    {
+      "epoch": 1.778469058993445,
+      "grad_norm": 0.3717402517795563,
+      "learning_rate": 0.0001,
+      "loss": 0.9298,
+      "step": 2001
+    },
+    {
+      "epoch": 1.7793578491278748,
+      "grad_norm": 0.4107438325881958,
+      "learning_rate": 0.0001,
+      "loss": 0.9007,
+      "step": 2002
+    },
+    {
+      "epoch": 1.7802466392623042,
+      "grad_norm": 0.355356901884079,
+      "learning_rate": 0.0001,
+      "loss": 0.8573,
+      "step": 2003
+    },
+    {
+      "epoch": 1.7811354293967336,
+      "grad_norm": 0.34428539872169495,
+      "learning_rate": 0.0001,
+      "loss": 0.9481,
+      "step": 2004
+    },
+    {
+      "epoch": 1.7820242195311633,
+      "grad_norm": 0.3407171666622162,
+      "learning_rate": 0.0001,
+      "loss": 0.9211,
+      "step": 2005
+    },
+    {
+      "epoch": 1.7829130096655927,
+      "grad_norm": 0.3513103425502777,
+      "learning_rate": 0.0001,
+      "loss": 0.9179,
+      "step": 2006
+    },
+    {
+      "epoch": 1.7838017998000222,
+      "grad_norm": 0.32477250695228577,
+      "learning_rate": 0.0001,
+      "loss": 0.8792,
+      "step": 2007
+    },
+    {
+      "epoch": 1.7846905899344518,
+      "grad_norm": 0.3490789234638214,
+      "learning_rate": 0.0001,
+      "loss": 0.9442,
+      "step": 2008
+    },
+    {
+      "epoch": 1.7855793800688813,
+      "grad_norm": 0.3724386692047119,
+      "learning_rate": 0.0001,
+      "loss": 0.9503,
+      "step": 2009
+    },
+    {
+      "epoch": 1.7864681702033107,
+      "grad_norm": 0.3316227197647095,
+      "learning_rate": 0.0001,
+      "loss": 0.9598,
+      "step": 2010
+    },
+    {
+      "epoch": 1.7873569603377404,
+      "grad_norm": 0.35122936964035034,
+      "learning_rate": 0.0001,
+      "loss": 0.8149,
+      "step": 2011
+    },
+    {
+      "epoch": 1.7882457504721696,
+      "grad_norm": 0.4029070734977722,
+      "learning_rate": 0.0001,
+      "loss": 0.9305,
+      "step": 2012
+    },
+    {
+      "epoch": 1.7891345406065993,
+      "grad_norm": 0.31067872047424316,
+      "learning_rate": 0.0001,
+      "loss": 0.8878,
+      "step": 2013
+    },
+    {
+      "epoch": 1.790023330741029,
+      "grad_norm": 0.34684422612190247,
+      "learning_rate": 0.0001,
+      "loss": 0.9143,
+      "step": 2014
+    },
+    {
+      "epoch": 1.7909121208754581,
+      "grad_norm": 0.32135358452796936,
+      "learning_rate": 0.0001,
+      "loss": 0.926,
+      "step": 2015
+    },
+    {
+      "epoch": 1.7918009110098878,
+      "grad_norm": 0.32894784212112427,
+      "learning_rate": 0.0001,
+      "loss": 0.9249,
+      "step": 2016
+    },
+    {
+      "epoch": 1.7926897011443172,
+      "grad_norm": 0.35362952947616577,
+      "learning_rate": 0.0001,
+      "loss": 0.9004,
+      "step": 2017
+    },
+    {
+      "epoch": 1.7935784912787467,
+      "grad_norm": 0.3030708432197571,
+      "learning_rate": 0.0001,
+      "loss": 1.0155,
+      "step": 2018
+    },
+    {
+      "epoch": 1.7944672814131764,
+      "grad_norm": 0.3144596219062805,
+      "learning_rate": 0.0001,
+      "loss": 0.9367,
+      "step": 2019
+    },
+    {
+      "epoch": 1.7953560715476058,
+      "grad_norm": 0.34851983189582825,
+      "learning_rate": 0.0001,
+      "loss": 0.8684,
+      "step": 2020
+    },
+    {
+      "epoch": 1.7962448616820352,
+      "grad_norm": 0.32248032093048096,
+      "learning_rate": 0.0001,
+      "loss": 0.8977,
+      "step": 2021
+    },
+    {
+      "epoch": 1.797133651816465,
+      "grad_norm": 0.3193919062614441,
+      "learning_rate": 0.0001,
+      "loss": 0.9297,
+      "step": 2022
+    },
+    {
+      "epoch": 1.7980224419508943,
+      "grad_norm": 0.3607577085494995,
+      "learning_rate": 0.0001,
+      "loss": 0.9774,
+      "step": 2023
+    },
+    {
+      "epoch": 1.7989112320853238,
+      "grad_norm": 0.30439522862434387,
+      "learning_rate": 0.0001,
+      "loss": 0.9313,
+      "step": 2024
+    },
+    {
+      "epoch": 1.7998000222197534,
+      "grad_norm": 0.31033584475517273,
+      "learning_rate": 0.0001,
+      "loss": 0.9535,
+      "step": 2025
+    },
+    {
+      "epoch": 1.8006888123541829,
+      "grad_norm": 0.34924376010894775,
+      "learning_rate": 0.0001,
+      "loss": 1.0392,
+      "step": 2026
+    },
+    {
+      "epoch": 1.8015776024886123,
+      "grad_norm": 0.3580887019634247,
+      "learning_rate": 0.0001,
+      "loss": 0.8772,
+      "step": 2027
+    },
+    {
+      "epoch": 1.802466392623042,
+      "grad_norm": 0.30029070377349854,
+      "learning_rate": 0.0001,
+      "loss": 0.904,
+      "step": 2028
+    },
+    {
+      "epoch": 1.8033551827574714,
+      "grad_norm": 0.36260783672332764,
+      "learning_rate": 0.0001,
+      "loss": 0.8598,
+      "step": 2029
+    },
+    {
+      "epoch": 1.8042439728919009,
+      "grad_norm": 0.3398006558418274,
+      "learning_rate": 0.0001,
+      "loss": 0.926,
+      "step": 2030
+    },
+    {
+      "epoch": 1.8051327630263305,
+      "grad_norm": 0.31382572650909424,
+      "learning_rate": 0.0001,
+      "loss": 0.871,
+      "step": 2031
+    },
+    {
+      "epoch": 1.80602155316076,
+      "grad_norm": 0.33854883909225464,
+      "learning_rate": 0.0001,
+      "loss": 0.9493,
+      "step": 2032
+    },
+    {
+      "epoch": 1.8069103432951894,
+      "grad_norm": 0.334824800491333,
+      "learning_rate": 0.0001,
+      "loss": 0.8451,
+      "step": 2033
+    },
+    {
+      "epoch": 1.807799133429619,
+      "grad_norm": 0.31903523206710815,
+      "learning_rate": 0.0001,
+      "loss": 0.8204,
+      "step": 2034
+    },
+    {
+      "epoch": 1.8086879235640483,
+      "grad_norm": 0.351361483335495,
+      "learning_rate": 0.0001,
+      "loss": 0.8484,
+      "step": 2035
+    },
+    {
+      "epoch": 1.809576713698478,
+      "grad_norm": 0.3341825604438782,
+      "learning_rate": 0.0001,
+      "loss": 0.9075,
+      "step": 2036
+    },
+    {
+      "epoch": 1.8104655038329076,
+      "grad_norm": 0.3398033678531647,
+      "learning_rate": 0.0001,
+      "loss": 0.9491,
+      "step": 2037
+    },
+    {
+      "epoch": 1.8113542939673368,
+      "grad_norm": 0.3391786813735962,
+      "learning_rate": 0.0001,
+      "loss": 0.9576,
+      "step": 2038
+    },
+    {
+      "epoch": 1.8122430841017665,
+      "grad_norm": 0.31477200984954834,
+      "learning_rate": 0.0001,
+      "loss": 0.8863,
+      "step": 2039
+    },
+    {
+      "epoch": 1.813131874236196,
+      "grad_norm": 0.31994014978408813,
+      "learning_rate": 0.0001,
+      "loss": 0.903,
+      "step": 2040
+    },
+    {
+      "epoch": 1.8140206643706254,
+      "grad_norm": 0.38486775755882263,
+      "learning_rate": 0.0001,
+      "loss": 0.9182,
+      "step": 2041
+    },
+    {
+      "epoch": 1.814909454505055,
+      "grad_norm": 0.3258659839630127,
+      "learning_rate": 0.0001,
+      "loss": 0.8561,
+      "step": 2042
+    },
+    {
+      "epoch": 1.8157982446394845,
+      "grad_norm": 0.36081910133361816,
+      "learning_rate": 0.0001,
+      "loss": 0.9326,
+      "step": 2043
+    },
+    {
+      "epoch": 1.816687034773914,
+      "grad_norm": 0.33326515555381775,
+      "learning_rate": 0.0001,
+      "loss": 0.941,
+      "step": 2044
+    },
+    {
+      "epoch": 1.8175758249083436,
+      "grad_norm": 0.3134666383266449,
+      "learning_rate": 0.0001,
+      "loss": 0.7828,
+      "step": 2045
+    },
+    {
+      "epoch": 1.818464615042773,
+      "grad_norm": 0.318588525056839,
+      "learning_rate": 0.0001,
+      "loss": 0.974,
+      "step": 2046
+    },
+    {
+      "epoch": 1.8193534051772025,
+      "grad_norm": 0.3159397840499878,
+      "learning_rate": 0.0001,
+      "loss": 0.8277,
+      "step": 2047
+    },
+    {
+      "epoch": 1.8202421953116321,
+      "grad_norm": 0.3912515938282013,
+      "learning_rate": 0.0001,
+      "loss": 0.9024,
+      "step": 2048
+    },
+    {
+      "epoch": 1.8211309854460616,
+      "grad_norm": 0.3366560935974121,
+      "learning_rate": 0.0001,
+      "loss": 0.9652,
+      "step": 2049
+    },
+    {
+      "epoch": 1.822019775580491,
+      "grad_norm": 0.33880704641342163,
+      "learning_rate": 0.0001,
+      "loss": 0.9178,
+      "step": 2050
+    },
+    {
+      "epoch": 1.8229085657149207,
+      "grad_norm": 0.3547055721282959,
+      "learning_rate": 0.0001,
+      "loss": 0.8965,
+      "step": 2051
+    },
+    {
+      "epoch": 1.82379735584935,
+      "grad_norm": 0.3428584337234497,
+      "learning_rate": 0.0001,
+      "loss": 0.8787,
+      "step": 2052
+    },
+    {
+      "epoch": 1.8246861459837795,
+      "grad_norm": 0.3745492398738861,
+      "learning_rate": 0.0001,
+      "loss": 0.9322,
+      "step": 2053
+    },
+    {
+      "epoch": 1.8255749361182092,
+      "grad_norm": 0.32607874274253845,
+      "learning_rate": 0.0001,
+      "loss": 0.8768,
+      "step": 2054
+    },
+    {
+      "epoch": 1.8264637262526386,
+      "grad_norm": 0.5843383073806763,
+      "learning_rate": 0.0001,
+      "loss": 0.9578,
+      "step": 2055
+    },
+    {
+      "epoch": 1.827352516387068,
+      "grad_norm": 0.3096837103366852,
+      "learning_rate": 0.0001,
+      "loss": 0.8705,
+      "step": 2056
+    },
+    {
+      "epoch": 1.8282413065214977,
+      "grad_norm": 0.32685980200767517,
+      "learning_rate": 0.0001,
+      "loss": 0.907,
+      "step": 2057
+    },
+    {
+      "epoch": 1.829130096655927,
+      "grad_norm": 0.3481847643852234,
+      "learning_rate": 0.0001,
+      "loss": 0.9312,
+      "step": 2058
+    },
+    {
+      "epoch": 1.8300188867903566,
+      "grad_norm": 0.3510027229785919,
+      "learning_rate": 0.0001,
+      "loss": 0.9272,
+      "step": 2059
+    },
+    {
+      "epoch": 1.8309076769247863,
+      "grad_norm": 0.3125600814819336,
+      "learning_rate": 0.0001,
+      "loss": 0.8919,
+      "step": 2060
+    },
+    {
+      "epoch": 1.8317964670592155,
+      "grad_norm": 0.343363881111145,
+      "learning_rate": 0.0001,
+      "loss": 0.9294,
+      "step": 2061
+    },
+    {
+      "epoch": 1.8326852571936452,
+      "grad_norm": 0.30843785405158997,
+      "learning_rate": 0.0001,
+      "loss": 0.9,
+      "step": 2062
+    },
+    {
+      "epoch": 1.8335740473280746,
+      "grad_norm": 0.38246282935142517,
+      "learning_rate": 0.0001,
