diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -1,5859 +1,265 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.252577319587629,
+  "epoch": 0.18041237113402062,
   "eval_steps": 97,
-  "global_step": 825,
+  "global_step": 35,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.005154639175257732,
-      "grad_norm": 0.8410804867744446,
+      "grad_norm": 14.970044136047363,
       "learning_rate": 2.0000000000000003e-06,
-      "loss": 0.9672,
+      "loss": 9.1577,
       "step": 1
     },
     {
       "epoch": 0.005154639175257732,
-      "eval_loss": 0.986353874206543,
-      "eval_runtime": 22.7571,
-      "eval_samples_per_second": 7.207,
-      "eval_steps_per_second": 1.802,
+      "eval_loss": 9.00115966796875,
+      "eval_runtime": 22.9959,
+      "eval_samples_per_second": 7.132,
+      "eval_steps_per_second": 1.783,
       "step": 1
     },
     {
       "epoch": 0.010309278350515464,
-      "grad_norm": 0.7938956618309021,
+      "grad_norm": 15.33629035949707,
       "learning_rate": 4.000000000000001e-06,
-      "loss": 1.0214,
+      "loss": 9.2777,
       "step": 2
     },
     {
       "epoch": 0.015463917525773196,
-      "grad_norm": 0.705802321434021,
+      "grad_norm": 16.92699432373047,
       "learning_rate": 6e-06,
-      "loss": 0.8948,
+      "loss": 8.744,
       "step": 3
     },
     {
       "epoch": 0.020618556701030927,
-      "grad_norm": 0.8201313614845276,
+      "grad_norm": 19.3622989654541,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 0.9046,
+      "loss": 8.545,
       "step": 4
     },
     {
       "epoch": 0.02577319587628866,
-      "grad_norm": 1.1557425260543823,
+      "grad_norm": 20.165861129760742,
       "learning_rate": 1e-05,
-      "loss": 1.0693,
+      "loss": 9.7134,
       "step": 5
     },
     {
       "epoch": 0.030927835051546393,
-      "grad_norm": 1.263604998588562,
+      "grad_norm": 18.959148406982422,
       "learning_rate": 1.2e-05,
-      "loss": 1.1302,
+      "loss": 9.7526,
       "step": 6
     },
     {
       "epoch": 0.03608247422680412,
-      "grad_norm": 0.7817522883415222,
+      "grad_norm": 16.62849235534668,
       "learning_rate": 1.4000000000000001e-05,
-      "loss": 0.9247,
+      "loss": 8.9248,
       "step": 7
     },
     {
       "epoch": 0.041237113402061855,
-      "grad_norm": 0.6655614376068115,
+      "grad_norm": 19.083742141723633,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 0.6942,
+      "loss": 9.0543,
       "step": 8
     },
     {
       "epoch": 0.04639175257731959,
-      "grad_norm": 0.9141671061515808,
+      "grad_norm": 16.922298431396484,
       "learning_rate": 1.8e-05,
-      "loss": 0.8889,
+      "loss": 8.7968,
       "step": 9
     },
     {
       "epoch": 0.05154639175257732,
-      "grad_norm": 0.8002082109451294,
+      "grad_norm": 15.013986587524414,
       "learning_rate": 2e-05,
-      "loss": 0.9334,
+      "loss": 8.8618,
       "step": 10
     },
     {
       "epoch": 0.05670103092783505,
-      "grad_norm": 1.0407661199569702,
+      "grad_norm": 16.929433822631836,
       "learning_rate": 2.2000000000000003e-05,
-      "loss": 0.9861,
+      "loss": 8.3198,
       "step": 11
     },
     {
       "epoch": 0.061855670103092786,
-      "grad_norm": 0.7072747945785522,
+      "grad_norm": 14.889703750610352,
       "learning_rate": 2.4e-05,
-      "loss": 1.0617,
+      "loss": 8.1047,
       "step": 12
     },
     {
       "epoch": 0.06701030927835051,
-      "grad_norm": 0.6469315886497498,
+      "grad_norm": 15.865534782409668,
       "learning_rate": 2.6000000000000002e-05,
-      "loss": 0.846,
+      "loss": 7.6863,
       "step": 13
     },
     {
       "epoch": 0.07216494845360824,
-      "grad_norm": 0.7403808236122131,
+      "grad_norm": 15.19003963470459,
       "learning_rate": 2.8000000000000003e-05,
-      "loss": 0.8555,
+      "loss": 7.779,
       "step": 14
     },
     {
       "epoch": 0.07731958762886598,
-      "grad_norm": 0.7198350429534912,
+      "grad_norm": 17.154193878173828,
       "learning_rate": 3e-05,
-      "loss": 0.9019,
+      "loss": 7.7273,
       "step": 15
     },
     {
       "epoch": 0.08247422680412371,
-      "grad_norm": 0.5983201265335083,
+      "grad_norm": 14.37552547454834,
       "learning_rate": 3.2000000000000005e-05,
-      "loss": 0.9096,
+      "loss": 7.1202,
       "step": 16
     },
     {
       "epoch": 0.08762886597938144,
-      "grad_norm": 0.6279081702232361,
+      "grad_norm": 13.732095718383789,
       "learning_rate": 3.4000000000000007e-05,
-      "loss": 0.945,
+      "loss": 6.9298,
       "step": 17
     },
     {
       "epoch": 0.09278350515463918,
-      "grad_norm": 0.4670332670211792,
+      "grad_norm": 13.2391357421875,
       "learning_rate": 3.6e-05,
-      "loss": 0.8456,
+      "loss": 6.308,
       "step": 18
     },
     {
       "epoch": 0.0979381443298969,
-      "grad_norm": 0.4307556748390198,
+      "grad_norm": 11.839639663696289,
       "learning_rate": 3.8e-05,
-      "loss": 0.8309,
+      "loss": 6.1335,
       "step": 19
     },
     {
       "epoch": 0.10309278350515463,
-      "grad_norm": 0.6864880919456482,
+      "grad_norm": 11.819024085998535,
       "learning_rate": 4e-05,
-      "loss": 1.1069,
+      "loss": 5.7396,
       "step": 20
     },
     {
       "epoch": 0.10824742268041238,
-      "grad_norm": 0.3861861526966095,
+      "grad_norm": 11.764089584350586,
       "learning_rate": 4.2e-05,
-      "loss": 0.864,
+      "loss": 5.1633,
       "step": 21
     },
     {
       "epoch": 0.1134020618556701,
-      "grad_norm": 0.5285366773605347,
+      "grad_norm": 12.526201248168945,
       "learning_rate": 4.4000000000000006e-05,
-      "loss": 0.8627,
+      "loss": 4.7586,
       "step": 22
     },
     {
       "epoch": 0.11855670103092783,
-      "grad_norm": 0.39096203446388245,
+      "grad_norm": 10.944100379943848,
       "learning_rate": 4.600000000000001e-05,
-      "loss": 0.8984,
+      "loss": 4.47,
       "step": 23
     },
     {
       "epoch": 0.12371134020618557,
-      "grad_norm": 0.39784133434295654,
+      "grad_norm": 14.008569717407227,
       "learning_rate": 4.8e-05,
-      "loss": 0.9631,
+      "loss": 4.2032,
       "step": 24
     },
     {
       "epoch": 0.12886597938144329,
-      "grad_norm": 0.37275853753089905,
+      "grad_norm": 14.705824851989746,
       "learning_rate": 5e-05,
-      "loss": 0.831,
+      "loss": 3.2385,
       "step": 25
     },
     {
       "epoch": 0.13402061855670103,
-      "grad_norm": 0.4517609179019928,
+      "grad_norm": 16.897640228271484,
       "learning_rate": 5.2000000000000004e-05,
-      "loss": 0.8904,
+      "loss": 3.0181,
       "step": 26
     },
     {
       "epoch": 0.13917525773195877,
-      "grad_norm": 0.4948658347129822,
+      "grad_norm": 15.038681983947754,
       "learning_rate": 5.4000000000000005e-05,
-      "loss": 0.8942,
+      "loss": 2.171,
       "step": 27
     },
     {
       "epoch": 0.14432989690721648,
-      "grad_norm": 0.43722641468048096,
+      "grad_norm": 15.12243938446045,
       "learning_rate": 5.6000000000000006e-05,
-      "loss": 0.8707,
+      "loss": 2.1848,
       "step": 28
     },
     {
       "epoch": 0.14948453608247422,
-      "grad_norm": 0.36861544847488403,
+      "grad_norm": 12.384464263916016,
       "learning_rate": 5.8e-05,
-      "loss": 0.7922,
+      "loss": 1.6113,
       "step": 29
     },
     {
       "epoch": 0.15463917525773196,
-      "grad_norm": 0.5646365880966187,
+      "grad_norm": 10.626302719116211,
       "learning_rate": 6e-05,
-      "loss": 0.9966,
+      "loss": 0.8058,
       "step": 30
     },
     {
       "epoch": 0.15979381443298968,
-      "grad_norm": 0.32113298773765564,
+      "grad_norm": 10.231700897216797,
       "learning_rate": 6.2e-05,
-      "loss": 0.6493,
+      "loss": 1.1714,
       "step": 31
     },
     {
       "epoch": 0.16494845360824742,
-      "grad_norm": 0.5166299939155579,
+      "grad_norm": 11.243683815002441,
       "learning_rate": 6.400000000000001e-05,
-      "loss": 0.9284,
+      "loss": 0.8837,
       "step": 32
     },
     {
       "epoch": 0.17010309278350516,
-      "grad_norm": 0.47344839572906494,
+      "grad_norm": 13.52505874633789,
       "learning_rate": 6.6e-05,
-      "loss": 0.718,
+      "loss": 0.9935,
       "step": 33
     },
     {
       "epoch": 0.17525773195876287,
-      "grad_norm": 0.4184200167655945,
+      "grad_norm": 11.277730941772461,
       "learning_rate": 6.800000000000001e-05,
-      "loss": 0.8632,
+      "loss": 0.7997,
       "step": 34
