diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20419 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 2500,
+  "global_step": 2911,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003435983335480823,
+      "grad_norm": 0.2608718276023865,
+      "learning_rate": 0.0,
+      "loss": 11.9378,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006871966670961646,
+      "grad_norm": 0.27573105692863464,
+      "learning_rate": 1.7123287671232875e-06,
+      "loss": 11.9422,
+      "step": 2
+    },
+    {
+      "epoch": 0.001030795000644247,
+      "grad_norm": 0.22586911916732788,
+      "learning_rate": 3.424657534246575e-06,
+      "loss": 11.9418,
+      "step": 3
+    },
+    {
+      "epoch": 0.0013743933341923292,
+      "grad_norm": 0.23945586383342743,
+      "learning_rate": 5.136986301369863e-06,
+      "loss": 11.9397,
+      "step": 4
+    },
+    {
+      "epoch": 0.0017179916677404115,
+      "grad_norm": 0.2580551207065582,
+      "learning_rate": 6.84931506849315e-06,
+      "loss": 11.9401,
+      "step": 5
+    },
+    {
+      "epoch": 0.002061590001288494,
+      "grad_norm": 0.24455508589744568,
+      "learning_rate": 8.561643835616438e-06,
+      "loss": 11.9391,
+      "step": 6
+    },
+    {
+      "epoch": 0.002405188334836576,
+      "grad_norm": 0.26748892664909363,
+      "learning_rate": 1.0273972602739726e-05,
+      "loss": 11.9398,
+      "step": 7
+    },
+    {
+      "epoch": 0.0027487866683846584,
+      "grad_norm": 0.26134631037712097,
+      "learning_rate": 1.1986301369863013e-05,
+      "loss": 11.9397,
+      "step": 8
+    },
+    {
+      "epoch": 0.0030923850019327404,
+      "grad_norm": 0.23617519438266754,
+      "learning_rate": 1.36986301369863e-05,
+      "loss": 11.9416,
+      "step": 9
+    },
+    {
+      "epoch": 0.003435983335480823,
+      "grad_norm": 0.24465899169445038,
+      "learning_rate": 1.541095890410959e-05,
+      "loss": 11.938,
+      "step": 10
+    },
+    {
+      "epoch": 0.0037795816690289054,
+      "grad_norm": 0.23813503980636597,
+      "learning_rate": 1.7123287671232875e-05,
+      "loss": 11.9394,
+      "step": 11
+    },
+    {
+      "epoch": 0.004123180002576988,
+      "grad_norm": 0.25183629989624023,
+      "learning_rate": 1.8835616438356162e-05,
+      "loss": 11.9392,
+      "step": 12
+    },
+    {
+      "epoch": 0.00446677833612507,
+      "grad_norm": 0.22182029485702515,
+      "learning_rate": 2.0547945205479453e-05,
+      "loss": 11.9393,
+      "step": 13
+    },
+    {
+      "epoch": 0.004810376669673152,
+      "grad_norm": 0.22089728713035583,
+      "learning_rate": 2.226027397260274e-05,
+      "loss": 11.9374,
+      "step": 14
+    },
+    {
+      "epoch": 0.005153975003221235,
+      "grad_norm": 0.22943605482578278,
+      "learning_rate": 2.3972602739726026e-05,
+      "loss": 11.9379,
+      "step": 15
+    },
+    {
+      "epoch": 0.005497573336769317,
+      "grad_norm": 0.25133973360061646,
+      "learning_rate": 2.5684931506849313e-05,
+      "loss": 11.936,
+      "step": 16
+    },
+    {
+      "epoch": 0.005841171670317399,
+      "grad_norm": 0.2602233290672302,
+      "learning_rate": 2.73972602739726e-05,
+      "loss": 11.9316,
+      "step": 17
+    },
+    {
+      "epoch": 0.006184770003865481,
+      "grad_norm": 0.23244044184684753,
+      "learning_rate": 2.910958904109589e-05,
+      "loss": 11.9329,
+      "step": 18
+    },
+    {
+      "epoch": 0.006528368337413564,
+      "grad_norm": 0.2674640417098999,
+      "learning_rate": 3.082191780821918e-05,
+      "loss": 11.9311,
+      "step": 19
+    },
+    {
+      "epoch": 0.006871966670961646,
+      "grad_norm": 0.24958398938179016,
+      "learning_rate": 3.2534246575342464e-05,
+      "loss": 11.9319,
+      "step": 20
+    },
+    {
+      "epoch": 0.007215565004509728,
+      "grad_norm": 0.28795570135116577,
+      "learning_rate": 3.424657534246575e-05,
+      "loss": 11.9309,
+      "step": 21
+    },
+    {
+      "epoch": 0.007559163338057811,
+      "grad_norm": 0.27296486496925354,
+      "learning_rate": 3.595890410958904e-05,
+      "loss": 11.9292,
+      "step": 22
+    },
+    {
+      "epoch": 0.007902761671605892,
+      "grad_norm": 0.29060548543930054,
+      "learning_rate": 3.7671232876712325e-05,
+      "loss": 11.93,
+      "step": 23
+    },
+    {
+      "epoch": 0.008246360005153976,
+      "grad_norm": 0.2740913927555084,
+      "learning_rate": 3.938356164383562e-05,
+      "loss": 11.9286,
+      "step": 24
+    },
+    {
+      "epoch": 0.008589958338702058,
+      "grad_norm": 0.29341545701026917,
+      "learning_rate": 4.1095890410958905e-05,
+      "loss": 11.9271,
+      "step": 25
+    },
+    {
+      "epoch": 0.00893355667225014,
+      "grad_norm": 0.323218435049057,
+      "learning_rate": 4.280821917808219e-05,
+      "loss": 11.9232,
+      "step": 26
+    },
+    {
+      "epoch": 0.009277155005798222,
+      "grad_norm": 0.3378582000732422,
+      "learning_rate": 4.452054794520548e-05,
+      "loss": 11.9198,
+      "step": 27
+    },
+    {
+      "epoch": 0.009620753339346304,
+      "grad_norm": 0.36083006858825684,
+      "learning_rate": 4.6232876712328766e-05,
+      "loss": 11.9189,
+      "step": 28
+    },
+    {
+      "epoch": 0.009964351672894386,
+      "grad_norm": 0.3552938401699066,
+      "learning_rate": 4.794520547945205e-05,
+      "loss": 11.9171,
+      "step": 29
+    },
+    {
+      "epoch": 0.01030795000644247,
+      "grad_norm": 0.3807588219642639,
+      "learning_rate": 4.965753424657534e-05,
+      "loss": 11.9156,
+      "step": 30
+    },
+    {
+      "epoch": 0.010651548339990552,
+      "grad_norm": 0.3817954957485199,
+      "learning_rate": 5.1369863013698626e-05,
+      "loss": 11.9133,
+      "step": 31
+    },
+    {
+      "epoch": 0.010995146673538634,
+      "grad_norm": 0.42127540707588196,
+      "learning_rate": 5.308219178082191e-05,
+      "loss": 11.9084,
+      "step": 32
+    },
+    {
+      "epoch": 0.011338745007086716,
+      "grad_norm": 0.42336493730545044,
+      "learning_rate": 5.47945205479452e-05,
+      "loss": 11.9081,
+      "step": 33
+    },
+    {
+      "epoch": 0.011682343340634798,
+      "grad_norm": 0.4692400097846985,
+      "learning_rate": 5.6506849315068494e-05,
+      "loss": 11.8979,
+      "step": 34
+    },
+    {
+      "epoch": 0.01202594167418288,
+      "grad_norm": 0.45885521173477173,
+      "learning_rate": 5.821917808219178e-05,
+      "loss": 11.898,
+      "step": 35
+    },
+    {
+      "epoch": 0.012369540007730962,
+      "grad_norm": 0.48935025930404663,
+      "learning_rate": 5.993150684931507e-05,
+      "loss": 11.8932,
+      "step": 36
+    },
+    {
+      "epoch": 0.012713138341279046,
+      "grad_norm": 0.5434125065803528,
+      "learning_rate": 6.164383561643835e-05,
+      "loss": 11.8905,
+      "step": 37
+    },
+    {
+      "epoch": 0.013056736674827128,
+      "grad_norm": 0.5894971489906311,
+      "learning_rate": 6.335616438356165e-05,
+      "loss": 11.8791,
+      "step": 38
+    },
+    {
+      "epoch": 0.01340033500837521,
+      "grad_norm": 0.5886083841323853,
+      "learning_rate": 6.506849315068493e-05,
+      "loss": 11.879,
+      "step": 39
+    },
+    {
+      "epoch": 0.013743933341923292,
+      "grad_norm": 0.5979152321815491,
+      "learning_rate": 6.678082191780822e-05,
+      "loss": 11.8734,
+      "step": 40
+    },
+    {
+      "epoch": 0.014087531675471374,
+      "grad_norm": 0.6167778372764587,
+      "learning_rate": 6.84931506849315e-05,
+      "loss": 11.8688,
+      "step": 41
+    },
+    {
+      "epoch": 0.014431130009019456,
+      "grad_norm": 0.6409335732460022,
+      "learning_rate": 7.02054794520548e-05,
+      "loss": 11.858,
+      "step": 42
+    },
+    {
+      "epoch": 0.014774728342567538,
+      "grad_norm": 0.6955194473266602,
+      "learning_rate": 7.191780821917808e-05,
+      "loss": 11.8506,
+      "step": 43
+    },
+    {
+      "epoch": 0.015118326676115621,
+      "grad_norm": 0.7092486619949341,
+      "learning_rate": 7.363013698630137e-05,
+      "loss": 11.8318,
+      "step": 44
+    },
+    {
+      "epoch": 0.015461925009663703,
+      "grad_norm": 0.6970604658126831,
+      "learning_rate": 7.534246575342465e-05,
+      "loss": 11.833,
+      "step": 45
+    },
+    {
+      "epoch": 0.015805523343211784,
+      "grad_norm": 0.7087641358375549,
+      "learning_rate": 7.705479452054794e-05,
+      "loss": 11.8274,
+      "step": 46
+    },
+    {
+      "epoch": 0.01614912167675987,
+      "grad_norm": 0.7028504014015198,
+      "learning_rate": 7.876712328767124e-05,
+      "loss": 11.8209,
+      "step": 47
+    },
+    {
+      "epoch": 0.01649272001030795,
+      "grad_norm": 0.6933362483978271,
+      "learning_rate": 8.047945205479453e-05,
+      "loss": 11.8089,
+      "step": 48
+    },
+    {
+      "epoch": 0.016836318343856033,
+      "grad_norm": 0.7153650522232056,
+      "learning_rate": 8.219178082191781e-05,
+      "loss": 11.7998,
+      "step": 49
+    },
+    {
+      "epoch": 0.017179916677404115,
+      "grad_norm": 0.7396632432937622,
+      "learning_rate": 8.39041095890411e-05,
+      "loss": 11.7903,
+      "step": 50
+    },
+    {
+      "epoch": 0.017523515010952197,
+      "grad_norm": 0.7029207348823547,
+      "learning_rate": 8.561643835616438e-05,
+      "loss": 11.7541,
+      "step": 51
+    },
+    {
+      "epoch": 0.01786711334450028,
+      "grad_norm": 0.6831297278404236,
+      "learning_rate": 8.732876712328768e-05,
+      "loss": 11.7451,
+      "step": 52
+    },
+    {
+      "epoch": 0.01821071167804836,
+      "grad_norm": 0.6773477792739868,
+      "learning_rate": 8.904109589041096e-05,
+      "loss": 11.7351,
+      "step": 53
+    },
+    {
+      "epoch": 0.018554310011596443,
+      "grad_norm": 0.6540944576263428,
+      "learning_rate": 9.075342465753425e-05,
+      "loss": 11.7275,
+      "step": 54
+    },
+    {
+      "epoch": 0.018897908345144526,
+      "grad_norm": 0.6842557191848755,
+      "learning_rate": 9.246575342465753e-05,
+      "loss": 11.7084,
+      "step": 55
+    },
+    {
+      "epoch": 0.019241506678692608,
+      "grad_norm": 0.6622518301010132,
+      "learning_rate": 9.417808219178083e-05,
+      "loss": 11.7025,
+      "step": 56
+    },
+    {
+      "epoch": 0.01958510501224069,
+      "grad_norm": 0.6573502421379089,
+      "learning_rate": 9.58904109589041e-05,
+      "loss": 11.6883,
+      "step": 57
+    },
+    {
+      "epoch": 0.01992870334578877,
+      "grad_norm": 0.6725475788116455,
+      "learning_rate": 9.76027397260274e-05,
+      "loss": 11.6727,
+      "step": 58
+    },
+    {
+      "epoch": 0.020272301679336854,
+      "grad_norm": 0.6889570355415344,
+      "learning_rate": 9.931506849315068e-05,
+      "loss": 11.6625,
+      "step": 59
+    },
+    {
+      "epoch": 0.02061590001288494,
+      "grad_norm": 0.672089695930481,
+      "learning_rate": 0.00010102739726027397,
+      "loss": 11.6495,
+      "step": 60
+    },
+    {
+      "epoch": 0.02095949834643302,
+      "grad_norm": 0.6492573618888855,
+      "learning_rate": 0.00010273972602739725,
+      "loss": 11.6344,
+      "step": 61
+    },
+    {
+      "epoch": 0.021303096679981103,
+      "grad_norm": 0.6624525189399719,
+      "learning_rate": 0.00010445205479452055,
+      "loss": 11.6246,
+      "step": 62
+    },
+    {
+      "epoch": 0.021646695013529185,
+      "grad_norm": 0.640524685382843,
+      "learning_rate": 0.00010616438356164383,
+      "loss": 11.6162,
+      "step": 63
+    },
+    {
+      "epoch": 0.021990293347077267,
+      "grad_norm": 0.6240212917327881,
+      "learning_rate": 0.00010787671232876712,
+      "loss": 11.601,
+      "step": 64
+    },
+    {
+      "epoch": 0.02233389168062535,
+      "grad_norm": 0.6419646739959717,
+      "learning_rate": 0.0001095890410958904,
+      "loss": 11.5904,
+      "step": 65
+    },
+    {
+      "epoch": 0.02267749001417343,
+      "grad_norm": 0.6217727065086365,
+      "learning_rate": 0.00011130136986301371,
+      "loss": 11.5754,
+      "step": 66
+    },
+    {
+      "epoch": 0.023021088347721513,
+      "grad_norm": 0.6366190314292908,
+      "learning_rate": 0.00011301369863013699,
+      "loss": 11.5616,
+      "step": 67
+    },
+    {
+      "epoch": 0.023364686681269595,
+      "grad_norm": 0.6308091282844543,
+      "learning_rate": 0.00011472602739726028,
+      "loss": 11.5476,
+      "step": 68
+    },
+    {
+      "epoch": 0.023708285014817677,
+      "grad_norm": 0.6248790621757507,
+      "learning_rate": 0.00011643835616438356,
+      "loss": 11.5356,
+      "step": 69
+    },
+    {
+      "epoch": 0.02405188334836576,
+      "grad_norm": 0.6040776968002319,
+      "learning_rate": 0.00011815068493150686,
+      "loss": 11.5316,
+      "step": 70
+    },
+    {
+      "epoch": 0.02439548168191384,
+      "grad_norm": 0.6206280589103699,
+      "learning_rate": 0.00011986301369863014,
+      "loss": 11.5134,
+      "step": 71
+    },
+    {
+      "epoch": 0.024739080015461924,
+      "grad_norm": 0.600658118724823,
+      "learning_rate": 0.00012157534246575343,
+      "loss": 11.5024,
+      "step": 72
+    },
+    {
+      "epoch": 0.02508267834901001,
+      "grad_norm": 0.6167108416557312,
+      "learning_rate": 0.0001232876712328767,
+      "loss": 11.4839,
+      "step": 73
+    },
+    {
+      "epoch": 0.02542627668255809,
+      "grad_norm": 0.6266405582427979,
+      "learning_rate": 0.000125,
+      "loss": 11.4643,
+      "step": 74
+    },
+    {
+      "epoch": 0.025769875016106173,
+      "grad_norm": 0.6280407309532166,
+      "learning_rate": 0.0001267123287671233,
+      "loss": 11.4495,
+      "step": 75
+    },
+    {
+      "epoch": 0.026113473349654255,
+      "grad_norm": 0.6044653654098511,
+      "learning_rate": 0.0001284246575342466,
+      "loss": 11.4406,
+      "step": 76
+    },
+    {
+      "epoch": 0.026457071683202337,
+      "grad_norm": 0.624096691608429,
+      "learning_rate": 0.00013013698630136986,
+      "loss": 11.4213,
+      "step": 77
+    },
+    {
+      "epoch": 0.02680067001675042,
+      "grad_norm": 0.6047064661979675,
+      "learning_rate": 0.00013184931506849315,
+      "loss": 11.4102,
+      "step": 78
+    },
+    {
+      "epoch": 0.0271442683502985,
+      "grad_norm": 0.6190263628959656,
+      "learning_rate": 0.00013356164383561644,
+      "loss": 11.3944,
+      "step": 79
+    },
+    {
+      "epoch": 0.027487866683846583,
+      "grad_norm": 0.6345274448394775,
+      "learning_rate": 0.00013527397260273974,
+      "loss": 11.3798,
+      "step": 80
+    },
+    {
+      "epoch": 0.027831465017394665,
+      "grad_norm": 0.6231955885887146,
+      "learning_rate": 0.000136986301369863,
+      "loss": 11.3585,
+      "step": 81
+    },
+    {
+      "epoch": 0.028175063350942747,
+      "grad_norm": 0.6305320262908936,
+      "learning_rate": 0.0001386986301369863,
+      "loss": 11.3541,
+      "step": 82
+    },
+    {
+      "epoch": 0.02851866168449083,
+      "grad_norm": 0.6369199156761169,
+      "learning_rate": 0.0001404109589041096,
+      "loss": 11.3318,
+      "step": 83
+    },
+    {
+      "epoch": 0.02886226001803891,
+      "grad_norm": 0.6159324049949646,
+      "learning_rate": 0.00014212328767123288,
+      "loss": 11.3296,
+      "step": 84
+    },
+    {
+      "epoch": 0.029205858351586993,
+      "grad_norm": 0.5997834205627441,
+      "learning_rate": 0.00014383561643835615,
+      "loss": 11.3051,
+      "step": 85
+    },
+    {
+      "epoch": 0.029549456685135075,
+      "grad_norm": 0.6137603521347046,
+      "learning_rate": 0.00014554794520547945,
+      "loss": 11.3053,
+      "step": 86
+    },
+    {
+      "epoch": 0.02989305501868316,
+      "grad_norm": 0.5946083068847656,
+      "learning_rate": 0.00014726027397260274,
+      "loss": 11.2886,
+      "step": 87
+    },
+    {
+      "epoch": 0.030236653352231243,
+      "grad_norm": 0.6125856041908264,
+      "learning_rate": 0.00014897260273972603,
+      "loss": 11.2621,
+      "step": 88
+    },
+    {
+      "epoch": 0.030580251685779325,
+      "grad_norm": 0.6170690059661865,
+      "learning_rate": 0.0001506849315068493,
+      "loss": 11.2379,
+      "step": 89
+    },
+    {
+      "epoch": 0.030923850019327407,
+      "grad_norm": 0.613112211227417,
+      "learning_rate": 0.0001523972602739726,
+      "loss": 11.2319,
+      "step": 90
+    },
+    {
+      "epoch": 0.03126744835287549,
+      "grad_norm": 0.6426445841789246,
+      "learning_rate": 0.00015410958904109589,
+      "loss": 11.217,
+      "step": 91
+    },
+    {
+      "epoch": 0.03161104668642357,
+      "grad_norm": 0.6466779708862305,
+      "learning_rate": 0.0001558219178082192,
+      "loss": 11.1905,
+      "step": 92
+    },
+    {
+      "epoch": 0.03195464501997165,
+      "grad_norm": 0.6395965814590454,
+      "learning_rate": 0.00015753424657534247,
+      "loss": 11.182,
+      "step": 93
+    },
+    {
+      "epoch": 0.03229824335351974,
+      "grad_norm": 0.6484614014625549,
+      "learning_rate": 0.00015924657534246577,
+      "loss": 11.166,
+      "step": 94
+    },
+    {
+      "epoch": 0.03264184168706782,
+      "grad_norm": 0.6011567115783691,
+      "learning_rate": 0.00016095890410958906,
+      "loss": 11.1601,
+      "step": 95
+    },
+    {
+      "epoch": 0.0329854400206159,
+      "grad_norm": 0.6242689490318298,
+      "learning_rate": 0.00016267123287671235,
+      "loss": 11.1447,
+      "step": 96
+    },
+    {
+      "epoch": 0.03332903835416398,
+      "grad_norm": 0.6144719123840332,
+      "learning_rate": 0.00016438356164383562,
+      "loss": 11.1382,
+      "step": 97
+    },
+    {
+      "epoch": 0.03367263668771207,
+      "grad_norm": 0.6248654127120972,
+      "learning_rate": 0.00016609589041095891,
+      "loss": 11.1246,
+      "step": 98
+    },
+    {
+      "epoch": 0.034016235021260145,
+      "grad_norm": 0.6292065978050232,
+      "learning_rate": 0.0001678082191780822,
+      "loss": 11.0922,
+      "step": 99
+    },
+    {
+      "epoch": 0.03435983335480823,
+      "grad_norm": 0.6909387707710266,
+      "learning_rate": 0.0001695205479452055,
+      "loss": 11.0901,
+      "step": 100
+    },
+    {
+      "epoch": 0.03470343168835631,
+      "grad_norm": 0.6519057750701904,
+      "learning_rate": 0.00017123287671232877,
+      "loss": 11.0416,
+      "step": 101
+    },
+    {
+      "epoch": 0.035047030021904395,
+      "grad_norm": 0.6702002882957458,
+      "learning_rate": 0.00017294520547945206,
+      "loss": 11.0272,
+      "step": 102
+    },
+    {
+      "epoch": 0.03539062835545247,
+      "grad_norm": 0.6443936228752136,
+      "learning_rate": 0.00017465753424657536,
+      "loss": 11.015,
+      "step": 103
+    },
+    {
+      "epoch": 0.03573422668900056,
+      "grad_norm": 0.6505857110023499,
+      "learning_rate": 0.00017636986301369865,
+      "loss": 10.999,
+      "step": 104
+    },
+    {
+      "epoch": 0.03607782502254864,
+      "grad_norm": 0.6232591867446899,
+      "learning_rate": 0.00017808219178082192,
+      "loss": 10.9871,
+      "step": 105
+    },
+    {
+      "epoch": 0.03642142335609672,
+      "grad_norm": 0.6446307897567749,
+      "learning_rate": 0.0001797945205479452,
+      "loss": 10.9577,
+      "step": 106
+    },
+    {
+      "epoch": 0.03676502168964481,
+      "grad_norm": 0.662112295627594,
+      "learning_rate": 0.0001815068493150685,
+      "loss": 10.9258,
+      "step": 107
+    },
+    {
+      "epoch": 0.03710862002319289,
+      "grad_norm": 0.6445570588111877,
+      "learning_rate": 0.0001832191780821918,
+      "loss": 10.9299,
+      "step": 108
+    },
+    {
+      "epoch": 0.03745221835674097,
+      "grad_norm": 0.6506933569908142,
+      "learning_rate": 0.00018493150684931506,
+      "loss": 10.9051,
+      "step": 109
+    },
+    {
+      "epoch": 0.03779581669028905,
+      "grad_norm": 0.6474794149398804,
+      "learning_rate": 0.00018664383561643836,
+      "loss": 10.8871,
+      "step": 110
+    },
+    {
+      "epoch": 0.03813941502383714,
+      "grad_norm": 0.6381330490112305,
+      "learning_rate": 0.00018835616438356165,
+      "loss": 10.8668,
+      "step": 111
+    },
+    {
+      "epoch": 0.038483013357385215,
+      "grad_norm": 0.625579833984375,
+      "learning_rate": 0.00019006849315068494,
+      "loss": 10.8491,
+      "step": 112
+    },
+    {
+      "epoch": 0.0388266116909333,
+      "grad_norm": 0.6460126042366028,
+      "learning_rate": 0.0001917808219178082,
+      "loss": 10.8417,
+      "step": 113
+    },
+    {
+      "epoch": 0.03917021002448138,
+      "grad_norm": 0.6264495253562927,
+      "learning_rate": 0.0001934931506849315,
+      "loss": 10.8322,
+      "step": 114
+    },
+    {
+      "epoch": 0.039513808358029465,
+      "grad_norm": 0.6536591053009033,
+      "learning_rate": 0.0001952054794520548,
+      "loss": 10.7748,
+      "step": 115
+    },
+    {
+      "epoch": 0.03985740669157754,
+      "grad_norm": 0.6263400912284851,
+      "learning_rate": 0.0001969178082191781,
+      "loss": 10.7766,
+      "step": 116
+    },
+    {
+      "epoch": 0.04020100502512563,
+      "grad_norm": 0.6201140284538269,
+      "learning_rate": 0.00019863013698630136,
+      "loss": 10.7611,
+      "step": 117
+    },
+    {
+      "epoch": 0.04054460335867371,
+      "grad_norm": 0.615506649017334,
+      "learning_rate": 0.00020034246575342465,
+      "loss": 10.7421,
+      "step": 118
+    },
+    {
+      "epoch": 0.04088820169222179,
+      "grad_norm": 0.6192501783370972,
+      "learning_rate": 0.00020205479452054795,
+      "loss": 10.72,
+      "step": 119
+    },
+    {
+      "epoch": 0.04123180002576988,
+      "grad_norm": 0.6386846899986267,
+      "learning_rate": 0.00020376712328767124,
+      "loss": 10.6968,
+      "step": 120
+    },
+    {
+      "epoch": 0.04157539835931796,
+      "grad_norm": 0.6488209962844849,
+      "learning_rate": 0.0002054794520547945,
+      "loss": 10.6839,
+      "step": 121
+    },
+    {
+      "epoch": 0.04191899669286604,
+      "grad_norm": 0.6438455581665039,
+      "learning_rate": 0.0002071917808219178,
+      "loss": 10.6481,
+      "step": 122
+    },
+    {
+      "epoch": 0.04226259502641412,
+      "grad_norm": 0.6512130498886108,
+      "learning_rate": 0.0002089041095890411,
+      "loss": 10.6226,
+      "step": 123
+    },
+    {
+      "epoch": 0.042606193359962206,
+      "grad_norm": 0.6471563577651978,
+      "learning_rate": 0.0002106164383561644,
+      "loss": 10.6165,
+      "step": 124
+    },
+    {
+      "epoch": 0.042949791693510285,
+      "grad_norm": 0.615242063999176,
+      "learning_rate": 0.00021232876712328765,
+      "loss": 10.6079,
+      "step": 125
+    },
+    {
+      "epoch": 0.04329339002705837,
+      "grad_norm": 0.6485576629638672,
+      "learning_rate": 0.00021404109589041095,
+      "loss": 10.577,
+      "step": 126
+    },
+    {
+      "epoch": 0.04363698836060645,
+      "grad_norm": 0.6325823664665222,
+      "learning_rate": 0.00021575342465753424,
+      "loss": 10.5585,
+      "step": 127
+    },
+    {
+      "epoch": 0.043980586694154535,
+      "grad_norm": 0.6436929702758789,
+      "learning_rate": 0.00021746575342465753,
+      "loss": 10.5302,
+      "step": 128
+    },
+    {
+      "epoch": 0.04432418502770261,
+      "grad_norm": 0.6149619817733765,
+      "learning_rate": 0.0002191780821917808,
+      "loss": 10.5413,
+      "step": 129
+    },
+    {
+      "epoch": 0.0446677833612507,
+      "grad_norm": 0.6164280772209167,
+      "learning_rate": 0.00022089041095890412,
+      "loss": 10.5123,
+      "step": 130
+    },
+    {
+      "epoch": 0.04501138169479878,
+      "grad_norm": 0.6501047015190125,
+      "learning_rate": 0.00022260273972602742,
+      "loss": 10.4795,
+      "step": 131
+    },
+    {
+      "epoch": 0.04535498002834686,
+      "grad_norm": 0.650057852268219,
+      "learning_rate": 0.0002243150684931507,
+      "loss": 10.4438,
+      "step": 132
+    },
+    {
+      "epoch": 0.04569857836189495,
+      "grad_norm": 0.6304041147232056,
+      "learning_rate": 0.00022602739726027398,
+      "loss": 10.4388,
+      "step": 133
+    },
+    {
+      "epoch": 0.04604217669544303,
+      "grad_norm": 0.6359322667121887,
+      "learning_rate": 0.00022773972602739727,
+      "loss": 10.4147,
+      "step": 134
+    },
+    {
+      "epoch": 0.04638577502899111,
+      "grad_norm": 0.6494120359420776,
+      "learning_rate": 0.00022945205479452056,
+      "loss": 10.3909,
+      "step": 135
+    },
+    {
+      "epoch": 0.04672937336253919,
+      "grad_norm": 0.6699193120002747,
+      "learning_rate": 0.00023116438356164386,
+      "loss": 10.3737,
+      "step": 136
+    },
+    {
+      "epoch": 0.047072971696087276,
+      "grad_norm": 0.6436437964439392,
+      "learning_rate": 0.00023287671232876712,
+      "loss": 10.3688,
+      "step": 137
+    },
+    {
+      "epoch": 0.047416570029635355,
+      "grad_norm": 0.6163201928138733,
+      "learning_rate": 0.00023458904109589042,
+      "loss": 10.3468,
+      "step": 138
+    },
+    {
+      "epoch": 0.04776016836318344,
+      "grad_norm": 0.6744312644004822,
+      "learning_rate": 0.0002363013698630137,
+      "loss": 10.3168,
+      "step": 139
+    },
+    {
+      "epoch": 0.04810376669673152,
+      "grad_norm": 0.6373304724693298,
+      "learning_rate": 0.000238013698630137,
+      "loss": 10.3134,
+      "step": 140
+    },
+    {
+      "epoch": 0.048447365030279604,
+      "grad_norm": 0.6340605616569519,
+      "learning_rate": 0.00023972602739726027,
+      "loss": 10.2961,
+      "step": 141
+    },
+    {
+      "epoch": 0.04879096336382768,
+      "grad_norm": 0.6478452682495117,
+      "learning_rate": 0.00024143835616438356,
+      "loss": 10.2708,
+      "step": 142
+    },
+    {
+      "epoch": 0.04913456169737577,
+      "grad_norm": 0.6576781272888184,
+      "learning_rate": 0.00024315068493150686,
+      "loss": 10.2501,
+      "step": 143
+    },
+    {
+      "epoch": 0.04947816003092385,
+      "grad_norm": 0.6450794339179993,
+      "learning_rate": 0.0002448630136986301,
+      "loss": 10.2429,
+      "step": 144
+    },
+    {
+      "epoch": 0.04982175836447193,
+      "grad_norm": 0.6385270953178406,
+      "learning_rate": 0.0002465753424657534,
+      "loss": 10.2387,
+      "step": 145
+    },
+    {
+      "epoch": 0.05016535669802002,
+      "grad_norm": 0.6406118869781494,
+      "learning_rate": 0.0002482876712328767,
+      "loss": 10.1705,
+      "step": 146
+    },
+    {
+      "epoch": 0.0505089550315681,
+      "grad_norm": 0.610375702381134,
+      "learning_rate": 0.00025,
+      "loss": 10.2039,
+      "step": 147
+    },
+    {
+      "epoch": 0.05085255336511618,
+      "grad_norm": 0.6268151998519897,
+      "learning_rate": 0.0002517123287671233,
+      "loss": 10.1923,
+      "step": 148
+    },
+    {
+      "epoch": 0.05119615169866426,
+      "grad_norm": 0.6328505277633667,
+      "learning_rate": 0.0002534246575342466,
+      "loss": 10.1562,
+      "step": 149
+    },
+    {
+      "epoch": 0.051539750032212346,
+      "grad_norm": 0.6489254236221313,
+      "learning_rate": 0.0002551369863013699,
+      "loss": 10.1535,
+      "step": 150
+    },
+    {
+      "epoch": 0.051883348365760425,
+      "grad_norm": 0.6568716764450073,
+      "learning_rate": 0.0002568493150684932,
+      "loss": 10.083,
+      "step": 151
+    },
+    {
+      "epoch": 0.05222694669930851,
+      "grad_norm": 0.6831304430961609,
+      "learning_rate": 0.0002585616438356164,
+      "loss": 10.0611,
+      "step": 152
+    },
+    {
+      "epoch": 0.05257054503285659,
+      "grad_norm": 0.6912304759025574,
+      "learning_rate": 0.0002602739726027397,
+      "loss": 10.0351,
+      "step": 153
+    },
+    {
+      "epoch": 0.052914143366404674,
+      "grad_norm": 0.6460905075073242,
+      "learning_rate": 0.000261986301369863,
+      "loss": 10.0702,
+      "step": 154
+    },
+    {
+      "epoch": 0.05325774169995275,
+      "grad_norm": 0.6496686339378357,
+      "learning_rate": 0.0002636986301369863,
+      "loss": 10.0229,
+      "step": 155
+    },
+    {
+      "epoch": 0.05360134003350084,
+      "grad_norm": 0.684745728969574,
+      "learning_rate": 0.0002654109589041096,
+      "loss": 9.986,
+      "step": 156
+    },
+    {
+      "epoch": 0.05394493836704892,
+      "grad_norm": 0.6830720901489258,
+      "learning_rate": 0.0002671232876712329,
+      "loss": 9.959,
+      "step": 157
+    },
+    {
+      "epoch": 0.054288536700597,
+      "grad_norm": 0.6738273501396179,
+      "learning_rate": 0.0002688356164383562,
+      "loss": 9.9602,
+      "step": 158
+    },
+    {
+      "epoch": 0.05463213503414508,
+      "grad_norm": 0.6997708082199097,
+      "learning_rate": 0.0002705479452054795,
+      "loss": 9.9037,
+      "step": 159
+    },
+    {
+      "epoch": 0.054975733367693166,
+      "grad_norm": 0.6702543497085571,
+      "learning_rate": 0.0002722602739726027,
+      "loss": 9.8971,
+      "step": 160
+    },
+    {
+      "epoch": 0.05531933170124125,
+      "grad_norm": 0.6914408206939697,
+      "learning_rate": 0.000273972602739726,
+      "loss": 9.8413,
+      "step": 161
+    },
+    {
+      "epoch": 0.05566293003478933,
+      "grad_norm": 0.670492947101593,
+      "learning_rate": 0.0002756849315068493,
+      "loss": 9.8841,
+      "step": 162
+    },
+    {
+      "epoch": 0.056006528368337416,
+      "grad_norm": 0.6712900996208191,
+      "learning_rate": 0.0002773972602739726,
+      "loss": 9.8381,
+      "step": 163
+    },
+    {
+      "epoch": 0.056350126701885495,
+      "grad_norm": 0.6814056038856506,
+      "learning_rate": 0.0002791095890410959,
+      "loss": 9.814,
+      "step": 164
+    },
+    {
+      "epoch": 0.05669372503543358,
+      "grad_norm": 0.6779478192329407,
+      "learning_rate": 0.0002808219178082192,
+      "loss": 9.7816,
+      "step": 165
+    },
+    {
+      "epoch": 0.05703732336898166,
+      "grad_norm": 0.6546109914779663,
+      "learning_rate": 0.0002825342465753425,
+      "loss": 9.7916,
+      "step": 166
+    },
+    {
+      "epoch": 0.057380921702529744,
+      "grad_norm": 0.6739956736564636,
+      "learning_rate": 0.00028424657534246577,
+      "loss": 9.7473,
+      "step": 167
+    },
+    {
+      "epoch": 0.05772452003607782,
+      "grad_norm": 0.6486943960189819,
+      "learning_rate": 0.000285958904109589,
+      "loss": 9.7394,
+      "step": 168
+    },
+    {
+      "epoch": 0.05806811836962591,
+      "grad_norm": 0.6602917313575745,
+      "learning_rate": 0.0002876712328767123,
+      "loss": 9.6986,
+      "step": 169
+    },
+    {
+      "epoch": 0.05841171670317399,
+      "grad_norm": 0.6496753096580505,
+      "learning_rate": 0.0002893835616438356,
+      "loss": 9.7017,
+      "step": 170
+    },
+    {
+      "epoch": 0.05875531503672207,
+      "grad_norm": 0.6671965718269348,
+      "learning_rate": 0.0002910958904109589,
+      "loss": 9.6521,
+      "step": 171
+    },
+    {
+      "epoch": 0.05909891337027015,
+      "grad_norm": 0.6669342517852783,
+      "learning_rate": 0.0002928082191780822,
+      "loss": 9.6483,
+      "step": 172
+    },
+    {
+      "epoch": 0.059442511703818236,
+      "grad_norm": 0.673161506652832,
+      "learning_rate": 0.0002945205479452055,
+      "loss": 9.6201,
+      "step": 173
+    },
+    {
+      "epoch": 0.05978611003736632,
+      "grad_norm": 0.6617433428764343,
+      "learning_rate": 0.00029623287671232877,
+      "loss": 9.6162,
+      "step": 174
+    },
+    {
+      "epoch": 0.0601297083709144,
+      "grad_norm": 0.6584163904190063,
+      "learning_rate": 0.00029794520547945206,
+      "loss": 9.5791,
+      "step": 175
+    },
+    {
+      "epoch": 0.060473306704462486,
+      "grad_norm": 0.6521568894386292,
+      "learning_rate": 0.0002996575342465753,
+      "loss": 9.6007,
+      "step": 176
+    },
+    {
+      "epoch": 0.060816905038010564,
+      "grad_norm": 0.6696915626525879,
+      "learning_rate": 0.0003013698630136986,
+      "loss": 9.5377,
+      "step": 177
+    },
+    {
+      "epoch": 0.06116050337155865,
+      "grad_norm": 0.6637017726898193,
+      "learning_rate": 0.0003030821917808219,
+      "loss": 9.5216,
+      "step": 178
+    },
+    {
+      "epoch": 0.06150410170510673,
+      "grad_norm": 0.6698155999183655,
+      "learning_rate": 0.0003047945205479452,
+      "loss": 9.4881,
+      "step": 179
+    },
+    {
+      "epoch": 0.061847700038654814,
+      "grad_norm": 0.6680699586868286,
+      "learning_rate": 0.0003065068493150685,
+      "loss": 9.4521,
+      "step": 180
+    },
+    {
+      "epoch": 0.06219129837220289,
+      "grad_norm": 0.6579115986824036,
+      "learning_rate": 0.00030821917808219177,
+      "loss": 9.4734,
+      "step": 181
+    },
+    {
+      "epoch": 0.06253489670575098,
+      "grad_norm": 0.6589128375053406,
+      "learning_rate": 0.00030993150684931507,
+      "loss": 9.4434,
+      "step": 182
+    },
+    {
+      "epoch": 0.06287849503929906,
+      "grad_norm": 0.6762176752090454,
+      "learning_rate": 0.0003116438356164384,
+      "loss": 9.3834,
+      "step": 183
+    },
+    {
+      "epoch": 0.06322209337284714,
+      "grad_norm": 0.6807367205619812,
+      "learning_rate": 0.0003133561643835616,
+      "loss": 9.3898,
+      "step": 184
+    },
+    {
+      "epoch": 0.06356569170639523,
+      "grad_norm": 0.6483362317085266,
+      "learning_rate": 0.00031506849315068495,
+      "loss": 9.3783,
+      "step": 185
+    },
+    {
+      "epoch": 0.0639092900399433,
+      "grad_norm": 0.6637298464775085,
+      "learning_rate": 0.00031678082191780824,
+      "loss": 9.3476,
+      "step": 186
+    },
+    {
+      "epoch": 0.06425288837349138,
+      "grad_norm": 0.6700524091720581,
+      "learning_rate": 0.00031849315068493153,
+      "loss": 9.3696,
+      "step": 187
+    },
+    {
+      "epoch": 0.06459648670703948,
+      "grad_norm": 0.6919601559638977,
+      "learning_rate": 0.00032020547945205483,
+      "loss": 9.2993,
+      "step": 188
+    },
+    {
+      "epoch": 0.06494008504058756,
+      "grad_norm": 0.6557508111000061,
+      "learning_rate": 0.0003219178082191781,
+      "loss": 9.316,
+      "step": 189
+    },
+    {
+      "epoch": 0.06528368337413563,
+      "grad_norm": 0.6749283671379089,
+      "learning_rate": 0.0003236301369863014,
+      "loss": 9.2672,
+      "step": 190
+    },
+    {
+      "epoch": 0.06562728170768371,
+      "grad_norm": 0.6584534049034119,
+      "learning_rate": 0.0003253424657534247,
+      "loss": 9.3218,
+      "step": 191
+    },
+    {
+      "epoch": 0.0659708800412318,
+      "grad_norm": 0.6565567851066589,
+      "learning_rate": 0.00032705479452054795,
+      "loss": 9.2853,
+      "step": 192
+    },
+    {
+      "epoch": 0.06631447837477988,
+      "grad_norm": 0.6454471349716187,
+      "learning_rate": 0.00032876712328767124,
+      "loss": 9.3203,
+      "step": 193
+    },
+    {
+      "epoch": 0.06665807670832796,
+      "grad_norm": 0.6428927183151245,
+      "learning_rate": 0.00033047945205479454,
+      "loss": 9.235,
+      "step": 194
+    },
+    {
+      "epoch": 0.06700167504187604,
+      "grad_norm": 0.6591072082519531,
+      "learning_rate": 0.00033219178082191783,
+      "loss": 9.2496,
+      "step": 195
+    },
+    {
+      "epoch": 0.06734527337542413,
+      "grad_norm": 0.666397750377655,
+      "learning_rate": 0.0003339041095890411,
+      "loss": 9.2136,
+      "step": 196
+    },
+    {
+      "epoch": 0.06768887170897221,
+      "grad_norm": 0.6873658895492554,
+      "learning_rate": 0.0003356164383561644,
+      "loss": 9.1737,
+      "step": 197
+    },
+    {
+      "epoch": 0.06803247004252029,
+      "grad_norm": 0.6527069211006165,
+      "learning_rate": 0.0003373287671232877,
+      "loss": 9.221,
+      "step": 198
+    },
+    {
+      "epoch": 0.06837606837606838,
+      "grad_norm": 0.6505078077316284,
+      "learning_rate": 0.000339041095890411,
+      "loss": 9.199,
+      "step": 199
+    },
+    {
+      "epoch": 0.06871966670961646,
+      "grad_norm": 0.6579967141151428,
+      "learning_rate": 0.00034075342465753424,
+      "loss": 9.2194,
+      "step": 200
+    },
+    {
+      "epoch": 0.06906326504316454,
+      "grad_norm": 0.7028538584709167,
+      "learning_rate": 0.00034246575342465754,
+      "loss": 9.0879,
+      "step": 201
+    },
+    {
+      "epoch": 0.06940686337671262,
+      "grad_norm": 0.7203247547149658,
+      "learning_rate": 0.00034417808219178083,
+      "loss": 9.0403,
+      "step": 202
+    },
+    {
+      "epoch": 0.06975046171026071,
+      "grad_norm": 0.6787813901901245,
+      "learning_rate": 0.0003458904109589041,
+      "loss": 9.0896,
+      "step": 203
+    },
+    {
+      "epoch": 0.07009406004380879,
+      "grad_norm": 0.6986070275306702,
+      "learning_rate": 0.0003476027397260274,
+      "loss": 9.025,
+      "step": 204
+    },
+    {
+      "epoch": 0.07043765837735687,
+      "grad_norm": 0.7031541466712952,
+      "learning_rate": 0.0003493150684931507,
+      "loss": 8.9863,
+      "step": 205
+    },
+    {
+      "epoch": 0.07078125671090495,
+      "grad_norm": 0.6880903244018555,
+      "learning_rate": 0.000351027397260274,
+      "loss": 9.0159,
+      "step": 206
+    },
+    {
+      "epoch": 0.07112485504445304,
+      "grad_norm": 0.6828157901763916,
+      "learning_rate": 0.0003527397260273973,
+      "loss": 8.9844,
+      "step": 207
+    },
+    {
+      "epoch": 0.07146845337800112,
+      "grad_norm": 0.6969785690307617,
+      "learning_rate": 0.00035445205479452054,
+      "loss": 8.9524,
+      "step": 208
+    },
+    {
+      "epoch": 0.0718120517115492,
+      "grad_norm": 0.6816834211349487,
+      "learning_rate": 0.00035616438356164383,
+      "loss": 8.9422,
+      "step": 209
+    },
+    {
+      "epoch": 0.07215565004509727,
+      "grad_norm": 0.7117534875869751,
+      "learning_rate": 0.0003578767123287671,
+      "loss": 8.8944,
+      "step": 210
+    },
+    {
+      "epoch": 0.07249924837864537,
+      "grad_norm": 0.685808002948761,
+      "learning_rate": 0.0003595890410958904,
+      "loss": 8.9264,
+      "step": 211
+    },
+    {
+      "epoch": 0.07284284671219345,
+      "grad_norm": 0.6689701080322266,
+      "learning_rate": 0.0003613013698630137,
+      "loss": 8.8904,
+      "step": 212
+    },
+    {
+      "epoch": 0.07318644504574152,
+      "grad_norm": 0.6558648347854614,
+      "learning_rate": 0.000363013698630137,
+      "loss": 8.8794,
+      "step": 213
+    },
+    {
+      "epoch": 0.07353004337928962,
+      "grad_norm": 0.6610527038574219,
+      "learning_rate": 0.0003647260273972603,
+      "loss": 8.8526,
+      "step": 214
+    },
+    {
+      "epoch": 0.0738736417128377,
+      "grad_norm": 0.6478385329246521,
+      "learning_rate": 0.0003664383561643836,
+      "loss": 8.8549,
+      "step": 215
+    },
+    {
+      "epoch": 0.07421724004638577,
+      "grad_norm": 0.6762006282806396,
+      "learning_rate": 0.00036815068493150683,
+      "loss": 8.8003,
+      "step": 216
+    },
+    {
+      "epoch": 0.07456083837993385,
+      "grad_norm": 0.6985549330711365,
+      "learning_rate": 0.0003698630136986301,
+      "loss": 8.7256,
+      "step": 217
+    },
+    {
+      "epoch": 0.07490443671348194,
+      "grad_norm": 0.647552490234375,
+      "learning_rate": 0.0003715753424657534,
+      "loss": 8.822,
+      "step": 218
+    },
+    {
+      "epoch": 0.07524803504703002,
+      "grad_norm": 0.6919633746147156,
+      "learning_rate": 0.0003732876712328767,
+      "loss": 8.6791,
+      "step": 219
+    },
+    {
+      "epoch": 0.0755916333805781,
+      "grad_norm": 0.6744142770767212,
+      "learning_rate": 0.000375,
+      "loss": 8.6984,
+      "step": 220
+    },
+    {
+      "epoch": 0.07593523171412618,
+      "grad_norm": 0.6745022535324097,
+      "learning_rate": 0.0003767123287671233,
+      "loss": 8.6842,
+      "step": 221
+    },
+    {
+      "epoch": 0.07627883004767427,
+      "grad_norm": 0.6558954119682312,
+      "learning_rate": 0.0003784246575342466,
+      "loss": 8.6935,
+      "step": 222
+    },
+    {
+      "epoch": 0.07662242838122235,
+      "grad_norm": 0.6784209609031677,
+      "learning_rate": 0.0003801369863013699,
+      "loss": 8.6508,
+      "step": 223
+    },
+    {
+      "epoch": 0.07696602671477043,
+      "grad_norm": 0.6614058613777161,
+      "learning_rate": 0.0003818493150684932,
+      "loss": 8.6713,
+      "step": 224
+    },
+    {
+      "epoch": 0.07730962504831851,
+      "grad_norm": 0.6582658886909485,
+      "learning_rate": 0.0003835616438356164,
+      "loss": 8.6104,
+      "step": 225
+    },
+    {
+      "epoch": 0.0776532233818666,
+      "grad_norm": 0.6616186499595642,
+      "learning_rate": 0.0003852739726027397,
+      "loss": 8.6167,
+      "step": 226
+    },
+    {
+      "epoch": 0.07799682171541468,
+      "grad_norm": 0.6750086545944214,
+      "learning_rate": 0.000386986301369863,
+      "loss": 8.6126,
+      "step": 227
+    },
+    {
+      "epoch": 0.07834042004896276,
+      "grad_norm": 0.6450734734535217,
+      "learning_rate": 0.0003886986301369863,
+      "loss": 8.6054,
+      "step": 228
+    },
+    {
+      "epoch": 0.07868401838251085,
+      "grad_norm": 0.6688063144683838,
+      "learning_rate": 0.0003904109589041096,
+      "loss": 8.5684,
+      "step": 229
+    },
+    {
+      "epoch": 0.07902761671605893,
+      "grad_norm": 0.6341917514801025,
+      "learning_rate": 0.0003921232876712329,
+      "loss": 8.5838,
+      "step": 230
+    },
+    {
+      "epoch": 0.07937121504960701,
+      "grad_norm": 0.6572481393814087,
+      "learning_rate": 0.0003938356164383562,
+      "loss": 8.5452,
+      "step": 231
+    },
+    {
+      "epoch": 0.07971481338315509,
+      "grad_norm": 0.6542115807533264,
+      "learning_rate": 0.0003955479452054795,
+      "loss": 8.5227,
+      "step": 232
+    },
+    {
+      "epoch": 0.08005841171670318,
+      "grad_norm": 0.6360100507736206,
+      "learning_rate": 0.0003972602739726027,
+      "loss": 8.5747,
+      "step": 233
+    },
+    {
+      "epoch": 0.08040201005025126,
+      "grad_norm": 0.6474318504333496,
+      "learning_rate": 0.000398972602739726,
+      "loss": 8.53,
+      "step": 234
+    },
+    {
+      "epoch": 0.08074560838379934,
+      "grad_norm": 0.639065682888031,
+      "learning_rate": 0.0004006849315068493,
+      "loss": 8.4835,
+      "step": 235
+    },
+    {
+      "epoch": 0.08108920671734741,
+      "grad_norm": 0.6288236379623413,
+      "learning_rate": 0.0004023972602739726,
+      "loss": 8.5055,
+      "step": 236
+    },
+    {
+      "epoch": 0.08143280505089551,
+      "grad_norm": 0.600935697555542,
+      "learning_rate": 0.0004041095890410959,
+      "loss": 8.5239,
+      "step": 237
+    },
+    {
+      "epoch": 0.08177640338444359,
+      "grad_norm": 0.6632112264633179,
+      "learning_rate": 0.0004058219178082192,
+      "loss": 8.4146,
+      "step": 238
+    },
+    {
+      "epoch": 0.08212000171799166,
+      "grad_norm": 0.6458113193511963,
+      "learning_rate": 0.0004075342465753425,
+      "loss": 8.4419,
+      "step": 239
+    },
+    {
+      "epoch": 0.08246360005153976,
+      "grad_norm": 0.6225025057792664,
+      "learning_rate": 0.00040924657534246577,
+      "loss": 8.4463,
+      "step": 240
+    },
+    {
+      "epoch": 0.08280719838508784,
+      "grad_norm": 0.5956367254257202,
+      "learning_rate": 0.000410958904109589,
+      "loss": 8.4468,
+      "step": 241
+    },
+    {
+      "epoch": 0.08315079671863591,
+      "grad_norm": 0.6125132441520691,
+      "learning_rate": 0.0004126712328767123,
+      "loss": 8.4347,
+      "step": 242
+    },
+    {
+      "epoch": 0.08349439505218399,
+      "grad_norm": 0.5911486744880676,
+      "learning_rate": 0.0004143835616438356,
+      "loss": 8.5122,
+      "step": 243
+    },
+    {
+      "epoch": 0.08383799338573208,
+      "grad_norm": 0.5973955392837524,
+      "learning_rate": 0.0004160958904109589,
+      "loss": 8.5174,
+      "step": 244
+    },
+    {
+      "epoch": 0.08418159171928016,
+      "grad_norm": 0.6203837394714355,
+      "learning_rate": 0.0004178082191780822,
+      "loss": 8.3962,
+      "step": 245
+    },
+    {
+      "epoch": 0.08452519005282824,
+      "grad_norm": 0.624750018119812,
+      "learning_rate": 0.0004195205479452055,
+      "loss": 8.3446,
+      "step": 246
+    },
+    {
+      "epoch": 0.08486878838637632,
+      "grad_norm": 0.613397479057312,
+      "learning_rate": 0.0004212328767123288,
+      "loss": 8.377,
+      "step": 247
+    },
+    {
+      "epoch": 0.08521238671992441,
+      "grad_norm": 0.6277963519096375,
+      "learning_rate": 0.00042294520547945207,
+      "loss": 8.3163,
+      "step": 248
+    },
+    {
+      "epoch": 0.08555598505347249,
+      "grad_norm": 0.6092117428779602,
+      "learning_rate": 0.0004246575342465753,
+      "loss": 8.4043,
+      "step": 249
+    },
+    {
+      "epoch": 0.08589958338702057,
+      "grad_norm": 0.6500337719917297,
+      "learning_rate": 0.0004263698630136986,
+      "loss": 8.3297,
+      "step": 250
+    },
+    {
+      "epoch": 0.08624318172056865,
+      "grad_norm": 0.6311240792274475,
+      "learning_rate": 0.0004280821917808219,
+      "loss": 8.3059,
+      "step": 251
+    },
+    {
+      "epoch": 0.08658678005411674,
+      "grad_norm": 0.6225205063819885,
+      "learning_rate": 0.0004297945205479452,
+      "loss": 8.2266,
+      "step": 252
+    },
+    {
+      "epoch": 0.08693037838766482,
+      "grad_norm": 0.6449727416038513,
+      "learning_rate": 0.0004315068493150685,
+      "loss": 8.2155,
+      "step": 253
+    },
+    {
+      "epoch": 0.0872739767212129,
+      "grad_norm": 0.6553236842155457,
+      "learning_rate": 0.0004332191780821918,
+      "loss": 8.1811,
+      "step": 254
+    },
+    {
+      "epoch": 0.08761757505476099,
+      "grad_norm": 0.6186061501502991,
+      "learning_rate": 0.00043493150684931507,
+      "loss": 8.21,
+      "step": 255
+    },
+    {
+      "epoch": 0.08796117338830907,
+      "grad_norm": 0.6524087190628052,
+      "learning_rate": 0.0004366438356164384,
+      "loss": 8.1486,
+      "step": 256
+    },
+    {
+      "epoch": 0.08830477172185715,
+      "grad_norm": 0.6334664225578308,
+      "learning_rate": 0.0004383561643835616,
+      "loss": 8.207,
+      "step": 257
+    },
+    {
+      "epoch": 0.08864837005540523,
+      "grad_norm": 0.597114086151123,
+      "learning_rate": 0.00044006849315068495,
+      "loss": 8.2354,
+      "step": 258
+    },
+    {
+      "epoch": 0.08899196838895332,
+      "grad_norm": 0.599537193775177,
+      "learning_rate": 0.00044178082191780824,
+      "loss": 8.1919,
+      "step": 259
+    },
+    {
+      "epoch": 0.0893355667225014,
+      "grad_norm": 0.6212196946144104,
+      "learning_rate": 0.00044349315068493154,
+      "loss": 8.1179,
+      "step": 260
+    },
+    {
+      "epoch": 0.08967916505604948,
+      "grad_norm": 0.5893779397010803,
+      "learning_rate": 0.00044520547945205483,
+      "loss": 8.2266,
+      "step": 261
+    },
+    {
+      "epoch": 0.09002276338959755,
+      "grad_norm": 0.5951864719390869,
+      "learning_rate": 0.0004469178082191781,
+      "loss": 8.1375,
+      "step": 262
+    },
+    {
+      "epoch": 0.09036636172314565,
+      "grad_norm": 0.6006582975387573,
+      "learning_rate": 0.0004486301369863014,
+      "loss": 8.1599,
+      "step": 263
+    },
+    {
+      "epoch": 0.09070996005669373,
+      "grad_norm": 0.5896690487861633,
+      "learning_rate": 0.0004503424657534247,
+      "loss": 8.08,
+      "step": 264
+    },
+    {
+      "epoch": 0.0910535583902418,
+      "grad_norm": 0.5560733675956726,
+      "learning_rate": 0.00045205479452054795,
+      "loss": 8.1453,
+      "step": 265
+    },
+    {
+      "epoch": 0.0913971567237899,
+      "grad_norm": 0.5687079429626465,
+      "learning_rate": 0.00045376712328767124,
+      "loss": 8.0543,
+      "step": 266
+    },
+    {
+      "epoch": 0.09174075505733797,
+      "grad_norm": 0.5450804829597473,
+      "learning_rate": 0.00045547945205479454,
+      "loss": 8.1676,
+      "step": 267
+    },
+    {
+      "epoch": 0.09208435339088605,
+      "grad_norm": 0.536425769329071,
+      "learning_rate": 0.00045719178082191783,
+      "loss": 8.0958,
+      "step": 268
+    },
+    {
+      "epoch": 0.09242795172443413,
+      "grad_norm": 0.5232003331184387,
+      "learning_rate": 0.0004589041095890411,
+      "loss": 8.1276,
+      "step": 269
+    },
+    {
+      "epoch": 0.09277155005798222,
+      "grad_norm": 0.556194543838501,
+      "learning_rate": 0.0004606164383561644,
+      "loss": 8.0365,
+      "step": 270
+    },
+    {
+      "epoch": 0.0931151483915303,
+      "grad_norm": 0.5368773341178894,
+      "learning_rate": 0.0004623287671232877,
+      "loss": 8.0283,
+      "step": 271
+    },
+    {
+      "epoch": 0.09345874672507838,
+      "grad_norm": 0.5228593945503235,
+      "learning_rate": 0.000464041095890411,
+      "loss": 8.0311,
+      "step": 272
+    },
+    {
+      "epoch": 0.09380234505862646,
+      "grad_norm": 0.5249463319778442,
+      "learning_rate": 0.00046575342465753425,
+      "loss": 7.9928,
+      "step": 273
+    },
+    {
+      "epoch": 0.09414594339217455,
+      "grad_norm": 0.511807918548584,
+      "learning_rate": 0.00046746575342465754,
+      "loss": 8.0101,
+      "step": 274
+    },
+    {
+      "epoch": 0.09448954172572263,
+      "grad_norm": 0.5214989185333252,
+      "learning_rate": 0.00046917808219178083,
+      "loss": 8.0251,
+      "step": 275
+    },
+    {
+      "epoch": 0.09483314005927071,
+      "grad_norm": 0.48177778720855713,
+      "learning_rate": 0.0004708904109589041,
+      "loss": 8.0669,
+      "step": 276
+    },
+    {
+      "epoch": 0.09517673839281879,
+      "grad_norm": 0.4860067069530487,
+      "learning_rate": 0.0004726027397260274,
+      "loss": 8.0312,
+      "step": 277
+    },
+    {
+      "epoch": 0.09552033672636688,
+      "grad_norm": 0.4770212769508362,
+      "learning_rate": 0.0004743150684931507,
+      "loss": 7.981,
+      "step": 278
+    },
+    {
+      "epoch": 0.09586393505991496,
+      "grad_norm": 0.4737316966056824,
+      "learning_rate": 0.000476027397260274,
+      "loss": 7.9322,
+      "step": 279
+    },
+    {
+      "epoch": 0.09620753339346304,
+      "grad_norm": 0.48152047395706177,
+      "learning_rate": 0.0004777397260273973,
+      "loss": 7.9734,
+      "step": 280
+    },
+    {
+      "epoch": 0.09655113172701113,
+      "grad_norm": 0.455615371465683,
+      "learning_rate": 0.00047945205479452054,
+      "loss": 7.9721,
+      "step": 281
+    },
+    {
+      "epoch": 0.09689473006055921,
+      "grad_norm": 0.45772621035575867,
+      "learning_rate": 0.00048116438356164383,
+      "loss": 8.0154,
+      "step": 282
+    },
+    {
+      "epoch": 0.09723832839410729,
+      "grad_norm": 0.4532890021800995,
+      "learning_rate": 0.00048287671232876713,
+      "loss": 7.9678,
+      "step": 283
+    },
+    {
+      "epoch": 0.09758192672765537,
+      "grad_norm": 0.46691393852233887,
+      "learning_rate": 0.0004845890410958904,
+      "loss": 7.8717,
+      "step": 284
+    },
+    {
+      "epoch": 0.09792552506120346,
+      "grad_norm": 0.4515172243118286,
+      "learning_rate": 0.0004863013698630137,
+      "loss": 7.9627,
+      "step": 285
+    },
+    {
+      "epoch": 0.09826912339475154,
+      "grad_norm": 0.4819401800632477,
+      "learning_rate": 0.000488013698630137,
+      "loss": 7.9424,
+      "step": 286
+    },
+    {
+      "epoch": 0.09861272172829962,
+      "grad_norm": 0.4063154458999634,
+      "learning_rate": 0.0004897260273972602,
+      "loss": 7.9792,
+      "step": 287
+    },
+    {
+      "epoch": 0.0989563200618477,
+      "grad_norm": 0.4167320132255554,
+      "learning_rate": 0.0004914383561643835,
+      "loss": 7.9856,
+      "step": 288
+    },
+    {
+      "epoch": 0.09929991839539579,
+      "grad_norm": 0.4385222792625427,
+      "learning_rate": 0.0004931506849315068,
+      "loss": 7.8981,
+      "step": 289
+    },
+    {
+      "epoch": 0.09964351672894387,
+      "grad_norm": 0.38443422317504883,
+      "learning_rate": 0.0004948630136986301,
+      "loss": 8.0099,
+      "step": 290
+    },
+    {
+      "epoch": 0.09998711506249194,
+      "grad_norm": 0.4247475564479828,
+      "learning_rate": 0.0004965753424657534,
+      "loss": 7.912,
+      "step": 291
+    },
+    {
+      "epoch": 0.10033071339604004,
+      "grad_norm": 0.3842032551765442,
+      "learning_rate": 0.0004982876712328767,
+      "loss": 7.9831,
+      "step": 292
+    },
+    {
+      "epoch": 0.10067431172958811,
+      "grad_norm": 0.40302786231040955,
+      "learning_rate": 0.0005,
+      "loss": 7.9525,
+      "step": 293
+    },
+    {
+      "epoch": 0.1010179100631362,
+      "grad_norm": 0.36673638224601746,
+      "learning_rate": 0.0004999998201382936,
+      "loss": 7.9676,
+      "step": 294
+    },
+    {
+      "epoch": 0.10136150839668427,
+      "grad_norm": 0.4072817862033844,
+      "learning_rate": 0.000499999280553433,
+      "loss": 7.8912,
+      "step": 295
+    },
+    {
+      "epoch": 0.10170510673023236,
+      "grad_norm": 0.3764428496360779,
+      "learning_rate": 0.0004999983812461949,
+      "loss": 8.003,
+      "step": 296
+    },
+    {
+      "epoch": 0.10204870506378044,
+      "grad_norm": 0.38769039511680603,
+      "learning_rate": 0.0004999971222178729,
+      "loss": 7.9691,
+      "step": 297
+    },
+    {
+      "epoch": 0.10239230339732852,
+      "grad_norm": 0.35378220677375793,
+      "learning_rate": 0.0004999955034702791,
+      "loss": 7.9757,
+      "step": 298
+    },
+    {
+      "epoch": 0.1027359017308766,
+      "grad_norm": 0.37706032395362854,
+      "learning_rate": 0.0004999935250057423,
+      "loss": 8.0697,
+      "step": 299
+    },
+    {
+      "epoch": 0.10307950006442469,
+      "grad_norm": 0.42491599917411804,
+      "learning_rate": 0.0004999911868271095,
+      "loss": 8.0201,
+      "step": 300
+    },
+    {
+      "epoch": 0.10342309839797277,
+      "grad_norm": 0.3899023234844208,
+      "learning_rate": 0.000499988488937745,
+      "loss": 7.7999,
+      "step": 301
+    },
+    {
+      "epoch": 0.10376669673152085,
+      "grad_norm": 0.4056253135204315,
+      "learning_rate": 0.0004999854313415308,
+      "loss": 7.7656,
+      "step": 302
+    },
+    {
+      "epoch": 0.10411029506506893,
+      "grad_norm": 0.34815019369125366,
+      "learning_rate": 0.0004999820140428665,
+      "loss": 7.8854,
+      "step": 303
+    },
+    {
+      "epoch": 0.10445389339861702,
+      "grad_norm": 0.32494989037513733,
+      "learning_rate": 0.0004999782370466693,
+      "loss": 7.9957,
+      "step": 304
+    },
+    {
+      "epoch": 0.1047974917321651,
+      "grad_norm": 0.3544636368751526,
+      "learning_rate": 0.0004999741003583737,
+      "loss": 7.8446,
+      "step": 305
+    },
+    {
+      "epoch": 0.10514109006571318,
+      "grad_norm": 0.34266841411590576,
+      "learning_rate": 0.000499969603983932,
+      "loss": 7.781,
+      "step": 306
+    },
+    {
+      "epoch": 0.10548468839926127,
+      "grad_norm": 0.3225366175174713,
+      "learning_rate": 0.0004999647479298142,
+      "loss": 7.8695,
+      "step": 307
+    },
+    {
+      "epoch": 0.10582828673280935,
+      "grad_norm": 0.3114038407802582,
+      "learning_rate": 0.0004999595322030074,
+      "loss": 7.8691,
+      "step": 308
+    },
+    {
+      "epoch": 0.10617188506635743,
+      "grad_norm": 0.30096235871315,
+      "learning_rate": 0.0004999539568110165,
+      "loss": 7.9136,
+      "step": 309
+    },
+    {
+      "epoch": 0.1065154833999055,
+      "grad_norm": 0.325207382440567,
+      "learning_rate": 0.0004999480217618641,
+      "loss": 7.8082,
+      "step": 310
+    },
+    {
+      "epoch": 0.1068590817334536,
+      "grad_norm": 0.29082146286964417,
+      "learning_rate": 0.0004999417270640899,
+      "loss": 7.8148,
+      "step": 311
+    },
+    {
+      "epoch": 0.10720268006700168,
+      "grad_norm": 0.29000920057296753,
+      "learning_rate": 0.0004999350727267515,
+      "loss": 7.8117,
+      "step": 312
+    },
+    {
+      "epoch": 0.10754627840054976,
+      "grad_norm": 0.290337473154068,
+      "learning_rate": 0.0004999280587594235,
+      "loss": 7.808,
+      "step": 313
+    },
+    {
+      "epoch": 0.10788987673409783,
+      "grad_norm": 0.2645377516746521,
+      "learning_rate": 0.0004999206851721985,
+      "loss": 7.8217,
+      "step": 314
+    },
+    {
+      "epoch": 0.10823347506764593,
+      "grad_norm": 0.23495185375213623,
+      "learning_rate": 0.0004999129519756862,
+      "loss": 7.8687,
+      "step": 315
+    },
+    {
+      "epoch": 0.108577073401194,
+      "grad_norm": 0.2429482489824295,
+      "learning_rate": 0.0004999048591810139,
+      "loss": 7.823,
+      "step": 316
+    },
+    {
+      "epoch": 0.10892067173474208,
+      "grad_norm": 0.22635473310947418,
+      "learning_rate": 0.0004998964067998262,
+      "loss": 7.8572,
+      "step": 317
+    },
+    {
+      "epoch": 0.10926427006829016,
+      "grad_norm": 0.2399221658706665,
+      "learning_rate": 0.0004998875948442852,
+      "loss": 7.7956,
+      "step": 318
+    },
+    {
+      "epoch": 0.10960786840183825,
+      "grad_norm": 0.24210013449192047,
+      "learning_rate": 0.0004998784233270705,
+      "loss": 7.8072,
+      "step": 319
+    },
+    {
+      "epoch": 0.10995146673538633,
+      "grad_norm": 0.224739670753479,
+      "learning_rate": 0.0004998688922613788,
+      "loss": 7.8351,
+      "step": 320
+    },
+    {
+      "epoch": 0.11029506506893441,
+      "grad_norm": 0.22617986798286438,
+      "learning_rate": 0.0004998590016609242,
+      "loss": 7.8284,
+      "step": 321
+    },
+    {
+      "epoch": 0.1106386634024825,
+      "grad_norm": 0.2693329453468323,
+      "learning_rate": 0.0004998487515399384,
+      "loss": 7.8181,
+      "step": 322
+    },
+    {
+      "epoch": 0.11098226173603058,
+      "grad_norm": 0.3233254551887512,
+      "learning_rate": 0.0004998381419131701,
+      "loss": 7.8468,
+      "step": 323
+    },
+    {
+      "epoch": 0.11132586006957866,
+      "grad_norm": 0.2308340072631836,
+      "learning_rate": 0.0004998271727958857,
+      "loss": 7.8013,
+      "step": 324
+    },
+    {
+      "epoch": 0.11166945840312674,
+      "grad_norm": 0.2707921862602234,
+      "learning_rate": 0.0004998158442038682,
+      "loss": 7.7526,
+      "step": 325
+    },
+    {
+      "epoch": 0.11201305673667483,
+      "grad_norm": 0.5553349852561951,
+      "learning_rate": 0.0004998041561534185,
+      "loss": 7.8472,
+      "step": 326
+    },
+    {
+      "epoch": 0.11235665507022291,
+      "grad_norm": 0.5412582159042358,
+      "learning_rate": 0.0004997921086613543,
+      "loss": 7.8249,
+      "step": 327
+    },
+    {
+      "epoch": 0.11270025340377099,
+      "grad_norm": 0.3885006606578827,
+      "learning_rate": 0.0004997797017450108,
+      "loss": 7.8228,
+      "step": 328
+    },
+    {
+      "epoch": 0.11304385173731907,
+      "grad_norm": 1.165300726890564,
+      "learning_rate": 0.0004997669354222401,
+      "loss": 7.8178,
+      "step": 329
+    },
+    {
+      "epoch": 0.11338745007086716,
+      "grad_norm": 0.4480549991130829,
+      "learning_rate": 0.0004997538097114118,
+      "loss": 7.8745,
+      "step": 330
+    },
+    {
+      "epoch": 0.11373104840441524,
+      "grad_norm": 0.5066674947738647,
+      "learning_rate": 0.0004997403246314123,
+      "loss": 7.8827,
+      "step": 331
+    },
+    {
+      "epoch": 0.11407464673796332,
+      "grad_norm": 0.2583351135253906,
+      "learning_rate": 0.000499726480201645,
+      "loss": 7.8056,
+      "step": 332
+    },
+    {
+      "epoch": 0.11441824507151141,
+      "grad_norm": 0.49098068475723267,
+      "learning_rate": 0.0004997122764420309,
+      "loss": 7.8203,
+      "step": 333
+    },
+    {
+      "epoch": 0.11476184340505949,
+      "grad_norm": 0.3366967439651489,
+      "learning_rate": 0.0004996977133730074,
+      "loss": 7.8302,
+      "step": 334
+    },
+    {
+      "epoch": 0.11510544173860757,
+      "grad_norm": 0.4575011134147644,
+      "learning_rate": 0.0004996827910155292,
+      "loss": 7.8429,
+      "step": 335
+    },
+    {
+      "epoch": 0.11544904007215565,
+      "grad_norm": 1.2171545028686523,
+      "learning_rate": 0.0004996675093910684,
+      "loss": 7.7702,
+      "step": 336
+    },
+    {
+      "epoch": 0.11579263840570374,
+      "grad_norm": 1.0319626331329346,
+      "learning_rate": 0.0004996518685216132,
+      "loss": 7.7597,
+      "step": 337
+    },
+    {
+      "epoch": 0.11613623673925182,
+      "grad_norm": 0.9371957182884216,
+      "learning_rate": 0.0004996358684296692,
+      "loss": 7.8775,
+      "step": 338
+    },
+    {
+      "epoch": 0.1164798350727999,
+      "grad_norm": 1.9794833660125732,
+      "learning_rate": 0.0004996195091382591,
+      "loss": 7.8503,
+      "step": 339
+    },
+    {
+      "epoch": 0.11682343340634797,
+      "grad_norm": 0.7713110446929932,
+      "learning_rate": 0.0004996027906709219,
+      "loss": 7.8266,
+      "step": 340
+    },
+    {
+      "epoch": 0.11716703173989607,
+      "grad_norm": 0.39992398023605347,
+      "learning_rate": 0.0004995857130517139,
+      "loss": 7.9142,
+      "step": 341
+    },
+    {
+      "epoch": 0.11751063007344414,
+      "grad_norm": 0.6665037274360657,
+      "learning_rate": 0.0004995682763052077,
+      "loss": 7.9238,
+      "step": 342
+    },
+    {
+      "epoch": 0.11785422840699222,
+      "grad_norm": 0.43217286467552185,
+      "learning_rate": 0.0004995504804564932,
+      "loss": 7.9302,
+      "step": 343
+    },
+    {
+      "epoch": 0.1181978267405403,
+      "grad_norm": 0.4050378203392029,
+      "learning_rate": 0.0004995323255311767,
+      "loss": 7.9248,
+      "step": 344
+    },
+    {
+      "epoch": 0.1185414250740884,
+      "grad_norm": 1.3695038557052612,
+      "learning_rate": 0.0004995138115553811,
+      "loss": 7.8124,
+      "step": 345
+    },
+    {
+      "epoch": 0.11888502340763647,
+      "grad_norm": 0.388114869594574,
+      "learning_rate": 0.0004994949385557461,
+      "loss": 7.8363,
+      "step": 346
+    },
+    {
+      "epoch": 0.11922862174118455,
+      "grad_norm": 1.1071730852127075,
+      "learning_rate": 0.0004994757065594279,
+      "loss": 7.8569,
+      "step": 347
+    },
+    {
+      "epoch": 0.11957222007473264,
+      "grad_norm": 0.5634297132492065,
+      "learning_rate": 0.0004994561155940994,
+      "loss": 7.9719,
+      "step": 348
+    },
+    {
+      "epoch": 0.11991581840828072,
+      "grad_norm": 0.4597019553184509,
+      "learning_rate": 0.0004994361656879497,
+      "loss": 7.8573,
+      "step": 349
+    },
+    {
+      "epoch": 0.1202594167418288,
+      "grad_norm": 1.1164665222167969,
+      "learning_rate": 0.0004994158568696848,
+      "loss": 7.7194,
+      "step": 350
+    },
+    {
+      "epoch": 0.12060301507537688,
+      "grad_norm": 0.6222408413887024,
+      "learning_rate": 0.0004993951891685269,
+      "loss": 7.8213,
+      "step": 351
+    },
+    {
+      "epoch": 0.12094661340892497,
+      "grad_norm": 0.3941010534763336,
+      "learning_rate": 0.0004993741626142145,
+      "loss": 7.8279,
+      "step": 352
+    },
+    {
+      "epoch": 0.12129021174247305,
+      "grad_norm": 0.46420833468437195,
+      "learning_rate": 0.0004993527772370028,
+      "loss": 7.6712,
+      "step": 353
+    },
+    {
+      "epoch": 0.12163381007602113,
+      "grad_norm": 0.28539156913757324,
+      "learning_rate": 0.0004993310330676629,
+      "loss": 7.8035,
+      "step": 354
+    },
+    {
+      "epoch": 0.12197740840956921,
+      "grad_norm": 0.32697808742523193,
+      "learning_rate": 0.0004993089301374823,
+      "loss": 7.6737,
+      "step": 355
+    },
+    {
+      "epoch": 0.1223210067431173,
+      "grad_norm": 0.33268988132476807,
+      "learning_rate": 0.0004992864684782649,
+      "loss": 7.8325,
+      "step": 356
+    },
+    {
+      "epoch": 0.12266460507666538,
+      "grad_norm": 0.5018799901008606,
+      "learning_rate": 0.0004992636481223306,
+      "loss": 7.6845,
+      "step": 357
+    },
+    {
+      "epoch": 0.12300820341021346,
+      "grad_norm": 0.31378456950187683,
+      "learning_rate": 0.0004992404691025155,
+      "loss": 7.7502,
+      "step": 358
+    },
+    {
+      "epoch": 0.12335180174376155,
+      "grad_norm": 0.512126088142395,
+      "learning_rate": 0.0004992169314521717,
+      "loss": 7.8296,
+      "step": 359
+    },
+    {
+      "epoch": 0.12369540007730963,
+      "grad_norm": 0.3105211555957794,
+      "learning_rate": 0.0004991930352051673,
+      "loss": 7.9009,
+      "step": 360
+    },
+    {
+      "epoch": 0.1240389984108577,
+      "grad_norm": 0.2352656126022339,
+      "learning_rate": 0.0004991687803958866,
+      "loss": 7.7176,
+      "step": 361
+    },
+    {
+      "epoch": 0.12438259674440579,
+      "grad_norm": 0.32939714193344116,
+      "learning_rate": 0.0004991441670592297,
+      "loss": 7.676,
+      "step": 362
+    },
+    {
+      "epoch": 0.12472619507795388,