+      "loss": 0.9301,
+      "step": 2063
+    },
+    {
+      "epoch": 1.834462837462504,
+      "grad_norm": 0.30538472533226013,
+      "learning_rate": 0.0001,
+      "loss": 1.0181,
+      "step": 2064
+    },
+    {
+      "epoch": 1.8353516275969337,
+      "grad_norm": 0.3176038861274719,
+      "learning_rate": 0.0001,
+      "loss": 0.8301,
+      "step": 2065
+    },
+    {
+      "epoch": 1.8362404177313632,
+      "grad_norm": 0.3275960683822632,
+      "learning_rate": 0.0001,
+      "loss": 0.8787,
+      "step": 2066
+    },
+    {
+      "epoch": 1.8371292078657926,
+      "grad_norm": 0.3405584394931793,
+      "learning_rate": 0.0001,
+      "loss": 0.8562,
+      "step": 2067
+    },
+    {
+      "epoch": 1.8380179980002223,
+      "grad_norm": 0.3240833580493927,
+      "learning_rate": 0.0001,
+      "loss": 0.8733,
+      "step": 2068
+    },
+    {
+      "epoch": 1.8389067881346517,
+      "grad_norm": 0.3166552484035492,
+      "learning_rate": 0.0001,
+      "loss": 0.9458,
+      "step": 2069
+    },
+    {
+      "epoch": 1.8397955782690811,
+      "grad_norm": 0.3474705219268799,
+      "learning_rate": 0.0001,
+      "loss": 0.8703,
+      "step": 2070
+    },
+    {
+      "epoch": 1.8406843684035108,
+      "grad_norm": 0.3441227376461029,
+      "learning_rate": 0.0001,
+      "loss": 0.9149,
+      "step": 2071
+    },
+    {
+      "epoch": 1.8415731585379402,
+      "grad_norm": 0.3015057146549225,
+      "learning_rate": 0.0001,
+      "loss": 0.8754,
+      "step": 2072
+    },
+    {
+      "epoch": 1.8424619486723697,
+      "grad_norm": 0.33986207842826843,
+      "learning_rate": 0.0001,
+      "loss": 0.9179,
+      "step": 2073
+    },
+    {
+      "epoch": 1.8433507388067993,
+      "grad_norm": 0.33093711733818054,
+      "learning_rate": 0.0001,
+      "loss": 0.9072,
+      "step": 2074
+    },
+    {
+      "epoch": 1.8442395289412288,
+      "grad_norm": 0.338056743144989,
+      "learning_rate": 0.0001,
+      "loss": 0.8816,
+      "step": 2075
+    },
+    {
+      "epoch": 1.8451283190756582,
+      "grad_norm": 0.32622647285461426,
+      "learning_rate": 0.0001,
+      "loss": 0.9017,
+      "step": 2076
+    },
+    {
+      "epoch": 1.8460171092100879,
+      "grad_norm": 0.33805835247039795,
+      "learning_rate": 0.0001,
+      "loss": 0.9135,
+      "step": 2077
+    },
+    {
+      "epoch": 1.8469058993445173,
+      "grad_norm": 0.3360745906829834,
+      "learning_rate": 0.0001,
+      "loss": 0.8238,
+      "step": 2078
+    },
+    {
+      "epoch": 1.8477946894789468,
+      "grad_norm": 0.3394176959991455,
+      "learning_rate": 0.0001,
+      "loss": 0.9539,
+      "step": 2079
+    },
+    {
+      "epoch": 1.8486834796133764,
+      "grad_norm": 0.3259996175765991,
+      "learning_rate": 0.0001,
+      "loss": 0.8874,
+      "step": 2080
+    },
+    {
+      "epoch": 1.8495722697478056,
+      "grad_norm": 0.30118152499198914,
+      "learning_rate": 0.0001,
+      "loss": 0.8875,
+      "step": 2081
+    },
+    {
+      "epoch": 1.8504610598822353,
+      "grad_norm": 0.3274901211261749,
+      "learning_rate": 0.0001,
+      "loss": 0.9579,
+      "step": 2082
+    },
+    {
+      "epoch": 1.851349850016665,
+      "grad_norm": 0.34811556339263916,
+      "learning_rate": 0.0001,
+      "loss": 0.9363,
+      "step": 2083
+    },
+    {
+      "epoch": 1.8522386401510942,
+      "grad_norm": 0.35961511731147766,
+      "learning_rate": 0.0001,
+      "loss": 0.8803,
+      "step": 2084
+    },
+    {
+      "epoch": 1.8531274302855238,
+      "grad_norm": 0.3167450726032257,
+      "learning_rate": 0.0001,
+      "loss": 0.8289,
+      "step": 2085
+    },
+    {
+      "epoch": 1.8540162204199533,
+      "grad_norm": 0.32763344049453735,
+      "learning_rate": 0.0001,
+      "loss": 0.8965,
+      "step": 2086
+    },
+    {
+      "epoch": 1.8549050105543827,
+      "grad_norm": 0.32265403866767883,
+      "learning_rate": 0.0001,
+      "loss": 0.9522,
+      "step": 2087
+    },
+    {
+      "epoch": 1.8557938006888124,
+      "grad_norm": 0.3623351752758026,
+      "learning_rate": 0.0001,
+      "loss": 0.9497,
+      "step": 2088
+    },
+    {
+      "epoch": 1.8566825908232418,
+      "grad_norm": 0.34022286534309387,
+      "learning_rate": 0.0001,
+      "loss": 0.9482,
+      "step": 2089
+    },
+    {
+      "epoch": 1.8575713809576713,
+      "grad_norm": 0.34971117973327637,
+      "learning_rate": 0.0001,
+      "loss": 0.9088,
+      "step": 2090
+    },
+    {
+      "epoch": 1.858460171092101,
+      "grad_norm": 0.35238999128341675,
+      "learning_rate": 0.0001,
+      "loss": 0.893,
+      "step": 2091
+    },
+    {
+      "epoch": 1.8593489612265304,
+      "grad_norm": 0.3588760197162628,
+      "learning_rate": 0.0001,
+      "loss": 0.9416,
+      "step": 2092
+    },
+    {
+      "epoch": 1.8602377513609598,
+      "grad_norm": 0.3537149131298065,
+      "learning_rate": 0.0001,
+      "loss": 0.9008,
+      "step": 2093
+    },
+    {
+      "epoch": 1.8611265414953895,
+      "grad_norm": 0.35885384678840637,
+      "learning_rate": 0.0001,
+      "loss": 0.9227,
+      "step": 2094
+    },
+    {
+      "epoch": 1.862015331629819,
+      "grad_norm": 0.44651269912719727,
+      "learning_rate": 0.0001,
+      "loss": 0.8945,
+      "step": 2095
+    },
+    {
+      "epoch": 1.8629041217642484,
+      "grad_norm": 0.8166044354438782,
+      "learning_rate": 0.0001,
+      "loss": 0.9453,
+      "step": 2096
+    },
+    {
+      "epoch": 1.863792911898678,
+      "grad_norm": 0.4329817593097687,
+      "learning_rate": 0.0001,
+      "loss": 0.9554,
+      "step": 2097
+    },
+    {
+      "epoch": 1.8646817020331075,
+      "grad_norm": 0.3563162684440613,
+      "learning_rate": 0.0001,
+      "loss": 0.9519,
+      "step": 2098
+    },
+    {
+      "epoch": 1.865570492167537,
+      "grad_norm": 0.3440234065055847,
+      "learning_rate": 0.0001,
+      "loss": 0.9086,
+      "step": 2099
+    },
+    {
+      "epoch": 1.8664592823019666,
+      "grad_norm": 0.5276257991790771,
+      "learning_rate": 0.0001,
+      "loss": 0.9137,
+      "step": 2100
+    },
+    {
+      "epoch": 1.867348072436396,
+      "grad_norm": 0.39348065853118896,
+      "learning_rate": 0.0001,
+      "loss": 0.9041,
+      "step": 2101
+    },
+    {
+      "epoch": 1.8682368625708254,
+      "grad_norm": 0.3502536416053772,
+      "learning_rate": 0.0001,
+      "loss": 0.9357,
+      "step": 2102
+    },
+    {
+      "epoch": 1.869125652705255,
+      "grad_norm": 0.3747837245464325,
+      "learning_rate": 0.0001,
+      "loss": 0.9268,
+      "step": 2103
+    },
+    {
+      "epoch": 1.8700144428396843,
+      "grad_norm": 0.3654274344444275,
+      "learning_rate": 0.0001,
+      "loss": 0.9001,
+      "step": 2104
+    },
+    {
+      "epoch": 1.870903232974114,
+      "grad_norm": 0.34981828927993774,
+      "learning_rate": 0.0001,
+      "loss": 0.8879,
+      "step": 2105
+    },
+    {
+      "epoch": 1.8717920231085436,
+      "grad_norm": 0.34973669052124023,
+      "learning_rate": 0.0001,
+      "loss": 0.9614,
+      "step": 2106
+    },
+    {
+      "epoch": 1.8726808132429729,
+      "grad_norm": 0.35739874839782715,
+      "learning_rate": 0.0001,
+      "loss": 0.9628,
+      "step": 2107
+    },
+    {
+      "epoch": 1.8735696033774025,
+      "grad_norm": 0.33049049973487854,
+      "learning_rate": 0.0001,
+      "loss": 0.8869,
+      "step": 2108
+    },
+    {
+      "epoch": 1.874458393511832,
+      "grad_norm": 0.3104158639907837,
+      "learning_rate": 0.0001,
+      "loss": 0.8905,
+      "step": 2109
+    },
+    {
+      "epoch": 1.8753471836462614,
+      "grad_norm": 0.3537830114364624,
+      "learning_rate": 0.0001,
+      "loss": 0.9178,
+      "step": 2110
+    },
+    {
+      "epoch": 1.876235973780691,
+      "grad_norm": 0.36299172043800354,
+      "learning_rate": 0.0001,
+      "loss": 0.8937,
+      "step": 2111
+    },
+    {
+      "epoch": 1.8771247639151205,
+      "grad_norm": 0.30172330141067505,
+      "learning_rate": 0.0001,
+      "loss": 0.8999,
+      "step": 2112
+    },
+    {
+      "epoch": 1.87801355404955,
+      "grad_norm": 0.34734198451042175,
+      "learning_rate": 0.0001,
+      "loss": 0.8742,
+      "step": 2113
+    },
+    {
+      "epoch": 1.8789023441839796,
+      "grad_norm": 0.3040066659450531,
+      "learning_rate": 0.0001,
+      "loss": 0.8276,
+      "step": 2114
+    },
+    {
+      "epoch": 1.879791134318409,
+      "grad_norm": 0.38280028104782104,
+      "learning_rate": 0.0001,
+      "loss": 0.8841,
+      "step": 2115
+    },
+    {
+      "epoch": 1.8806799244528385,
+      "grad_norm": 0.3382204473018646,
+      "learning_rate": 0.0001,
+      "loss": 0.9274,
+      "step": 2116
+    },
+    {
+      "epoch": 1.8815687145872682,
+      "grad_norm": 0.3138622045516968,
+      "learning_rate": 0.0001,
+      "loss": 0.8872,
+      "step": 2117
+    },
+    {
+      "epoch": 1.8824575047216976,
+      "grad_norm": 0.37905508279800415,
+      "learning_rate": 0.0001,
+      "loss": 0.9759,
+      "step": 2118
+    },
+    {
+      "epoch": 1.883346294856127,
+      "grad_norm": 0.36490878462791443,
+      "learning_rate": 0.0001,
+      "loss": 0.935,
+      "step": 2119
+    },
+    {
+      "epoch": 1.8842350849905567,
+      "grad_norm": 0.30611562728881836,
+      "learning_rate": 0.0001,
+      "loss": 0.8862,
+      "step": 2120
+    },
+    {
+      "epoch": 1.8851238751249861,
+      "grad_norm": 0.30281272530555725,
+      "learning_rate": 0.0001,
+      "loss": 0.9486,
+      "step": 2121
+    },
+    {
+      "epoch": 1.8860126652594156,
+      "grad_norm": 0.3309618830680847,
+      "learning_rate": 0.0001,
+      "loss": 0.9686,
+      "step": 2122
+    },
+    {
+      "epoch": 1.8869014553938452,
+      "grad_norm": 0.36201414465904236,
+      "learning_rate": 0.0001,
+      "loss": 0.8934,
+      "step": 2123
+    },
+    {
+      "epoch": 1.8877902455282747,
+      "grad_norm": 0.3028503656387329,
+      "learning_rate": 0.0001,
+      "loss": 0.916,
+      "step": 2124
+    },
+    {
+      "epoch": 1.8886790356627041,
+      "grad_norm": 0.3417774736881256,
+      "learning_rate": 0.0001,
+      "loss": 0.96,
+      "step": 2125
+    },
+    {