     },
     {
       "epoch": 0.18041237113402062,
-      "grad_norm": 0.5064208507537842,
+      "grad_norm": 8.148871421813965,
       "learning_rate": 7e-05,
-      "loss": 0.8827,
+      "loss": 0.5211,
       "step": 35
-    },
-    {
-      "epoch": 0.18556701030927836,
-      "grad_norm": 0.3908073902130127,
-      "learning_rate": 7.2e-05,
-      "loss": 0.7842,
-      "step": 36
-    },
-    {
-      "epoch": 0.19072164948453607,
-      "grad_norm": 0.4032430648803711,
-      "learning_rate": 7.4e-05,
-      "loss": 0.711,
-      "step": 37
-    },
-    {
-      "epoch": 0.1958762886597938,
-      "grad_norm": 0.44231921434402466,
-      "learning_rate": 7.6e-05,
-      "loss": 0.755,
-      "step": 38
-    },
-    {
-      "epoch": 0.20103092783505155,
-      "grad_norm": 0.4277024269104004,
-      "learning_rate": 7.800000000000001e-05,
-      "loss": 0.9275,
-      "step": 39
-    },
-    {
-      "epoch": 0.20618556701030927,
-      "grad_norm": 0.417645663022995,
-      "learning_rate": 8e-05,
-      "loss": 0.7142,
-      "step": 40
-    },
-    {
-      "epoch": 0.211340206185567,
-      "grad_norm": 0.5510072708129883,
-      "learning_rate": 8.2e-05,
-      "loss": 0.8196,
-      "step": 41
-    },
-    {
-      "epoch": 0.21649484536082475,
-      "grad_norm": 0.42206454277038574,
-      "learning_rate": 8.4e-05,
-      "loss": 0.8858,
-      "step": 42
-    },
-    {
-      "epoch": 0.22164948453608246,
-      "grad_norm": 0.4281873106956482,
-      "learning_rate": 8.6e-05,
-      "loss": 0.9484,
-      "step": 43
-    },
-    {
-      "epoch": 0.2268041237113402,
-      "grad_norm": 0.36399605870246887,
-      "learning_rate": 8.800000000000001e-05,
-      "loss": 0.662,
-      "step": 44
-    },
-    {
-      "epoch": 0.23195876288659795,
-      "grad_norm": 0.43412840366363525,
-      "learning_rate": 9e-05,
-      "loss": 0.7535,
-      "step": 45
-    },
-    {
-      "epoch": 0.23711340206185566,
-      "grad_norm": 0.5094694495201111,
-      "learning_rate": 9.200000000000001e-05,
-      "loss": 1.104,
-      "step": 46
-    },
-    {
-      "epoch": 0.2422680412371134,
-      "grad_norm": 0.5075117945671082,
-      "learning_rate": 9.4e-05,
-      "loss": 0.8194,
-      "step": 47
-    },
-    {
-      "epoch": 0.24742268041237114,
-      "grad_norm": 0.4164354205131531,
-      "learning_rate": 9.6e-05,
-      "loss": 0.8732,
-      "step": 48
-    },
-    {
-      "epoch": 0.25257731958762886,
-      "grad_norm": 0.4368288218975067,
-      "learning_rate": 9.8e-05,
-      "loss": 0.9469,
-      "step": 49
-    },
-    {
-      "epoch": 0.25773195876288657,
-      "grad_norm": 0.4410831928253174,
-      "learning_rate": 0.0001,
-      "loss": 0.9754,
-      "step": 50
-    },
-    {
-      "epoch": 0.26288659793814434,
-      "grad_norm": 0.3590613603591919,
-      "learning_rate": 9.999970848314005e-05,
-      "loss": 0.6932,
-      "step": 51
-    },
-    {
-      "epoch": 0.26804123711340205,
-      "grad_norm": 0.4591519832611084,
-      "learning_rate": 9.999883393595947e-05,
-      "loss": 0.8374,
-      "step": 52
-    },
-    {
-      "epoch": 0.27319587628865977,
-      "grad_norm": 0.3702682852745056,
-      "learning_rate": 9.999737636865609e-05,
-      "loss": 0.73,
-      "step": 53
-    },
-    {
-      "epoch": 0.27835051546391754,
-      "grad_norm": 0.42546412348747253,
-      "learning_rate": 9.99953357982261e-05,
-      "loss": 0.7748,
-      "step": 54
-    },
-    {
-      "epoch": 0.28350515463917525,
-      "grad_norm": 0.4516392946243286,
-      "learning_rate": 9.999271224846396e-05,
-      "loss": 0.8051,
-      "step": 55
-    },
-    {
-      "epoch": 0.28865979381443296,
-      "grad_norm": 0.34078386425971985,
-      "learning_rate": 9.998950574996199e-05,
-      "loss": 0.6558,
-      "step": 56
-    },
-    {
-      "epoch": 0.29381443298969073,
-      "grad_norm": 0.5125095844268799,
-      "learning_rate": 9.998571634011015e-05,
-      "loss": 0.9307,
-      "step": 57
-    },
-    {
-      "epoch": 0.29896907216494845,
-      "grad_norm": 0.3470001816749573,
-      "learning_rate": 9.998134406309554e-05,
-      "loss": 0.7473,
-      "step": 58
-    },
-    {
-      "epoch": 0.30412371134020616,
-      "grad_norm": 0.3486801087856293,
-      "learning_rate": 9.99763889699018e-05,
-      "loss": 0.6796,
-      "step": 59
-    },
-    {
-      "epoch": 0.30927835051546393,
-      "grad_norm": 0.4312107563018799,
-      "learning_rate": 9.99708511183087e-05,
-      "loss": 0.8765,
-      "step": 60
-    },
-    {
-      "epoch": 0.31443298969072164,
-      "grad_norm": 0.3910192847251892,
-      "learning_rate": 9.996473057289132e-05,
-      "loss": 0.849,
-      "step": 61
-    },
-    {
-      "epoch": 0.31958762886597936,
-      "grad_norm": 0.4039042890071869,
-      "learning_rate": 9.995802740501933e-05,
-      "loss": 0.8548,
-      "step": 62
-    },
-    {
-      "epoch": 0.3247422680412371,
-      "grad_norm": 0.4193078577518463,
-      "learning_rate": 9.99507416928562e-05,
-      "loss": 0.7689,
-      "step": 63
-    },
-    {
-      "epoch": 0.32989690721649484,
-      "grad_norm": 0.36549004912376404,
-      "learning_rate": 9.994287352135825e-05,
-      "loss": 0.8877,
-      "step": 64
-    },
-    {
-      "epoch": 0.33505154639175255,
-      "grad_norm": 0.3615829348564148,
-      "learning_rate": 9.993442298227365e-05,
-      "loss": 0.8102,
-      "step": 65
-    },
-    {
-      "epoch": 0.3402061855670103,
-      "grad_norm": 0.42232227325439453,
-      "learning_rate": 9.99253901741414e-05,
-      "loss": 0.7458,
-      "step": 66
-    },
-    {
-      "epoch": 0.34536082474226804,
-      "grad_norm": 0.38672903180122375,
-      "learning_rate": 9.991577520229014e-05,
-      "loss": 0.923,
-      "step": 67
-    },
-    {
-      "epoch": 0.35051546391752575,
-      "grad_norm": 0.3575025200843811,
-      "learning_rate": 9.99055781788369e-05,
-      "loss": 0.7854,
-      "step": 68
-    },
-    {
-      "epoch": 0.3556701030927835,
-      "grad_norm": 0.37855708599090576,
-      "learning_rate": 9.989479922268588e-05,
-      "loss": 0.6534,
-      "step": 69
-    },
-    {
-      "epoch": 0.36082474226804123,
-      "grad_norm": 0.3880573809146881,
-      "learning_rate": 9.988343845952697e-05,
-      "loss": 0.8377,
-      "step": 70
-    },
-    {
-      "epoch": 0.36597938144329895,
-      "grad_norm": 0.41330426931381226,
-      "learning_rate": 9.98714960218343e-05,
-      "loss": 0.8176,
-      "step": 71
-    },
-    {
-      "epoch": 0.3711340206185567,
-      "grad_norm": 0.3920232057571411,
-      "learning_rate": 9.985897204886481e-05,
-      "loss": 0.7983,
-      "step": 72
-    },
-    {
-      "epoch": 0.37628865979381443,
-      "grad_norm": 0.4510120451450348,
-      "learning_rate": 9.98458666866564e-05,
-      "loss": 0.8814,
-      "step": 73
-    },
-    {
-      "epoch": 0.38144329896907214,
-      "grad_norm": 0.4475913643836975,
-      "learning_rate": 9.983218008802648e-05,
-      "loss": 0.8086,
-      "step": 74
-    },
-    {
-      "epoch": 0.3865979381443299,
-      "grad_norm": 0.4115805923938751,
-      "learning_rate": 9.981791241257e-05,
-      "loss": 0.8963,
-      "step": 75
-    },
-    {
-      "epoch": 0.3917525773195876,
-      "grad_norm": 0.3336406648159027,
-      "learning_rate": 9.98030638266577e-05,
-      "loss": 0.6722,
-      "step": 76
-    },
-    {
-      "epoch": 0.39690721649484534,
-      "grad_norm": 0.4010220766067505,
-      "learning_rate": 9.978763450343407e-05,
-      "loss": 0.9867,
-      "step": 77
-    },
-    {
-      "epoch": 0.4020618556701031,
-      "grad_norm": 0.45894166827201843,
-      "learning_rate": 9.977162462281544e-05,
-      "loss": 0.8535,
-      "step": 78
-    },
-    {
-      "epoch": 0.4072164948453608,
-      "grad_norm": 0.36226922273635864,
-      "learning_rate": 9.975503437148783e-05,
-      "loss": 0.715,
-      "step": 79
-    },
-    {
-      "epoch": 0.41237113402061853,
-      "grad_norm": 0.4665627181529999,
-      "learning_rate": 9.973786394290474e-05,
-      "loss": 0.8986,
-      "step": 80
-    },
-    {
-      "epoch": 0.4175257731958763,
-      "grad_norm": 0.38297462463378906,
-      "learning_rate": 9.972011353728496e-05,
-      "loss": 0.6301,
-      "step": 81
-    },
-    {
-      "epoch": 0.422680412371134,
-      "grad_norm": 0.42398950457572937,
-      "learning_rate": 9.970178336161018e-05,
-      "loss": 0.8692,
-      "step": 82
-    },
-    {
-      "epoch": 0.42783505154639173,
-      "grad_norm": 0.43351009488105774,
-      "learning_rate": 9.968287362962264e-05,