+      "grad_norm": 0.3954201638698578,
+      "learning_rate": 0.0004991191952306124,
+      "loss": 7.6914,
+      "step": 363
+    },
+    {
+      "epoch": 0.12506979341150196,
+      "grad_norm": 0.3367353677749634,
+      "learning_rate": 0.0004990938649459667,
+      "loss": 7.6831,
+      "step": 364
+    },
+    {
+      "epoch": 0.12541339174505003,
+      "grad_norm": 0.2749853730201721,
+      "learning_rate": 0.00049906817624174,
+      "loss": 7.6483,
+      "step": 365
+    },
+    {
+      "epoch": 0.1257569900785981,
+      "grad_norm": 0.2436531037092209,
+      "learning_rate": 0.0004990421291548958,
+      "loss": 7.7626,
+      "step": 366
+    },
+    {
+      "epoch": 0.1261005884121462,
+      "grad_norm": 0.23887164890766144,
+      "learning_rate": 0.0004990157237229129,
+      "loss": 7.8766,
+      "step": 367
+    },
+    {
+      "epoch": 0.12644418674569427,
+      "grad_norm": 0.22134557366371155,
+      "learning_rate": 0.0004989889599837861,
+      "loss": 7.7312,
+      "step": 368
+    },
+    {
+      "epoch": 0.12678778507924238,
+      "grad_norm": 0.2769152820110321,
+      "learning_rate": 0.0004989618379760254,
+      "loss": 7.7004,
+      "step": 369
+    },
+    {
+      "epoch": 0.12713138341279046,
+      "grad_norm": 0.26756346225738525,
+      "learning_rate": 0.0004989343577386565,
+      "loss": 7.7231,
+      "step": 370
+    },
+    {
+      "epoch": 0.12747498174633853,
+      "grad_norm": 0.2157527357339859,
+      "learning_rate": 0.0004989065193112208,
+      "loss": 7.7513,
+      "step": 371
+    },
+    {
+      "epoch": 0.1278185800798866,
+      "grad_norm": 0.4620709717273712,
+      "learning_rate": 0.0004988783227337746,
+      "loss": 7.739,
+      "step": 372
+    },
+    {
+      "epoch": 0.1281621784134347,
+      "grad_norm": 0.24664214253425598,
+      "learning_rate": 0.0004988497680468898,
+      "loss": 7.7792,
+      "step": 373
+    },
+    {
+      "epoch": 0.12850577674698277,
+      "grad_norm": 0.45489999651908875,
+      "learning_rate": 0.0004988208552916534,
+      "loss": 7.7012,
+      "step": 374
+    },
+    {
+      "epoch": 0.12884937508053085,
+      "grad_norm": 0.2705993950366974,
+      "learning_rate": 0.0004987915845096683,
+      "loss": 7.7137,
+      "step": 375
+    },
+    {
+      "epoch": 0.12919297341407895,
+      "grad_norm": 0.26571300625801086,
+      "learning_rate": 0.0004987619557430513,
+      "loss": 7.6652,
+      "step": 376
+    },
+    {
+      "epoch": 0.12953657174762703,
+      "grad_norm": 0.2846947908401489,
+      "learning_rate": 0.0004987319690344358,
+      "loss": 7.6621,
+      "step": 377
+    },
+    {
+      "epoch": 0.1298801700811751,
+      "grad_norm": 0.253897100687027,
+      "learning_rate": 0.000498701624426969,
+      "loss": 7.6912,
+      "step": 378
+    },
+    {
+      "epoch": 0.1302237684147232,
+      "grad_norm": 0.21716997027397156,
+      "learning_rate": 0.0004986709219643136,
+      "loss": 7.8242,
+      "step": 379
+    },
+    {
+      "epoch": 0.13056736674827127,
+      "grad_norm": 0.2383873611688614,
+      "learning_rate": 0.0004986398616906474,
+      "loss": 7.6825,
+      "step": 380
+    },
+    {
+      "epoch": 0.13091096508181935,
+      "grad_norm": 0.26393526792526245,
+      "learning_rate": 0.0004986084436506625,
+      "loss": 7.6705,
+      "step": 381
+    },
+    {
+      "epoch": 0.13125456341536743,
+      "grad_norm": 0.22068330645561218,
+      "learning_rate": 0.0004985766678895665,
+      "loss": 7.7126,
+      "step": 382
+    },
+    {
+      "epoch": 0.1315981617489155,
+      "grad_norm": 0.269553542137146,
+      "learning_rate": 0.000498544534453081,
+      "loss": 7.8269,
+      "step": 383
+    },
+    {
+      "epoch": 0.1319417600824636,
+      "grad_norm": 0.38264012336730957,
+      "learning_rate": 0.0004985120433874429,
+      "loss": 7.6628,
+      "step": 384
+    },
+    {
+      "epoch": 0.1322853584160117,
+      "grad_norm": 0.3014664053916931,
+      "learning_rate": 0.0004984791947394032,
+      "loss": 7.6818,
+      "step": 385
+    },
+    {
+      "epoch": 0.13262895674955977,
+      "grad_norm": 0.24440672993659973,
+      "learning_rate": 0.0004984459885562277,
+      "loss": 7.6293,
+      "step": 386
+    },
+    {
+      "epoch": 0.13297255508310785,
+      "grad_norm": 0.2836483418941498,
+      "learning_rate": 0.0004984124248856964,
+      "loss": 7.7407,
+      "step": 387
+    },
+    {
+      "epoch": 0.13331615341665592,
+      "grad_norm": 0.26046499609947205,
+      "learning_rate": 0.0004983785037761041,
+      "loss": 7.7502,
+      "step": 388
+    },
+    {
+      "epoch": 0.133659751750204,
+      "grad_norm": 0.4465503692626953,
+      "learning_rate": 0.0004983442252762595,
+      "loss": 7.7441,
+      "step": 389
+    },
+    {
+      "epoch": 0.13400335008375208,
+      "grad_norm": 0.4302878975868225,
+      "learning_rate": 0.0004983095894354857,
+      "loss": 7.7906,
+      "step": 390
+    },
+    {
+      "epoch": 0.1343469484173002,
+      "grad_norm": 0.5368163585662842,
+      "learning_rate": 0.0004982745963036201,
+      "loss": 7.7609,
+      "step": 391
+    },
+    {
+      "epoch": 0.13469054675084827,
+      "grad_norm": 0.3662016689777374,
+      "learning_rate": 0.0004982392459310141,
+      "loss": 7.6962,
+      "step": 392
+    },
+    {
+      "epoch": 0.13503414508439635,
+      "grad_norm": 0.27907463908195496,
+      "learning_rate": 0.000498203538368533,
+      "loss": 7.7156,
+      "step": 393
+    },
+    {
+      "epoch": 0.13537774341794442,
+      "grad_norm": 0.6360219120979309,
+      "learning_rate": 0.0004981674736675563,
+      "loss": 7.8432,
+      "step": 394
+    },
+    {
+      "epoch": 0.1357213417514925,
+      "grad_norm": 0.38870713114738464,
+      "learning_rate": 0.0004981310518799772,
+      "loss": 7.6269,
+      "step": 395
+    },
+    {
+      "epoch": 0.13606494008504058,
+      "grad_norm": 0.3981183171272278,
+      "learning_rate": 0.0004980942730582028,
+      "loss": 7.7014,
+      "step": 396
+    },
+    {
+      "epoch": 0.13640853841858866,
+      "grad_norm": 0.48505455255508423,
+      "learning_rate": 0.0004980571372551538,
+      "loss": 7.8247,
+      "step": 397
+    },
+    {
+      "epoch": 0.13675213675213677,
+      "grad_norm": 0.3913789391517639,
+      "learning_rate": 0.0004980196445242651,
+      "loss": 7.8572,
+      "step": 398
+    },
+    {
+      "epoch": 0.13709573508568484,
+      "grad_norm": 0.6935904026031494,
+      "learning_rate": 0.0004979817949194842,
+      "loss": 7.8824,
+      "step": 399
+    },
+    {
+      "epoch": 0.13743933341923292,
+      "grad_norm": 0.7566063404083252,
+      "learning_rate": 0.000497943588495273,
+      "loss": 7.7401,
+      "step": 400
+    },
+    {
+      "epoch": 0.137782931752781,
+      "grad_norm": 0.6485404968261719,
+      "learning_rate": 0.0004979050253066063,
+      "loss": 7.7275,
+      "step": 401
+    },
+    {
+      "epoch": 0.13812653008632908,
+      "grad_norm": 1.2425336837768555,
+      "learning_rate": 0.0004978661054089726,
+      "loss": 7.6111,
+      "step": 402
+    },
+    {
+      "epoch": 0.13847012841987716,
+      "grad_norm": 0.5411668419837952,
+      "learning_rate": 0.0004978268288583733,
+      "loss": 7.6819,
+      "step": 403
+    },
+    {
+      "epoch": 0.13881372675342524,
+      "grad_norm": 0.7507022619247437,
+      "learning_rate": 0.0004977871957113233,
+      "loss": 7.5208,
+      "step": 404
+    },
+    {
+      "epoch": 0.13915732508697332,
+      "grad_norm": 0.7834251523017883,
+      "learning_rate": 0.0004977472060248505,
+      "loss": 7.6503,
+      "step": 405
+    },
+    {
+      "epoch": 0.13950092342052142,
+      "grad_norm": 0.37674951553344727,
+      "learning_rate": 0.0004977068598564957,
+      "loss": 7.6368,
+      "step": 406
+    },
+    {
+      "epoch": 0.1398445217540695,
+      "grad_norm": 0.4405427575111389,
+      "learning_rate": 0.0004976661572643128,
+      "loss": 7.6453,
+      "step": 407
+    },
+    {
+      "epoch": 0.14018812008761758,
+      "grad_norm": 0.40070560574531555,
+      "learning_rate": 0.0004976250983068687,
+      "loss": 7.6718,
+      "step": 408
+    },
+    {
+      "epoch": 0.14053171842116566,
+      "grad_norm": 0.49566248059272766,
+      "learning_rate": 0.0004975836830432425,
+      "loss": 7.6901,
+      "step": 409
+    },
+    {
+      "epoch": 0.14087531675471374,
+      "grad_norm": 0.5421594381332397,
+      "learning_rate": 0.0004975419115330267,
+      "loss": 7.6493,
+      "step": 410
+    },
+    {
+      "epoch": 0.14121891508826181,
+      "grad_norm": 0.33900225162506104,
+      "learning_rate": 0.0004974997838363258,
+      "loss": 7.6087,
+      "step": 411
+    },
+    {
+      "epoch": 0.1415625134218099,
+      "grad_norm": 0.32626527547836304,
+      "learning_rate": 0.0004974573000137572,
+      "loss": 7.6064,
+      "step": 412
+    },
+    {
+      "epoch": 0.141906111755358,
+      "grad_norm": 0.25622686743736267,
+      "learning_rate": 0.0004974144601264507,
+      "loss": 7.6272,
+      "step": 413
+    },
+    {
+      "epoch": 0.14224971008890608,
+      "grad_norm": 0.26424112915992737,
+      "learning_rate": 0.0004973712642360481,
+      "loss": 7.6451,
+      "step": 414
+    },
+    {
+      "epoch": 0.14259330842245416,
+      "grad_norm": 0.3264583647251129,
+      "learning_rate": 0.0004973277124047039,
+      "loss": 7.6348,
+      "step": 415
+    },
+    {
+      "epoch": 0.14293690675600224,
+      "grad_norm": 0.5043911933898926,
+      "learning_rate": 0.0004972838046950844,
+      "loss": 7.6353,
+      "step": 416
+    },
+    {
+      "epoch": 0.14328050508955031,
+      "grad_norm": 0.4086906313896179,
+      "learning_rate": 0.0004972395411703682,
+      "loss": 7.5676,
+      "step": 417
+    },
+    {
+      "epoch": 0.1436241034230984,
+      "grad_norm": 0.550713300704956,
+      "learning_rate": 0.0004971949218942459,
+      "loss": 7.6814,
+      "step": 418
+    },
+    {
+      "epoch": 0.14396770175664647,
+      "grad_norm": 0.44711998105049133,
+      "learning_rate": 0.0004971499469309196,
+      "loss": 7.5975,
+      "step": 419
+    },
+    {
+      "epoch": 0.14431130009019455,
+      "grad_norm": 0.5666840076446533,
+      "learning_rate": 0.0004971046163451039,
+      "loss": 7.5721,
+      "step": 420
+    },
+    {
+      "epoch": 0.14465489842374266,
+      "grad_norm": 0.8350973725318909,
+      "learning_rate": 0.0004970589302020244,
+      "loss": 7.5351,
+      "step": 421
+    },
+    {
+      "epoch": 0.14499849675729073,
+      "grad_norm": 0.22808456420898438,
+      "learning_rate": 0.0004970128885674188,
+      "loss": 7.6153,
+      "step": 422
+    },
+    {
+      "epoch": 0.1453420950908388,
+      "grad_norm": 0.7204859852790833,
+      "learning_rate": 0.0004969664915075358,
+      "loss": 7.5626,
+      "step": 423
+    },
+    {
+      "epoch": 0.1456856934243869,
+      "grad_norm": 0.34824058413505554,
+      "learning_rate": 0.0004969197390891361,
+      "loss": 7.6858,
+      "step": 424
+    },
+    {
+      "epoch": 0.14602929175793497,
+      "grad_norm": 0.4266470968723297,
+      "learning_rate": 0.0004968726313794914,
+      "loss": 7.6733,
+      "step": 425
+    },
+    {
+      "epoch": 0.14637289009148305,
+      "grad_norm": 0.42429882287979126,
+      "learning_rate": 0.0004968251684463847,
+      "loss": 7.6719,
+      "step": 426
+    },
+    {
+      "epoch": 0.14671648842503113,
+      "grad_norm": 0.3132624626159668,
+      "learning_rate": 0.0004967773503581101,
+      "loss": 7.7054,
+      "step": 427
+    },
+    {
+      "epoch": 0.14706008675857923,
+      "grad_norm": 0.5525080561637878,
+      "learning_rate": 0.0004967291771834727,
+      "loss": 7.5684,
+      "step": 428
+    },
+    {
+      "epoch": 0.1474036850921273,
+      "grad_norm": 0.3956605792045593,
+      "learning_rate": 0.0004966806489917886,
+      "loss": 7.5842,
+      "step": 429
+    },
+    {
+      "epoch": 0.1477472834256754,
+      "grad_norm": 0.5083109140396118,
+      "learning_rate": 0.0004966317658528847,
+      "loss": 7.5118,
+      "step": 430
+    },
+    {
+      "epoch": 0.14809088175922347,
+      "grad_norm": 0.4539259076118469,
+      "learning_rate": 0.0004965825278370987,
+      "loss": 7.5676,
+      "step": 431
+    },
+    {
+      "epoch": 0.14843448009277155,
+      "grad_norm": 0.6377333998680115,
+      "learning_rate": 0.0004965329350152788,
+      "loss": 7.6241,
+      "step": 432
+    },
+    {
+      "epoch": 0.14877807842631963,
+      "grad_norm": 0.32773837447166443,
+      "learning_rate": 0.0004964829874587838,
+      "loss": 7.579,
+      "step": 433
+    },
+    {
+      "epoch": 0.1491216767598677,
+      "grad_norm": 0.4683372378349304,
+      "learning_rate": 0.0004964326852394829,
+      "loss": 7.567,
+      "step": 434
+    },
+    {
+      "epoch": 0.14946527509341578,
+      "grad_norm": 0.5232347249984741,
+      "learning_rate": 0.0004963820284297558,
+      "loss": 7.554,
+      "step": 435
+    },
+    {
+      "epoch": 0.1498088734269639,
+      "grad_norm": 0.6766787171363831,
+      "learning_rate": 0.0004963310171024921,
+      "loss": 7.6316,
+      "step": 436
+    },
+    {
+      "epoch": 0.15015247176051197,
+      "grad_norm": 0.33900943398475647,
+      "learning_rate": 0.0004962796513310917,
+      "loss": 7.6034,
+      "step": 437
+    },
+    {
+      "epoch": 0.15049607009406005,
+      "grad_norm": 0.4721566140651703,
+      "learning_rate": 0.0004962279311894644,
+      "loss": 7.6699,
+      "step": 438
+    },
+    {
+      "epoch": 0.15083966842760813,
+      "grad_norm": 0.399872750043869,
+      "learning_rate": 0.0004961758567520302,
+      "loss": 7.6539,
+      "step": 439
+    },
+    {
+      "epoch": 0.1511832667611562,
+      "grad_norm": 0.4169398248195648,
+      "learning_rate": 0.0004961234280937185,
+      "loss": 7.6788,
+      "step": 440
+    },
+    {
+      "epoch": 0.15152686509470428,
+      "grad_norm": 0.4548245668411255,
+      "learning_rate": 0.0004960706452899687,
+      "loss": 7.6121,
+      "step": 441
+    },
+    {
+      "epoch": 0.15187046342825236,
+      "grad_norm": 0.7090252041816711,
+      "learning_rate": 0.0004960175084167296,
+      "loss": 7.5421,
+      "step": 442
+    },
+    {
+      "epoch": 0.15221406176180047,
+      "grad_norm": 0.431866854429245,
+      "learning_rate": 0.0004959640175504593,
+      "loss": 7.5093,
+      "step": 443
+    },
+    {
+      "epoch": 0.15255766009534855,
+      "grad_norm": 0.40838518738746643,
+      "learning_rate": 0.0004959101727681258,
+      "loss": 7.6853,
+      "step": 444
+    },
+    {
+      "epoch": 0.15290125842889662,
+      "grad_norm": 0.9189926981925964,
+      "learning_rate": 0.0004958559741472058,
+      "loss": 7.6081,
+      "step": 445
+    },
+    {
+      "epoch": 0.1532448567624447,
+      "grad_norm": 0.522569477558136,
+      "learning_rate": 0.0004958014217656855,
+      "loss": 7.5748,
+      "step": 446
+    },
+    {
+      "epoch": 0.15358845509599278,
+      "grad_norm": 0.9465051889419556,
+      "learning_rate": 0.0004957465157020598,
+      "loss": 7.6038,
+      "step": 447
+    },
+    {
+      "epoch": 0.15393205342954086,
+      "grad_norm": 0.7611620426177979,
+      "learning_rate": 0.0004956912560353327,
+      "loss": 7.6636,
+      "step": 448
+    },
+    {
+      "epoch": 0.15427565176308894,
+      "grad_norm": 0.6408162117004395,
+      "learning_rate": 0.0004956356428450171,
+      "loss": 7.6456,
+      "step": 449
+    },
+    {
+      "epoch": 0.15461925009663702,
+      "grad_norm": 0.8566272854804993,
+      "learning_rate": 0.0004955796762111345,
+      "loss": 7.6224,
+      "step": 450
+    },
+    {
+      "epoch": 0.15496284843018512,
+      "grad_norm": 0.9389132857322693,
+      "learning_rate": 0.0004955233562142148,
+      "loss": 7.5777,
+      "step": 451
+    },
+    {
+      "epoch": 0.1553064467637332,
+      "grad_norm": 1.3984068632125854,
+      "learning_rate": 0.0004954666829352966,
+      "loss": 7.4553,
+      "step": 452
+    },
+    {
+      "epoch": 0.15565004509728128,
+      "grad_norm": 0.6359997987747192,
+      "learning_rate": 0.0004954096564559267,
+      "loss": 7.5531,
+      "step": 453
+    },
+    {
+      "epoch": 0.15599364343082936,
+      "grad_norm": 1.2158923149108887,
+      "learning_rate": 0.00049535227685816,
+      "loss": 7.519,
+      "step": 454
+    },
+    {
+      "epoch": 0.15633724176437744,
+      "grad_norm": 0.7679694890975952,
+      "learning_rate": 0.0004952945442245598,
+      "loss": 7.5488,
+      "step": 455
+    },
+    {
+      "epoch": 0.15668084009792552,
+      "grad_norm": 1.019667387008667,
+      "learning_rate": 0.0004952364586381971,
+      "loss": 7.5456,
+      "step": 456
+    },
+    {
+      "epoch": 0.1570244384314736,
+      "grad_norm": 0.9500595927238464,
+      "learning_rate": 0.000495178020182651,
+      "loss": 7.4402,
+      "step": 457
+    },
+    {
+      "epoch": 0.1573680367650217,
+      "grad_norm": 0.764927089214325,
+      "learning_rate": 0.0004951192289420082,
+      "loss": 7.5489,
+      "step": 458
+    },
+    {
+      "epoch": 0.15771163509856978,
+      "grad_norm": 0.689294159412384,
+      "learning_rate": 0.0004950600850008629,
+      "loss": 7.3963,
+      "step": 459
+    },
+    {
+      "epoch": 0.15805523343211786,
+      "grad_norm": 0.9121485948562622,
+      "learning_rate": 0.0004950005884443171,
+      "loss": 7.3951,
+      "step": 460
+    },
+    {
+      "epoch": 0.15839883176566594,
+      "grad_norm": 0.3724419176578522,
+      "learning_rate": 0.00049494073935798,
+      "loss": 7.52,
+      "step": 461
+    },
+    {
+      "epoch": 0.15874243009921402,
+      "grad_norm": 1.2896360158920288,
+      "learning_rate": 0.0004948805378279681,
+      "loss": 7.5127,
+      "step": 462
+    },
+    {
+      "epoch": 0.1590860284327621,
+      "grad_norm": 0.5851932168006897,
+      "learning_rate": 0.0004948199839409047,
+      "loss": 7.4566,
+      "step": 463
+    },
+    {
+      "epoch": 0.15942962676631017,
+      "grad_norm": 0.8496333360671997,
+      "learning_rate": 0.0004947590777839209,
+      "loss": 7.4376,
+      "step": 464
+    },
+    {
+      "epoch": 0.15977322509985828,
+      "grad_norm": 0.40505436062812805,
+      "learning_rate": 0.0004946978194446538,
+      "loss": 7.5951,
+      "step": 465
+    },
+    {
+      "epoch": 0.16011682343340636,
+      "grad_norm": 0.7311084866523743,
+      "learning_rate": 0.0004946362090112479,
+      "loss": 7.5839,
+      "step": 466
+    },
+    {
+      "epoch": 0.16046042176695444,
+      "grad_norm": 0.3588482737541199,
+      "learning_rate": 0.0004945742465723537,
+      "loss": 7.4579,
+      "step": 467
+    },
+    {
+      "epoch": 0.16080402010050251,
+      "grad_norm": 0.5303255319595337,
+      "learning_rate": 0.000494511932217129,
+      "loss": 7.4116,
+      "step": 468
+    },
+    {
+      "epoch": 0.1611476184340506,
+      "grad_norm": 0.5300981998443604,
+      "learning_rate": 0.0004944492660352371,
+      "loss": 7.3889,
+      "step": 469
+    },
+    {
+      "epoch": 0.16149121676759867,
+      "grad_norm": 0.7826660871505737,
+      "learning_rate": 0.0004943862481168483,
+      "loss": 7.4045,
+      "step": 470
+    },
+    {
+      "epoch": 0.16183481510114675,
+      "grad_norm": 0.5411097407341003,
+      "learning_rate": 0.0004943228785526386,
+      "loss": 7.4755,
+      "step": 471
+    },
+    {
+      "epoch": 0.16217841343469483,
+      "grad_norm": 0.4609074294567108,
+      "learning_rate": 0.00049425915743379,
+      "loss": 7.3915,
+      "step": 472
+    },
+    {
+      "epoch": 0.16252201176824294,
+      "grad_norm": 0.6618660688400269,
+      "learning_rate": 0.0004941950848519904,
+      "loss": 7.494,
+      "step": 473
+    },
+    {
+      "epoch": 0.16286561010179101,
+      "grad_norm": 0.4540770351886749,
+      "learning_rate": 0.0004941306608994336,
+      "loss": 7.4365,
+      "step": 474
+    },
+    {
+      "epoch": 0.1632092084353391,
+      "grad_norm": 0.29459720849990845,
+      "learning_rate": 0.0004940658856688185,
+      "loss": 7.4786,
+      "step": 475
+    },
+    {
+      "epoch": 0.16355280676888717,
+      "grad_norm": 0.38014736771583557,
+      "learning_rate": 0.00049400075925335,
+      "loss": 7.5007,
+      "step": 476
+    },
+    {
+      "epoch": 0.16389640510243525,
+      "grad_norm": 0.3502187728881836,
+      "learning_rate": 0.0004939352817467382,
+      "loss": 7.3697,
+      "step": 477
+    },
+    {
+      "epoch": 0.16424000343598333,
+      "grad_norm": 0.539600133895874,
+      "learning_rate": 0.0004938694532431979,
+      "loss": 7.4328,
+      "step": 478
+    },
+    {
+      "epoch": 0.1645836017695314,
+      "grad_norm": 0.25308141112327576,
+      "learning_rate": 0.0004938032738374497,
+      "loss": 7.524,
+      "step": 479
+    },
+    {
+      "epoch": 0.1649272001030795,
+      "grad_norm": 0.5650960206985474,
+      "learning_rate": 0.0004937367436247186,
+      "loss": 7.5126,
+      "step": 480
+    },
+    {
+      "epoch": 0.1652707984366276,
+      "grad_norm": 0.27598950266838074,
+      "learning_rate": 0.0004936698627007343,
+      "loss": 7.4391,
+      "step": 481
+    },
+    {
+      "epoch": 0.16561439677017567,
+      "grad_norm": 0.4361298382282257,
+      "learning_rate": 0.0004936026311617315,
+      "loss": 7.5193,
+      "step": 482
+    },
+    {
+      "epoch": 0.16595799510372375,
+      "grad_norm": 0.3199003040790558,
+      "learning_rate": 0.0004935350491044493,
+      "loss": 7.4871,
+      "step": 483
+    },
+    {
+      "epoch": 0.16630159343727183,
+      "grad_norm": 0.4775875210762024,
+      "learning_rate": 0.000493467116626131,
+      "loss": 7.4474,
+      "step": 484
+    },
+    {
+      "epoch": 0.1666451917708199,
+      "grad_norm": 0.6670292019844055,
+      "learning_rate": 0.0004933988338245242,
+      "loss": 7.4594,
+      "step": 485
+    },
+    {
+      "epoch": 0.16698879010436798,
+      "grad_norm": 0.3038119375705719,
+      "learning_rate": 0.0004933302007978807,
+      "loss": 7.4544,
+      "step": 486
+    },
+    {
+      "epoch": 0.16733238843791606,
+      "grad_norm": 0.6222366690635681,
+      "learning_rate": 0.0004932612176449559,
+      "loss": 7.4342,
+      "step": 487
+    },
+    {
+      "epoch": 0.16767598677146417,
+      "grad_norm": 0.3004700541496277,
+      "learning_rate": 0.0004931918844650095,
+      "loss": 7.4134,
+      "step": 488
+    },
+    {
+      "epoch": 0.16801958510501225,
+      "grad_norm": 0.44555893540382385,
+      "learning_rate": 0.0004931222013578045,
+      "loss": 7.4252,
+      "step": 489
+    },
+    {
+      "epoch": 0.16836318343856033,
+      "grad_norm": 0.3072405159473419,
+      "learning_rate": 0.0004930521684236073,
+      "loss": 7.4277,
+      "step": 490
+    },
+    {
+      "epoch": 0.1687067817721084,
+      "grad_norm": 0.8164499998092651,
+      "learning_rate": 0.000492981785763188,
+      "loss": 7.4174,
+      "step": 491
+    },
+    {
+      "epoch": 0.16905038010565648,
+      "grad_norm": 0.6826271414756775,
+      "learning_rate": 0.0004929110534778197,
+      "loss": 7.4227,
+      "step": 492
+    },
+    {
+      "epoch": 0.16939397843920456,
+      "grad_norm": 0.6980798244476318,
+      "learning_rate": 0.0004928399716692787,
+      "loss": 7.3474,
+      "step": 493
+    },
+    {
+      "epoch": 0.16973757677275264,
+      "grad_norm": 0.6725314855575562,
+      "learning_rate": 0.0004927685404398441,
+      "loss": 7.5103,
+      "step": 494
+    },
+    {
+      "epoch": 0.17008117510630075,
+      "grad_norm": 0.38431742787361145,
+      "learning_rate": 0.000492696759892298,
+      "loss": 7.4468,
+      "step": 495
+    },
+    {
+      "epoch": 0.17042477343984883,
+      "grad_norm": 0.41020020842552185,
+      "learning_rate": 0.0004926246301299247,
+      "loss": 7.48,
+      "step": 496
+    },
+    {
+      "epoch": 0.1707683717733969,
+      "grad_norm": 0.5806956887245178,
+      "learning_rate": 0.0004925521512565114,
+      "loss": 7.4053,
+      "step": 497
+    },
+    {
+      "epoch": 0.17111197010694498,
+      "grad_norm": 0.7412662506103516,
+      "learning_rate": 0.0004924793233763476,
+      "loss": 7.329,
+      "step": 498
+    },
+    {
+      "epoch": 0.17145556844049306,
+      "grad_norm": 0.5328527092933655,
+      "learning_rate": 0.0004924061465942247,
+      "loss": 7.4516,
+      "step": 499
+    },
+    {
+      "epoch": 0.17179916677404114,
+      "grad_norm": 1.086364507675171,
+      "learning_rate": 0.0004923326210154364,
+      "loss": 7.6593,
+      "step": 500
+    },
+    {
+      "epoch": 0.17214276510758922,
+      "grad_norm": 0.6880685091018677,
+      "learning_rate": 0.0004922587467457781,
+      "loss": 7.3688,
+      "step": 501
+    },
+    {
+      "epoch": 0.1724863634411373,
+      "grad_norm": 1.1379863023757935,
+      "learning_rate": 0.0004921845238915472,
+      "loss": 7.3837,
+      "step": 502
+    },
+    {
+      "epoch": 0.1728299617746854,
+      "grad_norm": 0.8507835268974304,
+      "learning_rate": 0.0004921099525595423,
+      "loss": 7.339,
+      "step": 503
+    },
+    {
+      "epoch": 0.17317356010823348,
+      "grad_norm": 1.0241001844406128,
+      "learning_rate": 0.0004920350328570638,
+      "loss": 7.2608,
+      "step": 504
+    },
+    {
+      "epoch": 0.17351715844178156,
+      "grad_norm": 0.6690787672996521,
+      "learning_rate": 0.000491959764891913,
+      "loss": 7.3628,
+      "step": 505
+    },
+    {
+      "epoch": 0.17386075677532964,
+      "grad_norm": 0.8972269892692566,
+      "learning_rate": 0.0004918841487723926,
+      "loss": 7.2917,
+      "step": 506
+    },
+    {
+      "epoch": 0.17420435510887772,
+      "grad_norm": 0.33962926268577576,
+      "learning_rate": 0.0004918081846073059,
+      "loss": 7.4265,
+      "step": 507
+    },
+    {
+      "epoch": 0.1745479534424258,
+      "grad_norm": 0.9010359048843384,
+      "learning_rate": 0.0004917318725059577,
+      "loss": 7.3324,
+      "step": 508
+    },
+    {
+      "epoch": 0.17489155177597387,
+      "grad_norm": 0.42200517654418945,
+      "learning_rate": 0.0004916552125781528,
+      "loss": 7.3756,
+      "step": 509
+    },
+    {
+      "epoch": 0.17523515010952198,
+      "grad_norm": 0.8156759738922119,
+      "learning_rate": 0.0004915782049341967,
+      "loss": 7.2236,
+      "step": 510
+    },
+    {
+      "epoch": 0.17557874844307006,
+      "grad_norm": 0.3509763181209564,
+      "learning_rate": 0.0004915008496848951,
+      "loss": 7.2068,
+      "step": 511
+    },
+    {
+      "epoch": 0.17592234677661814,
+      "grad_norm": 0.7369383573532104,
+      "learning_rate": 0.000491423146941554,
+      "loss": 7.2954,
+      "step": 512
+    },
+    {
+      "epoch": 0.17626594511016622,
+      "grad_norm": 0.5369434356689453,
+      "learning_rate": 0.0004913450968159794,
+      "loss": 7.1138,
+      "step": 513
+    },
+    {
+      "epoch": 0.1766095434437143,
+      "grad_norm": 0.64871746301651,
+      "learning_rate": 0.0004912666994204773,
+      "loss": 7.2767,
+      "step": 514
+    },
+    {
+      "epoch": 0.17695314177726237,
+      "grad_norm": 0.5261757969856262,
+      "learning_rate": 0.0004911879548678531,
+      "loss": 7.2552,
+      "step": 515
+    },
+    {
+      "epoch": 0.17729674011081045,
+      "grad_norm": 0.8654578328132629,
+      "learning_rate": 0.0004911088632714117,
+      "loss": 7.4006,
+      "step": 516
+    },
+    {
+      "epoch": 0.17764033844435856,
+      "grad_norm": 0.7231168746948242,
+      "learning_rate": 0.0004910294247449576,
+      "loss": 7.2783,
+      "step": 517
+    },
+    {
+      "epoch": 0.17798393677790664,
+      "grad_norm": 0.6039850115776062,
+      "learning_rate": 0.0004909496394027945,
+      "loss": 7.3449,
+      "step": 518
+    },
+    {
+      "epoch": 0.17832753511145472,
+      "grad_norm": 0.5625784993171692,
+      "learning_rate": 0.0004908695073597249,
+      "loss": 7.1675,
+      "step": 519
+    },
+    {
+      "epoch": 0.1786711334450028,
+      "grad_norm": 0.39175331592559814,
+      "learning_rate": 0.0004907890287310504,
+      "loss": 7.3169,
+      "step": 520
+    },
+    {
+      "epoch": 0.17901473177855087,
+      "grad_norm": 0.6726620197296143,
+      "learning_rate": 0.000490708203632571,
+      "loss": 7.3171,
+      "step": 521
+    },
+    {
+      "epoch": 0.17935833011209895,
+      "grad_norm": 0.36819761991500854,
+      "learning_rate": 0.0004906270321805854,
+      "loss": 7.2229,
+      "step": 522
+    },
+    {
+      "epoch": 0.17970192844564703,
+      "grad_norm": 0.8770174384117126,
+      "learning_rate": 0.000490545514491891,
+      "loss": 7.2737,
+      "step": 523
+    },
+    {
+      "epoch": 0.1800455267791951,
+      "grad_norm": 0.3748573660850525,
+      "learning_rate": 0.0004904636506837828,
+      "loss": 7.2918,
+      "step": 524
+    },
+    {
+      "epoch": 0.18038912511274321,
+      "grad_norm": 1.1249299049377441,
+      "learning_rate": 0.0004903814408740543,
+      "loss": 7.2435,
+      "step": 525
+    },
+    {
+      "epoch": 0.1807327234462913,
+      "grad_norm": 0.3472743034362793,
+      "learning_rate": 0.0004902988851809965,
+      "loss": 7.3431,
+      "step": 526
+    },
+    {
+      "epoch": 0.18107632177983937,
+      "grad_norm": 0.8709349632263184,
+      "learning_rate": 0.0004902159837233984,
+      "loss": 7.3518,
+      "step": 527
+    },
+    {
+      "epoch": 0.18141992011338745,
+      "grad_norm": 0.818452775478363,
+      "learning_rate": 0.0004901327366205464,
+      "loss": 7.353,
+      "step": 528
+    },
+    {
+      "epoch": 0.18176351844693553,
+      "grad_norm": 0.6798619031906128,
+      "learning_rate": 0.000490049143992224,
+      "loss": 7.2756,
+      "step": 529
+    },
+    {
+      "epoch": 0.1821071167804836,
+      "grad_norm": 0.7426633238792419,
+      "learning_rate": 0.0004899652059587123,
+      "loss": 7.2484,
+      "step": 530
+    },
+    {
+      "epoch": 0.1824507151140317,
+      "grad_norm": 0.5532476305961609,
+      "learning_rate": 0.0004898809226407892,
+      "loss": 7.3746,
+      "step": 531
+    },
+    {
+      "epoch": 0.1827943134475798,
+      "grad_norm": 0.7635505795478821,
+      "learning_rate": 0.0004897962941597294,
+      "loss": 7.2973,
+      "step": 532
+    },
+    {
+      "epoch": 0.18313791178112787,
+      "grad_norm": 0.6389451622962952,
+      "learning_rate": 0.0004897113206373042,
+      "loss": 7.2043,
+      "step": 533
+    },
+    {
+      "epoch": 0.18348151011467595,
+      "grad_norm": 0.7844037413597107,
+      "learning_rate": 0.0004896260021957816,
+      "loss": 7.2125,
+      "step": 534
+    },
+    {
+      "epoch": 0.18382510844822403,
+      "grad_norm": 0.4252033829689026,
+      "learning_rate": 0.0004895403389579258,
+      "loss": 7.3679,
+      "step": 535
+    },
+    {
+      "epoch": 0.1841687067817721,
+      "grad_norm": 0.8579635620117188,
+      "learning_rate": 0.0004894543310469967,
+      "loss": 7.2665,
+      "step": 536
+    },
+    {
+      "epoch": 0.18451230511532019,
+      "grad_norm": 0.4492959678173065,
+      "learning_rate": 0.0004893679785867511,
+      "loss": 7.262,
+      "step": 537
+    },
+    {
+      "epoch": 0.18485590344886826,
+      "grad_norm": 0.6443487405776978,
+      "learning_rate": 0.0004892812817014407,
+      "loss": 7.3039,
+      "step": 538
+    },
+    {
+      "epoch": 0.18519950178241634,
+      "grad_norm": 0.35954129695892334,
+      "learning_rate": 0.000489194240515813,
+      "loss": 7.28,
+      "step": 539
+    },
+    {
+      "epoch": 0.18554310011596445,
+      "grad_norm": 0.5164822340011597,
+      "learning_rate": 0.0004891068551551112,
+      "loss": 7.4009,
+      "step": 540
+    },
+    {
+      "epoch": 0.18588669844951253,
+      "grad_norm": 0.38150545954704285,
+      "learning_rate": 0.0004890191257450736,
+      "loss": 7.3082,
+      "step": 541
+    },
+    {
+      "epoch": 0.1862302967830606,
+      "grad_norm": 0.7000799179077148,
+      "learning_rate": 0.0004889310524119331,
+      "loss": 7.3069,
+      "step": 542
+    },
+    {
+      "epoch": 0.18657389511660868,
+      "grad_norm": 0.45130738615989685,
+      "learning_rate": 0.0004888426352824184,
+      "loss": 7.3513,
+      "step": 543
+    },
+    {
+      "epoch": 0.18691749345015676,
+      "grad_norm": 0.7859145402908325,
+      "learning_rate": 0.000488753874483752,
+      "loss": 7.3586,
+      "step": 544
+    },
+    {
+      "epoch": 0.18726109178370484,
+      "grad_norm": 0.426577091217041,
+      "learning_rate": 0.0004886647701436513,
+      "loss": 7.3083,
+      "step": 545
+    },
+    {
+      "epoch": 0.18760469011725292,
+      "grad_norm": 0.41109466552734375,
+      "learning_rate": 0.0004885753223903281,
+      "loss": 7.3997,
+      "step": 546
+    },
+    {
+      "epoch": 0.18794828845080103,
+      "grad_norm": 0.552748441696167,
+      "learning_rate": 0.0004884855313524879,
+      "loss": 7.3883,
+      "step": 547
+    },
+    {
+      "epoch": 0.1882918867843491,
+      "grad_norm": 0.7718663811683655,
+      "learning_rate": 0.0004883953971593308,
+      "loss": 7.3032,
+      "step": 548
+    },
+    {
+      "epoch": 0.18863548511789718,
+      "grad_norm": 0.4850485622882843,
+      "learning_rate": 0.0004883049199405501,
+      "loss": 7.3862,
+      "step": 549
+    },
+    {
+      "epoch": 0.18897908345144526,
+      "grad_norm": 0.8388044238090515,
+      "learning_rate": 0.0004882140998263331,
+      "loss": 7.4342,
+      "step": 550
+    },
+    {
+      "epoch": 0.18932268178499334,
+      "grad_norm": 1.0273692607879639,
+      "learning_rate": 0.0004881229369473601,
+      "loss": 7.1913,
+      "step": 551
+    },
+    {
+      "epoch": 0.18966628011854142,
+      "grad_norm": 0.8978946208953857,
+      "learning_rate": 0.0004880314314348048,
+      "loss": 7.2436,
+      "step": 552
+    },
+    {
+      "epoch": 0.1900098784520895,
+      "grad_norm": 0.9823456406593323,
+      "learning_rate": 0.000487939583420334,
+      "loss": 7.2297,
+      "step": 553
+    },
+    {
+      "epoch": 0.19035347678563758,
+      "grad_norm": 1.137190580368042,
+      "learning_rate": 0.00048784739303610715,
+      "loss": 7.2173,
+      "step": 554
+    },
+    {
+      "epoch": 0.19069707511918568,
+      "grad_norm": 0.842155396938324,
+      "learning_rate": 0.00048775486041477645,
+      "loss": 7.1943,
+      "step": 555
+    },
+    {
+      "epoch": 0.19104067345273376,
+      "grad_norm": 0.9378860592842102,
+      "learning_rate": 0.0004876619856894864,
+      "loss": 7.1498,
+      "step": 556
+    },
+    {
+      "epoch": 0.19138427178628184,
+      "grad_norm": 0.8283404111862183,
+      "learning_rate": 0.000487568768993874,
+      "loss": 7.1507,
+      "step": 557
+    },
+    {
+      "epoch": 0.19172787011982992,
+      "grad_norm": 0.8430905342102051,
+      "learning_rate": 0.0004874752104620681,
+      "loss": 7.2152,
+      "step": 558
+    },
+    {
+      "epoch": 0.192071468453378,
+      "grad_norm": 0.5364805459976196,
+      "learning_rate": 0.00048738131022868947,
+      "loss": 7.2251,
+      "step": 559
+    },
+    {
+      "epoch": 0.19241506678692608,
+      "grad_norm": 0.550553023815155,
+      "learning_rate": 0.0004872870684288505,
+      "loss": 7.2291,
+      "step": 560
+    },
+    {
+      "epoch": 0.19275866512047415,
+      "grad_norm": 0.5642486810684204,
+      "learning_rate": 0.0004871924851981553,
+      "loss": 7.2261,
+      "step": 561
+    },
+    {
+      "epoch": 0.19310226345402226,
+      "grad_norm": 0.4002983868122101,
+      "learning_rate": 0.00048709756067269884,
+      "loss": 7.0912,
+      "step": 562
+    },
+    {
+      "epoch": 0.19344586178757034,
+      "grad_norm": 0.5690410137176514,
+      "learning_rate": 0.0004870022949890676,
+      "loss": 7.1634,
+      "step": 563
+    },
+    {
+      "epoch": 0.19378946012111842,
+      "grad_norm": 0.48423007130622864,
+      "learning_rate": 0.0004869066882843387,
+      "loss": 7.1851,
+      "step": 564
+    },
+    {
+      "epoch": 0.1941330584546665,
+      "grad_norm": 0.7870746850967407,
+      "learning_rate": 0.00048681074069608006,
+      "loss": 7.0857,
+      "step": 565
+    },
+    {
+      "epoch": 0.19447665678821457,
+      "grad_norm": 0.3157571852207184,
+      "learning_rate": 0.00048671445236234996,
+      "loss": 7.2261,
+      "step": 566
+    },
+    {
+      "epoch": 0.19482025512176265,
+      "grad_norm": 0.7632848024368286,
+      "learning_rate": 0.00048661782342169715,
+      "loss": 7.2091,
+      "step": 567
+    },
+    {
+      "epoch": 0.19516385345531073,
+      "grad_norm": 0.4711115062236786,
+      "learning_rate": 0.00048652085401316037,
+      "loss": 7.2743,
+      "step": 568
+    },
+    {
+      "epoch": 0.1955074517888588,
+      "grad_norm": 0.5823638439178467,
+      "learning_rate": 0.00048642354427626836,
+      "loss": 7.204,
+      "step": 569
+    },
+    {
+      "epoch": 0.19585105012240692,
+      "grad_norm": 0.6164575815200806,
+      "learning_rate": 0.00048632589435103937,
+      "loss": 7.1551,
+      "step": 570
+    },
+    {
+      "epoch": 0.196194648455955,
+      "grad_norm": 0.4023503363132477,
+      "learning_rate": 0.0004862279043779813,
+      "loss": 7.1331,
+      "step": 571
+    },
+    {
+      "epoch": 0.19653824678950307,
+      "grad_norm": 0.6359233856201172,
+      "learning_rate": 0.00048612957449809137,
+      "loss": 7.0852,
+      "step": 572
+    },
+    {
+      "epoch": 0.19688184512305115,
+      "grad_norm": 0.5544841289520264,
+      "learning_rate": 0.00048603090485285565,
+      "loss": 7.171,
+      "step": 573
+    },
+    {
+      "epoch": 0.19722544345659923,
+      "grad_norm": 0.6806702613830566,
+      "learning_rate": 0.00048593189558424944,
+      "loss": 7.154,
+      "step": 574
+    },
+    {
+      "epoch": 0.1975690417901473,
+      "grad_norm": 0.5554197430610657,
+      "learning_rate": 0.00048583254683473657,
+      "loss": 7.1088,
+      "step": 575
+    },
+    {
+      "epoch": 0.1979126401236954,
+      "grad_norm": 0.5139564871788025,
+      "learning_rate": 0.0004857328587472691,
+      "loss": 7.1426,
+      "step": 576
+    },
+    {
+      "epoch": 0.1982562384572435,
+      "grad_norm": 0.7411857843399048,
+      "learning_rate": 0.00048563283146528774,
+      "loss": 7.227,
+      "step": 577
+    },
+    {
+      "epoch": 0.19859983679079157,
+      "grad_norm": 0.533267617225647,
+      "learning_rate": 0.00048553246513272113,
+      "loss": 7.1424,
+      "step": 578
+    },
+    {
+      "epoch": 0.19894343512433965,
+      "grad_norm": 0.5299692749977112,
+      "learning_rate": 0.0004854317598939857,
+      "loss": 7.0561,
+      "step": 579
+    },
+    {
+      "epoch": 0.19928703345788773,
+      "grad_norm": 0.6946874856948853,
+      "learning_rate": 0.00048533071589398566,
+      "loss": 7.0598,
+      "step": 580
+    },
+    {
+      "epoch": 0.1996306317914358,
+      "grad_norm": 0.42671942710876465,
+      "learning_rate": 0.00048522933327811246,
+      "loss": 7.1416,
+      "step": 581
+    },
+    {
+      "epoch": 0.1999742301249839,
+      "grad_norm": 0.4139076769351959,
+      "learning_rate": 0.00048512761219224494,
+      "loss": 7.1578,
+      "step": 582
+    },
+    {
+      "epoch": 0.20031782845853197,
+      "grad_norm": 0.40493038296699524,
+      "learning_rate": 0.000485025552782749,
+      "loss": 7.1914,
+      "step": 583
+    },
+    {
+      "epoch": 0.20066142679208007,
+      "grad_norm": 0.4322982430458069,
+      "learning_rate": 0.0004849231551964771,
+      "loss": 7.1429,
+      "step": 584
+    },
+    {
+      "epoch": 0.20100502512562815,
+      "grad_norm": 0.5229026079177856,
+      "learning_rate": 0.0004848204195807687,
+      "loss": 7.1781,
+      "step": 585
+    },
+    {
+      "epoch": 0.20134862345917623,
+      "grad_norm": 0.3657551407814026,
+      "learning_rate": 0.00048471734608344924,
+      "loss": 7.1775,
+      "step": 586
+    },
+    {
+      "epoch": 0.2016922217927243,
+      "grad_norm": 0.6244291067123413,
+      "learning_rate": 0.0004846139348528307,
+      "loss": 7.234,
+      "step": 587
+    },
+    {
+      "epoch": 0.2020358201262724,
+      "grad_norm": 0.4564877450466156,
+      "learning_rate": 0.00048451018603771064,
+      "loss": 7.092,
+      "step": 588
+    },
+    {
+      "epoch": 0.20237941845982046,
+      "grad_norm": 0.3984292447566986,
+      "learning_rate": 0.0004844060997873727,
+      "loss": 7.2266,
+      "step": 589
+    },
+    {
+      "epoch": 0.20272301679336854,
+      "grad_norm": 0.4138945937156677,
+      "learning_rate": 0.00048430167625158595,
+      "loss": 7.1706,
+      "step": 590
+    },
+    {
+      "epoch": 0.20306661512691662,
+      "grad_norm": 0.5082581043243408,
+      "learning_rate": 0.00048419691558060466,
+      "loss": 7.0813,
+      "step": 591
+    },
+    {
+      "epoch": 0.20341021346046473,
+      "grad_norm": 0.48261651396751404,
+      "learning_rate": 0.0004840918179251683,
+      "loss": 7.1004,
+      "step": 592
+    },
+    {
+      "epoch": 0.2037538117940128,
+      "grad_norm": 0.4194554388523102,
+      "learning_rate": 0.0004839863834365013,
+      "loss": 7.0886,
+      "step": 593
+    },
+    {
+      "epoch": 0.20409741012756089,
+      "grad_norm": 0.5701256394386292,
+      "learning_rate": 0.00048388061226631264,
+      "loss": 7.188,
+      "step": 594
+    },
+    {
+      "epoch": 0.20444100846110896,
+      "grad_norm": 0.37624257802963257,
+      "learning_rate": 0.0004837745045667957,
+      "loss": 7.1977,
+      "step": 595
+    },
+    {
+      "epoch": 0.20478460679465704,
+      "grad_norm": 0.4813244640827179,
+      "learning_rate": 0.0004836680604906284,
+      "loss": 7.1103,
+      "step": 596
+    },
+    {
+      "epoch": 0.20512820512820512,
+      "grad_norm": 0.7968626022338867,
+      "learning_rate": 0.00048356128019097223,
+      "loss": 7.1366,
+      "step": 597
+    },
+    {
+      "epoch": 0.2054718034617532,
+      "grad_norm": 0.6953873038291931,
+      "learning_rate": 0.0004834541638214727,
+      "loss": 7.3475,
+      "step": 598
+    },
+    {
+      "epoch": 0.2058154017953013,
+      "grad_norm": 0.7357029318809509,
+      "learning_rate": 0.00048334671153625895,
+      "loss": 7.1789,
+      "step": 599
+    },
+    {
+      "epoch": 0.20615900012884938,
+      "grad_norm": 0.9581784605979919,
+      "learning_rate": 0.00048323892348994335,
+      "loss": 7.2079,
+      "step": 600
+    },
+    {
+      "epoch": 0.20650259846239746,
+      "grad_norm": 0.8259401917457581,
+      "learning_rate": 0.00048313079983762155,
+      "loss": 7.1046,
+      "step": 601
+    },
+    {
+      "epoch": 0.20684619679594554,
+      "grad_norm": 0.6607183814048767,
+      "learning_rate": 0.00048302234073487185,
+      "loss": 7.0981,
+      "step": 602
+    },
+    {
+      "epoch": 0.20718979512949362,
+      "grad_norm": 0.5796660780906677,
+      "learning_rate": 0.0004829135463377553,
+      "loss": 7.1793,
+      "step": 603
+    },
+    {
+      "epoch": 0.2075333934630417,
+      "grad_norm": 0.7962479591369629,
+      "learning_rate": 0.00048280441680281566,
+      "loss": 7.0865,
+      "step": 604
+    },
+    {
+      "epoch": 0.20787699179658978,
+      "grad_norm": 0.48541417717933655,
+      "learning_rate": 0.0004826949522870786,
+      "loss": 7.0426,
+      "step": 605
+    },
+    {
+      "epoch": 0.20822059013013786,
+      "grad_norm": 0.4835425913333893,
+      "learning_rate": 0.00048258515294805207,
+      "loss": 7.0629,
+      "step": 606
+    },
+    {
+      "epoch": 0.20856418846368596,
+      "grad_norm": 0.8537253737449646,
+      "learning_rate": 0.00048247501894372534,
+      "loss": 7.1095,
+      "step": 607
+    },
+    {
+      "epoch": 0.20890778679723404,
+      "grad_norm": 0.43013009428977966,
+      "learning_rate": 0.0004823645504325699,
+      "loss": 7.0569,
+      "step": 608
+    },
+    {
+      "epoch": 0.20925138513078212,
+      "grad_norm": 0.4540230929851532,
+      "learning_rate": 0.0004822537475735379,
+      "loss": 7.1163,
+      "step": 609
+    },
+    {
+      "epoch": 0.2095949834643302,
+      "grad_norm": 0.5654716491699219,
+      "learning_rate": 0.00048214261052606294,
+      "loss": 7.0631,
+      "step": 610
+    },
+    {
+      "epoch": 0.20993858179787828,
+      "grad_norm": 0.6178705096244812,
+      "learning_rate": 0.00048203113945005947,
+      "loss": 7.0843,
+      "step": 611
+    },
+    {
+      "epoch": 0.21028218013142636,
+      "grad_norm": 0.43849003314971924,
+      "learning_rate": 0.00048191933450592256,
+      "loss": 7.0351,
+      "step": 612
+    },
+    {
+      "epoch": 0.21062577846497443,
+      "grad_norm": 0.48836883902549744,
+      "learning_rate": 0.00048180719585452753,
+      "loss": 7.0211,
+      "step": 613
+    },
+    {
+      "epoch": 0.21096937679852254,
+      "grad_norm": 0.5587930679321289,
+      "learning_rate": 0.00048169472365723,
+      "loss": 7.0268,
+      "step": 614
+    },
+    {
+      "epoch": 0.21131297513207062,
+      "grad_norm": 0.41739344596862793,
+      "learning_rate": 0.00048158191807586546,
+      "loss": 6.9506,
+      "step": 615
+    },
+    {
+      "epoch": 0.2116565734656187,
+      "grad_norm": 0.4349384009838104,
+      "learning_rate": 0.0004814687792727493,
+      "loss": 7.0376,
+      "step": 616
+    },
+    {
+      "epoch": 0.21200017179916678,
+      "grad_norm": 0.6397560834884644,
+      "learning_rate": 0.00048135530741067606,
+      "loss": 7.0313,
+      "step": 617
+    },
+    {
+      "epoch": 0.21234377013271485,
+      "grad_norm": 0.36318764090538025,
+      "learning_rate": 0.00048124150265291976,
+      "loss": 7.0587,
+      "step": 618
+    },
+    {
+      "epoch": 0.21268736846626293,
+      "grad_norm": 0.7137599587440491,
+      "learning_rate": 0.0004811273651632333,
+      "loss": 6.9631,
+      "step": 619
+    },
+    {
+      "epoch": 0.213030966799811,
+      "grad_norm": 0.59110426902771,
+      "learning_rate": 0.00048101289510584845,
+      "loss": 7.0164,
+      "step": 620
+    },
+    {
+      "epoch": 0.2133745651333591,
+      "grad_norm": 0.4364396333694458,
+      "learning_rate": 0.00048089809264547533,
+      "loss": 7.063,
+      "step": 621
+    },
+    {
+      "epoch": 0.2137181634669072,
+      "grad_norm": 0.6641795039176941,
+      "learning_rate": 0.00048078295794730266,
+      "loss": 7.0567,
+      "step": 622
+    },
+    {
+      "epoch": 0.21406176180045527,
+      "grad_norm": 0.37408143281936646,
+      "learning_rate": 0.0004806674911769968,
+      "loss": 7.0209,
+      "step": 623
+    },
+    {
+      "epoch": 0.21440536013400335,
+      "grad_norm": 0.4407050907611847,
+      "learning_rate": 0.0004805516925007024,
+      "loss": 7.0024,
+      "step": 624
+    },
+    {
+      "epoch": 0.21474895846755143,
+      "grad_norm": 0.5581675171852112,
+      "learning_rate": 0.00048043556208504125,
+      "loss": 7.0305,
+      "step": 625
+    },
+    {
+      "epoch": 0.2150925568010995,
+      "grad_norm": 0.5932101011276245,
+      "learning_rate": 0.0004803191000971128,
+      "loss": 7.0775,
+      "step": 626
+    },
+    {
+      "epoch": 0.2154361551346476,
+      "grad_norm": 0.630458414554596,
+      "learning_rate": 0.0004802023067044933,
+      "loss": 7.0122,
+      "step": 627
+    },
+    {
+      "epoch": 0.21577975346819567,
+      "grad_norm": 0.6922136545181274,
+      "learning_rate": 0.0004800851820752361,
+      "loss": 7.0494,
+      "step": 628
+    },
+    {
+      "epoch": 0.21612335180174377,
+      "grad_norm": 0.5312909483909607,
+      "learning_rate": 0.00047996772637787124,
+      "loss": 6.9883,
+      "step": 629
+    },
+    {
+      "epoch": 0.21646695013529185,
+      "grad_norm": 0.635317862033844,
+      "learning_rate": 0.00047984993978140473,
+      "loss": 7.0347,
+      "step": 630
+    },
+    {
+      "epoch": 0.21681054846883993,
+      "grad_norm": 0.587139368057251,
+      "learning_rate": 0.0004797318224553191,
+      "loss": 7.0547,
+      "step": 631
+    },
+    {
+      "epoch": 0.217154146802388,
+      "grad_norm": 0.46918633580207825,
+      "learning_rate": 0.00047961337456957256,
+      "loss": 7.119,
+      "step": 632
+    },
+    {
+      "epoch": 0.2174977451359361,
+      "grad_norm": 0.9854417443275452,
+      "learning_rate": 0.0004794945962945991,
+      "loss": 6.9985,
+      "step": 633
+    },
+    {
+      "epoch": 0.21784134346948417,
+      "grad_norm": 0.5167178511619568,
+      "learning_rate": 0.0004793754878013079,
+      "loss": 7.1618,
+      "step": 634
+    },
+    {
+      "epoch": 0.21818494180303225,
+      "grad_norm": 0.8147865533828735,
+      "learning_rate": 0.00047925604926108355,
+      "loss": 7.1051,
+      "step": 635
+    },
+    {
+      "epoch": 0.21852854013658032,
+      "grad_norm": 0.4546651840209961,
+      "learning_rate": 0.0004791362808457854,
+      "loss": 7.0173,
+      "step": 636
+    },
+    {
+      "epoch": 0.21887213847012843,
+      "grad_norm": 0.5698447823524475,
+      "learning_rate": 0.0004790161827277473,
+      "loss": 7.0944,
+      "step": 637
+    },
+    {
+      "epoch": 0.2192157368036765,
+      "grad_norm": 0.9129824042320251,
+      "learning_rate": 0.0004788957550797778,
+      "loss": 7.0367,
+      "step": 638
+    },
+    {
+      "epoch": 0.2195593351372246,
+      "grad_norm": 0.46476656198501587,
+      "learning_rate": 0.0004787749980751595,
+      "loss": 6.9945,
+      "step": 639
+    },
+    {
+      "epoch": 0.21990293347077267,
+      "grad_norm": 0.8614557385444641,
+      "learning_rate": 0.00047865391188764883,
+      "loss": 7.0681,
+      "step": 640
+    },
+    {
+      "epoch": 0.22024653180432074,
+      "grad_norm": 0.4384600520133972,
+      "learning_rate": 0.0004785324966914759,
+      "loss": 7.0053,
+      "step": 641
+    },
+    {
+      "epoch": 0.22059013013786882,
+      "grad_norm": 0.4937174320220947,
+      "learning_rate": 0.00047841075266134435,
+      "loss": 7.1369,
+      "step": 642
+    },
+    {
+      "epoch": 0.2209337284714169,
+      "grad_norm": 0.43719929456710815,
+      "learning_rate": 0.00047828867997243085,
+      "loss": 7.0655,
+      "step": 643
+    },
+    {
+      "epoch": 0.221277326804965,
+      "grad_norm": 0.5568514466285706,
+      "learning_rate": 0.00047816627880038504,
+      "loss": 7.0752,
+      "step": 644
+    },
+    {
+      "epoch": 0.2216209251385131,
+      "grad_norm": 0.6036604046821594,
+      "learning_rate": 0.0004780435493213292,
+      "loss": 7.1172,
+      "step": 645
+    },
+    {
+      "epoch": 0.22196452347206116,
+      "grad_norm": 0.7314201593399048,
+      "learning_rate": 0.0004779204917118579,
+      "loss": 7.025,
+      "step": 646
+    },
+    {
+      "epoch": 0.22230812180560924,
+      "grad_norm": 0.8482968807220459,
+      "learning_rate": 0.00047779710614903804,
+      "loss": 7.1686,
+      "step": 647
+    },
+    {
+      "epoch": 0.22265172013915732,
+      "grad_norm": 0.3950001895427704,
+      "learning_rate": 0.00047767339281040835,
+      "loss": 7.0538,
+      "step": 648
+    },
+    {
+      "epoch": 0.2229953184727054,
+      "grad_norm": 0.9423245787620544,
+      "learning_rate": 0.00047754935187397914,
+      "loss": 6.9874,
+      "step": 649
+    },
+    {
+      "epoch": 0.22333891680625348,
+      "grad_norm": 0.7840802073478699,
+      "learning_rate": 0.000477424983518232,
+      "loss": 7.1969,
+      "step": 650
+    },
+    {
+      "epoch": 0.22368251513980159,
+      "grad_norm": 0.7974851131439209,
+      "learning_rate": 0.00047730028792212,
+      "loss": 6.9681,
+      "step": 651
+    },
+    {
+      "epoch": 0.22402611347334966,
+      "grad_norm": 1.3284245729446411,
+      "learning_rate": 0.00047717526526506673,
+      "loss": 6.8882,
+      "step": 652
+    },
+    {
+      "epoch": 0.22436971180689774,
+      "grad_norm": 1.194890022277832,
+      "learning_rate": 0.0004770499157269664,
+      "loss": 7.0421,
+      "step": 653
+    },
+    {
+      "epoch": 0.22471331014044582,
+      "grad_norm": 0.7155433893203735,
+      "learning_rate": 0.0004769242394881838,
+      "loss": 6.875,
+      "step": 654
+    },
+    {
+      "epoch": 0.2250569084739939,
+      "grad_norm": 1.2595062255859375,
+      "learning_rate": 0.00047679823672955356,
+      "loss": 6.9402,
+      "step": 655
+    },
+    {
+      "epoch": 0.22540050680754198,
+      "grad_norm": 0.8444681763648987,
+      "learning_rate": 0.0004766719076323804,
+      "loss": 6.8736,
+      "step": 656
+    },
+    {
+      "epoch": 0.22574410514109006,
+      "grad_norm": 0.7984046339988708,
+      "learning_rate": 0.00047654525237843834,
+      "loss": 6.9932,
+      "step": 657
+    },
+    {
+      "epoch": 0.22608770347463814,
+      "grad_norm": 0.8302541971206665,
+      "learning_rate": 0.00047641827114997085,
+      "loss": 7.0802,
+      "step": 658
+    },
+    {
+      "epoch": 0.22643130180818624,
+      "grad_norm": 0.8889338374137878,
+      "learning_rate": 0.0004762909641296904,
+      "loss": 7.016,
+      "step": 659
+    },
+    {
+      "epoch": 0.22677490014173432,
+      "grad_norm": 0.5691213011741638,
+      "learning_rate": 0.00047616333150077826,
+      "loss": 6.9966,
+      "step": 660
+    },
+    {
+      "epoch": 0.2271184984752824,
+      "grad_norm": 0.5026600360870361,
+      "learning_rate": 0.00047603537344688423,
+      "loss": 6.9897,
+      "step": 661
+    },
+    {
+      "epoch": 0.22746209680883048,
+      "grad_norm": 0.4827994704246521,
+      "learning_rate": 0.00047590709015212635,
+      "loss": 6.8955,
+      "step": 662
+    },
+    {
+      "epoch": 0.22780569514237856,
+      "grad_norm": 0.6659157872200012,
+      "learning_rate": 0.0004757784818010906,
+      "loss": 6.952,
+      "step": 663
+    },
+    {
+      "epoch": 0.22814929347592663,
+      "grad_norm": 0.5329311490058899,
+      "learning_rate": 0.00047564954857883077,
+      "loss": 6.9433,
+      "step": 664
+    },
+    {
+      "epoch": 0.2284928918094747,
+      "grad_norm": 0.40088188648223877,
+      "learning_rate": 0.000475520290670868,
+      "loss": 6.8927,
+      "step": 665
+    },
+    {
+      "epoch": 0.22883649014302282,
+      "grad_norm": 0.6760165691375732,
+      "learning_rate": 0.0004753907082631906,
+      "loss": 6.9534,
+      "step": 666
+    },
+    {
+      "epoch": 0.2291800884765709,
+      "grad_norm": 0.4778375029563904,
+      "learning_rate": 0.0004752608015422541,
+      "loss": 7.0038,
+      "step": 667
+    },
+    {
+      "epoch": 0.22952368681011898,
+      "grad_norm": 0.3798183500766754,
+      "learning_rate": 0.0004751305706949803,
+      "loss": 7.0103,
+      "step": 668
+    },
+    {
+      "epoch": 0.22986728514366706,
+      "grad_norm": 0.6319564580917358,
+      "learning_rate": 0.00047500001590875755,
+      "loss": 7.0189,
+      "step": 669
+    },
+    {
+      "epoch": 0.23021088347721513,
+      "grad_norm": 0.4286358952522278,
+      "learning_rate": 0.0004748691373714403,
+      "loss": 6.9399,
+      "step": 670
+    },
+    {
+      "epoch": 0.2305544818107632,
+      "grad_norm": 0.5430161356925964,
+      "learning_rate": 0.00047473793527134884,
+      "loss": 6.9448,
+      "step": 671
+    },
+    {
+      "epoch": 0.2308980801443113,
+      "grad_norm": 0.45105573534965515,
+      "learning_rate": 0.0004746064097972691,
+      "loss": 6.9236,
+      "step": 672
+    },
+    {
+      "epoch": 0.23124167847785937,
+      "grad_norm": 0.587155818939209,
+      "learning_rate": 0.00047447456113845223,
+      "loss": 6.9462,
+      "step": 673
+    },
+    {
+      "epoch": 0.23158527681140748,
+      "grad_norm": 0.33924758434295654,
+      "learning_rate": 0.00047434238948461437,
+      "loss": 6.8088,
+      "step": 674
+    },
+    {
+      "epoch": 0.23192887514495555,
+      "grad_norm": 0.45958212018013,
+      "learning_rate": 0.0004742098950259365,
+      "loss": 6.9157,
+      "step": 675
+    },
+    {
+      "epoch": 0.23227247347850363,
+      "grad_norm": 0.44224873185157776,
+      "learning_rate": 0.0004740770779530641,
+      "loss": 7.0208,
+      "step": 676
+    },
+    {
+      "epoch": 0.2326160718120517,
+      "grad_norm": 0.48070213198661804,
+      "learning_rate": 0.00047394393845710684,
+      "loss": 6.9281,
+      "step": 677
+    },
+    {
+      "epoch": 0.2329596701455998,
+      "grad_norm": 0.3841937482357025,
+      "learning_rate": 0.00047381047672963815,
+      "loss": 6.9006,
+      "step": 678
+    },
+    {
+      "epoch": 0.23330326847914787,
+      "grad_norm": 0.3501332700252533,
+      "learning_rate": 0.0004736766929626954,
+      "loss": 6.8771,
+      "step": 679
+    },
+    {
+      "epoch": 0.23364686681269595,
+      "grad_norm": 0.4568793773651123,
+      "learning_rate": 0.00047354258734877907,
+      "loss": 6.828,
+      "step": 680
+    },
+    {
+      "epoch": 0.23399046514624405,
+      "grad_norm": 0.4417133033275604,
+      "learning_rate": 0.00047340816008085306,
+      "loss": 6.8497,
+      "step": 681
+    },
+    {
+      "epoch": 0.23433406347979213,
+      "grad_norm": 0.32319939136505127,
+      "learning_rate": 0.0004732734113523438,
+      "loss": 6.9765,
+      "step": 682
+    },
+    {
+      "epoch": 0.2346776618133402,
+      "grad_norm": 0.41723182797431946,
+      "learning_rate": 0.0004731383413571404,
+      "loss": 6.9349,
+      "step": 683
+    },
+    {
+      "epoch": 0.2350212601468883,
+      "grad_norm": 0.5126868486404419,
+      "learning_rate": 0.0004730029502895942,
+      "loss": 6.8654,
+      "step": 684
+    },
+    {
+      "epoch": 0.23536485848043637,
+      "grad_norm": 0.4400496184825897,
+      "learning_rate": 0.0004728672383445185,
+      "loss": 6.9556,
+      "step": 685
+    },
+    {
+      "epoch": 0.23570845681398445,
+      "grad_norm": 0.513038158416748,
+      "learning_rate": 0.0004727312057171884,
+      "loss": 6.9809,
+      "step": 686
+    },
+    {
+      "epoch": 0.23605205514753252,
+      "grad_norm": 0.370490163564682,
+      "learning_rate": 0.0004725948526033405,
+      "loss": 6.9601,
+      "step": 687
+    },
+    {
+      "epoch": 0.2363956534810806,
+      "grad_norm": 0.42959868907928467,
+      "learning_rate": 0.00047245817919917225,
+      "loss": 6.8961,
+      "step": 688
+    },
+    {
+      "epoch": 0.2367392518146287,
+      "grad_norm": 0.4754343628883362,
+      "learning_rate": 0.0004723211857013423,
+      "loss": 6.9749,
+      "step": 689
+    },
+    {
+      "epoch": 0.2370828501481768,
+      "grad_norm": 0.5382852554321289,
+      "learning_rate": 0.0004721838723069696,
+      "loss": 6.9439,
+      "step": 690
+    },
+    {
+      "epoch": 0.23742644848172487,
+      "grad_norm": 0.5841006636619568,
+      "learning_rate": 0.00047204623921363355,
+      "loss": 6.8801,
+      "step": 691
+    },
+    {
+      "epoch": 0.23777004681527295,
+      "grad_norm": 0.3350447416305542,
+      "learning_rate": 0.0004719082866193736,
+      "loss": 6.9135,
+      "step": 692
+    },
+    {
+      "epoch": 0.23811364514882102,
+      "grad_norm": 0.7295602560043335,
+      "learning_rate": 0.0004717700147226887,
+      "loss": 7.0857,
+      "step": 693
+    },
+    {
+      "epoch": 0.2384572434823691,
+      "grad_norm": 0.656107485294342,
+      "learning_rate": 0.00047163142372253766,
+      "loss": 7.0187,
+      "step": 694
+    },
+    {
+      "epoch": 0.23880084181591718,
+      "grad_norm": 0.48529568314552307,
+      "learning_rate": 0.0004714925138183379,
+      "loss": 7.0068,
+      "step": 695
+    },
+    {
+      "epoch": 0.2391444401494653,
+      "grad_norm": 0.6135896444320679,
+      "learning_rate": 0.0004713532852099663,
+      "loss": 7.0369,
+      "step": 696
+    },
+    {
+      "epoch": 0.23948803848301337,
+      "grad_norm": 0.8559364080429077,
+      "learning_rate": 0.00047121373809775783,
+      "loss": 6.8712,
+      "step": 697
+    },
+    {
+      "epoch": 0.23983163681656144,
+      "grad_norm": 0.509367823600769,
+      "learning_rate": 0.0004710738726825059,
+      "loss": 7.0357,
+      "step": 698
+    },
+    {
+      "epoch": 0.24017523515010952,
+      "grad_norm": 0.878773033618927,
+      "learning_rate": 0.0004709336891654621,
+      "loss": 7.0097,
+      "step": 699
+    },
+    {
+      "epoch": 0.2405188334836576,
+      "grad_norm": 0.6598408818244934,
+      "learning_rate": 0.00047079318774833555,
+      "loss": 6.9585,
+      "step": 700
+    },
+    {
+      "epoch": 0.24086243181720568,
+      "grad_norm": 1.2100400924682617,
+      "learning_rate": 0.00047065236863329284,
+      "loss": 6.876,
+      "step": 701
+    },
+    {
+      "epoch": 0.24120603015075376,
+      "grad_norm": 0.577090322971344,
+      "learning_rate": 0.00047051123202295777,
+      "loss": 6.9374,
+      "step": 702
+    },
+    {
+      "epoch": 0.24154962848430186,
+      "grad_norm": 0.682905912399292,
+      "learning_rate": 0.0004703697781204108,
+      "loss": 6.6889,
+      "step": 703
+    },
+    {
+      "epoch": 0.24189322681784994,
+      "grad_norm": 0.8273911476135254,
+      "learning_rate": 0.0004702280071291891,
+      "loss": 6.7998,
+      "step": 704
+    },
+    {
+      "epoch": 0.24223682515139802,
+      "grad_norm": 0.39249464869499207,
+      "learning_rate": 0.00047008591925328614,
+      "loss": 6.9258,
+      "step": 705
+    },
+    {
+      "epoch": 0.2425804234849461,
+      "grad_norm": 0.8825654983520508,
+      "learning_rate": 0.00046994351469715107,
+      "loss": 6.8986,
+      "step": 706
+    },
+    {
+      "epoch": 0.24292402181849418,
+      "grad_norm": 0.4094250202178955,
+      "learning_rate": 0.0004698007936656891,
+      "loss": 6.8182,
+      "step": 707
+    },
+    {
+      "epoch": 0.24326762015204226,
+      "grad_norm": 0.6193078756332397,
+      "learning_rate": 0.00046965775636426046,
+      "loss": 6.952,
+      "step": 708
+    },
+    {
+      "epoch": 0.24361121848559034,
+      "grad_norm": 0.8103482723236084,
+      "learning_rate": 0.0004695144029986807,
+      "loss": 6.8076,
+      "step": 709
+    },
+    {
+      "epoch": 0.24395481681913841,
+      "grad_norm": 0.655996561050415,
+      "learning_rate": 0.00046937073377522004,
+      "loss": 6.7956,
+      "step": 710
+    },
+    {
+      "epoch": 0.24429841515268652,
+      "grad_norm": 0.6959365606307983,
+      "learning_rate": 0.00046922674890060326,
+      "loss": 6.8264,
+      "step": 711
+    },
+    {
+      "epoch": 0.2446420134862346,
+      "grad_norm": 0.5749850273132324,
+      "learning_rate": 0.0004690824485820092,
+      "loss": 6.7739,
+      "step": 712
+    },
+    {
+      "epoch": 0.24498561181978268,
+      "grad_norm": 0.840958833694458,
+      "learning_rate": 0.0004689378330270707,
+      "loss": 6.8343,
+      "step": 713
+    },
+    {
+      "epoch": 0.24532921015333076,
+      "grad_norm": 0.5014116168022156,
+      "learning_rate": 0.0004687929024438742,
+      "loss": 6.8268,
+      "step": 714
+    },
+    {
+      "epoch": 0.24567280848687884,
+      "grad_norm": 0.7229539155960083,
+      "learning_rate": 0.0004686476570409594,
+      "loss": 6.8121,
+      "step": 715
+    },
+    {
+      "epoch": 0.24601640682042691,
+      "grad_norm": 0.4928436279296875,
+      "learning_rate": 0.00046850209702731894,
+      "loss": 6.8857,
+      "step": 716
+    },
+    {
+      "epoch": 0.246360005153975,
+      "grad_norm": 0.5194845199584961,
+      "learning_rate": 0.00046835622261239825,
+      "loss": 6.7582,
+      "step": 717
+    },
+    {
+      "epoch": 0.2467036034875231,
+      "grad_norm": 0.4320314824581146,
+      "learning_rate": 0.0004682100340060951,
+      "loss": 6.9288,
+      "step": 718
+    },
+    {
+      "epoch": 0.24704720182107118,
+      "grad_norm": 0.5804980397224426,
+      "learning_rate": 0.0004680635314187592,
+      "loss": 6.8817,
+      "step": 719
+    },
+    {
+      "epoch": 0.24739080015461926,
+      "grad_norm": 0.458049476146698,
+      "learning_rate": 0.0004679167150611924,
+      "loss": 6.8723,
+      "step": 720
+    },
+    {
+      "epoch": 0.24773439848816733,
+      "grad_norm": 0.4215126931667328,
+      "learning_rate": 0.00046776958514464773,
+      "loss": 6.8758,
+      "step": 721
+    },
+    {
+      "epoch": 0.2480779968217154,
+      "grad_norm": 0.40314817428588867,
+      "learning_rate": 0.0004676221418808295,
+      "loss": 6.7961,
+      "step": 722
+    },
+    {
+      "epoch": 0.2484215951552635,
+      "grad_norm": 0.4778897166252136,
+      "learning_rate": 0.00046747438548189294,
+      "loss": 6.8339,
+      "step": 723
+    },
+    {
+      "epoch": 0.24876519348881157,
+      "grad_norm": 0.341545969247818,
+      "learning_rate": 0.00046732631616044364,
+      "loss": 6.8159,
+      "step": 724
+    },
+    {
+      "epoch": 0.24910879182235965,
+      "grad_norm": 0.3407266438007355,
+      "learning_rate": 0.0004671779341295378,
+      "loss": 6.8255,
+      "step": 725
+    },
+    {
+      "epoch": 0.24945239015590775,
+      "grad_norm": 0.407026469707489,
+      "learning_rate": 0.0004670292396026812,
+      "loss": 6.8622,
+      "step": 726
+    },
+    {
+      "epoch": 0.24979598848945583,