+      "epoch": 1.8895678257971338,
+      "grad_norm": 0.3560897707939148,
+      "learning_rate": 0.0001,
+      "loss": 0.9264,
+      "step": 2126
+    },
+    {
+      "epoch": 1.890456615931563,
+      "grad_norm": 0.33282026648521423,
+      "learning_rate": 0.0001,
+      "loss": 0.9066,
+      "step": 2127
+    },
+    {
+      "epoch": 1.8913454060659927,
+      "grad_norm": 0.3309515416622162,
+      "learning_rate": 0.0001,
+      "loss": 0.8867,
+      "step": 2128
+    },
+    {
+      "epoch": 1.8922341962004223,
+      "grad_norm": 0.32819992303848267,
+      "learning_rate": 0.0001,
+      "loss": 0.9102,
+      "step": 2129
+    },
+    {
+      "epoch": 1.8931229863348515,
+      "grad_norm": 0.3443058133125305,
+      "learning_rate": 0.0001,
+      "loss": 0.9559,
+      "step": 2130
+    },
+    {
+      "epoch": 1.8940117764692812,
+      "grad_norm": 0.3516992926597595,
+      "learning_rate": 0.0001,
+      "loss": 0.917,
+      "step": 2131
+    },
+    {
+      "epoch": 1.8949005666037106,
+      "grad_norm": 0.32620400190353394,
+      "learning_rate": 0.0001,
+      "loss": 0.8944,
+      "step": 2132
+    },
+    {
+      "epoch": 1.89578935673814,
+      "grad_norm": 0.34699490666389465,
+      "learning_rate": 0.0001,
+      "loss": 0.87,
+      "step": 2133
+    },
+    {
+      "epoch": 1.8966781468725697,
+      "grad_norm": 0.32767391204833984,
+      "learning_rate": 0.0001,
+      "loss": 0.9523,
+      "step": 2134
+    },
+    {
+      "epoch": 1.8975669370069992,
+      "grad_norm": 0.34350037574768066,
+      "learning_rate": 0.0001,
+      "loss": 0.9449,
+      "step": 2135
+    },
+    {
+      "epoch": 1.8984557271414286,
+      "grad_norm": 0.36585548520088196,
+      "learning_rate": 0.0001,
+      "loss": 0.9338,
+      "step": 2136
+    },
+    {
+      "epoch": 1.8993445172758583,
+      "grad_norm": 0.3513524830341339,
+      "learning_rate": 0.0001,
+      "loss": 0.9429,
+      "step": 2137
+    },
+    {
+      "epoch": 1.9002333074102877,
+      "grad_norm": 0.3541501462459564,
+      "learning_rate": 0.0001,
+      "loss": 0.8976,
+      "step": 2138
+    },
+    {
+      "epoch": 1.9011220975447172,
+      "grad_norm": 0.3008817434310913,
+      "learning_rate": 0.0001,
+      "loss": 0.8689,
+      "step": 2139
+    },
+    {
+      "epoch": 1.9020108876791468,
+      "grad_norm": 0.3268721401691437,
+      "learning_rate": 0.0001,
+      "loss": 0.8712,
+      "step": 2140
+    },
+    {
+      "epoch": 1.9028996778135763,
+      "grad_norm": 0.3459062874317169,
+      "learning_rate": 0.0001,
+      "loss": 0.9399,
+      "step": 2141
+    },
+    {
+      "epoch": 1.9037884679480057,
+      "grad_norm": 0.3577088713645935,
+      "learning_rate": 0.0001,
+      "loss": 0.9029,
+      "step": 2142
+    },
+    {
+      "epoch": 1.9046772580824354,
+      "grad_norm": 0.34948304295539856,
+      "learning_rate": 0.0001,
+      "loss": 0.8714,
+      "step": 2143
+    },
+    {
+      "epoch": 1.9055660482168648,
+      "grad_norm": 0.34985479712486267,
+      "learning_rate": 0.0001,
+      "loss": 0.902,
+      "step": 2144
+    },
+    {
+      "epoch": 1.9064548383512943,
+      "grad_norm": 0.32033130526542664,
+      "learning_rate": 0.0001,
+      "loss": 0.8456,
+      "step": 2145
+    },
+    {
+      "epoch": 1.907343628485724,
+      "grad_norm": 0.32239830493927,
+      "learning_rate": 0.0001,
+      "loss": 0.9245,
+      "step": 2146
+    },
+    {
+      "epoch": 1.9082324186201534,
+      "grad_norm": 0.35271769762039185,
+      "learning_rate": 0.0001,
+      "loss": 0.9427,
+      "step": 2147
+    },
+    {
+      "epoch": 1.9091212087545828,
+      "grad_norm": 0.34471169114112854,
+      "learning_rate": 0.0001,
+      "loss": 0.9369,
+      "step": 2148
+    },
+    {
+      "epoch": 1.9100099988890125,
+      "grad_norm": 0.2854446470737457,
+      "learning_rate": 0.0001,
+      "loss": 0.8902,
+      "step": 2149
+    },
+    {
+      "epoch": 1.9108987890234417,
+      "grad_norm": 0.32251161336898804,
+      "learning_rate": 0.0001,
+      "loss": 0.8536,
+      "step": 2150
+    },
+    {
+      "epoch": 1.9117875791578713,
+      "grad_norm": 0.33492353558540344,
+      "learning_rate": 0.0001,
+      "loss": 0.8761,
+      "step": 2151
+    },
+    {
+      "epoch": 1.912676369292301,
+      "grad_norm": 0.3147629201412201,
+      "learning_rate": 0.0001,
+      "loss": 0.8802,
+      "step": 2152
+    },
+    {
+      "epoch": 1.9135651594267302,
+      "grad_norm": 0.32801344990730286,
+      "learning_rate": 0.0001,
+      "loss": 0.8123,
+      "step": 2153
+    },
+    {
+      "epoch": 1.9144539495611599,
+      "grad_norm": 0.2836940586566925,
+      "learning_rate": 0.0001,
+      "loss": 0.9012,
+      "step": 2154
+    },
+    {
+      "epoch": 1.9153427396955893,
+      "grad_norm": 0.3101726770401001,
+      "learning_rate": 0.0001,
+      "loss": 0.8274,
+      "step": 2155
+    },
+    {
+      "epoch": 1.9162315298300188,
+      "grad_norm": 0.3067977726459503,
+      "learning_rate": 0.0001,
+      "loss": 0.8986,
+      "step": 2156
+    },
+    {
+      "epoch": 1.9171203199644484,
+      "grad_norm": 0.27377212047576904,
+      "learning_rate": 0.0001,
+      "loss": 0.8844,
+      "step": 2157
+    },
+    {
+      "epoch": 1.9180091100988779,
+      "grad_norm": 0.36923712491989136,
+      "learning_rate": 0.0001,
+      "loss": 0.9264,
+      "step": 2158
+    },
+    {
+      "epoch": 1.9188979002333073,
+      "grad_norm": 0.3664684295654297,
+      "learning_rate": 0.0001,
+      "loss": 0.953,
+      "step": 2159
+    },
+    {
+      "epoch": 1.919786690367737,
+      "grad_norm": 0.3705950081348419,
+      "learning_rate": 0.0001,
+      "loss": 0.873,
+      "step": 2160
+    },
+    {
+      "epoch": 1.9206754805021664,
+      "grad_norm": 0.33334994316101074,
+      "learning_rate": 0.0001,
+      "loss": 0.8896,
+      "step": 2161
+    },
+    {
+      "epoch": 1.9215642706365959,
+      "grad_norm": 0.32042577862739563,
+      "learning_rate": 0.0001,
+      "loss": 0.832,
+      "step": 2162
+    },
+    {
+      "epoch": 1.9224530607710255,
+      "grad_norm": 0.2969491183757782,
+      "learning_rate": 0.0001,
+      "loss": 0.8963,
+      "step": 2163
+    },
+    {
+      "epoch": 1.923341850905455,
+      "grad_norm": 0.3007069528102875,
+      "learning_rate": 0.0001,
+      "loss": 0.8812,
+      "step": 2164
+    },
+    {
+      "epoch": 1.9242306410398844,
+      "grad_norm": 0.34228307008743286,
+      "learning_rate": 0.0001,
+      "loss": 0.9252,
+      "step": 2165
+    },
+    {
+      "epoch": 1.925119431174314,
+      "grad_norm": 0.34786513447761536,
+      "learning_rate": 0.0001,
+      "loss": 0.8791,
+      "step": 2166
+    },
+    {
+      "epoch": 1.9260082213087435,
+      "grad_norm": 0.30037039518356323,
+      "learning_rate": 0.0001,
+      "loss": 0.8397,
+      "step": 2167
+    },
+    {
+      "epoch": 1.926897011443173,
+      "grad_norm": 0.29265427589416504,
+      "learning_rate": 0.0001,
+      "loss": 0.8792,
+      "step": 2168
+    },
+    {
+      "epoch": 1.9277858015776026,
+      "grad_norm": 0.35532552003860474,
+      "learning_rate": 0.0001,
+      "loss": 0.9429,
+      "step": 2169
+    },
+    {
+      "epoch": 1.928674591712032,
+      "grad_norm": 0.3534153997898102,
+      "learning_rate": 0.0001,
+      "loss": 0.8774,
+      "step": 2170
+    },
+    {
+      "epoch": 1.9295633818464615,
+      "grad_norm": 0.29537761211395264,
+      "learning_rate": 0.0001,
+      "loss": 0.9146,
+      "step": 2171
+    },
+    {
+      "epoch": 1.9304521719808911,
+      "grad_norm": 0.34193140268325806,
+      "learning_rate": 0.0001,
+      "loss": 0.9506,
+      "step": 2172
+    },
+    {
+      "epoch": 1.9313409621153204,
+      "grad_norm": 0.33322104811668396,
+      "learning_rate": 0.0001,
+      "loss": 0.965,
+      "step": 2173
+    },
+    {
+      "epoch": 1.93222975224975,
+      "grad_norm": 0.34304308891296387,
+      "learning_rate": 0.0001,
+      "loss": 0.9704,
+      "step": 2174
+    },
+    {
+      "epoch": 1.9331185423841797,
+      "grad_norm": 0.3284062445163727,
+      "learning_rate": 0.0001,
+      "loss": 0.7752,
+      "step": 2175
+    },
+    {
+      "epoch": 1.934007332518609,
+      "grad_norm": 0.34800034761428833,
+      "learning_rate": 0.0001,
+      "loss": 0.9484,
+      "step": 2176
+    },
+    {
+      "epoch": 1.9348961226530386,
+      "grad_norm": 0.34290215373039246,
+      "learning_rate": 0.0001,
+      "loss": 0.9032,
+      "step": 2177
+    },
+    {
+      "epoch": 1.935784912787468,
+      "grad_norm": 0.34737300872802734,
+      "learning_rate": 0.0001,
+      "loss": 0.9158,
+      "step": 2178
+    },
+    {
+      "epoch": 1.9366737029218974,
+      "grad_norm": 0.34877312183380127,
+      "learning_rate": 0.0001,
+      "loss": 0.8934,
+      "step": 2179
+    },
+    {
+      "epoch": 1.937562493056327,
+      "grad_norm": 0.3398269712924957,
+      "learning_rate": 0.0001,
+      "loss": 0.9713,
+      "step": 2180
+    },
+    {
+      "epoch": 1.9384512831907565,
+      "grad_norm": 0.33079251646995544,
+      "learning_rate": 0.0001,
+      "loss": 1.01,
+      "step": 2181
+    },
+    {
+      "epoch": 1.939340073325186,
+      "grad_norm": 0.3067607581615448,
+      "learning_rate": 0.0001,
+      "loss": 0.8631,
+      "step": 2182
+    },
+    {
+      "epoch": 1.9402288634596156,
+      "grad_norm": 0.34091418981552124,
+      "learning_rate": 0.0001,
+      "loss": 0.9544,
+      "step": 2183
+    },
+    {
+      "epoch": 1.941117653594045,
+      "grad_norm": 0.30287304520606995,
+      "learning_rate": 0.0001,
+      "loss": 0.941,
+      "step": 2184
+    },
+    {
+      "epoch": 1.9420064437284745,
+      "grad_norm": 0.2976100742816925,
+      "learning_rate": 0.0001,
+      "loss": 0.9028,
+      "step": 2185
+    },
+    {
+      "epoch": 1.9428952338629042,
+      "grad_norm": 0.3236287236213684,
+      "learning_rate": 0.0001,
+      "loss": 0.9633,
+      "step": 2186
+    },
+    {
+      "epoch": 1.9437840239973336,
+      "grad_norm": 0.35301899909973145,
+      "learning_rate": 0.0001,
+      "loss": 0.9464,
+      "step": 2187
+    },
+    {
+      "epoch": 1.944672814131763,
+      "grad_norm": 0.35564854741096497,
+      "learning_rate": 0.0001,