-      "loss": 0.8294,
-      "step": 83
-    },
-    {
-      "epoch": 0.4329896907216495,
-      "grad_norm": 0.38560381531715393,
-      "learning_rate": 9.96633845618225e-05,
-      "loss": 0.6438,
-      "step": 84
-    },
-    {
-      "epoch": 0.4381443298969072,
-      "grad_norm": 0.38818269968032837,
-      "learning_rate": 9.96433163854655e-05,
-      "loss": 0.8707,
-      "step": 85
-    },
-    {
-      "epoch": 0.44329896907216493,
-      "grad_norm": 0.42916783690452576,
-      "learning_rate": 9.962266933456008e-05,
-      "loss": 0.8905,
-      "step": 86
-    },
-    {
-      "epoch": 0.4484536082474227,
-      "grad_norm": 0.4095560908317566,
-      "learning_rate": 9.96014436498648e-05,
-      "loss": 0.7936,
-      "step": 87
-    },
-    {
-      "epoch": 0.4536082474226804,
-      "grad_norm": 0.3557605445384979,
-      "learning_rate": 9.957963957888542e-05,
-      "loss": 0.7018,
-      "step": 88
-    },
-    {
-      "epoch": 0.4587628865979381,
-      "grad_norm": 0.3835409879684448,
-      "learning_rate": 9.955725737587214e-05,
-      "loss": 0.7601,
-      "step": 89
-    },
-    {
-      "epoch": 0.4639175257731959,
-      "grad_norm": 0.43281444907188416,
-      "learning_rate": 9.953429730181653e-05,
-      "loss": 0.8437,
-      "step": 90
-    },
-    {
-      "epoch": 0.4690721649484536,
-      "grad_norm": 0.5132317543029785,
-      "learning_rate": 9.951075962444856e-05,
-      "loss": 0.9834,
-      "step": 91
-    },
-    {
-      "epoch": 0.4742268041237113,
-      "grad_norm": 0.48250970244407654,
-      "learning_rate": 9.94866446182334e-05,
-      "loss": 0.9003,
-      "step": 92
-    },
-    {
-      "epoch": 0.4793814432989691,
-      "grad_norm": 0.4286979138851166,
-      "learning_rate": 9.94619525643683e-05,
-      "loss": 0.9507,
-      "step": 93
-    },
-    {
-      "epoch": 0.4845360824742268,
-      "grad_norm": 0.3736780285835266,
-      "learning_rate": 9.943668375077925e-05,
-      "loss": 0.8167,
-      "step": 94
-    },
-    {
-      "epoch": 0.4896907216494845,
-      "grad_norm": 0.39715173840522766,
-      "learning_rate": 9.941083847211765e-05,
-      "loss": 0.7804,
-      "step": 95
-    },
-    {
-      "epoch": 0.4948453608247423,
-      "grad_norm": 0.4031369686126709,
-      "learning_rate": 9.938441702975689e-05,
-      "loss": 0.8113,
-      "step": 96
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 0.39510655403137207,
-      "learning_rate": 9.93574197317888e-05,
-      "loss": 0.7941,
-      "step": 97
-    },
-    {
-      "epoch": 0.5,
-      "eval_loss": 0.8026774525642395,
-      "eval_runtime": 23.1438,
-      "eval_samples_per_second": 7.086,
-      "eval_steps_per_second": 1.772,
-      "step": 97
-    },
-    {
-      "epoch": 0.5051546391752577,
-      "grad_norm": 0.5623717308044434,
-      "learning_rate": 9.93298468930201e-05,
-      "loss": 0.9629,
-      "step": 98
-    },
-    {
-      "epoch": 0.5103092783505154,
-      "grad_norm": 0.3663593530654907,
-      "learning_rate": 9.930169883496867e-05,
-      "loss": 0.7471,
-      "step": 99
-    },
-    {
-      "epoch": 0.5154639175257731,
-      "grad_norm": 0.41299688816070557,
-      "learning_rate": 9.927297588585984e-05,
-      "loss": 0.6697,
-      "step": 100
-    },
-    {
-      "epoch": 0.520618556701031,
-      "grad_norm": 0.41344979405403137,
-      "learning_rate": 9.924367838062259e-05,
-      "loss": 0.6884,
-      "step": 101
-    },
-    {
-      "epoch": 0.5257731958762887,
-      "grad_norm": 0.4803592562675476,
-      "learning_rate": 9.921380666088558e-05,
-      "loss": 0.807,
-      "step": 102
-    },
-    {
-      "epoch": 0.5309278350515464,
-      "grad_norm": 0.3681694567203522,
-      "learning_rate": 9.91833610749732e-05,
-      "loss": 0.8595,
-      "step": 103
-    },
-    {
-      "epoch": 0.5360824742268041,
-      "grad_norm": 0.3941933810710907,
-      "learning_rate": 9.915234197790152e-05,
-      "loss": 0.7553,
-      "step": 104
-    },
-    {
-      "epoch": 0.5412371134020618,
-      "grad_norm": 0.4864148199558258,
-      "learning_rate": 9.912074973137412e-05,
-      "loss": 0.9059,
-      "step": 105
-    },
-    {
-      "epoch": 0.5463917525773195,
-      "grad_norm": 0.4248103201389313,
-      "learning_rate": 9.908858470377793e-05,
-      "loss": 0.7479,
-      "step": 106
-    },
-    {
-      "epoch": 0.5515463917525774,
-      "grad_norm": 0.39183953404426575,
-      "learning_rate": 9.905584727017884e-05,
-      "loss": 0.8723,
-      "step": 107
-    },
-    {
-      "epoch": 0.5567010309278351,
-      "grad_norm": 0.35667285323143005,
-      "learning_rate": 9.90225378123174e-05,
-      "loss": 0.7337,
-      "step": 108
-    },
-    {
-      "epoch": 0.5618556701030928,
-      "grad_norm": 0.33024296164512634,
-      "learning_rate": 9.898865671860438e-05,
-      "loss": 0.6999,
-      "step": 109
-    },
-    {
-      "epoch": 0.5670103092783505,
-      "grad_norm": 0.5383896827697754,
-      "learning_rate": 9.895420438411616e-05,
-      "loss": 0.7962,
-      "step": 110
-    },
-    {
-      "epoch": 0.5721649484536082,
-      "grad_norm": 0.41683122515678406,
-      "learning_rate": 9.891918121059019e-05,
-      "loss": 0.8638,
-      "step": 111
-    },
-    {
-      "epoch": 0.5773195876288659,
-      "grad_norm": 0.37739497423171997,
-      "learning_rate": 9.888358760642029e-05,
-      "loss": 0.65,
-      "step": 112
-    },
-    {
-      "epoch": 0.5824742268041238,
-      "grad_norm": 0.5009469985961914,
-      "learning_rate": 9.884742398665191e-05,
-      "loss": 0.8071,
-      "step": 113
-    },
-    {
-      "epoch": 0.5876288659793815,
-      "grad_norm": 0.38047826290130615,
-      "learning_rate": 9.881069077297723e-05,
-      "loss": 0.702,
-      "step": 114
-    },
-    {
-      "epoch": 0.5927835051546392,
-      "grad_norm": 0.42374080419540405,
-      "learning_rate": 9.877338839373032e-05,
-      "loss": 0.761,
-      "step": 115
-    },
-    {
-      "epoch": 0.5979381443298969,
-      "grad_norm": 0.45142003893852234,
-      "learning_rate": 9.873551728388203e-05,
-      "loss": 0.7846,
-      "step": 116
-    },
-    {
-      "epoch": 0.6030927835051546,
-      "grad_norm": 0.4653826653957367,
-      "learning_rate": 9.869707788503508e-05,
-      "loss": 0.8521,
-      "step": 117
-    },
-    {
-      "epoch": 0.6082474226804123,
-      "grad_norm": 0.40280038118362427,
-      "learning_rate": 9.865807064541877e-05,
-      "loss": 0.7892,
-      "step": 118
-    },
-    {
-      "epoch": 0.6134020618556701,
-      "grad_norm": 0.38454410433769226,
-      "learning_rate": 9.861849601988383e-05,
-      "loss": 0.7835,
-      "step": 119
-    },
-    {
-      "epoch": 0.6185567010309279,
-      "grad_norm": 0.43196186423301697,
-      "learning_rate": 9.857835446989707e-05,
-      "loss": 0.7327,
-      "step": 120
-    },
-    {
-      "epoch": 0.6237113402061856,
-      "grad_norm": 0.4396250247955322,
-      "learning_rate": 9.853764646353605e-05,
-      "loss": 0.9224,
-      "step": 121
-    },
-    {
-      "epoch": 0.6288659793814433,
-      "grad_norm": 0.34583795070648193,
-      "learning_rate": 9.849637247548356e-05,
-      "loss": 0.6383,
-      "step": 122
-    },
-    {
-      "epoch": 0.634020618556701,
-      "grad_norm": 0.49787506461143494,
-      "learning_rate": 9.845453298702216e-05,
-      "loss": 0.8184,
-      "step": 123
-    },
-    {
-      "epoch": 0.6391752577319587,
-      "grad_norm": 0.37390121817588806,
-      "learning_rate": 9.841212848602846e-05,
-      "loss": 0.6893,
-      "step": 124
-    },
-    {
-      "epoch": 0.6443298969072165,
-      "grad_norm": 0.4788094460964203,
-      "learning_rate": 9.836915946696759e-05,
-      "loss": 0.7379,
-      "step": 125
-    },
-    {
-      "epoch": 0.6494845360824743,
-      "grad_norm": 0.5014426112174988,
-      "learning_rate": 9.832562643088724e-05,
-      "loss": 0.8205,
-      "step": 126
-    },
-    {
-      "epoch": 0.654639175257732,
-      "grad_norm": 0.3642871081829071,
-      "learning_rate": 9.828152988541201e-05,
-      "loss": 0.6653,
-      "step": 127
-    },
-    {
-      "epoch": 0.6597938144329897,
-      "grad_norm": 0.3835623264312744,
-      "learning_rate": 9.823687034473735e-05,
-      "loss": 0.7903,
-      "step": 128
-    },
-    {
-      "epoch": 0.6649484536082474,
-      "grad_norm": 0.36520278453826904,
-      "learning_rate": 9.81916483296236e-05,
-      "loss": 0.5541,
-      "step": 129
-    },
-    {
-      "epoch": 0.6701030927835051,
-      "grad_norm": 0.382394403219223,