+      "grad_norm": 0.4541730582714081,
+      "learning_rate": 0.00046688023279382965,
+      "loss": 6.8587,
+      "step": 727
+    },
+    {
+      "epoch": 0.2501395868230039,
+      "grad_norm": 0.37031111121177673,
+      "learning_rate": 0.0004667309139173879,
+      "loss": 6.7835,
+      "step": 728
+    },
+    {
+      "epoch": 0.250483185156552,
+      "grad_norm": 0.47327926754951477,
+      "learning_rate": 0.00046658128318821,
+      "loss": 6.8738,
+      "step": 729
+    },
+    {
+      "epoch": 0.25082678349010007,
+      "grad_norm": 0.47790077328681946,
+      "learning_rate": 0.00046643134082159876,
+      "loss": 6.8492,
+      "step": 730
+    },
+    {
+      "epoch": 0.25117038182364815,
+      "grad_norm": 0.6300373673439026,
+      "learning_rate": 0.0004662810870333053,
+      "loss": 6.8715,
+      "step": 731
+    },
+    {
+      "epoch": 0.2515139801571962,
+      "grad_norm": 0.32731807231903076,
+      "learning_rate": 0.0004661305220395286,
+      "loss": 6.8528,
+      "step": 732
+    },
+    {
+      "epoch": 0.2518575784907443,
+      "grad_norm": 0.5586317181587219,
+      "learning_rate": 0.0004659796460569159,
+      "loss": 6.8647,
+      "step": 733
+    },
+    {
+      "epoch": 0.2522011768242924,
+      "grad_norm": 0.5391055941581726,
+      "learning_rate": 0.0004658284593025617,
+      "loss": 6.8655,
+      "step": 734
+    },
+    {
+      "epoch": 0.25254477515784046,
+      "grad_norm": 0.44915974140167236,
+      "learning_rate": 0.0004656769619940075,
+      "loss": 6.8233,
+      "step": 735
+    },
+    {
+      "epoch": 0.25288837349138854,
+      "grad_norm": 0.47281840443611145,
+      "learning_rate": 0.00046552515434924194,
+      "loss": 6.7832,
+      "step": 736
+    },
+    {
+      "epoch": 0.2532319718249367,
+      "grad_norm": 0.37644147872924805,
+      "learning_rate": 0.0004653730365867,
+      "loss": 6.8604,
+      "step": 737
+    },
+    {
+      "epoch": 0.25357557015848475,
+      "grad_norm": 0.35747960209846497,
+      "learning_rate": 0.0004652206089252631,
+      "loss": 6.7951,
+      "step": 738
+    },
+    {
+      "epoch": 0.25391916849203283,
+      "grad_norm": 0.4939366579055786,
+      "learning_rate": 0.00046506787158425827,
+      "loss": 7.0186,
+      "step": 739
+    },
+    {
+      "epoch": 0.2542627668255809,
+      "grad_norm": 0.4213995933532715,
+      "learning_rate": 0.00046491482478345836,
+      "loss": 6.944,
+      "step": 740
+    },
+    {
+      "epoch": 0.254606365159129,
+      "grad_norm": 0.41876089572906494,
+      "learning_rate": 0.00046476146874308157,
+      "loss": 6.8617,
+      "step": 741
+    },
+    {
+      "epoch": 0.25494996349267707,
+      "grad_norm": 0.5702718496322632,
+      "learning_rate": 0.00046460780368379076,
+      "loss": 6.9404,
+      "step": 742
+    },
+    {
+      "epoch": 0.25529356182622515,
+      "grad_norm": 0.6110662817955017,
+      "learning_rate": 0.0004644538298266936,
+      "loss": 6.7365,
+      "step": 743
+    },
+    {
+      "epoch": 0.2556371601597732,
+      "grad_norm": 0.5077341794967651,
+      "learning_rate": 0.0004642995473933422,
+      "loss": 6.9497,
+      "step": 744
+    },
+    {
+      "epoch": 0.2559807584933213,
+      "grad_norm": 0.612311601638794,
+      "learning_rate": 0.0004641449566057325,
+      "loss": 7.0145,
+      "step": 745
+    },
+    {
+      "epoch": 0.2563243568268694,
+      "grad_norm": 0.6061681509017944,
+      "learning_rate": 0.00046399005768630425,
+      "loss": 6.965,
+      "step": 746
+    },
+    {
+      "epoch": 0.25666795516041746,
+      "grad_norm": 0.6442392468452454,
+      "learning_rate": 0.0004638348508579405,
+      "loss": 6.9068,
+      "step": 747
+    },
+    {
+      "epoch": 0.25701155349396554,
+      "grad_norm": 0.4608883857727051,
+      "learning_rate": 0.0004636793363439674,
+      "loss": 6.9511,
+      "step": 748
+    },
+    {
+      "epoch": 0.2573551518275136,
+      "grad_norm": 0.7235476970672607,
+      "learning_rate": 0.0004635235143681538,
+      "loss": 6.9665,
+      "step": 749
+    },
+    {
+      "epoch": 0.2576987501610617,
+      "grad_norm": 0.9052413702011108,
+      "learning_rate": 0.00046336738515471087,
+      "loss": 6.7848,
+      "step": 750
+    },
+    {
+      "epoch": 0.2580423484946098,
+      "grad_norm": 1.7460522651672363,
+      "learning_rate": 0.00046321094892829204,
+      "loss": 6.8786,
+      "step": 751
+    },
+    {
+      "epoch": 0.2583859468281579,
+      "grad_norm": 0.9062784314155579,
+      "learning_rate": 0.0004630542059139923,
+      "loss": 6.7366,
+      "step": 752
+    },
+    {
+      "epoch": 0.258729545161706,
+      "grad_norm": 0.9159809947013855,
+      "learning_rate": 0.0004628971563373483,
+      "loss": 6.7801,
+      "step": 753
+    },
+    {
+      "epoch": 0.25907314349525407,
+      "grad_norm": 1.137810230255127,
+      "learning_rate": 0.0004627398004243376,
+      "loss": 6.8466,
+      "step": 754
+    },
+    {
+      "epoch": 0.25941674182880214,
+      "grad_norm": 0.7727288007736206,
+      "learning_rate": 0.00046258213840137864,
+      "loss": 6.7646,
+      "step": 755
+    },
+    {
+      "epoch": 0.2597603401623502,
+      "grad_norm": 1.0215615034103394,
+      "learning_rate": 0.0004624241704953304,
+      "loss": 6.7423,
+      "step": 756
+    },
+    {
+      "epoch": 0.2601039384958983,
+      "grad_norm": 0.9777013063430786,
+      "learning_rate": 0.0004622658969334916,
+      "loss": 6.7828,
+      "step": 757
+    },
+    {
+      "epoch": 0.2604475368294464,
+      "grad_norm": 0.5539371371269226,
+      "learning_rate": 0.0004621073179436015,
+      "loss": 6.7059,
+      "step": 758
+    },
+    {
+      "epoch": 0.26079113516299446,
+      "grad_norm": 1.1476919651031494,
+      "learning_rate": 0.00046194843375383797,
+      "loss": 6.8811,
+      "step": 759
+    },
+    {
+      "epoch": 0.26113473349654254,
+      "grad_norm": 0.58660888671875,
+      "learning_rate": 0.0004617892445928188,
+      "loss": 6.7705,
+      "step": 760
+    },
+    {
+      "epoch": 0.2614783318300906,
+      "grad_norm": 0.6147243976593018,
+      "learning_rate": 0.00046162975068960013,
+      "loss": 6.7306,
+      "step": 761
+    },
+    {
+      "epoch": 0.2618219301636387,
+      "grad_norm": 0.8718441128730774,
+      "learning_rate": 0.00046146995227367663,
+      "loss": 6.7174,
+      "step": 762
+    },
+    {
+      "epoch": 0.2621655284971868,
+      "grad_norm": 0.6512937545776367,
+      "learning_rate": 0.00046130984957498135,
+      "loss": 6.8326,
+      "step": 763
+    },
+    {
+      "epoch": 0.26250912683073485,
+      "grad_norm": 0.7064685821533203,
+      "learning_rate": 0.00046114944282388504,
+      "loss": 6.83,
+      "step": 764
+    },
+    {
+      "epoch": 0.26285272516428293,
+      "grad_norm": 0.6572036743164062,
+      "learning_rate": 0.0004609887322511959,
+      "loss": 6.8905,
+      "step": 765
+    },
+    {
+      "epoch": 0.263196323497831,
+      "grad_norm": 0.44737884402275085,
+      "learning_rate": 0.0004608277180881594,
+      "loss": 6.8373,
+      "step": 766
+    },
+    {
+      "epoch": 0.26353992183137914,
+      "grad_norm": 0.8529016971588135,
+      "learning_rate": 0.00046066640056645775,
+      "loss": 6.8329,
+      "step": 767
+    },
+    {
+      "epoch": 0.2638835201649272,
+      "grad_norm": 0.4838736355304718,
+      "learning_rate": 0.0004605047799182097,
+      "loss": 6.8027,
+      "step": 768
+    },
+    {
+      "epoch": 0.2642271184984753,
+      "grad_norm": 0.6361299753189087,
+      "learning_rate": 0.0004603428563759703,
+      "loss": 6.8504,
+      "step": 769
+    },
+    {
+      "epoch": 0.2645707168320234,
+      "grad_norm": 0.5523536205291748,
+      "learning_rate": 0.0004601806301727302,
+      "loss": 6.7116,
+      "step": 770
+    },
+    {
+      "epoch": 0.26491431516557146,
+      "grad_norm": 0.683148980140686,
+      "learning_rate": 0.00046001810154191564,
+      "loss": 6.8294,
+      "step": 771
+    },
+    {
+      "epoch": 0.26525791349911954,
+      "grad_norm": 0.5535314083099365,
+      "learning_rate": 0.0004598552707173881,
+      "loss": 6.86,
+      "step": 772
+    },
+    {
+      "epoch": 0.2656015118326676,
+      "grad_norm": 0.41244781017303467,
+      "learning_rate": 0.0004596921379334438,
+      "loss": 6.7168,
+      "step": 773
+    },
+    {
+      "epoch": 0.2659451101662157,
+      "grad_norm": 0.6043175458908081,
+      "learning_rate": 0.0004595287034248134,
+      "loss": 6.7497,
+      "step": 774
+    },
+    {
+      "epoch": 0.26628870849976377,
+      "grad_norm": 0.6349027752876282,
+      "learning_rate": 0.0004593649674266619,
+      "loss": 6.8369,
+      "step": 775
+    },
+    {
+      "epoch": 0.26663230683331185,
+      "grad_norm": 0.4858072102069855,
+      "learning_rate": 0.00045920093017458785,
+      "loss": 6.7509,
+      "step": 776
+    },
+    {
+      "epoch": 0.26697590516685993,
+      "grad_norm": 0.6944791674613953,
+      "learning_rate": 0.0004590365919046235,
+      "loss": 6.8669,
+      "step": 777
+    },
+    {
+      "epoch": 0.267319503500408,
+      "grad_norm": 0.41476762294769287,
+      "learning_rate": 0.0004588719528532341,
+      "loss": 6.8717,
+      "step": 778
+    },
+    {
+      "epoch": 0.2676631018339561,
+      "grad_norm": 0.4901805520057678,
+      "learning_rate": 0.0004587070132573178,
+      "loss": 6.7902,
+      "step": 779
+    },
+    {
+      "epoch": 0.26800670016750416,
+      "grad_norm": 0.7249867916107178,
+      "learning_rate": 0.0004585417733542051,
+      "loss": 6.8388,
+      "step": 780
+    },
+    {
+      "epoch": 0.26835029850105224,
+      "grad_norm": 0.4491178095340729,
+      "learning_rate": 0.00045837623338165865,
+      "loss": 6.7337,
+      "step": 781
+    },
+    {
+      "epoch": 0.2686938968346004,
+      "grad_norm": 0.6511254906654358,
+      "learning_rate": 0.0004582103935778728,
+      "loss": 6.8176,
+      "step": 782
+    },
+    {
+      "epoch": 0.26903749516814845,
+      "grad_norm": 0.7600392699241638,
+      "learning_rate": 0.0004580442541814735,
+      "loss": 6.7143,
+      "step": 783
+    },
+    {
+      "epoch": 0.26938109350169653,
+      "grad_norm": 0.5467169284820557,
+      "learning_rate": 0.00045787781543151765,
+      "loss": 6.6721,
+      "step": 784
+    },
+    {
+      "epoch": 0.2697246918352446,
+      "grad_norm": 0.655737578868866,
+      "learning_rate": 0.0004577110775674928,
+      "loss": 6.8681,
+      "step": 785
+    },
+    {
+      "epoch": 0.2700682901687927,
+      "grad_norm": 0.7441399693489075,
+      "learning_rate": 0.00045754404082931714,
+      "loss": 6.7865,
+      "step": 786
+    },
+    {
+      "epoch": 0.27041188850234077,
+      "grad_norm": 0.42660608887672424,
+      "learning_rate": 0.00045737670545733866,
+      "loss": 6.7737,
+      "step": 787
+    },
+    {
+      "epoch": 0.27075548683588885,
+      "grad_norm": 0.6970313191413879,
+      "learning_rate": 0.0004572090716923353,
+      "loss": 6.7858,
+      "step": 788
+    },
+    {
+      "epoch": 0.2710990851694369,
+      "grad_norm": 0.4984079897403717,
+      "learning_rate": 0.0004570411397755141,
+      "loss": 6.8268,
+      "step": 789
+    },
+    {
+      "epoch": 0.271442683502985,
+      "grad_norm": 0.35249513387680054,
+      "learning_rate": 0.0004568729099485114,
+      "loss": 6.8356,
+      "step": 790
+    },
+    {
+      "epoch": 0.2717862818365331,
+      "grad_norm": 0.553500771522522,
+      "learning_rate": 0.00045670438245339176,
+      "loss": 6.759,
+      "step": 791
+    },
+    {
+      "epoch": 0.27212988017008116,
+      "grad_norm": 0.5268304944038391,
+      "learning_rate": 0.0004565355575326485,
+      "loss": 6.9307,
+      "step": 792
+    },
+    {
+      "epoch": 0.27247347850362924,
+      "grad_norm": 0.5753699541091919,
+      "learning_rate": 0.0004563664354292027,
+      "loss": 6.8501,
+      "step": 793
+    },
+    {
+      "epoch": 0.2728170768371773,
+      "grad_norm": 0.5838276147842407,
+      "learning_rate": 0.0004561970163864031,
+      "loss": 6.7411,
+      "step": 794
+    },
+    {
+      "epoch": 0.2731606751707254,
+      "grad_norm": 0.6358978152275085,
+      "learning_rate": 0.0004560273006480256,
+      "loss": 6.8322,
+      "step": 795
+    },
+    {
+      "epoch": 0.27350427350427353,
+      "grad_norm": 0.75990229845047,
+      "learning_rate": 0.0004558572884582732,
+      "loss": 6.7258,
+      "step": 796
+    },
+    {
+      "epoch": 0.2738478718378216,
+      "grad_norm": 0.6735451817512512,
+      "learning_rate": 0.00045568698006177535,
+      "loss": 6.7351,
+      "step": 797
+    },
+    {
+      "epoch": 0.2741914701713697,
+      "grad_norm": 0.5910247564315796,
+      "learning_rate": 0.0004555163757035876,
+      "loss": 6.8092,
+      "step": 798
+    },
+    {
+      "epoch": 0.27453506850491777,
+      "grad_norm": 0.6522710919380188,
+      "learning_rate": 0.0004553454756291916,
+      "loss": 6.866,
+      "step": 799
+    },
+    {
+      "epoch": 0.27487866683846585,
+      "grad_norm": 0.7660143375396729,
+      "learning_rate": 0.00045517428008449436,
+      "loss": 6.9458,
+      "step": 800
+    },
+    {
+      "epoch": 0.2752222651720139,
+      "grad_norm": 1.4569878578186035,
+      "learning_rate": 0.00045500278931582806,
+      "loss": 6.7737,
+      "step": 801
+    },
+    {
+      "epoch": 0.275565863505562,
+      "grad_norm": 0.972352147102356,
+      "learning_rate": 0.00045483100356994967,
+      "loss": 6.8857,
+      "step": 802
+    },
+    {
+      "epoch": 0.2759094618391101,
+      "grad_norm": 1.2153319120407104,
+      "learning_rate": 0.00045465892309404064,
+      "loss": 6.7493,
+      "step": 803
+    },
+    {
+      "epoch": 0.27625306017265816,
+      "grad_norm": 1.4174124002456665,
+      "learning_rate": 0.0004544865481357064,
+      "loss": 6.7461,
+      "step": 804
+    },
+    {
+      "epoch": 0.27659665850620624,
+      "grad_norm": 0.7363278269767761,
+      "learning_rate": 0.00045431387894297626,
+      "loss": 6.7705,
+      "step": 805
+    },
+    {
+      "epoch": 0.2769402568397543,
+      "grad_norm": 1.0926170349121094,
+      "learning_rate": 0.0004541409157643027,
+      "loss": 6.581,
+      "step": 806
+    },
+    {
+      "epoch": 0.2772838551733024,
+      "grad_norm": 1.3416601419448853,
+      "learning_rate": 0.00045396765884856154,
+      "loss": 6.7194,
+      "step": 807
+    },
+    {
+      "epoch": 0.2776274535068505,
+      "grad_norm": 0.5606733560562134,
+      "learning_rate": 0.0004537941084450509,
+      "loss": 6.8035,
+      "step": 808
+    },
+    {
+      "epoch": 0.27797105184039855,
+      "grad_norm": 0.7519741058349609,
+      "learning_rate": 0.0004536202648034914,
+      "loss": 6.8173,
+      "step": 809
+    },
+    {
+      "epoch": 0.27831465017394663,
+      "grad_norm": 0.9418850541114807,
+      "learning_rate": 0.0004534461281740255,
+      "loss": 6.7941,
+      "step": 810
+    },
+    {
+      "epoch": 0.27865824850749477,
+      "grad_norm": 0.6259487867355347,
+      "learning_rate": 0.0004532716988072175,
+      "loss": 6.6247,
+      "step": 811
+    },
+    {
+      "epoch": 0.27900184684104284,
+      "grad_norm": 0.8542924523353577,
+      "learning_rate": 0.00045309697695405243,
+      "loss": 6.7618,
+      "step": 812
+    },
+    {
+      "epoch": 0.2793454451745909,
+      "grad_norm": 0.8797979354858398,
+      "learning_rate": 0.0004529219628659366,
+      "loss": 6.7305,
+      "step": 813
+    },
+    {
+      "epoch": 0.279689043508139,
+      "grad_norm": 0.484947144985199,
+      "learning_rate": 0.00045274665679469666,
+      "loss": 6.7074,
+      "step": 814
+    },
+    {
+      "epoch": 0.2800326418416871,
+      "grad_norm": 0.940768301486969,
+      "learning_rate": 0.0004525710589925794,
+      "loss": 6.8354,
+      "step": 815
+    },
+    {
+      "epoch": 0.28037624017523516,
+      "grad_norm": 0.640664279460907,
+      "learning_rate": 0.0004523951697122514,
+      "loss": 6.7316,
+      "step": 816
+    },
+    {
+      "epoch": 0.28071983850878324,
+      "grad_norm": 0.3233809173107147,
+      "learning_rate": 0.0004522189892067985,
+      "loss": 6.8219,
+      "step": 817
+    },
+    {
+      "epoch": 0.2810634368423313,
+      "grad_norm": 0.71434485912323,
+      "learning_rate": 0.00045204251772972595,
+      "loss": 6.6404,
+      "step": 818
+    },
+    {
+      "epoch": 0.2814070351758794,
+      "grad_norm": 0.6310349702835083,
+      "learning_rate": 0.00045186575553495716,
+      "loss": 6.6152,
+      "step": 819
+    },
+    {
+      "epoch": 0.2817506335094275,
+      "grad_norm": 0.40379568934440613,
+      "learning_rate": 0.0004516887028768344,
+      "loss": 6.6638,
+      "step": 820
+    },
+    {
+      "epoch": 0.28209423184297555,
+      "grad_norm": 0.46431466937065125,
+      "learning_rate": 0.0004515113600101174,
+      "loss": 6.8226,
+      "step": 821
+    },
+    {
+      "epoch": 0.28243783017652363,
+      "grad_norm": 0.6690576672554016,
+      "learning_rate": 0.0004513337271899838,
+      "loss": 6.7364,
+      "step": 822
+    },
+    {
+      "epoch": 0.2827814285100717,
+      "grad_norm": 0.3921510875225067,
+      "learning_rate": 0.00045115580467202835,
+      "loss": 6.6688,
+      "step": 823
+    },
+    {
+      "epoch": 0.2831250268436198,
+      "grad_norm": 0.4708666205406189,
+      "learning_rate": 0.0004509775927122625,
+      "loss": 6.688,
+      "step": 824
+    },
+    {
+      "epoch": 0.28346862517716787,
+      "grad_norm": 0.6789883971214294,
+      "learning_rate": 0.00045079909156711445,
+      "loss": 6.5988,
+      "step": 825
+    },
+    {
+      "epoch": 0.283812223510716,
+      "grad_norm": 0.5630450248718262,
+      "learning_rate": 0.0004506203014934283,
+      "loss": 6.7569,
+      "step": 826
+    },
+    {
+      "epoch": 0.2841558218442641,
+      "grad_norm": 0.4851526916027069,
+      "learning_rate": 0.000450441222748464,
+      "loss": 6.7702,
+      "step": 827
+    },
+    {
+      "epoch": 0.28449942017781216,
+      "grad_norm": 0.6094221472740173,
+      "learning_rate": 0.00045026185558989676,
+      "loss": 6.864,
+      "step": 828
+    },
+    {
+      "epoch": 0.28484301851136024,
+      "grad_norm": 0.5844842195510864,
+      "learning_rate": 0.0004500822002758169,
+      "loss": 6.6631,
+      "step": 829
+    },
+    {
+      "epoch": 0.2851866168449083,
+      "grad_norm": 0.5621864199638367,
+      "learning_rate": 0.0004499022570647292,
+      "loss": 6.6732,
+      "step": 830
+    },
+    {
+      "epoch": 0.2855302151784564,
+      "grad_norm": 0.5564741492271423,
+      "learning_rate": 0.00044972202621555295,
+      "loss": 6.8367,
+      "step": 831
+    },
+    {
+      "epoch": 0.28587381351200447,
+      "grad_norm": 0.6472800970077515,
+      "learning_rate": 0.0004495415079876211,
+      "loss": 6.8898,
+      "step": 832
+    },
+    {
+      "epoch": 0.28621741184555255,
+      "grad_norm": 0.6958934664726257,
+      "learning_rate": 0.00044936070264068017,
+      "loss": 6.7203,
+      "step": 833
+    },
+    {
+      "epoch": 0.28656101017910063,
+      "grad_norm": 0.5514652132987976,
+      "learning_rate": 0.00044917961043488994,
+      "loss": 6.8128,
+      "step": 834
+    },
+    {
+      "epoch": 0.2869046085126487,
+      "grad_norm": 0.7259712815284729,
+      "learning_rate": 0.00044899823163082264,
+      "loss": 6.7471,
+      "step": 835
+    },
+    {
+      "epoch": 0.2872482068461968,
+      "grad_norm": 0.4060599207878113,
+      "learning_rate": 0.00044881656648946324,
+      "loss": 6.7305,
+      "step": 836
+    },
+    {
+      "epoch": 0.28759180517974486,
+      "grad_norm": 0.697080671787262,
+      "learning_rate": 0.0004486346152722085,
+      "loss": 6.6904,
+      "step": 837
+    },
+    {
+      "epoch": 0.28793540351329294,
+      "grad_norm": 0.46781861782073975,
+      "learning_rate": 0.0004484523782408668,
+      "loss": 6.6904,
+      "step": 838
+    },
+    {
+      "epoch": 0.288279001846841,
+      "grad_norm": 0.5770555734634399,
+      "learning_rate": 0.000448269855657658,
+      "loss": 6.7446,
+      "step": 839
+    },
+    {
+      "epoch": 0.2886226001803891,
+      "grad_norm": 0.458122193813324,
+      "learning_rate": 0.0004480870477852126,
+      "loss": 6.7125,
+      "step": 840
+    },
+    {
+      "epoch": 0.28896619851393723,
+      "grad_norm": 0.8331108093261719,
+      "learning_rate": 0.00044790395488657165,
+      "loss": 6.7704,
+      "step": 841
+    },
+    {
+      "epoch": 0.2893097968474853,
+      "grad_norm": 0.47524768114089966,
+      "learning_rate": 0.00044772057722518646,
+      "loss": 6.7999,
+      "step": 842
+    },
+    {
+      "epoch": 0.2896533951810334,
+      "grad_norm": 0.6736469864845276,
+      "learning_rate": 0.00044753691506491783,
+      "loss": 6.8244,
+      "step": 843
+    },
+    {
+      "epoch": 0.28999699351458147,
+      "grad_norm": 0.4914340078830719,
+      "learning_rate": 0.00044735296867003625,
+      "loss": 6.7941,
+      "step": 844
+    },
+    {
+      "epoch": 0.29034059184812955,
+      "grad_norm": 0.6356337070465088,
+      "learning_rate": 0.0004471687383052209,
+      "loss": 6.7074,
+      "step": 845
+    },
+    {
+      "epoch": 0.2906841901816776,
+      "grad_norm": 0.4956357777118683,
+      "learning_rate": 0.0004469842242355598,
+      "loss": 6.6277,
+      "step": 846
+    },
+    {
+      "epoch": 0.2910277885152257,
+      "grad_norm": 0.7174055576324463,
+      "learning_rate": 0.00044679942672654896,
+      "loss": 6.7337,
+      "step": 847
+    },
+    {
+      "epoch": 0.2913713868487738,
+      "grad_norm": 0.5987393856048584,
+      "learning_rate": 0.0004466143460440923,
+      "loss": 6.7455,
+      "step": 848
+    },
+    {
+      "epoch": 0.29171498518232186,
+      "grad_norm": 0.7290389537811279,
+      "learning_rate": 0.00044642898245450134,
+      "loss": 6.9492,
+      "step": 849
+    },
+    {
+      "epoch": 0.29205858351586994,
+      "grad_norm": 0.9655451774597168,
+      "learning_rate": 0.0004462433362244946,
+      "loss": 6.8043,
+      "step": 850
+    },
+    {
+      "epoch": 0.292402181849418,
+      "grad_norm": 0.908883810043335,
+      "learning_rate": 0.0004460574076211973,
+      "loss": 6.5924,
+      "step": 851
+    },
+    {
+      "epoch": 0.2927457801829661,
+      "grad_norm": 0.8863041996955872,
+      "learning_rate": 0.00044587119691214075,
+      "loss": 6.729,
+      "step": 852
+    },
+    {
+      "epoch": 0.2930893785165142,
+      "grad_norm": 0.6901432275772095,
+      "learning_rate": 0.0004456847043652624,
+      "loss": 6.7657,
+      "step": 853
+    },
+    {
+      "epoch": 0.29343297685006225,
+      "grad_norm": 1.0988612174987793,
+      "learning_rate": 0.00044549793024890535,
+      "loss": 6.5863,
+      "step": 854
+    },
+    {
+      "epoch": 0.29377657518361033,
+      "grad_norm": 0.700564980506897,
+      "learning_rate": 0.00044531087483181753,
+      "loss": 6.6539,
+      "step": 855
+    },
+    {
+      "epoch": 0.29412017351715847,
+      "grad_norm": 0.5881510376930237,
+      "learning_rate": 0.00044512353838315177,
+      "loss": 6.6578,
+      "step": 856
+    },
+    {
+      "epoch": 0.29446377185070655,
+      "grad_norm": 0.8495005369186401,
+      "learning_rate": 0.00044493592117246544,
+      "loss": 6.753,
+      "step": 857
+    },
+    {
+      "epoch": 0.2948073701842546,
+      "grad_norm": 0.5816578269004822,
+      "learning_rate": 0.00044474802346971973,
+      "loss": 6.6681,
+      "step": 858
+    },
+    {
+      "epoch": 0.2951509685178027,
+      "grad_norm": 0.6731425523757935,
+      "learning_rate": 0.00044455984554527927,
+      "loss": 6.7042,
+      "step": 859
+    },
+    {
+      "epoch": 0.2954945668513508,
+      "grad_norm": 0.5258037447929382,
+      "learning_rate": 0.0004443713876699124,
+      "loss": 6.6961,
+      "step": 860
+    },
+    {
+      "epoch": 0.29583816518489886,
+      "grad_norm": 0.6853783130645752,
+      "learning_rate": 0.00044418265011478964,
+      "loss": 6.7168,
+      "step": 861
+    },
+    {
+      "epoch": 0.29618176351844694,
+      "grad_norm": 0.5510942935943604,
+      "learning_rate": 0.0004439936331514844,
+      "loss": 6.5608,
+      "step": 862
+    },
+    {
+      "epoch": 0.296525361851995,
+      "grad_norm": 0.7021049857139587,
+      "learning_rate": 0.000443804337051972,
+      "loss": 6.6402,
+      "step": 863
+    },
+    {
+      "epoch": 0.2968689601855431,
+      "grad_norm": 0.522904634475708,
+      "learning_rate": 0.0004436147620886294,
+      "loss": 6.7086,
+      "step": 864
+    },
+    {
+      "epoch": 0.2972125585190912,
+      "grad_norm": 0.6038668155670166,
+      "learning_rate": 0.00044342490853423476,
+      "loss": 6.7431,
+      "step": 865
+    },
+    {
+      "epoch": 0.29755615685263925,
+      "grad_norm": 0.5508298277854919,
+      "learning_rate": 0.0004432347766619672,
+      "loss": 6.6429,
+      "step": 866
+    },
+    {
+      "epoch": 0.29789975518618733,
+      "grad_norm": 0.4411887526512146,
+      "learning_rate": 0.00044304436674540626,
+      "loss": 6.6424,
+      "step": 867
+    },
+    {
+      "epoch": 0.2982433535197354,
+      "grad_norm": 0.692892849445343,
+      "learning_rate": 0.0004428536790585315,
+      "loss": 6.601,
+      "step": 868
+    },
+    {
+      "epoch": 0.2985869518532835,
+      "grad_norm": 0.38894644379615784,
+      "learning_rate": 0.00044266271387572234,
+      "loss": 6.7866,
+      "step": 869
+    },
+    {
+      "epoch": 0.29893055018683157,
+      "grad_norm": 0.5061523914337158,
+      "learning_rate": 0.00044247147147175725,
+      "loss": 6.7596,
+      "step": 870
+    },
+    {
+      "epoch": 0.2992741485203797,
+      "grad_norm": 0.5069543123245239,
+      "learning_rate": 0.00044227995212181375,
+      "loss": 6.7587,
+      "step": 871
+    },
+    {
+      "epoch": 0.2996177468539278,
+      "grad_norm": 0.48187875747680664,
+      "learning_rate": 0.0004420881561014679,
+      "loss": 6.5468,
+      "step": 872
+    },
+    {
+      "epoch": 0.29996134518747586,
+      "grad_norm": 0.5325146317481995,
+      "learning_rate": 0.00044189608368669364,
+      "loss": 6.7372,
+      "step": 873
+    },
+    {
+      "epoch": 0.30030494352102394,
+      "grad_norm": 0.5697482228279114,
+      "learning_rate": 0.0004417037351538628,
+      "loss": 6.6634,
+      "step": 874
+    },
+    {
+      "epoch": 0.300648541854572,
+      "grad_norm": 0.4286175072193146,
+      "learning_rate": 0.0004415111107797445,
+      "loss": 6.6235,
+      "step": 875
+    },
+    {
+      "epoch": 0.3009921401881201,
+      "grad_norm": 0.6160731911659241,
+      "learning_rate": 0.0004413182108415047,
+      "loss": 6.5816,
+      "step": 876
+    },
+    {
+      "epoch": 0.3013357385216682,
+      "grad_norm": 0.4158060550689697,
+      "learning_rate": 0.00044112503561670593,
+      "loss": 6.6958,
+      "step": 877
+    },
+    {
+      "epoch": 0.30167933685521625,
+      "grad_norm": 0.43943968415260315,
+      "learning_rate": 0.00044093158538330675,
+      "loss": 6.7403,
+      "step": 878
+    },
+    {
+      "epoch": 0.30202293518876433,
+      "grad_norm": 0.45871615409851074,
+      "learning_rate": 0.0004407378604196615,
+      "loss": 6.6959,
+      "step": 879
+    },
+    {
+      "epoch": 0.3023665335223124,
+      "grad_norm": 0.35182371735572815,
+      "learning_rate": 0.00044054386100451974,
+      "loss": 6.7437,
+      "step": 880
+    },
+    {
+      "epoch": 0.3027101318558605,
+      "grad_norm": 0.5132912397384644,
+      "learning_rate": 0.0004403495874170261,
+      "loss": 6.6109,
+      "step": 881
+    },
+    {
+      "epoch": 0.30305373018940857,
+      "grad_norm": 0.4913354814052582,
+      "learning_rate": 0.00044015503993671953,
+      "loss": 6.6739,
+      "step": 882
+    },
+    {
+      "epoch": 0.30339732852295664,
+      "grad_norm": 0.37906894087791443,
+      "learning_rate": 0.0004399602188435332,
+      "loss": 6.5922,
+      "step": 883
+    },
+    {
+      "epoch": 0.3037409268565047,
+      "grad_norm": 0.4221290946006775,
+      "learning_rate": 0.0004397651244177939,
+      "loss": 6.6821,
+      "step": 884
+    },
+    {
+      "epoch": 0.3040845251900528,
+      "grad_norm": 0.40234696865081787,
+      "learning_rate": 0.0004395697569402218,
+      "loss": 6.5751,
+      "step": 885
+    },
+    {
+      "epoch": 0.30442812352360094,
+      "grad_norm": 0.6360386610031128,
+      "learning_rate": 0.00043937411669192996,
+      "loss": 6.6542,
+      "step": 886
+    },
+    {
+      "epoch": 0.304771721857149,
+      "grad_norm": 0.4311862885951996,
+      "learning_rate": 0.0004391782039544238,
+      "loss": 6.6606,
+      "step": 887
+    },
+    {
+      "epoch": 0.3051153201906971,
+      "grad_norm": 0.5377070903778076,
+      "learning_rate": 0.000438982019009601,
+      "loss": 6.7238,
+      "step": 888
+    },
+    {
+      "epoch": 0.30545891852424517,
+      "grad_norm": 0.441566526889801,
+      "learning_rate": 0.0004387855621397508,
+      "loss": 6.5752,
+      "step": 889
+    },
+    {
+      "epoch": 0.30580251685779325,
+      "grad_norm": 0.42087745666503906,
+      "learning_rate": 0.00043858883362755377,
+      "loss": 6.6638,
+      "step": 890
+    },
+    {
+      "epoch": 0.30614611519134133,
+      "grad_norm": 0.39083683490753174,
+      "learning_rate": 0.00043839183375608115,
+      "loss": 6.6812,
+      "step": 891
+    },
+    {
+      "epoch": 0.3064897135248894,
+      "grad_norm": 0.5337326526641846,
+      "learning_rate": 0.0004381945628087951,
+      "loss": 6.6301,
+      "step": 892
+    },
+    {
+      "epoch": 0.3068333118584375,
+      "grad_norm": 0.43058067560195923,
+      "learning_rate": 0.0004379970210695473,
+      "loss": 6.6669,
+      "step": 893
+    },
+    {
+      "epoch": 0.30717691019198556,
+      "grad_norm": 0.5187793374061584,
+      "learning_rate": 0.0004377992088225794,
+      "loss": 6.6735,
+      "step": 894
+    },
+    {
+      "epoch": 0.30752050852553364,
+      "grad_norm": 0.4057919681072235,
+      "learning_rate": 0.0004376011263525221,
+      "loss": 6.7691,
+      "step": 895
+    },
+    {
+      "epoch": 0.3078641068590817,
+      "grad_norm": 0.5876466631889343,
+      "learning_rate": 0.0004374027739443952,
+      "loss": 6.8599,
+      "step": 896
+    },
+    {
+      "epoch": 0.3082077051926298,
+      "grad_norm": 0.8425361514091492,
+      "learning_rate": 0.00043720415188360645,
+      "loss": 6.8207,
+      "step": 897
+    },
+    {
+      "epoch": 0.3085513035261779,
+      "grad_norm": 0.6548316478729248,
+      "learning_rate": 0.00043700526045595213,
+      "loss": 6.7563,
+      "step": 898
+    },
+    {
+      "epoch": 0.30889490185972596,
+      "grad_norm": 0.8088213801383972,
+      "learning_rate": 0.00043680609994761565,
+      "loss": 6.7839,
+      "step": 899
+    },
+    {
+      "epoch": 0.30923850019327404,
+      "grad_norm": 0.962527334690094,
+      "learning_rate": 0.00043660667064516795,
+      "loss": 6.8026,
+      "step": 900
+    },
+    {
+      "epoch": 0.30958209852682217,
+      "grad_norm": 1.5333473682403564,
+      "learning_rate": 0.0004364069728355665,
+      "loss": 6.5689,
+      "step": 901
+    },
+    {
+      "epoch": 0.30992569686037025,
+      "grad_norm": 0.7353837490081787,
+      "learning_rate": 0.0004362070068061553,
+      "loss": 6.7109,
+      "step": 902
+    },
+    {
+      "epoch": 0.3102692951939183,
+      "grad_norm": 1.1564072370529175,
+      "learning_rate": 0.00043600677284466404,
+      "loss": 6.7302,
+      "step": 903
+    },
+    {
+      "epoch": 0.3106128935274664,
+      "grad_norm": 0.8054890632629395,
+      "learning_rate": 0.00043580627123920824,
+      "loss": 6.5836,
+      "step": 904
+    },
+    {
+      "epoch": 0.3109564918610145,
+      "grad_norm": 0.8645017743110657,
+      "learning_rate": 0.0004356055022782884,
+      "loss": 6.5373,
+      "step": 905
+    },
+    {
+      "epoch": 0.31130009019456256,
+      "grad_norm": 1.007243037223816,
+      "learning_rate": 0.00043540446625078957,
+      "loss": 6.6522,
+      "step": 906
+    },
+    {
+      "epoch": 0.31164368852811064,
+      "grad_norm": 0.5772066116333008,
+      "learning_rate": 0.0004352031634459813,
+      "loss": 6.5941,
+      "step": 907
+    },
+    {
+      "epoch": 0.3119872868616587,
+      "grad_norm": 1.151341199874878,
+      "learning_rate": 0.00043500159415351693,
+      "loss": 6.71,
+      "step": 908
+    },
+    {
+      "epoch": 0.3123308851952068,
+      "grad_norm": 0.7844879031181335,
+      "learning_rate": 0.00043479975866343316,
+      "loss": 6.564,
+      "step": 909
+    },
+    {
+      "epoch": 0.3126744835287549,
+      "grad_norm": 0.6750956177711487,
+      "learning_rate": 0.0004345976572661499,
+      "loss": 6.52,
+      "step": 910
+    },
+    {
+      "epoch": 0.31301808186230295,
+      "grad_norm": 0.9015241861343384,
+      "learning_rate": 0.0004343952902524695,
+      "loss": 6.641,
+      "step": 911
+    },
+    {
+      "epoch": 0.31336168019585103,
+      "grad_norm": 0.6968957185745239,
+      "learning_rate": 0.00043419265791357656,
+      "loss": 6.5191,
+      "step": 912
+    },
+    {
+      "epoch": 0.3137052785293991,
+      "grad_norm": 0.6399244070053101,
+      "learning_rate": 0.00043398976054103756,
+      "loss": 6.4583,
+      "step": 913
+    },
+    {
+      "epoch": 0.3140488768629472,
+      "grad_norm": 0.7823150157928467,
+      "learning_rate": 0.0004337865984268001,
+      "loss": 6.6004,
+      "step": 914
+    },
+    {
+      "epoch": 0.3143924751964953,
+      "grad_norm": 0.6562049984931946,
+      "learning_rate": 0.000433583171863193,
+      "loss": 6.6225,
+      "step": 915
+    },
+    {
+      "epoch": 0.3147360735300434,
+      "grad_norm": 0.5780950784683228,
+      "learning_rate": 0.0004333794811429253,
+      "loss": 6.682,
+      "step": 916
+    },
+    {
+      "epoch": 0.3150796718635915,
+      "grad_norm": 0.7009182572364807,
+      "learning_rate": 0.0004331755265590864,
+      "loss": 6.5454,
+      "step": 917
+    },
+    {
+      "epoch": 0.31542327019713956,
+      "grad_norm": 0.4939836263656616,
+      "learning_rate": 0.0004329713084051452,
+      "loss": 6.6302,
+      "step": 918
+    },
+    {
+      "epoch": 0.31576686853068764,
+      "grad_norm": 0.7301641702651978,
+      "learning_rate": 0.00043276682697494995,
+      "loss": 6.5314,
+      "step": 919
+    },
+    {
+      "epoch": 0.3161104668642357,
+      "grad_norm": 0.4727785289287567,
+      "learning_rate": 0.00043256208256272765,
+      "loss": 6.6969,
+      "step": 920
+    },
+    {
+      "epoch": 0.3164540651977838,
+      "grad_norm": 0.4740373194217682,
+      "learning_rate": 0.0004323570754630838,
+      "loss": 6.6621,
+      "step": 921
+    },
+    {
+      "epoch": 0.3167976635313319,
+      "grad_norm": 0.5652904510498047,
+      "learning_rate": 0.00043215180597100167,
+      "loss": 6.5914,
+      "step": 922
+    },
+    {
+      "epoch": 0.31714126186487995,
+      "grad_norm": 0.678132176399231,
+      "learning_rate": 0.0004319462743818424,
+      "loss": 6.6089,
+      "step": 923
+    },
+    {
+      "epoch": 0.31748486019842803,
+      "grad_norm": 0.3767012357711792,
+      "learning_rate": 0.0004317404809913439,
+      "loss": 6.5765,
+      "step": 924
+    },
+    {
+      "epoch": 0.3178284585319761,
+      "grad_norm": 0.6184304356575012,
+      "learning_rate": 0.00043153442609562115,
+      "loss": 6.6968,
+      "step": 925
+    },
+    {
+      "epoch": 0.3181720568655242,
+      "grad_norm": 0.529187798500061,
+      "learning_rate": 0.00043132810999116513,
+      "loss": 6.6796,
+      "step": 926
+    },
+    {
+      "epoch": 0.31851565519907227,
+      "grad_norm": 0.486133873462677,
+      "learning_rate": 0.0004311215329748428,
+      "loss": 6.6639,
+      "step": 927
+    },
+    {
+      "epoch": 0.31885925353262035,
+      "grad_norm": 0.5016087293624878,
+      "learning_rate": 0.0004309146953438966,
+      "loss": 6.5808,
+      "step": 928
+    },
+    {
+      "epoch": 0.3192028518661684,
+      "grad_norm": 0.5674017071723938,
+      "learning_rate": 0.00043070759739594365,
+      "loss": 6.5018,
+      "step": 929
+    },
+    {
+      "epoch": 0.31954645019971656,
+      "grad_norm": 0.4067033529281616,
+      "learning_rate": 0.0004305002394289762,
+      "loss": 6.6528,
+      "step": 930
+    },
+    {
+      "epoch": 0.31989004853326464,
+      "grad_norm": 0.5798845887184143,
+      "learning_rate": 0.00043029262174136,
+      "loss": 6.5737,
+      "step": 931
+    },
+    {
+      "epoch": 0.3202336468668127,
+      "grad_norm": 0.6043611168861389,
+      "learning_rate": 0.000430084744631835,
+      "loss": 6.6468,
+      "step": 932
+    },
+    {
+      "epoch": 0.3205772452003608,
+      "grad_norm": 0.4977371394634247,
+      "learning_rate": 0.00042987660839951424,
+      "loss": 6.6475,
+      "step": 933
+    },
+    {
+      "epoch": 0.3209208435339089,
+      "grad_norm": 0.5232419371604919,
+      "learning_rate": 0.0004296682133438836,
+      "loss": 6.5821,
+      "step": 934
+    },
+    {
+      "epoch": 0.32126444186745695,
+      "grad_norm": 0.5253320932388306,
+      "learning_rate": 0.0004294595597648014,
+      "loss": 6.6007,
+      "step": 935
+    },
+    {
+      "epoch": 0.32160804020100503,
+      "grad_norm": 0.48890867829322815,
+      "learning_rate": 0.0004292506479624979,
+      "loss": 6.5654,
+      "step": 936
+    },
+    {
+      "epoch": 0.3219516385345531,
+      "grad_norm": 0.5480407476425171,
+      "learning_rate": 0.00042904147823757504,
+      "loss": 6.5487,
+      "step": 937
+    },
+    {
+      "epoch": 0.3222952368681012,
+      "grad_norm": 0.44250285625457764,
+      "learning_rate": 0.00042883205089100574,
+      "loss": 6.6104,
+      "step": 938
+    },
+    {
+      "epoch": 0.32263883520164927,
+      "grad_norm": 0.5121859908103943,
+      "learning_rate": 0.00042862236622413384,
+      "loss": 6.5706,
+      "step": 939
+    },
+    {
+      "epoch": 0.32298243353519734,
+      "grad_norm": 0.4506629407405853,
+      "learning_rate": 0.00042841242453867313,
+      "loss": 6.5679,
+      "step": 940
+    },
+    {
+      "epoch": 0.3233260318687454,
+      "grad_norm": 0.5592184662818909,
+      "learning_rate": 0.0004282022261367073,
+      "loss": 6.6794,
+      "step": 941
+    },
+    {
+      "epoch": 0.3236696302022935,
+      "grad_norm": 0.48790469765663147,
+      "learning_rate": 0.0004279917713206897,
+      "loss": 6.7813,
+      "step": 942
+    },
+    {
+      "epoch": 0.3240132285358416,
+      "grad_norm": 0.7107426524162292,
+      "learning_rate": 0.00042778106039344227,
+      "loss": 6.7156,
+      "step": 943
+    },
+    {
+      "epoch": 0.32435682686938966,
+      "grad_norm": 0.42893269658088684,
+      "learning_rate": 0.00042757009365815567,
+      "loss": 6.6114,
+      "step": 944
+    },
+    {
+      "epoch": 0.3247004252029378,
+      "grad_norm": 0.5906190276145935,
+      "learning_rate": 0.0004273588714183887,
+      "loss": 6.6893,
+      "step": 945
+    },
+    {
+      "epoch": 0.32504402353648587,
+      "grad_norm": 0.8306535482406616,
+      "learning_rate": 0.00042714739397806746,
+      "loss": 6.7016,
+      "step": 946
+    },
+    {
+      "epoch": 0.32538762187003395,
+      "grad_norm": 0.5476863384246826,
+      "learning_rate": 0.00042693566164148577,
+      "loss": 6.8143,
+      "step": 947
+    },
+    {
+      "epoch": 0.32573122020358203,
+      "grad_norm": 1.0128917694091797,
+      "learning_rate": 0.00042672367471330373,
+      "loss": 6.6837,
+      "step": 948
+    },
+    {
+      "epoch": 0.3260748185371301,
+      "grad_norm": 0.6552304625511169,
+      "learning_rate": 0.00042651143349854817,
+      "loss": 6.797,
+      "step": 949
+    },
+    {
+      "epoch": 0.3264184168706782,
+      "grad_norm": 0.8041029572486877,
+      "learning_rate": 0.0004262989383026115,
+      "loss": 6.8178,
+      "step": 950
+    },
+    {
+      "epoch": 0.32676201520422626,
+      "grad_norm": 1.0014656782150269,
+      "learning_rate": 0.00042608618943125166,
+      "loss": 6.5204,
+      "step": 951
+    },
+    {
+      "epoch": 0.32710561353777434,
+      "grad_norm": 0.7724018096923828,
+      "learning_rate": 0.00042587318719059176,
+      "loss": 6.3653,
+      "step": 952
+    },
+    {
+      "epoch": 0.3274492118713224,
+      "grad_norm": 0.7663524746894836,
+      "learning_rate": 0.00042565993188711934,
+      "loss": 6.5467,
+      "step": 953
+    },
+    {
+      "epoch": 0.3277928102048705,
+      "grad_norm": 0.9170653820037842,
+      "learning_rate": 0.00042544642382768606,
+      "loss": 6.4953,
+      "step": 954
+    },
+    {
+      "epoch": 0.3281364085384186,
+      "grad_norm": 0.462952584028244,
+      "learning_rate": 0.00042523266331950745,
+      "loss": 6.6712,
+      "step": 955
+    },
+    {
+      "epoch": 0.32848000687196666,
+      "grad_norm": 0.7020468711853027,
+      "learning_rate": 0.000425018650670162,
+      "loss": 6.5491,
+      "step": 956
+    },
+    {
+      "epoch": 0.32882360520551474,
+      "grad_norm": 0.5389872193336487,
+      "learning_rate": 0.0004248043861875912,
+      "loss": 6.499,
+      "step": 957
+    },
+    {
+      "epoch": 0.3291672035390628,
+      "grad_norm": 0.41495102643966675,
+      "learning_rate": 0.0004245898701800989,
+      "loss": 6.5433,
+      "step": 958
+    },
+    {
+      "epoch": 0.3295108018726109,
+      "grad_norm": 0.5482673048973083,
+      "learning_rate": 0.00042437510295635075,
+      "loss": 6.5594,
+      "step": 959
+    },
+    {
+      "epoch": 0.329854400206159,
+      "grad_norm": 0.5409672260284424,
+      "learning_rate": 0.0004241600848253739,
+      "loss": 6.5728,
+      "step": 960
+    },
+    {
+      "epoch": 0.3301979985397071,
+      "grad_norm": 0.3423202633857727,
+      "learning_rate": 0.0004239448160965567,
+      "loss": 6.6138,
+      "step": 961
+    },
+    {
+      "epoch": 0.3305415968732552,
+      "grad_norm": 0.7785842418670654,
+      "learning_rate": 0.00042372929707964796,
+      "loss": 6.4957,
+      "step": 962
+    },
+    {
+      "epoch": 0.33088519520680326,
+      "grad_norm": 0.49811336398124695,
+      "learning_rate": 0.0004235135280847565,
+      "loss": 6.5921,
+      "step": 963
+    },
+    {
+      "epoch": 0.33122879354035134,
+      "grad_norm": 0.4862115979194641,
+      "learning_rate": 0.0004232975094223511,
+      "loss": 6.5412,
+      "step": 964
+    },
+    {
+      "epoch": 0.3315723918738994,
+      "grad_norm": 0.5971689224243164,
+      "learning_rate": 0.0004230812414032595,
+      "loss": 6.4801,
+      "step": 965
+    },
+    {
+      "epoch": 0.3319159902074475,
+      "grad_norm": 0.48547112941741943,
+      "learning_rate": 0.0004228647243386685,
+      "loss": 6.5879,
+      "step": 966
+    },
+    {
+      "epoch": 0.3322595885409956,
+      "grad_norm": 0.36358118057250977,
+      "learning_rate": 0.000422647958540123,
+      "loss": 6.6918,
+      "step": 967
+    },
+    {
+      "epoch": 0.33260318687454365,
+      "grad_norm": 0.5594439506530762,
+      "learning_rate": 0.0004224309443195261,
+      "loss": 6.618,
+      "step": 968
+    },
+    {
+      "epoch": 0.33294678520809173,
+      "grad_norm": 0.4119882583618164,
+      "learning_rate": 0.000422213681989138,
+      "loss": 6.5929,
+      "step": 969
+    },
+    {
+      "epoch": 0.3332903835416398,
+      "grad_norm": 0.3595133423805237,
+      "learning_rate": 0.00042199617186157624,
+      "loss": 6.4484,
+      "step": 970
+    },
+    {
+      "epoch": 0.3336339818751879,
+      "grad_norm": 0.4845230281352997,
+      "learning_rate": 0.00042177841424981467,
+      "loss": 6.5066,
+      "step": 971
+    },
+    {
+      "epoch": 0.33397758020873597,
+      "grad_norm": 0.467210054397583,
+      "learning_rate": 0.00042156040946718344,
+      "loss": 6.5665,
+      "step": 972
+    },
+    {
+      "epoch": 0.33432117854228405,
+      "grad_norm": 0.537044107913971,
+      "learning_rate": 0.00042134215782736804,
+      "loss": 6.5144,
+      "step": 973
+    },
+    {
+      "epoch": 0.3346647768758321,
+      "grad_norm": 0.5238211750984192,
+      "learning_rate": 0.00042112365964440965,
+      "loss": 6.4788,
+      "step": 974
+    },
+    {
+      "epoch": 0.33500837520938026,
+      "grad_norm": 0.4372142255306244,
+      "learning_rate": 0.0004209049152327037,
+      "loss": 6.5712,
+      "step": 975
+    },
+    {
+      "epoch": 0.33535197354292834,
+      "grad_norm": 0.38211849331855774,
+      "learning_rate": 0.0004206859249070002,
+      "loss": 6.6268,
+      "step": 976
+    },
+    {
+      "epoch": 0.3356955718764764,
+      "grad_norm": 0.3685896098613739,
+      "learning_rate": 0.00042046668898240296,
+      "loss": 6.5762,
+      "step": 977
+    },
+    {
+      "epoch": 0.3360391702100245,
+      "grad_norm": 0.4509871006011963,
+      "learning_rate": 0.0004202472077743692,
+      "loss": 6.5991,
+      "step": 978
+    },
+    {
+      "epoch": 0.3363827685435726,
+      "grad_norm": 0.3491871953010559,
+      "learning_rate": 0.00042002748159870895,
+      "loss": 6.5537,
+      "step": 979
+    },
+    {
+      "epoch": 0.33672636687712065,
+      "grad_norm": 0.38078513741493225,
+      "learning_rate": 0.00041980751077158487,
+      "loss": 6.5054,
+      "step": 980
+    },
+    {
+      "epoch": 0.33706996521066873,
+      "grad_norm": 0.4195086658000946,
+      "learning_rate": 0.0004195872956095115,
+      "loss": 6.5264,
+      "step": 981
+    },
+    {
+      "epoch": 0.3374135635442168,
+      "grad_norm": 0.3654896318912506,
+      "learning_rate": 0.00041936683642935515,
+      "loss": 6.4927,
+      "step": 982
+    },
+    {
+      "epoch": 0.3377571618777649,
+      "grad_norm": 0.4824683368206024,
+      "learning_rate": 0.000419146133548333,
+      "loss": 6.5299,
+      "step": 983
+    },
+    {
+      "epoch": 0.33810076021131297,
+      "grad_norm": 0.4073243737220764,
+      "learning_rate": 0.00041892518728401317,
+      "loss": 6.5118,
+      "step": 984
+    },
+    {
+      "epoch": 0.33844435854486105,
+      "grad_norm": 0.37707698345184326,
+      "learning_rate": 0.0004187039979543138,
+      "loss": 6.5245,
+      "step": 985
+    },
+    {
+      "epoch": 0.3387879568784091,
+      "grad_norm": 0.40587034821510315,
+      "learning_rate": 0.0004184825658775027,
+      "loss": 6.6768,
+      "step": 986
+    },
+    {
+      "epoch": 0.3391315552119572,
+      "grad_norm": 0.4814893901348114,
+      "learning_rate": 0.00041826089137219724,
+      "loss": 6.5214,
+      "step": 987
+    },
+    {
+      "epoch": 0.3394751535455053,
+      "grad_norm": 0.38695332407951355,
+      "learning_rate": 0.0004180389747573634,
+      "loss": 6.6829,
+      "step": 988
+    },
+    {
+      "epoch": 0.33981875187905336,
+      "grad_norm": 0.4304019808769226,
+      "learning_rate": 0.00041781681635231555,
+      "loss": 6.6049,
+      "step": 989
+    },
+    {
+      "epoch": 0.3401623502126015,
+      "grad_norm": 0.41044285893440247,
+      "learning_rate": 0.00041759441647671604,
+      "loss": 6.56,
+      "step": 990
+    },
+    {
+      "epoch": 0.3405059485461496,
+      "grad_norm": 0.5040400624275208,
+      "learning_rate": 0.00041737177545057456,
+      "loss": 6.6315,
+      "step": 991
+    },
+    {
+      "epoch": 0.34084954687969765,
+      "grad_norm": 0.5537315607070923,
+      "learning_rate": 0.000417148893594248,
+      "loss": 6.5066,
+      "step": 992
+    },
+    {
+      "epoch": 0.34119314521324573,
+      "grad_norm": 0.3936966061592102,
+      "learning_rate": 0.00041692577122843963,
+      "loss": 6.572,
+      "step": 993
+    },
+    {
+      "epoch": 0.3415367435467938,
+      "grad_norm": 0.5949533581733704,
+      "learning_rate": 0.0004167024086741987,
+      "loss": 6.5359,
+      "step": 994
+    },
+    {
+      "epoch": 0.3418803418803419,
+      "grad_norm": 0.6033600568771362,
+      "learning_rate": 0.0004164788062529203,
+      "loss": 6.5554,
+      "step": 995
+    },
+    {
+      "epoch": 0.34222394021388997,
+      "grad_norm": 0.5860453248023987,
+      "learning_rate": 0.0004162549642863445,
+      "loss": 6.6641,
+      "step": 996
+    },
+    {
+      "epoch": 0.34256753854743804,
+      "grad_norm": 0.5877623558044434,
+      "learning_rate": 0.0004160308830965559,
+      "loss": 6.7018,
+      "step": 997
+    },
+    {
+      "epoch": 0.3429111368809861,
+      "grad_norm": 0.73051917552948,
+      "learning_rate": 0.00041580656300598375,
+      "loss": 6.5305,
+      "step": 998
+    },
+    {
+      "epoch": 0.3432547352145342,
+      "grad_norm": 0.6099729537963867,
+      "learning_rate": 0.00041558200433740067,
+      "loss": 6.6973,
+      "step": 999
+    },
+    {
+      "epoch": 0.3435983335480823,
+      "grad_norm": 0.9224446415901184,
+      "learning_rate": 0.0004153572074139228,
+      "loss": 6.6335,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34394193188163036,
+      "grad_norm": 1.092505693435669,
+      "learning_rate": 0.00041513217255900893,
+      "loss": 6.5569,
+      "step": 1001
+    },
+    {
+      "epoch": 0.34428553021517844,
+      "grad_norm": 0.5701151490211487,
+      "learning_rate": 0.00041490690009646024,
+      "loss": 6.4043,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3446291285487265,
+      "grad_norm": 0.838062047958374,
+      "learning_rate": 0.00041468139035042003,
+      "loss": 6.5534,
+      "step": 1003
+    },
+    {
+      "epoch": 0.3449727268822746,
+      "grad_norm": 0.7914153337478638,
+      "learning_rate": 0.0004144556436453727,
+      "loss": 6.526,
+      "step": 1004
+    },
+    {
+      "epoch": 0.34531632521582273,
+      "grad_norm": 0.47454750537872314,
+      "learning_rate": 0.00041422966030614375,
+      "loss": 6.5954,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3456599235493708,
+      "grad_norm": 0.8763797283172607,
+      "learning_rate": 0.0004140034406578991,
+      "loss": 6.5447,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3460035218829189,
+      "grad_norm": 0.7527329325675964,
+      "learning_rate": 0.000413776985026145,
+      "loss": 6.423,
+      "step": 1007
+    },
+    {
+      "epoch": 0.34634712021646696,
+      "grad_norm": 0.5256273150444031,
+      "learning_rate": 0.0004135502937367268,
+      "loss": 6.5776,
+      "step": 1008
+    },
+    {
+      "epoch": 0.34669071855001504,
+      "grad_norm": 0.8483496308326721,
+      "learning_rate": 0.00041332336711582916,
+      "loss": 6.4731,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3470343168835631,
+      "grad_norm": 0.5697401165962219,
+      "learning_rate": 0.00041309620548997557,
+      "loss": 6.5361,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3473779152171112,
+      "grad_norm": 0.4558808505535126,
+      "learning_rate": 0.0004128688091860273,
+      "loss": 6.4667,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3477215135506593,
+      "grad_norm": 0.7396731376647949,
+      "learning_rate": 0.00041264117853118343,
+      "loss": 6.5441,
+      "step": 1012
+    },
+    {
+      "epoch": 0.34806511188420736,
+      "grad_norm": 0.5605579614639282,
+      "learning_rate": 0.0004124133138529803,
+      "loss": 6.561,
+      "step": 1013
+    },
+    {
+      "epoch": 0.34840871021775544,
+      "grad_norm": 0.5781577229499817,
+      "learning_rate": 0.00041218521547929096,
+      "loss": 6.5182,
+      "step": 1014
+    },
+    {
+      "epoch": 0.3487523085513035,
+      "grad_norm": 0.5698264837265015,
+      "learning_rate": 0.00041195688373832465,
+      "loss": 6.4753,
+      "step": 1015
+    },
+    {
+      "epoch": 0.3490959068848516,
+      "grad_norm": 0.6196632385253906,
+      "learning_rate": 0.0004117283189586265,
+      "loss": 6.6474,
+      "step": 1016
+    },
+    {
+      "epoch": 0.34943950521839967,
+      "grad_norm": 0.631287693977356,
+      "learning_rate": 0.00041149952146907697,
+      "loss": 6.508,
+      "step": 1017
+    },
+    {
+      "epoch": 0.34978310355194775,
+      "grad_norm": 0.6482592821121216,
+      "learning_rate": 0.0004112704915988913,
+      "loss": 6.4538,
+      "step": 1018
+    },
+    {
+      "epoch": 0.35012670188549583,
+      "grad_norm": 0.45326462388038635,
+      "learning_rate": 0.00041104122967761906,
+      "loss": 6.4599,
+      "step": 1019
+    },
+    {
+      "epoch": 0.35047030021904396,
+      "grad_norm": 0.5699672698974609,
+      "learning_rate": 0.0004108117360351438,
+      "loss": 6.5104,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35081389855259204,
+      "grad_norm": 0.5698271989822388,
+      "learning_rate": 0.0004105820110016825,
+      "loss": 6.4758,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3511574968861401,
+      "grad_norm": 0.48133784532546997,
+      "learning_rate": 0.000410352054907785,
+      "loss": 6.4689,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3515010952196882,
+      "grad_norm": 0.5399252772331238,
+      "learning_rate": 0.00041012186808433364,
+      "loss": 6.5425,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3518446935532363,
+      "grad_norm": 0.4314304292201996,
+      "learning_rate": 0.00040989145086254295,
+      "loss": 6.5328,
+      "step": 1024
+    },
+    {
+      "epoch": 0.35218829188678435,
+      "grad_norm": 0.5915732979774475,
+      "learning_rate": 0.0004096608035739585,
+      "loss": 6.504,
+      "step": 1025
+    },
+    {
+      "epoch": 0.35253189022033243,
+      "grad_norm": 0.5009291768074036,
+      "learning_rate": 0.0004094299265504575,
+      "loss": 6.4955,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3528754885538805,
+      "grad_norm": 0.3504527807235718,
+      "learning_rate": 0.00040919882012424737,
+      "loss": 6.433,
+      "step": 1027
+    },
+    {
+      "epoch": 0.3532190868874286,
+      "grad_norm": 0.4587783217430115,
+      "learning_rate": 0.0004089674846278656,
+      "loss": 6.4936,
+      "step": 1028
+    },
+    {
+      "epoch": 0.35356268522097667,
+      "grad_norm": 0.5252232551574707,
+      "learning_rate": 0.00040873592039417935,
+      "loss": 6.5655,
+      "step": 1029
+    },
+    {
+      "epoch": 0.35390628355452475,
+      "grad_norm": 0.491263747215271,
+      "learning_rate": 0.000408504127756385,
+      "loss": 6.5125,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3542498818880728,
+      "grad_norm": 0.4879305362701416,
+      "learning_rate": 0.0004082721070480075,
+      "loss": 6.623,
+      "step": 1031
+    },
+    {
+      "epoch": 0.3545934802216209,
+      "grad_norm": 0.6748274564743042,
+      "learning_rate": 0.00040803985860289995,
+      "loss": 6.5595,
+      "step": 1032
+    },
+    {
+      "epoch": 0.354937078555169,
+      "grad_norm": 0.356755793094635,
+      "learning_rate": 0.0004078073827552432,
+      "loss": 6.5241,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3552806768887171,
+      "grad_norm": 0.6783245205879211,
+      "learning_rate": 0.0004075746798395452,
+      "loss": 6.5518,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3556242752222652,
+      "grad_norm": 0.5618282556533813,
+      "learning_rate": 0.0004073417501906407,
+      "loss": 6.526,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3559678735558133,
+      "grad_norm": 0.4738239347934723,
+      "learning_rate": 0.0004071085941436908,
+      "loss": 6.6759,
+      "step": 1036
+    },
+    {
+      "epoch": 0.35631147188936135,
+      "grad_norm": 0.7224136590957642,
+      "learning_rate": 0.00040687521203418216,
+      "loss": 6.519,
+      "step": 1037
+    },
+    {
+      "epoch": 0.35665507022290943,
+      "grad_norm": 0.638771116733551,
+      "learning_rate": 0.00040664160419792684,
+      "loss": 6.466,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3569986685564575,
+      "grad_norm": 0.562751293182373,
+      "learning_rate": 0.00040640777097106164,
+      "loss": 6.6204,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3573422668900056,
+      "grad_norm": 0.7001228928565979,
+      "learning_rate": 0.00040617371269004783,
+      "loss": 6.6701,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35768586522355367,
+      "grad_norm": 0.7808020114898682,
+      "learning_rate": 0.0004059394296916702,
+      "loss": 6.4939,
+      "step": 1041
+    },
+    {
+      "epoch": 0.35802946355710175,
+      "grad_norm": 0.8399685025215149,
+      "learning_rate": 0.00040570492231303725,
+      "loss": 6.5546,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3583730618906498,
+      "grad_norm": 0.659149169921875,
+      "learning_rate": 0.00040547019089158006,
+      "loss": 6.5908,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3587166602241979,
+      "grad_norm": 0.7955062389373779,
+      "learning_rate": 0.00040523523576505217,
+      "loss": 6.5679,
+      "step": 1044
+    },
+    {
+      "epoch": 0.359060258557746,
+      "grad_norm": 0.7686776518821716,
+      "learning_rate": 0.000405000057271529,
+      "loss": 6.4661,
+      "step": 1045
+    },
+    {
+      "epoch": 0.35940385689129406,
+      "grad_norm": 0.5270580053329468,
+      "learning_rate": 0.0004047646557494076,
+      "loss": 6.4503,
+      "step": 1046
+    },
+    {
+      "epoch": 0.35974745522484214,
+      "grad_norm": 0.7081395983695984,
+      "learning_rate": 0.0004045290315374054,
+      "loss": 6.5489,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3600910535583902,
+      "grad_norm": 0.8378854393959045,
+      "learning_rate": 0.00040429318497456075,
+      "loss": 6.7946,
+      "step": 1048
+    },
+    {
+      "epoch": 0.36043465189193835,
+      "grad_norm": 0.59283846616745,
+      "learning_rate": 0.00040405711640023183,
+      "loss": 6.6438,
+      "step": 1049
+    },
+    {
+      "epoch": 0.36077825022548643,
+      "grad_norm": 0.998915433883667,
+      "learning_rate": 0.0004038208261540961,
+      "loss": 6.4583,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3611218485590345,
+      "grad_norm": 1.1409025192260742,
+      "learning_rate": 0.0004035843145761502,
+      "loss": 6.3864,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3614654468925826,
+      "grad_norm": 0.7622621059417725,
+      "learning_rate": 0.0004033475820067091,
+      "loss": 6.552,
+      "step": 1052
+    },
+    {
+      "epoch": 0.36180904522613067,
+      "grad_norm": 0.8285176753997803,
+      "learning_rate": 0.0004031106287864057,
+      "loss": 6.4924,
+      "step": 1053
+    },
+    {
+      "epoch": 0.36215264355967874,
+      "grad_norm": 0.8217036724090576,
+      "learning_rate": 0.0004028734552561906,
+      "loss": 6.5873,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3624962418932268,
+      "grad_norm": 0.797195553779602,
+      "learning_rate": 0.00040263606175733124,
+      "loss": 6.4403,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3628398402267749,
+      "grad_norm": 0.6831180453300476,
+      "learning_rate": 0.00040239844863141163,
+      "loss": 6.5032,
+      "step": 1056
+    },
+    {
+      "epoch": 0.363183438560323,
+      "grad_norm": 0.6810872554779053,
+      "learning_rate": 0.0004021606162203318,
+      "loss": 6.4218,
+      "step": 1057
+    },
+    {
+      "epoch": 0.36352703689387106,
+      "grad_norm": 0.6327046155929565,
+      "learning_rate": 0.00040192256486630724,
+      "loss": 6.4264,
+      "step": 1058
+    },
+    {
+      "epoch": 0.36387063522741914,
+      "grad_norm": 0.5407282114028931,
+      "learning_rate": 0.0004016842949118686,
+      "loss": 6.4026,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3642142335609672,
+      "grad_norm": 0.7327606081962585,
+      "learning_rate": 0.000401445806699861,
+      "loss": 6.5312,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3645578318945153,
+      "grad_norm": 0.7049651145935059,
+      "learning_rate": 0.0004012071005734435,
+      "loss": 6.3978,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3649014302280634,
+      "grad_norm": 0.6945012211799622,
+      "learning_rate": 0.000400968176876089,
+      "loss": 6.4396,
+      "step": 1062
+    },
+    {
+      "epoch": 0.36524502856161145,
+      "grad_norm": 0.7001397609710693,
+      "learning_rate": 0.0004007290359515832,
+      "loss": 6.5293,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3655886268951596,
+      "grad_norm": 0.5994911193847656,
+      "learning_rate": 0.0004004896781440244,
+      "loss": 6.5305,
+      "step": 1064
+    },
+    {
+      "epoch": 0.36593222522870766,
+      "grad_norm": 0.4745972752571106,
+      "learning_rate": 0.0004002501037978232,
+      "loss": 6.5981,
+      "step": 1065
+    },
+    {
+      "epoch": 0.36627582356225574,
+      "grad_norm": 0.5303031802177429,
+      "learning_rate": 0.0004000103132577014,
+      "loss": 6.4155,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3666194218958038,
+      "grad_norm": 0.4650964140892029,
+      "learning_rate": 0.0003997703068686923,
+      "loss": 6.4578,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3669630202293519,
+      "grad_norm": 0.5369909405708313,
+      "learning_rate": 0.0003995300849761394,
+      "loss": 6.4916,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3673066185629,
+      "grad_norm": 0.4537830650806427,
+      "learning_rate": 0.00039928964792569654,
+      "loss": 6.4896,
+      "step": 1069
+    },
+    {
+      "epoch": 0.36765021689644806,
+      "grad_norm": 0.467717707157135,
+      "learning_rate": 0.0003990489960633271,
+      "loss": 6.4669,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36799381522999614,
+      "grad_norm": 0.5717526078224182,
+      "learning_rate": 0.00039880812973530335,
+      "loss": 6.39,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3683374135635442,
+      "grad_norm": 0.3943920135498047,
+      "learning_rate": 0.0003985670492882065,
+      "loss": 6.4894,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3686810118970923,
+      "grad_norm": 0.4640418291091919,
+      "learning_rate": 0.00039832575506892556,
+      "loss": 6.5758,
+      "step": 1073
+    },
+    {
+      "epoch": 0.36902461023064037,
+      "grad_norm": 0.527631938457489,
+      "learning_rate": 0.0003980842474246573,
+      "loss": 6.5324,
+      "step": 1074
+    },
+    {
+      "epoch": 0.36936820856418845,
+      "grad_norm": 0.5745725035667419,
+      "learning_rate": 0.00039784252670290555,
+      "loss": 6.4548,
+      "step": 1075
+    },
+    {
+      "epoch": 0.36971180689773653,
+      "grad_norm": 0.3347758948802948,
+      "learning_rate": 0.00039760059325148067,
+      "loss": 6.406,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3700554052312846,
+      "grad_norm": 0.42385727167129517,
+      "learning_rate": 0.0003973584474184992,
+      "loss": 6.5037,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3703990035648327,
+      "grad_norm": 0.3935813009738922,
+      "learning_rate": 0.00039711608955238334,
+      "loss": 6.5546,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3707426018983808,
+      "grad_norm": 0.6405526995658875,
+      "learning_rate": 0.00039687352000186005,
+      "loss": 6.4536,
+      "step": 1079
+    },
+    {
+      "epoch": 0.3710862002319289,
+      "grad_norm": 0.4569138288497925,
+      "learning_rate": 0.00039663073911596134,
+      "loss": 6.5994,
+      "step": 1080
+    },
+    {
+      "epoch": 0.371429798565477,
+      "grad_norm": 0.5814416408538818,
+      "learning_rate": 0.00039638774724402295,
+      "loss": 6.5257,
+      "step": 1081
+    },
+    {
+      "epoch": 0.37177339689902505,
+      "grad_norm": 0.604465126991272,
+      "learning_rate": 0.0003961445447356844,
+      "loss": 6.4128,
+      "step": 1082
+    },
+    {
+      "epoch": 0.37211699523257313,
+      "grad_norm": 0.45002633333206177,
+      "learning_rate": 0.00039590113194088827,
+      "loss": 6.4803,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3724605935661212,
+      "grad_norm": 0.6199561953544617,
+      "learning_rate": 0.00039565750920987966,
+      "loss": 6.5573,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3728041918996693,
+      "grad_norm": 0.45244845747947693,
+      "learning_rate": 0.00039541367689320566,
+      "loss": 6.5621,
+      "step": 1085
+    },
+    {
+      "epoch": 0.37314779023321737,
+      "grad_norm": 0.6887307167053223,
+      "learning_rate": 0.0003951696353417152,
+      "loss": 6.5448,
+      "step": 1086
+    },
+    {
+      "epoch": 0.37349138856676545,
+      "grad_norm": 0.5372146368026733,
+      "learning_rate": 0.000394925384906558,
+      "loss": 6.5137,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3738349869003135,
+      "grad_norm": 0.530670702457428,
+      "learning_rate": 0.0003946809259391846,
+      "loss": 6.5199,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3741785852338616,
+      "grad_norm": 0.7162120342254639,
+      "learning_rate": 0.00039443625879134525,
+      "loss": 6.473,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3745221835674097,
+      "grad_norm": 0.5712336897850037,
+      "learning_rate": 0.0003941913838150902,
+      "loss": 6.6048,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37486578190095776,