+      "loss": 1.0202,
+      "step": 2188
+    },
+    {
+      "epoch": 1.9455616042661927,
+      "grad_norm": 0.31409159302711487,
+      "learning_rate": 0.0001,
+      "loss": 0.965,
+      "step": 2189
+    },
+    {
+      "epoch": 1.9464503944006222,
+      "grad_norm": 0.3617132306098938,
+      "learning_rate": 0.0001,
+      "loss": 0.9124,
+      "step": 2190
+    },
+    {
+      "epoch": 1.9473391845350516,
+      "grad_norm": 0.2927257716655731,
+      "learning_rate": 0.0001,
+      "loss": 0.9333,
+      "step": 2191
+    },
+    {
+      "epoch": 1.9482279746694813,
+      "grad_norm": 0.3173227906227112,
+      "learning_rate": 0.0001,
+      "loss": 0.8952,
+      "step": 2192
+    },
+    {
+      "epoch": 1.9491167648039107,
+      "grad_norm": 0.29953786730766296,
+      "learning_rate": 0.0001,
+      "loss": 0.8728,
+      "step": 2193
+    },
+    {
+      "epoch": 1.9500055549383402,
+      "grad_norm": 0.36522650718688965,
+      "learning_rate": 0.0001,
+      "loss": 0.8915,
+      "step": 2194
+    },
+    {
+      "epoch": 1.9508943450727698,
+      "grad_norm": 0.36083757877349854,
+      "learning_rate": 0.0001,
+      "loss": 0.9012,
+      "step": 2195
+    },
+    {
+      "epoch": 1.951783135207199,
+      "grad_norm": 0.3177006244659424,
+      "learning_rate": 0.0001,
+      "loss": 0.8705,
+      "step": 2196
+    },
+    {
+      "epoch": 1.9526719253416287,
+      "grad_norm": 0.33240652084350586,
+      "learning_rate": 0.0001,
+      "loss": 0.9201,
+      "step": 2197
+    },
+    {
+      "epoch": 1.9535607154760584,
+      "grad_norm": 0.33079612255096436,
+      "learning_rate": 0.0001,
+      "loss": 0.8434,
+      "step": 2198
+    },
+    {
+      "epoch": 1.9544495056104876,
+      "grad_norm": 0.3984503149986267,
+      "learning_rate": 0.0001,
+      "loss": 0.9963,
+      "step": 2199
+    },
+    {
+      "epoch": 1.9553382957449172,
+      "grad_norm": 0.33371084928512573,
+      "learning_rate": 0.0001,
+      "loss": 0.9055,
+      "step": 2200
+    },
+    {
+      "epoch": 1.956227085879347,
+      "grad_norm": 0.3662305772304535,
+      "learning_rate": 0.0001,
+      "loss": 0.8911,
+      "step": 2201
+    },
+    {
+      "epoch": 1.9571158760137761,
+      "grad_norm": 0.3294123411178589,
+      "learning_rate": 0.0001,
+      "loss": 0.9329,
+      "step": 2202
+    },
+    {
+      "epoch": 1.9580046661482058,
+      "grad_norm": 0.32030513882637024,
+      "learning_rate": 0.0001,
+      "loss": 0.9697,
+      "step": 2203
+    },
+    {
+      "epoch": 1.9588934562826352,
+      "grad_norm": 0.3227706551551819,
+      "learning_rate": 0.0001,
+      "loss": 0.8552,
+      "step": 2204
+    },
+    {
+      "epoch": 1.9597822464170647,
+      "grad_norm": 0.3211442828178406,
+      "learning_rate": 0.0001,
+      "loss": 0.8235,
+      "step": 2205
+    },
+    {
+      "epoch": 1.9606710365514943,
+      "grad_norm": 0.3925526440143585,
+      "learning_rate": 0.0001,
+      "loss": 0.8938,
+      "step": 2206
+    },
+    {
+      "epoch": 1.9615598266859238,
+      "grad_norm": 0.2847107946872711,
+      "learning_rate": 0.0001,
+      "loss": 0.8749,
+      "step": 2207
+    },
+    {
+      "epoch": 1.9624486168203532,
+      "grad_norm": 0.35088127851486206,
+      "learning_rate": 0.0001,
+      "loss": 0.9328,
+      "step": 2208
+    },
+    {
+      "epoch": 1.9633374069547829,
+      "grad_norm": 0.3825710713863373,
+      "learning_rate": 0.0001,
+      "loss": 0.8982,
+      "step": 2209
+    },
+    {
+      "epoch": 1.9642261970892123,
+      "grad_norm": 0.3291476368904114,
+      "learning_rate": 0.0001,
+      "loss": 0.8782,
+      "step": 2210
+    },
+    {
+      "epoch": 1.9651149872236418,
+      "grad_norm": 0.3212360441684723,
+      "learning_rate": 0.0001,
+      "loss": 0.9278,
+      "step": 2211
+    },
+    {
+      "epoch": 1.9660037773580714,
+      "grad_norm": 0.3502598702907562,
+      "learning_rate": 0.0001,
+      "loss": 0.8873,
+      "step": 2212
+    },
+    {
+      "epoch": 1.9668925674925009,
+      "grad_norm": 0.3515768051147461,
+      "learning_rate": 0.0001,
+      "loss": 0.8633,
+      "step": 2213
+    },
+    {
+      "epoch": 1.9677813576269303,
+      "grad_norm": 0.340701162815094,
+      "learning_rate": 0.0001,
+      "loss": 0.9178,
+      "step": 2214
+    },
+    {
+      "epoch": 1.96867014776136,
+      "grad_norm": 0.3128719925880432,
+      "learning_rate": 0.0001,
+      "loss": 0.8933,
+      "step": 2215
+    },
+    {
+      "epoch": 1.9695589378957894,
+      "grad_norm": 0.34365659952163696,
+      "learning_rate": 0.0001,
+      "loss": 0.9546,
+      "step": 2216
+    },
+    {
+      "epoch": 1.9704477280302188,
+      "grad_norm": 0.31693172454833984,
+      "learning_rate": 0.0001,
+      "loss": 0.8905,
+      "step": 2217
+    },
+    {
+      "epoch": 1.9713365181646485,
+      "grad_norm": 0.36978334188461304,
+      "learning_rate": 0.0001,
+      "loss": 0.8712,
+      "step": 2218
+    },
+    {
+      "epoch": 1.9722253082990777,
+      "grad_norm": 0.31684744358062744,
+      "learning_rate": 0.0001,
+      "loss": 0.9069,
+      "step": 2219
+    },
+    {
+      "epoch": 1.9731140984335074,
+      "grad_norm": 0.3603816330432892,
+      "learning_rate": 0.0001,
+      "loss": 0.9562,
+      "step": 2220
+    },
+    {
+      "epoch": 1.974002888567937,
+      "grad_norm": 0.3792661428451538,
+      "learning_rate": 0.0001,
+      "loss": 0.9382,
+      "step": 2221
+    },
+    {
+      "epoch": 1.9748916787023663,
+      "grad_norm": 0.4025600254535675,
+      "learning_rate": 0.0001,
+      "loss": 0.9571,
+      "step": 2222
+    },
+    {
+      "epoch": 1.975780468836796,
+      "grad_norm": 0.3250136077404022,
+      "learning_rate": 0.0001,
+      "loss": 0.849,
+      "step": 2223
+    },
+    {
+      "epoch": 1.9766692589712256,
+      "grad_norm": 0.3059399425983429,
+      "learning_rate": 0.0001,
+      "loss": 0.9364,
+      "step": 2224
+    },
+    {
+      "epoch": 1.9775580491056548,
+      "grad_norm": 0.2995017468929291,
+      "learning_rate": 0.0001,
+      "loss": 0.8323,
+      "step": 2225
+    },
+    {
+      "epoch": 1.9784468392400845,
+      "grad_norm": 0.33151498436927795,
+      "learning_rate": 0.0001,
+      "loss": 0.9296,
+      "step": 2226
+    },
+    {
+      "epoch": 1.979335629374514,
+      "grad_norm": 0.36364221572875977,
+      "learning_rate": 0.0001,
+      "loss": 0.9274,
+      "step": 2227
+    },
+    {
+      "epoch": 1.9802244195089433,
+      "grad_norm": 0.3598885238170624,
+      "learning_rate": 0.0001,
+      "loss": 0.9747,
+      "step": 2228
+    },
+    {
+      "epoch": 1.981113209643373,
+      "grad_norm": 0.3541521430015564,
+      "learning_rate": 0.0001,
+      "loss": 0.9024,
+      "step": 2229
+    },
+    {
+      "epoch": 1.9820019997778024,
+      "grad_norm": 0.312953382730484,
+      "learning_rate": 0.0001,
+      "loss": 0.8793,
+      "step": 2230
+    },
+    {
+      "epoch": 1.9828907899122319,
+      "grad_norm": 0.3080136477947235,
+      "learning_rate": 0.0001,
+      "loss": 0.8732,
+      "step": 2231
+    },
+    {
+      "epoch": 1.9837795800466616,
+      "grad_norm": 0.32917580008506775,
+      "learning_rate": 0.0001,
+      "loss": 0.8911,
+      "step": 2232
+    },
+    {
+      "epoch": 1.984668370181091,
+      "grad_norm": 0.34910324215888977,
+      "learning_rate": 0.0001,
+      "loss": 0.9626,
+      "step": 2233
+    },
+    {
+      "epoch": 1.9855571603155204,
+      "grad_norm": 0.38250458240509033,
+      "learning_rate": 0.0001,
+      "loss": 0.9059,
+      "step": 2234
+    },
+    {
+      "epoch": 1.98644595044995,
+      "grad_norm": 0.34895434975624084,
+      "learning_rate": 0.0001,
+      "loss": 0.9963,
+      "step": 2235
+    },
+    {
+      "epoch": 1.9873347405843795,
+      "grad_norm": 0.3334400951862335,
+      "learning_rate": 0.0001,
+      "loss": 0.8768,
+      "step": 2236
+    },
+    {
+      "epoch": 1.988223530718809,
+      "grad_norm": 0.33048558235168457,
+      "learning_rate": 0.0001,
+      "loss": 0.9229,
+      "step": 2237
+    },
+    {
+      "epoch": 1.9891123208532386,
+      "grad_norm": 0.3428661823272705,
+      "learning_rate": 0.0001,
+      "loss": 0.9175,
+      "step": 2238
+    },
+    {
+      "epoch": 1.990001110987668,
+      "grad_norm": 0.3587013781070709,
+      "learning_rate": 0.0001,
+      "loss": 0.9277,
+      "step": 2239
+    },
+    {
+      "epoch": 1.9908899011220975,
+      "grad_norm": 0.3457068204879761,
+      "learning_rate": 0.0001,
+      "loss": 0.8732,
+      "step": 2240
+    },
+    {
+      "epoch": 1.9917786912565272,
+      "grad_norm": 0.3348812162876129,
+      "learning_rate": 0.0001,
+      "loss": 0.903,
+      "step": 2241
+    },
+    {
+      "epoch": 1.9926674813909564,
+      "grad_norm": 0.3357104957103729,
+      "learning_rate": 0.0001,
+      "loss": 0.898,
+      "step": 2242
+    },
+    {
+      "epoch": 1.993556271525386,
+      "grad_norm": 0.346883088350296,
+      "learning_rate": 0.0001,
+      "loss": 0.9087,
+      "step": 2243
+    },
+    {
+      "epoch": 1.9944450616598157,
+      "grad_norm": 0.30482059717178345,
+      "learning_rate": 0.0001,
+      "loss": 0.9312,
+      "step": 2244
+    },
+    {
+      "epoch": 1.995333851794245,
+      "grad_norm": 0.34484198689460754,
+      "learning_rate": 0.0001,
+      "loss": 0.8785,
+      "step": 2245
+    },
+    {
+      "epoch": 1.9962226419286746,
+      "grad_norm": 0.2776843309402466,
+      "learning_rate": 0.0001,
+      "loss": 0.8985,
+      "step": 2246
+    },
+    {
+      "epoch": 1.9971114320631043,
+      "grad_norm": 0.33060091733932495,
+      "learning_rate": 0.0001,
+      "loss": 0.9854,
+      "step": 2247
+    },
+    {
+      "epoch": 1.9980002221975335,
+      "grad_norm": 0.31561362743377686,
+      "learning_rate": 0.0001,
+      "loss": 0.8927,
+      "step": 2248
+    },
+    {
+      "epoch": 1.9988890123319631,
+      "grad_norm": 0.29034098982810974,
+      "learning_rate": 0.0001,
+      "loss": 0.868,
+      "step": 2249
+    },
+    {
+      "epoch": 1.9997778024663926,
+      "grad_norm": 0.3135145604610443,
+      "learning_rate": 0.0001,
+      "loss": 0.7991,
+      "step": 2250
+    },
+    {
+      "epoch": 2.000666592600822,