-      "learning_rate": 9.814586436738998e-05,
-      "loss": 0.8159,
-      "step": 130
-    },
-    {
-      "epoch": 0.6752577319587629,
-      "grad_norm": 0.4563412070274353,
-      "learning_rate": 9.809951899190835e-05,
-      "loss": 0.9263,
-      "step": 131
-    },
-    {
-      "epoch": 0.6804123711340206,
-      "grad_norm": 0.43446069955825806,
-      "learning_rate": 9.805261274359705e-05,
-      "loss": 0.8323,
-      "step": 132
-    },
-    {
-      "epoch": 0.6855670103092784,
-      "grad_norm": 0.3788876235485077,
-      "learning_rate": 9.800514616941457e-05,
-      "loss": 0.8373,
-      "step": 133
-    },
-    {
-      "epoch": 0.6907216494845361,
-      "grad_norm": 0.32987335324287415,
-      "learning_rate": 9.795711982285316e-05,
-      "loss": 0.5943,
-      "step": 134
-    },
-    {
-      "epoch": 0.6958762886597938,
-      "grad_norm": 0.41327905654907227,
-      "learning_rate": 9.790853426393245e-05,
-      "loss": 0.7817,
-      "step": 135
-    },
-    {
-      "epoch": 0.7010309278350515,
-      "grad_norm": 0.40496692061424255,
-      "learning_rate": 9.785939005919278e-05,
-      "loss": 0.7946,
-      "step": 136
-    },
-    {
-      "epoch": 0.7061855670103093,
-      "grad_norm": 0.305164635181427,
-      "learning_rate": 9.780968778168874e-05,
-      "loss": 0.6239,
-      "step": 137
-    },
-    {
-      "epoch": 0.711340206185567,
-      "grad_norm": 0.5270521640777588,
-      "learning_rate": 9.77594280109824e-05,
-      "loss": 0.8353,
-      "step": 138
-    },
-    {
-      "epoch": 0.7164948453608248,
-      "grad_norm": 0.3462916314601898,
-      "learning_rate": 9.77086113331366e-05,
-      "loss": 0.7364,
-      "step": 139
-    },
-    {
-      "epoch": 0.7216494845360825,
-      "grad_norm": 0.4270078241825104,
-      "learning_rate": 9.765723834070804e-05,
-      "loss": 0.6089,
-      "step": 140
-    },
-    {
-      "epoch": 0.7268041237113402,
-      "grad_norm": 0.42253026366233826,
-      "learning_rate": 9.760530963274048e-05,
-      "loss": 0.6619,
-      "step": 141
-    },
-    {
-      "epoch": 0.7319587628865979,
-      "grad_norm": 0.4140520691871643,
-      "learning_rate": 9.755282581475769e-05,
-      "loss": 0.7179,
-      "step": 142
-    },
-    {
-      "epoch": 0.7371134020618557,
-      "grad_norm": 0.4326084852218628,
-      "learning_rate": 9.749978749875635e-05,
-      "loss": 0.7465,
-      "step": 143
-    },
-    {
-      "epoch": 0.7422680412371134,
-      "grad_norm": 0.44547709822654724,
-      "learning_rate": 9.744619530319899e-05,
-      "loss": 0.7606,
-      "step": 144
-    },
-    {
-      "epoch": 0.7474226804123711,
-      "grad_norm": 0.4291188418865204,
-      "learning_rate": 9.739204985300679e-05,
-      "loss": 0.8603,
-      "step": 145
-    },
-    {
-      "epoch": 0.7525773195876289,
-      "grad_norm": 0.38336077332496643,
-      "learning_rate": 9.733735177955219e-05,
-      "loss": 0.7124,
-      "step": 146
-    },
-    {
-      "epoch": 0.7577319587628866,
-      "grad_norm": 0.49138304591178894,
-      "learning_rate": 9.728210172065162e-05,
-      "loss": 0.9644,
-      "step": 147
-    },
-    {
-      "epoch": 0.7628865979381443,
-      "grad_norm": 0.4559629261493683,
-      "learning_rate": 9.722630032055803e-05,
-      "loss": 0.8036,
-      "step": 148
-    },
-    {
-      "epoch": 0.7680412371134021,
-      "grad_norm": 0.45178917050361633,
-      "learning_rate": 9.716994822995338e-05,
-      "loss": 0.9369,
-      "step": 149
-    },
-    {
-      "epoch": 0.7731958762886598,
-      "grad_norm": 0.3612707257270813,
-      "learning_rate": 9.711304610594104e-05,
-      "loss": 0.6617,
-      "step": 150
-    },
-    {
-      "epoch": 0.7783505154639175,
-      "grad_norm": 0.41716015338897705,
-      "learning_rate": 9.705559461203815e-05,
-      "loss": 0.8865,
-      "step": 151
-    },
-    {
-      "epoch": 0.7835051546391752,
-      "grad_norm": 0.4099236726760864,
-      "learning_rate": 9.699759441816787e-05,
-      "loss": 0.7566,
-      "step": 152
-    },
-    {
-      "epoch": 0.788659793814433,
-      "grad_norm": 0.42841625213623047,
-      "learning_rate": 9.69390462006516e-05,
-      "loss": 0.7696,
-      "step": 153
-    },
-    {
-      "epoch": 0.7938144329896907,
-      "grad_norm": 0.33566102385520935,
-      "learning_rate": 9.687995064220102e-05,
-      "loss": 0.7742,
-      "step": 154
-    },
-    {
-      "epoch": 0.7989690721649485,
-      "grad_norm": 0.4562363624572754,
-      "learning_rate": 9.682030843191022e-05,
-      "loss": 0.8731,
-      "step": 155
-    },
-    {
-      "epoch": 0.8041237113402062,
-      "grad_norm": 0.43077826499938965,
-      "learning_rate": 9.676012026524755e-05,
-      "loss": 0.7728,
-      "step": 156
-    },
-    {
-      "epoch": 0.8092783505154639,
-      "grad_norm": 0.4811527729034424,
-      "learning_rate": 9.669938684404766e-05,
-      "loss": 0.9621,
-      "step": 157
-    },
-    {
-      "epoch": 0.8144329896907216,
-      "grad_norm": 0.340797483921051,
-      "learning_rate": 9.663810887650318e-05,
-      "loss": 0.6092,
-      "step": 158
-    },
-    {
-      "epoch": 0.8195876288659794,
-      "grad_norm": 0.399932324886322,
-      "learning_rate": 9.657628707715655e-05,
-      "loss": 0.6474,
-      "step": 159
-    },
-    {
-      "epoch": 0.8247422680412371,
-      "grad_norm": 0.4639035761356354,
-      "learning_rate": 9.651392216689165e-05,
-      "loss": 0.9095,
-      "step": 160
-    },
-    {
-      "epoch": 0.8298969072164949,
-      "grad_norm": 0.39397144317626953,
-      "learning_rate": 9.645101487292539e-05,
-      "loss": 0.7839,
-      "step": 161
-    },
-    {
-      "epoch": 0.8350515463917526,
-      "grad_norm": 0.45253580808639526,
-      "learning_rate": 9.638756592879922e-05,
-      "loss": 0.7803,
-      "step": 162
-    },
-    {
-      "epoch": 0.8402061855670103,
-      "grad_norm": 0.39097675681114197,
-      "learning_rate": 9.632357607437065e-05,
-      "loss": 0.7402,
-      "step": 163
-    },
-    {
-      "epoch": 0.845360824742268,
-      "grad_norm": 0.3595625162124634,
-      "learning_rate": 9.625904605580452e-05,
-      "loss": 0.6507,
-      "step": 164
-    },
-    {
-      "epoch": 0.8505154639175257,
-      "grad_norm": 0.3774010241031647,
-      "learning_rate": 9.619397662556435e-05,
-      "loss": 0.7891,
-      "step": 165
-    },
-    {
-      "epoch": 0.8556701030927835,
-      "grad_norm": 0.43031737208366394,
-      "learning_rate": 9.612836854240358e-05,
-      "loss": 0.8141,
-      "step": 166
-    },
-    {
-      "epoch": 0.8608247422680413,
-      "grad_norm": 0.43795299530029297,
-      "learning_rate": 9.606222257135675e-05,
-      "loss": 0.7407,
-      "step": 167
-    },
-    {
-      "epoch": 0.865979381443299,
-      "grad_norm": 0.43038204312324524,
-      "learning_rate": 9.599553948373045e-05,
-      "loss": 0.5853,
-      "step": 168
-    },
-    {
-      "epoch": 0.8711340206185567,
-      "grad_norm": 0.4552323818206787,
-      "learning_rate": 9.592832005709448e-05,
-      "loss": 0.8141,
-      "step": 169
-    },
-    {
-      "epoch": 0.8762886597938144,
-      "grad_norm": 0.4247092604637146,
-      "learning_rate": 9.586056507527266e-05,
-      "loss": 0.8336,
-      "step": 170
-    },
-    {
-      "epoch": 0.8814432989690721,
-      "grad_norm": 0.4271308481693268,
-      "learning_rate": 9.579227532833377e-05,
-      "loss": 0.8686,
-      "step": 171
-    },
-    {
-      "epoch": 0.8865979381443299,
-      "grad_norm": 0.49759241938591003,
-      "learning_rate": 9.572345161258235e-05,
-      "loss": 0.8467,
-      "step": 172
-    },
-    {
-      "epoch": 0.8917525773195877,
-      "grad_norm": 0.38598212599754333,
-      "learning_rate": 9.565409473054932e-05,
-      "loss": 0.6457,
-      "step": 173
-    },
-    {
-      "epoch": 0.8969072164948454,
-      "grad_norm": 0.3519401550292969,
-      "learning_rate": 9.558420549098268e-05,
-      "loss": 0.8855,
-      "step": 174
-    },
-    {
-      "epoch": 0.9020618556701031,
-      "grad_norm": 0.4537753462791443,
-      "learning_rate": 9.551378470883812e-05,
-      "loss": 0.7518,
-      "step": 175
-    },
-    {
-      "epoch": 0.9072164948453608,
-      "grad_norm": 0.4422832429409027,
-      "learning_rate": 9.544283320526943e-05,
-      "loss": 0.8546,
-      "step": 176
-    },
-    {
-      "epoch": 0.9123711340206185,
-      "grad_norm": 0.43420276045799255,
-      "learning_rate": 9.537135180761903e-05,
-      "loss": 0.8066,
-      "step": 177
-    },
-    {
-      "epoch": 0.9175257731958762,