+      "grad_norm": 0.6182492971420288,
+      "learning_rate": 0.0003939463013627683,
+      "loss": 6.5619,
+      "step": 1091
+    },
+    {
+      "epoch": 0.37520938023450584,
+      "grad_norm": 0.6004253625869751,
+      "learning_rate": 0.00039370101178702724,
+      "loss": 6.5003,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3755529785680539,
+      "grad_norm": 0.7150577306747437,
+      "learning_rate": 0.00039345551544081256,
+      "loss": 6.5916,
+      "step": 1093
+    },
+    {
+      "epoch": 0.37589657690160205,
+      "grad_norm": 0.8572820425033569,
+      "learning_rate": 0.00039320981267736745,
+      "loss": 6.5192,
+      "step": 1094
+    },
+    {
+      "epoch": 0.37624017523515013,
+      "grad_norm": 0.6833814382553101,
+      "learning_rate": 0.00039296390385023204,
+      "loss": 6.5086,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3765837735686982,
+      "grad_norm": 0.7316805720329285,
+      "learning_rate": 0.00039271778931324297,
+      "loss": 6.5474,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3769273719022463,
+      "grad_norm": 0.6537583470344543,
+      "learning_rate": 0.00039247146942053297,
+      "loss": 6.4364,
+      "step": 1097
+    },
+    {
+      "epoch": 0.37727097023579437,
+      "grad_norm": 0.8147690892219543,
+      "learning_rate": 0.00039222494452653006,
+      "loss": 6.6332,
+      "step": 1098
+    },
+    {
+      "epoch": 0.37761456856934245,
+      "grad_norm": 0.6788108348846436,
+      "learning_rate": 0.00039197821498595744,
+      "loss": 6.6515,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3779581669028905,
+      "grad_norm": 0.886707603931427,
+      "learning_rate": 0.0003917312811538325,
+      "loss": 6.5627,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3783017652364386,
+      "grad_norm": 0.9681865572929382,
+      "learning_rate": 0.000391484143385467,
+      "loss": 6.4266,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3786453635699867,
+      "grad_norm": 0.6177529096603394,
+      "learning_rate": 0.0003912368020364657,
+      "loss": 6.3702,
+      "step": 1102
+    },
+    {
+      "epoch": 0.37898896190353476,
+      "grad_norm": 1.0723011493682861,
+      "learning_rate": 0.0003909892574627266,
+      "loss": 6.4111,
+      "step": 1103
+    },
+    {
+      "epoch": 0.37933256023708284,
+      "grad_norm": 0.870075523853302,
+      "learning_rate": 0.0003907415100204401,
+      "loss": 6.5443,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3796761585706309,
+      "grad_norm": 0.5722936987876892,
+      "learning_rate": 0.0003904935600660883,
+      "loss": 6.452,
+      "step": 1105
+    },
+    {
+      "epoch": 0.380019756904179,
+      "grad_norm": 0.8501458168029785,
+      "learning_rate": 0.0003902454079564447,
+      "loss": 6.3924,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3803633552377271,
+      "grad_norm": 0.7085288763046265,
+      "learning_rate": 0.0003899970540485741,
+      "loss": 6.3751,
+      "step": 1107
+    },
+    {
+      "epoch": 0.38070695357127515,
+      "grad_norm": 0.509347140789032,
+      "learning_rate": 0.00038974849869983114,
+      "loss": 6.4594,
+      "step": 1108
+    },
+    {
+      "epoch": 0.3810505519048233,
+      "grad_norm": 0.6944195032119751,
+      "learning_rate": 0.00038949974226786053,
+      "loss": 6.2957,
+      "step": 1109
+    },
+    {
+      "epoch": 0.38139415023837137,
+      "grad_norm": 0.5596153736114502,
+      "learning_rate": 0.0003892507851105965,
+      "loss": 6.3734,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38173774857191944,
+      "grad_norm": 0.5919499397277832,
+      "learning_rate": 0.0003890016275862618,
+      "loss": 6.4215,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3820813469054675,
+      "grad_norm": 0.4953160583972931,
+      "learning_rate": 0.0003887522700533675,
+      "loss": 6.3711,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3824249452390156,
+      "grad_norm": 0.580381453037262,
+      "learning_rate": 0.0003885027128707127,
+      "loss": 6.5075,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3827685435725637,
+      "grad_norm": 0.4716396629810333,
+      "learning_rate": 0.0003882529563973837,
+      "loss": 6.4973,
+      "step": 1114
+    },
+    {
+      "epoch": 0.38311214190611176,
+      "grad_norm": 0.5329322814941406,
+      "learning_rate": 0.00038800300099275345,
+      "loss": 6.4449,
+      "step": 1115
+    },
+    {
+      "epoch": 0.38345574023965984,
+      "grad_norm": 0.5385730862617493,
+      "learning_rate": 0.00038775284701648115,
+      "loss": 6.3797,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3837993385732079,
+      "grad_norm": 0.592823326587677,
+      "learning_rate": 0.00038750249482851184,
+      "loss": 6.579,
+      "step": 1117
+    },
+    {
+      "epoch": 0.384142936906756,
+      "grad_norm": 0.5093126893043518,
+      "learning_rate": 0.00038725194478907556,
+      "loss": 6.4524,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3844865352403041,
+      "grad_norm": 0.42565739154815674,
+      "learning_rate": 0.00038700119725868735,
+      "loss": 6.5539,
+      "step": 1119
+    },
+    {
+      "epoch": 0.38483013357385215,
+      "grad_norm": 0.4704289734363556,
+      "learning_rate": 0.00038675025259814606,
+      "loss": 6.3764,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38517373190740023,
+      "grad_norm": 0.46135643124580383,
+      "learning_rate": 0.00038649911116853456,
+      "loss": 6.4639,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3855173302409483,
+      "grad_norm": 0.5629793405532837,
+      "learning_rate": 0.0003862477733312185,
+      "loss": 6.3743,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3858609285744964,
+      "grad_norm": 0.41762539744377136,
+      "learning_rate": 0.0003859962394478464,
+      "loss": 6.443,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3862045269080445,
+      "grad_norm": 0.5560463666915894,
+      "learning_rate": 0.0003857445098803487,
+      "loss": 6.5066,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3865481252415926,
+      "grad_norm": 0.5698534250259399,
+      "learning_rate": 0.00038549258499093756,
+      "loss": 6.5049,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3868917235751407,
+      "grad_norm": 0.47976475954055786,
+      "learning_rate": 0.000385240465142106,
+      "loss": 6.4264,
+      "step": 1126
+    },
+    {
+      "epoch": 0.38723532190868876,
+      "grad_norm": 0.5374791026115417,
+      "learning_rate": 0.00038498815069662766,
+      "loss": 6.4272,
+      "step": 1127
+    },
+    {
+      "epoch": 0.38757892024223684,
+      "grad_norm": 0.4447578191757202,
+      "learning_rate": 0.0003847356420175564,
+      "loss": 6.3393,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3879225185757849,
+      "grad_norm": 0.5105568766593933,
+      "learning_rate": 0.0003844829394682251,
+      "loss": 6.4803,
+      "step": 1129
+    },
+    {
+      "epoch": 0.388266116909333,
+      "grad_norm": 0.5288437604904175,
+      "learning_rate": 0.00038423004341224597,
+      "loss": 6.4587,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38860971524288107,
+      "grad_norm": 0.4225218594074249,
+      "learning_rate": 0.00038397695421350954,
+      "loss": 6.4213,
+      "step": 1131
+    },
+    {
+      "epoch": 0.38895331357642915,
+      "grad_norm": 0.5961380004882812,
+      "learning_rate": 0.0003837236722361842,
+      "loss": 6.5326,
+      "step": 1132
+    },
+    {
+      "epoch": 0.38929691190997723,
+      "grad_norm": 0.527696967124939,
+      "learning_rate": 0.00038347019784471594,
+      "loss": 6.3996,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3896405102435253,
+      "grad_norm": 0.5186758041381836,
+      "learning_rate": 0.00038321653140382727,
+      "loss": 6.3858,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3899841085770734,
+      "grad_norm": 0.45041173696517944,
+      "learning_rate": 0.00038296267327851723,
+      "loss": 6.561,
+      "step": 1135
+    },
+    {
+      "epoch": 0.39032770691062146,
+      "grad_norm": 0.6820952892303467,
+      "learning_rate": 0.0003827086238340608,
+      "loss": 6.4386,
+      "step": 1136
+    },
+    {
+      "epoch": 0.39067130524416954,
+      "grad_norm": 0.6062747240066528,
+      "learning_rate": 0.00038245438343600805,
+      "loss": 6.3137,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3910149035777176,
+      "grad_norm": 0.4744030237197876,
+      "learning_rate": 0.0003821999524501837,
+      "loss": 6.3566,
+      "step": 1138
+    },
+    {
+      "epoch": 0.39135850191126575,
+      "grad_norm": 0.4075278639793396,
+      "learning_rate": 0.00038194533124268716,
+      "loss": 6.4366,
+      "step": 1139
+    },
+    {
+      "epoch": 0.39170210024481383,
+      "grad_norm": 0.5966638922691345,
+      "learning_rate": 0.0003816905201798912,
+      "loss": 6.4666,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3920456985783619,
+      "grad_norm": 0.6207997798919678,
+      "learning_rate": 0.0003814355196284417,
+      "loss": 6.5711,
+      "step": 1141
+    },
+    {
+      "epoch": 0.39238929691191,
+      "grad_norm": 0.5189718008041382,
+      "learning_rate": 0.00038118032995525746,
+      "loss": 6.4682,
+      "step": 1142
+    },
+    {
+      "epoch": 0.39273289524545807,
+      "grad_norm": 0.5805226564407349,
+      "learning_rate": 0.0003809249515275293,
+      "loss": 6.4373,
+      "step": 1143
+    },
+    {
+      "epoch": 0.39307649357900615,
+      "grad_norm": 0.5662277340888977,
+      "learning_rate": 0.00038066938471271946,
+      "loss": 6.5011,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3934200919125542,
+      "grad_norm": 0.6074782609939575,
+      "learning_rate": 0.00038041362987856155,
+      "loss": 6.5347,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3937636902461023,
+      "grad_norm": 0.6265222430229187,
+      "learning_rate": 0.00038015768739305946,
+      "loss": 6.5338,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3941072885796504,
+      "grad_norm": 0.6739917993545532,
+      "learning_rate": 0.0003799015576244874,
+      "loss": 6.4666,
+      "step": 1147
+    },
+    {
+      "epoch": 0.39445088691319846,
+      "grad_norm": 0.6248368620872498,
+      "learning_rate": 0.0003796452409413887,
+      "loss": 6.5219,
+      "step": 1148
+    },
+    {
+      "epoch": 0.39479448524674654,
+      "grad_norm": 0.7328528761863708,
+      "learning_rate": 0.00037938873771257585,
+      "loss": 6.4904,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3951380835802946,
+      "grad_norm": 0.9122981429100037,
+      "learning_rate": 0.0003791320483071298,
+      "loss": 6.5959,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3954816819138427,
+      "grad_norm": 0.8380439877510071,
+      "learning_rate": 0.0003788751730943991,
+      "loss": 6.5559,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3958252802473908,
+      "grad_norm": 0.6608699560165405,
+      "learning_rate": 0.0003786181124440001,
+      "loss": 6.6044,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3961688785809389,
+      "grad_norm": 0.7114856839179993,
+      "learning_rate": 0.0003783608667258156,
+      "loss": 6.5022,
+      "step": 1153
+    },
+    {
+      "epoch": 0.396512476914487,
+      "grad_norm": 0.5558148622512817,
+      "learning_rate": 0.0003781034363099949,
+      "loss": 6.5019,
+      "step": 1154
+    },
+    {
+      "epoch": 0.39685607524803507,
+      "grad_norm": 0.612568736076355,
+      "learning_rate": 0.00037784582156695284,
+      "loss": 6.4562,
+      "step": 1155
+    },
+    {
+      "epoch": 0.39719967358158315,
+      "grad_norm": 0.6382710337638855,
+      "learning_rate": 0.0003775880228673699,
+      "loss": 6.3982,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3975432719151312,
+      "grad_norm": 0.6008266806602478,
+      "learning_rate": 0.00037733004058219076,
+      "loss": 6.4837,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3978868702486793,
+      "grad_norm": 0.5972487926483154,
+      "learning_rate": 0.0003770718750826246,
+      "loss": 6.4001,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3982304685822274,
+      "grad_norm": 0.6197747588157654,
+      "learning_rate": 0.0003768135267401441,
+      "loss": 6.3853,
+      "step": 1159
+    },
+    {
+      "epoch": 0.39857406691577546,
+      "grad_norm": 0.7038761377334595,
+      "learning_rate": 0.00037655499592648513,
+      "loss": 6.5378,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39891766524932354,
+      "grad_norm": 0.5936171412467957,
+      "learning_rate": 0.0003762962830136458,
+      "loss": 6.2935,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3992612635828716,
+      "grad_norm": 0.7182568907737732,
+      "learning_rate": 0.00037603738837388667,
+      "loss": 6.2826,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3996048619164197,
+      "grad_norm": 0.655178427696228,
+      "learning_rate": 0.0003757783123797297,
+      "loss": 6.3054,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3999484602499678,
+      "grad_norm": 0.40302804112434387,
+      "learning_rate": 0.00037551905540395735,
+      "loss": 6.3942,
+      "step": 1164
+    },
+    {
+      "epoch": 0.40029205858351585,
+      "grad_norm": 0.7752393484115601,
+      "learning_rate": 0.0003752596178196131,
+      "loss": 6.252,
+      "step": 1165
+    },
+    {
+      "epoch": 0.40063565691706393,
+      "grad_norm": 0.5546431541442871,
+      "learning_rate": 0.000375,
+      "loss": 6.3494,
+      "step": 1166
+    },
+    {
+      "epoch": 0.400979255250612,
+      "grad_norm": 0.6124146580696106,
+      "learning_rate": 0.00037474020231868045,
+      "loss": 6.4457,
+      "step": 1167
+    },
+    {
+      "epoch": 0.40132285358416014,
+      "grad_norm": 0.5023613572120667,
+      "learning_rate": 0.00037448022514947573,
+      "loss": 6.3672,
+      "step": 1168
+    },
+    {
+      "epoch": 0.4016664519177082,
+      "grad_norm": 0.6657715439796448,
+      "learning_rate": 0.0003742200688664653,
+      "loss": 6.3568,
+      "step": 1169
+    },
+    {
+      "epoch": 0.4020100502512563,
+      "grad_norm": 0.44071656465530396,
+      "learning_rate": 0.00037395973384398627,
+      "loss": 6.5611,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4023536485848044,
+      "grad_norm": 0.5543152689933777,
+      "learning_rate": 0.00037369922045663327,
+      "loss": 6.3697,
+      "step": 1171
+    },
+    {
+      "epoch": 0.40269724691835246,
+      "grad_norm": 0.473361074924469,
+      "learning_rate": 0.0003734385290792573,
+      "loss": 6.4666,
+      "step": 1172
+    },
+    {
+      "epoch": 0.40304084525190054,
+      "grad_norm": 0.4800427556037903,
+      "learning_rate": 0.00037317766008696543,
+      "loss": 6.5445,
+      "step": 1173
+    },
+    {
+      "epoch": 0.4033844435854486,
+      "grad_norm": 0.6011155247688293,
+      "learning_rate": 0.0003729166138551204,
+      "loss": 6.4005,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4037280419189967,
+      "grad_norm": 0.47439199686050415,
+      "learning_rate": 0.00037265539075934014,
+      "loss": 6.4418,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4040716402525448,
+      "grad_norm": 0.7657064199447632,
+      "learning_rate": 0.00037239399117549676,
+      "loss": 6.4249,
+      "step": 1176
+    },
+    {
+      "epoch": 0.40441523858609285,
+      "grad_norm": 0.49792230129241943,
+      "learning_rate": 0.00037213241547971645,
+      "loss": 6.3721,
+      "step": 1177
+    },
+    {
+      "epoch": 0.40475883691964093,
+      "grad_norm": 0.7585155963897705,
+      "learning_rate": 0.0003718706640483789,
+      "loss": 6.4348,
+      "step": 1178
+    },
+    {
+      "epoch": 0.405102435253189,
+      "grad_norm": 0.4420984089374542,
+      "learning_rate": 0.0003716087372581165,
+      "loss": 6.3232,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4054460335867371,
+      "grad_norm": 0.8976798057556152,
+      "learning_rate": 0.0003713466354858141,
+      "loss": 6.2834,
+      "step": 1180
+    },
+    {
+      "epoch": 0.40578963192028517,
+      "grad_norm": 0.5315551161766052,
+      "learning_rate": 0.0003710843591086083,
+      "loss": 6.3475,
+      "step": 1181
+    },
+    {
+      "epoch": 0.40613323025383324,
+      "grad_norm": 0.8557246923446655,
+      "learning_rate": 0.0003708219085038869,
+      "loss": 6.2522,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4064768285873814,
+      "grad_norm": 0.7426571846008301,
+      "learning_rate": 0.0003705592840492883,
+      "loss": 6.416,
+      "step": 1183
+    },
+    {
+      "epoch": 0.40682042692092946,
+      "grad_norm": 0.5289968848228455,
+      "learning_rate": 0.0003702964861227013,
+      "loss": 6.3237,
+      "step": 1184
+    },
+    {
+      "epoch": 0.40716402525447754,
+      "grad_norm": 0.7140244841575623,
+      "learning_rate": 0.00037003351510226415,
+      "loss": 6.4297,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4075076235880256,
+      "grad_norm": 0.5616683959960938,
+      "learning_rate": 0.00036977037136636404,
+      "loss": 6.3243,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4078512219215737,
+      "grad_norm": 0.6869580745697021,
+      "learning_rate": 0.000369507055293637,
+      "loss": 6.3222,
+      "step": 1187
+    },
+    {
+      "epoch": 0.40819482025512177,
+      "grad_norm": 0.8005693554878235,
+      "learning_rate": 0.00036924356726296674,
+      "loss": 6.4359,
+      "step": 1188
+    },
+    {
+      "epoch": 0.40853841858866985,
+      "grad_norm": 0.6318978667259216,
+      "learning_rate": 0.00036897990765348467,
+      "loss": 6.4053,
+      "step": 1189
+    },
+    {
+      "epoch": 0.40888201692221793,
+      "grad_norm": 0.6056877374649048,
+      "learning_rate": 0.0003687160768445688,
+      "loss": 6.4243,
+      "step": 1190
+    },
+    {
+      "epoch": 0.409225615255766,
+      "grad_norm": 0.7091792225837708,
+      "learning_rate": 0.00036845207521584355,
+      "loss": 6.3087,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4095692135893141,
+      "grad_norm": 0.7672974467277527,
+      "learning_rate": 0.00036818790314717935,
+      "loss": 6.409,
+      "step": 1192
+    },
+    {
+      "epoch": 0.40991281192286216,
+      "grad_norm": 0.9131182432174683,
+      "learning_rate": 0.00036792356101869155,
+      "loss": 6.4735,
+      "step": 1193
+    },
+    {
+      "epoch": 0.41025641025641024,
+      "grad_norm": 0.5129940509796143,
+      "learning_rate": 0.00036765904921074046,
+      "loss": 6.5078,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4106000085899583,
+      "grad_norm": 1.1640437841415405,
+      "learning_rate": 0.0003673943681039305,
+      "loss": 6.3134,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4109436069235064,
+      "grad_norm": 0.6941258907318115,
+      "learning_rate": 0.0003671295180791094,
+      "loss": 6.4081,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4112872052570545,
+      "grad_norm": 0.8544447422027588,
+      "learning_rate": 0.0003668644995173684,
+      "loss": 6.3699,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4116308035906026,
+      "grad_norm": 0.8141912221908569,
+      "learning_rate": 0.000366599312800041,
+      "loss": 6.537,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4119744019241507,
+      "grad_norm": 0.9887378811836243,
+      "learning_rate": 0.0003663339583087025,
+      "loss": 6.6481,
+      "step": 1199
+    },
+    {
+      "epoch": 0.41231800025769877,
+      "grad_norm": 1.1278897523880005,
+      "learning_rate": 0.0003660684364251701,
+      "loss": 6.3531,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41266159859124685,
+      "grad_norm": 0.8282272219657898,
+      "learning_rate": 0.00036580274753150125,
+      "loss": 6.3874,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4130051969247949,
+      "grad_norm": 0.870682954788208,
+      "learning_rate": 0.00036553689200999426,
+      "loss": 6.3558,
+      "step": 1202
+    },
+    {
+      "epoch": 0.413348795258343,
+      "grad_norm": 0.9290775656700134,
+      "learning_rate": 0.00036527087024318676,
+      "loss": 6.3082,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4136923935918911,
+      "grad_norm": 0.6926900148391724,
+      "learning_rate": 0.0003650046826138559,
+      "loss": 6.4447,
+      "step": 1204
+    },
+    {
+      "epoch": 0.41403599192543916,
+      "grad_norm": 0.69936203956604,
+      "learning_rate": 0.0003647383295050173,
+      "loss": 6.3996,
+      "step": 1205
+    },
+    {
+      "epoch": 0.41437959025898724,
+      "grad_norm": 0.7394657731056213,
+      "learning_rate": 0.0003644718112999249,
+      "loss": 6.4317,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4147231885925353,
+      "grad_norm": 0.708120584487915,
+      "learning_rate": 0.0003642051283820699,
+      "loss": 6.3382,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4150667869260834,
+      "grad_norm": 0.6249948740005493,
+      "learning_rate": 0.00036393828113518063,
+      "loss": 6.4358,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4154103852596315,
+      "grad_norm": 0.6756424903869629,
+      "learning_rate": 0.00036367126994322195,
+      "loss": 6.3744,
+      "step": 1209
+    },
+    {
+      "epoch": 0.41575398359317955,
+      "grad_norm": 0.6610203385353088,
+      "learning_rate": 0.00036340409519039463,
+      "loss": 6.1553,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41609758192672763,
+      "grad_norm": 0.7560092210769653,
+      "learning_rate": 0.0003631367572611348,
+      "loss": 6.2694,
+      "step": 1211
+    },
+    {
+      "epoch": 0.4164411802602757,
+      "grad_norm": 0.5730945467948914,
+      "learning_rate": 0.00036286925654011303,
+      "loss": 6.397,
+      "step": 1212
+    },
+    {
+      "epoch": 0.41678477859382385,
+      "grad_norm": 0.6729722619056702,
+      "learning_rate": 0.0003626015934122346,
+      "loss": 6.5225,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4171283769273719,
+      "grad_norm": 0.5002995729446411,
+      "learning_rate": 0.0003623337682626383,
+      "loss": 6.4229,
+      "step": 1214
+    },
+    {
+      "epoch": 0.41747197526092,
+      "grad_norm": 0.5038610100746155,
+      "learning_rate": 0.000362065781476696,
+      "loss": 6.3297,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4178155735944681,
+      "grad_norm": 0.5231453776359558,
+      "learning_rate": 0.00036179763344001216,
+      "loss": 6.3065,
+      "step": 1216
+    },
+    {
+      "epoch": 0.41815917192801616,
+      "grad_norm": 0.541223406791687,
+      "learning_rate": 0.00036152932453842324,
+      "loss": 6.2675,
+      "step": 1217
+    },
+    {
+      "epoch": 0.41850277026156424,
+      "grad_norm": 0.48831766843795776,
+      "learning_rate": 0.00036126085515799744,
+      "loss": 6.4614,
+      "step": 1218
+    },
+    {
+      "epoch": 0.4188463685951123,
+      "grad_norm": 0.480375736951828,
+      "learning_rate": 0.00036099222568503335,
+      "loss": 6.3863,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4191899669286604,
+      "grad_norm": 0.4671798050403595,
+      "learning_rate": 0.00036072343650606043,
+      "loss": 6.3354,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4195335652622085,
+      "grad_norm": 0.7875792384147644,
+      "learning_rate": 0.00036045448800783766,
+      "loss": 6.351,
+      "step": 1221
+    },
+    {
+      "epoch": 0.41987716359575655,
+      "grad_norm": 0.5600533485412598,
+      "learning_rate": 0.0003601853805773533,
+      "loss": 6.3025,
+      "step": 1222
+    },
+    {
+      "epoch": 0.42022076192930463,
+      "grad_norm": 0.7269924879074097,
+      "learning_rate": 0.0003599161146018243,
+      "loss": 6.3016,
+      "step": 1223
+    },
+    {
+      "epoch": 0.4205643602628527,
+      "grad_norm": 0.4470710754394531,
+      "learning_rate": 0.00035964669046869587,
+      "loss": 6.4315,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4209079585964008,
+      "grad_norm": 0.6437981724739075,
+      "learning_rate": 0.00035937710856564055,
+      "loss": 6.4414,
+      "step": 1225
+    },
+    {
+      "epoch": 0.42125155692994887,
+      "grad_norm": 0.5562970042228699,
+      "learning_rate": 0.0003591073692805581,
+      "loss": 6.3088,
+      "step": 1226
+    },
+    {
+      "epoch": 0.42159515526349695,
+      "grad_norm": 0.37384918332099915,
+      "learning_rate": 0.00035883747300157463,
+      "loss": 6.4137,
+      "step": 1227
+    },
+    {
+      "epoch": 0.4219387535970451,
+      "grad_norm": 0.6116818785667419,
+      "learning_rate": 0.00035856742011704224,
+      "loss": 6.4572,
+      "step": 1228
+    },
+    {
+      "epoch": 0.42228235193059316,
+      "grad_norm": 0.5554793477058411,
+      "learning_rate": 0.0003582972110155383,
+      "loss": 6.414,
+      "step": 1229
+    },
+    {
+      "epoch": 0.42262595026414124,
+      "grad_norm": 0.4585476815700531,
+      "learning_rate": 0.0003580268460858649,
+      "loss": 6.3572,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4229695485976893,
+      "grad_norm": 0.4323978126049042,
+      "learning_rate": 0.00035775632571704853,
+      "loss": 6.3846,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4233131469312374,
+      "grad_norm": 0.725913405418396,
+      "learning_rate": 0.0003574856502983392,
+      "loss": 6.4419,
+      "step": 1232
+    },
+    {
+      "epoch": 0.4236567452647855,
+      "grad_norm": 0.5197131037712097,
+      "learning_rate": 0.00035721482021920995,
+      "loss": 6.3584,
+      "step": 1233
+    },
+    {
+      "epoch": 0.42400034359833355,
+      "grad_norm": 0.5760481357574463,
+      "learning_rate": 0.00035694383586935656,
+      "loss": 6.33,
+      "step": 1234
+    },
+    {
+      "epoch": 0.42434394193188163,
+      "grad_norm": 0.6064772009849548,
+      "learning_rate": 0.0003566726976386967,
+      "loss": 6.4748,
+      "step": 1235
+    },
+    {
+      "epoch": 0.4246875402654297,
+      "grad_norm": 0.5518282055854797,
+      "learning_rate": 0.0003564014059173694,
+      "loss": 6.4234,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4250311385989778,
+      "grad_norm": 0.637048065662384,
+      "learning_rate": 0.0003561299610957346,
+      "loss": 6.3062,
+      "step": 1237
+    },
+    {
+      "epoch": 0.42537473693252587,
+      "grad_norm": 0.6192359328269958,
+      "learning_rate": 0.00035585836356437264,
+      "loss": 6.4022,
+      "step": 1238
+    },
+    {
+      "epoch": 0.42571833526607394,
+      "grad_norm": 0.46911513805389404,
+      "learning_rate": 0.00035558661371408326,
+      "loss": 6.4925,
+      "step": 1239
+    },
+    {
+      "epoch": 0.426061933599622,
+      "grad_norm": 0.5996941328048706,
+      "learning_rate": 0.00035531471193588575,
+      "loss": 6.4118,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4264055319331701,
+      "grad_norm": 0.5280793905258179,
+      "learning_rate": 0.0003550426586210178,
+      "loss": 6.3713,
+      "step": 1241
+    },
+    {
+      "epoch": 0.4267491302667182,
+      "grad_norm": 0.5926524996757507,
+      "learning_rate": 0.0003547704541609353,
+      "loss": 6.2895,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4270927286002663,
+      "grad_norm": 0.7197567224502563,
+      "learning_rate": 0.00035449809894731136,
+      "loss": 6.6008,
+      "step": 1243
+    },
+    {
+      "epoch": 0.4274363269338144,
+      "grad_norm": 0.6966863870620728,
+      "learning_rate": 0.0003542255933720363,
+      "loss": 6.3742,
+      "step": 1244
+    },
+    {
+      "epoch": 0.42777992526736247,
+      "grad_norm": 0.8108692765235901,
+      "learning_rate": 0.0003539529378272166,
+      "loss": 6.4517,
+      "step": 1245
+    },
+    {
+      "epoch": 0.42812352360091055,
+      "grad_norm": 0.7402332425117493,
+      "learning_rate": 0.0003536801327051746,
+      "loss": 6.3478,
+      "step": 1246
+    },
+    {
+      "epoch": 0.42846712193445863,
+      "grad_norm": 0.7129719853401184,
+      "learning_rate": 0.0003534071783984479,
+      "loss": 6.3975,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4288107202680067,
+      "grad_norm": 0.7612306475639343,
+      "learning_rate": 0.0003531340752997886,
+      "loss": 6.3858,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4291543186015548,
+      "grad_norm": 0.8239126801490784,
+      "learning_rate": 0.00035286082380216313,
+      "loss": 6.4085,
+      "step": 1249
+    },
+    {
+      "epoch": 0.42949791693510286,
+      "grad_norm": 1.0187656879425049,
+      "learning_rate": 0.00035258742429875137,
+      "loss": 6.4928,
+      "step": 1250
+    },
+    {
+      "epoch": 0.42984151526865094,
+      "grad_norm": 0.8988327980041504,
+      "learning_rate": 0.00035231387718294595,
+      "loss": 6.2908,
+      "step": 1251
+    },
+    {
+      "epoch": 0.430185113602199,
+      "grad_norm": 0.941773533821106,
+      "learning_rate": 0.00035204018284835226,
+      "loss": 6.3999,
+      "step": 1252
+    },
+    {
+      "epoch": 0.4305287119357471,
+      "grad_norm": 0.879536509513855,
+      "learning_rate": 0.00035176634168878723,
+      "loss": 6.3897,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4308723102692952,
+      "grad_norm": 0.658165454864502,
+      "learning_rate": 0.0003514923540982793,
+      "loss": 6.2647,
+      "step": 1254
+    },
+    {
+      "epoch": 0.43121590860284326,
+      "grad_norm": 0.9977064728736877,
+      "learning_rate": 0.0003512182204710673,
+      "loss": 6.2759,
+      "step": 1255
+    },
+    {
+      "epoch": 0.43155950693639133,
+      "grad_norm": 0.8180080652236938,
+      "learning_rate": 0.0003509439412016004,
+      "loss": 6.3759,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4319031052699394,
+      "grad_norm": 0.8202301859855652,
+      "learning_rate": 0.00035066951668453745,
+      "loss": 6.4117,
+      "step": 1257
+    },
+    {
+      "epoch": 0.43224670360348755,
+      "grad_norm": 0.6717396378517151,
+      "learning_rate": 0.000350394947314746,
+      "loss": 6.2032,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4325903019370356,
+      "grad_norm": 0.7441490888595581,
+      "learning_rate": 0.0003501202334873021,
+      "loss": 6.3116,
+      "step": 1259
+    },
+    {
+      "epoch": 0.4329339002705837,
+      "grad_norm": 0.7404760122299194,
+      "learning_rate": 0.00034984537559749,
+      "loss": 6.3653,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4332774986041318,
+      "grad_norm": 0.5555077791213989,
+      "learning_rate": 0.0003495703740408008,
+      "loss": 6.3558,
+      "step": 1261
+    },
+    {
+      "epoch": 0.43362109693767986,
+      "grad_norm": 0.8754700422286987,
+      "learning_rate": 0.00034929522921293244,
+      "loss": 6.3524,
+      "step": 1262
+    },
+    {
+      "epoch": 0.43396469527122794,
+      "grad_norm": 0.5776944756507874,
+      "learning_rate": 0.00034901994150978924,
+      "loss": 6.2332,
+      "step": 1263
+    },
+    {
+      "epoch": 0.434308293604776,
+      "grad_norm": 0.5266901850700378,
+      "learning_rate": 0.00034874451132748074,
+      "loss": 6.3565,
+      "step": 1264
+    },
+    {
+      "epoch": 0.4346518919383241,
+      "grad_norm": 0.5814431309700012,
+      "learning_rate": 0.0003484689390623218,
+      "loss": 6.4379,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4349954902718722,
+      "grad_norm": 0.6413130164146423,
+      "learning_rate": 0.0003481932251108316,
+      "loss": 6.3527,
+      "step": 1266
+    },
+    {
+      "epoch": 0.43533908860542025,
+      "grad_norm": 0.5977810025215149,
+      "learning_rate": 0.0003479173698697331,
+      "loss": 6.2953,
+      "step": 1267
+    },
+    {
+      "epoch": 0.43568268693896833,
+      "grad_norm": 0.5503946542739868,
+      "learning_rate": 0.0003476413737359527,
+      "loss": 6.362,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4360262852725164,
+      "grad_norm": 0.6731773614883423,
+      "learning_rate": 0.00034736523710661964,
+      "loss": 6.2602,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4363698836060645,
+      "grad_norm": 0.5576619505882263,
+      "learning_rate": 0.000347088960379065,
+      "loss": 6.3608,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43671348193961257,
+      "grad_norm": 0.580730676651001,
+      "learning_rate": 0.00034681254395082156,
+      "loss": 6.3143,
+      "step": 1271
+    },
+    {
+      "epoch": 0.43705708027316065,
+      "grad_norm": 0.5770998001098633,
+      "learning_rate": 0.0003465359882196233,
+      "loss": 6.364,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4374006786067088,
+      "grad_norm": 0.6039701700210571,
+      "learning_rate": 0.0003462592935834044,
+      "loss": 6.2892,
+      "step": 1273
+    },
+    {
+      "epoch": 0.43774427694025686,
+      "grad_norm": 0.6770427823066711,
+      "learning_rate": 0.00034598246044029906,
+      "loss": 6.2845,
+      "step": 1274
+    },
+    {
+      "epoch": 0.43808787527380494,
+      "grad_norm": 0.4633253216743469,
+      "learning_rate": 0.00034570548918864074,
+      "loss": 6.3737,
+      "step": 1275
+    },
+    {
+      "epoch": 0.438431473607353,
+      "grad_norm": 0.6166013479232788,
+      "learning_rate": 0.0003454283802269617,
+      "loss": 6.3048,
+      "step": 1276
+    },
+    {
+      "epoch": 0.4387750719409011,
+      "grad_norm": 0.6878900527954102,
+      "learning_rate": 0.0003451511339539921,
+      "loss": 6.1519,
+      "step": 1277
+    },
+    {
+      "epoch": 0.4391186702744492,
+      "grad_norm": 0.5486677289009094,
+      "learning_rate": 0.0003448737507686599,
+      "loss": 6.2707,
+      "step": 1278
+    },
+    {
+      "epoch": 0.43946226860799725,
+      "grad_norm": 0.6651487946510315,
+      "learning_rate": 0.00034459623107009006,
+      "loss": 6.443,
+      "step": 1279
+    },
+    {
+      "epoch": 0.43980586694154533,
+      "grad_norm": 0.6062619090080261,
+      "learning_rate": 0.00034431857525760385,
+      "loss": 6.2852,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4401494652750934,
+      "grad_norm": 0.5092039704322815,
+      "learning_rate": 0.00034404078373071845,
+      "loss": 6.4106,
+      "step": 1281
+    },
+    {
+      "epoch": 0.4404930636086415,
+      "grad_norm": 0.4990077316761017,
+      "learning_rate": 0.00034376285688914645,
+      "loss": 6.4131,
+      "step": 1282
+    },
+    {
+      "epoch": 0.44083666194218957,
+      "grad_norm": 0.6572175621986389,
+      "learning_rate": 0.00034348479513279486,
+      "loss": 6.2638,
+      "step": 1283
+    },
+    {
+      "epoch": 0.44118026027573765,
+      "grad_norm": 0.5944482088088989,
+      "learning_rate": 0.000343206598861765,
+      "loss": 6.4179,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4415238586092857,
+      "grad_norm": 0.6007838249206543,
+      "learning_rate": 0.0003429282684763519,
+      "loss": 6.268,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4418674569428338,
+      "grad_norm": 0.528721034526825,
+      "learning_rate": 0.0003426498043770432,
+      "loss": 6.4174,
+      "step": 1286
+    },
+    {
+      "epoch": 0.44221105527638194,
+      "grad_norm": 0.515328586101532,
+      "learning_rate": 0.00034237120696451904,
+      "loss": 6.3786,
+      "step": 1287
+    },
+    {
+      "epoch": 0.44255465360993,
+      "grad_norm": 0.7291128635406494,
+      "learning_rate": 0.0003420924766396517,
+      "loss": 6.3848,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4428982519434781,
+      "grad_norm": 0.5647246837615967,
+      "learning_rate": 0.0003418136138035044,
+      "loss": 6.3112,
+      "step": 1289
+    },
+    {
+      "epoch": 0.4432418502770262,
+      "grad_norm": 0.6985518336296082,
+      "learning_rate": 0.000341534618857331,
+      "loss": 6.3603,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44358544861057425,
+      "grad_norm": 0.6250963807106018,
+      "learning_rate": 0.0003412554922025756,
+      "loss": 6.4102,
+      "step": 1291
+    },
+    {
+      "epoch": 0.44392904694412233,
+      "grad_norm": 0.8003494739532471,
+      "learning_rate": 0.0003409762342408719,
+      "loss": 6.279,
+      "step": 1292
+    },
+    {
+      "epoch": 0.4442726452776704,
+      "grad_norm": 0.6796652674674988,
+      "learning_rate": 0.0003406968453740423,
+      "loss": 6.4095,
+      "step": 1293
+    },
+    {
+      "epoch": 0.4446162436112185,
+      "grad_norm": 0.7780187726020813,
+      "learning_rate": 0.0003404173260040976,
+      "loss": 6.4159,
+      "step": 1294
+    },
+    {
+      "epoch": 0.44495984194476657,
+      "grad_norm": 0.9151715040206909,
+      "learning_rate": 0.0003401376765332366,
+      "loss": 6.3968,
+      "step": 1295
+    },
+    {
+      "epoch": 0.44530344027831464,
+      "grad_norm": 0.8557326197624207,
+      "learning_rate": 0.000339857897363845,
+      "loss": 6.3813,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4456470386118627,
+      "grad_norm": 0.8565590977668762,
+      "learning_rate": 0.0003395779888984954,
+      "loss": 6.3745,
+      "step": 1297
+    },
+    {
+      "epoch": 0.4459906369454108,
+      "grad_norm": 0.7628725171089172,
+      "learning_rate": 0.00033929795153994624,
+      "loss": 6.4808,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4463342352789589,
+      "grad_norm": 0.8995116353034973,
+      "learning_rate": 0.00033901778569114154,
+      "loss": 6.4562,
+      "step": 1299
+    },
+    {
+      "epoch": 0.44667783361250696,
+      "grad_norm": 1.1376203298568726,
+      "learning_rate": 0.0003387374917552101,
+      "loss": 6.4389,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44702143194605504,
+      "grad_norm": 0.8454490303993225,
+      "learning_rate": 0.0003384570701354652,
+      "loss": 6.2257,
+      "step": 1301
+    },
+    {
+      "epoch": 0.44736503027960317,
+      "grad_norm": 0.9702429175376892,
+      "learning_rate": 0.0003381765212354036,
+      "loss": 6.2489,
+      "step": 1302
+    },
+    {
+      "epoch": 0.44770862861315125,
+      "grad_norm": 0.8661314845085144,
+      "learning_rate": 0.0003378958454587054,
+      "loss": 6.2794,
+      "step": 1303
+    },
+    {
+      "epoch": 0.44805222694669933,
+      "grad_norm": 0.6850778460502625,
+      "learning_rate": 0.00033761504320923316,
+      "loss": 6.2974,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4483958252802474,
+      "grad_norm": 0.8128910660743713,
+      "learning_rate": 0.0003373341148910315,
+      "loss": 6.3021,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4487394236137955,
+      "grad_norm": 0.6577335596084595,
+      "learning_rate": 0.00033705306090832626,
+      "loss": 6.3377,
+      "step": 1306
+    },
+    {
+      "epoch": 0.44908302194734356,
+      "grad_norm": 0.7702295184135437,
+      "learning_rate": 0.0003367718816655244,
+      "loss": 6.2775,
+      "step": 1307
+    },
+    {
+      "epoch": 0.44942662028089164,
+      "grad_norm": 0.7849611043930054,
+      "learning_rate": 0.0003364905775672129,
+      "loss": 6.1687,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4497702186144397,
+      "grad_norm": 0.5363289713859558,
+      "learning_rate": 0.00033620914901815835,
+      "loss": 6.3946,
+      "step": 1309
+    },
+    {
+      "epoch": 0.4501138169479878,
+      "grad_norm": 0.8963726162910461,
+      "learning_rate": 0.0003359275964233066,
+      "loss": 6.3973,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4504574152815359,
+      "grad_norm": 0.6057178974151611,
+      "learning_rate": 0.0003356459201877819,
+      "loss": 6.1898,
+      "step": 1311
+    },
+    {
+      "epoch": 0.45080101361508396,
+      "grad_norm": 0.7307539582252502,
+      "learning_rate": 0.00033536412071688635,
+      "loss": 6.2201,
+      "step": 1312
+    },
+    {
+      "epoch": 0.45114461194863203,
+      "grad_norm": 0.7102933526039124,
+      "learning_rate": 0.0003350821984160994,
+      "loss": 6.3939,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4514882102821801,
+      "grad_norm": 0.41668689250946045,
+      "learning_rate": 0.00033480015369107734,
+      "loss": 6.3618,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4518318086157282,
+      "grad_norm": 0.5837572813034058,
+      "learning_rate": 0.00033451798694765256,
+      "loss": 6.2348,
+      "step": 1315
+    },
+    {
+      "epoch": 0.45217540694927627,
+      "grad_norm": 0.5444297790527344,
+      "learning_rate": 0.00033423569859183277,
+      "loss": 6.2863,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4525190052828244,
+      "grad_norm": 0.5512393116950989,
+      "learning_rate": 0.00033395328902980113,
+      "loss": 6.227,
+      "step": 1317
+    },
+    {
+      "epoch": 0.4528626036163725,
+      "grad_norm": 0.6816386580467224,
+      "learning_rate": 0.00033367075866791484,
+      "loss": 6.2635,
+      "step": 1318
+    },
+    {
+      "epoch": 0.45320620194992056,
+      "grad_norm": 0.5417824387550354,
+      "learning_rate": 0.0003333881079127052,
+      "loss": 6.2983,
+      "step": 1319
+    },
+    {
+      "epoch": 0.45354980028346864,
+      "grad_norm": 0.5983556509017944,
+      "learning_rate": 0.00033310533717087633,
+      "loss": 6.2582,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4538933986170167,
+      "grad_norm": 0.5117073059082031,
+      "learning_rate": 0.00033282244684930553,
+      "loss": 6.356,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4542369969505648,
+      "grad_norm": 0.5673801302909851,
+      "learning_rate": 0.0003325394373550416,
+      "loss": 6.2926,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4545805952841129,
+      "grad_norm": 0.6815798282623291,
+      "learning_rate": 0.00033225630909530535,
+      "loss": 6.3458,
+      "step": 1323
+    },
+    {
+      "epoch": 0.45492419361766095,
+      "grad_norm": 0.6034033894538879,
+      "learning_rate": 0.0003319730624774881,
+      "loss": 6.3044,
+      "step": 1324
+    },
+    {
+      "epoch": 0.45526779195120903,
+      "grad_norm": 0.6391438245773315,
+      "learning_rate": 0.0003316896979091517,
+      "loss": 6.312,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4556113902847571,
+      "grad_norm": 0.5472264289855957,
+      "learning_rate": 0.0003314062157980275,
+      "loss": 6.279,
+      "step": 1326
+    },
+    {
+      "epoch": 0.4559549886183052,
+      "grad_norm": 0.5441725850105286,
+      "learning_rate": 0.0003311226165520163,
+      "loss": 6.2368,
+      "step": 1327
+    },
+    {
+      "epoch": 0.45629858695185327,
+      "grad_norm": 0.5921383500099182,
+      "learning_rate": 0.00033083890057918714,
+      "loss": 6.2127,
+      "step": 1328
+    },
+    {
+      "epoch": 0.45664218528540135,
+      "grad_norm": 0.5410550832748413,
+      "learning_rate": 0.0003305550682877771,
+      "loss": 6.3174,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4569857836189494,
+      "grad_norm": 0.7097800374031067,
+      "learning_rate": 0.0003302711200861907,
+      "loss": 6.2349,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4573293819524975,
+      "grad_norm": 0.6367291808128357,
+      "learning_rate": 0.00032998705638299925,
+      "loss": 6.2174,
+      "step": 1331
+    },
+    {
+      "epoch": 0.45767298028604564,
+      "grad_norm": 0.5397276878356934,
+      "learning_rate": 0.0003297028775869401,
+      "loss": 6.412,
+      "step": 1332
+    },
+    {
+      "epoch": 0.4580165786195937,
+      "grad_norm": 0.7171114683151245,
+      "learning_rate": 0.0003294185841069165,
+      "loss": 6.3053,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4583601769531418,
+      "grad_norm": 0.5611921548843384,
+      "learning_rate": 0.0003291341763519963,
+      "loss": 6.2494,
+      "step": 1334
+    },
+    {
+      "epoch": 0.4587037752866899,
+      "grad_norm": 0.5535762906074524,
+      "learning_rate": 0.0003288496547314122,
+      "loss": 6.3483,
+      "step": 1335
+    },
+    {
+      "epoch": 0.45904737362023795,
+      "grad_norm": 0.5840194225311279,
+      "learning_rate": 0.00032856501965456043,
+      "loss": 6.4088,
+      "step": 1336
+    },
+    {
+      "epoch": 0.45939097195378603,
+      "grad_norm": 0.5516507029533386,
+      "learning_rate": 0.00032828027153100067,
+      "loss": 6.362,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4597345702873341,
+      "grad_norm": 0.5980287194252014,
+      "learning_rate": 0.0003279954107704551,
+      "loss": 6.2421,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4600781686208822,
+      "grad_norm": 0.6924148201942444,
+      "learning_rate": 0.00032771043778280826,
+      "loss": 6.1814,
+      "step": 1339
+    },
+    {
+      "epoch": 0.46042176695443027,
+      "grad_norm": 0.5325207114219666,
+      "learning_rate": 0.00032742535297810573,
+      "loss": 6.3799,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46076536528797835,
+      "grad_norm": 0.7110170722007751,
+      "learning_rate": 0.0003271401567665544,
+      "loss": 6.4423,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4611089636215264,
+      "grad_norm": 0.7545775771141052,
+      "learning_rate": 0.0003268548495585212,
+      "loss": 6.3796,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4614525619550745,
+      "grad_norm": 0.6151261329650879,
+      "learning_rate": 0.0003265694317645328,
+      "loss": 6.3873,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4617961602886226,
+      "grad_norm": 0.5837677121162415,
+      "learning_rate": 0.00032628390379527524,
+      "loss": 6.5048,
+      "step": 1344
+    },
+    {
+      "epoch": 0.46213975862217066,
+      "grad_norm": 0.6211174130439758,
+      "learning_rate": 0.0003259982660615927,
+      "loss": 6.6132,
+      "step": 1345
+    },
+    {
+      "epoch": 0.46248335695571874,
+      "grad_norm": 0.6927919387817383,
+      "learning_rate": 0.00032571251897448765,
+      "loss": 6.3105,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4628269552892669,
+      "grad_norm": 0.796061098575592,
+      "learning_rate": 0.0003254266629451198,
+      "loss": 6.4112,
+      "step": 1347
+    },
+    {
+      "epoch": 0.46317055362281495,
+      "grad_norm": 0.8083237409591675,
+      "learning_rate": 0.00032514069838480536,
+      "loss": 6.3347,
+      "step": 1348
+    },
+    {
+      "epoch": 0.46351415195636303,
+      "grad_norm": 0.7419441938400269,
+      "learning_rate": 0.0003248546257050171,
+      "loss": 6.3444,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4638577502899111,
+      "grad_norm": 0.9289708733558655,
+      "learning_rate": 0.00032456844531738313,
+      "loss": 6.4263,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4642013486234592,
+      "grad_norm": 0.8992305397987366,
+      "learning_rate": 0.00032428215763368655,
+      "loss": 6.366,
+      "step": 1351
+    },
+    {
+      "epoch": 0.46454494695700727,
+      "grad_norm": 0.6328170895576477,
+      "learning_rate": 0.00032399576306586493,
+      "loss": 6.2029,
+      "step": 1352
+    },
+    {
+      "epoch": 0.46488854529055534,
+      "grad_norm": 0.48920828104019165,
+      "learning_rate": 0.0003237092620260096,
+      "loss": 6.2901,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4652321436241034,
+      "grad_norm": 0.6819382309913635,
+      "learning_rate": 0.0003234226549263651,
+      "loss": 6.1887,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4655757419576515,
+      "grad_norm": 0.6389537453651428,
+      "learning_rate": 0.0003231359421793286,
+      "loss": 6.2041,
+      "step": 1355
+    },
+    {
+      "epoch": 0.4659193402911996,
+      "grad_norm": 0.6682074666023254,
+      "learning_rate": 0.00032284912419744904,
+      "loss": 6.1585,
+      "step": 1356
+    },
+    {
+      "epoch": 0.46626293862474766,
+      "grad_norm": 0.5480344891548157,
+      "learning_rate": 0.0003225622013934273,
+      "loss": 6.2923,
+      "step": 1357
+    },
+    {
+      "epoch": 0.46660653695829574,
+      "grad_norm": 0.6636848449707031,
+      "learning_rate": 0.00032227517418011457,
+      "loss": 6.1032,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4669501352918438,
+      "grad_norm": 0.7700697779655457,
+      "learning_rate": 0.00032198804297051256,
+      "loss": 6.2676,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4672937336253919,
+      "grad_norm": 0.776321291923523,
+      "learning_rate": 0.0003217008081777726,
+      "loss": 6.1532,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46763733195894,
+      "grad_norm": 0.5545940399169922,
+      "learning_rate": 0.00032141347021519485,
+      "loss": 6.2717,
+      "step": 1361
+    },
+    {
+      "epoch": 0.4679809302924881,
+      "grad_norm": 0.5399784445762634,
+      "learning_rate": 0.0003211260294962282,
+      "loss": 6.3534,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4683245286260362,
+      "grad_norm": 0.6080695390701294,
+      "learning_rate": 0.00032083848643446936,
+      "loss": 6.338,
+      "step": 1363
+    },
+    {
+      "epoch": 0.46866812695958426,
+      "grad_norm": 0.6340306997299194,
+      "learning_rate": 0.00032055084144366194,
+      "loss": 6.4141,
+      "step": 1364
+    },
+    {
+      "epoch": 0.46901172529313234,
+      "grad_norm": 0.5792441368103027,
+      "learning_rate": 0.0003202630949376968,
+      "loss": 6.2975,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4693553236266804,
+      "grad_norm": 0.7110378742218018,
+      "learning_rate": 0.00031997524733061027,
+      "loss": 6.3343,
+      "step": 1366
+    },
+    {
+      "epoch": 0.4696989219602285,
+      "grad_norm": 0.5993596911430359,
+      "learning_rate": 0.0003196872990365847,
+      "loss": 6.2514,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4700425202937766,
+      "grad_norm": 0.5569903254508972,
+      "learning_rate": 0.00031939925046994686,
+      "loss": 6.3676,
+      "step": 1368
+    },
+    {
+      "epoch": 0.47038611862732466,
+      "grad_norm": 0.5726590156555176,
+      "learning_rate": 0.0003191111020451682,
+      "loss": 6.2824,
+      "step": 1369
+    },
+    {
+      "epoch": 0.47072971696087273,
+      "grad_norm": 0.5852641463279724,
+      "learning_rate": 0.00031882285417686354,
+      "loss": 6.1269,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4710733152944208,
+      "grad_norm": 0.5540098547935486,
+      "learning_rate": 0.0003185345072797909,
+      "loss": 6.2548,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4714169136279689,
+      "grad_norm": 0.6843763589859009,
+      "learning_rate": 0.0003182460617688508,
+      "loss": 6.2628,
+      "step": 1372
+    },
+    {
+      "epoch": 0.47176051196151697,
+      "grad_norm": 0.6013829708099365,
+      "learning_rate": 0.00031795751805908576,
+      "loss": 6.1731,
+      "step": 1373
+    },
+    {
+      "epoch": 0.47210411029506505,
+      "grad_norm": 0.667677104473114,
+      "learning_rate": 0.0003176688765656793,
+      "loss": 6.2667,
+      "step": 1374
+    },
+    {
+      "epoch": 0.47244770862861313,
+      "grad_norm": 0.4927704334259033,
+      "learning_rate": 0.000317380137703956,
+      "loss": 6.1375,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4727913069621612,
+      "grad_norm": 0.4976654052734375,
+      "learning_rate": 0.0003170913018893804,
+      "loss": 6.3665,
+      "step": 1376
+    },
+    {
+      "epoch": 0.47313490529570934,
+      "grad_norm": 0.5618501901626587,
+      "learning_rate": 0.0003168023695375563,
+      "loss": 6.2159,
+      "step": 1377
+    },
+    {
+      "epoch": 0.4734785036292574,
+      "grad_norm": 0.48412713408470154,
+      "learning_rate": 0.0003165133410642268,
+      "loss": 6.3496,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4738221019628055,
+      "grad_norm": 0.44460970163345337,
+      "learning_rate": 0.0003162242168852732,
+      "loss": 6.3743,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4741657002963536,
+      "grad_norm": 0.4922606647014618,
+      "learning_rate": 0.0003159349974167143,
+      "loss": 6.3456,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47450929862990165,
+      "grad_norm": 0.5359041690826416,
+      "learning_rate": 0.00031564568307470615,
+      "loss": 6.2156,
+      "step": 1381
+    },
+    {
+      "epoch": 0.47485289696344973,
+      "grad_norm": 0.5539677739143372,
+      "learning_rate": 0.00031535627427554144,
+      "loss": 6.1331,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4751964952969978,
+      "grad_norm": 0.5184550881385803,
+      "learning_rate": 0.00031506677143564856,
+      "loss": 6.2359,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4755400936305459,
+      "grad_norm": 0.5449308156967163,
+      "learning_rate": 0.00031477717497159133,
+      "loss": 6.2612,
+      "step": 1384
+    },
+    {
+      "epoch": 0.47588369196409397,
+      "grad_norm": 0.6895928978919983,
+      "learning_rate": 0.0003144874853000682,
+      "loss": 6.2665,
+      "step": 1385
+    },
+    {
+      "epoch": 0.47622729029764205,
+      "grad_norm": 0.5198205709457397,
+      "learning_rate": 0.000314197702837912,
+      "loss": 6.2982,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4765708886311901,
+      "grad_norm": 0.622516393661499,
+      "learning_rate": 0.00031390782800208865,
+      "loss": 6.3573,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4769144869647382,
+      "grad_norm": 0.7482896447181702,
+      "learning_rate": 0.00031361786120969734,
+      "loss": 6.3066,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4772580852982863,
+      "grad_norm": 0.6616333723068237,
+      "learning_rate": 0.0003133278028779695,
+      "loss": 6.2663,
+      "step": 1389
+    },
+    {
+      "epoch": 0.47760168363183436,
+      "grad_norm": 0.6296412944793701,
+      "learning_rate": 0.000313037653424268,
+      "loss": 6.2826,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47794528196538244,
+      "grad_norm": 0.7290968298912048,
+      "learning_rate": 0.0003127474132660872,
+      "loss": 6.4081,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4782888802989306,
+      "grad_norm": 0.6888962388038635,
+      "learning_rate": 0.0003124570828210518,
+      "loss": 6.3042,
+      "step": 1392
+    },
+    {
+      "epoch": 0.47863247863247865,
+      "grad_norm": 0.5802988409996033,
+      "learning_rate": 0.0003121666625069165,
+      "loss": 6.4235,
+      "step": 1393
+    },
+    {
+      "epoch": 0.47897607696602673,
+      "grad_norm": 0.7673103213310242,
+      "learning_rate": 0.0003118761527415651,
+      "loss": 6.3122,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4793196752995748,
+      "grad_norm": 0.9014639854431152,
+      "learning_rate": 0.0003115855539430104,
+      "loss": 6.1529,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4796632736331229,
+      "grad_norm": 0.7389999032020569,
+      "learning_rate": 0.0003112948665293931,
+      "loss": 6.2241,
+      "step": 1396
+    },
+    {
+      "epoch": 0.48000687196667097,
+      "grad_norm": 0.8546797633171082,
+      "learning_rate": 0.0003110040909189815,
+      "loss": 6.457,
+      "step": 1397
+    },
+    {
+      "epoch": 0.48035047030021905,
+      "grad_norm": 0.8488891124725342,
+      "learning_rate": 0.0003107132275301707,
+      "loss": 6.4161,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4806940686337671,
+      "grad_norm": 0.7810699939727783,
+      "learning_rate": 0.0003104222767814823,
+      "loss": 6.434,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4810376669673152,
+      "grad_norm": 1.0687366724014282,
+      "learning_rate": 0.00031013123909156344,
+      "loss": 6.5133,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4813812653008633,
+      "grad_norm": 0.9212234020233154,
+      "learning_rate": 0.0003098401148791863,
+      "loss": 6.223,
+      "step": 1401
+    },
+    {
+      "epoch": 0.48172486363441136,
+      "grad_norm": 0.7973336577415466,
+      "learning_rate": 0.0003095489045632479,
+      "loss": 6.1975,
+      "step": 1402
+    },
+    {
+      "epoch": 0.48206846196795944,
+      "grad_norm": 0.6696678400039673,
+      "learning_rate": 0.00030925760856276866,
+      "loss": 6.3295,
+      "step": 1403
+    },
+    {
+      "epoch": 0.4824120603015075,
+      "grad_norm": 0.5218870043754578,
+      "learning_rate": 0.00030896622729689266,
+      "loss": 6.2556,
+      "step": 1404
+    },
+    {
+      "epoch": 0.4827556586350556,
+      "grad_norm": 0.663040816783905,
+      "learning_rate": 0.00030867476118488654,
+      "loss": 6.1849,
+      "step": 1405
+    },
+    {
+      "epoch": 0.48309925696860373,
+      "grad_norm": 0.7346043586730957,
+      "learning_rate": 0.0003083832106461391,
+      "loss": 6.3262,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4834428553021518,
+      "grad_norm": 0.6023099422454834,
+      "learning_rate": 0.0003080915761001605,
+      "loss": 6.2441,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4837864536356999,
+      "grad_norm": 0.6472316980361938,
+      "learning_rate": 0.00030779985796658177,
+      "loss": 6.3055,
+      "step": 1408
+    },
+    {
+      "epoch": 0.48413005196924797,
+      "grad_norm": 0.6640135049819946,
+      "learning_rate": 0.0003075080566651544,
+      "loss": 6.245,
+      "step": 1409
+    },
+    {
+      "epoch": 0.48447365030279604,
+      "grad_norm": 0.6156014204025269,
+      "learning_rate": 0.0003072161726157494,
+      "loss": 6.1676,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4848172486363441,
+      "grad_norm": 0.798943042755127,
+      "learning_rate": 0.0003069242062383569,
+      "loss": 6.3021,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4851608469698922,
+      "grad_norm": 0.5238396525382996,
+      "learning_rate": 0.00030663215795308533,
+      "loss": 6.2686,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4855044453034403,
+      "grad_norm": 0.5713295340538025,
+      "learning_rate": 0.0003063400281801613,
+      "loss": 6.0774,
+      "step": 1413
+    },
+    {
+      "epoch": 0.48584804363698836,
+      "grad_norm": 0.6762855052947998,
+      "learning_rate": 0.0003060478173399283,
+      "loss": 6.2609,
+      "step": 1414
+    },
+    {
+      "epoch": 0.48619164197053644,
+      "grad_norm": 0.589787483215332,
+      "learning_rate": 0.00030575552585284684,
+      "loss": 6.2896,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4865352403040845,
+      "grad_norm": 0.6290336847305298,
+      "learning_rate": 0.0003054631541394932,
+      "loss": 6.1697,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4868788386376326,
+      "grad_norm": 0.5550079345703125,
+      "learning_rate": 0.00030517070262055907,
+      "loss": 6.1668,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4872224369711807,
+      "grad_norm": 0.5955278873443604,
+      "learning_rate": 0.00030487817171685126,
+      "loss": 6.2552,
+      "step": 1418
+    },
+    {
+      "epoch": 0.48756603530472875,
+      "grad_norm": 0.5142794251441956,
+      "learning_rate": 0.0003045855618492905,
+      "loss": 6.3203,
+      "step": 1419
+    },
+    {
+      "epoch": 0.48790963363827683,
+      "grad_norm": 0.6616760492324829,
+      "learning_rate": 0.0003042928734389114,
+      "loss": 6.3663,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48825323197182496,
+      "grad_norm": 0.5484029650688171,
+      "learning_rate": 0.0003040001069068613,
+      "loss": 6.1665,
+      "step": 1421
+    },
+    {
+      "epoch": 0.48859683030537304,
+      "grad_norm": 0.5682584047317505,
+      "learning_rate": 0.0003037072626744003,
+      "loss": 6.3342,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4889404286389211,
+      "grad_norm": 0.5823968648910522,
+      "learning_rate": 0.00030341434116289997,
+      "loss": 6.1182,
+      "step": 1423
+    },
+    {
+      "epoch": 0.4892840269724692,
+      "grad_norm": 0.5125210285186768,
+      "learning_rate": 0.00030312134279384317,
+      "loss": 6.3474,
+      "step": 1424
+    },
+    {
+      "epoch": 0.4896276253060173,
+      "grad_norm": 0.5110132098197937,
+      "learning_rate": 0.00030282826798882356,
+      "loss": 6.2739,
+      "step": 1425
+    },
+    {
+      "epoch": 0.48997122363956536,
+      "grad_norm": 0.56996750831604,
+      "learning_rate": 0.0003025351171695444,
+      "loss": 6.1881,
+      "step": 1426
+    },
+    {
+      "epoch": 0.49031482197311343,
+      "grad_norm": 0.5609503984451294,
+      "learning_rate": 0.0003022418907578188,
+      "loss": 6.3131,
+      "step": 1427
+    },
+    {
+      "epoch": 0.4906584203066615,
+      "grad_norm": 0.5459311604499817,
+      "learning_rate": 0.00030194858917556816,
+      "loss": 6.1978,
+      "step": 1428
+    },
+    {
+      "epoch": 0.4910020186402096,
+      "grad_norm": 0.5984681248664856,
+      "learning_rate": 0.0003016552128448224,
+      "loss": 6.321,
+      "step": 1429
+    },
+    {
+      "epoch": 0.49134561697375767,
+      "grad_norm": 0.5124654769897461,
+      "learning_rate": 0.00030136176218771875,
+      "loss": 6.2282,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49168921530730575,
+      "grad_norm": 0.5804280638694763,
+      "learning_rate": 0.00030106823762650163,
+      "loss": 6.2156,
+      "step": 1431
+    },
+    {
+      "epoch": 0.49203281364085383,
+      "grad_norm": 0.7151570916175842,
+      "learning_rate": 0.0003007746395835215,
+      "loss": 6.2783,
+      "step": 1432
+    },
+    {
+      "epoch": 0.4923764119744019,
+      "grad_norm": 0.6895557641983032,
+      "learning_rate": 0.00030048096848123493,
+      "loss": 6.1935,
+      "step": 1433
+    },
+    {
+      "epoch": 0.49272001030795,
+      "grad_norm": 0.7361437082290649,
+      "learning_rate": 0.0003001872247422032,
+      "loss": 6.2602,
+      "step": 1434
+    },
+    {
+      "epoch": 0.49306360864149806,
+      "grad_norm": 0.7202491760253906,
+      "learning_rate": 0.00029989340878909244,
+      "loss": 6.2125,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4934072069750462,
+      "grad_norm": 0.6291714906692505,
+      "learning_rate": 0.00029959952104467247,
+      "loss": 6.2358,
+      "step": 1436
+    },
+    {
+      "epoch": 0.4937508053085943,
+      "grad_norm": 0.8956025242805481,
+      "learning_rate": 0.0002993055619318166,
+      "loss": 6.2331,
+      "step": 1437
+    },
+    {
+      "epoch": 0.49409440364214235,
+      "grad_norm": 0.7390561103820801,
+      "learning_rate": 0.0002990115318735007,
+      "loss": 6.3176,
+      "step": 1438
+    },
+    {
+      "epoch": 0.49443800197569043,
+      "grad_norm": 0.6197091937065125,
+      "learning_rate": 0.00029871743129280273,
+      "loss": 6.1561,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4947816003092385,
+      "grad_norm": 0.7448058128356934,
+      "learning_rate": 0.00029842326061290205,
+      "loss": 6.344,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4951251986427866,
+      "grad_norm": 0.7358593940734863,
+      "learning_rate": 0.0002981290202570792,
+      "loss": 6.2683,
+      "step": 1441
+    },
+    {
+      "epoch": 0.49546879697633467,
+      "grad_norm": 0.8008904457092285,
+      "learning_rate": 0.0002978347106487146,
+      "loss": 6.2716,
+      "step": 1442
+    },
+    {
+      "epoch": 0.49581239530988275,
+      "grad_norm": 0.7615576386451721,
+      "learning_rate": 0.00029754033221128864,
+      "loss": 6.3032,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4961559936434308,
+      "grad_norm": 0.7408410906791687,
+      "learning_rate": 0.0002972458853683803,
+      "loss": 6.352,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4964995919769789,
+      "grad_norm": 0.6674975752830505,
+      "learning_rate": 0.0002969513705436676,
+      "loss": 6.3652,
+      "step": 1445
+    },
+    {
+      "epoch": 0.496843190310527,
+      "grad_norm": 0.9053188562393188,
+      "learning_rate": 0.0002966567881609258,
+      "loss": 6.1545,
+      "step": 1446
+    },
+    {
+      "epoch": 0.49718678864407506,
+      "grad_norm": 0.7032760381698608,
+      "learning_rate": 0.0002963621386440277,
+      "loss": 6.3121,
+      "step": 1447
+    },
+    {
+      "epoch": 0.49753038697762314,
+      "grad_norm": 0.9544557332992554,
+      "learning_rate": 0.0002960674224169427,
+      "loss": 6.2407,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4978739853111712,
+      "grad_norm": 0.8960500955581665,
+      "learning_rate": 0.00029577263990373593,
+      "loss": 6.38,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4982175836447193,
+      "grad_norm": 1.0645229816436768,
+      "learning_rate": 0.00029547779152856827,
+      "loss": 6.2797,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49856118197826743,
+      "grad_norm": 1.2595845460891724,
+      "learning_rate": 0.0002951828777156951,
+      "loss": 6.2422,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4989047803118155,
+      "grad_norm": 1.0342768430709839,
+      "learning_rate": 0.000294887898889466,
+      "loss": 6.2663,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4992483786453636,
+      "grad_norm": 0.9279031753540039,
+      "learning_rate": 0.0002945928554743241,
+      "loss": 6.2004,
+      "step": 1453
+    },
+    {
+      "epoch": 0.49959197697891167,
+      "grad_norm": 0.8587821125984192,
+      "learning_rate": 0.0002942977478948057,