+      "grad_norm": 0.36708906292915344,
+      "learning_rate": 0.0001,
+      "loss": 0.8307,
+      "step": 2251
+    },
+    {
+      "epoch": 2.0015553827352517,
+      "grad_norm": 0.3554810881614685,
+      "learning_rate": 0.0001,
+      "loss": 0.9073,
+      "step": 2252
+    },
+    {
+      "epoch": 2.0024441728696813,
+      "grad_norm": 0.3140956461429596,
+      "learning_rate": 0.0001,
+      "loss": 0.8779,
+      "step": 2253
+    },
+    {
+      "epoch": 2.0033329630041106,
+      "grad_norm": 0.3324379622936249,
+      "learning_rate": 0.0001,
+      "loss": 0.8612,
+      "step": 2254
+    },
+    {
+      "epoch": 2.0042217531385402,
+      "grad_norm": 0.3427555561065674,
+      "learning_rate": 0.0001,
+      "loss": 0.8843,
+      "step": 2255
+    },
+    {
+      "epoch": 2.00511054327297,
+      "grad_norm": 0.298017293214798,
+      "learning_rate": 0.0001,
+      "loss": 0.8393,
+      "step": 2256
+    },
+    {
+      "epoch": 2.005999333407399,
+      "grad_norm": 0.3473068177700043,
+      "learning_rate": 0.0001,
+      "loss": 0.9666,
+      "step": 2257
+    },
+    {
+      "epoch": 2.0068881235418288,
+      "grad_norm": 0.32553350925445557,
+      "learning_rate": 0.0001,
+      "loss": 0.9294,
+      "step": 2258
+    },
+    {
+      "epoch": 2.007776913676258,
+      "grad_norm": 0.29851260781288147,
+      "learning_rate": 0.0001,
+      "loss": 0.9186,
+      "step": 2259
+    },
+    {
+      "epoch": 2.0086657038106877,
+      "grad_norm": 0.3403533101081848,
+      "learning_rate": 0.0001,
+      "loss": 0.8777,
+      "step": 2260
+    },
+    {
+      "epoch": 2.0095544939451173,
+      "grad_norm": 0.31564658880233765,
+      "learning_rate": 0.0001,
+      "loss": 0.8203,
+      "step": 2261
+    },
+    {
+      "epoch": 2.0104432840795465,
+      "grad_norm": 0.34301063418388367,
+      "learning_rate": 0.0001,
+      "loss": 0.8448,
+      "step": 2262
+    },
+    {
+      "epoch": 2.011332074213976,
+      "grad_norm": 0.34970736503601074,
+      "learning_rate": 0.0001,
+      "loss": 0.8925,
+      "step": 2263
+    },
+    {
+      "epoch": 2.012220864348406,
+      "grad_norm": 0.3418786823749542,
+      "learning_rate": 0.0001,
+      "loss": 0.9322,
+      "step": 2264
+    },
+    {
+      "epoch": 2.013109654482835,
+      "grad_norm": 0.3406841456890106,
+      "learning_rate": 0.0001,
+      "loss": 0.8289,
+      "step": 2265
+    },
+    {
+      "epoch": 2.0139984446172647,
+      "grad_norm": 0.35538503527641296,
+      "learning_rate": 0.0001,
+      "loss": 0.9544,
+      "step": 2266
+    },
+    {
+      "epoch": 2.0148872347516944,
+      "grad_norm": 0.35400450229644775,
+      "learning_rate": 0.0001,
+      "loss": 0.9106,
+      "step": 2267
+    },
+    {
+      "epoch": 2.0157760248861236,
+      "grad_norm": 0.3589840829372406,
+      "learning_rate": 0.0001,
+      "loss": 0.9033,
+      "step": 2268
+    },
+    {
+      "epoch": 2.0166648150205533,
+      "grad_norm": 0.3830343782901764,
+      "learning_rate": 0.0001,
+      "loss": 0.9254,
+      "step": 2269
+    },
+    {
+      "epoch": 2.017553605154983,
+      "grad_norm": 0.3211663067340851,
+      "learning_rate": 0.0001,
+      "loss": 0.8952,
+      "step": 2270
+    },
+    {
+      "epoch": 2.018442395289412,
+      "grad_norm": 0.37262219190597534,
+      "learning_rate": 0.0001,
+      "loss": 0.9151,
+      "step": 2271
+    },
+    {
+      "epoch": 2.019331185423842,
+      "grad_norm": 0.34682542085647583,
+      "learning_rate": 0.0001,
+      "loss": 0.8174,
+      "step": 2272
+    },
+    {
+      "epoch": 2.0202199755582715,
+      "grad_norm": 0.3054455816745758,
+      "learning_rate": 0.0001,
+      "loss": 0.8983,
+      "step": 2273
+    },
+    {
+      "epoch": 2.0211087656927007,
+      "grad_norm": 0.2940855622291565,
+      "learning_rate": 0.0001,
+      "loss": 0.8573,
+      "step": 2274
+    },
+    {
+      "epoch": 2.0219975558271304,
+      "grad_norm": 0.37542182207107544,
+      "learning_rate": 0.0001,
+      "loss": 0.9686,
+      "step": 2275
+    },
+    {
+      "epoch": 2.02288634596156,
+      "grad_norm": 0.29696211218833923,
+      "learning_rate": 0.0001,
+      "loss": 0.8948,
+      "step": 2276
+    },
+    {
+      "epoch": 2.0237751360959892,
+      "grad_norm": 0.35127562284469604,
+      "learning_rate": 0.0001,
+      "loss": 0.9528,
+      "step": 2277
+    },
+    {
+      "epoch": 2.024663926230419,
+      "grad_norm": 0.3054238259792328,
+      "learning_rate": 0.0001,
+      "loss": 0.8935,
+      "step": 2278
+    },
+    {
+      "epoch": 2.0255527163648486,
+      "grad_norm": 0.33190521597862244,
+      "learning_rate": 0.0001,
+      "loss": 0.8616,
+      "step": 2279
+    },
+    {
+      "epoch": 2.026441506499278,
+      "grad_norm": 0.3622536361217499,
+      "learning_rate": 0.0001,
+      "loss": 0.9324,
+      "step": 2280
+    },
+    {
+      "epoch": 2.0273302966337075,
+      "grad_norm": 0.32910165190696716,
+      "learning_rate": 0.0001,
+      "loss": 0.8391,
+      "step": 2281
+    },
+    {
+      "epoch": 2.0282190867681367,
+      "grad_norm": 0.31704333424568176,
+      "learning_rate": 0.0001,
+      "loss": 0.8393,
+      "step": 2282
+    },
+    {
+      "epoch": 2.0291078769025663,
+      "grad_norm": 0.39209648966789246,
+      "learning_rate": 0.0001,
+      "loss": 0.8637,
+      "step": 2283
+    },
+    {
+      "epoch": 2.029996667036996,
+      "grad_norm": 0.3536166846752167,
+      "learning_rate": 0.0001,
+      "loss": 0.8957,
+      "step": 2284
+    },
+    {
+      "epoch": 2.030885457171425,
+      "grad_norm": 0.33245381712913513,
+      "learning_rate": 0.0001,
+      "loss": 0.8915,
+      "step": 2285
+    },
+    {
+      "epoch": 2.031774247305855,
+      "grad_norm": 0.3466053307056427,
+      "learning_rate": 0.0001,
+      "loss": 0.8671,
+      "step": 2286
+    },
+    {
+      "epoch": 2.0326630374402845,
+      "grad_norm": 0.32629281282424927,
+      "learning_rate": 0.0001,
+      "loss": 0.93,
+      "step": 2287
+    },
+    {
+      "epoch": 2.0335518275747138,
+      "grad_norm": 0.321918785572052,
+      "learning_rate": 0.0001,
+      "loss": 0.8497,
+      "step": 2288
+    },
+    {
+      "epoch": 2.0344406177091434,
+      "grad_norm": 0.3338482081890106,
+      "learning_rate": 0.0001,
+      "loss": 0.8984,
+      "step": 2289
+    },
+    {
+      "epoch": 2.035329407843573,
+      "grad_norm": 0.32754746079444885,
+      "learning_rate": 0.0001,
+      "loss": 0.8787,
+      "step": 2290
+    },
+    {
+      "epoch": 2.0362181979780023,
+      "grad_norm": 0.34894460439682007,
+      "learning_rate": 0.0001,
+      "loss": 0.8761,
+      "step": 2291
+    },
+    {
+      "epoch": 2.037106988112432,
+      "grad_norm": 0.34283894300460815,
+      "learning_rate": 0.0001,
+      "loss": 0.9411,
+      "step": 2292
+    },
+    {
+      "epoch": 2.0379957782468616,
+      "grad_norm": 0.3812194764614105,
+      "learning_rate": 0.0001,
+      "loss": 0.9111,
+      "step": 2293
+    },
+    {
+      "epoch": 2.038884568381291,
+      "grad_norm": 0.323310911655426,
+      "learning_rate": 0.0001,
+      "loss": 0.8984,
+      "step": 2294
+    },
+    {
+      "epoch": 2.0397733585157205,
+      "grad_norm": 0.376645565032959,
+      "learning_rate": 0.0001,
+      "loss": 0.8604,
+      "step": 2295
+    },
+    {
+      "epoch": 2.04066214865015,
+      "grad_norm": 0.3399569094181061,
+      "learning_rate": 0.0001,
+      "loss": 0.8495,
+      "step": 2296
+    },
+    {
+      "epoch": 2.0415509387845794,
+      "grad_norm": 0.3920535147190094,
+      "learning_rate": 0.0001,
+      "loss": 0.8972,
+      "step": 2297
+    },
+    {
+      "epoch": 2.042439728919009,
+      "grad_norm": 0.3221544623374939,
+      "learning_rate": 0.0001,
+      "loss": 0.9215,
+      "step": 2298
+    },
+    {
+      "epoch": 2.0433285190534387,
+      "grad_norm": 0.32029077410697937,
+      "learning_rate": 0.0001,
+      "loss": 0.8555,
+      "step": 2299
+    },
+    {
+      "epoch": 2.044217309187868,
+      "grad_norm": 0.3429443836212158,
+      "learning_rate": 0.0001,
+      "loss": 0.8572,
+      "step": 2300
+    },
+    {
+      "epoch": 2.0451060993222976,
+      "grad_norm": 0.32114291191101074,
+      "learning_rate": 0.0001,
+      "loss": 0.8466,
+      "step": 2301
+    },
+    {
+      "epoch": 2.0459948894567273,
+      "grad_norm": 0.35281145572662354,
+      "learning_rate": 0.0001,
+      "loss": 0.9114,
+      "step": 2302
+    },
+    {
+      "epoch": 2.0468836795911565,
+      "grad_norm": 0.32281407713890076,
+      "learning_rate": 0.0001,
+      "loss": 0.9293,
+      "step": 2303
+    },
+    {
+      "epoch": 2.047772469725586,
+      "grad_norm": 0.3905143439769745,
+      "learning_rate": 0.0001,
+      "loss": 0.9392,
+      "step": 2304
+    },
+    {
+      "epoch": 2.0486612598600153,
+      "grad_norm": 0.34570732712745667,
+      "learning_rate": 0.0001,
+      "loss": 0.8764,
+      "step": 2305
+    },
+    {
+      "epoch": 2.049550049994445,
+      "grad_norm": 0.34093308448791504,
+      "learning_rate": 0.0001,
+      "loss": 0.9282,
+      "step": 2306
+    },
+    {
+      "epoch": 2.0504388401288747,
+      "grad_norm": 0.3642049729824066,
+      "learning_rate": 0.0001,
+      "loss": 0.8468,
+      "step": 2307
+    },
+    {
+      "epoch": 2.051327630263304,
+      "grad_norm": 0.3626710772514343,
+      "learning_rate": 0.0001,
+      "loss": 0.8742,
+      "step": 2308
+    },
+    {
+      "epoch": 2.0522164203977336,
+      "grad_norm": 0.3760134279727936,
+      "learning_rate": 0.0001,
+      "loss": 0.8199,
+      "step": 2309
+    },
+    {
+      "epoch": 2.053105210532163,
+      "grad_norm": 0.3483445346355438,
+      "learning_rate": 0.0001,
+      "loss": 0.9071,
+      "step": 2310
+    },
+    {
+      "epoch": 2.0539940006665924,
+      "grad_norm": 0.3375813663005829,
+      "learning_rate": 0.0001,
+      "loss": 0.8971,
+      "step": 2311
+    },
+    {
+      "epoch": 2.054882790801022,
+      "grad_norm": 0.3339730501174927,
+      "learning_rate": 0.0001,
+      "loss": 0.8204,
+      "step": 2312
+    },
+    {
+      "epoch": 2.0557715809354518,
+      "grad_norm": 0.36098751425743103,
+      "learning_rate": 0.0001,
+      "loss": 0.8845,