-      "grad_norm": 0.5223539471626282,
-      "learning_rate": 9.52993413494082e-05,
-      "loss": 0.922,
-      "step": 178
-    },
-    {
-      "epoch": 0.9226804123711341,
-      "grad_norm": 0.32053807377815247,
-      "learning_rate": 9.522680267032742e-05,
-      "loss": 0.6183,
-      "step": 179
-    },
-    {
-      "epoch": 0.9278350515463918,
-      "grad_norm": 0.4255862236022949,
-      "learning_rate": 9.515373661622664e-05,
-      "loss": 0.8405,
-      "step": 180
-    },
-    {
-      "epoch": 0.9329896907216495,
-      "grad_norm": 0.4862018823623657,
-      "learning_rate": 9.508014403910533e-05,
-      "loss": 0.8749,
-      "step": 181
-    },
-    {
-      "epoch": 0.9381443298969072,
-      "grad_norm": 0.38821473717689514,
-      "learning_rate": 9.500602579710256e-05,
-      "loss": 0.8354,
-      "step": 182
-    },
-    {
-      "epoch": 0.9432989690721649,
-      "grad_norm": 0.416502982378006,
-      "learning_rate": 9.4931382754487e-05,
-      "loss": 0.7721,
-      "step": 183
-    },
-    {
-      "epoch": 0.9484536082474226,
-      "grad_norm": 0.38249465823173523,
-      "learning_rate": 9.485621578164689e-05,
-      "loss": 0.6713,
-      "step": 184
-    },
-    {
-      "epoch": 0.9536082474226805,
-      "grad_norm": 0.4525269567966461,
-      "learning_rate": 9.478052575507982e-05,
-      "loss": 0.7183,
-      "step": 185
-    },
-    {
-      "epoch": 0.9587628865979382,
-      "grad_norm": 0.35123562812805176,
-      "learning_rate": 9.470431355738257e-05,
-      "loss": 0.7076,
-      "step": 186
-    },
-    {
-      "epoch": 0.9639175257731959,
-      "grad_norm": 0.3771204948425293,
-      "learning_rate": 9.46275800772407e-05,
-      "loss": 0.827,
-      "step": 187
-    },
-    {
-      "epoch": 0.9690721649484536,
-      "grad_norm": 0.39616143703460693,
-      "learning_rate": 9.45503262094184e-05,
-      "loss": 0.7204,
-      "step": 188
-    },
-    {
-      "epoch": 0.9742268041237113,
-      "grad_norm": 0.40495437383651733,
-      "learning_rate": 9.447255285474783e-05,
-      "loss": 0.739,
-      "step": 189
-    },
-    {
-      "epoch": 0.979381443298969,
-      "grad_norm": 0.44291606545448303,
-      "learning_rate": 9.439426092011875e-05,
-      "loss": 0.9957,
-      "step": 190
-    },
-    {
-      "epoch": 0.9845360824742269,
-      "grad_norm": 0.39156320691108704,
-      "learning_rate": 9.431545131846797e-05,
-      "loss": 0.8605,
-      "step": 191
-    },
-    {
-      "epoch": 0.9896907216494846,
-      "grad_norm": 0.5261005163192749,
-      "learning_rate": 9.423612496876855e-05,
-      "loss": 0.7107,
-      "step": 192
-    },
-    {
-      "epoch": 0.9948453608247423,
-      "grad_norm": 0.39191609621047974,
-      "learning_rate": 9.415628279601923e-05,
-      "loss": 0.7519,
-      "step": 193
-    },
-    {
-      "epoch": 1.0,
-      "grad_norm": 0.5396644473075867,
-      "learning_rate": 9.407592573123358e-05,
-      "loss": 0.7442,
-      "step": 194
-    },
-    {
-      "epoch": 1.0,
-      "eval_loss": 0.7816203236579895,
-      "eval_runtime": 23.1628,
-      "eval_samples_per_second": 7.08,
-      "eval_steps_per_second": 1.77,
-      "step": 194
-    },
-    {
-      "epoch": 1.0051546391752577,
-      "grad_norm": 0.37258586287498474,
-      "learning_rate": 9.39950547114292e-05,
-      "loss": 0.8192,
-      "step": 195
-    },
-    {
-      "epoch": 1.0103092783505154,
-      "grad_norm": 0.3827040493488312,
-      "learning_rate": 9.39136706796167e-05,
-      "loss": 0.6509,
-      "step": 196
-    },
-    {
-      "epoch": 1.0154639175257731,
-      "grad_norm": 0.3988623023033142,
-      "learning_rate": 9.383177458478878e-05,
-      "loss": 0.7298,
-      "step": 197
-    },
-    {
-      "epoch": 1.0206185567010309,
-      "grad_norm": 0.40072330832481384,
-      "learning_rate": 9.374936738190914e-05,
-      "loss": 0.7147,
-      "step": 198
-    },
-    {
-      "epoch": 1.0257731958762886,
-      "grad_norm": 0.45648422837257385,
-      "learning_rate": 9.366645003190132e-05,
-      "loss": 0.7607,
-      "step": 199
-    },
-    {
-      "epoch": 1.0309278350515463,
-      "grad_norm": 0.45081812143325806,
-      "learning_rate": 9.358302350163757e-05,
-      "loss": 0.9211,
-      "step": 200
-    },
-    {
-      "epoch": 1.0360824742268042,
-      "grad_norm": 0.4077489674091339,
-      "learning_rate": 9.349908876392748e-05,
-      "loss": 0.6459,
-      "step": 201
-    },
-    {
-      "epoch": 1.041237113402062,
-      "grad_norm": 0.47869938611984253,
-      "learning_rate": 9.341464679750669e-05,
-      "loss": 0.7938,
-      "step": 202
-    },
-    {
-      "epoch": 1.0463917525773196,
-      "grad_norm": 0.45740288496017456,
-      "learning_rate": 9.33296985870255e-05,
-      "loss": 0.7555,
-      "step": 203
-    },
-    {
-      "epoch": 1.0515463917525774,
-      "grad_norm": 0.3793458938598633,
-      "learning_rate": 9.32442451230373e-05,
-      "loss": 0.6282,
-      "step": 204
-    },
-    {
-      "epoch": 1.056701030927835,
-      "grad_norm": 0.4383585751056671,
-      "learning_rate": 9.315828740198714e-05,
-      "loss": 0.808,
-      "step": 205
-    },
-    {
-      "epoch": 1.0618556701030928,
-      "grad_norm": 0.36687490344047546,
-      "learning_rate": 9.30718264262e-05,
-      "loss": 0.8306,
-      "step": 206
-    },
-    {
-      "epoch": 1.0670103092783505,
-      "grad_norm": 0.3878268599510193,
-      "learning_rate": 9.298486320386919e-05,
-      "loss": 0.6411,
-      "step": 207
-    },
-    {
-      "epoch": 1.0721649484536082,
-      "grad_norm": 0.41478773951530457,
-      "learning_rate": 9.289739874904449e-05,
-      "loss": 0.6118,
-      "step": 208
-    },
-    {
-      "epoch": 1.077319587628866,
-      "grad_norm": 0.3920111358165741,
-      "learning_rate": 9.280943408162046e-05,
-      "loss": 0.6645,
-      "step": 209
-    },
-    {
-      "epoch": 1.0824742268041236,
-      "grad_norm": 0.44977059960365295,
-      "learning_rate": 9.272097022732443e-05,
-      "loss": 0.7551,
-      "step": 210
-    },
-    {
-      "epoch": 1.0876288659793814,
-      "grad_norm": 0.4331608712673187,
-      "learning_rate": 9.263200821770461e-05,
-      "loss": 0.7164,
-      "step": 211
-    },
-    {
-      "epoch": 1.0927835051546393,
-      "grad_norm": 0.47665220499038696,
-      "learning_rate": 9.254254909011804e-05,
-      "loss": 0.8104,
-      "step": 212
-    },
-    {
-      "epoch": 1.097938144329897,
-      "grad_norm": 0.42196914553642273,
-      "learning_rate": 9.245259388771845e-05,
-      "loss": 0.7194,
-      "step": 213
-    },
-    {
-      "epoch": 1.1030927835051547,
-      "grad_norm": 0.42440083622932434,
-      "learning_rate": 9.236214365944418e-05,
-      "loss": 0.7578,
-      "step": 214
-    },
-    {
-      "epoch": 1.1082474226804124,
-      "grad_norm": 0.38077446818351746,
-      "learning_rate": 9.22711994600059e-05,
-      "loss": 0.6378,
-      "step": 215
-    },
-    {
-      "epoch": 1.1134020618556701,
-      "grad_norm": 0.3987003564834595,
-      "learning_rate": 9.217976234987428e-05,
-      "loss": 0.7715,
-      "step": 216
-    },
-    {
-      "epoch": 1.1185567010309279,
-      "grad_norm": 0.3989262282848358,
-      "learning_rate": 9.208783339526773e-05,
-      "loss": 0.5646,
-      "step": 217
-    },
-    {
-      "epoch": 1.1237113402061856,
-      "grad_norm": 0.3729444742202759,
-      "learning_rate": 9.199541366813982e-05,
-      "loss": 0.7266,
-      "step": 218
-    },
-    {
-      "epoch": 1.1288659793814433,
-      "grad_norm": 0.487589031457901,
-      "learning_rate": 9.190250424616693e-05,
-      "loss": 0.7127,
-      "step": 219
-    },
-    {
-      "epoch": 1.134020618556701,
-      "grad_norm": 0.47167909145355225,
-      "learning_rate": 9.180910621273555e-05,
-      "loss": 0.681,
-      "step": 220
-    },
-    {
-      "epoch": 1.1391752577319587,
-      "grad_norm": 0.40272852778434753,
-      "learning_rate": 9.171522065692975e-05,
-      "loss": 0.7188,
-      "step": 221
-    },
-    {
-      "epoch": 1.1443298969072164,
-      "grad_norm": 0.48004817962646484,
-      "learning_rate": 9.162084867351842e-05,
-      "loss": 0.6727,
-      "step": 222
-    },
-    {
-      "epoch": 1.1494845360824741,
-      "grad_norm": 0.35861465334892273,
-      "learning_rate": 9.152599136294253e-05,
-      "loss": 0.5481,
-      "step": 223
-    },
-    {
-      "epoch": 1.1546391752577319,
-      "grad_norm": 0.4914831221103668,
-      "learning_rate": 9.14306498313023e-05,
-      "loss": 0.7459,
-      "step": 224
-    },