+      "loss": 6.1789,
+      "step": 1454
+    },
+    {
+      "epoch": 0.49993557531245975,
+      "grad_norm": 0.880979597568512,
+      "learning_rate": 0.00029400257657553896,
+      "loss": 6.1842,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5002791736460078,
+      "grad_norm": 0.9389069676399231,
+      "learning_rate": 0.0002937073419412442,
+      "loss": 6.18,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5006227719795558,
+      "grad_norm": 0.5919857025146484,
+      "learning_rate": 0.00029341204441673266,
+      "loss": 6.1496,
+      "step": 1457
+    },
+    {
+      "epoch": 0.500966370313104,
+      "grad_norm": 0.8019484281539917,
+      "learning_rate": 0.0002931166844269059,
+      "loss": 6.0585,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5013099686466521,
+      "grad_norm": 0.6207021474838257,
+      "learning_rate": 0.0002928212623967556,
+      "loss": 6.1253,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5016535669802001,
+      "grad_norm": 0.7272038459777832,
+      "learning_rate": 0.0002925257787513628,
+      "loss": 6.1487,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5019971653137483,
+      "grad_norm": 0.7108120918273926,
+      "learning_rate": 0.00029223023391589695,
+      "loss": 6.1344,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5023407636472963,
+      "grad_norm": 0.51841801404953,
+      "learning_rate": 0.0002919346283156155,
+      "loss": 6.3044,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5026843619808444,
+      "grad_norm": 0.7992899417877197,
+      "learning_rate": 0.0002916389623758636,
+      "loss": 6.1096,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5030279603143925,
+      "grad_norm": 0.8045459985733032,
+      "learning_rate": 0.0002913432365220732,
+      "loss": 6.1531,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5033715586479406,
+      "grad_norm": 0.4957735240459442,
+      "learning_rate": 0.0002910474511797621,
+      "loss": 6.2309,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5037151569814886,
+      "grad_norm": 0.5887153148651123,
+      "learning_rate": 0.00029075160677453416,
+      "loss": 6.2778,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5040587553150367,
+      "grad_norm": 0.7119141221046448,
+      "learning_rate": 0.00029045570373207794,
+      "loss": 6.1357,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5044023536485848,
+      "grad_norm": 0.5873328447341919,
+      "learning_rate": 0.0002901597424781664,
+      "loss": 6.1697,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5047459519821329,
+      "grad_norm": 0.5278820395469666,
+      "learning_rate": 0.00028986372343865643,
+      "loss": 6.1599,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5050895503156809,
+      "grad_norm": 0.5881029963493347,
+      "learning_rate": 0.00028956764703948787,
+      "loss": 6.2839,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5054331486492291,
+      "grad_norm": 0.6216078996658325,
+      "learning_rate": 0.0002892715137066831,
+      "loss": 6.1502,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5057767469827771,
+      "grad_norm": 0.548704981803894,
+      "learning_rate": 0.00028897532386634663,
+      "loss": 6.2749,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5061203453163252,
+      "grad_norm": 0.6772187352180481,
+      "learning_rate": 0.00028867907794466403,
+      "loss": 6.1506,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5064639436498733,
+      "grad_norm": 0.6250154972076416,
+      "learning_rate": 0.00028838277636790183,
+      "loss": 6.3769,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5068075419834214,
+      "grad_norm": 0.6144253015518188,
+      "learning_rate": 0.0002880864195624063,
+      "loss": 6.2031,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5071511403169695,
+      "grad_norm": 0.5683775544166565,
+      "learning_rate": 0.0002877900079546035,
+      "loss": 6.2813,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5074947386505175,
+      "grad_norm": 0.6378828883171082,
+      "learning_rate": 0.0002874935419709982,
+      "loss": 6.2685,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5078383369840657,
+      "grad_norm": 0.5964968800544739,
+      "learning_rate": 0.0002871970220381733,
+      "loss": 6.2304,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5081819353176137,
+      "grad_norm": 0.6776195764541626,
+      "learning_rate": 0.0002869004485827896,
+      "loss": 6.3116,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5085255336511618,
+      "grad_norm": 0.6209731101989746,
+      "learning_rate": 0.0002866038220315847,
+      "loss": 6.215,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5088691319847098,
+      "grad_norm": 0.6367368698120117,
+      "learning_rate": 0.0002863071428113726,
+      "loss": 6.2344,
+      "step": 1481
+    },
+    {
+      "epoch": 0.509212730318258,
+      "grad_norm": 0.6730502247810364,
+      "learning_rate": 0.0002860104113490432,
+      "loss": 6.2209,
+      "step": 1482
+    },
+    {
+      "epoch": 0.509556328651806,
+      "grad_norm": 0.529813289642334,
+      "learning_rate": 0.0002857136280715616,
+      "loss": 6.1866,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5098999269853541,
+      "grad_norm": 0.7928996682167053,
+      "learning_rate": 0.0002854167934059672,
+      "loss": 6.2018,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5102435253189022,
+      "grad_norm": 0.5787724852561951,
+      "learning_rate": 0.0002851199077793736,
+      "loss": 6.276,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5105871236524503,
+      "grad_norm": 0.6694987416267395,
+      "learning_rate": 0.0002848229716189678,
+      "loss": 6.4032,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5109307219859983,
+      "grad_norm": 0.604768693447113,
+      "learning_rate": 0.0002845259853520091,
+      "loss": 6.2687,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5112743203195464,
+      "grad_norm": 0.6459552049636841,
+      "learning_rate": 0.00028422894940582927,
+      "loss": 6.2607,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5116179186530946,
+      "grad_norm": 0.992656409740448,
+      "learning_rate": 0.00028393186420783145,
+      "loss": 6.3237,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5119615169866426,
+      "grad_norm": 0.823799192905426,
+      "learning_rate": 0.0002836347301854897,
+      "loss": 6.092,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5123051153201907,
+      "grad_norm": 0.8477516174316406,
+      "learning_rate": 0.0002833375477663481,
+      "loss": 6.223,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5126487136537388,
+      "grad_norm": 0.7959129214286804,
+      "learning_rate": 0.00028304031737802076,
+      "loss": 6.2709,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5129923119872869,
+      "grad_norm": 0.8560094237327576,
+      "learning_rate": 0.00028274303944819044,
+      "loss": 6.3429,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5133359103208349,
+      "grad_norm": 0.8726269602775574,
+      "learning_rate": 0.0002824457144046086,
+      "loss": 6.244,
+      "step": 1494
+    },
+    {
+      "epoch": 0.513679508654383,
+      "grad_norm": 0.8484057188034058,
+      "learning_rate": 0.0002821483426750942,
+      "loss": 6.2978,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5140231069879311,
+      "grad_norm": 0.7470120787620544,
+      "learning_rate": 0.00028185092468753373,
+      "loss": 6.3663,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5143667053214792,
+      "grad_norm": 0.9047187566757202,
+      "learning_rate": 0.0002815534608698798,
+      "loss": 6.5125,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5147103036550272,
+      "grad_norm": 0.9526438117027283,
+      "learning_rate": 0.00028125595165015137,
+      "loss": 6.2855,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5150539019885754,
+      "grad_norm": 1.011681318283081,
+      "learning_rate": 0.0002809583974564326,
+      "loss": 6.3073,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5153975003221234,
+      "grad_norm": 0.8152887225151062,
+      "learning_rate": 0.0002806607987168722,
+      "loss": 6.4397,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5157410986556715,
+      "grad_norm": 1.0889192819595337,
+      "learning_rate": 0.0002803631558596832,
+      "loss": 6.2249,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5160846969892195,
+      "grad_norm": 1.135572075843811,
+      "learning_rate": 0.000280065469313142,
+      "loss": 6.0517,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5164282953227677,
+      "grad_norm": 0.6644673347473145,
+      "learning_rate": 0.0002797677395055879,
+      "loss": 6.1306,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5167718936563158,
+      "grad_norm": 0.7205954194068909,
+      "learning_rate": 0.0002794699668654223,
+      "loss": 6.1675,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5171154919898638,
+      "grad_norm": 0.6675506830215454,
+      "learning_rate": 0.00027917215182110853,
+      "loss": 6.1734,
+      "step": 1505
+    },
+    {
+      "epoch": 0.517459090323412,
+      "grad_norm": 0.7752920389175415,
+      "learning_rate": 0.00027887429480117075,
+      "loss": 6.2684,
+      "step": 1506
+    },
+    {
+      "epoch": 0.51780268865696,
+      "grad_norm": 0.7607707381248474,
+      "learning_rate": 0.00027857639623419346,
+      "loss": 6.1148,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5181462869905081,
+      "grad_norm": 0.49518853425979614,
+      "learning_rate": 0.0002782784565488211,
+      "loss": 6.1403,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5184898853240562,
+      "grad_norm": 0.7124084830284119,
+      "learning_rate": 0.0002779804761737571,
+      "loss": 6.1645,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5188334836576043,
+      "grad_norm": 0.7854694724082947,
+      "learning_rate": 0.00027768245553776356,
+      "loss": 6.2317,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5191770819911523,
+      "grad_norm": 0.6847150325775146,
+      "learning_rate": 0.00027738439506966046,
+      "loss": 6.005,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5195206803247004,
+      "grad_norm": 0.6421884894371033,
+      "learning_rate": 0.00027708629519832516,
+      "loss": 6.1717,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5198642786582485,
+      "grad_norm": 0.7636914253234863,
+      "learning_rate": 0.0002767881563526917,
+      "loss": 6.2715,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5202078769917966,
+      "grad_norm": 0.5291933417320251,
+      "learning_rate": 0.00027648997896175003,
+      "loss": 6.3303,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5205514753253446,
+      "grad_norm": 0.5756345987319946,
+      "learning_rate": 0.00027619176345454585,
+      "loss": 6.1258,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5208950736588928,
+      "grad_norm": 0.7027370929718018,
+      "learning_rate": 0.0002758935102601796,
+      "loss": 6.214,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5212386719924408,
+      "grad_norm": 0.7267154455184937,
+      "learning_rate": 0.00027559521980780564,
+      "loss": 6.2464,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5215822703259889,
+      "grad_norm": 0.5030806064605713,
+      "learning_rate": 0.0002752968925266325,
+      "loss": 6.2359,
+      "step": 1518
+    },
+    {
+      "epoch": 0.521925868659537,
+      "grad_norm": 0.6831111311912537,
+      "learning_rate": 0.0002749985288459213,
+      "loss": 6.0984,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5222694669930851,
+      "grad_norm": 0.6685227751731873,
+      "learning_rate": 0.00027470012919498567,
+      "loss": 6.0867,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5226130653266332,
+      "grad_norm": 0.5604387521743774,
+      "learning_rate": 0.00027440169400319087,
+      "loss": 6.3091,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5229566636601812,
+      "grad_norm": 0.6780193448066711,
+      "learning_rate": 0.00027410322369995357,
+      "loss": 6.2407,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5233002619937294,
+      "grad_norm": 0.7392048835754395,
+      "learning_rate": 0.0002738047187147406,
+      "loss": 6.2447,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5236438603272774,
+      "grad_norm": 0.6794151067733765,
+      "learning_rate": 0.00027350617947706913,
+      "loss": 6.3431,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5239874586608255,
+      "grad_norm": 0.6256364583969116,
+      "learning_rate": 0.0002732076064165052,
+      "loss": 6.2444,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5243310569943735,
+      "grad_norm": 0.6079226136207581,
+      "learning_rate": 0.0002729089999626637,
+      "loss": 6.2164,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5246746553279217,
+      "grad_norm": 0.6453949809074402,
+      "learning_rate": 0.0002726103605452075,
+      "loss": 6.1986,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5250182536614697,
+      "grad_norm": 0.7797389626502991,
+      "learning_rate": 0.0002723116885938472,
+      "loss": 6.1708,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5253618519950178,
+      "grad_norm": 0.5642460584640503,
+      "learning_rate": 0.00027201298453833977,
+      "loss": 6.0971,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5257054503285659,
+      "grad_norm": 0.5858938694000244,
+      "learning_rate": 0.00027171424880848867,
+      "loss": 6.3316,
+      "step": 1530
+    },
+    {
+      "epoch": 0.526049048662114,
+      "grad_norm": 0.8247479796409607,
+      "learning_rate": 0.00027141548183414274,
+      "loss": 6.1299,
+      "step": 1531
+    },
+    {
+      "epoch": 0.526392646995662,
+      "grad_norm": 0.6232000589370728,
+      "learning_rate": 0.00027111668404519604,
+      "loss": 6.2575,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5267362453292102,
+      "grad_norm": 0.5557352900505066,
+      "learning_rate": 0.0002708178558715866,
+      "loss": 6.2561,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5270798436627583,
+      "grad_norm": 0.7393467426300049,
+      "learning_rate": 0.00027051899774329665,
+      "loss": 6.1768,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5274234419963063,
+      "grad_norm": 0.6794891357421875,
+      "learning_rate": 0.00027022011009035106,
+      "loss": 6.2293,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5277670403298544,
+      "grad_norm": 0.7217805981636047,
+      "learning_rate": 0.0002699211933428174,
+      "loss": 6.1596,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5281106386634025,
+      "grad_norm": 0.7262641787528992,
+      "learning_rate": 0.00026962224793080513,
+      "loss": 6.1514,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5284542369969506,
+      "grad_norm": 0.7474170327186584,
+      "learning_rate": 0.0002693232742844649,
+      "loss": 6.2244,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5287978353304986,
+      "grad_norm": 0.7755861878395081,
+      "learning_rate": 0.00026902427283398796,
+      "loss": 6.1124,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5291414336640468,
+      "grad_norm": 0.7614656090736389,
+      "learning_rate": 0.00026872524400960564,
+      "loss": 6.3099,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5294850319975948,
+      "grad_norm": 0.6895480751991272,
+      "learning_rate": 0.0002684261882415886,
+      "loss": 6.2277,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5298286303311429,
+      "grad_norm": 0.7481386065483093,
+      "learning_rate": 0.0002681271059602462,
+      "loss": 6.4332,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5301722286646909,
+      "grad_norm": 0.6497515439987183,
+      "learning_rate": 0.0002678279975959261,
+      "loss": 6.2423,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5305158269982391,
+      "grad_norm": 0.7895376086235046,
+      "learning_rate": 0.00026752886357901353,
+      "loss": 6.4106,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5308594253317871,
+      "grad_norm": 0.7385738492012024,
+      "learning_rate": 0.0002672297043399304,
+      "loss": 6.33,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5312030236653352,
+      "grad_norm": 0.6905698776245117,
+      "learning_rate": 0.0002669305203091351,
+      "loss": 6.1804,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5315466219988833,
+      "grad_norm": 0.9432231783866882,
+      "learning_rate": 0.0002666313119171216,
+      "loss": 6.2444,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5318902203324314,
+      "grad_norm": 1.0359256267547607,
+      "learning_rate": 0.000266332079594419,
+      "loss": 6.2369,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5322338186659795,
+      "grad_norm": 0.7891038656234741,
+      "learning_rate": 0.0002660328237715907,
+      "loss": 6.3167,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5325774169995275,
+      "grad_norm": 1.0773836374282837,
+      "learning_rate": 0.000265733544879234,
+      "loss": 6.3635,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5329210153330757,
+      "grad_norm": 1.258744478225708,
+      "learning_rate": 0.00026543424334797956,
+      "loss": 6.1315,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5332646136666237,
+      "grad_norm": 1.075838565826416,
+      "learning_rate": 0.0002651349196084903,
+      "loss": 6.1109,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5336082120001718,
+      "grad_norm": 0.7799094915390015,
+      "learning_rate": 0.0002648355740914613,
+      "loss": 6.1875,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5339518103337199,
+      "grad_norm": 0.872112512588501,
+      "learning_rate": 0.00026453620722761897,
+      "loss": 6.1078,
+      "step": 1554
+    },
+    {
+      "epoch": 0.534295408667268,
+      "grad_norm": 0.8411591649055481,
+      "learning_rate": 0.00026423681944772034,
+      "loss": 6.0248,
+      "step": 1555
+    },
+    {
+      "epoch": 0.534639007000816,
+      "grad_norm": 0.8713768720626831,
+      "learning_rate": 0.00026393741118255253,
+      "loss": 6.1603,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5349826053343641,
+      "grad_norm": 1.0349767208099365,
+      "learning_rate": 0.00026363798286293226,
+      "loss": 6.1652,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5353262036679122,
+      "grad_norm": 0.7198126912117004,
+      "learning_rate": 0.0002633385349197051,
+      "loss": 6.125,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5356698020014603,
+      "grad_norm": 0.8494895696640015,
+      "learning_rate": 0.0002630390677837447,
+      "loss": 6.2175,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5360134003350083,
+      "grad_norm": 0.8133593201637268,
+      "learning_rate": 0.00026273958188595235,
+      "loss": 6.3068,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5363569986685565,
+      "grad_norm": 0.7584961652755737,
+      "learning_rate": 0.0002624400776572566,
+      "loss": 6.1282,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5367005970021045,
+      "grad_norm": 0.7136414051055908,
+      "learning_rate": 0.0002621405555286121,
+      "loss": 6.22,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5370441953356526,
+      "grad_norm": 0.8710498809814453,
+      "learning_rate": 0.0002618410159309992,
+      "loss": 6.16,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5373877936692008,
+      "grad_norm": 0.8503127098083496,
+      "learning_rate": 0.00026154145929542386,
+      "loss": 6.2239,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5377313920027488,
+      "grad_norm": 0.8023079037666321,
+      "learning_rate": 0.0002612418860529158,
+      "loss": 6.1692,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5380749903362969,
+      "grad_norm": 0.9042754769325256,
+      "learning_rate": 0.00026094229663452934,
+      "loss": 6.1975,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5384185886698449,
+      "grad_norm": 0.7970142960548401,
+      "learning_rate": 0.0002606426914713418,
+      "loss": 6.2393,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5387621870033931,
+      "grad_norm": 0.607698917388916,
+      "learning_rate": 0.00026034307099445295,
+      "loss": 6.19,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5391057853369411,
+      "grad_norm": 0.8219443559646606,
+      "learning_rate": 0.0002600434356349849,
+      "loss": 6.3123,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5394493836704892,
+      "grad_norm": 0.6942794919013977,
+      "learning_rate": 0.0002597437858240812,
+      "loss": 6.1638,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5397929820040372,
+      "grad_norm": 0.7742037773132324,
+      "learning_rate": 0.00025944412199290585,
+      "loss": 6.1596,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5401365803375854,
+      "grad_norm": 0.8137544989585876,
+      "learning_rate": 0.00025914444457264334,
+      "loss": 6.2533,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5404801786711334,
+      "grad_norm": 0.6724058985710144,
+      "learning_rate": 0.0002588447539944976,
+      "loss": 6.2323,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5408237770046815,
+      "grad_norm": 0.7243574857711792,
+      "learning_rate": 0.0002585450506896915,
+      "loss": 6.2136,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5411673753382296,
+      "grad_norm": 0.734247624874115,
+      "learning_rate": 0.00025824533508946615,
+      "loss": 6.1987,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5415109736717777,
+      "grad_norm": 0.6768348813056946,
+      "learning_rate": 0.00025794560762508044,
+      "loss": 6.2893,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5418545720053258,
+      "grad_norm": 0.5567851662635803,
+      "learning_rate": 0.00025764586872781053,
+      "loss": 6.2841,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5421981703388739,
+      "grad_norm": 0.6665927767753601,
+      "learning_rate": 0.00025734611882894857,
+      "loss": 6.3247,
+      "step": 1578
+    },
+    {
+      "epoch": 0.542541768672422,
+      "grad_norm": 0.6924320459365845,
+      "learning_rate": 0.0002570463583598028,
+      "loss": 6.2716,
+      "step": 1579
+    },
+    {
+      "epoch": 0.54288536700597,
+      "grad_norm": 0.6488571166992188,
+      "learning_rate": 0.0002567465877516968,
+      "loss": 6.191,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5432289653395181,
+      "grad_norm": 0.5639288425445557,
+      "learning_rate": 0.0002564468074359684,
+      "loss": 6.1643,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5435725636730662,
+      "grad_norm": 0.6971966624259949,
+      "learning_rate": 0.0002561470178439698,
+      "loss": 6.1649,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5439161620066143,
+      "grad_norm": 0.7037007808685303,
+      "learning_rate": 0.0002558472194070662,
+      "loss": 6.1952,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5442597603401623,
+      "grad_norm": 0.5069451332092285,
+      "learning_rate": 0.00025554741255663584,
+      "loss": 6.0759,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5446033586737105,
+      "grad_norm": 0.6529680490493774,
+      "learning_rate": 0.00025524759772406865,
+      "loss": 6.2216,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5449469570072585,
+      "grad_norm": 0.5853424072265625,
+      "learning_rate": 0.00025494777534076647,
+      "loss": 6.1679,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5452905553408066,
+      "grad_norm": 0.4783673882484436,
+      "learning_rate": 0.00025464794583814174,
+      "loss": 6.2046,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5456341536743546,
+      "grad_norm": 0.5306436419487,
+      "learning_rate": 0.00025434810964761726,
+      "loss": 6.1886,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5459777520079028,
+      "grad_norm": 0.6045244932174683,
+      "learning_rate": 0.0002540482672006254,
+      "loss": 6.2545,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5463213503414508,
+      "grad_norm": 0.6048717498779297,
+      "learning_rate": 0.0002537484189286076,
+      "loss": 6.3085,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5466649486749989,
+      "grad_norm": 0.5896782279014587,
+      "learning_rate": 0.0002534485652630135,
+      "loss": 6.2663,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5470085470085471,
+      "grad_norm": 0.5969629883766174,
+      "learning_rate": 0.0002531487066353008,
+      "loss": 6.1941,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5473521453420951,
+      "grad_norm": 0.6015769243240356,
+      "learning_rate": 0.00025284884347693415,
+      "loss": 6.0663,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5476957436756432,
+      "grad_norm": 0.6984557509422302,
+      "learning_rate": 0.0002525489762193847,
+      "loss": 6.2395,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5480393420091912,
+      "grad_norm": 0.7120092511177063,
+      "learning_rate": 0.0002522491052941295,
+      "loss": 6.2486,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5483829403427394,
+      "grad_norm": 0.6783263683319092,
+      "learning_rate": 0.00025194923113265095,
+      "loss": 6.2436,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5487265386762874,
+      "grad_norm": 0.7355448007583618,
+      "learning_rate": 0.0002516493541664362,
+      "loss": 6.1035,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5490701370098355,
+      "grad_norm": 0.6941987872123718,
+      "learning_rate": 0.00025134947482697613,
+      "loss": 6.3763,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5494137353433836,
+      "grad_norm": 0.873782753944397,
+      "learning_rate": 0.0002510495935457653,
+      "loss": 6.3246,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5497573336769317,
+      "grad_norm": 1.0551037788391113,
+      "learning_rate": 0.00025074971075430104,
+      "loss": 6.2419,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5501009320104797,
+      "grad_norm": 0.7868710160255432,
+      "learning_rate": 0.0002504498268840826,
+      "loss": 6.1377,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5504445303440278,
+      "grad_norm": 0.6369834542274475,
+      "learning_rate": 0.00025014994236661125,
+      "loss": 6.1627,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5507881286775759,
+      "grad_norm": 0.6547728180885315,
+      "learning_rate": 0.00024985005763338876,
+      "loss": 6.0642,
+      "step": 1603
+    },
+    {
+      "epoch": 0.551131727011124,
+      "grad_norm": 0.5977075099945068,
+      "learning_rate": 0.0002495501731159174,
+      "loss": 6.0824,
+      "step": 1604
+    },
+    {
+      "epoch": 0.551475325344672,
+      "grad_norm": 0.8198515772819519,
+      "learning_rate": 0.0002492502892456991,
+      "loss": 6.0344,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5518189236782202,
+      "grad_norm": 0.7595130801200867,
+      "learning_rate": 0.0002489504064542347,
+      "loss": 6.039,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5521625220117683,
+      "grad_norm": 0.5898275375366211,
+      "learning_rate": 0.00024865052517302394,
+      "loss": 6.1183,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5525061203453163,
+      "grad_norm": 0.6673107743263245,
+      "learning_rate": 0.0002483506458335639,
+      "loss": 6.0777,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5528497186788645,
+      "grad_norm": 0.8127065300941467,
+      "learning_rate": 0.00024805076886734906,
+      "loss": 6.0912,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5531933170124125,
+      "grad_norm": 0.6119105815887451,
+      "learning_rate": 0.00024775089470587057,
+      "loss": 6.2048,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5535369153459606,
+      "grad_norm": 0.5875973105430603,
+      "learning_rate": 0.00024745102378061543,
+      "loss": 6.0123,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5538805136795086,
+      "grad_norm": 0.837634265422821,
+      "learning_rate": 0.00024715115652306586,
+      "loss": 6.1173,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5542241120130568,
+      "grad_norm": 0.7039463520050049,
+      "learning_rate": 0.0002468512933646992,
+      "loss": 6.1897,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5545677103466048,
+      "grad_norm": 0.6563109159469604,
+      "learning_rate": 0.00024655143473698655,
+      "loss": 6.2105,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5549113086801529,
+      "grad_norm": 0.5650578737258911,
+      "learning_rate": 0.00024625158107139246,
+      "loss": 6.1727,
+      "step": 1615
+    },
+    {
+      "epoch": 0.555254907013701,
+      "grad_norm": 0.6338276863098145,
+      "learning_rate": 0.0002459517327993746,
+      "loss": 6.0853,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5555985053472491,
+      "grad_norm": 0.8130798935890198,
+      "learning_rate": 0.0002456518903523828,
+      "loss": 6.2192,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5559421036807971,
+      "grad_norm": 0.47761037945747375,
+      "learning_rate": 0.00024535205416185827,
+      "loss": 6.2542,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5562857020143452,
+      "grad_norm": 0.5506361126899719,
+      "learning_rate": 0.00024505222465923354,
+      "loss": 6.1728,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5566293003478933,
+      "grad_norm": 0.6886653304100037,
+      "learning_rate": 0.0002447524022759313,
+      "loss": 6.1607,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5569728986814414,
+      "grad_norm": 0.5948178768157959,
+      "learning_rate": 0.0002444525874433642,
+      "loss": 6.1875,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5573164970149895,
+      "grad_norm": 0.5316332578659058,
+      "learning_rate": 0.0002441527805929338,
+      "loss": 6.1377,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5576600953485376,
+      "grad_norm": 0.6047093868255615,
+      "learning_rate": 0.00024385298215603017,
+      "loss": 6.1663,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5580036936820857,
+      "grad_norm": 0.7805117964744568,
+      "learning_rate": 0.00024355319256403156,
+      "loss": 6.2092,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5583472920156337,
+      "grad_norm": 0.5755521655082703,
+      "learning_rate": 0.0002432534122483033,
+      "loss": 6.1914,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5586908903491818,
+      "grad_norm": 0.6552025675773621,
+      "learning_rate": 0.0002429536416401972,
+      "loss": 6.1984,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5590344886827299,
+      "grad_norm": 0.6080207824707031,
+      "learning_rate": 0.00024265388117105153,
+      "loss": 6.1819,
+      "step": 1627
+    },
+    {
+      "epoch": 0.559378087016278,
+      "grad_norm": 0.6865605711936951,
+      "learning_rate": 0.0002423541312721896,
+      "loss": 6.1663,
+      "step": 1628
+    },
+    {
+      "epoch": 0.559721685349826,
+      "grad_norm": 0.6106628179550171,
+      "learning_rate": 0.00024205439237491949,
+      "loss": 6.2782,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5600652836833742,
+      "grad_norm": 0.5570806264877319,
+      "learning_rate": 0.00024175466491053392,
+      "loss": 6.1906,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5604088820169222,
+      "grad_norm": 0.6362578272819519,
+      "learning_rate": 0.0002414549493103086,
+      "loss": 6.2011,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5607524803504703,
+      "grad_norm": 0.7689515352249146,
+      "learning_rate": 0.00024115524600550243,
+      "loss": 6.0611,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5610960786840183,
+      "grad_norm": 0.5351979732513428,
+      "learning_rate": 0.0002408555554273567,
+      "loss": 6.1061,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5614396770175665,
+      "grad_norm": 0.5828037261962891,
+      "learning_rate": 0.0002405558780070942,
+      "loss": 6.2728,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5617832753511145,
+      "grad_norm": 0.6618168354034424,
+      "learning_rate": 0.00024025621417591886,
+      "loss": 6.2246,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5621268736846626,
+      "grad_norm": 0.7272453904151917,
+      "learning_rate": 0.0002399565643650151,
+      "loss": 6.0693,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5624704720182108,
+      "grad_norm": 0.7130446434020996,
+      "learning_rate": 0.00023965692900554712,
+      "loss": 6.2137,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5628140703517588,
+      "grad_norm": 0.4798155128955841,
+      "learning_rate": 0.0002393573085286583,
+      "loss": 6.2844,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5631576686853069,
+      "grad_norm": 0.6884756088256836,
+      "learning_rate": 0.0002390577033654707,
+      "loss": 6.1711,
+      "step": 1639
+    },
+    {
+      "epoch": 0.563501267018855,
+      "grad_norm": 0.682988703250885,
+      "learning_rate": 0.0002387581139470843,
+      "loss": 6.2269,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5638448653524031,
+      "grad_norm": 0.7200033068656921,
+      "learning_rate": 0.00023845854070457623,
+      "loss": 6.2101,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5641884636859511,
+      "grad_norm": 0.6693586111068726,
+      "learning_rate": 0.0002381589840690008,
+      "loss": 6.2992,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5645320620194992,
+      "grad_norm": 0.6025054454803467,
+      "learning_rate": 0.000237859444471388,
+      "loss": 6.2408,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5648756603530473,
+      "grad_norm": 0.7449381351470947,
+      "learning_rate": 0.0002375599223427434,
+      "loss": 6.2862,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5652192586865954,
+      "grad_norm": 1.0429128408432007,
+      "learning_rate": 0.00023726041811404766,
+      "loss": 6.2646,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5655628570201434,
+      "grad_norm": 0.7181035876274109,
+      "learning_rate": 0.00023696093221625532,
+      "loss": 6.1514,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5659064553536916,
+      "grad_norm": 0.838463306427002,
+      "learning_rate": 0.0002366614650802949,
+      "loss": 6.4686,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5662500536872396,
+      "grad_norm": 1.0257136821746826,
+      "learning_rate": 0.00023636201713706772,
+      "loss": 6.3583,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5665936520207877,
+      "grad_norm": 0.8851714134216309,
+      "learning_rate": 0.00023606258881744745,
+      "loss": 6.3048,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5669372503543357,
+      "grad_norm": 1.000500202178955,
+      "learning_rate": 0.00023576318055227975,
+      "loss": 6.3964,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5672808486878839,
+      "grad_norm": 1.0033082962036133,
+      "learning_rate": 0.00023546379277238105,
+      "loss": 6.1371,
+      "step": 1651
+    },
+    {
+      "epoch": 0.567624447021432,
+      "grad_norm": 0.7857053875923157,
+      "learning_rate": 0.0002351644259085387,
+      "loss": 6.0755,
+      "step": 1652
+    },
+    {
+      "epoch": 0.56796804535498,
+      "grad_norm": 0.7347760200500488,
+      "learning_rate": 0.00023486508039150976,
+      "loss": 6.1028,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5683116436885282,
+      "grad_norm": 0.5923128128051758,
+      "learning_rate": 0.00023456575665202053,
+      "loss": 6.0312,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5686552420220762,
+      "grad_norm": 0.7084698677062988,
+      "learning_rate": 0.000234266455120766,
+      "loss": 6.0676,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5689988403556243,
+      "grad_norm": 0.7293703556060791,
+      "learning_rate": 0.0002339671762284094,
+      "loss": 6.1787,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5693424386891723,
+      "grad_norm": 0.604590654373169,
+      "learning_rate": 0.00023366792040558113,
+      "loss": 6.1567,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5696860370227205,
+      "grad_norm": 0.6981744766235352,
+      "learning_rate": 0.00023336868808287843,
+      "loss": 5.9918,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5700296353562685,
+      "grad_norm": 0.6478411555290222,
+      "learning_rate": 0.00023306947969086494,
+      "loss": 6.1123,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5703732336898166,
+      "grad_norm": 0.6853176951408386,
+      "learning_rate": 0.00023277029566006965,
+      "loss": 6.1282,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5707168320233647,
+      "grad_norm": 0.7653744220733643,
+      "learning_rate": 0.00023247113642098648,
+      "loss": 6.0782,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5710604303569128,
+      "grad_norm": 0.6941654086112976,
+      "learning_rate": 0.00023217200240407387,
+      "loss": 6.0358,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5714040286904608,
+      "grad_norm": 0.6170705556869507,
+      "learning_rate": 0.0002318728940397539,
+      "loss": 6.2664,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5717476270240089,
+      "grad_norm": 0.6748530864715576,
+      "learning_rate": 0.00023157381175841144,
+      "loss": 6.1793,
+      "step": 1664
+    },
+    {
+      "epoch": 0.572091225357557,
+      "grad_norm": 0.7513784170150757,
+      "learning_rate": 0.0002312747559903944,
+      "loss": 6.1495,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5724348236911051,
+      "grad_norm": 0.7271863222122192,
+      "learning_rate": 0.0002309757271660121,
+      "loss": 6.1181,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5727784220246532,
+      "grad_norm": 0.7138211131095886,
+      "learning_rate": 0.00023067672571553514,
+      "loss": 6.2254,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5731220203582013,
+      "grad_norm": 0.7385585904121399,
+      "learning_rate": 0.00023037775206919493,
+      "loss": 6.2181,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5734656186917494,
+      "grad_norm": 0.6244990229606628,
+      "learning_rate": 0.00023007880665718263,
+      "loss": 6.2382,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5738092170252974,
+      "grad_norm": 0.6192176938056946,
+      "learning_rate": 0.00022977988990964898,
+      "loss": 6.2845,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5741528153588455,
+      "grad_norm": 0.6722596287727356,
+      "learning_rate": 0.0002294810022567034,
+      "loss": 6.2133,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5744964136923936,
+      "grad_norm": 0.7086919546127319,
+      "learning_rate": 0.0002291821441284133,
+      "loss": 6.0343,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5748400120259417,
+      "grad_norm": 0.7796761989593506,
+      "learning_rate": 0.000228883315954804,
+      "loss": 6.1843,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5751836103594897,
+      "grad_norm": 0.4925467371940613,
+      "learning_rate": 0.0002285845181658573,
+      "loss": 6.2025,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5755272086930379,
+      "grad_norm": 0.6159957051277161,
+      "learning_rate": 0.00022828575119151134,
+      "loss": 6.1741,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5758708070265859,
+      "grad_norm": 0.5439791083335876,
+      "learning_rate": 0.00022798701546166024,
+      "loss": 6.2161,
+      "step": 1676
+    },
+    {
+      "epoch": 0.576214405360134,
+      "grad_norm": 0.6240994930267334,
+      "learning_rate": 0.00022768831140615285,
+      "loss": 6.2162,
+      "step": 1677
+    },
+    {
+      "epoch": 0.576558003693682,
+      "grad_norm": 0.5322961807250977,
+      "learning_rate": 0.0002273896394547924,
+      "loss": 6.1087,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5769016020272302,
+      "grad_norm": 0.6332693696022034,
+      "learning_rate": 0.00022709100003733636,
+      "loss": 6.1368,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5772452003607782,
+      "grad_norm": 0.6597732901573181,
+      "learning_rate": 0.0002267923935834949,
+      "loss": 6.1741,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5775887986943263,
+      "grad_norm": 0.6494370102882385,
+      "learning_rate": 0.0002264938205229309,
+      "loss": 6.1665,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5779323970278745,
+      "grad_norm": 0.565851628780365,
+      "learning_rate": 0.0002261952812852594,
+      "loss": 6.1571,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5782759953614225,
+      "grad_norm": 0.4857479929924011,
+      "learning_rate": 0.0002258967763000465,
+      "loss": 6.2028,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5786195936949706,
+      "grad_norm": 0.6660141944885254,
+      "learning_rate": 0.00022559830599680914,
+      "loss": 6.1034,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5789631920285186,
+      "grad_norm": 0.7688015103340149,
+      "learning_rate": 0.0002252998708050144,
+      "loss": 6.1557,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5793067903620668,
+      "grad_norm": 0.5798430442810059,
+      "learning_rate": 0.0002250014711540788,
+      "loss": 6.1569,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5796503886956148,
+      "grad_norm": 0.8817176222801208,
+      "learning_rate": 0.0002247031074733675,
+      "loss": 6.0853,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5799939870291629,
+      "grad_norm": 0.6133841872215271,
+      "learning_rate": 0.00022440478019219437,
+      "loss": 6.1425,
+      "step": 1688
+    },
+    {
+      "epoch": 0.580337585362711,
+      "grad_norm": 0.7280254364013672,
+      "learning_rate": 0.00022410648973982057,
+      "loss": 6.1649,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5806811836962591,
+      "grad_norm": 0.7156420946121216,
+      "learning_rate": 0.00022380823654545416,
+      "loss": 6.0785,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5810247820298071,
+      "grad_norm": 0.7227360010147095,
+      "learning_rate": 0.00022351002103825003,
+      "loss": 6.1744,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5813683803633553,
+      "grad_norm": 0.6410605311393738,
+      "learning_rate": 0.00022321184364730847,
+      "loss": 6.2956,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5817119786969033,
+      "grad_norm": 0.7240776419639587,
+      "learning_rate": 0.00022291370480167485,
+      "loss": 6.2184,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5820555770304514,
+      "grad_norm": 0.7092165946960449,
+      "learning_rate": 0.0002226156049303396,
+      "loss": 6.3391,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5823991753639994,
+      "grad_norm": 0.692663311958313,
+      "learning_rate": 0.00022231754446223656,
+      "loss": 6.1169,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5827427736975476,
+      "grad_norm": 0.6744980812072754,
+      "learning_rate": 0.00022201952382624294,
+      "loss": 6.2275,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5830863720310957,
+      "grad_norm": 0.6746431589126587,
+      "learning_rate": 0.00022172154345117894,
+      "loss": 6.2227,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5834299703646437,
+      "grad_norm": 0.9164922833442688,
+      "learning_rate": 0.0002214236037658065,
+      "loss": 6.2579,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5837735686981919,
+      "grad_norm": 0.8660106658935547,
+      "learning_rate": 0.00022112570519882923,
+      "loss": 6.2453,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5841171670317399,
+      "grad_norm": 1.112679123878479,
+      "learning_rate": 0.00022082784817889148,
+      "loss": 6.1501,
+      "step": 1700
+    },
+    {
+      "epoch": 0.584460765365288,
+      "grad_norm": 1.1230827569961548,
+      "learning_rate": 0.00022053003313457763,
+      "loss": 6.0107,
+      "step": 1701
+    },
+    {
+      "epoch": 0.584804363698836,
+      "grad_norm": 1.0337865352630615,
+      "learning_rate": 0.00022023226049441218,
+      "loss": 6.1094,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5851479620323842,
+      "grad_norm": 0.7783079743385315,
+      "learning_rate": 0.0002199345306868581,
+      "loss": 5.9931,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5854915603659322,
+      "grad_norm": 0.6345077157020569,
+      "learning_rate": 0.0002196368441403168,
+      "loss": 5.9898,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5858351586994803,
+      "grad_norm": 0.8554926514625549,
+      "learning_rate": 0.00021933920128312784,
+      "loss": 6.1712,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5861787570330284,
+      "grad_norm": 0.8038747906684875,
+      "learning_rate": 0.0002190416025435675,
+      "loss": 6.1559,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5865223553665765,
+      "grad_norm": 0.63972008228302,
+      "learning_rate": 0.0002187440483498486,
+      "loss": 6.1405,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5868659537001245,
+      "grad_norm": 0.7269260883331299,
+      "learning_rate": 0.00021844653913012026,
+      "loss": 5.9317,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5872095520336726,
+      "grad_norm": 0.6801608204841614,
+      "learning_rate": 0.00021814907531246642,
+      "loss": 6.08,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5875531503672207,
+      "grad_norm": 0.6652011275291443,
+      "learning_rate": 0.0002178516573249058,
+      "loss": 6.1866,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5878967487007688,
+      "grad_norm": 0.7540636658668518,
+      "learning_rate": 0.00021755428559539145,
+      "loss": 6.1272,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5882403470343169,
+      "grad_norm": 0.7460317611694336,
+      "learning_rate": 0.0002172569605518096,
+      "loss": 6.0443,
+      "step": 1712
+    },
+    {
+      "epoch": 0.588583945367865,
+      "grad_norm": 0.5961695909500122,
+      "learning_rate": 0.00021695968262197928,
+      "loss": 5.9874,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5889275437014131,
+      "grad_norm": 0.7832739353179932,
+      "learning_rate": 0.00021666245223365193,
+      "loss": 6.1076,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5892711420349611,
+      "grad_norm": 0.7737155556678772,
+      "learning_rate": 0.00021636526981451038,
+      "loss": 6.0668,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5896147403685092,
+      "grad_norm": 0.7129299640655518,
+      "learning_rate": 0.00021606813579216856,
+      "loss": 6.2231,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5899583387020573,
+      "grad_norm": 0.7148540019989014,
+      "learning_rate": 0.00021577105059417077,
+      "loss": 6.0411,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5903019370356054,
+      "grad_norm": 0.7563623785972595,
+      "learning_rate": 0.000215474014647991,
+      "loss": 6.2887,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5906455353691534,
+      "grad_norm": 0.569948136806488,
+      "learning_rate": 0.00021517702838103224,
+      "loss": 6.1062,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5909891337027016,
+      "grad_norm": 0.5353204011917114,
+      "learning_rate": 0.00021488009222062637,
+      "loss": 6.1717,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5913327320362496,
+      "grad_norm": 0.592048168182373,
+      "learning_rate": 0.00021458320659403289,
+      "loss": 6.1654,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5916763303697977,
+      "grad_norm": 0.526983380317688,
+      "learning_rate": 0.00021428637192843843,
+      "loss": 6.0983,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5920199287033457,
+      "grad_norm": 0.5798513889312744,
+      "learning_rate": 0.00021398958865095682,
+      "loss": 6.1792,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5923635270368939,
+      "grad_norm": 0.5458791255950928,
+      "learning_rate": 0.0002136928571886275,
+      "loss": 6.1895,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5927071253704419,
+      "grad_norm": 0.5700400471687317,
+      "learning_rate": 0.00021339617796841534,
+      "loss": 6.1609,
+      "step": 1725
+    },
+    {
+      "epoch": 0.59305072370399,
+      "grad_norm": 0.4680607318878174,
+      "learning_rate": 0.00021309955141721044,
+      "loss": 6.1472,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5933943220375382,
+      "grad_norm": 0.5086695551872253,
+      "learning_rate": 0.00021280297796182667,
+      "loss": 6.236,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5937379203710862,
+      "grad_norm": 0.4671761691570282,
+      "learning_rate": 0.00021250645802900183,
+      "loss": 6.1338,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5940815187046343,
+      "grad_norm": 0.5281257033348083,
+      "learning_rate": 0.0002122099920453965,
+      "loss": 6.1005,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5944251170381823,
+      "grad_norm": 0.5001750588417053,
+      "learning_rate": 0.00021191358043759368,
+      "loss": 6.162,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5947687153717305,
+      "grad_norm": 0.46611008048057556,
+      "learning_rate": 0.0002116172236320982,
+      "loss": 6.0662,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5951123137052785,
+      "grad_norm": 0.5261972546577454,
+      "learning_rate": 0.00021132092205533598,
+      "loss": 6.2476,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5954559120388266,
+      "grad_norm": 0.557052731513977,
+      "learning_rate": 0.00021102467613365336,
+      "loss": 6.0415,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5957995103723747,
+      "grad_norm": 0.6007571816444397,
+      "learning_rate": 0.00021072848629331693,
+      "loss": 6.1562,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5961431087059228,
+      "grad_norm": 0.5048114061355591,
+      "learning_rate": 0.00021043235296051225,
+      "loss": 6.1526,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5964867070394708,
+      "grad_norm": 0.5392985939979553,
+      "learning_rate": 0.0002101362765613436,
+      "loss": 6.1723,
+      "step": 1736
+    },
+    {
+      "epoch": 0.596830305373019,
+      "grad_norm": 0.6697172522544861,
+      "learning_rate": 0.00020984025752183365,
+      "loss": 6.1767,
+      "step": 1737
+    },
+    {
+      "epoch": 0.597173903706567,
+      "grad_norm": 0.5006392598152161,
+      "learning_rate": 0.00020954429626792215,
+      "loss": 6.1836,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5975175020401151,
+      "grad_norm": 0.665244996547699,
+      "learning_rate": 0.00020924839322546585,
+      "loss": 6.1117,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5978611003736631,
+      "grad_norm": 0.6159089207649231,
+      "learning_rate": 0.00020895254882023791,
+      "loss": 6.1428,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5982046987072113,
+      "grad_norm": 0.5167410969734192,
+      "learning_rate": 0.00020865676347792692,
+      "loss": 6.1017,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5985482970407594,
+      "grad_norm": 0.6677618622779846,
+      "learning_rate": 0.0002083610376241364,
+      "loss": 6.1412,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5988918953743074,
+      "grad_norm": 0.6453651785850525,
+      "learning_rate": 0.00020806537168438456,
+      "loss": 6.2316,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5992354937078556,
+      "grad_norm": 0.7127484679222107,
+      "learning_rate": 0.00020776976608410317,
+      "loss": 6.1927,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5995790920414036,
+      "grad_norm": 0.8041994571685791,
+      "learning_rate": 0.00020747422124863725,
+      "loss": 6.2754,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5999226903749517,
+      "grad_norm": 0.6847991347312927,
+      "learning_rate": 0.00020717873760324443,
+      "loss": 6.2191,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6002662887084997,
+      "grad_norm": 0.7265855073928833,
+      "learning_rate": 0.0002068833155730942,
+      "loss": 6.2995,
+      "step": 1747
+    },
+    {
+      "epoch": 0.6006098870420479,
+      "grad_norm": 0.7762460708618164,
+      "learning_rate": 0.00020658795558326743,
+      "loss": 6.2802,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6009534853755959,
+      "grad_norm": 0.972196102142334,
+      "learning_rate": 0.00020629265805875585,
+      "loss": 6.3219,
+      "step": 1749
+    },
+    {
+      "epoch": 0.601297083709144,
+      "grad_norm": 1.1169794797897339,
+      "learning_rate": 0.0002059974234244611,
+      "loss": 6.2547,
+      "step": 1750
+    },
+    {
+      "epoch": 0.601640682042692,
+      "grad_norm": 0.9174626469612122,
+      "learning_rate": 0.00020570225210519432,
+      "loss": 6.1758,
+      "step": 1751
+    },
+    {
+      "epoch": 0.6019842803762402,
+      "grad_norm": 0.9274932146072388,
+      "learning_rate": 0.00020540714452567589,
+      "loss": 6.0432,
+      "step": 1752
+    },
+    {
+      "epoch": 0.6023278787097882,
+      "grad_norm": 0.7927635908126831,
+      "learning_rate": 0.000205112101110534,
+      "loss": 6.0737,
+      "step": 1753
+    },
+    {
+      "epoch": 0.6026714770433363,
+      "grad_norm": 0.7221388220787048,
+      "learning_rate": 0.00020481712228430493,
+      "loss": 6.1436,
+      "step": 1754
+    },
+    {
+      "epoch": 0.6030150753768844,
+      "grad_norm": 0.8973913192749023,
+      "learning_rate": 0.00020452220847143177,
+      "loss": 6.0908,
+      "step": 1755
+    },
+    {
+      "epoch": 0.6033586737104325,
+      "grad_norm": 1.0738279819488525,
+      "learning_rate": 0.00020422736009626405,
+      "loss": 6.1207,
+      "step": 1756
+    },
+    {
+      "epoch": 0.6037022720439806,
+      "grad_norm": 0.872310221195221,
+      "learning_rate": 0.00020393257758305738,
+      "loss": 6.0342,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6040458703775287,
+      "grad_norm": 0.7840533256530762,
+      "learning_rate": 0.00020363786135597236,
+      "loss": 6.0474,
+      "step": 1758
+    },
+    {
+      "epoch": 0.6043894687110768,
+      "grad_norm": 0.6296777725219727,
+      "learning_rate": 0.00020334321183907428,
+      "loss": 6.1445,
+      "step": 1759
+    },
+    {
+      "epoch": 0.6047330670446248,
+      "grad_norm": 0.9561694860458374,
+      "learning_rate": 0.0002030486294563325,
+      "loss": 6.017,
+      "step": 1760
+    },
+    {
+      "epoch": 0.605076665378173,
+      "grad_norm": 0.6782475113868713,
+      "learning_rate": 0.0002027541146316197,
+      "loss": 6.2128,
+      "step": 1761
+    },
+    {
+      "epoch": 0.605420263711721,
+      "grad_norm": 0.8592099547386169,
+      "learning_rate": 0.00020245966778871145,
+      "loss": 5.9947,
+      "step": 1762
+    },
+    {
+      "epoch": 0.6057638620452691,
+      "grad_norm": 0.681970477104187,
+      "learning_rate": 0.00020216528935128542,
+      "loss": 5.98,
+      "step": 1763
+    },
+    {
+      "epoch": 0.6061074603788171,
+      "grad_norm": 0.7359707355499268,
+      "learning_rate": 0.00020187097974292087,
+      "loss": 6.0558,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6064510587123653,
+      "grad_norm": 0.6512293815612793,
+      "learning_rate": 0.00020157673938709793,
+      "loss": 6.0842,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6067946570459133,
+      "grad_norm": 0.7809913158416748,
+      "learning_rate": 0.00020128256870719736,
+      "loss": 6.0179,
+      "step": 1766
+    },
+    {
+      "epoch": 0.6071382553794614,
+      "grad_norm": 0.6424116492271423,
+      "learning_rate": 0.0002009884681264994,
+      "loss": 6.089,
+      "step": 1767
+    },
+    {
+      "epoch": 0.6074818537130094,
+      "grad_norm": 0.46616989374160767,
+      "learning_rate": 0.00020069443806818339,
+      "loss": 6.1271,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6078254520465576,
+      "grad_norm": 0.6025314331054688,
+      "learning_rate": 0.00020040047895532754,
+      "loss": 6.153,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6081690503801056,
+      "grad_norm": 0.7727947235107422,
+      "learning_rate": 0.00020010659121090765,
+      "loss": 6.0856,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6085126487136537,
+      "grad_norm": 0.6243574619293213,
+      "learning_rate": 0.00019981277525779682,
+      "loss": 6.1613,
+      "step": 1771
+    },
+    {
+      "epoch": 0.6088562470472019,
+      "grad_norm": 0.5759406685829163,
+      "learning_rate": 0.00019951903151876516,
+      "loss": 6.1344,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6091998453807499,
+      "grad_norm": 0.5405790209770203,
+      "learning_rate": 0.00019922536041647854,
+      "loss": 6.0394,
+      "step": 1773
+    },
+    {
+      "epoch": 0.609543443714298,
+      "grad_norm": 0.6683335304260254,
+      "learning_rate": 0.00019893176237349838,
+      "loss": 6.055,
+      "step": 1774
+    },
+    {
+      "epoch": 0.609887042047846,
+      "grad_norm": 0.6566500067710876,
+      "learning_rate": 0.00019863823781228127,
+      "loss": 5.9496,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6102306403813942,
+      "grad_norm": 0.4817597270011902,
+      "learning_rate": 0.00019834478715517767,
+      "loss": 6.0593,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6105742387149422,
+      "grad_norm": 0.5774630904197693,
+      "learning_rate": 0.00019805141082443188,
+      "loss": 6.1018,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6109178370484903,
+      "grad_norm": 0.6785969138145447,
+      "learning_rate": 0.00019775810924218125,
+      "loss": 6.199,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6112614353820384,
+      "grad_norm": 0.7792773842811584,
+      "learning_rate": 0.0001974648828304556,
+      "loss": 6.09,
+      "step": 1779
+    },
+    {
+      "epoch": 0.6116050337155865,
+      "grad_norm": 0.5020087361335754,
+      "learning_rate": 0.0001971717320111765,
+      "loss": 6.0291,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6119486320491345,
+      "grad_norm": 0.764399528503418,
+      "learning_rate": 0.0001968786572061569,
+      "loss": 6.0347,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6122922303826827,
+      "grad_norm": 0.656197726726532,
+      "learning_rate": 0.00019658565883710005,
+      "loss": 6.0684,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6126358287162307,
+      "grad_norm": 0.6660633683204651,
+      "learning_rate": 0.00019629273732559973,
+      "loss": 6.0719,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6129794270497788,
+      "grad_norm": 0.6405670642852783,
+      "learning_rate": 0.0001959998930931387,
+      "loss": 6.0928,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6133230253833268,
+      "grad_norm": 0.6059765219688416,
+      "learning_rate": 0.0001957071265610886,
+      "loss": 6.2629,
+      "step": 1785
+    },
+    {
+      "epoch": 0.613666623716875,
+      "grad_norm": 0.6842978000640869,
+      "learning_rate": 0.00019541443815070952,
+      "loss": 6.3053,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6140102220504231,
+      "grad_norm": 0.6792712211608887,
+      "learning_rate": 0.00019512182828314883,
+      "loss": 6.1212,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6143538203839711,
+      "grad_norm": 0.6276842951774597,
+      "learning_rate": 0.00019482929737944094,
+      "loss": 6.1975,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6146974187175193,
+      "grad_norm": 0.6486376523971558,
+      "learning_rate": 0.00019453684586050692,
+      "loss": 6.0693,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6150410170510673,
+      "grad_norm": 0.6076637506484985,
+      "learning_rate": 0.00019424447414715323,
+      "loss": 6.1387,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.7046791911125183,
+      "learning_rate": 0.0001939521826600717,
+      "loss": 6.1665,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6157282137181634,
+      "grad_norm": 0.7905498147010803,
+      "learning_rate": 0.00019365997181983874,
+      "loss": 6.2077,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6160718120517116,
+      "grad_norm": 0.6241534352302551,
+      "learning_rate": 0.0001933678420469147,
+      "loss": 6.1993,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6164154103852596,
+      "grad_norm": 0.7369571328163147,
+      "learning_rate": 0.00019307579376164313,
+      "loss": 6.2529,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6167590087188077,
+      "grad_norm": 0.8129594326019287,
+      "learning_rate": 0.00019278382738425063,
+      "loss": 6.305,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6171026070523558,
+      "grad_norm": 0.750092089176178,
+      "learning_rate": 0.00019249194333484566,
+      "loss": 6.3213,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6174462053859039,
+      "grad_norm": 0.9499820470809937,
+      "learning_rate": 0.00019220014203341824,
+      "loss": 6.2478,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6177898037194519,
+      "grad_norm": 0.764100193977356,
+      "learning_rate": 0.0001919084238998396,
+      "loss": 6.2935,
+      "step": 1798
+    },
+    {
+      "epoch": 0.618133402053,
+      "grad_norm": 0.8200209140777588,
+      "learning_rate": 0.00019161678935386098,
+      "loss": 6.2591,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6184770003865481,
+      "grad_norm": 1.2245675325393677,
+      "learning_rate": 0.00019132523881511344,
+      "loss": 6.0823,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6188205987200962,
+      "grad_norm": 0.81374192237854,
+      "learning_rate": 0.0001910337727031074,
+      "loss": 6.0214,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6191641970536443,
+      "grad_norm": 0.6380704641342163,
+      "learning_rate": 0.00019074239143723144,
+      "loss": 5.9451,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6195077953871924,
+      "grad_norm": 0.7890001535415649,
+      "learning_rate": 0.00019045109543675215,
+      "loss": 6.1033,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6198513937207405,
+      "grad_norm": 0.7154576778411865,
+      "learning_rate": 0.00019015988512081369,
+      "loss": 6.0945,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6201949920542885,
+      "grad_norm": 0.6763778924942017,
+      "learning_rate": 0.00018986876090843667,
+      "loss": 5.9776,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6205385903878367,
+      "grad_norm": 0.6738643050193787,
+      "learning_rate": 0.00018957772321851767,
+      "loss": 6.0804,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6208821887213847,
+      "grad_norm": 0.7186247706413269,
+      "learning_rate": 0.0001892867724698293,
+      "loss": 5.9855,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6212257870549328,
+      "grad_norm": 0.556603729724884,
+      "learning_rate": 0.00018899590908101851,
+      "loss": 6.1443,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6215693853884808,
+      "grad_norm": 0.5243987441062927,
+      "learning_rate": 0.0001887051334706069,
+      "loss": 6.032,
+      "step": 1809
+    },
+    {
+      "epoch": 0.621912983722029,
+      "grad_norm": 0.7323964834213257,
+      "learning_rate": 0.0001884144460569896,
+      "loss": 6.0841,
+      "step": 1810
+    },
+    {
+      "epoch": 0.622256582055577,
+      "grad_norm": 0.6673109531402588,
+      "learning_rate": 0.00018812384725843488,
+      "loss": 6.0665,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6226001803891251,
+      "grad_norm": 0.5664012432098389,
+      "learning_rate": 0.00018783333749308357,
+      "loss": 6.0865,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6229437787226731,
+      "grad_norm": 0.6237545013427734,
+      "learning_rate": 0.00018754291717894826,
+      "loss": 6.1261,
+      "step": 1813
+    },
+    {
+      "epoch": 0.6232873770562213,
+      "grad_norm": 0.7225657105445862,
+      "learning_rate": 0.00018725258673391281,
+      "loss": 6.1274,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6236309753897693,
+      "grad_norm": 0.5414710640907288,
+      "learning_rate": 0.00018696234657573208,
+      "loss": 6.1915,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6239745737233174,
+      "grad_norm": 0.587091326713562,
+      "learning_rate": 0.00018667219712203064,
+      "loss": 6.088,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6243181720568656,
+      "grad_norm": 0.6084412932395935,
+      "learning_rate": 0.00018638213879030265,
+      "loss": 6.1797,
+      "step": 1817
+    },
+    {
+      "epoch": 0.6246617703904136,
+      "grad_norm": 0.6020777821540833,
+      "learning_rate": 0.00018609217199791136,
+      "loss": 6.1391,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6250053687239617,
+      "grad_norm": 0.5605757832527161,
+      "learning_rate": 0.00018580229716208806,