+      "step": 2313
+    },
+    {
+      "epoch": 2.056660371069881,
+      "grad_norm": 0.34348028898239136,
+      "learning_rate": 0.0001,
+      "loss": 0.8818,
+      "step": 2314
+    },
+    {
+      "epoch": 2.0575491612043106,
+      "grad_norm": 0.36092281341552734,
+      "learning_rate": 0.0001,
+      "loss": 0.9322,
+      "step": 2315
+    },
+    {
+      "epoch": 2.0584379513387403,
+      "grad_norm": 0.4044939875602722,
+      "learning_rate": 0.0001,
+      "loss": 0.868,
+      "step": 2316
+    },
+    {
+      "epoch": 2.0593267414731695,
+      "grad_norm": 0.3279127776622772,
+      "learning_rate": 0.0001,
+      "loss": 0.886,
+      "step": 2317
+    },
+    {
+      "epoch": 2.060215531607599,
+      "grad_norm": 0.3734513819217682,
+      "learning_rate": 0.0001,
+      "loss": 0.8934,
+      "step": 2318
+    },
+    {
+      "epoch": 2.061104321742029,
+      "grad_norm": 0.3154412508010864,
+      "learning_rate": 0.0001,
+      "loss": 0.8556,
+      "step": 2319
+    },
+    {
+      "epoch": 2.061993111876458,
+      "grad_norm": 0.35183727741241455,
+      "learning_rate": 0.0001,
+      "loss": 0.9645,
+      "step": 2320
+    },
+    {
+      "epoch": 2.0628819020108877,
+      "grad_norm": 0.35789453983306885,
+      "learning_rate": 0.0001,
+      "loss": 0.8836,
+      "step": 2321
+    },
+    {
+      "epoch": 2.0637706921453174,
+      "grad_norm": 0.3455105423927307,
+      "learning_rate": 0.0001,
+      "loss": 0.9554,
+      "step": 2322
+    },
+    {
+      "epoch": 2.0646594822797466,
+      "grad_norm": 0.35150641202926636,
+      "learning_rate": 0.0001,
+      "loss": 0.8408,
+      "step": 2323
+    },
+    {
+      "epoch": 2.0655482724141763,
+      "grad_norm": 0.3445546627044678,
+      "learning_rate": 0.0001,
+      "loss": 0.9087,
+      "step": 2324
+    },
+    {
+      "epoch": 2.066437062548606,
+      "grad_norm": 0.3782254755496979,
+      "learning_rate": 0.0001,
+      "loss": 0.9708,
+      "step": 2325
+    },
+    {
+      "epoch": 2.067325852683035,
+      "grad_norm": 0.30935153365135193,
+      "learning_rate": 0.0001,
+      "loss": 0.8883,
+      "step": 2326
+    },
+    {
+      "epoch": 2.068214642817465,
+      "grad_norm": 0.34302398562431335,
+      "learning_rate": 0.0001,
+      "loss": 0.9016,
+      "step": 2327
+    },
+    {
+      "epoch": 2.069103432951894,
+      "grad_norm": 0.38530564308166504,
+      "learning_rate": 0.0001,
+      "loss": 0.8869,
+      "step": 2328
+    },
+    {
+      "epoch": 2.0699922230863237,
+      "grad_norm": 0.34200990200042725,
+      "learning_rate": 0.0001,
+      "loss": 0.8758,
+      "step": 2329
+    },
+    {
+      "epoch": 2.0708810132207534,
+      "grad_norm": 0.36103734374046326,
+      "learning_rate": 0.0001,
+      "loss": 0.9055,
+      "step": 2330
+    },
+    {
+      "epoch": 2.0717698033551826,
+      "grad_norm": 0.36420273780822754,
+      "learning_rate": 0.0001,
+      "loss": 0.9257,
+      "step": 2331
+    },
+    {
+      "epoch": 2.0726585934896122,
+      "grad_norm": 0.3854941725730896,
+      "learning_rate": 0.0001,
+      "loss": 0.9891,
+      "step": 2332
+    },
+    {
+      "epoch": 2.073547383624042,
+      "grad_norm": 0.33720192313194275,
+      "learning_rate": 0.0001,
+      "loss": 0.8197,
+      "step": 2333
+    },
+    {
+      "epoch": 2.074436173758471,
+      "grad_norm": 0.3435489237308502,
+      "learning_rate": 0.0001,
+      "loss": 0.8943,
+      "step": 2334
+    },
+    {
+      "epoch": 2.0753249638929008,
+      "grad_norm": 0.33030179142951965,
+      "learning_rate": 0.0001,
+      "loss": 0.8799,
+      "step": 2335
+    },
+    {
+      "epoch": 2.0762137540273304,
+      "grad_norm": 0.48584482073783875,
+      "learning_rate": 0.0001,
+      "loss": 0.8402,
+      "step": 2336
+    },
+    {
+      "epoch": 2.0771025441617597,
+      "grad_norm": 0.39869070053100586,
+      "learning_rate": 0.0001,
+      "loss": 0.8668,
+      "step": 2337
+    },
+    {
+      "epoch": 2.0779913342961893,
+      "grad_norm": 0.3875821530818939,
+      "learning_rate": 0.0001,
+      "loss": 0.9544,
+      "step": 2338
+    },
+    {
+      "epoch": 2.078880124430619,
+      "grad_norm": 0.3594411611557007,
+      "learning_rate": 0.0001,
+      "loss": 0.9415,
+      "step": 2339
+    },
+    {
+      "epoch": 2.079768914565048,
+      "grad_norm": 0.34104394912719727,
+      "learning_rate": 0.0001,
+      "loss": 0.8698,
+      "step": 2340
+    },
+    {
+      "epoch": 2.080657704699478,
+      "grad_norm": 0.3437696695327759,
+      "learning_rate": 0.0001,
+      "loss": 0.8795,
+      "step": 2341
+    },
+    {
+      "epoch": 2.0815464948339075,
+      "grad_norm": 0.3021574318408966,
+      "learning_rate": 0.0001,
+      "loss": 0.8896,
+      "step": 2342
+    },
+    {
+      "epoch": 2.0824352849683367,
+      "grad_norm": 0.306111216545105,
+      "learning_rate": 0.0001,
+      "loss": 0.8579,
+      "step": 2343
+    },
+    {
+      "epoch": 2.0833240751027664,
+      "grad_norm": 0.3199199140071869,
+      "learning_rate": 0.0001,
+      "loss": 0.8107,
+      "step": 2344
+    },
+    {
+      "epoch": 2.084212865237196,
+      "grad_norm": 0.34308746457099915,
+      "learning_rate": 0.0001,
+      "loss": 0.9077,
+      "step": 2345
+    },
+    {
+      "epoch": 2.0851016553716253,
+      "grad_norm": 0.366335928440094,
+      "learning_rate": 0.0001,
+      "loss": 0.9323,
+      "step": 2346
+    },
+    {
+      "epoch": 2.085990445506055,
+      "grad_norm": 0.3313388228416443,
+      "learning_rate": 0.0001,
+      "loss": 0.8682,
+      "step": 2347
+    },
+    {
+      "epoch": 2.0868792356404846,
+      "grad_norm": 0.31360548734664917,
+      "learning_rate": 0.0001,
+      "loss": 0.8854,
+      "step": 2348
+    },
+    {
+      "epoch": 2.087768025774914,
+      "grad_norm": 0.3268912434577942,
+      "learning_rate": 0.0001,
+      "loss": 0.867,
+      "step": 2349
+    },
+    {
+      "epoch": 2.0886568159093435,
+      "grad_norm": 0.33978188037872314,
+      "learning_rate": 0.0001,
+      "loss": 0.9178,
+      "step": 2350
+    },
+    {
+      "epoch": 2.0895456060437727,
+      "grad_norm": 0.3470538258552551,
+      "learning_rate": 0.0001,
+      "loss": 0.9443,
+      "step": 2351
+    },
+    {
+      "epoch": 2.0904343961782024,
+      "grad_norm": 0.35913875699043274,
+      "learning_rate": 0.0001,
+      "loss": 0.943,
+      "step": 2352
+    },
+    {
+      "epoch": 2.091323186312632,
+      "grad_norm": 0.37776345014572144,
+      "learning_rate": 0.0001,
+      "loss": 0.8927,
+      "step": 2353
+    },
+    {
+      "epoch": 2.0922119764470613,
+      "grad_norm": 0.39539778232574463,
+      "learning_rate": 0.0001,
+      "loss": 0.8981,
+      "step": 2354
+    },
+    {
+      "epoch": 2.093100766581491,
+      "grad_norm": 0.352658748626709,
+      "learning_rate": 0.0001,
+      "loss": 0.8996,
+      "step": 2355
+    },
+    {
+      "epoch": 2.0939895567159206,
+      "grad_norm": 0.31923985481262207,
+      "learning_rate": 0.0001,
+      "loss": 0.9008,
+      "step": 2356
+    },
+    {
+      "epoch": 2.09487834685035,
+      "grad_norm": 0.36446240544319153,
+      "learning_rate": 0.0001,
+      "loss": 0.9393,
+      "step": 2357
+    },
+    {
+      "epoch": 2.0957671369847795,
+      "grad_norm": 0.3254162073135376,
+      "learning_rate": 0.0001,
+      "loss": 0.7878,
+      "step": 2358
+    },
+    {
+      "epoch": 2.096655927119209,
+      "grad_norm": 0.3216734826564789,
+      "learning_rate": 0.0001,
+      "loss": 0.8606,
+      "step": 2359
+    },
+    {
+      "epoch": 2.0975447172536383,
+      "grad_norm": 0.3242090344429016,
+      "learning_rate": 0.0001,
+      "loss": 0.8695,
+      "step": 2360
+    },
+    {
+      "epoch": 2.098433507388068,
+      "grad_norm": 0.29055359959602356,
+      "learning_rate": 0.0001,
+      "loss": 0.8997,
+      "step": 2361
+    },
+    {
+      "epoch": 2.0993222975224977,
+      "grad_norm": 0.3655925691127777,
+      "learning_rate": 0.0001,
+      "loss": 0.9162,
+      "step": 2362
+    },
+    {
+      "epoch": 2.100211087656927,
+      "grad_norm": 0.3490038514137268,
+      "learning_rate": 0.0001,
+      "loss": 0.9202,
+      "step": 2363
+    },
+    {
+      "epoch": 2.1010998777913565,
+      "grad_norm": 0.3455352485179901,
+      "learning_rate": 0.0001,
+      "loss": 0.8759,
+      "step": 2364
+    },
+    {
+      "epoch": 2.101988667925786,
+      "grad_norm": 0.2887704074382782,
+      "learning_rate": 0.0001,
+      "loss": 0.8869,
+      "step": 2365
+    },
+    {
+      "epoch": 2.1028774580602154,
+      "grad_norm": 0.32117751240730286,
+      "learning_rate": 0.0001,
+      "loss": 0.8057,
+      "step": 2366
+    },
+    {
+      "epoch": 2.103766248194645,
+      "grad_norm": 0.33989858627319336,
+      "learning_rate": 0.0001,
+      "loss": 1.0061,
+      "step": 2367
+    },
+    {
+      "epoch": 2.1046550383290747,
+      "grad_norm": 0.3154880106449127,
+      "learning_rate": 0.0001,
+      "loss": 0.8518,
+      "step": 2368
+    },
+    {
+      "epoch": 2.105543828463504,
+      "grad_norm": 0.2925361692905426,
+      "learning_rate": 0.0001,
+      "loss": 0.8935,
+      "step": 2369
+    },
+    {
+      "epoch": 2.1064326185979336,
+      "grad_norm": 0.3639398217201233,
+      "learning_rate": 0.0001,
+      "loss": 0.8115,
+      "step": 2370
+    },
+    {
+      "epoch": 2.1073214087323633,
+      "grad_norm": 0.3522791266441345,
+      "learning_rate": 0.0001,
+      "loss": 0.8857,
+      "step": 2371
+    },
+    {
+      "epoch": 2.1082101988667925,
+      "grad_norm": 0.3359151780605316,
+      "learning_rate": 0.0001,
+      "loss": 0.8931,
+      "step": 2372
+    },
+    {
+      "epoch": 2.109098989001222,
+      "grad_norm": 0.33028051257133484,
+      "learning_rate": 0.0001,
+      "loss": 0.8142,
+      "step": 2373
+    },
+    {
+      "epoch": 2.1099877791356514,
+      "grad_norm": 0.3378293514251709,
+      "learning_rate": 0.0001,
+      "loss": 0.9244,
+      "step": 2374
+    },
+    {
+      "epoch": 2.110876569270081,
+      "grad_norm": 0.36423346400260925,
+      "learning_rate": 0.0001,
+      "loss": 0.8713,
+      "step": 2375
+    },
+    {
+      "epoch": 2.1117653594045107,
+      "grad_norm": 0.32155025005340576,