-    {
-      "epoch": 1.1597938144329896,
-      "grad_norm": 0.4673561751842499,
-      "learning_rate": 9.133482519034428e-05,
-      "loss": 0.7004,
-      "step": 225
-    },
-    {
-      "epoch": 1.1649484536082475,
-      "grad_norm": 0.4495156407356262,
-      "learning_rate": 9.123851855744843e-05,
-      "loss": 0.751,
-      "step": 226
-    },
-    {
-      "epoch": 1.1701030927835052,
-      "grad_norm": 0.4275054335594177,
-      "learning_rate": 9.114173105561501e-05,
-      "loss": 0.7959,
-      "step": 227
-    },
-    {
-      "epoch": 1.175257731958763,
-      "grad_norm": 0.3814803659915924,
-      "learning_rate": 9.104446381345159e-05,
-      "loss": 0.7069,
-      "step": 228
-    },
-    {
-      "epoch": 1.1804123711340206,
-      "grad_norm": 0.3911239206790924,
-      "learning_rate": 9.094671796515978e-05,
-      "loss": 0.627,
-      "step": 229
-    },
-    {
-      "epoch": 1.1855670103092784,
-      "grad_norm": 0.5213860273361206,
-      "learning_rate": 9.08484946505221e-05,
-      "loss": 0.7112,
-      "step": 230
-    },
-    {
-      "epoch": 1.190721649484536,
-      "grad_norm": 0.41048097610473633,
-      "learning_rate": 9.074979501488867e-05,
-      "loss": 0.588,
-      "step": 231
-    },
-    {
-      "epoch": 1.1958762886597938,
-      "grad_norm": 0.40846139192581177,
-      "learning_rate": 9.065062020916377e-05,
-      "loss": 0.6034,
-      "step": 232
-    },
-    {
-      "epoch": 1.2010309278350515,
-      "grad_norm": 0.4402838945388794,
-      "learning_rate": 9.055097138979252e-05,
-      "loss": 0.7185,
-      "step": 233
-    },
-    {
-      "epoch": 1.2061855670103092,
-      "grad_norm": 0.4315119683742523,
-      "learning_rate": 9.045084971874738e-05,
-      "loss": 0.6632,
-      "step": 234
-    },
-    {
-      "epoch": 1.211340206185567,
-      "grad_norm": 0.5140849947929382,
-      "learning_rate": 9.035025636351452e-05,
-      "loss": 0.6369,
-      "step": 235
-    },
-    {
-      "epoch": 1.2164948453608249,
-      "grad_norm": 0.45563650131225586,
-      "learning_rate": 9.024919249708035e-05,
-      "loss": 0.7947,
-      "step": 236
-    },
-    {
-      "epoch": 1.2216494845360826,
-      "grad_norm": 0.5599696636199951,
-      "learning_rate": 9.014765929791768e-05,
-      "loss": 0.7507,
-      "step": 237
-    },
-    {
-      "epoch": 1.2268041237113403,
-      "grad_norm": 0.4429771602153778,
-      "learning_rate": 9.004565794997209e-05,
-      "loss": 0.6726,
-      "step": 238
-    },
-    {
-      "epoch": 1.231958762886598,
-      "grad_norm": 0.3993242084980011,
-      "learning_rate": 8.994318964264809e-05,
-      "loss": 0.6158,
-      "step": 239
-    },
-    {
-      "epoch": 1.2371134020618557,
-      "grad_norm": 0.4936448931694031,
-      "learning_rate": 8.984025557079523e-05,
-      "loss": 0.752,
-      "step": 240
-    },
-    {
-      "epoch": 1.2422680412371134,
-      "grad_norm": 0.5145292282104492,
-      "learning_rate": 8.973685693469423e-05,
-      "loss": 0.7791,
-      "step": 241
-    },
-    {
-      "epoch": 1.2474226804123711,
-      "grad_norm": 0.4600600600242615,
-      "learning_rate": 8.963299494004291e-05,
-      "loss": 0.7412,
-      "step": 242
-    },
-    {
-      "epoch": 1.2525773195876289,
-      "grad_norm": 0.3841022849082947,
-      "learning_rate": 8.952867079794218e-05,
-      "loss": 0.5547,
-      "step": 243
-    },
-    {
-      "epoch": 1.2577319587628866,
-      "grad_norm": 0.42398545145988464,
-      "learning_rate": 8.942388572488187e-05,
-      "loss": 0.8999,
-      "step": 244
-    },
-    {
-      "epoch": 1.2628865979381443,
-      "grad_norm": 0.461775004863739,
-      "learning_rate": 8.931864094272663e-05,
-      "loss": 0.5883,
-      "step": 245
-    },
-    {
-      "epoch": 1.268041237113402,
-      "grad_norm": 0.5585225820541382,
-      "learning_rate": 8.921293767870157e-05,
-      "loss": 0.7303,
-      "step": 246
-    },
-    {
-      "epoch": 1.2731958762886597,
-      "grad_norm": 0.45500585436820984,
-      "learning_rate": 8.910677716537806e-05,
-      "loss": 0.8066,
-      "step": 247
-    },
-    {
-      "epoch": 1.2783505154639174,
-      "grad_norm": 0.49855995178222656,
-      "learning_rate": 8.900016064065923e-05,
-      "loss": 0.7256,
-      "step": 248
-    },
-    {
-      "epoch": 1.2835051546391751,
-      "grad_norm": 0.5525998473167419,
-      "learning_rate": 8.889308934776572e-05,
-      "loss": 0.8411,
-      "step": 249
-    },
-    {
-      "epoch": 1.2886597938144329,
-      "grad_norm": 0.49777719378471375,
-      "learning_rate": 8.8785564535221e-05,
-      "loss": 0.8178,
-      "step": 250
-    },
-    {
-      "epoch": 1.2938144329896908,
-      "grad_norm": 0.48288318514823914,
-      "learning_rate": 8.867758745683687e-05,
-      "loss": 0.8036,
-      "step": 251
-    },
-    {
-      "epoch": 1.2989690721649485,
-      "grad_norm": 0.5231608152389526,
-      "learning_rate": 8.85691593716989e-05,
-      "loss": 0.8048,
-      "step": 252
-    },
-    {
-      "epoch": 1.3041237113402062,
-      "grad_norm": 0.4988687038421631,
-      "learning_rate": 8.84602815441517e-05,
-      "loss": 0.7114,
-      "step": 253
-    },
-    {
-      "epoch": 1.309278350515464,
-      "grad_norm": 0.47835099697113037,
-      "learning_rate": 8.835095524378414e-05,
-      "loss": 0.8064,
-      "step": 254
-    },
-    {
-      "epoch": 1.3144329896907216,
-      "grad_norm": 0.4335271716117859,
-      "learning_rate": 8.824118174541464e-05,
-      "loss": 0.6845,
-      "step": 255
-    },
-    {
-      "epoch": 1.3195876288659794,
-      "grad_norm": 0.5307462215423584,
-      "learning_rate": 8.81309623290762e-05,
-      "loss": 0.8478,
-      "step": 256
-    },
-    {
-      "epoch": 1.324742268041237,
-      "grad_norm": 0.4708554446697235,
-      "learning_rate": 8.802029828000156e-05,
-      "loss": 0.8098,
-      "step": 257
-    },
-    {
-      "epoch": 1.3298969072164948,
-      "grad_norm": 0.4892447590827942,
-      "learning_rate": 8.790919088860814e-05,
-      "loss": 0.7612,
-      "step": 258
-    },
-    {
-      "epoch": 1.3350515463917525,
-      "grad_norm": 0.5239393711090088,
-      "learning_rate": 8.779764145048308e-05,
-      "loss": 0.7195,
-      "step": 259
-    },
-    {
-      "epoch": 1.3402061855670104,
-      "grad_norm": 0.5320796966552734,
-      "learning_rate": 8.768565126636806e-05,
-      "loss": 0.7106,
-      "step": 260
-    },
-    {
-      "epoch": 1.3453608247422681,
-      "grad_norm": 0.48917362093925476,
-      "learning_rate": 8.757322164214413e-05,
-      "loss": 0.6933,
-      "step": 261
-    },
-    {
-      "epoch": 1.3505154639175259,
-      "grad_norm": 0.5254843831062317,
-      "learning_rate": 8.746035388881655e-05,
-      "loss": 0.8247,
-      "step": 262
-    },
-    {
-      "epoch": 1.3556701030927836,
-      "grad_norm": 0.4948481619358063,
-      "learning_rate": 8.734704932249944e-05,
-      "loss": 0.9323,
-      "step": 263
-    },
-    {
-      "epoch": 1.3608247422680413,
-      "grad_norm": 0.5369058847427368,
-      "learning_rate": 8.723330926440045e-05,
-      "loss": 0.8482,
-      "step": 264
-    },
-    {
-      "epoch": 1.365979381443299,
-      "grad_norm": 0.48337045311927795,
-      "learning_rate": 8.711913504080534e-05,
-      "loss": 0.6482,
-      "step": 265
-    },
-    {
-      "epoch": 1.3711340206185567,
-      "grad_norm": 0.4217570722103119,
-      "learning_rate": 8.70045279830626e-05,
-      "loss": 0.6593,
-      "step": 266
-    },
-    {
-      "epoch": 1.3762886597938144,
-      "grad_norm": 0.43802469968795776,
-      "learning_rate": 8.688948942756778e-05,
-      "loss": 0.6567,
-      "step": 267
-    },
-    {
-      "epoch": 1.3814432989690721,
-      "grad_norm": 0.5340169072151184,
-      "learning_rate": 8.677402071574805e-05,
-      "loss": 0.6976,
-      "step": 268
-    },
-    {
-      "epoch": 1.3865979381443299,
-      "grad_norm": 0.5338581800460815,
-      "learning_rate": 8.665812319404643e-05,
-      "loss": 0.6684,
-      "step": 269
-    },
-    {
-      "epoch": 1.3917525773195876,
-      "grad_norm": 0.5673936009407043,
-      "learning_rate": 8.654179821390621e-05,
-      "loss": 0.8238,
-      "step": 270
-    },
-    {
-      "epoch": 1.3969072164948453,
-      "grad_norm": 0.3845212459564209,
-      "learning_rate": 8.642504713175508e-05,
-      "loss": 0.7176,
-      "step": 271
-    },
-    {
-      "epoch": 1.402061855670103,
-      "grad_norm": 0.5132887959480286,
-      "learning_rate": 8.630787130898943e-05,
-      "loss": 0.819,