+      "loss": 6.1469,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6253489670575098,
+      "grad_norm": 0.634785532951355,
+      "learning_rate": 0.00018551251469993175,
+      "loss": 6.044,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6256925653910579,
+      "grad_norm": 0.5692383050918579,
+      "learning_rate": 0.00018522282502840873,
+      "loss": 6.1452,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6260361637246059,
+      "grad_norm": 0.5231955051422119,
+      "learning_rate": 0.00018493322856435155,
+      "loss": 6.0816,
+      "step": 1822
+    },
+    {
+      "epoch": 0.626379762058154,
+      "grad_norm": 0.5089519619941711,
+      "learning_rate": 0.00018464372572445865,
+      "loss": 6.1039,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6267233603917021,
+      "grad_norm": 0.5622355341911316,
+      "learning_rate": 0.00018435431692529386,
+      "loss": 6.1048,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6270669587252502,
+      "grad_norm": 0.543976366519928,
+      "learning_rate": 0.0001840650025832858,
+      "loss": 6.132,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6274105570587982,
+      "grad_norm": 0.609881579875946,
+      "learning_rate": 0.00018377578311472683,
+      "loss": 6.0715,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6277541553923464,
+      "grad_norm": 0.44123703241348267,
+      "learning_rate": 0.0001834866589357732,
+      "loss": 6.1373,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6280977537258944,
+      "grad_norm": 0.6479792594909668,
+      "learning_rate": 0.0001831976304624438,
+      "loss": 6.1722,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6284413520594425,
+      "grad_norm": 0.4903275966644287,
+      "learning_rate": 0.00018290869811061968,
+      "loss": 6.2031,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6287849503929906,
+      "grad_norm": 0.4645865261554718,
+      "learning_rate": 0.00018261986229604402,
+      "loss": 6.1344,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6291285487265387,
+      "grad_norm": 0.5052070617675781,
+      "learning_rate": 0.00018233112343432077,
+      "loss": 6.0615,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6294721470600868,
+      "grad_norm": 0.6264731287956238,
+      "learning_rate": 0.00018204248194091428,
+      "loss": 6.1618,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6298157453936348,
+      "grad_norm": 0.5955216884613037,
+      "learning_rate": 0.0001817539382311492,
+      "loss": 6.1757,
+      "step": 1833
+    },
+    {
+      "epoch": 0.630159343727183,
+      "grad_norm": 0.523066520690918,
+      "learning_rate": 0.00018146549272020918,
+      "loss": 6.1458,
+      "step": 1834
+    },
+    {
+      "epoch": 0.630502942060731,
+      "grad_norm": 0.583475649356842,
+      "learning_rate": 0.0001811771458231365,
+      "loss": 6.1725,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6308465403942791,
+      "grad_norm": 0.6626928448677063,
+      "learning_rate": 0.00018088889795483184,
+      "loss": 6.1341,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6311901387278271,
+      "grad_norm": 0.6540782451629639,
+      "learning_rate": 0.00018060074953005307,
+      "loss": 6.2446,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6315337370613753,
+      "grad_norm": 0.6838895082473755,
+      "learning_rate": 0.00018031270096341534,
+      "loss": 6.1862,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6318773353949233,
+      "grad_norm": 0.6299639940261841,
+      "learning_rate": 0.00018002475266938977,
+      "loss": 6.19,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6322209337284714,
+      "grad_norm": 0.6744266152381897,
+      "learning_rate": 0.0001797369050623033,
+      "loss": 6.2666,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6325645320620195,
+      "grad_norm": 0.5881931185722351,
+      "learning_rate": 0.0001794491585563381,
+      "loss": 6.0901,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6329081303955676,
+      "grad_norm": 0.5639910697937012,
+      "learning_rate": 0.00017916151356553073,
+      "loss": 6.0917,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6332517287291156,
+      "grad_norm": 0.8729662895202637,
+      "learning_rate": 0.0001788739705037718,
+      "loss": 6.1398,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6335953270626637,
+      "grad_norm": 0.6799281239509583,
+      "learning_rate": 0.00017858652978480516,
+      "loss": 6.1564,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6339389253962119,
+      "grad_norm": 0.7408438324928284,
+      "learning_rate": 0.00017829919182222752,
+      "loss": 6.1424,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6342825237297599,
+      "grad_norm": 0.6003269553184509,
+      "learning_rate": 0.00017801195702948742,
+      "loss": 6.1305,
+      "step": 1846
+    },
+    {
+      "epoch": 0.634626122063308,
+      "grad_norm": 0.8122835755348206,
+      "learning_rate": 0.00017772482581988544,
+      "loss": 6.2311,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6349697203968561,
+      "grad_norm": 0.8266429901123047,
+      "learning_rate": 0.0001774377986065728,
+      "loss": 6.1746,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6353133187304042,
+      "grad_norm": 0.9495203495025635,
+      "learning_rate": 0.0001771508758025509,
+      "loss": 6.2391,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6356569170639522,
+      "grad_norm": 1.0640305280685425,
+      "learning_rate": 0.0001768640578206715,
+      "loss": 6.1779,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6360005153975004,
+      "grad_norm": 1.1024693250656128,
+      "learning_rate": 0.00017657734507363498,
+      "loss": 5.9968,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6363441137310484,
+      "grad_norm": 1.1050010919570923,
+      "learning_rate": 0.00017629073797399036,
+      "loss": 6.1062,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6366877120645965,
+      "grad_norm": 0.9328210949897766,
+      "learning_rate": 0.00017600423693413508,
+      "loss": 6.0665,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6370313103981445,
+      "grad_norm": 0.6804933547973633,
+      "learning_rate": 0.00017571784236631351,
+      "loss": 5.9694,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6373749087316927,
+      "grad_norm": 0.836593747138977,
+      "learning_rate": 0.00017543155468261696,
+      "loss": 5.8933,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6377185070652407,
+      "grad_norm": 0.7389679551124573,
+      "learning_rate": 0.00017514537429498295,
+      "loss": 6.0701,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6380621053987888,
+      "grad_norm": 0.769934356212616,
+      "learning_rate": 0.0001748593016151947,
+      "loss": 6.0735,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6384057037323368,
+      "grad_norm": 0.7202440500259399,
+      "learning_rate": 0.00017457333705488026,
+      "loss": 5.9477,
+      "step": 1858
+    },
+    {
+      "epoch": 0.638749302065885,
+      "grad_norm": 0.7156488299369812,
+      "learning_rate": 0.00017428748102551236,
+      "loss": 6.1103,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6390929003994331,
+      "grad_norm": 0.7631789445877075,
+      "learning_rate": 0.00017400173393840735,
+      "loss": 6.0724,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6394364987329811,
+      "grad_norm": 0.6912445425987244,
+      "learning_rate": 0.00017371609620472477,
+      "loss": 5.9004,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6397800970665293,
+      "grad_norm": 0.6902914047241211,
+      "learning_rate": 0.00017343056823546725,
+      "loss": 6.1013,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6401236954000773,
+      "grad_norm": 0.5509893894195557,
+      "learning_rate": 0.00017314515044147884,
+      "loss": 6.0479,
+      "step": 1863
+    },
+    {
+      "epoch": 0.6404672937336254,
+      "grad_norm": 0.5896221995353699,
+      "learning_rate": 0.00017285984323344567,
+      "loss": 6.0818,
+      "step": 1864
+    },
+    {
+      "epoch": 0.6408108920671735,
+      "grad_norm": 0.6542482972145081,
+      "learning_rate": 0.00017257464702189433,
+      "loss": 6.1131,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6411544904007216,
+      "grad_norm": 0.5327386856079102,
+      "learning_rate": 0.00017228956221719178,
+      "loss": 6.1203,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6414980887342696,
+      "grad_norm": 0.5413427352905273,
+      "learning_rate": 0.00017200458922954486,
+      "loss": 6.2124,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6418416870678177,
+      "grad_norm": 0.4703226685523987,
+      "learning_rate": 0.00017171972846899942,
+      "loss": 6.1323,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6421852854013658,
+      "grad_norm": 0.4837634563446045,
+      "learning_rate": 0.00017143498034543958,
+      "loss": 6.0326,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6425288837349139,
+      "grad_norm": 0.6183565258979797,
+      "learning_rate": 0.00017115034526858785,
+      "loss": 6.0834,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6428724820684619,
+      "grad_norm": 0.6575108170509338,
+      "learning_rate": 0.00017086582364800375,
+      "loss": 6.0392,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6432160804020101,
+      "grad_norm": 0.5329729318618774,
+      "learning_rate": 0.00017058141589308356,
+      "loss": 6.2179,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6435596787355581,
+      "grad_norm": 0.5143067240715027,
+      "learning_rate": 0.0001702971224130599,
+      "loss": 5.9996,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6439032770691062,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001700129436170008,
+      "loss": 6.1188,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6442468754026544,
+      "grad_norm": 0.5242812037467957,
+      "learning_rate": 0.0001697288799138093,
+      "loss": 6.0961,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6445904737362024,
+      "grad_norm": 0.45744621753692627,
+      "learning_rate": 0.00016944493171222296,
+      "loss": 6.0911,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6449340720697505,
+      "grad_norm": 0.4443244934082031,
+      "learning_rate": 0.00016916109942081292,
+      "loss": 6.1137,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6452776704032985,
+      "grad_norm": 0.668917179107666,
+      "learning_rate": 0.0001688773834479837,
+      "loss": 5.9674,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6456212687368467,
+      "grad_norm": 0.5518603920936584,
+      "learning_rate": 0.00016859378420197246,
+      "loss": 6.0861,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6459648670703947,
+      "grad_norm": 0.5576813817024231,
+      "learning_rate": 0.0001683103020908484,
+      "loss": 6.0854,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6463084654039428,
+      "grad_norm": 0.5512084364891052,
+      "learning_rate": 0.00016802693752251187,
+      "loss": 5.9926,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6466520637374908,
+      "grad_norm": 0.7495642304420471,
+      "learning_rate": 0.0001677436909046947,
+      "loss": 6.1177,
+      "step": 1882
+    },
+    {
+      "epoch": 0.646995662071039,
+      "grad_norm": 0.6886982917785645,
+      "learning_rate": 0.00016746056264495846,
+      "loss": 6.1576,
+      "step": 1883
+    },
+    {
+      "epoch": 0.647339260404587,
+      "grad_norm": 0.5660991072654724,
+      "learning_rate": 0.00016717755315069456,
+      "loss": 6.015,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6476828587381351,
+      "grad_norm": 0.6272467374801636,
+      "learning_rate": 0.00016689466282912368,
+      "loss": 6.2081,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6480264570716832,
+      "grad_norm": 0.6678543090820312,
+      "learning_rate": 0.0001666118920872949,
+      "loss": 6.1756,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6483700554052313,
+      "grad_norm": 0.6672770977020264,
+      "learning_rate": 0.00016632924133208515,
+      "loss": 6.1607,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6487136537387793,
+      "grad_norm": 0.8277221918106079,
+      "learning_rate": 0.00016604671097019885,
+      "loss": 6.0318,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6490572520723275,
+      "grad_norm": 0.600071370601654,
+      "learning_rate": 0.00016576430140816716,
+      "loss": 6.1854,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6494008504058756,
+      "grad_norm": 0.6369357705116272,
+      "learning_rate": 0.0001654820130523475,
+      "loss": 6.2219,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6497444487394236,
+      "grad_norm": 0.6288464665412903,
+      "learning_rate": 0.00016519984630892264,
+      "loss": 6.0491,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6500880470729717,
+      "grad_norm": 0.7976536750793457,
+      "learning_rate": 0.0001649178015839005,
+      "loss": 6.0651,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6504316454065198,
+      "grad_norm": 0.7433154582977295,
+      "learning_rate": 0.00016463587928311363,
+      "loss": 6.2634,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6507752437400679,
+      "grad_norm": 0.8158239722251892,
+      "learning_rate": 0.0001643540798122181,
+      "loss": 6.0948,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6511188420736159,
+      "grad_norm": 0.9072747230529785,
+      "learning_rate": 0.00016407240357669333,
+      "loss": 6.1638,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6514624404071641,
+      "grad_norm": 0.8244097828865051,
+      "learning_rate": 0.00016379085098184166,
+      "loss": 6.1314,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6518060387407121,
+      "grad_norm": 0.7535924911499023,
+      "learning_rate": 0.0001635094224327872,
+      "loss": 6.1109,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6521496370742602,
+      "grad_norm": 1.0342903137207031,
+      "learning_rate": 0.0001632281183344756,
+      "loss": 6.2378,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6524932354078082,
+      "grad_norm": 0.8878806233406067,
+      "learning_rate": 0.00016294693909167378,
+      "loss": 6.2589,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6528368337413564,
+      "grad_norm": 0.8649120330810547,
+      "learning_rate": 0.00016266588510896864,
+      "loss": 6.335,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6531804320749044,
+      "grad_norm": 1.5248429775238037,
+      "learning_rate": 0.00016238495679076688,
+      "loss": 6.0537,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6535240304084525,
+      "grad_norm": 1.4273432493209839,
+      "learning_rate": 0.00016210415454129463,
+      "loss": 5.9864,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6538676287420006,
+      "grad_norm": 0.9812554717063904,
+      "learning_rate": 0.00016182347876459648,
+      "loss": 6.0068,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6542112270755487,
+      "grad_norm": 0.7029387354850769,
+      "learning_rate": 0.00016154292986453485,
+      "loss": 6.1058,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6545548254090968,
+      "grad_norm": 1.040303111076355,
+      "learning_rate": 0.0001612625082447899,
+      "loss": 6.0546,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6548984237426448,
+      "grad_norm": 1.0513960123062134,
+      "learning_rate": 0.00016098221430885844,
+      "loss": 5.9321,
+      "step": 1906
+    },
+    {
+      "epoch": 0.655242022076193,
+      "grad_norm": 0.7275641560554504,
+      "learning_rate": 0.00016070204846005374,
+      "loss": 6.0139,
+      "step": 1907
+    },
+    {
+      "epoch": 0.655585620409741,
+      "grad_norm": 0.8591195940971375,
+      "learning_rate": 0.0001604220111015046,
+      "loss": 5.992,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6559292187432891,
+      "grad_norm": 0.9006431698799133,
+      "learning_rate": 0.00016014210263615505,
+      "loss": 5.9939,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6562728170768372,
+      "grad_norm": 0.611177384853363,
+      "learning_rate": 0.00015986232346676345,
+      "loss": 5.9714,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6566164154103853,
+      "grad_norm": 0.555124819278717,
+      "learning_rate": 0.00015958267399590243,
+      "loss": 5.9323,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6569600137439333,
+      "grad_norm": 0.6982326507568359,
+      "learning_rate": 0.0001593031546259578,
+      "loss": 5.9022,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6573036120774814,
+      "grad_norm": 0.7364804744720459,
+      "learning_rate": 0.00015902376575912814,
+      "loss": 5.9534,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6576472104110295,
+      "grad_norm": 0.6368590593338013,
+      "learning_rate": 0.0001587445077974244,
+      "loss": 6.0573,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6579908087445776,
+      "grad_norm": 0.596760094165802,
+      "learning_rate": 0.00015846538114266912,
+      "loss": 6.1416,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6583344070781256,
+      "grad_norm": 0.576379656791687,
+      "learning_rate": 0.00015818638619649567,
+      "loss": 6.0386,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6586780054116738,
+      "grad_norm": 0.5851263999938965,
+      "learning_rate": 0.00015790752336034835,
+      "loss": 6.1689,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6590216037452218,
+      "grad_norm": 0.6037353873252869,
+      "learning_rate": 0.00015762879303548094,
+      "loss": 6.1273,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6593652020787699,
+      "grad_norm": 0.533187747001648,
+      "learning_rate": 0.00015735019562295688,
+      "loss": 5.9624,
+      "step": 1919
+    },
+    {
+      "epoch": 0.659708800412318,
+      "grad_norm": 0.6187282204627991,
+      "learning_rate": 0.00015707173152364816,
+      "loss": 6.1505,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6600523987458661,
+      "grad_norm": 0.5342614054679871,
+      "learning_rate": 0.00015679340113823495,
+      "loss": 6.2201,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6603959970794142,
+      "grad_norm": 0.536963939666748,
+      "learning_rate": 0.00015651520486720515,
+      "loss": 6.1204,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6607395954129622,
+      "grad_norm": 0.5395856499671936,
+      "learning_rate": 0.00015623714311085364,
+      "loss": 5.9842,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6610831937465104,
+      "grad_norm": 0.5402804613113403,
+      "learning_rate": 0.0001559592162692815,
+      "loss": 6.1283,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6614267920800584,
+      "grad_norm": 0.4991236925125122,
+      "learning_rate": 0.00015568142474239622,
+      "loss": 5.9874,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6617703904136065,
+      "grad_norm": 0.5533581376075745,
+      "learning_rate": 0.00015540376892991004,
+      "loss": 6.0261,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6621139887471545,
+      "grad_norm": 0.549507737159729,
+      "learning_rate": 0.0001551262492313401,
+      "loss": 6.097,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6624575870807027,
+      "grad_norm": 0.5345736742019653,
+      "learning_rate": 0.00015484886604600796,
+      "loss": 6.1086,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6628011854142507,
+      "grad_norm": 0.4900307357311249,
+      "learning_rate": 0.0001545716197730384,
+      "loss": 6.0988,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6631447837477988,
+      "grad_norm": 0.468268483877182,
+      "learning_rate": 0.00015429451081135922,
+      "loss": 6.0424,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6634883820813469,
+      "grad_norm": 0.5394709706306458,
+      "learning_rate": 0.00015401753955970095,
+      "loss": 6.1305,
+      "step": 1931
+    },
+    {
+      "epoch": 0.663831980414895,
+      "grad_norm": 0.5436084866523743,
+      "learning_rate": 0.00015374070641659566,
+      "loss": 6.0656,
+      "step": 1932
+    },
+    {
+      "epoch": 0.664175578748443,
+      "grad_norm": 0.5724718570709229,
+      "learning_rate": 0.00015346401178037672,
+      "loss": 5.9819,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6645191770819912,
+      "grad_norm": 0.5867025256156921,
+      "learning_rate": 0.00015318745604917848,
+      "loss": 6.1125,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6648627754155393,
+      "grad_norm": 0.5411561131477356,
+      "learning_rate": 0.0001529110396209351,
+      "loss": 6.1245,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6652063737490873,
+      "grad_norm": 0.5726490616798401,
+      "learning_rate": 0.0001526347628933804,
+      "loss": 6.1578,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6655499720826354,
+      "grad_norm": 0.615553081035614,
+      "learning_rate": 0.00015235862626404727,
+      "loss": 6.0445,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6658935704161835,
+      "grad_norm": 0.5673866868019104,
+      "learning_rate": 0.00015208263013026692,
+      "loss": 6.1863,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6662371687497316,
+      "grad_norm": 0.5758171677589417,
+      "learning_rate": 0.00015180677488916845,
+      "loss": 6.1639,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6665807670832796,
+      "grad_norm": 0.5809786319732666,
+      "learning_rate": 0.00015153106093767827,
+      "loss": 6.0779,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6669243654168278,
+      "grad_norm": 0.6912338733673096,
+      "learning_rate": 0.00015125548867251935,
+      "loss": 6.134,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6672679637503758,
+      "grad_norm": 0.7445138096809387,
+      "learning_rate": 0.0001509800584902108,
+      "loss": 6.1787,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6676115620839239,
+      "grad_norm": 0.665468156337738,
+      "learning_rate": 0.00015070477078706757,
+      "loss": 6.1019,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6679551604174719,
+      "grad_norm": 0.6977120041847229,
+      "learning_rate": 0.00015042962595919918,
+      "loss": 6.1073,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6682987587510201,
+      "grad_norm": 0.6922124028205872,
+      "learning_rate": 0.00015015462440250997,
+      "loss": 6.21,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6686423570845681,
+      "grad_norm": 0.6119797825813293,
+      "learning_rate": 0.00014987976651269788,
+      "loss": 6.154,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6689859554181162,
+      "grad_norm": 0.6788223385810852,
+      "learning_rate": 0.000149605052685254,
+      "loss": 6.0747,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6693295537516643,
+      "grad_norm": 0.9458711743354797,
+      "learning_rate": 0.00014933048331546258,
+      "loss": 6.2254,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6696731520852124,
+      "grad_norm": 0.9792423844337463,
+      "learning_rate": 0.0001490560587983996,
+      "loss": 6.296,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6700167504187605,
+      "grad_norm": 1.0456666946411133,
+      "learning_rate": 0.00014878177952893276,
+      "loss": 6.3253,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6703603487523085,
+      "grad_norm": 0.954923689365387,
+      "learning_rate": 0.00014850764590172076,
+      "loss": 5.9976,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6707039470858567,
+      "grad_norm": 0.8935602307319641,
+      "learning_rate": 0.00014823365831121278,
+      "loss": 6.1013,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6710475454194047,
+      "grad_norm": 0.7865163087844849,
+      "learning_rate": 0.00014795981715164775,
+      "loss": 5.9458,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6713911437529528,
+      "grad_norm": 0.7201636433601379,
+      "learning_rate": 0.00014768612281705406,
+      "loss": 6.1155,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6717347420865009,
+      "grad_norm": 0.7660732269287109,
+      "learning_rate": 0.00014741257570124875,
+      "loss": 6.0565,
+      "step": 1955
+    },
+    {
+      "epoch": 0.672078340420049,
+      "grad_norm": 0.6856514811515808,
+      "learning_rate": 0.00014713917619783685,
+      "loss": 6.012,
+      "step": 1956
+    },
+    {
+      "epoch": 0.672421938753597,
+      "grad_norm": 0.716894805431366,
+      "learning_rate": 0.00014686592470021143,
+      "loss": 6.1342,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6727655370871451,
+      "grad_norm": 0.715986430644989,
+      "learning_rate": 0.00014659282160155222,
+      "loss": 6.0448,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6731091354206932,
+      "grad_norm": 0.6127437949180603,
+      "learning_rate": 0.0001463198672948254,
+      "loss": 6.1887,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6734527337542413,
+      "grad_norm": 0.7476091384887695,
+      "learning_rate": 0.00014604706217278345,
+      "loss": 6.022,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6737963320877893,
+      "grad_norm": 0.5875628590583801,
+      "learning_rate": 0.0001457744066279637,
+      "loss": 6.0752,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6741399304213375,
+      "grad_norm": 0.6040765047073364,
+      "learning_rate": 0.00014550190105268863,
+      "loss": 6.0603,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6744835287548855,
+      "grad_norm": 0.5486308336257935,
+      "learning_rate": 0.0001452295458390648,
+      "loss": 6.2029,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6748271270884336,
+      "grad_norm": 0.5594898462295532,
+      "learning_rate": 0.00014495734137898227,
+      "loss": 6.077,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6751707254219818,
+      "grad_norm": 0.5191746354103088,
+      "learning_rate": 0.0001446852880641143,
+      "loss": 6.0956,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6755143237555298,
+      "grad_norm": 0.5390771627426147,
+      "learning_rate": 0.0001444133862859168,
+      "loss": 6.0423,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6758579220890779,
+      "grad_norm": 0.48739349842071533,
+      "learning_rate": 0.00014414163643562756,
+      "loss": 6.0822,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6762015204226259,
+      "grad_norm": 0.5168871283531189,
+      "learning_rate": 0.00014387003890426538,
+      "loss": 6.106,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6765451187561741,
+      "grad_norm": 0.5096775889396667,
+      "learning_rate": 0.00014359859408263068,
+      "loss": 6.1346,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6768887170897221,
+      "grad_norm": 0.48305463790893555,
+      "learning_rate": 0.00014332730236130337,
+      "loss": 6.122,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6772323154232702,
+      "grad_norm": 0.5125235915184021,
+      "learning_rate": 0.00014305616413064345,
+      "loss": 6.2059,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6775759137568182,
+      "grad_norm": 0.5879007577896118,
+      "learning_rate": 0.00014278517978079006,
+      "loss": 5.9098,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6779195120903664,
+      "grad_norm": 0.48073089122772217,
+      "learning_rate": 0.00014251434970166083,
+      "loss": 6.0738,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6782631104239144,
+      "grad_norm": 0.5409465432167053,
+      "learning_rate": 0.00014224367428295143,
+      "loss": 5.9913,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6786067087574625,
+      "grad_norm": 0.5101212859153748,
+      "learning_rate": 0.00014197315391413512,
+      "loss": 6.0706,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6789503070910106,
+      "grad_norm": 0.540166974067688,
+      "learning_rate": 0.00014170278898446175,
+      "loss": 6.094,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6792939054245587,
+      "grad_norm": 0.5894446969032288,
+      "learning_rate": 0.00014143257988295777,
+      "loss": 6.1305,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6796375037581067,
+      "grad_norm": 0.5489010214805603,
+      "learning_rate": 0.00014116252699842546,
+      "loss": 6.1303,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6799811020916549,
+      "grad_norm": 0.5311551690101624,
+      "learning_rate": 0.00014089263071944192,
+      "loss": 6.1514,
+      "step": 1979
+    },
+    {
+      "epoch": 0.680324700425203,
+      "grad_norm": 0.5869585871696472,
+      "learning_rate": 0.00014062289143435957,
+      "loss": 6.0273,
+      "step": 1980
+    },
+    {
+      "epoch": 0.680668298758751,
+      "grad_norm": 0.6781054735183716,
+      "learning_rate": 0.00014035330953130422,
+      "loss": 6.0957,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6810118970922991,
+      "grad_norm": 0.6006012558937073,
+      "learning_rate": 0.00014008388539817575,
+      "loss": 6.0071,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6813554954258472,
+      "grad_norm": 0.5926790833473206,
+      "learning_rate": 0.00013981461942264673,
+      "loss": 6.1698,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6816990937593953,
+      "grad_norm": 0.596321165561676,
+      "learning_rate": 0.00013954551199216246,
+      "loss": 6.046,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6820426920929433,
+      "grad_norm": 0.5697970986366272,
+      "learning_rate": 0.00013927656349393952,
+      "loss": 6.1458,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6823862904264915,
+      "grad_norm": 0.6854418516159058,
+      "learning_rate": 0.00013900777431496666,
+      "loss": 6.0886,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6827298887600395,
+      "grad_norm": 0.7039015889167786,
+      "learning_rate": 0.00013873914484200262,
+      "loss": 6.1631,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6830734870935876,
+      "grad_norm": 0.6003180146217346,
+      "learning_rate": 0.00013847067546157672,
+      "loss": 6.0595,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6834170854271356,
+      "grad_norm": 0.6838335394859314,
+      "learning_rate": 0.00013820236655998785,
+      "loss": 6.1989,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6837606837606838,
+      "grad_norm": 0.6573136448860168,
+      "learning_rate": 0.0001379342185233041,
+      "loss": 6.0901,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6841042820942318,
+      "grad_norm": 0.5662709474563599,
+      "learning_rate": 0.00013766623173736177,
+      "loss": 6.1881,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6844478804277799,
+      "grad_norm": 0.6747244596481323,
+      "learning_rate": 0.0001373984065877654,
+      "loss": 6.0781,
+      "step": 1992
+    },
+    {
+      "epoch": 0.684791478761328,
+      "grad_norm": 0.624338686466217,
+      "learning_rate": 0.000137130743459887,
+      "loss": 6.1607,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6851350770948761,
+      "grad_norm": 0.7191410064697266,
+      "learning_rate": 0.0001368632427388653,
+      "loss": 6.0574,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6854786754284242,
+      "grad_norm": 0.6128062009811401,
+      "learning_rate": 0.00013659590480960543,
+      "loss": 6.1612,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6858222737619722,
+      "grad_norm": 0.7554376721382141,
+      "learning_rate": 0.0001363287300567781,
+      "loss": 6.1125,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6861658720955204,
+      "grad_norm": 0.7863120436668396,
+      "learning_rate": 0.00013606171886481943,
+      "loss": 6.1154,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6865094704290684,
+      "grad_norm": 0.7225969433784485,
+      "learning_rate": 0.00013579487161793018,
+      "loss": 6.1776,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6868530687626165,
+      "grad_norm": 0.9673846364021301,
+      "learning_rate": 0.00013552818870007514,
+      "loss": 6.1753,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6871966670961646,
+      "grad_norm": 1.0428494215011597,
+      "learning_rate": 0.00013526167049498263,
+      "loss": 6.1562,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6875402654297127,
+      "grad_norm": 0.9498250484466553,
+      "learning_rate": 0.00013499531738614414,
+      "loss": 6.0596,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6878838637632607,
+      "grad_norm": 0.8821714520454407,
+      "learning_rate": 0.00013472912975681317,
+      "loss": 6.052,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6882274620968089,
+      "grad_norm": 0.7349355220794678,
+      "learning_rate": 0.00013446310799000578,
+      "loss": 6.0121,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6885710604303569,
+      "grad_norm": 0.5704671740531921,
+      "learning_rate": 0.00013419725246849873,
+      "loss": 6.0056,
+      "step": 2004
+    },
+    {
+      "epoch": 0.688914658763905,
+      "grad_norm": 0.5243590474128723,
+      "learning_rate": 0.00013393156357482993,
+      "loss": 6.0066,
+      "step": 2005
+    },
+    {
+      "epoch": 0.689258257097453,
+      "grad_norm": 0.679142415523529,
+      "learning_rate": 0.00013366604169129742,
+      "loss": 5.9651,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6896018554310012,
+      "grad_norm": 0.7514729499816895,
+      "learning_rate": 0.00013340068719995912,
+      "loss": 6.0792,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6899454537645492,
+      "grad_norm": 0.7526649832725525,
+      "learning_rate": 0.00013313550048263168,
+      "loss": 5.9845,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6902890520980973,
+      "grad_norm": 0.7030274868011475,
+      "learning_rate": 0.00013287048192089064,
+      "loss": 5.9931,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6906326504316455,
+      "grad_norm": 0.7123146653175354,
+      "learning_rate": 0.0001326056318960697,
+      "loss": 5.9611,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6909762487651935,
+      "grad_norm": 0.7373788356781006,
+      "learning_rate": 0.00013234095078925952,
+      "loss": 5.9859,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6913198470987416,
+      "grad_norm": 0.6924092769622803,
+      "learning_rate": 0.00013207643898130854,
+      "loss": 6.1251,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6916634454322896,
+      "grad_norm": 0.6907500624656677,
+      "learning_rate": 0.00013181209685282074,
+      "loss": 5.9978,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6920070437658378,
+      "grad_norm": 0.4841921329498291,
+      "learning_rate": 0.00013154792478415646,
+      "loss": 6.1378,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6923506420993858,
+      "grad_norm": 0.5584970712661743,
+      "learning_rate": 0.00013128392315543125,
+      "loss": 6.0776,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6926942404329339,
+      "grad_norm": 0.6407040357589722,
+      "learning_rate": 0.00013102009234651542,
+      "loss": 6.0172,
+      "step": 2016
+    },
+    {
+      "epoch": 0.693037838766482,
+      "grad_norm": 0.663905918598175,
+      "learning_rate": 0.00013075643273703316,
+      "loss": 5.8666,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6933814371000301,
+      "grad_norm": 0.6265025734901428,
+      "learning_rate": 0.00013049294470636303,
+      "loss": 6.1306,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6937250354335781,
+      "grad_norm": 0.5826771259307861,
+      "learning_rate": 0.00013022962863363597,
+      "loss": 6.0611,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6940686337671262,
+      "grad_norm": 0.515347421169281,
+      "learning_rate": 0.00012996648489773595,
+      "loss": 6.1146,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6944122321006743,
+      "grad_norm": 0.623602569103241,
+      "learning_rate": 0.00012970351387729873,
+      "loss": 6.0535,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6947558304342224,
+      "grad_norm": 0.4509178102016449,
+      "learning_rate": 0.0001294407159507118,
+      "loss": 6.0843,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6950994287677704,
+      "grad_norm": 0.5675387382507324,
+      "learning_rate": 0.00012917809149611323,
+      "loss": 6.0791,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6954430271013186,
+      "grad_norm": 0.4827669858932495,
+      "learning_rate": 0.0001289156408913918,
+      "loss": 6.0356,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6957866254348667,
+      "grad_norm": 0.558542788028717,
+      "learning_rate": 0.00012865336451418593,
+      "loss": 6.0506,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6961302237684147,
+      "grad_norm": 0.5373152494430542,
+      "learning_rate": 0.00012839126274188353,
+      "loss": 5.9674,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6964738221019628,
+      "grad_norm": 0.5774347186088562,
+      "learning_rate": 0.00012812933595162125,
+      "loss": 6.0237,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6968174204355109,
+      "grad_norm": 0.5224197506904602,
+      "learning_rate": 0.00012786758452028354,
+      "loss": 6.1522,
+      "step": 2028
+    },
+    {
+      "epoch": 0.697161018769059,
+      "grad_norm": 0.6040776371955872,
+      "learning_rate": 0.00012760600882450335,
+      "loss": 6.1,
+      "step": 2029
+    },
+    {
+      "epoch": 0.697504617102607,
+      "grad_norm": 0.44605669379234314,
+      "learning_rate": 0.00012734460924065992,
+      "loss": 6.1146,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6978482154361552,
+      "grad_norm": 0.5897433161735535,
+      "learning_rate": 0.00012708338614487958,
+      "loss": 5.9674,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6981918137697032,
+      "grad_norm": 0.5256454348564148,
+      "learning_rate": 0.00012682233991303458,
+      "loss": 6.1621,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6985354121032513,
+      "grad_norm": 0.566554069519043,
+      "learning_rate": 0.00012656147092074277,
+      "loss": 6.0676,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6988790104367993,
+      "grad_norm": 0.4864864945411682,
+      "learning_rate": 0.00012630077954336666,
+      "loss": 6.1398,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6992226087703475,
+      "grad_norm": 0.5105878710746765,
+      "learning_rate": 0.0001260402661560137,
+      "loss": 6.1083,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6995662071038955,
+      "grad_norm": 0.5783962607383728,
+      "learning_rate": 0.00012577993113353474,
+      "loss": 5.9497,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6999098054374436,
+      "grad_norm": 0.5311532616615295,
+      "learning_rate": 0.00012551977485052428,
+      "loss": 6.0643,
+      "step": 2037
+    },
+    {
+      "epoch": 0.7002534037709917,
+      "grad_norm": 0.550070583820343,
+      "learning_rate": 0.0001252597976813195,
+      "loss": 6.0993,
+      "step": 2038
+    },
+    {
+      "epoch": 0.7005970021045398,
+      "grad_norm": 0.7595521211624146,
+      "learning_rate": 0.00012500000000000006,
+      "loss": 6.2794,
+      "step": 2039
+    },
+    {
+      "epoch": 0.7009406004380879,
+      "grad_norm": 0.575773298740387,
+      "learning_rate": 0.00012474038218038695,
+      "loss": 5.9778,
+      "step": 2040
+    },
+    {
+      "epoch": 0.701284198771636,
+      "grad_norm": 0.6314275860786438,
+      "learning_rate": 0.00012448094459604266,
+      "loss": 6.0772,
+      "step": 2041
+    },
+    {
+      "epoch": 0.7016277971051841,
+      "grad_norm": 0.565686821937561,
+      "learning_rate": 0.0001242216876202705,
+      "loss": 6.1339,
+      "step": 2042
+    },
+    {
+      "epoch": 0.7019713954387321,
+      "grad_norm": 0.6472426652908325,
+      "learning_rate": 0.0001239626116261133,
+      "loss": 6.0905,
+      "step": 2043
+    },
+    {
+      "epoch": 0.7023149937722802,
+      "grad_norm": 0.5603044629096985,
+      "learning_rate": 0.00012370371698635426,
+      "loss": 6.1049,
+      "step": 2044
+    },
+    {
+      "epoch": 0.7026585921058283,
+      "grad_norm": 0.573658287525177,
+      "learning_rate": 0.000123445004073515,
+      "loss": 6.1619,
+      "step": 2045
+    },
+    {
+      "epoch": 0.7030021904393764,
+      "grad_norm": 0.7000228762626648,
+      "learning_rate": 0.00012318647325985593,
+      "loss": 6.1155,
+      "step": 2046
+    },
+    {
+      "epoch": 0.7033457887729244,
+      "grad_norm": 0.7718038558959961,
+      "learning_rate": 0.00012292812491737542,
+      "loss": 6.1795,
+      "step": 2047
+    },
+    {
+      "epoch": 0.7036893871064726,
+      "grad_norm": 0.7203500866889954,
+      "learning_rate": 0.00012266995941780933,
+      "loss": 6.1553,
+      "step": 2048
+    },
+    {
+      "epoch": 0.7040329854400206,
+      "grad_norm": 0.9468965530395508,
+      "learning_rate": 0.0001224119771326301,
+      "loss": 6.3386,
+      "step": 2049
+    },
+    {
+      "epoch": 0.7043765837735687,
+      "grad_norm": 1.2192009687423706,
+      "learning_rate": 0.0001221541784330472,
+      "loss": 6.2315,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7047201821071167,
+      "grad_norm": 0.7734485268592834,
+      "learning_rate": 0.00012189656369000518,
+      "loss": 6.0635,
+      "step": 2051
+    },
+    {
+      "epoch": 0.7050637804406649,
+      "grad_norm": 0.8719174265861511,
+      "learning_rate": 0.00012163913327418443,
+      "loss": 6.0625,
+      "step": 2052
+    },
+    {
+      "epoch": 0.7054073787742129,
+      "grad_norm": 0.7755087614059448,
+      "learning_rate": 0.00012138188755599994,
+      "loss": 5.9011,
+      "step": 2053
+    },
+    {
+      "epoch": 0.705750977107761,
+      "grad_norm": 0.5290982723236084,
+      "learning_rate": 0.00012112482690560089,
+      "loss": 5.9983,
+      "step": 2054
+    },
+    {
+      "epoch": 0.7060945754413092,
+      "grad_norm": 0.6337197422981262,
+      "learning_rate": 0.00012086795169287032,
+      "loss": 6.0545,
+      "step": 2055
+    },
+    {
+      "epoch": 0.7064381737748572,
+      "grad_norm": 0.7570812702178955,
+      "learning_rate": 0.00012061126228742419,
+      "loss": 5.9831,
+      "step": 2056
+    },
+    {
+      "epoch": 0.7067817721084053,
+      "grad_norm": 0.5576236248016357,
+      "learning_rate": 0.00012035475905861134,
+      "loss": 5.9773,
+      "step": 2057
+    },
+    {
+      "epoch": 0.7071253704419533,
+      "grad_norm": 0.544528067111969,
+      "learning_rate": 0.00012009844237551265,
+      "loss": 6.0095,
+      "step": 2058
+    },
+    {
+      "epoch": 0.7074689687755015,
+      "grad_norm": 0.6879101395606995,
+      "learning_rate": 0.00011984231260694061,
+      "loss": 5.9356,
+      "step": 2059
+    },
+    {
+      "epoch": 0.7078125671090495,
+      "grad_norm": 0.5554669499397278,
+      "learning_rate": 0.00011958637012143847,
+      "loss": 6.0969,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7081561654425976,
+      "grad_norm": 0.6475436687469482,
+      "learning_rate": 0.00011933061528728062,
+      "loss": 5.9936,
+      "step": 2061
+    },
+    {
+      "epoch": 0.7084997637761457,
+      "grad_norm": 0.6078757643699646,
+      "learning_rate": 0.00011907504847247081,
+      "loss": 5.9177,
+      "step": 2062
+    },
+    {
+      "epoch": 0.7088433621096938,
+      "grad_norm": 0.5807406306266785,
+      "learning_rate": 0.00011881967004474257,
+      "loss": 5.9959,
+      "step": 2063
+    },
+    {
+      "epoch": 0.7091869604432418,
+      "grad_norm": 0.7721190452575684,
+      "learning_rate": 0.00011856448037155828,
+      "loss": 6.0585,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7095305587767899,
+      "grad_norm": 0.6295862793922424,
+      "learning_rate": 0.00011830947982010889,
+      "loss": 6.0483,
+      "step": 2065
+    },
+    {
+      "epoch": 0.709874157110338,
+      "grad_norm": 0.5060104131698608,
+      "learning_rate": 0.00011805466875731277,
+      "loss": 6.0596,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7102177554438861,
+      "grad_norm": 0.5339775681495667,
+      "learning_rate": 0.0001178000475498163,
+      "loss": 6.1276,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7105613537774342,
+      "grad_norm": 0.6627702713012695,
+      "learning_rate": 0.00011754561656399204,
+      "loss": 6.0535,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7109049521109823,
+      "grad_norm": 0.49420684576034546,
+      "learning_rate": 0.00011729137616593922,
+      "loss": 6.1306,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7112485504445304,
+      "grad_norm": 0.6272643804550171,
+      "learning_rate": 0.00011703732672148274,
+      "loss": 6.0656,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7115921487780784,
+      "grad_norm": 0.4226469099521637,
+      "learning_rate": 0.00011678346859617283,
+      "loss": 6.0244,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7119357471116265,
+      "grad_norm": 0.5564938187599182,
+      "learning_rate": 0.00011652980215528414,
+      "loss": 6.0609,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7122793454451746,
+      "grad_norm": 0.47332972288131714,
+      "learning_rate": 0.00011627632776381577,
+      "loss": 6.1597,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7126229437787227,
+      "grad_norm": 0.5332189798355103,
+      "learning_rate": 0.00011602304578649056,
+      "loss": 6.0788,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7129665421122707,
+      "grad_norm": 0.5555863976478577,
+      "learning_rate": 0.00011576995658775405,
+      "loss": 6.0669,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7133101404458189,
+      "grad_norm": 0.5416020154953003,
+      "learning_rate": 0.000115517060531775,
+      "loss": 6.1068,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7136537387793669,
+      "grad_norm": 0.5631720423698425,
+      "learning_rate": 0.0001152643579824437,
+      "loss": 6.058,
+      "step": 2077
+    },
+    {
+      "epoch": 0.713997337112915,
+      "grad_norm": 0.5092825293540955,
+      "learning_rate": 0.00011501184930337235,
+      "loss": 6.0472,
+      "step": 2078
+    },
+    {
+      "epoch": 0.714340935446463,
+      "grad_norm": 0.5118112564086914,
+      "learning_rate": 0.00011475953485789406,
+      "loss": 6.188,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7146845337800112,
+      "grad_norm": 0.5366310477256775,
+      "learning_rate": 0.00011450741500906248,
+      "loss": 6.1434,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7150281321135592,
+      "grad_norm": 0.5202789902687073,
+      "learning_rate": 0.00011425549011965128,
+      "loss": 5.9162,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7153717304471073,
+      "grad_norm": 0.5178118944168091,
+      "learning_rate": 0.00011400376055215367,
+      "loss": 6.096,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7157153287806555,
+      "grad_norm": 0.6584262251853943,
+      "learning_rate": 0.00011375222666878143,
+      "loss": 6.0834,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7160589271142035,
+      "grad_norm": 0.6938084363937378,
+      "learning_rate": 0.00011350088883146548,
+      "loss": 6.0291,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7164025254477516,
+      "grad_norm": 0.5790725350379944,
+      "learning_rate": 0.00011324974740185392,
+      "loss": 6.0404,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7167461237812996,
+      "grad_norm": 0.571308970451355,
+      "learning_rate": 0.00011299880274131269,
+      "loss": 6.0675,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7170897221148478,
+      "grad_norm": 0.6977917551994324,
+      "learning_rate": 0.00011274805521092452,
+      "loss": 6.1419,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7174333204483958,
+      "grad_norm": 0.5644780397415161,
+      "learning_rate": 0.00011249750517148826,
+      "loss": 6.1105,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7177769187819439,
+      "grad_norm": 0.6917959451675415,
+      "learning_rate": 0.00011224715298351889,
+      "loss": 6.2568,
+      "step": 2089
+    },
+    {
+      "epoch": 0.718120517115492,
+      "grad_norm": 0.7160729765892029,
+      "learning_rate": 0.00011199699900724659,
+      "loss": 6.0182,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7184641154490401,
+      "grad_norm": 0.5881943106651306,
+      "learning_rate": 0.00011174704360261636,
+      "loss": 6.1496,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7188077137825881,
+      "grad_norm": 0.6279813647270203,
+      "learning_rate": 0.00011149728712928724,
+      "loss": 6.165,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7191513121161363,
+      "grad_norm": 0.6775481700897217,
+      "learning_rate": 0.00011124772994663257,
+      "loss": 6.165,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7194949104496843,
+      "grad_norm": 0.7559806704521179,
+      "learning_rate": 0.00011099837241373831,
+      "loss": 6.2015,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7198385087832324,
+      "grad_norm": 0.7788186073303223,
+      "learning_rate": 0.00011074921488940353,
+      "loss": 6.111,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7201821071167804,
+      "grad_norm": 0.6424675583839417,
+      "learning_rate": 0.00011050025773213943,
+      "loss": 6.179,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7205257054503286,
+      "grad_norm": 0.7416768670082092,
+      "learning_rate": 0.00011025150130016895,
+      "loss": 6.1499,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7208693037838767,
+      "grad_norm": 0.7270972728729248,
+      "learning_rate": 0.00011000294595142591,
+      "loss": 6.1627,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7212129021174247,
+      "grad_norm": 0.8351125121116638,
+      "learning_rate": 0.00010975459204355531,
+      "loss": 6.1789,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7215565004509729,
+      "grad_norm": 1.0598890781402588,
+      "learning_rate": 0.0001095064399339118,
+      "loss": 6.4582,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7219000987845209,
+      "grad_norm": 1.0137444734573364,
+      "learning_rate": 0.00010925848997955995,
+      "loss": 6.0051,
+      "step": 2101
+    },
+    {
+      "epoch": 0.722243697118069,
+      "grad_norm": 0.8496506810188293,
+      "learning_rate": 0.00010901074253727336,
+      "loss": 6.0438,
+      "step": 2102
+    },
+    {
+      "epoch": 0.722587295451617,
+      "grad_norm": 0.6422009468078613,
+      "learning_rate": 0.00010876319796353437,
+      "loss": 6.0723,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7229308937851652,
+      "grad_norm": 0.6015517711639404,
+      "learning_rate": 0.00010851585661453309,
+      "loss": 5.9828,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7232744921187132,
+      "grad_norm": 0.5879690051078796,
+      "learning_rate": 0.00010826871884616751,
+      "loss": 5.9908,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7236180904522613,
+      "grad_norm": 0.5813199281692505,
+      "learning_rate": 0.00010802178501404272,
+      "loss": 6.1323,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7239616887858094,
+      "grad_norm": 0.7803506851196289,
+      "learning_rate": 0.00010777505547346994,
+      "loss": 5.8687,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7243052871193575,
+      "grad_norm": 0.6409358382225037,
+      "learning_rate": 0.0001075285305794671,
+      "loss": 6.0781,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7246488854529055,
+      "grad_norm": 0.5260980129241943,
+      "learning_rate": 0.00010728221068675695,
+      "loss": 5.8364,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7249924837864536,
+      "grad_norm": 0.5710437893867493,
+      "learning_rate": 0.00010703609614976798,
+      "loss": 5.9737,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7253360821200017,
+      "grad_norm": 0.5062072277069092,
+      "learning_rate": 0.00010679018732263257,
+      "loss": 6.0028,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7256796804535498,
+      "grad_norm": 0.6027816534042358,
+      "learning_rate": 0.00010654448455918747,
+      "loss": 6.0024,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7260232787870979,
+      "grad_norm": 0.584230363368988,
+      "learning_rate": 0.00010629898821297279,
+      "loss": 6.0539,
+      "step": 2113
+    },
+    {
+      "epoch": 0.726366877120646,
+      "grad_norm": 0.5171947479248047,
+      "learning_rate": 0.0001060536986372318,
+      "loss": 6.0555,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7267104754541941,
+      "grad_norm": 0.5520440936088562,
+      "learning_rate": 0.0001058086161849098,
+      "loss": 6.0871,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7270540737877421,
+      "grad_norm": 0.5782375335693359,
+      "learning_rate": 0.00010556374120865477,
+      "loss": 5.9704,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7273976721212903,
+      "grad_norm": 0.4811157286167145,
+      "learning_rate": 0.00010531907406081548,
+      "loss": 6.08,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7277412704548383,
+      "grad_norm": 0.5225687623023987,
+      "learning_rate": 0.00010507461509344199,
+      "loss": 6.0993,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7280848687883864,
+      "grad_norm": 0.5056168437004089,
+      "learning_rate": 0.00010483036465828492,
+      "loss": 6.0458,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7284284671219344,
+      "grad_norm": 0.5187274217605591,
+      "learning_rate": 0.00010458632310679439,
+      "loss": 5.953,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7287720654554826,
+      "grad_norm": 0.5479382276535034,
+      "learning_rate": 0.00010434249079012043,
+      "loss": 6.0058,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7291156637890306,
+      "grad_norm": 0.4890539050102234,
+      "learning_rate": 0.00010409886805911175,
+      "loss": 6.0959,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7294592621225787,
+      "grad_norm": 0.6457909941673279,
+      "learning_rate": 0.00010385545526431567,
+      "loss": 6.0905,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7298028604561267,
+      "grad_norm": 0.66217440366745,
+      "learning_rate": 0.00010361225275597702,
+      "loss": 6.1174,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7301464587896749,
+      "grad_norm": 0.6101366281509399,
+      "learning_rate": 0.00010336926088403873,
+      "loss": 6.0577,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7304900571232229,
+      "grad_norm": 0.49836745858192444,
+      "learning_rate": 0.00010312647999813998,
+      "loss": 6.0062,
+      "step": 2126
+    },
+    {
+      "epoch": 0.730833655456771,
+      "grad_norm": 0.5868933200836182,
+      "learning_rate": 0.00010288391044761675,
+      "loss": 5.9208,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7311772537903192,
+      "grad_norm": 0.479012131690979,
+      "learning_rate": 0.00010264155258150079,
+      "loss": 6.0414,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7315208521238672,
+      "grad_norm": 0.5518100261688232,
+      "learning_rate": 0.00010239940674851941,
+      "loss": 6.116,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7318644504574153,
+      "grad_norm": 0.5689438581466675,
+      "learning_rate": 0.00010215747329709446,
+      "loss": 6.0154,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7322080487909634,
+      "grad_norm": 0.5302237868309021,
+      "learning_rate": 0.00010191575257534277,
+      "loss": 6.0576,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7325516471245115,
+      "grad_norm": 0.6240813136100769,
+      "learning_rate": 0.00010167424493107449,
+      "loss": 6.0377,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7328952454580595,
+      "grad_norm": 0.5226022601127625,
+      "learning_rate": 0.00010143295071179357,
+      "loss": 5.975,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7332388437916076,
+      "grad_norm": 0.501981258392334,
+      "learning_rate": 0.00010119187026469668,
+      "loss": 5.9928,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7335824421251557,
+      "grad_norm": 0.47182825207710266,
+      "learning_rate": 0.00010095100393667294,
+      "loss": 6.0488,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7339260404587038,
+      "grad_norm": 0.6065984964370728,
+      "learning_rate": 0.00010071035207430351,
+      "loss": 5.9399,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7342696387922518,
+      "grad_norm": 0.5377668142318726,
+      "learning_rate": 0.00010046991502386063,
+      "loss": 6.1058,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7346132371258,
+      "grad_norm": 0.617766261100769,
+      "learning_rate": 0.00010022969313130773,
+      "loss": 6.1166,
+      "step": 2138
+    },
+    {
+      "epoch": 0.734956835459348,
+      "grad_norm": 0.6241163015365601,
+      "learning_rate": 9.998968674229855e-05,
+      "loss": 6.0711,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7353004337928961,
+      "grad_norm": 0.5719399452209473,
+      "learning_rate": 9.974989620217689e-05,
+      "loss": 6.1984,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7356440321264441,
+      "grad_norm": 0.6313961148262024,
+      "learning_rate": 9.951032185597553e-05,
+      "loss": 6.0932,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7359876304599923,
+      "grad_norm": 0.7259103655815125,
+      "learning_rate": 9.927096404841688e-05,
+      "loss": 6.127,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7363312287935404,
+      "grad_norm": 0.6876815557479858,
+      "learning_rate": 9.903182312391104e-05,
+      "loss": 6.0632,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7366748271270884,
+      "grad_norm": 0.678907036781311,
+      "learning_rate": 9.87928994265565e-05,
+      "loss": 6.1833,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7370184254606366,
+      "grad_norm": 0.7564148902893066,
+      "learning_rate": 9.855419330013904e-05,
+      "loss": 6.1087,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7373620237941846,
+      "grad_norm": 0.7315529584884644,
+      "learning_rate": 9.831570508813149e-05,
+      "loss": 6.1149,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7377056221277327,
+      "grad_norm": 0.8211150765419006,
+      "learning_rate": 9.807743513369271e-05,
+      "loss": 6.27,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7380492204612807,
+      "grad_norm": 0.8686621189117432,
+      "learning_rate": 9.783938377966825e-05,
+      "loss": 6.142,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7383928187948289,
+      "grad_norm": 0.8419855237007141,
+      "learning_rate": 9.760155136858839e-05,
+      "loss": 6.1696,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7387364171283769,
+      "grad_norm": 1.0530507564544678,
+      "learning_rate": 9.736393824266876e-05,
+      "loss": 5.9706,
+      "step": 2150
+    },
+    {
+      "epoch": 0.739080015461925,
+      "grad_norm": 0.8615719079971313,
+      "learning_rate": 9.712654474380947e-05,
+      "loss": 5.9201,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7394236137954731,
+      "grad_norm": 0.7670858502388,
+      "learning_rate": 9.688937121359434e-05,
+      "loss": 5.991,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7397672121290212,
+      "grad_norm": 0.7319602966308594,
+      "learning_rate": 9.665241799329098e-05,
+      "loss": 6.0728,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7401108104625692,
+      "grad_norm": 0.6953431963920593,
+      "learning_rate": 9.641568542384982e-05,
+      "loss": 6.0657,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7404544087961173,
+      "grad_norm": 0.6169705390930176,
+      "learning_rate": 9.617917384590397e-05,
+      "loss": 6.0031,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7407980071296654,
+      "grad_norm": 0.6087819933891296,
+      "learning_rate": 9.594288359976817e-05,
+      "loss": 5.9738,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7411416054632135,
+      "grad_norm": 0.7109917998313904,
+      "learning_rate": 9.570681502543929e-05,
+      "loss": 6.0986,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7414852037967616,
+      "grad_norm": 0.725594162940979,
+      "learning_rate": 9.547096846259467e-05,
+      "loss": 6.0714,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7418288021303097,
+      "grad_norm": 0.6579272150993347,
+      "learning_rate": 9.523534425059252e-05,
+      "loss": 6.0939,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7421724004638578,
+      "grad_norm": 0.61977618932724,
+      "learning_rate": 9.499994272847099e-05,
+      "loss": 5.9499,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7425159987974058,
+      "grad_norm": 0.5611152052879333,
+      "learning_rate": 9.476476423494792e-05,
+      "loss": 6.0905,
+      "step": 2161
+    },
+    {
+      "epoch": 0.742859597130954,
+      "grad_norm": 0.5907820463180542,
+      "learning_rate": 9.452980910841993e-05,
+      "loss": 5.9445,
+      "step": 2162
+    },
+    {
+      "epoch": 0.743203195464502,
+      "grad_norm": 0.45404481887817383,
+      "learning_rate": 9.42950776869628e-05,
+      "loss": 6.0217,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7435467937980501,
+      "grad_norm": 0.5082436800003052,
+      "learning_rate": 9.40605703083298e-05,
+      "loss": 5.9823,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7438903921315981,
+      "grad_norm": 0.5696332454681396,
+      "learning_rate": 9.382628730995222e-05,
+      "loss": 6.1404,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7442339904651463,
+      "grad_norm": 0.5783550143241882,
+      "learning_rate": 9.359222902893832e-05,
+      "loss": 6.0203,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7445775887986943,
+      "grad_norm": 0.516828179359436,
+      "learning_rate": 9.335839580207317e-05,
+      "loss": 6.099,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7449211871322424,
+      "grad_norm": 0.5123862624168396,
+      "learning_rate": 9.312478796581792e-05,
+      "loss": 6.0458,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7452647854657904,
+      "grad_norm": 0.5932062864303589,
+      "learning_rate": 9.289140585630926e-05,
+      "loss": 6.0813,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7456083837993386,
+      "grad_norm": 0.4866209626197815,
+      "learning_rate": 9.265824980935933e-05,
+      "loss": 6.1021,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7459519821328866,
+      "grad_norm": 0.5570813417434692,
+      "learning_rate": 9.242532016045485e-05,
+      "loss": 6.1042,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7462955804664347,
+      "grad_norm": 0.5078079104423523,
+      "learning_rate": 9.219261724475692e-05,
+      "loss": 6.0633,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7466391787999829,
+      "grad_norm": 0.419669508934021,
+      "learning_rate": 9.196014139710005e-05,
+      "loss": 6.0652,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7469827771335309,
+      "grad_norm": 0.511519193649292,
+      "learning_rate": 9.172789295199255e-05,
+      "loss": 6.1272,
+      "step": 2174
+    },
+    {
+      "epoch": 0.747326375467079,
+      "grad_norm": 0.4874584972858429,
+      "learning_rate": 9.149587224361503e-05,
+      "loss": 6.1074,
+      "step": 2175
+    },
+    {
+      "epoch": 0.747669973800627,
+      "grad_norm": 0.4667309820652008,
+      "learning_rate": 9.126407960582067e-05,
+      "loss": 6.0939,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7480135721341752,
+      "grad_norm": 0.498703271150589,
+      "learning_rate": 9.103251537213445e-05,
+      "loss": 5.9496,
+      "step": 2177
+    },
+    {
+      "epoch": 0.7483571704677232,
+      "grad_norm": 0.536368727684021,
+      "learning_rate": 9.080117987575271e-05,
+      "loss": 5.9855,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7487007688012713,
+      "grad_norm": 0.5683017373085022,
+      "learning_rate": 9.057007344954244e-05,
+      "loss": 6.095,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7490443671348194,
+      "grad_norm": 0.6125717163085938,
+      "learning_rate": 9.033919642604149e-05,
+      "loss": 6.13,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7493879654683675,
+      "grad_norm": 0.5337144732475281,
+      "learning_rate": 9.010854913745712e-05,
+      "loss": 6.0017,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7497315638019155,
+      "grad_norm": 0.4696179926395416,
+      "learning_rate": 8.987813191566632e-05,
+      "loss": 6.0913,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7500751621354637,
+      "grad_norm": 0.6470948457717896,
+      "learning_rate": 8.964794509221508e-05,
+      "loss": 6.0583,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7504187604690117,
+      "grad_norm": 0.5472028851509094,