+      "learning_rate": 0.0001,
+      "loss": 0.8168,
+      "step": 2376
+    },
+    {
+      "epoch": 2.11265414953894,
+      "grad_norm": 0.627085268497467,
+      "learning_rate": 0.0001,
+      "loss": 0.8542,
+      "step": 2377
+    },
+    {
+      "epoch": 2.1135429396733696,
+      "grad_norm": 0.32278314232826233,
+      "learning_rate": 0.0001,
+      "loss": 0.8938,
+      "step": 2378
+    },
+    {
+      "epoch": 2.1144317298077993,
+      "grad_norm": 0.3442274332046509,
+      "learning_rate": 0.0001,
+      "loss": 0.9269,
+      "step": 2379
+    },
+    {
+      "epoch": 2.1153205199422285,
+      "grad_norm": 0.3448584973812103,
+      "learning_rate": 0.0001,
+      "loss": 0.8243,
+      "step": 2380
+    },
+    {
+      "epoch": 2.116209310076658,
+      "grad_norm": 0.3396780490875244,
+      "learning_rate": 0.0001,
+      "loss": 0.8122,
+      "step": 2381
+    },
+    {
+      "epoch": 2.117098100211088,
+      "grad_norm": 0.3797731101512909,
+      "learning_rate": 0.0001,
+      "loss": 0.9614,
+      "step": 2382
+    },
+    {
+      "epoch": 2.117986890345517,
+      "grad_norm": 0.41133296489715576,
+      "learning_rate": 0.0001,
+      "loss": 0.9597,
+      "step": 2383
+    },
+    {
+      "epoch": 2.1188756804799467,
+      "grad_norm": 0.4031218886375427,
+      "learning_rate": 0.0001,
+      "loss": 0.9289,
+      "step": 2384
+    },
+    {
+      "epoch": 2.1197644706143763,
+      "grad_norm": 0.3527853488922119,
+      "learning_rate": 0.0001,
+      "loss": 0.9029,
+      "step": 2385
+    },
+    {
+      "epoch": 2.1206532607488056,
+      "grad_norm": 0.3235529363155365,
+      "learning_rate": 0.0001,
+      "loss": 0.9195,
+      "step": 2386
+    },
+    {
+      "epoch": 2.121542050883235,
+      "grad_norm": 0.31495201587677,
+      "learning_rate": 0.0001,
+      "loss": 0.8524,
+      "step": 2387
+    },
+    {
+      "epoch": 2.122430841017665,
+      "grad_norm": 0.378337562084198,
+      "learning_rate": 0.0001,
+      "loss": 0.8896,
+      "step": 2388
+    },
+    {
+      "epoch": 2.123319631152094,
+      "grad_norm": 0.34486186504364014,
+      "learning_rate": 0.0001,
+      "loss": 0.9178,
+      "step": 2389
+    },
+    {
+      "epoch": 2.1242084212865238,
+      "grad_norm": 0.36222806572914124,
+      "learning_rate": 0.0001,
+      "loss": 0.8327,
+      "step": 2390
+    },
+    {
+      "epoch": 2.1250972114209534,
+      "grad_norm": 0.3327309191226959,
+      "learning_rate": 0.0001,
+      "loss": 0.8055,
+      "step": 2391
+    },
+    {
+      "epoch": 2.1259860015553826,
+      "grad_norm": 0.3839544951915741,
+      "learning_rate": 0.0001,
+      "loss": 0.9453,
+      "step": 2392
+    },
+    {
+      "epoch": 2.1268747916898123,
+      "grad_norm": 0.33836036920547485,
+      "learning_rate": 0.0001,
+      "loss": 0.8824,
+      "step": 2393
+    },
+    {
+      "epoch": 2.127763581824242,
+      "grad_norm": 0.3374810218811035,
+      "learning_rate": 0.0001,
+      "loss": 0.8261,
+      "step": 2394
+    },
+    {
+      "epoch": 2.128652371958671,
+      "grad_norm": 0.3791443109512329,
+      "learning_rate": 0.0001,
+      "loss": 0.9758,
+      "step": 2395
+    },
+    {
+      "epoch": 2.129541162093101,
+      "grad_norm": 0.3412923216819763,
+      "learning_rate": 0.0001,
+      "loss": 0.8655,
+      "step": 2396
+    },
+    {
+      "epoch": 2.13042995222753,
+      "grad_norm": 0.3597002625465393,
+      "learning_rate": 0.0001,
+      "loss": 0.8728,
+      "step": 2397
+    },
+    {
+      "epoch": 2.1313187423619597,
+      "grad_norm": 0.35584625601768494,
+      "learning_rate": 0.0001,
+      "loss": 0.8583,
+      "step": 2398
+    },
+    {
+      "epoch": 2.1322075324963894,
+      "grad_norm": 0.3240017890930176,
+      "learning_rate": 0.0001,
+      "loss": 0.9065,
+      "step": 2399
+    },
+    {
+      "epoch": 2.1330963226308186,
+      "grad_norm": 0.31349146366119385,
+      "learning_rate": 0.0001,
+      "loss": 0.8752,
+      "step": 2400
+    },
+    {
+      "epoch": 2.1339851127652483,
+      "grad_norm": 0.3268539011478424,
+      "learning_rate": 0.0001,
+      "loss": 0.867,
+      "step": 2401
+    },
+    {
+      "epoch": 2.134873902899678,
+      "grad_norm": 1.1880651712417603,
+      "learning_rate": 0.0001,
+      "loss": 0.9108,
+      "step": 2402
+    },
+    {
+      "epoch": 2.135762693034107,
+      "grad_norm": 0.32068613171577454,
+      "learning_rate": 0.0001,
+      "loss": 0.865,
+      "step": 2403
+    },
+    {
+      "epoch": 2.136651483168537,
+      "grad_norm": 0.3671189248561859,
+      "learning_rate": 0.0001,
+      "loss": 0.922,
+      "step": 2404
+    },
+    {
+      "epoch": 2.1375402733029665,
+      "grad_norm": 0.3628632724285126,
+      "learning_rate": 0.0001,
+      "loss": 0.8714,
+      "step": 2405
+    },
+    {
+      "epoch": 2.1384290634373957,
+      "grad_norm": 0.38792744278907776,
+      "learning_rate": 0.0001,
+      "loss": 0.9345,
+      "step": 2406
+    },
+    {
+      "epoch": 2.1393178535718254,
+      "grad_norm": 0.38119903206825256,
+      "learning_rate": 0.0001,
+      "loss": 0.8692,
+      "step": 2407
+    },
+    {
+      "epoch": 2.140206643706255,
+      "grad_norm": 0.431945264339447,
+      "learning_rate": 0.0001,
+      "loss": 0.9405,
+      "step": 2408
+    },
+    {
+      "epoch": 2.1410954338406842,
+      "grad_norm": 0.3509438633918762,
+      "learning_rate": 0.0001,
+      "loss": 0.9128,
+      "step": 2409
+    },
+    {
+      "epoch": 2.141984223975114,
+      "grad_norm": 0.35793623328208923,
+      "learning_rate": 0.0001,
+      "loss": 0.8824,
+      "step": 2410
+    },
+    {
+      "epoch": 2.1428730141095436,
+      "grad_norm": 0.6159213185310364,
+      "learning_rate": 0.0001,
+      "loss": 0.8774,
+      "step": 2411
+    },
+    {
+      "epoch": 2.143761804243973,
+      "grad_norm": 0.36963677406311035,
+      "learning_rate": 0.0001,
+      "loss": 0.8521,
+      "step": 2412
+    },
+    {
+      "epoch": 2.1446505943784024,
+      "grad_norm": 0.36160334944725037,
+      "learning_rate": 0.0001,
+      "loss": 0.8255,
+      "step": 2413
+    },
+    {
+      "epoch": 2.145539384512832,
+      "grad_norm": 0.35096341371536255,
+      "learning_rate": 0.0001,
+      "loss": 0.955,
+      "step": 2414
+    },
+    {
+      "epoch": 2.1464281746472613,
+      "grad_norm": 0.5263632535934448,
+      "learning_rate": 0.0001,
+      "loss": 0.9105,
+      "step": 2415
+    },
+    {
+      "epoch": 2.147316964781691,
+      "grad_norm": 0.37115469574928284,
+      "learning_rate": 0.0001,
+      "loss": 0.8369,
+      "step": 2416
+    },
+    {
+      "epoch": 2.1482057549161206,
+      "grad_norm": 0.3329596519470215,
+      "learning_rate": 0.0001,
+      "loss": 0.8283,
+      "step": 2417
+    },
+    {
+      "epoch": 2.14909454505055,
+      "grad_norm": 0.5003162026405334,
+      "learning_rate": 0.0001,
+      "loss": 0.9056,
+      "step": 2418
+    },
+    {
+      "epoch": 2.1499833351849795,
+      "grad_norm": 0.3137914538383484,
+      "learning_rate": 0.0001,
+      "loss": 0.8761,
+      "step": 2419
+    },
+    {
+      "epoch": 2.1508721253194087,
+      "grad_norm": 0.35285452008247375,
+      "learning_rate": 0.0001,
+      "loss": 0.8635,
+      "step": 2420
+    },
+    {
+      "epoch": 2.1517609154538384,
+      "grad_norm": 0.35029417276382446,
+      "learning_rate": 0.0001,
+      "loss": 0.8181,
+      "step": 2421
+    },
+    {
+      "epoch": 2.152649705588268,
+      "grad_norm": 0.36478570103645325,
+      "learning_rate": 0.0001,
+      "loss": 0.8536,
+      "step": 2422
+    },
+    {
+      "epoch": 2.1535384957226973,
+      "grad_norm": 0.43607574701309204,
+      "learning_rate": 0.0001,
+      "loss": 0.8965,
+      "step": 2423
+    },
+    {
+      "epoch": 2.154427285857127,
+      "grad_norm": 0.36862248182296753,
+      "learning_rate": 0.0001,
+      "loss": 0.8031,
+      "step": 2424
+    },
+    {
+      "epoch": 2.1553160759915566,
+      "grad_norm": 0.49188151955604553,
+      "learning_rate": 0.0001,
+      "loss": 0.9334,
+      "step": 2425
+    },
+    {
+      "epoch": 2.156204866125986,
+      "grad_norm": 0.7494956254959106,
+      "learning_rate": 0.0001,
+      "loss": 0.9039,
+      "step": 2426
+    },
+    {
+      "epoch": 2.1570936562604155,
+      "grad_norm": 0.9286481738090515,
+      "learning_rate": 0.0001,
+      "loss": 0.9374,
+      "step": 2427
+    },
+    {
+      "epoch": 2.157982446394845,
+      "grad_norm": 0.4706733822822571,
+      "learning_rate": 0.0001,
+      "loss": 0.9537,
+      "step": 2428
+    },
+    {
+      "epoch": 2.1588712365292744,
+      "grad_norm": 0.4074293076992035,
+      "learning_rate": 0.0001,
+      "loss": 0.8819,
+      "step": 2429
+    },
+    {
+      "epoch": 2.159760026663704,
+      "grad_norm": 0.3723183870315552,
+      "learning_rate": 0.0001,
+      "loss": 0.8569,
+      "step": 2430
+    },
+    {
+      "epoch": 2.1606488167981337,
+      "grad_norm": 0.5987865924835205,
+      "learning_rate": 0.0001,
+      "loss": 0.8495,
+      "step": 2431
+    },
+    {
+      "epoch": 2.161537606932563,
+      "grad_norm": 0.3815048038959503,
+      "learning_rate": 0.0001,
+      "loss": 0.8312,
+      "step": 2432
+    },
+    {
+      "epoch": 2.1624263970669926,
+      "grad_norm": 0.33688196539878845,
+      "learning_rate": 0.0001,
+      "loss": 0.8509,
+      "step": 2433
+    },
+    {
+      "epoch": 2.1633151872014222,
+      "grad_norm": 0.3331957161426544,
+      "learning_rate": 0.0001,
+      "loss": 0.8446,
+      "step": 2434
+    },
+    {
+      "epoch": 2.1642039773358515,
+      "grad_norm": 0.8897263407707214,
+      "learning_rate": 0.0001,
+      "loss": 0.7964,
+      "step": 2435
+    },
+    {
+      "epoch": 2.165092767470281,
+      "grad_norm": 0.3500574231147766,
+      "learning_rate": 0.0001,
+      "loss": 0.8289,
+      "step": 2436
+    },
+    {
+      "epoch": 2.165981557604711,
+      "grad_norm": 0.45759961009025574,
+      "learning_rate": 0.0001,
+      "loss": 0.8902,
+      "step": 2437
+    },
+    {
+      "epoch": 2.16687034773914,
+      "grad_norm": 0.34242385625839233,
+      "learning_rate": 0.0001,
+      "loss": 0.8415,
+      "step": 2438
+    },
+    {