-      "step": 272
-    },
-    {
-      "epoch": 1.4072164948453607,
-      "grad_norm": 0.4670240879058838,
-      "learning_rate": 8.619027211195836e-05,
-      "loss": 0.759,
-      "step": 273
-    },
-    {
-      "epoch": 1.4123711340206184,
-      "grad_norm": 0.4744627773761749,
-      "learning_rate": 8.607225091194779e-05,
-      "loss": 0.6824,
-      "step": 274
-    },
-    {
-      "epoch": 1.4175257731958764,
-      "grad_norm": 0.4141714870929718,
-      "learning_rate": 8.595380908516454e-05,
-      "loss": 0.5199,
-      "step": 275
-    },
-    {
-      "epoch": 1.422680412371134,
-      "grad_norm": 0.4950573444366455,
-      "learning_rate": 8.583494801272018e-05,
-      "loss": 0.7358,
-      "step": 276
-    },
-    {
-      "epoch": 1.4278350515463918,
-      "grad_norm": 0.42697805166244507,
-      "learning_rate": 8.571566908061497e-05,
-      "loss": 0.6716,
-      "step": 277
-    },
-    {
-      "epoch": 1.4329896907216495,
-      "grad_norm": 0.5071122646331787,
-      "learning_rate": 8.559597367972168e-05,
-      "loss": 0.7642,
-      "step": 278
-    },
-    {
-      "epoch": 1.4381443298969072,
-      "grad_norm": 0.5294380187988281,
-      "learning_rate": 8.547586320576945e-05,
-      "loss": 0.7594,
-      "step": 279
-    },
-    {
-      "epoch": 1.443298969072165,
-      "grad_norm": 0.4704250395298004,
-      "learning_rate": 8.535533905932738e-05,
-      "loss": 0.6536,
-      "step": 280
-    },
-    {
-      "epoch": 1.4484536082474226,
-      "grad_norm": 0.4988890290260315,
-      "learning_rate": 8.52344026457883e-05,
-      "loss": 0.752,
-      "step": 281
-    },
-    {
-      "epoch": 1.4536082474226804,
-      "grad_norm": 0.5117019414901733,
-      "learning_rate": 8.511305537535237e-05,
-      "loss": 0.6501,
-      "step": 282
-    },
-    {
-      "epoch": 1.458762886597938,
-      "grad_norm": 0.48722848296165466,
-      "learning_rate": 8.499129866301057e-05,
-      "loss": 0.6892,
-      "step": 283
-    },
-    {
-      "epoch": 1.463917525773196,
-      "grad_norm": 0.4610782563686371,
-      "learning_rate": 8.48691339285283e-05,
-      "loss": 0.7259,
-      "step": 284
-    },
-    {
-      "epoch": 1.4690721649484537,
-      "grad_norm": 0.5337876081466675,
-      "learning_rate": 8.474656259642873e-05,
-      "loss": 0.7162,
-      "step": 285
-    },
-    {
-      "epoch": 1.4742268041237114,
-      "grad_norm": 0.6001775860786438,
-      "learning_rate": 8.46235860959763e-05,
-      "loss": 0.8537,
-      "step": 286
-    },
-    {
-      "epoch": 1.4793814432989691,
-      "grad_norm": 0.5563547015190125,
-      "learning_rate": 8.450020586115987e-05,
-      "loss": 0.7413,
-      "step": 287
-    },
-    {
-      "epoch": 1.4845360824742269,
-      "grad_norm": 0.49136775732040405,
-      "learning_rate": 8.437642333067625e-05,
-      "loss": 0.7277,
-      "step": 288
-    },
-    {
-      "epoch": 1.4896907216494846,
-      "grad_norm": 0.6892454624176025,
-      "learning_rate": 8.42522399479132e-05,
-      "loss": 0.7893,
-      "step": 289
-    },
-    {
-      "epoch": 1.4948453608247423,
-      "grad_norm": 0.41972535848617554,
-      "learning_rate": 8.412765716093272e-05,
-      "loss": 0.6497,
-      "step": 290
-    },
-    {
-      "epoch": 1.5,
-      "grad_norm": 0.5736141800880432,
-      "learning_rate": 8.40026764224541e-05,
-      "loss": 0.8443,
-      "step": 291
-    },
-    {
-      "epoch": 1.5,
-      "eval_loss": 0.775231659412384,
-      "eval_runtime": 23.1141,
-      "eval_samples_per_second": 7.095,
-      "eval_steps_per_second": 1.774,
-      "step": 291
-    },
-    {
-      "epoch": 1.5051546391752577,
-      "grad_norm": 0.47275158762931824,
-      "learning_rate": 8.387729918983706e-05,
-      "loss": 0.5539,
-      "step": 292
-    },
-    {
-      "epoch": 1.5103092783505154,
-      "grad_norm": 0.5868181586265564,
-      "learning_rate": 8.375152692506468e-05,
-      "loss": 0.9621,
-      "step": 293
-    },
-    {
-      "epoch": 1.5154639175257731,
-      "grad_norm": 0.560245931148529,
-      "learning_rate": 8.362536109472636e-05,
-      "loss": 0.8364,
-      "step": 294
-    },
-    {
-      "epoch": 1.5206185567010309,
-      "grad_norm": 0.5294224619865417,
-      "learning_rate": 8.349880317000082e-05,
-      "loss": 0.6673,
-      "step": 295
-    },
-    {
-      "epoch": 1.5257731958762886,
-      "grad_norm": 0.47846174240112305,
-      "learning_rate": 8.337185462663878e-05,
-      "loss": 0.6843,
-      "step": 296
-    },
-    {
-      "epoch": 1.5309278350515463,
-      "grad_norm": 0.5413031578063965,
-      "learning_rate": 8.32445169449459e-05,
-      "loss": 0.9445,
-      "step": 297
-    },
-    {
-      "epoch": 1.536082474226804,
-      "grad_norm": 0.5185613036155701,
-      "learning_rate": 8.311679160976539e-05,
-      "loss": 0.8915,
-      "step": 298
-    },
-    {
-      "epoch": 1.5412371134020617,
-      "grad_norm": 0.5198041796684265,
-      "learning_rate": 8.29886801104608e-05,
-      "loss": 0.7735,
-      "step": 299
-    },
-    {
-      "epoch": 1.5463917525773194,
-      "grad_norm": 0.5014578104019165,
-      "learning_rate": 8.286018394089863e-05,
-      "loss": 0.7663,
-      "step": 300
-    },
-    {
-      "epoch": 1.5515463917525774,
-      "grad_norm": 0.5157457590103149,
-      "learning_rate": 8.273130459943086e-05,
-      "loss": 0.741,
-      "step": 301
-    },
-    {
-      "epoch": 1.556701030927835,
-      "grad_norm": 0.6345148086547852,
-      "learning_rate": 8.260204358887754e-05,
-      "loss": 0.9313,
-      "step": 302
-    },
-    {
-      "epoch": 1.5618556701030928,
-      "grad_norm": 0.46231186389923096,
-      "learning_rate": 8.247240241650918e-05,
-      "loss": 0.6452,
-      "step": 303
-    },
-    {
-      "epoch": 1.5670103092783505,
-      "grad_norm": 0.5200813412666321,
-      "learning_rate": 8.234238259402935e-05,
-      "loss": 0.723,
-      "step": 304
-    },
-    {
-      "epoch": 1.5721649484536082,
-      "grad_norm": 0.49244582653045654,
-      "learning_rate": 8.221198563755682e-05,
-      "loss": 0.6612,
-      "step": 305
-    },
-    {
-      "epoch": 1.577319587628866,
-      "grad_norm": 0.4823918640613556,
-      "learning_rate": 8.208121306760805e-05,
-      "loss": 0.7428,
-      "step": 306
-    },
-    {
-      "epoch": 1.5824742268041239,
-      "grad_norm": 0.4022305905818939,
-      "learning_rate": 8.195006640907942e-05,
-      "loss": 0.5851,
-      "step": 307
-    },
-    {
-      "epoch": 1.5876288659793816,
-      "grad_norm": 0.5390604734420776,
-      "learning_rate": 8.181854719122939e-05,
-      "loss": 0.7401,
-      "step": 308
-    },
-    {
-      "epoch": 1.5927835051546393,
-      "grad_norm": 0.5512533783912659,
-      "learning_rate": 8.168665694766073e-05,
-      "loss": 0.8259,
-      "step": 309
-    },
-    {
-      "epoch": 1.597938144329897,
-      "grad_norm": 0.4972156882286072,
-      "learning_rate": 8.155439721630264e-05,
-      "loss": 0.7784,
-      "step": 310
-    },
-    {
-      "epoch": 1.6030927835051547,
-      "grad_norm": 0.4425565302371979,
-      "learning_rate": 8.142176953939279e-05,
-      "loss": 0.6065,
-      "step": 311
-    },
-    {
-      "epoch": 1.6082474226804124,
-      "grad_norm": 0.4272889196872711,
-      "learning_rate": 8.128877546345933e-05,
-      "loss": 0.5773,
-      "step": 312
-    },
-    {
-      "epoch": 1.6134020618556701,
-      "grad_norm": 0.5591496229171753,
-      "learning_rate": 8.115541653930286e-05,
-      "loss": 0.6459,
-      "step": 313
-    },
-    {
-      "epoch": 1.6185567010309279,
-      "grad_norm": 0.4658234119415283,
-      "learning_rate": 8.102169432197842e-05,
-      "loss": 0.7003,
-      "step": 314
-    },
-    {
-      "epoch": 1.6237113402061856,
-      "grad_norm": 0.4371519684791565,
-      "learning_rate": 8.088761037077718e-05,
-      "loss": 0.8428,
-      "step": 315
-    },
-    {
-      "epoch": 1.6288659793814433,
-      "grad_norm": 0.40482646226882935,
-      "learning_rate": 8.075316624920848e-05,
-      "loss": 0.5668,
-      "step": 316
-    },
-    {
-      "epoch": 1.634020618556701,
-      "grad_norm": 0.5180542469024658,
-      "learning_rate": 8.061836352498145e-05,
-      "loss": 0.7009,
-      "step": 317
-    },
-    {
-      "epoch": 1.6391752577319587,
-      "grad_norm": 0.5090633630752563,
-      "learning_rate": 8.048320376998673e-05,
-      "loss": 0.6369,
-      "step": 318
-    },
-    {
-      "epoch": 1.6443298969072164,
-      "grad_norm": 0.5216387510299683,
-      "learning_rate": 8.034768856027826e-05,