+      "learning_rate": 8.941798899831757e-05,
+      "loss": 6.0463,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7507623588025598,
+      "grad_norm": 0.5236582159996033,
+      "learning_rate": 8.918826396485624e-05,
+      "loss": 6.0631,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7511059571361078,
+      "grad_norm": 0.5017586946487427,
+      "learning_rate": 8.895877032238095e-05,
+      "loss": 6.1308,
+      "step": 2186
+    },
+    {
+      "epoch": 0.751449555469656,
+      "grad_norm": 0.736363410949707,
+      "learning_rate": 8.872950840110879e-05,
+      "loss": 6.0674,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7517931538032041,
+      "grad_norm": 0.5078322291374207,
+      "learning_rate": 8.8500478530923e-05,
+      "loss": 6.0312,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7521367521367521,
+      "grad_norm": 0.564947783946991,
+      "learning_rate": 8.827168104137353e-05,
+      "loss": 6.1036,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7524803504703003,
+      "grad_norm": 0.6403129696846008,
+      "learning_rate": 8.804311626167533e-05,
+      "loss": 6.0244,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7528239488038483,
+      "grad_norm": 0.675612211227417,
+      "learning_rate": 8.781478452070912e-05,
+      "loss": 6.142,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7531675471373964,
+      "grad_norm": 0.6955165266990662,
+      "learning_rate": 8.758668614701973e-05,
+      "loss": 5.9662,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7535111454709444,
+      "grad_norm": 0.58688884973526,
+      "learning_rate": 8.735882146881661e-05,
+      "loss": 5.996,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7538547438044926,
+      "grad_norm": 0.7645385265350342,
+      "learning_rate": 8.713119081397273e-05,
+      "loss": 6.0944,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7541983421380406,
+      "grad_norm": 0.8334187269210815,
+      "learning_rate": 8.690379451002448e-05,
+      "loss": 6.0435,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7545419404715887,
+      "grad_norm": 0.7340932488441467,
+      "learning_rate": 8.667663288417082e-05,
+      "loss": 6.26,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7548855388051368,
+      "grad_norm": 0.6961662173271179,
+      "learning_rate": 8.644970626327329e-05,
+      "loss": 6.204,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7552291371386849,
+      "grad_norm": 0.87010657787323,
+      "learning_rate": 8.622301497385507e-05,
+      "loss": 6.2819,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7555727354722329,
+      "grad_norm": 0.9306342601776123,
+      "learning_rate": 8.599655934210088e-05,
+      "loss": 6.1762,
+      "step": 2199
+    },
+    {
+      "epoch": 0.755916333805781,
+      "grad_norm": 1.0517386198043823,
+      "learning_rate": 8.577033969385639e-05,
+      "loss": 6.3565,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7562599321393291,
+      "grad_norm": 0.8607641458511353,
+      "learning_rate": 8.55443563546274e-05,
+      "loss": 5.9011,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7566035304728772,
+      "grad_norm": 0.789115846157074,
+      "learning_rate": 8.531860964958002e-05,
+      "loss": 5.9841,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7569471288064253,
+      "grad_norm": 0.7245513796806335,
+      "learning_rate": 8.509309990353973e-05,
+      "loss": 6.0235,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7572907271399734,
+      "grad_norm": 0.5847095251083374,
+      "learning_rate": 8.486782744099117e-05,
+      "loss": 5.931,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7576343254735215,
+      "grad_norm": 0.5829930901527405,
+      "learning_rate": 8.464279258607718e-05,
+      "loss": 5.9971,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7579779238070695,
+      "grad_norm": 0.5951849818229675,
+      "learning_rate": 8.441799566259937e-05,
+      "loss": 6.0391,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7583215221406177,
+      "grad_norm": 0.6464923024177551,
+      "learning_rate": 8.41934369940163e-05,
+      "loss": 5.9902,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7586651204741657,
+      "grad_norm": 0.6821208000183105,
+      "learning_rate": 8.396911690344411e-05,
+      "loss": 5.9381,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7590087188077138,
+      "grad_norm": 0.651321530342102,
+      "learning_rate": 8.37450357136556e-05,
+      "loss": 5.9518,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7593523171412618,
+      "grad_norm": 0.5200663805007935,
+      "learning_rate": 8.352119374707978e-05,
+      "loss": 5.8669,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75969591547481,
+      "grad_norm": 0.5066912174224854,
+      "learning_rate": 8.329759132580126e-05,
+      "loss": 5.8757,
+      "step": 2211
+    },
+    {
+      "epoch": 0.760039513808358,
+      "grad_norm": 0.5652316212654114,
+      "learning_rate": 8.30742287715604e-05,
+      "loss": 5.8752,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7603831121419061,
+      "grad_norm": 0.6964173316955566,
+      "learning_rate": 8.285110640575199e-05,
+      "loss": 5.9425,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7607267104754541,
+      "grad_norm": 0.5349513292312622,
+      "learning_rate": 8.262822454942542e-05,
+      "loss": 6.1252,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7610703088090023,
+      "grad_norm": 0.5608907341957092,
+      "learning_rate": 8.240558352328406e-05,
+      "loss": 6.0566,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7614139071425503,
+      "grad_norm": 0.6297080516815186,
+      "learning_rate": 8.218318364768451e-05,
+      "loss": 6.0614,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7617575054760984,
+      "grad_norm": 0.5819994807243347,
+      "learning_rate": 8.196102524263666e-05,
+      "loss": 5.9989,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7621011038096466,
+      "grad_norm": 0.463117778301239,
+      "learning_rate": 8.173910862780275e-05,
+      "loss": 6.1422,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7624447021431946,
+      "grad_norm": 0.5110581517219543,
+      "learning_rate": 8.15174341224973e-05,
+      "loss": 5.9919,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7627883004767427,
+      "grad_norm": 0.5026494264602661,
+      "learning_rate": 8.129600204568624e-05,
+      "loss": 6.0571,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7631318988102908,
+      "grad_norm": 0.5578264594078064,
+      "learning_rate": 8.10748127159869e-05,
+      "loss": 6.0543,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7634754971438389,
+      "grad_norm": 0.6005034446716309,
+      "learning_rate": 8.085386645166699e-05,
+      "loss": 6.054,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7638190954773869,
+      "grad_norm": 0.5268194675445557,
+      "learning_rate": 8.063316357064496e-05,
+      "loss": 6.0657,
+      "step": 2223
+    },
+    {
+      "epoch": 0.764162693810935,
+      "grad_norm": 0.43706896901130676,
+      "learning_rate": 8.041270439048857e-05,
+      "loss": 5.9922,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7645062921444831,
+      "grad_norm": 0.4229617416858673,
+      "learning_rate": 8.019248922841518e-05,
+      "loss": 5.9731,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7648498904780312,
+      "grad_norm": 0.44867685437202454,
+      "learning_rate": 7.997251840129105e-05,
+      "loss": 6.084,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7651934888115792,
+      "grad_norm": 0.6207936406135559,
+      "learning_rate": 7.975279222563086e-05,
+      "loss": 6.0844,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7655370871451274,
+      "grad_norm": 0.5367064476013184,
+      "learning_rate": 7.953331101759707e-05,
+      "loss": 6.0374,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7658806854786754,
+      "grad_norm": 0.5211548805236816,
+      "learning_rate": 7.931407509299982e-05,
+      "loss": 6.1057,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7662242838122235,
+      "grad_norm": 0.4982808530330658,
+      "learning_rate": 7.909508476729632e-05,
+      "loss": 5.9777,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7665678821457715,
+      "grad_norm": 0.5660802721977234,
+      "learning_rate": 7.887634035559036e-05,
+      "loss": 6.0857,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7669114804793197,
+      "grad_norm": 0.6563755869865417,
+      "learning_rate": 7.865784217263197e-05,
+      "loss": 6.051,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7672550788128678,
+      "grad_norm": 0.5743918418884277,
+      "learning_rate": 7.843959053281663e-05,
+      "loss": 6.0887,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7675986771464158,
+      "grad_norm": 0.4557611644268036,
+      "learning_rate": 7.822158575018534e-05,
+      "loss": 6.0547,
+      "step": 2234
+    },
+    {
+      "epoch": 0.767942275479964,
+      "grad_norm": 0.6070594787597656,
+      "learning_rate": 7.800382813842377e-05,
+      "loss": 6.0169,
+      "step": 2235
+    },
+    {
+      "epoch": 0.768285873813512,
+      "grad_norm": 0.5235620141029358,
+      "learning_rate": 7.778631801086209e-05,
+      "loss": 6.086,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7686294721470601,
+      "grad_norm": 0.5600701570510864,
+      "learning_rate": 7.756905568047393e-05,
+      "loss": 5.9655,
+      "step": 2237
+    },
+    {
+      "epoch": 0.7689730704806081,
+      "grad_norm": 0.6043455004692078,
+      "learning_rate": 7.735204145987704e-05,
+      "loss": 6.0636,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7693166688141563,
+      "grad_norm": 0.618334949016571,
+      "learning_rate": 7.713527566133158e-05,
+      "loss": 6.1404,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7696602671477043,
+      "grad_norm": 0.5381041169166565,
+      "learning_rate": 7.691875859674053e-05,
+      "loss": 6.1488,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7700038654812524,
+      "grad_norm": 0.6507089734077454,
+      "learning_rate": 7.670249057764894e-05,
+      "loss": 6.1051,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7703474638148005,
+      "grad_norm": 0.5733596086502075,
+      "learning_rate": 7.648647191524355e-05,
+      "loss": 6.1027,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7706910621483486,
+      "grad_norm": 0.6554251313209534,
+      "learning_rate": 7.627070292035201e-05,
+      "loss": 6.012,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7710346604818966,
+      "grad_norm": 0.6997339129447937,
+      "learning_rate": 7.605518390344333e-05,
+      "loss": 6.0476,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7713782588154448,
+      "grad_norm": 0.5664951205253601,
+      "learning_rate": 7.58399151746261e-05,
+      "loss": 6.0055,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7717218571489928,
+      "grad_norm": 0.8235125541687012,
+      "learning_rate": 7.56248970436493e-05,
+      "loss": 6.1593,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7720654554825409,
+      "grad_norm": 0.7746663689613342,
+      "learning_rate": 7.541012981990122e-05,
+      "loss": 6.0848,
+      "step": 2247
+    },
+    {
+      "epoch": 0.772409053816089,
+      "grad_norm": 0.7530941367149353,
+      "learning_rate": 7.519561381240878e-05,
+      "loss": 6.2576,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7727526521496371,
+      "grad_norm": 0.9261403679847717,
+      "learning_rate": 7.498134932983805e-05,
+      "loss": 6.2184,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7730962504831852,
+      "grad_norm": 1.2425649166107178,
+      "learning_rate": 7.476733668049259e-05,
+      "loss": 6.1733,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7734398488167332,
+      "grad_norm": 1.1316070556640625,
+      "learning_rate": 7.455357617231392e-05,
+      "loss": 5.9885,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7737834471502814,
+      "grad_norm": 0.9721357822418213,
+      "learning_rate": 7.434006811288069e-05,
+      "loss": 5.9354,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7741270454838294,
+      "grad_norm": 0.9695913791656494,
+      "learning_rate": 7.412681280940834e-05,
+      "loss": 5.8127,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7744706438173775,
+      "grad_norm": 0.6883681416511536,
+      "learning_rate": 7.391381056874835e-05,
+      "loss": 6.0319,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7748142421509255,
+      "grad_norm": 0.6744367480278015,
+      "learning_rate": 7.37010616973886e-05,
+      "loss": 6.0038,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7751578404844737,
+      "grad_norm": 0.5574299097061157,
+      "learning_rate": 7.348856650145188e-05,
+      "loss": 5.9949,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7755014388180217,
+      "grad_norm": 0.7782508730888367,
+      "learning_rate": 7.327632528669625e-05,
+      "loss": 5.9506,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7758450371515698,
+      "grad_norm": 0.6571906805038452,
+      "learning_rate": 7.306433835851423e-05,
+      "loss": 5.9716,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7761886354851179,
+      "grad_norm": 0.6952961683273315,
+      "learning_rate": 7.285260602193256e-05,
+      "loss": 5.9193,
+      "step": 2259
+    },
+    {
+      "epoch": 0.776532233818666,
+      "grad_norm": 0.6619623303413391,
+      "learning_rate": 7.264112858161137e-05,
+      "loss": 6.0133,
+      "step": 2260
+    },
+    {
+      "epoch": 0.776875832152214,
+      "grad_norm": 0.6246268153190613,
+      "learning_rate": 7.242990634184432e-05,
+      "loss": 5.9537,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7772194304857621,
+      "grad_norm": 0.5367864370346069,
+      "learning_rate": 7.221893960655773e-05,
+      "loss": 5.9849,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7775630288193103,
+      "grad_norm": 0.5722734928131104,
+      "learning_rate": 7.200822867931032e-05,
+      "loss": 5.9398,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7779066271528583,
+      "grad_norm": 0.6350281834602356,
+      "learning_rate": 7.179777386329276e-05,
+      "loss": 6.0259,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7782502254864064,
+      "grad_norm": 0.6227221488952637,
+      "learning_rate": 7.158757546132696e-05,
+      "loss": 6.0207,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7785938238199545,
+      "grad_norm": 0.6056457757949829,
+      "learning_rate": 7.13776337758662e-05,
+      "loss": 6.1297,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7789374221535026,
+      "grad_norm": 0.571882963180542,
+      "learning_rate": 7.116794910899424e-05,
+      "loss": 6.1116,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7792810204870506,
+      "grad_norm": 0.4469764530658722,
+      "learning_rate": 7.095852176242503e-05,
+      "loss": 6.0374,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7796246188205987,
+      "grad_norm": 0.6148665547370911,
+      "learning_rate": 7.07493520375021e-05,
+      "loss": 5.9959,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7799682171541468,
+      "grad_norm": 0.6256452798843384,
+      "learning_rate": 7.05404402351987e-05,
+      "loss": 5.9674,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7803118154876949,
+      "grad_norm": 0.6558529138565063,
+      "learning_rate": 7.033178665611639e-05,
+      "loss": 6.028,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7806554138212429,
+      "grad_norm": 0.6320913434028625,
+      "learning_rate": 7.012339160048578e-05,
+      "loss": 5.9793,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7809990121547911,
+      "grad_norm": 0.5069714188575745,
+      "learning_rate": 6.991525536816498e-05,
+      "loss": 6.0383,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7813426104883391,
+      "grad_norm": 0.6668805480003357,
+      "learning_rate": 6.970737825863999e-05,
+      "loss": 6.1826,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7816862088218872,
+      "grad_norm": 0.5856713652610779,
+      "learning_rate": 6.949976057102384e-05,
+      "loss": 5.9855,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7820298071554352,
+      "grad_norm": 0.5079368948936462,
+      "learning_rate": 6.929240260405634e-05,
+      "loss": 6.1041,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7823734054889834,
+      "grad_norm": 0.5541518330574036,
+      "learning_rate": 6.908530465610347e-05,
+      "loss": 6.0746,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7827170038225315,
+      "grad_norm": 0.49500662088394165,
+      "learning_rate": 6.887846702515718e-05,
+      "loss": 6.0113,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7830606021560795,
+      "grad_norm": 0.7125208973884583,
+      "learning_rate": 6.867189000883495e-05,
+      "loss": 5.8986,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7834042004896277,
+      "grad_norm": 0.49747544527053833,
+      "learning_rate": 6.846557390437883e-05,
+      "loss": 6.203,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7837477988231757,
+      "grad_norm": 0.4951108694076538,
+      "learning_rate": 6.825951900865612e-05,
+      "loss": 6.0585,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7840913971567238,
+      "grad_norm": 0.6161938905715942,
+      "learning_rate": 6.805372561815768e-05,
+      "loss": 5.951,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7844349954902718,
+      "grad_norm": 0.5577975511550903,
+      "learning_rate": 6.784819402899833e-05,
+      "loss": 6.0114,
+      "step": 2283
+    },
+    {
+      "epoch": 0.78477859382382,
+      "grad_norm": 0.5856155157089233,
+      "learning_rate": 6.764292453691622e-05,
+      "loss": 6.0278,
+      "step": 2284
+    },
+    {
+      "epoch": 0.785122192157368,
+      "grad_norm": 0.5486255288124084,
+      "learning_rate": 6.74379174372724e-05,
+      "loss": 6.1078,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7854657904909161,
+      "grad_norm": 0.6295291185379028,
+      "learning_rate": 6.723317302505e-05,
+      "loss": 6.1005,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7858093888244642,
+      "grad_norm": 0.5545480251312256,
+      "learning_rate": 6.702869159485481e-05,
+      "loss": 6.1214,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7861529871580123,
+      "grad_norm": 0.5096721649169922,
+      "learning_rate": 6.682447344091364e-05,
+      "loss": 6.0126,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7864965854915603,
+      "grad_norm": 0.6050777435302734,
+      "learning_rate": 6.66205188570747e-05,
+      "loss": 6.1272,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7868401838251085,
+      "grad_norm": 0.5570022463798523,
+      "learning_rate": 6.641682813680705e-05,
+      "loss": 6.156,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7871837821586565,
+      "grad_norm": 0.5773416757583618,
+      "learning_rate": 6.621340157319997e-05,
+      "loss": 6.0434,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7875273804922046,
+      "grad_norm": 0.6278853416442871,
+      "learning_rate": 6.60102394589625e-05,
+      "loss": 6.1401,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7878709788257527,
+      "grad_norm": 0.8446879386901855,
+      "learning_rate": 6.580734208642344e-05,
+      "loss": 6.2182,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7882145771593008,
+      "grad_norm": 0.6405084133148193,
+      "learning_rate": 6.560470974753053e-05,
+      "loss": 6.1942,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7885581754928489,
+      "grad_norm": 0.6068698763847351,
+      "learning_rate": 6.54023427338501e-05,
+      "loss": 6.1913,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7889017738263969,
+      "grad_norm": 0.6595099568367004,
+      "learning_rate": 6.520024133656687e-05,
+      "loss": 6.062,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7892453721599451,
+      "grad_norm": 0.7970278859138489,
+      "learning_rate": 6.499840584648315e-05,
+      "loss": 6.1699,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7895889704934931,
+      "grad_norm": 0.8106145262718201,
+      "learning_rate": 6.479683655401875e-05,
+      "loss": 6.2438,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7899325688270412,
+      "grad_norm": 1.0167194604873657,
+      "learning_rate": 6.459553374921045e-05,
+      "loss": 6.4,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7902761671605892,
+      "grad_norm": 1.1539579629898071,
+      "learning_rate": 6.439449772171163e-05,
+      "loss": 6.2757,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7906197654941374,
+      "grad_norm": 0.8855049014091492,
+      "learning_rate": 6.419372876079174e-05,
+      "loss": 6.0337,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7909633638276854,
+      "grad_norm": 1.005344271659851,
+      "learning_rate": 6.399322715533601e-05,
+      "loss": 5.8928,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7913069621612335,
+      "grad_norm": 0.9247390627861023,
+      "learning_rate": 6.379299319384471e-05,
+      "loss": 5.9571,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7916505604947816,
+      "grad_norm": 0.7353301644325256,
+      "learning_rate": 6.359302716443352e-05,
+      "loss": 6.0767,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7919941588283297,
+      "grad_norm": 0.5871204733848572,
+      "learning_rate": 6.339332935483206e-05,
+      "loss": 5.9553,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7923377571618778,
+      "grad_norm": 0.5191056728363037,
+      "learning_rate": 6.319390005238432e-05,
+      "loss": 6.0189,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7926813554954258,
+      "grad_norm": 0.5726730227470398,
+      "learning_rate": 6.299473954404788e-05,
+      "loss": 5.9178,
+      "step": 2307
+    },
+    {
+      "epoch": 0.793024953828974,
+      "grad_norm": 0.6066359877586365,
+      "learning_rate": 6.279584811639357e-05,
+      "loss": 5.9549,
+      "step": 2308
+    },
+    {
+      "epoch": 0.793368552162522,
+      "grad_norm": 0.5796366930007935,
+      "learning_rate": 6.259722605560488e-05,
+      "loss": 6.0092,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7937121504960701,
+      "grad_norm": 0.5898374915122986,
+      "learning_rate": 6.23988736474779e-05,
+      "loss": 5.9802,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7940557488296182,
+      "grad_norm": 0.7097289562225342,
+      "learning_rate": 6.220079117742064e-05,
+      "loss": 6.0136,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7943993471631663,
+      "grad_norm": 0.669428288936615,
+      "learning_rate": 6.20029789304527e-05,
+      "loss": 5.8873,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7947429454967143,
+      "grad_norm": 0.6120542287826538,
+      "learning_rate": 6.180543719120496e-05,
+      "loss": 6.0431,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7950865438302624,
+      "grad_norm": 0.5299326777458191,
+      "learning_rate": 6.160816624391886e-05,
+      "loss": 5.9385,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7954301421638105,
+      "grad_norm": 0.49728134274482727,
+      "learning_rate": 6.141116637244631e-05,
+      "loss": 6.1362,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7957737404973586,
+      "grad_norm": 0.6035153269767761,
+      "learning_rate": 6.121443786024921e-05,
+      "loss": 5.9053,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7961173388309066,
+      "grad_norm": 0.5510128736495972,
+      "learning_rate": 6.101798099039907e-05,
+      "loss": 6.0034,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7964609371644548,
+      "grad_norm": 0.49103474617004395,
+      "learning_rate": 6.082179604557617e-05,
+      "loss": 6.0575,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7968045354980028,
+      "grad_norm": 0.5344441533088684,
+      "learning_rate": 6.062588330807009e-05,
+      "loss": 6.1826,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7971481338315509,
+      "grad_norm": 0.5215442776679993,
+      "learning_rate": 6.043024305977823e-05,
+      "loss": 6.0573,
+      "step": 2320
+    },
+    {
+      "epoch": 0.797491732165099,
+      "grad_norm": 0.4892186224460602,
+      "learning_rate": 6.023487558220614e-05,
+      "loss": 6.0188,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7978353304986471,
+      "grad_norm": 0.549307644367218,
+      "learning_rate": 6.003978115646683e-05,
+      "loss": 6.0237,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7981789288321952,
+      "grad_norm": 0.5364143252372742,
+      "learning_rate": 5.984496006328055e-05,
+      "loss": 6.0695,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7985225271657432,
+      "grad_norm": 0.4920170307159424,
+      "learning_rate": 5.965041258297396e-05,
+      "loss": 5.9426,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7988661254992914,
+      "grad_norm": 0.6157858371734619,
+      "learning_rate": 5.94561389954803e-05,
+      "loss": 6.0046,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7992097238328394,
+      "grad_norm": 0.47127848863601685,
+      "learning_rate": 5.926213958033855e-05,
+      "loss": 6.125,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7995533221663875,
+      "grad_norm": 0.4696691632270813,
+      "learning_rate": 5.9068414616693264e-05,
+      "loss": 6.0563,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7998969204999355,
+      "grad_norm": 0.5054667592048645,
+      "learning_rate": 5.887496438329412e-05,
+      "loss": 6.0715,
+      "step": 2328
+    },
+    {
+      "epoch": 0.8002405188334837,
+      "grad_norm": 0.5439390540122986,
+      "learning_rate": 5.868178915849526e-05,
+      "loss": 5.9687,
+      "step": 2329
+    },
+    {
+      "epoch": 0.8005841171670317,
+      "grad_norm": 0.5408337712287903,
+      "learning_rate": 5.848888922025553e-05,
+      "loss": 5.9605,
+      "step": 2330
+    },
+    {
+      "epoch": 0.8009277155005798,
+      "grad_norm": 0.6336498856544495,
+      "learning_rate": 5.82962648461372e-05,
+      "loss": 5.9145,
+      "step": 2331
+    },
+    {
+      "epoch": 0.8012713138341279,
+      "grad_norm": 0.46218758821487427,
+      "learning_rate": 5.810391631330639e-05,
+      "loss": 6.0208,
+      "step": 2332
+    },
+    {
+      "epoch": 0.801614912167676,
+      "grad_norm": 0.495760053396225,
+      "learning_rate": 5.791184389853213e-05,
+      "loss": 6.0446,
+      "step": 2333
+    },
+    {
+      "epoch": 0.801958510501224,
+      "grad_norm": 0.6889287829399109,
+      "learning_rate": 5.77200478781863e-05,
+      "loss": 5.9667,
+      "step": 2334
+    },
+    {
+      "epoch": 0.8023021088347722,
+      "grad_norm": 0.5764407515525818,
+      "learning_rate": 5.752852852824275e-05,
+      "loss": 6.0924,
+      "step": 2335
+    },
+    {
+      "epoch": 0.8026457071683203,
+      "grad_norm": 0.5324522256851196,
+      "learning_rate": 5.733728612427772e-05,
+      "loss": 6.0446,
+      "step": 2336
+    },
+    {
+      "epoch": 0.8029893055018683,
+      "grad_norm": 0.5219273567199707,
+      "learning_rate": 5.7146320941468515e-05,
+      "loss": 6.0894,
+      "step": 2337
+    },
+    {
+      "epoch": 0.8033329038354164,
+      "grad_norm": 0.6107932329177856,
+      "learning_rate": 5.695563325459377e-05,
+      "loss": 6.0145,
+      "step": 2338
+    },
+    {
+      "epoch": 0.8036765021689645,
+      "grad_norm": 0.74781733751297,
+      "learning_rate": 5.6765223338032804e-05,
+      "loss": 5.9561,
+      "step": 2339
+    },
+    {
+      "epoch": 0.8040201005025126,
+      "grad_norm": 0.5779476165771484,
+      "learning_rate": 5.6575091465765313e-05,
+      "loss": 6.1911,
+      "step": 2340
+    },
+    {
+      "epoch": 0.8043636988360606,
+      "grad_norm": 0.5426329970359802,
+      "learning_rate": 5.6385237911370654e-05,
+      "loss": 6.1045,
+      "step": 2341
+    },
+    {
+      "epoch": 0.8047072971696088,
+      "grad_norm": 0.6034143567085266,
+      "learning_rate": 5.6195662948028024e-05,
+      "loss": 6.0011,
+      "step": 2342
+    },
+    {
+      "epoch": 0.8050508955031568,
+      "grad_norm": 0.6380503177642822,
+      "learning_rate": 5.600636684851562e-05,
+      "loss": 6.0992,
+      "step": 2343
+    },
+    {
+      "epoch": 0.8053944938367049,
+      "grad_norm": 0.6098598837852478,
+      "learning_rate": 5.5817349885210395e-05,
+      "loss": 6.1604,
+      "step": 2344
+    },
+    {
+      "epoch": 0.8057380921702529,
+      "grad_norm": 0.7238516211509705,
+      "learning_rate": 5.562861233008773e-05,
+      "loss": 6.1236,
+      "step": 2345
+    },
+    {
+      "epoch": 0.8060816905038011,
+      "grad_norm": 0.6985032558441162,
+      "learning_rate": 5.5440154454720726e-05,
+      "loss": 6.0547,
+      "step": 2346
+    },
+    {
+      "epoch": 0.8064252888373491,
+      "grad_norm": 0.8203953504562378,
+      "learning_rate": 5.525197653028033e-05,
+      "loss": 6.0589,
+      "step": 2347
+    },
+    {
+      "epoch": 0.8067688871708972,
+      "grad_norm": 0.9466792941093445,
+      "learning_rate": 5.506407882753456e-05,
+      "loss": 6.2681,
+      "step": 2348
+    },
+    {
+      "epoch": 0.8071124855044453,
+      "grad_norm": 0.886099100112915,
+      "learning_rate": 5.4876461616848256e-05,
+      "loss": 6.2795,
+      "step": 2349
+    },
+    {
+      "epoch": 0.8074560838379934,
+      "grad_norm": 1.1105183362960815,
+      "learning_rate": 5.4689125168182504e-05,
+      "loss": 5.997,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8077996821715415,
+      "grad_norm": 0.8322456479072571,
+      "learning_rate": 5.450206975109473e-05,
+      "loss": 6.0151,
+      "step": 2351
+    },
+    {
+      "epoch": 0.8081432805050895,
+      "grad_norm": 0.8582592010498047,
+      "learning_rate": 5.431529563473758e-05,
+      "loss": 6.0131,
+      "step": 2352
+    },
+    {
+      "epoch": 0.8084868788386377,
+      "grad_norm": 0.7677584290504456,
+      "learning_rate": 5.412880308785928e-05,
+      "loss": 5.9629,
+      "step": 2353
+    },
+    {
+      "epoch": 0.8088304771721857,
+      "grad_norm": 0.6369113922119141,
+      "learning_rate": 5.394259237880272e-05,
+      "loss": 6.0089,
+      "step": 2354
+    },
+    {
+      "epoch": 0.8091740755057338,
+      "grad_norm": 0.5916951298713684,
+      "learning_rate": 5.375666377550534e-05,
+      "loss": 6.0501,
+      "step": 2355
+    },
+    {
+      "epoch": 0.8095176738392819,
+      "grad_norm": 0.5984398722648621,
+      "learning_rate": 5.357101754549864e-05,
+      "loss": 5.9582,
+      "step": 2356
+    },
+    {
+      "epoch": 0.80986127217283,
+      "grad_norm": 0.6078851222991943,
+      "learning_rate": 5.338565395590772e-05,
+      "loss": 5.9114,
+      "step": 2357
+    },
+    {
+      "epoch": 0.810204870506378,
+      "grad_norm": 0.4389530420303345,
+      "learning_rate": 5.320057327345112e-05,
+      "loss": 5.9859,
+      "step": 2358
+    },
+    {
+      "epoch": 0.8105484688399262,
+      "grad_norm": 0.6114437580108643,
+      "learning_rate": 5.301577576444025e-05,
+      "loss": 6.1864,
+      "step": 2359
+    },
+    {
+      "epoch": 0.8108920671734742,
+      "grad_norm": 0.6425684690475464,
+      "learning_rate": 5.2831261694779144e-05,
+      "loss": 5.974,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8112356655070223,
+      "grad_norm": 0.5639946460723877,
+      "learning_rate": 5.264703132996376e-05,
+      "loss": 5.9636,
+      "step": 2361
+    },
+    {
+      "epoch": 0.8115792638405703,
+      "grad_norm": 0.5804059505462646,
+      "learning_rate": 5.24630849350822e-05,
+      "loss": 5.9089,
+      "step": 2362
+    },
+    {
+      "epoch": 0.8119228621741185,
+      "grad_norm": 0.5132018327713013,
+      "learning_rate": 5.2279422774813624e-05,
+      "loss": 5.8329,
+      "step": 2363
+    },
+    {
+      "epoch": 0.8122664605076665,
+      "grad_norm": 0.5105260014533997,
+      "learning_rate": 5.2096045113428385e-05,
+      "loss": 6.1746,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8126100588412146,
+      "grad_norm": 0.5083038806915283,
+      "learning_rate": 5.191295221478745e-05,
+      "loss": 6.044,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8129536571747628,
+      "grad_norm": 0.7395703792572021,
+      "learning_rate": 5.1730144342342076e-05,
+      "loss": 6.0428,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8132972555083108,
+      "grad_norm": 0.4767551124095917,
+      "learning_rate": 5.1547621759133165e-05,
+      "loss": 6.0285,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8136408538418589,
+      "grad_norm": 0.4748813807964325,
+      "learning_rate": 5.136538472779156e-05,
+      "loss": 6.1975,
+      "step": 2368
+    },
+    {
+      "epoch": 0.8139844521754069,
+      "grad_norm": 0.5720754861831665,
+      "learning_rate": 5.118343351053681e-05,
+      "loss": 6.0841,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8143280505089551,
+      "grad_norm": 0.49053552746772766,
+      "learning_rate": 5.100176836917736e-05,
+      "loss": 5.9318,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8146716488425031,
+      "grad_norm": 0.5451865792274475,
+      "learning_rate": 5.0820389565110095e-05,
+      "loss": 6.0171,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8150152471760512,
+      "grad_norm": 0.4565944969654083,
+      "learning_rate": 5.063929735931985e-05,
+      "loss": 6.1135,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8153588455095993,
+      "grad_norm": 0.5404878854751587,
+      "learning_rate": 5.045849201237893e-05,
+      "loss": 6.024,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8157024438431474,
+      "grad_norm": 0.46480605006217957,
+      "learning_rate": 5.027797378444707e-05,
+      "loss": 6.0778,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8160460421766954,
+      "grad_norm": 0.5615720748901367,
+      "learning_rate": 5.0097742935270776e-05,
+      "loss": 6.095,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8163896405102435,
+      "grad_norm": 0.5113908648490906,
+      "learning_rate": 4.991779972418315e-05,
+      "loss": 6.0162,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8167332388437916,
+      "grad_norm": 0.4773457646369934,
+      "learning_rate": 4.97381444101033e-05,
+      "loss": 6.0966,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8170768371773397,
+      "grad_norm": 0.4736880660057068,
+      "learning_rate": 4.9558777251536043e-05,
+      "loss": 5.9657,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8174204355108877,
+      "grad_norm": 0.49745383858680725,
+      "learning_rate": 4.9379698506571705e-05,
+      "loss": 6.0504,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8177640338444359,
+      "grad_norm": 0.49785807728767395,
+      "learning_rate": 4.920090843288558e-05,
+      "loss": 5.9924,
+      "step": 2380
+    },
+    {
+      "epoch": 0.818107632177984,
+      "grad_norm": 0.4761577546596527,
+      "learning_rate": 4.902240728773749e-05,
+      "loss": 6.0144,
+      "step": 2381
+    },
+    {
+      "epoch": 0.818451230511532,
+      "grad_norm": 0.5067473649978638,
+      "learning_rate": 4.884419532797169e-05,
+      "loss": 5.9129,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8187948288450801,
+      "grad_norm": 0.5061054229736328,
+      "learning_rate": 4.866627281001626e-05,
+      "loss": 6.008,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8191384271786282,
+      "grad_norm": 0.5565149784088135,
+      "learning_rate": 4.8488639989882596e-05,
+      "loss": 6.1808,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8194820255121763,
+      "grad_norm": 0.5727932453155518,
+      "learning_rate": 4.8311297123165676e-05,
+      "loss": 6.1582,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8198256238457243,
+      "grad_norm": 0.49524804949760437,
+      "learning_rate": 4.813424446504283e-05,
+      "loss": 6.0394,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8201692221792725,
+      "grad_norm": 0.5881836414337158,
+      "learning_rate": 4.7957482270274106e-05,
+      "loss": 5.9767,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8205128205128205,
+      "grad_norm": 0.5645787715911865,
+      "learning_rate": 4.778101079320152e-05,
+      "loss": 6.1487,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8208564188463686,
+      "grad_norm": 0.5424336194992065,
+      "learning_rate": 4.760483028774868e-05,
+      "loss": 6.0081,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8212000171799166,
+      "grad_norm": 0.6262788772583008,
+      "learning_rate": 4.742894100742062e-05,
+      "loss": 6.1402,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8215436155134648,
+      "grad_norm": 0.6485010385513306,
+      "learning_rate": 4.725334320530333e-05,
+      "loss": 6.0652,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8218872138470128,
+      "grad_norm": 0.657599925994873,
+      "learning_rate": 4.707803713406344e-05,
+      "loss": 6.1892,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8222308121805609,
+      "grad_norm": 0.6848873496055603,
+      "learning_rate": 4.6903023045947544e-05,
+      "loss": 6.0109,
+      "step": 2393
+    },
+    {
+      "epoch": 0.822574410514109,
+      "grad_norm": 0.6060543060302734,
+      "learning_rate": 4.672830119278257e-05,
+      "loss": 6.0803,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8229180088476571,
+      "grad_norm": 0.681524395942688,
+      "learning_rate": 4.655387182597445e-05,
+      "loss": 6.1263,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8232616071812052,
+      "grad_norm": 0.689354419708252,
+      "learning_rate": 4.6379735196508596e-05,
+      "loss": 6.225,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8236052055147532,
+      "grad_norm": 0.7622163891792297,
+      "learning_rate": 4.620589155494911e-05,
+      "loss": 6.2745,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8239488038483014,
+      "grad_norm": 0.9404433369636536,
+      "learning_rate": 4.6032341151438536e-05,
+      "loss": 6.1798,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8242924021818494,
+      "grad_norm": 0.8369949460029602,
+      "learning_rate": 4.585908423569724e-05,
+      "loss": 6.2033,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8246360005153975,
+      "grad_norm": 1.1605521440505981,
+      "learning_rate": 4.5686121057023797e-05,
+      "loss": 6.1901,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8249795988489456,
+      "grad_norm": 0.7132769823074341,
+      "learning_rate": 4.551345186429362e-05,
+      "loss": 5.9783,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8253231971824937,
+      "grad_norm": 0.6558107137680054,
+      "learning_rate": 4.534107690595937e-05,
+      "loss": 5.9746,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8256667955160417,
+      "grad_norm": 0.6440469026565552,
+      "learning_rate": 4.516899643005032e-05,
+      "loss": 6.0365,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8260103938495899,
+      "grad_norm": 0.6780853271484375,
+      "learning_rate": 4.499721068417198e-05,
+      "loss": 5.869,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8263539921831379,
+      "grad_norm": 0.5479860305786133,
+      "learning_rate": 4.482571991550566e-05,
+      "loss": 6.0475,
+      "step": 2405
+    },
+    {
+      "epoch": 0.826697590516686,
+      "grad_norm": 0.4628962576389313,
+      "learning_rate": 4.4654524370808415e-05,
+      "loss": 5.9255,
+      "step": 2406
+    },
+    {
+      "epoch": 0.827041188850234,
+      "grad_norm": 0.4671342670917511,
+      "learning_rate": 4.4483624296412425e-05,
+      "loss": 5.908,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8273847871837822,
+      "grad_norm": 0.5757843255996704,
+      "learning_rate": 4.4313019938224703e-05,
+      "loss": 5.8512,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8277283855173302,
+      "grad_norm": 0.5119126439094543,
+      "learning_rate": 4.414271154172686e-05,
+      "loss": 6.0938,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8280719838508783,
+      "grad_norm": 0.49391672015190125,
+      "learning_rate": 4.3972699351974374e-05,
+      "loss": 5.9021,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8284155821844265,
+      "grad_norm": 0.5547550916671753,
+      "learning_rate": 4.380298361359697e-05,
+      "loss": 6.0918,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8287591805179745,
+      "grad_norm": 0.555704653263092,
+      "learning_rate": 4.363356457079734e-05,
+      "loss": 6.0062,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8291027788515226,
+      "grad_norm": 0.6617487668991089,
+      "learning_rate": 4.346444246735151e-05,
+      "loss": 5.8591,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8294463771850706,
+      "grad_norm": 0.5502761602401733,
+      "learning_rate": 4.329561754660827e-05,
+      "loss": 5.9567,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8297899755186188,
+      "grad_norm": 0.5624343752861023,
+      "learning_rate": 4.312709005148871e-05,
+      "loss": 5.9204,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8301335738521668,
+      "grad_norm": 0.5430840849876404,
+      "learning_rate": 4.295886022448583e-05,
+      "loss": 6.0215,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8304771721857149,
+      "grad_norm": 0.49531492590904236,
+      "learning_rate": 4.279092830766471e-05,
+      "loss": 6.0789,
+      "step": 2417
+    },
+    {
+      "epoch": 0.830820770519263,
+      "grad_norm": 0.5341028571128845,
+      "learning_rate": 4.262329454266131e-05,
+      "loss": 5.973,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8311643688528111,
+      "grad_norm": 0.4793427884578705,
+      "learning_rate": 4.2455959170682874e-05,
+      "loss": 5.9221,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8315079671863591,
+      "grad_norm": 0.5773627161979675,
+      "learning_rate": 4.228892243250726e-05,
+      "loss": 5.9938,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8318515655199072,
+      "grad_norm": 0.5644697546958923,
+      "learning_rate": 4.212218456848243e-05,
+      "loss": 6.0866,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8321951638534553,
+      "grad_norm": 0.5478450059890747,
+      "learning_rate": 4.195574581852654e-05,
+      "loss": 5.9969,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8325387621870034,
+      "grad_norm": 0.4832424819469452,
+      "learning_rate": 4.178960642212723e-05,
+      "loss": 5.9835,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8328823605205514,
+      "grad_norm": 0.5325506925582886,
+      "learning_rate": 4.162376661834147e-05,
+      "loss": 6.0678,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8332259588540996,
+      "grad_norm": 0.4787379503250122,
+      "learning_rate": 4.145822664579491e-05,
+      "loss": 6.122,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8335695571876477,
+      "grad_norm": 0.45453977584838867,
+      "learning_rate": 4.129298674268226e-05,
+      "loss": 6.0835,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8339131555211957,
+      "grad_norm": 0.6428323984146118,
+      "learning_rate": 4.112804714676593e-05,
+      "loss": 5.9736,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8342567538547438,
+      "grad_norm": 0.49018287658691406,
+      "learning_rate": 4.096340809537655e-05,
+      "loss": 5.9848,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8346003521882919,
+      "grad_norm": 0.5041020512580872,
+      "learning_rate": 4.0799069825412176e-05,
+      "loss": 6.096,
+      "step": 2429
+    },
+    {
+      "epoch": 0.83494395052184,
+      "grad_norm": 0.4534924030303955,
+      "learning_rate": 4.06350325733382e-05,
+      "loss": 6.0244,
+      "step": 2430
+    },
+    {
+      "epoch": 0.835287548855388,
+      "grad_norm": 0.45006832480430603,
+      "learning_rate": 4.047129657518658e-05,
+      "loss": 6.0738,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8356311471889362,
+      "grad_norm": 0.5864267945289612,
+      "learning_rate": 4.030786206655626e-05,
+      "loss": 6.0116,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8359747455224842,
+      "grad_norm": 0.49015316367149353,
+      "learning_rate": 4.014472928261193e-05,
+      "loss": 6.035,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8363183438560323,
+      "grad_norm": 0.6772815585136414,
+      "learning_rate": 3.998189845808437e-05,
+      "loss": 6.1424,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8366619421895803,
+      "grad_norm": 0.5230357646942139,
+      "learning_rate": 3.98193698272698e-05,
+      "loss": 6.0487,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8370055405231285,
+      "grad_norm": 0.5348901748657227,
+      "learning_rate": 3.9657143624029665e-05,
+      "loss": 6.0033,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8373491388566765,
+      "grad_norm": 0.5133422613143921,
+      "learning_rate": 3.9495220081790297e-05,
+      "loss": 6.1628,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8376927371902246,
+      "grad_norm": 0.4899173974990845,
+      "learning_rate": 3.9333599433542284e-05,
+      "loss": 6.0278,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8380363355237727,
+      "grad_norm": 0.4960936903953552,
+      "learning_rate": 3.9172281911840636e-05,
+      "loss": 6.0515,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8383799338573208,
+      "grad_norm": 0.6296806931495667,
+      "learning_rate": 3.901126774880412e-05,
+      "loss": 6.104,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8387235321908689,
+      "grad_norm": 0.5749461054801941,
+      "learning_rate": 3.885055717611505e-05,
+      "loss": 6.0408,
+      "step": 2441
+    },
+    {
+      "epoch": 0.839067130524417,
+      "grad_norm": 0.5432246327400208,
+      "learning_rate": 3.869015042501864e-05,
+      "loss": 6.146,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8394107288579651,
+      "grad_norm": 0.6659320592880249,
+      "learning_rate": 3.85300477263234e-05,
+      "loss": 6.2118,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8397543271915131,
+      "grad_norm": 0.6398926973342896,
+      "learning_rate": 3.8370249310399955e-05,
+      "loss": 6.1283,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8400979255250612,
+      "grad_norm": 0.6013005375862122,
+      "learning_rate": 3.821075540718122e-05,
+      "loss": 5.9936,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8404415238586093,
+      "grad_norm": 0.7379624247550964,
+      "learning_rate": 3.805156624616199e-05,
+      "loss": 6.2026,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8407851221921574,
+      "grad_norm": 0.7793958187103271,
+      "learning_rate": 3.789268205639859e-05,
+      "loss": 6.1806,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8411287205257054,
+      "grad_norm": 0.8797317147254944,
+      "learning_rate": 3.773410306650832e-05,
+      "loss": 6.1008,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8414723188592536,
+      "grad_norm": 1.1056629419326782,
+      "learning_rate": 3.757582950466967e-05,
+      "loss": 6.2515,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8418159171928016,
+      "grad_norm": 1.1044436693191528,
+      "learning_rate": 3.7417861598621345e-05,
+      "loss": 6.3385,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8421595155263497,
+      "grad_norm": 0.6891606450080872,
+      "learning_rate": 3.72601995756624e-05,
+      "loss": 5.9451,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8425031138598977,
+      "grad_norm": 0.7626943588256836,
+      "learning_rate": 3.710284366265168e-05,
+      "loss": 6.032,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8428467121934459,
+      "grad_norm": 0.6951205730438232,
+      "learning_rate": 3.69457940860077e-05,
+      "loss": 5.9814,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8431903105269939,
+      "grad_norm": 0.6206390857696533,
+      "learning_rate": 3.6789051071708016e-05,
+      "loss": 5.9554,
+      "step": 2454
+    },
+    {
+      "epoch": 0.843533908860542,
+      "grad_norm": 0.6567794680595398,
+      "learning_rate": 3.6632614845289154e-05,
+      "loss": 5.8821,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8438775071940902,
+      "grad_norm": 0.5018622875213623,
+      "learning_rate": 3.64764856318463e-05,
+      "loss": 5.9425,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8442211055276382,
+      "grad_norm": 0.5784042477607727,
+      "learning_rate": 3.632066365603259e-05,
+      "loss": 5.925,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8445647038611863,
+      "grad_norm": 0.5264183878898621,
+      "learning_rate": 3.616514914205954e-05,
+      "loss": 5.9962,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8449083021947343,
+      "grad_norm": 0.49773308634757996,
+      "learning_rate": 3.600994231369578e-05,
+      "loss": 6.0242,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8452519005282825,
+      "grad_norm": 0.4520735740661621,
+      "learning_rate": 3.585504339426754e-05,
+      "loss": 5.9737,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8455954988618305,
+      "grad_norm": 0.4381536841392517,
+      "learning_rate": 3.5700452606657855e-05,
+      "loss": 6.0548,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8459390971953786,
+      "grad_norm": 0.5472382307052612,
+      "learning_rate": 3.5546170173306444e-05,
+      "loss": 6.0331,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8462826955289267,
+      "grad_norm": 0.5395277738571167,
+      "learning_rate": 3.53921963162093e-05,
+      "loss": 5.9856,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8466262938624748,
+      "grad_norm": 0.5448597073554993,
+      "learning_rate": 3.5238531256918506e-05,
+      "loss": 6.0873,
+      "step": 2464
+    },
+    {
+      "epoch": 0.8469698921960228,
+      "grad_norm": 0.5200133323669434,
+      "learning_rate": 3.5085175216541614e-05,
+      "loss": 6.0775,
+      "step": 2465
+    },
+    {
+      "epoch": 0.847313490529571,
+      "grad_norm": 0.5575097799301147,
+      "learning_rate": 3.493212841574173e-05,
+      "loss": 6.0285,
+      "step": 2466
+    },
+    {
+      "epoch": 0.847657088863119,
+      "grad_norm": 0.5278059840202332,
+      "learning_rate": 3.4779391074736905e-05,
+      "loss": 6.0,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8480006871966671,
+      "grad_norm": 0.5051273107528687,
+      "learning_rate": 3.462696341329996e-05,
+      "loss": 5.9783,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8483442855302151,
+      "grad_norm": 0.45483526587486267,
+      "learning_rate": 3.4474845650758094e-05,
+      "loss": 6.0531,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8486878838637633,
+      "grad_norm": 0.5154264569282532,
+      "learning_rate": 3.432303800599254e-05,
+      "loss": 6.053,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8490314821973114,
+      "grad_norm": 0.4888317584991455,
+      "learning_rate": 3.4171540697438356e-05,
+      "loss": 6.1426,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8493750805308594,
+      "grad_norm": 0.5659269690513611,
+      "learning_rate": 3.4020353943084087e-05,
+      "loss": 5.9778,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8497186788644076,
+      "grad_norm": 0.4910965859889984,
+      "learning_rate": 3.386947796047144e-05,
+      "loss": 6.0154,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8500622771979556,
+      "grad_norm": 0.5661794543266296,
+      "learning_rate": 3.371891296669474e-05,
+      "loss": 6.0507,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8504058755315037,
+      "grad_norm": 0.4793378710746765,
+      "learning_rate": 3.356865917840124e-05,
+      "loss": 6.1263,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8507494738650517,
+      "grad_norm": 0.5714482069015503,
+      "learning_rate": 3.3418716811789956e-05,
+      "loss": 6.0975,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8510930721985999,
+      "grad_norm": 0.5254719257354736,
+      "learning_rate": 3.326908608261212e-05,
+      "loss": 6.033,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8514366705321479,
+      "grad_norm": 0.40919145941734314,
+      "learning_rate": 3.311976720617038e-05,
+      "loss": 5.9691,
+      "step": 2478
+    },
+    {
+      "epoch": 0.851780268865696,
+      "grad_norm": 0.5050671100616455,
+      "learning_rate": 3.297076039731883e-05,
+      "loss": 6.1006,
+      "step": 2479
+    },
+    {
+      "epoch": 0.852123867199244,
+      "grad_norm": 0.49542513489723206,
+      "learning_rate": 3.2822065870462214e-05,
+      "loss": 6.0539,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8524674655327922,
+      "grad_norm": 0.5702757239341736,
+      "learning_rate": 3.2673683839556376e-05,
+      "loss": 6.1042,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8528110638663402,
+      "grad_norm": 0.5085723400115967,
+      "learning_rate": 3.252561451810712e-05,
+      "loss": 5.9759,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8531546621998883,
+      "grad_norm": 0.5009546279907227,
+      "learning_rate": 3.237785811917049e-05,
+      "loss": 6.0489,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8534982605334364,
+      "grad_norm": 0.489520400762558,
+      "learning_rate": 3.223041485535225e-05,
+      "loss": 5.9973,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8538418588669845,
+      "grad_norm": 0.5538635849952698,
+      "learning_rate": 3.208328493880763e-05,
+      "loss": 6.1599,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8541854572005326,
+      "grad_norm": 0.5288501977920532,
+      "learning_rate": 3.19364685812408e-05,
+      "loss": 6.0023,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8545290555340807,
+      "grad_norm": 0.5383173823356628,
+      "learning_rate": 3.178996599390499e-05,
+      "loss": 6.0028,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8548726538676288,
+      "grad_norm": 0.5179246664047241,
+      "learning_rate": 3.164377738760182e-05,
+      "loss": 6.1045,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8552162522011768,
+      "grad_norm": 0.5900654792785645,
+      "learning_rate": 3.149790297268107e-05,
+      "loss": 6.1438,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8555598505347249,
+      "grad_norm": 0.6226531863212585,
+      "learning_rate": 3.135234295904066e-05,
+      "loss": 6.0877,
+      "step": 2490
+    },
+    {
+      "epoch": 0.855903448868273,
+      "grad_norm": 0.639117419719696,
+      "learning_rate": 3.1207097556125777e-05,
+      "loss": 6.0413,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8562470472018211,
+      "grad_norm": 0.6743376851081848,
+      "learning_rate": 3.106216697292932e-05,
+      "loss": 6.1535,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8565906455353691,
+      "grad_norm": 0.6806616187095642,
+      "learning_rate": 3.0917551417990854e-05,
+      "loss": 6.0679,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8569342438689173,
+      "grad_norm": 0.6853029131889343,
+      "learning_rate": 3.0773251099396773e-05,
+      "loss": 6.0792,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8572778422024653,
+      "grad_norm": 0.6363233923912048,
+      "learning_rate": 3.062926622477996e-05,
+      "loss": 6.1392,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8576214405360134,
+      "grad_norm": 0.6565764546394348,
+      "learning_rate": 3.0485597001319366e-05,
+      "loss": 6.0897,
+      "step": 2496
+    },
+    {
+      "epoch": 0.8579650388695614,
+      "grad_norm": 0.8439285755157471,
+      "learning_rate": 3.0342243635739593e-05,
+      "loss": 6.0918,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8583086372031096,
+      "grad_norm": 0.7531304955482483,
+      "learning_rate": 3.0199206334310948e-05,
+      "loss": 6.1271,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8586522355366576,
+      "grad_norm": 0.8806336522102356,
+      "learning_rate": 3.0056485302848934e-05,
+      "loss": 6.2187,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8589958338702057,
+      "grad_norm": 1.2116683721542358,
+      "learning_rate": 2.9914080746713896e-05,
+      "loss": 6.2619,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8589958338702057,
+      "eval_loss": 6.026234149932861,
+      "eval_runtime": 724.3408,
+      "eval_samples_per_second": 25.713,
+      "eval_steps_per_second": 6.429,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8593394322037539,
+      "grad_norm": 0.7180514335632324,
+      "learning_rate": 2.9771992870810894e-05,
+      "loss": 5.8551,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8596830305373019,
+      "grad_norm": 0.6997862458229065,
+      "learning_rate": 2.963022187958922e-05,
+      "loss": 5.9469,
+      "step": 2502
+    },
+    {
+      "epoch": 0.86002662887085,
+      "grad_norm": 0.7439790368080139,
+      "learning_rate": 2.9488767977042253e-05,
+      "loss": 5.9154,
+      "step": 2503
+    },
+    {
+      "epoch": 0.860370227204398,
+      "grad_norm": 0.670836865901947,
+      "learning_rate": 2.9347631366707124e-05,
+      "loss": 6.1126,
+      "step": 2504
+    },
+    {
+      "epoch": 0.8607138255379462,
+      "grad_norm": 0.664583683013916,
+      "learning_rate": 2.9206812251664492e-05,
+      "loss": 5.935,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8610574238714942,
+      "grad_norm": 0.5612172484397888,
+      "learning_rate": 2.9066310834537867e-05,
+      "loss": 6.044,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8614010222050423,
+      "grad_norm": 0.5193923115730286,
+      "learning_rate": 2.892612731749414e-05,
+      "loss": 5.9687,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8617446205385904,
+      "grad_norm": 0.4159911572933197,
+      "learning_rate": 2.8786261902242232e-05,
+      "loss": 5.9301,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8620882188721385,
+      "grad_norm": 0.442753404378891,
+      "learning_rate": 2.8646714790033752e-05,
+      "loss": 5.9968,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8624318172056865,
+      "grad_norm": 0.5828900337219238,
+      "learning_rate": 2.8507486181662075e-05,
+      "loss": 5.9093,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8627754155392346,
+      "grad_norm": 0.6334620118141174,
+      "learning_rate": 2.8368576277462422e-05,
+      "loss": 5.9837,
+      "step": 2511
+    },
+    {
+      "epoch": 0.8631190138727827,
+      "grad_norm": 0.5630239844322205,
+      "learning_rate": 2.822998527731127e-05,
+      "loss": 5.9947,
+      "step": 2512
+    },
+    {
+      "epoch": 0.8634626122063308,
+      "grad_norm": 0.6125161647796631,
+      "learning_rate": 2.8091713380626492e-05,
+      "loss": 5.9299,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8638062105398788,
+      "grad_norm": 0.6197602152824402,
+      "learning_rate": 2.7953760786366493e-05,
+      "loss": 6.0298,
+      "step": 2514
+    },
+    {
+      "epoch": 0.864149808873427,
+      "grad_norm": 0.6317358016967773,
+      "learning_rate": 2.7816127693030462e-05,
+      "loss": 6.0517,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8644934072069751,
+      "grad_norm": 0.6413782835006714,
+      "learning_rate": 2.7678814298657734e-05,
+      "loss": 6.0724,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8648370055405231,
+      "grad_norm": 0.5042860507965088,
+      "learning_rate": 2.7541820800827733e-05,
+      "loss": 6.0285,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8651806038740713,
+      "grad_norm": 0.49634793400764465,
+      "learning_rate": 2.7405147396659557e-05,
+      "loss": 6.0163,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8655242022076193,
+      "grad_norm": 0.4745235741138458,
+      "learning_rate": 2.7268794282811595e-05,
+      "loss": 6.0094,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8658678005411674,
+      "grad_norm": 0.48123598098754883,
+      "learning_rate": 2.7132761655481537e-05,
+      "loss": 6.1046,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8662113988747154,
+      "grad_norm": 0.7108414769172668,
+      "learning_rate": 2.699704971040587e-05,
+      "loss": 5.9763,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8665549972082636,
+      "grad_norm": 0.538608968257904,
+      "learning_rate": 2.6861658642859693e-05,
+      "loss": 5.9979,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8668985955418116,
+      "grad_norm": 0.49514704942703247,
+      "learning_rate": 2.6726588647656204e-05,
+      "loss": 5.9548,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8672421938753597,
+      "grad_norm": 0.589779794216156,
+      "learning_rate": 2.659183991914696e-05,
+      "loss": 6.0988,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8675857922089077,
+      "grad_norm": 0.5593236684799194,
+      "learning_rate": 2.6457412651220896e-05,
+      "loss": 6.1345,
+      "step": 2525
+    },
+    {
+      "epoch": 0.8679293905424559,
+      "grad_norm": 0.49659085273742676,
+      "learning_rate": 2.6323307037304624e-05,
+      "loss": 6.073,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8682729888760039,
+      "grad_norm": 0.6141493320465088,
+      "learning_rate": 2.6189523270361865e-05,
+      "loss": 6.0561,
+      "step": 2527
+    },
+    {
+      "epoch": 0.868616587209552,
+      "grad_norm": 0.4329065978527069,
+      "learning_rate": 2.605606154289322e-05,
+      "loss": 5.9661,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8689601855431001,
+      "grad_norm": 0.4631238281726837,
+      "learning_rate": 2.5922922046935914e-05,
+      "loss": 5.9962,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8693037838766482,
+      "grad_norm": 0.5284918546676636,
+      "learning_rate": 2.5790104974063505e-05,
+      "loss": 6.0944,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8696473822101963,
+      "grad_norm": 0.5931186676025391,
+      "learning_rate": 2.5657610515385647e-05,
+      "loss": 6.0207,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8699909805437444,
+      "grad_norm": 0.5561034679412842,
+      "learning_rate": 2.552543886154779e-05,
+      "loss": 6.044,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8703345788772925,
+      "grad_norm": 0.5106163620948792,
+      "learning_rate": 2.539359020273094e-05,
+      "loss": 5.9788,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8706781772108405,
+      "grad_norm": 0.5758001804351807,
+      "learning_rate": 2.5262064728651197e-05,
+      "loss": 6.0796,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8710217755443886,
+      "grad_norm": 0.4941878020763397,
+      "learning_rate": 2.5130862628559765e-05,
+      "loss": 6.0645,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8713653738779367,
+      "grad_norm": 0.6312124133110046,
+      "learning_rate": 2.499998409124252e-05,
+      "loss": 5.9077,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8717089722114848,
+      "grad_norm": 0.6568105816841125,
+      "learning_rate": 2.4869429305019785e-05,
+      "loss": 6.0868,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8720525705450328,
+      "grad_norm": 0.5947930812835693,
+      "learning_rate": 2.4739198457745915e-05,
+      "loss": 6.1068,
+      "step": 2538
+    },
+    {
+      "epoch": 0.872396168878581,
+      "grad_norm": 0.589493989944458,
+      "learning_rate": 2.4609291736809397e-05,
+      "loss": 5.9835,
+      "step": 2539
+    },
+    {
+      "epoch": 0.872739767212129,
+      "grad_norm": 0.6117597818374634,
+      "learning_rate": 2.447970932913207e-05,
+      "loss": 6.1012,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8730833655456771,
+      "grad_norm": 0.6265408396720886,
+      "learning_rate": 2.435045142116929e-05,
+      "loss": 6.0829,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8734269638792251,
+      "grad_norm": 0.5803533792495728,
+      "learning_rate": 2.4221518198909415e-05,
+      "loss": 6.0835,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8737705622127733,
+      "grad_norm": 0.5373610854148865,
+      "learning_rate": 2.409290984787371e-05,
+      "loss": 6.0519,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8741141605463213,
+      "grad_norm": 0.6814734935760498,
+      "learning_rate": 2.3964626553115766e-05,
+      "loss": 5.9762,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8744577588798694,
+      "grad_norm": 0.7417898774147034,
+      "learning_rate": 2.3836668499221752e-05,
+      "loss": 6.1121,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8748013572134176,
+      "grad_norm": 0.5553413033485413,
+      "learning_rate": 2.370903587030965e-05,
+      "loss": 6.1261,
+      "step": 2546
+    },
+    {
+      "epoch": 0.8751449555469656,
+      "grad_norm": 0.8009063005447388,
+      "learning_rate": 2.3581728850029182e-05,
+      "loss": 6.2087,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8754885538805137,
+      "grad_norm": 0.7052872180938721,
+      "learning_rate": 2.345474762156169e-05,