+      "epoch": 2.1677591378735697,
+      "grad_norm": 0.4210834205150604,
+      "learning_rate": 0.0001,
+      "loss": 0.9238,
+      "step": 2439
+    },
+    {
+      "epoch": 2.1686479280079993,
+      "grad_norm": 0.6454192399978638,
+      "learning_rate": 0.0001,
+      "loss": 0.8155,
+      "step": 2440
+    },
+    {
+      "epoch": 2.1695367181424285,
+      "grad_norm": 0.34665006399154663,
+      "learning_rate": 0.0001,
+      "loss": 0.8507,
+      "step": 2441
+    },
+    {
+      "epoch": 2.170425508276858,
+      "grad_norm": 0.3612930178642273,
+      "learning_rate": 0.0001,
+      "loss": 0.8899,
+      "step": 2442
+    },
+    {
+      "epoch": 2.1713142984112874,
+      "grad_norm": 0.36009481549263,
+      "learning_rate": 0.0001,
+      "loss": 0.8985,
+      "step": 2443
+    },
+    {
+      "epoch": 2.172203088545717,
+      "grad_norm": 0.392411470413208,
+      "learning_rate": 0.0001,
+      "loss": 0.947,
+      "step": 2444
+    },
+    {
+      "epoch": 2.1730918786801467,
+      "grad_norm": 0.346246600151062,
+      "learning_rate": 0.0001,
+      "loss": 0.8688,
+      "step": 2445
+    },
+    {
+      "epoch": 2.173980668814576,
+      "grad_norm": 0.3527586758136749,
+      "learning_rate": 0.0001,
+      "loss": 0.8966,
+      "step": 2446
+    },
+    {
+      "epoch": 2.1748694589490056,
+      "grad_norm": 0.3666391968727112,
+      "learning_rate": 0.0001,
+      "loss": 0.945,
+      "step": 2447
+    },
+    {
+      "epoch": 2.1757582490834353,
+      "grad_norm": 0.36609384417533875,
+      "learning_rate": 0.0001,
+      "loss": 0.844,
+      "step": 2448
+    },
+    {
+      "epoch": 2.1766470392178645,
+      "grad_norm": 0.33666878938674927,
+      "learning_rate": 0.0001,
+      "loss": 0.7822,
+      "step": 2449
+    },
+    {
+      "epoch": 2.177535829352294,
+      "grad_norm": 0.41050276160240173,
+      "learning_rate": 0.0001,
+      "loss": 0.9306,
+      "step": 2450
+    },
+    {
+      "epoch": 2.178424619486724,
+      "grad_norm": 0.33751270174980164,
+      "learning_rate": 0.0001,
+      "loss": 0.8642,
+      "step": 2451
+    },
+    {
+      "epoch": 2.179313409621153,
+      "grad_norm": 0.32639411091804504,
+      "learning_rate": 0.0001,
+      "loss": 0.8057,
+      "step": 2452
+    },
+    {
+      "epoch": 2.1802021997555827,
+      "grad_norm": 0.34592291712760925,
+      "learning_rate": 0.0001,
+      "loss": 0.9396,
+      "step": 2453
+    },
+    {
+      "epoch": 2.1810909898900124,
+      "grad_norm": 0.32606473565101624,
+      "learning_rate": 0.0001,
+      "loss": 0.8235,
+      "step": 2454
+    },
+    {
+      "epoch": 2.1819797800244416,
+      "grad_norm": 0.33683574199676514,
+      "learning_rate": 0.0001,
+      "loss": 0.8805,
+      "step": 2455
+    },
+    {
+      "epoch": 2.1828685701588713,
+      "grad_norm": 0.36452654004096985,
+      "learning_rate": 0.0001,
+      "loss": 0.8843,
+      "step": 2456
+    },
+    {
+      "epoch": 2.183757360293301,
+      "grad_norm": 0.3444773256778717,
+      "learning_rate": 0.0001,
+      "loss": 0.9527,
+      "step": 2457
+    },
+    {
+      "epoch": 2.18464615042773,
+      "grad_norm": 0.32133087515830994,
+      "learning_rate": 0.0001,
+      "loss": 0.8259,
+      "step": 2458
+    },
+    {
+      "epoch": 2.18553494056216,
+      "grad_norm": 0.3569784462451935,
+      "learning_rate": 0.0001,
+      "loss": 0.9475,
+      "step": 2459
+    },
+    {
+      "epoch": 2.1864237306965895,
+      "grad_norm": 0.3464709222316742,
+      "learning_rate": 0.0001,
+      "loss": 0.8372,
+      "step": 2460
+    },
+    {
+      "epoch": 2.1873125208310187,
+      "grad_norm": 0.30103108286857605,
+      "learning_rate": 0.0001,
+      "loss": 0.843,
+      "step": 2461
+    },
+    {
+      "epoch": 2.1882013109654483,
+      "grad_norm": 0.31629863381385803,
+      "learning_rate": 0.0001,
+      "loss": 0.8342,
+      "step": 2462
+    },
+    {
+      "epoch": 2.189090101099878,
+      "grad_norm": 0.31571292877197266,
+      "learning_rate": 0.0001,
+      "loss": 0.8531,
+      "step": 2463
+    },
+    {
+      "epoch": 2.1899788912343072,
+      "grad_norm": 0.36305055022239685,
+      "learning_rate": 0.0001,
+      "loss": 0.8899,
+      "step": 2464
+    },
+    {
+      "epoch": 2.190867681368737,
+      "grad_norm": 0.31181925535202026,
+      "learning_rate": 0.0001,
+      "loss": 0.8869,
+      "step": 2465
+    },
+    {
+      "epoch": 2.191756471503166,
+      "grad_norm": 0.33491456508636475,
+      "learning_rate": 0.0001,
+      "loss": 0.911,
+      "step": 2466
+    },
+    {
+      "epoch": 2.1926452616375958,
+      "grad_norm": 0.3482362926006317,
+      "learning_rate": 0.0001,
+      "loss": 0.8501,
+      "step": 2467
+    },
+    {
+      "epoch": 2.1935340517720254,
+      "grad_norm": 0.3489706516265869,
+      "learning_rate": 0.0001,
+      "loss": 0.854,
+      "step": 2468
+    },
+    {
+      "epoch": 2.1944228419064546,
+      "grad_norm": 0.33438587188720703,
+      "learning_rate": 0.0001,
+      "loss": 0.8383,
+      "step": 2469
+    },
+    {
+      "epoch": 2.1953116320408843,
+      "grad_norm": 0.3124372363090515,
+      "learning_rate": 0.0001,
+      "loss": 0.8203,
+      "step": 2470
+    },
+    {
+      "epoch": 2.196200422175314,
+      "grad_norm": 0.349856972694397,
+      "learning_rate": 0.0001,
+      "loss": 0.9404,
+      "step": 2471
+    },
+    {
+      "epoch": 2.197089212309743,
+      "grad_norm": 0.3826901316642761,
+      "learning_rate": 0.0001,
+      "loss": 0.9455,
+      "step": 2472
+    },
+    {
+      "epoch": 2.197978002444173,
+      "grad_norm": 0.34957224130630493,
+      "learning_rate": 0.0001,
+      "loss": 0.938,
+      "step": 2473
+    },
+    {
+      "epoch": 2.1988667925786025,
+      "grad_norm": 0.28016459941864014,
+      "learning_rate": 0.0001,
+      "loss": 0.9001,
+      "step": 2474
+    },
+    {
+      "epoch": 2.1997555827130317,
+      "grad_norm": 0.39419126510620117,
+      "learning_rate": 0.0001,
+      "loss": 0.9849,
+      "step": 2475
+    },
+    {
+      "epoch": 2.2006443728474614,
+      "grad_norm": 0.3508966267108917,
+      "learning_rate": 0.0001,
+      "loss": 0.8562,
+      "step": 2476
+    },
+    {
+      "epoch": 2.201533162981891,
+      "grad_norm": 0.32400640845298767,
+      "learning_rate": 0.0001,
+      "loss": 0.8503,
+      "step": 2477
+    },
+    {
+      "epoch": 2.2024219531163203,
+      "grad_norm": 0.31029975414276123,
+      "learning_rate": 0.0001,
+      "loss": 0.9593,
+      "step": 2478
+    },
+    {
+      "epoch": 2.20331074325075,
+      "grad_norm": 0.34035494923591614,
+      "learning_rate": 0.0001,
+      "loss": 1.0073,
+      "step": 2479
+    },
+    {
+      "epoch": 2.2041995333851796,
+      "grad_norm": 0.34316977858543396,
+      "learning_rate": 0.0001,
+      "loss": 0.8212,
+      "step": 2480
+    },
+    {
+      "epoch": 2.205088323519609,
+      "grad_norm": 0.3366788625717163,
+      "learning_rate": 0.0001,
+      "loss": 0.8808,
+      "step": 2481
+    },
+    {
+      "epoch": 2.2059771136540385,
+      "grad_norm": 0.36008214950561523,
+      "learning_rate": 0.0001,
+      "loss": 0.8823,
+      "step": 2482
+    },
+    {
+      "epoch": 2.206865903788468,
+      "grad_norm": 0.3107230067253113,
+      "learning_rate": 0.0001,
+      "loss": 0.9265,
+      "step": 2483
+    },
+    {
+      "epoch": 2.2077546939228974,
+      "grad_norm": 0.328205406665802,
+      "learning_rate": 0.0001,
+      "loss": 0.8796,
+      "step": 2484
+    },
+    {
+      "epoch": 2.208643484057327,
+      "grad_norm": 0.33115193247795105,
+      "learning_rate": 0.0001,
+      "loss": 0.9549,
+      "step": 2485
+    },
+    {
+      "epoch": 2.2095322741917567,
+      "grad_norm": 0.31935736536979675,
+      "learning_rate": 0.0001,
+      "loss": 0.9173,
+      "step": 2486
+    },
+    {
+      "epoch": 2.210421064326186,
+      "grad_norm": 0.3087455630302429,
+      "learning_rate": 0.0001,
+      "loss": 0.8084,
+      "step": 2487
+    },
+    {
+      "epoch": 2.2113098544606156,
+      "grad_norm": 0.38220590353012085,
+      "learning_rate": 0.0001,
+      "loss": 1.0343,
+      "step": 2488
+    },
+    {
+      "epoch": 2.212198644595045,
+      "grad_norm": 0.3595533072948456,
+      "learning_rate": 0.0001,
+      "loss": 0.9296,
+      "step": 2489
+    },
+    {
+      "epoch": 2.2130874347294744,
+      "grad_norm": 0.35236823558807373,
+      "learning_rate": 0.0001,
+      "loss": 0.9115,
+      "step": 2490
+    },
+    {
+      "epoch": 2.213976224863904,
+      "grad_norm": 0.2661668360233307,
+      "learning_rate": 0.0001,
+      "loss": 0.829,
+      "step": 2491
+    },
+    {
+      "epoch": 2.2148650149983333,
+      "grad_norm": 0.3272991478443146,
+      "learning_rate": 0.0001,
+      "loss": 0.8619,
+      "step": 2492
+    },
+    {
+      "epoch": 2.215753805132763,
+      "grad_norm": 0.3380112648010254,
+      "learning_rate": 0.0001,
+      "loss": 0.8567,
+      "step": 2493
+    },
+    {
+      "epoch": 2.2166425952671927,
+      "grad_norm": 0.36716368794441223,
+      "learning_rate": 0.0001,
+      "loss": 0.9482,
+      "step": 2494
+    },
+    {
+      "epoch": 2.217531385401622,
+      "grad_norm": 0.32290229201316833,
+      "learning_rate": 0.0001,
+      "loss": 0.8206,
+      "step": 2495
+    },
+    {
+      "epoch": 2.2184201755360515,
+      "grad_norm": 0.3434554934501648,
+      "learning_rate": 0.0001,
+      "loss": 0.829,
+      "step": 2496
+    },
+    {
+      "epoch": 2.219308965670481,
+      "grad_norm": 0.3272440433502197,
+      "learning_rate": 0.0001,
+      "loss": 0.9398,
+      "step": 2497
+    },
+    {
+      "epoch": 2.2201977558049104,
+      "grad_norm": 0.35004308819770813,
+      "learning_rate": 0.0001,
+      "loss": 0.9388,
+      "step": 2498
+    },
+    {
+      "epoch": 2.22108654593934,
+      "grad_norm": 0.38155871629714966,
+      "learning_rate": 0.0001,
+      "loss": 0.8243,
+      "step": 2499
+    },
+    {
+      "epoch": 2.2219753360737697,
+      "grad_norm": 0.36860787868499756,
+      "learning_rate": 0.0001,
+      "loss": 0.9036,
+      "step": 2500
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 1.3367420220730245e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}