+      "loss": 6.2921,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8758321522140617,
+      "grad_norm": 0.9335289597511292,
+      "learning_rate": 2.3328092367619596e-05,
+      "loss": 6.1382,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8761757505476099,
+      "grad_norm": 1.0865561962127686,
+      "learning_rate": 2.3201763270446457e-05,
+      "loss": 6.0753,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8765193488811579,
+      "grad_norm": 0.7006704807281494,
+      "learning_rate": 2.3075760511816257e-05,
+      "loss": 5.7694,
+      "step": 2551
+    },
+    {
+      "epoch": 0.876862947214706,
+      "grad_norm": 0.6684394478797913,
+      "learning_rate": 2.2950084273033632e-05,
+      "loss": 5.835,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8772065455482541,
+      "grad_norm": 0.6840022206306458,
+      "learning_rate": 2.2824734734933322e-05,
+      "loss": 5.8673,
+      "step": 2553
+    },
+    {
+      "epoch": 0.8775501438818022,
+      "grad_norm": 0.6259867548942566,
+      "learning_rate": 2.2699712077880046e-05,
+      "loss": 6.0112,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8778937422153502,
+      "grad_norm": 0.5849865674972534,
+      "learning_rate": 2.2575016481767936e-05,
+      "loss": 5.988,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8782373405488983,
+      "grad_norm": 0.5193759799003601,
+      "learning_rate": 2.2450648126020907e-05,
+      "loss": 5.9885,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8785809388824464,
+      "grad_norm": 0.4761309325695038,
+      "learning_rate": 2.2326607189591676e-05,
+      "loss": 5.8873,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8789245372159945,
+      "grad_norm": 0.5957377552986145,
+      "learning_rate": 2.220289385096194e-05,
+      "loss": 5.8765,
+      "step": 2558
+    },
+    {
+      "epoch": 0.8792681355495426,
+      "grad_norm": 0.5031549334526062,
+      "learning_rate": 2.2079508288142092e-05,
+      "loss": 6.0119,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8796117338830907,
+      "grad_norm": 0.4540199041366577,
+      "learning_rate": 2.195645067867086e-05,
+      "loss": 5.9878,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8799553322166388,
+      "grad_norm": 0.48119738698005676,
+      "learning_rate": 2.183372119961499e-05,
+      "loss": 5.9127,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8802989305501868,
+      "grad_norm": 0.5656386613845825,
+      "learning_rate": 2.171132002756915e-05,
+      "loss": 6.0097,
+      "step": 2562
+    },
+    {
+      "epoch": 0.880642528883735,
+      "grad_norm": 0.49384355545043945,
+      "learning_rate": 2.1589247338655666e-05,
+      "loss": 5.9723,
+      "step": 2563
+    },
+    {
+      "epoch": 0.880986127217283,
+      "grad_norm": 0.5116087794303894,
+      "learning_rate": 2.1467503308524096e-05,
+      "loss": 6.0294,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8813297255508311,
+      "grad_norm": 0.511194109916687,
+      "learning_rate": 2.1346088112351252e-05,
+      "loss": 5.9535,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8816733238843791,
+      "grad_norm": 0.4660123586654663,
+      "learning_rate": 2.122500192484056e-05,
+      "loss": 6.0129,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8820169222179273,
+      "grad_norm": 0.507530689239502,
+      "learning_rate": 2.1104244920222226e-05,
+      "loss": 5.9942,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8823605205514753,
+      "grad_norm": 0.4880560338497162,
+      "learning_rate": 2.0983817272252737e-05,
+      "loss": 5.9319,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8827041188850234,
+      "grad_norm": 0.47986316680908203,
+      "learning_rate": 2.08637191542147e-05,
+      "loss": 6.0592,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8830477172185714,
+      "grad_norm": 0.4974256157875061,
+      "learning_rate": 2.074395073891644e-05,
+      "loss": 6.1117,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8833913155521196,
+      "grad_norm": 0.41185539960861206,
+      "learning_rate": 2.06245121986921e-05,
+      "loss": 6.0907,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8837349138856676,
+      "grad_norm": 0.47288328409194946,
+      "learning_rate": 2.0505403705400883e-05,
+      "loss": 6.0898,
+      "step": 2572
+    },
+    {
+      "epoch": 0.8840785122192157,
+      "grad_norm": 0.4536161422729492,
+      "learning_rate": 2.0386625430427436e-05,
+      "loss": 6.0654,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8844221105527639,
+      "grad_norm": 0.4487610459327698,
+      "learning_rate": 2.026817754468091e-05,
+      "loss": 6.0253,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8847657088863119,
+      "grad_norm": 0.4847327768802643,
+      "learning_rate": 2.015006021859528e-05,
+      "loss": 6.1225,
+      "step": 2575
+    },
+    {
+      "epoch": 0.88510930721986,
+      "grad_norm": 0.5354999899864197,
+      "learning_rate": 2.0032273622128784e-05,
+      "loss": 6.0677,
+      "step": 2576
+    },
+    {
+      "epoch": 0.885452905553408,
+      "grad_norm": 0.6489529013633728,
+      "learning_rate": 1.9914817924763878e-05,
+      "loss": 6.0719,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8857965038869562,
+      "grad_norm": 0.4547826051712036,
+      "learning_rate": 1.9797693295506735e-05,
+      "loss": 6.0456,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8861401022205042,
+      "grad_norm": 0.48801037669181824,
+      "learning_rate": 1.9680899902887266e-05,
+      "loss": 6.0169,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8864837005540523,
+      "grad_norm": 0.5203239917755127,
+      "learning_rate": 1.9564437914958765e-05,
+      "loss": 6.0425,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8868272988876004,
+      "grad_norm": 0.501476526260376,
+      "learning_rate": 1.94483074992976e-05,
+      "loss": 6.0669,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8871708972211485,
+      "grad_norm": 0.49509742856025696,
+      "learning_rate": 1.9332508823003192e-05,
+      "loss": 5.947,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8875144955546965,
+      "grad_norm": 0.5070309042930603,
+      "learning_rate": 1.9217042052697393e-05,
+      "loss": 6.1098,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8878580938882447,
+      "grad_norm": 0.47886112332344055,
+      "learning_rate": 1.910190735452466e-05,
+      "loss": 6.0617,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8882016922217927,
+      "grad_norm": 0.6764759421348572,
+      "learning_rate": 1.8987104894151592e-05,
+      "loss": 5.942,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8885452905553408,
+      "grad_norm": 0.7424550652503967,
+      "learning_rate": 1.8872634836766768e-05,
+      "loss": 6.0808,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.49124762415885925,
+      "learning_rate": 1.8758497347080266e-05,
+      "loss": 5.9537,
+      "step": 2587
+    },
+    {
+      "epoch": 0.889232487222437,
+      "grad_norm": 0.5248420834541321,
+      "learning_rate": 1.864469258932397e-05,
+      "loss": 6.011,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8895760855559851,
+      "grad_norm": 0.5532990097999573,
+      "learning_rate": 1.853122072725072e-05,
+      "loss": 6.0058,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8899196838895331,
+      "grad_norm": 0.5646762847900391,
+      "learning_rate": 1.8418081924134494e-05,
+      "loss": 6.1114,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8902632822230813,
+      "grad_norm": 0.5970567464828491,
+      "learning_rate": 1.8305276342770015e-05,
+      "loss": 6.0876,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8906068805566293,
+      "grad_norm": 0.6650955080986023,
+      "learning_rate": 1.8192804145472502e-05,
+      "loss": 6.0022,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8909504788901774,
+      "grad_norm": 0.6694637537002563,
+      "learning_rate": 1.8080665494077468e-05,
+      "loss": 6.1372,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8912940772237254,
+      "grad_norm": 0.6882001757621765,
+      "learning_rate": 1.7968860549940512e-05,
+      "loss": 6.0342,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8916376755572736,
+      "grad_norm": 0.6639662384986877,
+      "learning_rate": 1.7857389473937058e-05,
+      "loss": 6.13,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8919812738908216,
+      "grad_norm": 0.7884045243263245,
+      "learning_rate": 1.7746252426462134e-05,
+      "loss": 6.2426,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8923248722243697,
+      "grad_norm": 0.8145104050636292,
+      "learning_rate": 1.7635449567430185e-05,
+      "loss": 6.1495,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8926684705579178,
+      "grad_norm": 0.8474352955818176,
+      "learning_rate": 1.7524981056274647e-05,
+      "loss": 6.2624,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8930120688914659,
+      "grad_norm": 0.9839531183242798,
+      "learning_rate": 1.7414847051948012e-05,
+      "loss": 6.2734,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8933556672250139,
+      "grad_norm": 1.2112486362457275,
+      "learning_rate": 1.730504771292138e-05,
+      "loss": 6.0494,
+      "step": 2600
+    },
+    {
+      "epoch": 0.893699265558562,
+      "grad_norm": 0.7058858275413513,
+      "learning_rate": 1.719558319718434e-05,
+      "loss": 5.9493,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8940428638921101,
+      "grad_norm": 0.7136659622192383,
+      "learning_rate": 1.7086453662244678e-05,
+      "loss": 5.9466,
+      "step": 2602
+    },
+    {
+      "epoch": 0.8943864622256582,
+      "grad_norm": 0.6175331473350525,
+      "learning_rate": 1.697765926512823e-05,
+      "loss": 6.0478,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8947300605592063,
+      "grad_norm": 0.7400083541870117,
+      "learning_rate": 1.6869200162378474e-05,
+      "loss": 5.9836,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8950736588927544,
+      "grad_norm": 0.6715840697288513,
+      "learning_rate": 1.6761076510056623e-05,
+      "loss": 5.8835,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8954172572263025,
+      "grad_norm": 0.6418355703353882,
+      "learning_rate": 1.665328846374106e-05,
+      "loss": 6.082,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8957608555598505,
+      "grad_norm": 0.6705430150032043,
+      "learning_rate": 1.6545836178527313e-05,
+      "loss": 5.9216,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8961044538933987,
+      "grad_norm": 0.5584115386009216,
+      "learning_rate": 1.6438719809027806e-05,
+      "loss": 5.9343,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8964480522269467,
+      "grad_norm": 0.6551584005355835,
+      "learning_rate": 1.6331939509371647e-05,
+      "loss": 5.989,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8967916505604948,
+      "grad_norm": 0.5179411172866821,
+      "learning_rate": 1.6225495433204256e-05,
+      "loss": 5.9579,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8971352488940428,
+      "grad_norm": 0.5057214498519897,
+      "learning_rate": 1.6119387733687374e-05,
+      "loss": 6.0373,
+      "step": 2611
+    },
+    {
+      "epoch": 0.897478847227591,
+      "grad_norm": 0.4302932918071747,
+      "learning_rate": 1.60136165634987e-05,
+      "loss": 5.9966,
+      "step": 2612
+    },
+    {
+      "epoch": 0.897822445561139,
+      "grad_norm": 0.5576332807540894,
+      "learning_rate": 1.590818207483169e-05,
+      "loss": 5.9207,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8981660438946871,
+      "grad_norm": 0.5023398995399475,
+      "learning_rate": 1.58030844193954e-05,
+      "loss": 5.8995,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8985096422282352,
+      "grad_norm": 0.555558443069458,
+      "learning_rate": 1.5698323748414122e-05,
+      "loss": 6.0222,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8988532405617833,
+      "grad_norm": 0.46205613017082214,
+      "learning_rate": 1.5593900212627326e-05,
+      "loss": 6.1039,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8991968388953313,
+      "grad_norm": 0.5434067249298096,
+      "learning_rate": 1.548981396228938e-05,
+      "loss": 5.9208,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8995404372288794,
+      "grad_norm": 0.5215374231338501,
+      "learning_rate": 1.5386065147169392e-05,
+      "loss": 5.9905,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8998840355624276,
+      "grad_norm": 0.5159684419631958,
+      "learning_rate": 1.528265391655076e-05,
+      "loss": 5.9615,
+      "step": 2619
+    },
+    {
+      "epoch": 0.9002276338959756,
+      "grad_norm": 0.5324832797050476,
+      "learning_rate": 1.5179580419231371e-05,
+      "loss": 5.9843,
+      "step": 2620
+    },
+    {
+      "epoch": 0.9005712322295237,
+      "grad_norm": 0.49988093972206116,
+      "learning_rate": 1.5076844803522921e-05,
+      "loss": 6.0481,
+      "step": 2621
+    },
+    {
+      "epoch": 0.9009148305630718,
+      "grad_norm": 0.4379015266895294,
+      "learning_rate": 1.497444721725108e-05,
+      "loss": 6.0757,
+      "step": 2622
+    },
+    {
+      "epoch": 0.9012584288966199,
+      "grad_norm": 0.45072677731513977,
+      "learning_rate": 1.4872387807755072e-05,
+      "loss": 6.0539,
+      "step": 2623
+    },
+    {
+      "epoch": 0.9016020272301679,
+      "grad_norm": 0.5357469916343689,
+      "learning_rate": 1.4770666721887622e-05,
+      "loss": 6.1215,
+      "step": 2624
+    },
+    {
+      "epoch": 0.901945625563716,
+      "grad_norm": 0.48705291748046875,
+      "learning_rate": 1.4669284106014369e-05,
+      "loss": 6.0271,
+      "step": 2625
+    },
+    {
+      "epoch": 0.9022892238972641,
+      "grad_norm": 0.4232194721698761,
+      "learning_rate": 1.4568240106014291e-05,
+      "loss": 5.9916,
+      "step": 2626
+    },
+    {
+      "epoch": 0.9026328222308122,
+      "grad_norm": 0.49429526925086975,
+      "learning_rate": 1.4467534867278864e-05,
+      "loss": 5.9512,
+      "step": 2627
+    },
+    {
+      "epoch": 0.9029764205643602,
+      "grad_norm": 0.47070807218551636,
+      "learning_rate": 1.436716853471226e-05,
+      "loss": 6.0667,
+      "step": 2628
+    },
+    {
+      "epoch": 0.9033200188979084,
+      "grad_norm": 0.5034219026565552,
+      "learning_rate": 1.4267141252730958e-05,
+      "loss": 6.0875,
+      "step": 2629
+    },
+    {
+      "epoch": 0.9036636172314564,
+      "grad_norm": 0.49576908349990845,
+      "learning_rate": 1.4167453165263495e-05,
+      "loss": 5.9769,
+      "step": 2630
+    },
+    {
+      "epoch": 0.9040072155650045,
+      "grad_norm": 0.6180304288864136,
+      "learning_rate": 1.4068104415750572e-05,
+      "loss": 6.1032,
+      "step": 2631
+    },
+    {
+      "epoch": 0.9043508138985525,
+      "grad_norm": 0.49888142943382263,
+      "learning_rate": 1.3969095147144339e-05,
+      "loss": 6.1171,
+      "step": 2632
+    },
+    {
+      "epoch": 0.9046944122321007,
+      "grad_norm": 0.5140848755836487,
+      "learning_rate": 1.3870425501908674e-05,
+      "loss": 6.1556,
+      "step": 2633
+    },
+    {
+      "epoch": 0.9050380105656488,
+      "grad_norm": 0.5649397969245911,
+      "learning_rate": 1.3772095622018698e-05,
+      "loss": 6.0178,
+      "step": 2634
+    },
+    {
+      "epoch": 0.9053816088991968,
+      "grad_norm": 0.5044685006141663,
+      "learning_rate": 1.3674105648960683e-05,
+      "loss": 6.1763,
+      "step": 2635
+    },
+    {
+      "epoch": 0.905725207232745,
+      "grad_norm": 0.59349524974823,
+      "learning_rate": 1.3576455723731645e-05,
+      "loss": 6.0187,
+      "step": 2636
+    },
+    {
+      "epoch": 0.906068805566293,
+      "grad_norm": 0.6226266622543335,
+      "learning_rate": 1.3479145986839636e-05,
+      "loss": 6.0291,
+      "step": 2637
+    },
+    {
+      "epoch": 0.9064124038998411,
+      "grad_norm": 0.6408131718635559,
+      "learning_rate": 1.3382176578302846e-05,
+      "loss": 6.0087,
+      "step": 2638
+    },
+    {
+      "epoch": 0.9067560022333891,
+      "grad_norm": 0.5794918537139893,
+      "learning_rate": 1.3285547637650052e-05,
+      "loss": 6.0192,
+      "step": 2639
+    },
+    {
+      "epoch": 0.9070996005669373,
+      "grad_norm": 0.5764051675796509,
+      "learning_rate": 1.3189259303919954e-05,
+      "loss": 6.1282,
+      "step": 2640
+    },
+    {
+      "epoch": 0.9074431989004853,
+      "grad_norm": 0.5949744582176208,
+      "learning_rate": 1.3093311715661304e-05,
+      "loss": 6.0212,
+      "step": 2641
+    },
+    {
+      "epoch": 0.9077867972340334,
+      "grad_norm": 0.5786007046699524,
+      "learning_rate": 1.2997705010932393e-05,
+      "loss": 6.059,
+      "step": 2642
+    },
+    {
+      "epoch": 0.9081303955675815,
+      "grad_norm": 0.6125741600990295,
+      "learning_rate": 1.2902439327301146e-05,
+      "loss": 6.076,
+      "step": 2643
+    },
+    {
+      "epoch": 0.9084739939011296,
+      "grad_norm": 0.5794528722763062,
+      "learning_rate": 1.2807514801844723e-05,
+      "loss": 6.0601,
+      "step": 2644
+    },
+    {
+      "epoch": 0.9088175922346776,
+      "grad_norm": 0.6618969440460205,
+      "learning_rate": 1.2712931571149444e-05,
+      "loss": 6.0093,
+      "step": 2645
+    },
+    {
+      "epoch": 0.9091611905682258,
+      "grad_norm": 0.7055357694625854,
+      "learning_rate": 1.261868977131056e-05,
+      "loss": 6.2117,
+      "step": 2646
+    },
+    {
+      "epoch": 0.9095047889017738,
+      "grad_norm": 0.6832464933395386,
+      "learning_rate": 1.252478953793193e-05,
+      "loss": 6.1221,
+      "step": 2647
+    },
+    {
+      "epoch": 0.9098483872353219,
+      "grad_norm": 0.8896463513374329,
+      "learning_rate": 1.2431231006126003e-05,
+      "loss": 6.1605,
+      "step": 2648
+    },
+    {
+      "epoch": 0.91019198556887,
+      "grad_norm": 0.8412816524505615,
+      "learning_rate": 1.2338014310513596e-05,
+      "loss": 6.1005,
+      "step": 2649
+    },
+    {
+      "epoch": 0.9105355839024181,
+      "grad_norm": 1.1724843978881836,
+      "learning_rate": 1.2245139585223636e-05,
+      "loss": 6.2482,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9108791822359662,
+      "grad_norm": 0.6163187026977539,
+      "learning_rate": 1.2152606963892864e-05,
+      "loss": 5.7662,
+      "step": 2651
+    },
+    {
+      "epoch": 0.9112227805695142,
+      "grad_norm": 0.5164371132850647,
+      "learning_rate": 1.2060416579666028e-05,
+      "loss": 5.9042,
+      "step": 2652
+    },
+    {
+      "epoch": 0.9115663789030624,
+      "grad_norm": 0.5768353939056396,
+      "learning_rate": 1.1968568565195182e-05,
+      "loss": 5.9316,
+      "step": 2653
+    },
+    {
+      "epoch": 0.9119099772366104,
+      "grad_norm": 0.5464223623275757,
+      "learning_rate": 1.1877063052639913e-05,
+      "loss": 5.8956,
+      "step": 2654
+    },
+    {
+      "epoch": 0.9122535755701585,
+      "grad_norm": 0.5711444616317749,
+      "learning_rate": 1.1785900173666903e-05,
+      "loss": 6.0076,
+      "step": 2655
+    },
+    {
+      "epoch": 0.9125971739037065,
+      "grad_norm": 0.4955626130104065,
+      "learning_rate": 1.169508005944983e-05,
+      "loss": 6.0148,
+      "step": 2656
+    },
+    {
+      "epoch": 0.9129407722372547,
+      "grad_norm": 0.40896841883659363,
+      "learning_rate": 1.1604602840669164e-05,
+      "loss": 6.0712,
+      "step": 2657
+    },
+    {
+      "epoch": 0.9132843705708027,
+      "grad_norm": 0.5791770219802856,
+      "learning_rate": 1.1514468647512072e-05,
+      "loss": 5.7973,
+      "step": 2658
+    },
+    {
+      "epoch": 0.9136279689043508,
+      "grad_norm": 0.5330086350440979,
+      "learning_rate": 1.1424677609671974e-05,
+      "loss": 5.939,
+      "step": 2659
+    },
+    {
+      "epoch": 0.9139715672378989,
+      "grad_norm": 0.44915223121643066,
+      "learning_rate": 1.1335229856348689e-05,
+      "loss": 5.9109,
+      "step": 2660
+    },
+    {
+      "epoch": 0.914315165571447,
+      "grad_norm": 0.49018964171409607,
+      "learning_rate": 1.124612551624804e-05,
+      "loss": 5.9027,
+      "step": 2661
+    },
+    {
+      "epoch": 0.914658763904995,
+      "grad_norm": 0.4549656808376312,
+      "learning_rate": 1.115736471758158e-05,
+      "loss": 5.9732,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9150023622385431,
+      "grad_norm": 0.471829891204834,
+      "learning_rate": 1.1068947588066813e-05,
+      "loss": 6.0014,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9153459605720913,
+      "grad_norm": 0.49442002177238464,
+      "learning_rate": 1.09808742549265e-05,
+      "loss": 5.9958,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9156895589056393,
+      "grad_norm": 0.4691806733608246,
+      "learning_rate": 1.0893144844888791e-05,
+      "loss": 6.0186,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9160331572391874,
+      "grad_norm": 0.49828535318374634,
+      "learning_rate": 1.0805759484186994e-05,
+      "loss": 6.0637,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9163767555727355,
+      "grad_norm": 0.48614802956581116,
+      "learning_rate": 1.0718718298559389e-05,
+      "loss": 6.1132,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9167203539062836,
+      "grad_norm": 0.4641305208206177,
+      "learning_rate": 1.06320214132489e-05,
+      "loss": 6.0557,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9170639522398316,
+      "grad_norm": 0.4682670831680298,
+      "learning_rate": 1.054566895300324e-05,
+      "loss": 5.9688,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9174075505733797,
+      "grad_norm": 0.4661727547645569,
+      "learning_rate": 1.0459661042074326e-05,
+      "loss": 6.13,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9177511489069278,
+      "grad_norm": 0.4203856885433197,
+      "learning_rate": 1.0373997804218411e-05,
+      "loss": 6.0238,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9180947472404759,
+      "grad_norm": 0.4406087100505829,
+      "learning_rate": 1.0288679362695786e-05,
+      "loss": 6.0538,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9184383455740239,
+      "grad_norm": 0.5143980383872986,
+      "learning_rate": 1.0203705840270665e-05,
+      "loss": 6.1826,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9187819439075721,
+      "grad_norm": 0.44605106115341187,
+      "learning_rate": 1.0119077359210832e-05,
+      "loss": 5.9918,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9191255422411201,
+      "grad_norm": 0.43419891595840454,
+      "learning_rate": 1.0034794041287709e-05,
+      "loss": 6.0827,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9194691405746682,
+      "grad_norm": 0.4553866982460022,
+      "learning_rate": 9.950856007776011e-06,
+      "loss": 6.0581,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9198127389082162,
+      "grad_norm": 0.46902498602867126,
+      "learning_rate": 9.867263379453677e-06,
+      "loss": 6.0497,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9201563372417644,
+      "grad_norm": 0.4728301167488098,
+      "learning_rate": 9.78401627660161e-06,
+      "loss": 6.0556,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9204999355753125,
+      "grad_norm": 0.4771122634410858,
+      "learning_rate": 9.701114819003486e-06,
+      "loss": 5.9773,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9208435339088605,
+      "grad_norm": 0.470584511756897,
+      "learning_rate": 9.618559125945725e-06,
+      "loss": 6.0586,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9211871322424087,
+      "grad_norm": 0.4982456862926483,
+      "learning_rate": 9.536349316217163e-06,
+      "loss": 6.0214,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9215307305759567,
+      "grad_norm": 0.47169244289398193,
+      "learning_rate": 9.454485508109012e-06,
+      "loss": 6.1075,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9218743289095048,
+      "grad_norm": 0.4510200321674347,
+      "learning_rate": 9.372967819414547e-06,
+      "loss": 6.1195,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9222179272430528,
+      "grad_norm": 0.4280446171760559,
+      "learning_rate": 9.291796367429107e-06,
+      "loss": 5.9874,
+      "step": 2684
+    },
+    {
+      "epoch": 0.922561525576601,
+      "grad_norm": 0.5145493149757385,
+      "learning_rate": 9.21097126894968e-06,
+      "loss": 6.0329,
+      "step": 2685
+    },
+    {
+      "epoch": 0.922905123910149,
+      "grad_norm": 0.5257851481437683,
+      "learning_rate": 9.130492640275129e-06,
+      "loss": 6.0867,
+      "step": 2686
+    },
+    {
+      "epoch": 0.9232487222436971,
+      "grad_norm": 0.5830054879188538,
+      "learning_rate": 9.050360597205515e-06,
+      "loss": 5.9751,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9235923205772452,
+      "grad_norm": 0.652542769908905,
+      "learning_rate": 8.970575255042385e-06,
+      "loss": 5.9654,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9239359189107933,
+      "grad_norm": 0.5067824125289917,
+      "learning_rate": 8.891136728588323e-06,
+      "loss": 6.0476,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9242795172443413,
+      "grad_norm": 0.5378184914588928,
+      "learning_rate": 8.812045132147007e-06,
+      "loss": 5.9726,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9246231155778895,
+      "grad_norm": 0.5374602675437927,
+      "learning_rate": 8.733300579522707e-06,
+      "loss": 6.141,
+      "step": 2691
+    },
+    {
+      "epoch": 0.9249667139114375,
+      "grad_norm": 0.6461105942726135,
+      "learning_rate": 8.654903184020568e-06,
+      "loss": 5.9509,
+      "step": 2692
+    },
+    {
+      "epoch": 0.9253103122449856,
+      "grad_norm": 0.6112871766090393,
+      "learning_rate": 8.576853058446077e-06,
+      "loss": 5.9735,
+      "step": 2693
+    },
+    {
+      "epoch": 0.9256539105785337,
+      "grad_norm": 0.5804199576377869,
+      "learning_rate": 8.49915031510498e-06,
+      "loss": 6.1318,
+      "step": 2694
+    },
+    {
+      "epoch": 0.9259975089120818,
+      "grad_norm": 0.7698377966880798,
+      "learning_rate": 8.4217950658034e-06,
+      "loss": 6.1104,
+      "step": 2695
+    },
+    {
+      "epoch": 0.9263411072456299,
+      "grad_norm": 0.7038789391517639,
+      "learning_rate": 8.344787421847217e-06,
+      "loss": 6.1166,
+      "step": 2696
+    },
+    {
+      "epoch": 0.9266847055791779,
+      "grad_norm": 0.7374392151832581,
+      "learning_rate": 8.268127494042265e-06,
+      "loss": 6.1314,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9270283039127261,
+      "grad_norm": 0.7966888546943665,
+      "learning_rate": 8.191815392694035e-06,
+      "loss": 6.1515,
+      "step": 2698
+    },
+    {
+      "epoch": 0.9273719022462741,
+      "grad_norm": 0.9581894278526306,
+      "learning_rate": 8.115851227607552e-06,
+      "loss": 6.2336,
+      "step": 2699
+    },
+    {
+      "epoch": 0.9277155005798222,
+      "grad_norm": 1.1803251504898071,
+      "learning_rate": 8.040235108087075e-06,
+      "loss": 6.0879,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9280590989133702,
+      "grad_norm": 0.4471254050731659,
+      "learning_rate": 7.964967142936263e-06,
+      "loss": 6.0322,
+      "step": 2701
+    },
+    {
+      "epoch": 0.9284026972469184,
+      "grad_norm": 0.6071027517318726,
+      "learning_rate": 7.890047440457682e-06,
+      "loss": 5.865,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9287462955804664,
+      "grad_norm": 0.5928388833999634,
+      "learning_rate": 7.81547610845279e-06,
+      "loss": 6.0027,
+      "step": 2703
+    },
+    {
+      "epoch": 0.9290898939140145,
+      "grad_norm": 0.5783106088638306,
+      "learning_rate": 7.741253254221841e-06,
+      "loss": 5.8402,
+      "step": 2704
+    },
+    {
+      "epoch": 0.9294334922475626,
+      "grad_norm": 0.508571207523346,
+      "learning_rate": 7.6673789845636e-06,
+      "loss": 5.8952,
+      "step": 2705
+    },
+    {
+      "epoch": 0.9297770905811107,
+      "grad_norm": 0.585649311542511,
+      "learning_rate": 7.593853405775286e-06,
+      "loss": 5.9212,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9301206889146587,
+      "grad_norm": 0.5307704210281372,
+      "learning_rate": 7.520676623652411e-06,
+      "loss": 5.8958,
+      "step": 2707
+    },
+    {
+      "epoch": 0.9304642872482068,
+      "grad_norm": 0.48460954427719116,
+      "learning_rate": 7.4478487434885554e-06,
+      "loss": 6.0424,
+      "step": 2708
+    },
+    {
+      "epoch": 0.930807885581755,
+      "grad_norm": 0.45228397846221924,
+      "learning_rate": 7.3753698700753105e-06,
+      "loss": 6.0324,
+      "step": 2709
+    },
+    {
+      "epoch": 0.931151483915303,
+      "grad_norm": 0.5303856730461121,
+      "learning_rate": 7.303240107702086e-06,
+      "loss": 6.0299,
+      "step": 2710
+    },
+    {
+      "epoch": 0.9314950822488511,
+      "grad_norm": 0.46378111839294434,
+      "learning_rate": 7.2314595601558346e-06,
+      "loss": 5.9614,
+      "step": 2711
+    },
+    {
+      "epoch": 0.9318386805823992,
+      "grad_norm": 0.5033019185066223,
+      "learning_rate": 7.160028330721297e-06,
+      "loss": 6.0099,
+      "step": 2712
+    },
+    {
+      "epoch": 0.9321822789159473,
+      "grad_norm": 0.5241475701332092,
+      "learning_rate": 7.088946522180284e-06,
+      "loss": 6.0135,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9325258772494953,
+      "grad_norm": 0.44836217164993286,
+      "learning_rate": 7.018214236812009e-06,
+      "loss": 5.9512,
+      "step": 2714
+    },
+    {
+      "epoch": 0.9328694755830435,
+      "grad_norm": 0.40336859226226807,
+      "learning_rate": 6.947831576392727e-06,
+      "loss": 5.9761,
+      "step": 2715
+    },
+    {
+      "epoch": 0.9332130739165915,
+      "grad_norm": 0.38975638151168823,
+      "learning_rate": 6.877798642195565e-06,
+      "loss": 5.9835,
+      "step": 2716
+    },
+    {
+      "epoch": 0.9335566722501396,
+      "grad_norm": 0.5378215312957764,
+      "learning_rate": 6.808115534990445e-06,
+      "loss": 6.0024,
+      "step": 2717
+    },
+    {
+      "epoch": 0.9339002705836876,
+      "grad_norm": 0.4532417953014374,
+      "learning_rate": 6.7387823550440485e-06,
+      "loss": 6.0132,
+      "step": 2718
+    },
+    {
+      "epoch": 0.9342438689172358,
+      "grad_norm": 0.4508761763572693,
+      "learning_rate": 6.669799202119353e-06,
+      "loss": 6.1408,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9345874672507838,
+      "grad_norm": 0.48313215374946594,
+      "learning_rate": 6.601166175475792e-06,
+      "loss": 5.9936,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9349310655843319,
+      "grad_norm": 0.520293653011322,
+      "learning_rate": 6.532883373869009e-06,
+      "loss": 6.0829,
+      "step": 2721
+    },
+    {
+      "epoch": 0.93527466391788,
+      "grad_norm": 0.45849609375,
+      "learning_rate": 6.464950895550742e-06,
+      "loss": 5.9805,
+      "step": 2722
+    },
+    {
+      "epoch": 0.9356182622514281,
+      "grad_norm": 0.5732299089431763,
+      "learning_rate": 6.3973688382684965e-06,
+      "loss": 6.0153,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9359618605849762,
+      "grad_norm": 0.4997122585773468,
+      "learning_rate": 6.330137299265737e-06,
+      "loss": 5.9576,
+      "step": 2724
+    },
+    {
+      "epoch": 0.9363054589185242,
+      "grad_norm": 0.4114673435688019,
+      "learning_rate": 6.263256375281523e-06,
+      "loss": 6.0066,
+      "step": 2725
+    },
+    {
+      "epoch": 0.9366490572520724,
+      "grad_norm": 0.4480920135974884,
+      "learning_rate": 6.196726162550292e-06,
+      "loss": 6.0656,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9369926555856204,
+      "grad_norm": 0.4622270166873932,
+      "learning_rate": 6.130546756802053e-06,
+      "loss": 5.9519,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9373362539191685,
+      "grad_norm": 0.44009390473365784,
+      "learning_rate": 6.064718253261852e-06,
+      "loss": 6.0834,
+      "step": 2728
+    },
+    {
+      "epoch": 0.9376798522527166,
+      "grad_norm": 0.5133486986160278,
+      "learning_rate": 5.999240746649953e-06,
+      "loss": 6.1036,
+      "step": 2729
+    },
+    {
+      "epoch": 0.9380234505862647,
+      "grad_norm": 0.4361947774887085,
+      "learning_rate": 5.93411433118149e-06,
+      "loss": 6.0289,
+      "step": 2730
+    },
+    {
+      "epoch": 0.9383670489198127,
+      "grad_norm": 0.503173291683197,
+      "learning_rate": 5.8693391005665005e-06,
+      "loss": 5.9717,
+      "step": 2731
+    },
+    {
+      "epoch": 0.9387106472533608,
+      "grad_norm": 0.5202142000198364,
+      "learning_rate": 5.804915148009571e-06,
+      "loss": 6.0946,
+      "step": 2732
+    },
+    {
+      "epoch": 0.9390542455869089,
+      "grad_norm": 0.45355695486068726,
+      "learning_rate": 5.740842566210019e-06,
+      "loss": 6.0567,
+      "step": 2733
+    },
+    {
+      "epoch": 0.939397843920457,
+      "grad_norm": 0.47078660130500793,
+      "learning_rate": 5.677121447361405e-06,
+      "loss": 6.0976,
+      "step": 2734
+    },
+    {
+      "epoch": 0.939741442254005,
+      "grad_norm": 0.57381671667099,
+      "learning_rate": 5.613751883151663e-06,
+      "loss": 6.1569,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9400850405875532,
+      "grad_norm": 0.5258800983428955,
+      "learning_rate": 5.550733964762855e-06,
+      "loss": 6.0503,
+      "step": 2736
+    },
+    {
+      "epoch": 0.9404286389211012,
+      "grad_norm": 0.5274262428283691,
+      "learning_rate": 5.488067782871059e-06,
+      "loss": 5.9954,
+      "step": 2737
+    },
+    {
+      "epoch": 0.9407722372546493,
+      "grad_norm": 0.55703204870224,
+      "learning_rate": 5.425753427646257e-06,
+      "loss": 6.0501,
+      "step": 2738
+    },
+    {
+      "epoch": 0.9411158355881974,
+      "grad_norm": 0.6053146123886108,
+      "learning_rate": 5.3637909887521705e-06,
+      "loss": 6.0434,
+      "step": 2739
+    },
+    {
+      "epoch": 0.9414594339217455,
+      "grad_norm": 0.5558188557624817,
+      "learning_rate": 5.302180555346175e-06,
+      "loss": 5.9769,
+      "step": 2740
+    },
+    {
+      "epoch": 0.9418030322552936,
+      "grad_norm": 0.6423223614692688,
+      "learning_rate": 5.24092221607908e-06,
+      "loss": 6.0606,
+      "step": 2741
+    },
+    {
+      "epoch": 0.9421466305888416,
+      "grad_norm": 0.7581862211227417,
+      "learning_rate": 5.18001605909521e-06,
+      "loss": 6.0086,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9424902289223898,
+      "grad_norm": 0.7525710463523865,
+      "learning_rate": 5.119462172031963e-06,
+      "loss": 6.0993,
+      "step": 2743
+    },
+    {
+      "epoch": 0.9428338272559378,
+      "grad_norm": 0.6466197371482849,
+      "learning_rate": 5.059260642020003e-06,
+      "loss": 6.1109,
+      "step": 2744
+    },
+    {
+      "epoch": 0.9431774255894859,
+      "grad_norm": 0.6755435466766357,
+      "learning_rate": 4.99941155568287e-06,
+      "loss": 6.1111,
+      "step": 2745
+    },
+    {
+      "epoch": 0.9435210239230339,
+      "grad_norm": 0.6645384430885315,
+      "learning_rate": 4.939914999137096e-06,
+      "loss": 6.0707,
+      "step": 2746
+    },
+    {
+      "epoch": 0.9438646222565821,
+      "grad_norm": 0.770700216293335,
+      "learning_rate": 4.8807710579918394e-06,
+      "loss": 6.1119,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9442082205901301,
+      "grad_norm": 0.8000778555870056,
+      "learning_rate": 4.8219798173490255e-06,
+      "loss": 5.9846,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9445518189236782,
+      "grad_norm": 0.9796259999275208,
+      "learning_rate": 4.763541361802875e-06,
+      "loss": 6.3486,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9448954172572263,
+      "grad_norm": 1.1820958852767944,
+      "learning_rate": 4.705455775440237e-06,
+      "loss": 6.272,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9452390155907744,
+      "grad_norm": 0.48769789934158325,
+      "learning_rate": 4.647723141840033e-06,
+      "loss": 5.9491,
+      "step": 2751
+    },
+    {
+      "epoch": 0.9455826139243224,
+      "grad_norm": 0.5634806156158447,
+      "learning_rate": 4.590343544073367e-06,
+      "loss": 6.0842,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9459262122578705,
+      "grad_norm": 0.5970801711082458,
+      "learning_rate": 4.533317064703391e-06,
+      "loss": 5.836,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9462698105914187,
+      "grad_norm": 0.4978366792201996,
+      "learning_rate": 4.476643785785162e-06,
+      "loss": 5.9155,
+      "step": 2754
+    },
+    {
+      "epoch": 0.9466134089249667,
+      "grad_norm": 0.4581436216831207,
+      "learning_rate": 4.420323788865476e-06,
+      "loss": 5.9773,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9469570072585148,
+      "grad_norm": 0.630322277545929,
+      "learning_rate": 4.364357154982846e-06,
+      "loss": 5.8643,
+      "step": 2756
+    },
+    {
+      "epoch": 0.9473006055920629,
+      "grad_norm": 0.6597774028778076,
+      "learning_rate": 4.308743964667294e-06,
+      "loss": 6.0093,
+      "step": 2757
+    },
+    {
+      "epoch": 0.947644203925611,
+      "grad_norm": 0.5160709619522095,
+      "learning_rate": 4.2534842979402575e-06,
+      "loss": 5.8845,
+      "step": 2758
+    },
+    {
+      "epoch": 0.947987802259159,
+      "grad_norm": 0.47501340508461,
+      "learning_rate": 4.198578234314604e-06,
+      "loss": 5.9749,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9483314005927072,
+      "grad_norm": 0.4054076075553894,
+      "learning_rate": 4.14402585279422e-06,
+      "loss": 6.036,
+      "step": 2760
+    },
+    {
+      "epoch": 0.9486749989262552,
+      "grad_norm": 0.487505704164505,
+      "learning_rate": 4.0898272318742324e-06,
+      "loss": 5.9956,
+      "step": 2761
+    },
+    {
+      "epoch": 0.9490185972598033,
+      "grad_norm": 0.4308890402317047,
+      "learning_rate": 4.035982449540676e-06,
+      "loss": 5.9623,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9493621955933513,
+      "grad_norm": 0.44951295852661133,
+      "learning_rate": 3.982491583270492e-06,
+      "loss": 5.9991,
+      "step": 2763
+    },
+    {
+      "epoch": 0.9497057939268995,
+      "grad_norm": 0.45850709080696106,
+      "learning_rate": 3.9293547100313075e-06,
+      "loss": 6.0219,
+      "step": 2764
+    },
+    {
+      "epoch": 0.9500493922604475,
+      "grad_norm": 0.4028855860233307,
+      "learning_rate": 3.87657190628149e-06,
+      "loss": 6.0092,
+      "step": 2765
+    },
+    {
+      "epoch": 0.9503929905939956,
+      "grad_norm": 0.3813883662223816,
+      "learning_rate": 3.824143247969813e-06,
+      "loss": 6.1796,
+      "step": 2766
+    },
+    {
+      "epoch": 0.9507365889275436,
+      "grad_norm": 0.4750194847583771,
+      "learning_rate": 3.7720688105356005e-06,
+      "loss": 6.0574,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9510801872610918,
+      "grad_norm": 0.47151845693588257,
+      "learning_rate": 3.7203486689083854e-06,
+      "loss": 6.0058,
+      "step": 2768
+    },
+    {
+      "epoch": 0.9514237855946399,
+      "grad_norm": 0.41912469267845154,
+      "learning_rate": 3.668982897507972e-06,
+      "loss": 5.9304,
+      "step": 2769
+    },
+    {
+      "epoch": 0.9517673839281879,
+      "grad_norm": 0.4504321217536926,
+      "learning_rate": 3.617971570244266e-06,
+      "loss": 5.9236,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9521109822617361,
+      "grad_norm": 0.49030429124832153,
+      "learning_rate": 3.5673147605170817e-06,
+      "loss": 6.0394,
+      "step": 2771
+    },
+    {
+      "epoch": 0.9524545805952841,
+      "grad_norm": 0.4937245845794678,
+      "learning_rate": 3.5170125412162247e-06,
+      "loss": 5.9422,
+      "step": 2772
+    },
+    {
+      "epoch": 0.9527981789288322,
+      "grad_norm": 0.4641237258911133,
+      "learning_rate": 3.467064984721241e-06,
+      "loss": 6.0364,
+      "step": 2773
+    },
+    {
+      "epoch": 0.9531417772623803,
+      "grad_norm": 0.44841110706329346,
+      "learning_rate": 3.417472162901336e-06,
+      "loss": 5.9775,
+      "step": 2774
+    },
+    {
+      "epoch": 0.9534853755959284,
+      "grad_norm": 0.4379451870918274,
+      "learning_rate": 3.36823414711529e-06,
+      "loss": 5.992,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9538289739294764,
+      "grad_norm": 0.46499103307724,
+      "learning_rate": 3.3193510082114297e-06,
+      "loss": 5.9937,
+      "step": 2776
+    },
+    {
+      "epoch": 0.9541725722630245,
+      "grad_norm": 0.49435579776763916,
+      "learning_rate": 3.2708228165273244e-06,
+      "loss": 6.0961,
+      "step": 2777
+    },
+    {
+      "epoch": 0.9545161705965726,
+      "grad_norm": 0.47890639305114746,
+      "learning_rate": 3.2226496418899244e-06,
+      "loss": 6.0472,
+      "step": 2778
+    },
+    {
+      "epoch": 0.9548597689301207,
+      "grad_norm": 0.5395154356956482,
+      "learning_rate": 3.1748315536153094e-06,
+      "loss": 5.9907,
+      "step": 2779
+    },
+    {
+      "epoch": 0.9552033672636687,
+      "grad_norm": 0.45388686656951904,
+      "learning_rate": 3.127368620508608e-06,
+      "loss": 5.9788,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9555469655972169,
+      "grad_norm": 0.44846445322036743,
+      "learning_rate": 3.0802609108638858e-06,
+      "loss": 6.0075,
+      "step": 2781
+    },
+    {
+      "epoch": 0.9558905639307649,
+      "grad_norm": 0.5502305626869202,
+      "learning_rate": 3.0335084924642263e-06,
+      "loss": 6.0372,
+      "step": 2782
+    },
+    {
+      "epoch": 0.956234162264313,
+      "grad_norm": 0.4417821168899536,
+      "learning_rate": 2.987111432581291e-06,
+      "loss": 5.9533,
+      "step": 2783
+    },
+    {
+      "epoch": 0.9565777605978611,
+      "grad_norm": 0.5430399775505066,
+      "learning_rate": 2.9410697979755928e-06,
+      "loss": 6.0111,
+      "step": 2784
+    },
+    {
+      "epoch": 0.9569213589314092,
+      "grad_norm": 0.5501893162727356,
+      "learning_rate": 2.8953836548960834e-06,
+      "loss": 5.9851,
+      "step": 2785
+    },
+    {
+      "epoch": 0.9572649572649573,
+      "grad_norm": 0.6305771470069885,
+      "learning_rate": 2.850053069080344e-06,
+      "loss": 6.0965,
+      "step": 2786
+    },
+    {
+      "epoch": 0.9576085555985053,
+      "grad_norm": 0.6068539023399353,
+      "learning_rate": 2.805078105754172e-06,
+      "loss": 6.0559,
+      "step": 2787
+    },
+    {
+      "epoch": 0.9579521539320535,
+      "grad_norm": 0.5408775806427002,
+      "learning_rate": 2.760458829631801e-06,
+      "loss": 5.9709,
+      "step": 2788
+    },
+    {
+      "epoch": 0.9582957522656015,
+      "grad_norm": 0.5502400398254395,
+      "learning_rate": 2.716195304915653e-06,
+      "loss": 6.0905,
+      "step": 2789
+    },
+    {
+      "epoch": 0.9586393505991496,
+      "grad_norm": 0.534099817276001,
+      "learning_rate": 2.672287595296169e-06,
+      "loss": 6.0971,
+      "step": 2790
+    },
+    {
+      "epoch": 0.9589829489326976,
+      "grad_norm": 0.5350020527839661,
+      "learning_rate": 2.6287357639519504e-06,
+      "loss": 6.1302,
+      "step": 2791
+    },
+    {
+      "epoch": 0.9593265472662458,
+      "grad_norm": 0.6089175939559937,
+      "learning_rate": 2.5855398735493697e-06,
+      "loss": 6.0019,
+      "step": 2792
+    },
+    {
+      "epoch": 0.9596701455997938,
+      "grad_norm": 0.6447976231575012,
+      "learning_rate": 2.5426999862427914e-06,
+      "loss": 5.9295,
+      "step": 2793
+    },
+    {
+      "epoch": 0.9600137439333419,
+      "grad_norm": 0.5977621674537659,
+      "learning_rate": 2.5002161636742125e-06,
+      "loss": 6.0715,
+      "step": 2794
+    },
+    {
+      "epoch": 0.96035734226689,
+      "grad_norm": 0.7416554689407349,
+      "learning_rate": 2.458088466973346e-06,
+      "loss": 6.1905,
+      "step": 2795
+    },
+    {
+      "epoch": 0.9607009406004381,
+      "grad_norm": 0.7069945931434631,
+      "learning_rate": 2.4163169567574526e-06,
+      "loss": 6.1577,
+      "step": 2796
+    },
+    {
+      "epoch": 0.9610445389339862,
+      "grad_norm": 0.8261065483093262,
+      "learning_rate": 2.3749016931313426e-06,
+      "loss": 6.1972,
+      "step": 2797
+    },
+    {
+      "epoch": 0.9613881372675342,
+      "grad_norm": 0.9036763310432434,
+      "learning_rate": 2.333842735687097e-06,
+      "loss": 6.0924,
+      "step": 2798
+    },
+    {
+      "epoch": 0.9617317356010824,
+      "grad_norm": 0.9265655279159546,
+      "learning_rate": 2.293140143504291e-06,
+      "loss": 6.1676,
+      "step": 2799
+    },
+    {
+      "epoch": 0.9620753339346304,
+      "grad_norm": 1.5305782556533813,
+      "learning_rate": 2.2527939751495476e-06,
+      "loss": 6.1288,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9624189322681785,
+      "grad_norm": 0.5339880585670471,
+      "learning_rate": 2.212804288676706e-06,
+      "loss": 5.9065,
+      "step": 2801
+    },
+    {
+      "epoch": 0.9627625306017266,
+      "grad_norm": 0.5260118842124939,
+      "learning_rate": 2.1731711416267396e-06,
+      "loss": 5.9645,
+      "step": 2802
+    },
+    {
+      "epoch": 0.9631061289352747,
+      "grad_norm": 0.587617814540863,
+      "learning_rate": 2.1338945910274743e-06,
+      "loss": 5.8985,
+      "step": 2803
+    },
+    {
+      "epoch": 0.9634497272688227,
+      "grad_norm": 0.5828947424888611,
+      "learning_rate": 2.094974693393731e-06,
+      "loss": 5.9318,
+      "step": 2804
+    },
+    {
+      "epoch": 0.9637933256023709,
+      "grad_norm": 0.4665927290916443,
+      "learning_rate": 2.0564115047270458e-06,
+      "loss": 6.1154,
+      "step": 2805
+    },
+    {
+      "epoch": 0.9641369239359189,
+      "grad_norm": 0.5141355991363525,
+      "learning_rate": 2.0182050805158115e-06,
+      "loss": 5.991,
+      "step": 2806
+    },
+    {
+      "epoch": 0.964480522269467,
+      "grad_norm": 0.5290352702140808,
+      "learning_rate": 1.9803554757349685e-06,
+      "loss": 5.9607,
+      "step": 2807
+    },
+    {
+      "epoch": 0.964824120603015,
+      "grad_norm": 0.48139309883117676,
+      "learning_rate": 1.942862744846091e-06,
+      "loss": 5.9097,
+      "step": 2808
+    },
+    {
+      "epoch": 0.9651677189365632,
+      "grad_norm": 0.6123693585395813,
+      "learning_rate": 1.90572694179722e-06,
+      "loss": 5.9149,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9655113172701112,
+      "grad_norm": 0.4561925530433655,
+      "learning_rate": 1.8689481200228064e-06,
+      "loss": 5.8467,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9658549156036593,
+      "grad_norm": 0.526841402053833,
+      "learning_rate": 1.8325263324437125e-06,
+      "loss": 6.0141,
+      "step": 2811
+    },
+    {
+      "epoch": 0.9661985139372075,
+      "grad_norm": 0.49907776713371277,
+      "learning_rate": 1.796461631466989e-06,
+      "loss": 6.0447,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9665421122707555,
+      "grad_norm": 0.6142306923866272,
+      "learning_rate": 1.7607540689859035e-06,
+      "loss": 5.8905,
+      "step": 2813
+    },
+    {
+      "epoch": 0.9668857106043036,
+      "grad_norm": 0.45012393593788147,
+      "learning_rate": 1.7254036963798569e-06,
+      "loss": 6.0046,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9672293089378516,
+      "grad_norm": 0.4007306396961212,
+      "learning_rate": 1.6904105645142442e-06,
+      "loss": 6.0132,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9675729072713998,
+      "grad_norm": 0.3214866816997528,
+      "learning_rate": 1.6557747237405107e-06,
+      "loss": 6.032,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9679165056049478,
+      "grad_norm": 0.4845171868801117,
+      "learning_rate": 1.621496223895902e-06,
+      "loss": 5.9634,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9682601039384959,
+      "grad_norm": 0.4626264274120331,
+      "learning_rate": 1.5875751143035465e-06,
+      "loss": 5.8288,
+      "step": 2818
+    },
+    {
+      "epoch": 0.968603702272044,
+      "grad_norm": 0.42864149808883667,
+      "learning_rate": 1.5540114437723185e-06,
+      "loss": 6.0096,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9689473006055921,
+      "grad_norm": 0.3970453143119812,
+      "learning_rate": 1.5208052605967804e-06,
+      "loss": 5.9898,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9692908989391401,
+      "grad_norm": 0.3872493505477905,
+      "learning_rate": 1.4879566125570732e-06,
+      "loss": 6.0657,
+      "step": 2821
+    },
+    {
+      "epoch": 0.9696344972726882,
+      "grad_norm": 0.4018920660018921,
+      "learning_rate": 1.4554655469189438e-06,
+      "loss": 5.975,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9699780956062363,
+      "grad_norm": 0.45553165674209595,
+      "learning_rate": 1.4233321104335506e-06,
+      "loss": 6.0283,
+      "step": 2823
+    },
+    {
+      "epoch": 0.9703216939397844,
+      "grad_norm": 0.46554648876190186,
+      "learning_rate": 1.391556349337464e-06,
+      "loss": 6.0846,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9706652922733324,
+      "grad_norm": 0.4684000313282013,
+      "learning_rate": 1.3601383093526931e-06,
+      "loss": 6.0688,
+      "step": 2825
+    },
+    {
+      "epoch": 0.9710088906068806,
+      "grad_norm": 0.4713064134120941,
+      "learning_rate": 1.3290780356864374e-06,
+      "loss": 6.0044,
+      "step": 2826
+    },
+    {
+      "epoch": 0.9713524889404287,
+      "grad_norm": 0.4442633092403412,
+      "learning_rate": 1.2983755730310854e-06,
+      "loss": 6.0409,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9716960872739767,
+      "grad_norm": 0.4553023874759674,
+      "learning_rate": 1.2680309655642431e-06,
+      "loss": 6.0703,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9720396856075249,
+      "grad_norm": 0.41858962178230286,
+      "learning_rate": 1.238044256948595e-06,
+      "loss": 6.0488,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9723832839410729,
+      "grad_norm": 0.4985685348510742,
+      "learning_rate": 1.2084154903317934e-06,
+      "loss": 6.0478,
+      "step": 2830
+    },
+    {
+      "epoch": 0.972726882274621,
+      "grad_norm": 0.5059686899185181,
+      "learning_rate": 1.1791447083465134e-06,
+      "loss": 5.9557,
+      "step": 2831
+    },
+    {
+      "epoch": 0.973070480608169,
+      "grad_norm": 0.5063678026199341,
+      "learning_rate": 1.150231953110259e-06,
+      "loss": 6.1021,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9734140789417172,
+      "grad_norm": 0.4942188858985901,
+      "learning_rate": 1.1216772662254182e-06,
+      "loss": 6.0075,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9737576772752652,
+      "grad_norm": 0.5138784646987915,
+      "learning_rate": 1.0934806887791803e-06,
+      "loss": 5.968,
+      "step": 2834
+    },
+    {
+      "epoch": 0.9741012756088133,
+      "grad_norm": 0.4592101275920868,
+      "learning_rate": 1.065642261343397e-06,
+      "loss": 6.1217,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9744448739423613,
+      "grad_norm": 0.5657181739807129,
+      "learning_rate": 1.0381620239746093e-06,
+      "loss": 5.928,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9747884722759095,
+      "grad_norm": 0.5013971924781799,
+      "learning_rate": 1.0110400162139377e-06,
+      "loss": 5.9431,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9751320706094575,
+      "grad_norm": 0.5016094446182251,
+      "learning_rate": 9.842762770871094e-07,
+      "loss": 6.0844,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9754756689430056,
+      "grad_norm": 0.5339999794960022,
+      "learning_rate": 9.57870845104264e-07,
+      "loss": 5.9993,
+      "step": 2839
+    },
+    {
+      "epoch": 0.9758192672765537,
+      "grad_norm": 0.5596013069152832,
+      "learning_rate": 9.318237582600086e-07,
+      "loss": 6.081,
+      "step": 2840
+    },
+    {
+      "epoch": 0.9761628656101018,
+      "grad_norm": 0.6841441988945007,
+      "learning_rate": 9.061350540333635e-07,
+      "loss": 5.903,
+      "step": 2841
+    },
+    {
+      "epoch": 0.9765064639436499,
+      "grad_norm": 0.6830129623413086,
+      "learning_rate": 8.80804769387622e-07,
+      "loss": 6.1143,
+      "step": 2842
+    },
+    {
+      "epoch": 0.976850062277198,
+      "grad_norm": 0.6157656311988831,
+      "learning_rate": 8.558329407703514e-07,
+      "loss": 6.0526,
+      "step": 2843
+    },
+    {
+      "epoch": 0.9771936606107461,
+      "grad_norm": 0.6276822686195374,
+      "learning_rate": 8.312196041133923e-07,
+      "loss": 6.06,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9775372589442941,
+      "grad_norm": 0.6318298578262329,
+      "learning_rate": 8.069647948326653e-07,
+      "loss": 6.0182,
+      "step": 2845
+    },
+    {
+      "epoch": 0.9778808572778422,
+      "grad_norm": 0.73749178647995,
+      "learning_rate": 7.830685478283362e-07,
+      "loss": 6.1867,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9782244556113903,
+      "grad_norm": 0.8540349006652832,
+      "learning_rate": 7.595308974845117e-07,
+      "loss": 6.0417,
+      "step": 2847
+    },
+    {
+      "epoch": 0.9785680539449384,
+      "grad_norm": 0.9874235987663269,
+      "learning_rate": 7.363518776694056e-07,
+      "loss": 6.0435,
+      "step": 2848
+    },
+    {
+      "epoch": 0.9789116522784864,
+      "grad_norm": 0.8798337578773499,
+      "learning_rate": 7.135315217350891e-07,
+      "loss": 6.1899,
+      "step": 2849
+    },
+    {
+      "epoch": 0.9792552506120346,
+      "grad_norm": 1.329888939857483,
+      "learning_rate": 6.910698625177126e-07,
+      "loss": 6.2289,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9795988489455826,
+      "grad_norm": 0.4526410400867462,
+      "learning_rate": 6.689669323371728e-07,
+      "loss": 6.0513,
+      "step": 2851
+    },
+    {
+      "epoch": 0.9799424472791307,
+      "grad_norm": 0.4708828628063202,
+      "learning_rate": 6.472227629972239e-07,
+      "loss": 6.0642,
+      "step": 2852
+    },
+    {
+      "epoch": 0.9802860456126787,
+      "grad_norm": 0.3837355077266693,
+      "learning_rate": 6.258373857854494e-07,
+      "loss": 5.9422,
+      "step": 2853
+    },
+    {
+      "epoch": 0.9806296439462269,
+      "grad_norm": 0.47510474920272827,
+      "learning_rate": 6.04810831473096e-07,
+      "loss": 5.9057,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9809732422797749,
+      "grad_norm": 0.42741143703460693,
+      "learning_rate": 5.841431303151845e-07,
+      "loss": 5.9465,
+      "step": 2855
+    },
+    {
+      "epoch": 0.981316840613323,
+      "grad_norm": 0.5458832383155823,
+      "learning_rate": 5.638343120502598e-07,
+      "loss": 6.0429,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9816604389468712,
+      "grad_norm": 0.46792909502983093,
+      "learning_rate": 5.438844059006409e-07,
+      "loss": 6.0272,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9820040372804192,
+      "grad_norm": 0.6238328218460083,
+      "learning_rate": 5.242934405720879e-07,
+      "loss": 5.818,
+      "step": 2858
+    },
+    {
+      "epoch": 0.9823476356139673,
+      "grad_norm": 0.5959128737449646,
+      "learning_rate": 5.050614442538848e-07,
+      "loss": 5.7514,
+      "step": 2859
+    },
+    {
+      "epoch": 0.9826912339475153,
+      "grad_norm": 0.45926153659820557,
+      "learning_rate": 4.86188444618868e-07,
+      "loss": 5.8914,
+      "step": 2860
+    },
+    {
+      "epoch": 0.9830348322810635,
+      "grad_norm": 0.39194220304489136,
+      "learning_rate": 4.6767446882328703e-07,
+      "loss": 6.0244,
+      "step": 2861
+    },
+    {
+      "epoch": 0.9833784306146115,
+      "grad_norm": 0.4128912389278412,
+      "learning_rate": 4.495195435067212e-07,
+      "loss": 6.0298,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9837220289481596,
+      "grad_norm": 0.5608128905296326,
+      "learning_rate": 4.317236947922465e-07,
+      "loss": 5.8793,
+      "step": 2863
+    },
+    {
+      "epoch": 0.9840656272817077,
+      "grad_norm": 0.42504721879959106,
+      "learning_rate": 4.142869482861578e-07,
+      "loss": 5.8194,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9844092256152558,
+      "grad_norm": 0.33117401599884033,
+      "learning_rate": 3.972093290781076e-07,
+      "loss": 6.0172,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9847528239488038,
+      "grad_norm": 0.4305660128593445,
+      "learning_rate": 3.8049086174093973e-07,
+      "loss": 6.0165,
+      "step": 2866
+    },
+    {
+      "epoch": 0.985096422282352,
+      "grad_norm": 0.37048450112342834,
+      "learning_rate": 3.641315703307724e-07,
+      "loss": 5.973,
+      "step": 2867
+    },
+    {
+      "epoch": 0.9854400206159,
+      "grad_norm": 0.4533330500125885,
+      "learning_rate": 3.481314783868594e-07,
+      "loss": 6.0333,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9857836189494481,
+      "grad_norm": 0.3933621048927307,
+      "learning_rate": 3.324906089316737e-07,
+      "loss": 6.0991,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9861272172829961,
+      "grad_norm": 0.4046175479888916,
+      "learning_rate": 3.172089844707404e-07,
+      "loss": 6.0109,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9864708156165443,
+      "grad_norm": 0.42547526955604553,
+      "learning_rate": 3.0228662699266494e-07,
+      "loss": 5.8854,
+      "step": 2871
+    },
+    {
+      "epoch": 0.9868144139500924,
+      "grad_norm": 0.4869686961174011,
+      "learning_rate": 2.8772355796918836e-07,
+      "loss": 5.8482,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9871580122836404,
+      "grad_norm": 0.5427522659301758,
+      "learning_rate": 2.7351979835496534e-07,
+      "loss": 6.1293,
+      "step": 2873
+    },
+    {
+      "epoch": 0.9875016106171886,
+      "grad_norm": 0.408896267414093,
+      "learning_rate": 2.596753685877584e-07,
+      "loss": 6.0766,
+      "step": 2874
+    },
+    {
+      "epoch": 0.9878452089507366,
+      "grad_norm": 0.45009753108024597,
+      "learning_rate": 2.461902885881606e-07,
+      "loss": 6.1401,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9881888072842847,
+      "grad_norm": 0.45075657963752747,
+      "learning_rate": 2.3306457775981728e-07,
+      "loss": 5.9555,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9885324056178327,
+      "grad_norm": 0.4548248052597046,
+      "learning_rate": 2.202982549892041e-07,
+      "loss": 6.0435,
+      "step": 2877
+    },
+    {
+      "epoch": 0.9888760039513809,
+      "grad_norm": 0.44134655594825745,
+      "learning_rate": 2.0789133864571042e-07,
+      "loss": 5.957,
+      "step": 2878
+    },
+    {
+      "epoch": 0.9892196022849289,
+      "grad_norm": 0.44243818521499634,
+      "learning_rate": 1.9584384658158371e-07,
+      "loss": 6.0225,
+      "step": 2879
+    },
+    {
+      "epoch": 0.989563200618477,
+      "grad_norm": 0.5458806157112122,
+      "learning_rate": 1.841557961318463e-07,
+      "loss": 5.9614,
+      "step": 2880
+    },
+    {
+      "epoch": 0.989906798952025,
+      "grad_norm": 0.4366854131221771,
+      "learning_rate": 1.7282720411437858e-07,
+      "loss": 6.0524,
+      "step": 2881
+    },
+    {
+      "epoch": 0.9902503972855732,
+      "grad_norm": 0.4618584215641022,
+      "learning_rate": 1.6185808682986358e-07,
+      "loss": 6.0012,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9905939956191212,
+      "grad_norm": 0.49428001046180725,
+      "learning_rate": 1.512484600616204e-07,
+      "loss": 6.0846,
+      "step": 2883
+    },
+    {
+      "epoch": 0.9909375939526693,
+      "grad_norm": 0.5537979602813721,
+      "learning_rate": 1.4099833907582627e-07,
+      "loss": 5.92,
+      "step": 2884
+    },
+    {
+      "epoch": 0.9912811922862174,
+      "grad_norm": 0.45787203311920166,
+      "learning_rate": 1.3110773862126668e-07,
+      "loss": 6.1011,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9916247906197655,
+      "grad_norm": 0.48102542757987976,
+      "learning_rate": 1.2157667292952978e-07,
+      "loss": 5.9676,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9919683889533136,
+      "grad_norm": 0.518682062625885,
+      "learning_rate": 1.1240515571475651e-07,
+      "loss": 6.0037,
+      "step": 2887
+    },
+    {
+      "epoch": 0.9923119872868617,
+      "grad_norm": 0.5656102299690247,
+      "learning_rate": 1.0359320017377937e-07,
+      "loss": 5.9474,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9926555856204098,
+      "grad_norm": 0.4793003797531128,
+      "learning_rate": 9.514081898612247e-08,
+      "loss": 6.1492,
+      "step": 2889
+    },
+    {
+      "epoch": 0.9929991839539578,
+      "grad_norm": 0.6143734455108643,
+      "learning_rate": 8.704802431377945e-08,
+      "loss": 5.9602,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9933427822875059,
+      "grad_norm": 0.5799821019172668,
+      "learning_rate": 7.931482780149102e-08,
+      "loss": 6.0795,
+      "step": 2891
+    },
+    {
+      "epoch": 0.993686380621054,
+      "grad_norm": 0.6480079889297485,
+      "learning_rate": 7.194124057649521e-08,
+      "loss": 6.1525,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9940299789546021,
+      "grad_norm": 0.5958256125450134,
+      "learning_rate": 6.492727324855508e-08,
+      "loss": 6.0316,
+      "step": 2893
+    },
+    {
+      "epoch": 0.9943735772881501,
+      "grad_norm": 0.7720398902893066,
+      "learning_rate": 5.8272935910069765e-08,
+      "loss": 6.2484,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9947171756216983,
+      "grad_norm": 0.5963339805603027,
+      "learning_rate": 5.1978238135907965e-08,
+      "loss": 6.1121,
+      "step": 2895
+    },
+    {
+      "epoch": 0.9950607739552463,
+      "grad_norm": 0.8526974320411682,
+      "learning_rate": 4.604318898346338e-08,
+      "loss": 6.0864,
+      "step": 2896
+    },
+    {
+      "epoch": 0.9954043722887944,
+      "grad_norm": 0.820821225643158,
+      "learning_rate": 4.0467796992654795e-08,
+      "loss": 6.1486,
+      "step": 2897
+    },
+    {
+      "epoch": 0.9957479706223424,
+      "grad_norm": 0.8554714918136597,
+      "learning_rate": 3.5252070185870514e-08,
+      "loss": 6.2766,
+      "step": 2898
+    },
+    {
+      "epoch": 0.9960915689558906,
+      "grad_norm": 0.914206862449646,
+      "learning_rate": 3.039601606796838e-08,
+      "loss": 6.222,
+      "step": 2899
+    },
+    {
+      "epoch": 0.9964351672894386,
+      "grad_norm": 1.1582309007644653,
+      "learning_rate": 2.5899641626331295e-08,
+      "loss": 6.2757,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9967787656229867,
+      "grad_norm": 0.46385082602500916,
+      "learning_rate": 2.1762953330728417e-08,
+      "loss": 5.9457,
+      "step": 2901
+    },
+    {
+      "epoch": 0.9971223639565349,
+      "grad_norm": 0.5638375282287598,
+      "learning_rate": 1.798595713342621e-08,
+      "loss": 5.8336,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9974659622900829,
+      "grad_norm": 0.42126670479774475,
+      "learning_rate": 1.4568658469132912e-08,
+      "loss": 5.9258,
+      "step": 2903
+    },
+    {
+      "epoch": 0.997809560623631,
+      "grad_norm": 0.44771361351013184,
+      "learning_rate": 1.1511062254970784e-08,
+      "loss": 5.9978,
+      "step": 2904
+    },
+    {
+      "epoch": 0.998153158957179,
+      "grad_norm": 0.44968634843826294,
+      "learning_rate": 8.813172890503874e-09,
+      "loss": 6.0372,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9984967572907272,
+      "grad_norm": 0.48755761981010437,
+      "learning_rate": 6.4749942576824986e-09,
+      "loss": 5.919,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9988403556242752,
+      "grad_norm": 0.5411795973777771,
+      "learning_rate": 4.496529720926512e-09,
+      "loss": 6.0362,
+      "step": 2907
+    },
+    {
+      "epoch": 0.9991839539578233,
+      "grad_norm": 0.5617440938949585,
+      "learning_rate": 2.8777821270142835e-09,
+      "loss": 6.0123,
+      "step": 2908
+    },
+    {
+      "epoch": 0.9995275522913714,
+      "grad_norm": 0.5796970725059509,
+      "learning_rate": 1.6187538051382067e-09,
+      "loss": 6.0778,
+      "step": 2909
+    },
+    {
+      "epoch": 0.9998711506249195,
+      "grad_norm": 0.6862174868583679,
+      "learning_rate": 7.194465669602135e-10,
+      "loss": 6.1623,
+      "step": 2910
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5478452444076538,
+      "learning_rate": 1.7986170644523901e-10,
+      "loss": 6.0827,
+      "step": 2911
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2911,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 102226707072000.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}