-      "loss": 0.8043,
-      "step": 319
-    },
-    {
-      "epoch": 1.6494845360824741,
-      "grad_norm": 0.6093303561210632,
-      "learning_rate": 8.021181947605473e-05,
-      "loss": 0.793,
-      "step": 320
-    },
-    {
-      "epoch": 1.6546391752577319,
-      "grad_norm": 0.477040559053421,
-      "learning_rate": 8.007559810164133e-05,
-      "loss": 0.7163,
-      "step": 321
-    },
-    {
-      "epoch": 1.6597938144329896,
-      "grad_norm": 0.5674880146980286,
-      "learning_rate": 7.993902602547113e-05,
-      "loss": 0.6498,
-      "step": 322
-    },
-    {
-      "epoch": 1.6649484536082473,
-      "grad_norm": 0.5792298913002014,
-      "learning_rate": 7.980210484006666e-05,
-      "loss": 0.7753,
-      "step": 323
-    },
-    {
-      "epoch": 1.670103092783505,
-      "grad_norm": 0.4890528619289398,
-      "learning_rate": 7.966483614202128e-05,
-      "loss": 0.79,
-      "step": 324
-    },
-    {
-      "epoch": 1.675257731958763,
-      "grad_norm": 0.4526829719543457,
-      "learning_rate": 7.952722153198054e-05,
-      "loss": 0.572,
-      "step": 325
-    },
-    {
-      "epoch": 1.6804123711340206,
-      "grad_norm": 0.5673860907554626,
-      "learning_rate": 7.938926261462366e-05,
-      "loss": 0.706,
-      "step": 326
-    },
-    {
-      "epoch": 1.6855670103092784,
-      "grad_norm": 0.4477851092815399,
-      "learning_rate": 7.925096099864464e-05,
-      "loss": 0.553,
-      "step": 327
-    },
-    {
-      "epoch": 1.690721649484536,
-      "grad_norm": 0.5353184342384338,
-      "learning_rate": 7.911231829673356e-05,
-      "loss": 0.7405,
-      "step": 328
-    },
-    {
-      "epoch": 1.6958762886597938,
-      "grad_norm": 0.4556489586830139,
-      "learning_rate": 7.897333612555785e-05,
-      "loss": 0.5462,
-      "step": 329
-    },
-    {
-      "epoch": 1.7010309278350515,
-      "grad_norm": 0.5018810033798218,
-      "learning_rate": 7.883401610574336e-05,
-      "loss": 0.7072,
-      "step": 330
-    },
-    {
-      "epoch": 1.7061855670103094,
-      "grad_norm": 0.5862541794776917,
-      "learning_rate": 7.869435986185547e-05,
-      "loss": 0.8802,
-      "step": 331
-    },
-    {
-      "epoch": 1.7113402061855671,
-      "grad_norm": 0.491769939661026,
-      "learning_rate": 7.855436902238017e-05,
-      "loss": 0.832,
-      "step": 332
-    },
-    {
-      "epoch": 1.7164948453608249,
-      "grad_norm": 0.46636489033699036,
-      "learning_rate": 7.841404521970505e-05,
-      "loss": 0.5585,
-      "step": 333
-    },
-    {
-      "epoch": 1.7216494845360826,
-      "grad_norm": 0.4456227123737335,
-      "learning_rate": 7.82733900901003e-05,
-      "loss": 0.4713,
-      "step": 334
-    },
-    {
-      "epoch": 1.7268041237113403,
-      "grad_norm": 0.4480153024196625,
-      "learning_rate": 7.813240527369959e-05,
-      "loss": 0.6575,
-      "step": 335
-    },
-    {
-      "epoch": 1.731958762886598,
-      "grad_norm": 0.4899657368659973,
-      "learning_rate": 7.799109241448091e-05,
-      "loss": 0.6507,
-      "step": 336
-    },
-    {
-      "epoch": 1.7371134020618557,
-      "grad_norm": 0.552197277545929,
-      "learning_rate": 7.784945316024756e-05,
-      "loss": 0.7384,
-      "step": 337
-    },
-    {
-      "epoch": 1.7422680412371134,
-      "grad_norm": 0.4821660816669464,
-      "learning_rate": 7.770748916260875e-05,
-      "loss": 0.72,
-      "step": 338
-    },
-    {
-      "epoch": 1.7474226804123711,
-      "grad_norm": 0.5010591745376587,
-      "learning_rate": 7.756520207696041e-05,
-      "loss": 0.7132,
-      "step": 339
-    },
-    {
-      "epoch": 1.7525773195876289,
-      "grad_norm": 0.4646522104740143,
-      "learning_rate": 7.742259356246593e-05,
-      "loss": 0.6314,
-      "step": 340
-    },
-    {
-      "epoch": 1.7577319587628866,
-      "grad_norm": 0.48994216322898865,
-      "learning_rate": 7.727966528203678e-05,
-      "loss": 0.5914,
-      "step": 341
-    },
-    {
-      "epoch": 1.7628865979381443,
-      "grad_norm": 0.4667370319366455,
-      "learning_rate": 7.71364189023131e-05,
-      "loss": 0.7351,
-      "step": 342
-    },
-    {
-      "epoch": 1.768041237113402,
-      "grad_norm": 0.5273435115814209,
-      "learning_rate": 7.699285609364424e-05,
-      "loss": 0.7077,
-      "step": 343
-    },
-    {
-      "epoch": 1.7731958762886597,
-      "grad_norm": 0.6191979050636292,
-      "learning_rate": 7.68489785300694e-05,
-      "loss": 0.8277,
-      "step": 344
-    },
-    {
-      "epoch": 1.7783505154639174,
-      "grad_norm": 0.5085694789886475,
-      "learning_rate": 7.670478788929802e-05,
-      "loss": 0.8608,
-      "step": 345
-    },
-    {
-      "epoch": 1.7835051546391751,
-      "grad_norm": 0.47608375549316406,
-      "learning_rate": 7.656028585269018e-05,
-      "loss": 0.6082,
-      "step": 346
-    },
-    {
-      "epoch": 1.7886597938144329,
-      "grad_norm": 0.5139414072036743,
-      "learning_rate": 7.641547410523709e-05,
-      "loss": 0.6301,
-      "step": 347
-    },
-    {
-      "epoch": 1.7938144329896906,
-      "grad_norm": 0.45298558473587036,
-      "learning_rate": 7.627035433554138e-05,
-      "loss": 0.695,
-      "step": 348
-    },
-    {
-      "epoch": 1.7989690721649485,
-      "grad_norm": 0.5817694664001465,
-      "learning_rate": 7.612492823579745e-05,
-      "loss": 0.7001,
-      "step": 349
-    },
-    {
-      "epoch": 1.8041237113402062,
-      "grad_norm": 0.481673926115036,
-      "learning_rate": 7.597919750177168e-05,
-      "loss": 0.6867,
-      "step": 350
-    },
-    {
-      "epoch": 1.809278350515464,
-      "grad_norm": 0.5240175724029541,
-      "learning_rate": 7.583316383278273e-05,
-      "loss": 0.6443,
-      "step": 351
-    },
-    {
-      "epoch": 1.8144329896907216,
-      "grad_norm": 0.5637550950050354,
-      "learning_rate": 7.568682893168164e-05,
-      "loss": 0.7145,
-      "step": 352
-    },
-    {
-      "epoch": 1.8195876288659794,
-      "grad_norm": 0.4581276476383209,
-      "learning_rate": 7.554019450483208e-05,
-      "loss": 0.7877,
-      "step": 353
-    },
-    {
-      "epoch": 1.824742268041237,
-      "grad_norm": 0.5422832369804382,
-      "learning_rate": 7.539326226209031e-05,
-      "loss": 0.6451,
-      "step": 354
-    },
-    {
-      "epoch": 1.829896907216495,
-      "grad_norm": 0.45504269003868103,
-      "learning_rate": 7.524603391678541e-05,
-      "loss": 0.816,
-      "step": 355
-    },
-    {
-      "epoch": 1.8350515463917527,
-      "grad_norm": 0.5947327017784119,
-      "learning_rate": 7.509851118569915e-05,
-      "loss": 0.674,
-      "step": 356
-    },
-    {
-      "epoch": 1.8402061855670104,
-      "grad_norm": 0.38084176182746887,
-      "learning_rate": 7.495069578904608e-05,
-      "loss": 0.5898,
-      "step": 357
-    },
-    {
-      "epoch": 1.8453608247422681,
-      "grad_norm": 0.4741254448890686,
-      "learning_rate": 7.48025894504534e-05,
-      "loss": 0.6011,
-      "step": 358
-    },
-    {
-      "epoch": 1.8505154639175259,
-      "grad_norm": 0.5517462491989136,
-      "learning_rate": 7.465419389694092e-05,
-      "loss": 0.7738,
-      "step": 359
-    },
-    {
-      "epoch": 1.8556701030927836,
-      "grad_norm": 0.5621674656867981,
-      "learning_rate": 7.450551085890087e-05,
-      "loss": 0.8294,
-      "step": 360
-    },
-    {
-      "epoch": 1.8608247422680413,
-      "grad_norm": 0.4534311592578888,
-      "learning_rate": 7.435654207007773e-05,
-      "loss": 0.635,
-      "step": 361
-    },
-    {
-      "epoch": 1.865979381443299,
-      "grad_norm": 0.5818427205085754,
-      "learning_rate": 7.420728926754803e-05,
-      "loss": 0.824,
-      "step": 362
-    },
-    {
-      "epoch": 1.8711340206185567,
-      "grad_norm": 0.43529012799263,
-      "learning_rate": 7.405775419170014e-05,
-      "loss": 0.5617,
-      "step": 363
-    },
-    {
-      "epoch": 1.8762886597938144,
-      "grad_norm": 0.39464038610458374,
-      "learning_rate": 7.390793858621386e-05,
-      "loss": 0.6568,
-      "step": 364
-    },
-    {
-      "epoch": 1.8814432989690721,
-      "grad_norm": 0.5439510941505432,
-      "learning_rate": 7.375784419804019e-05,
-      "loss": 0.6493,
-      "step": 365
-    },
-    {
-      "epoch": 1.8865979381443299,
-      "grad_norm": 0.44549572467803955,
-      "learning_rate": 7.360747277738094e-05,
-      "loss": 0.6491,
-      "step": 366
-    },
-    {
-      "epoch": 1.8917525773195876,
-      "grad_norm": 0.5053935050964355,
-      "learning_rate": 7.345682607766826e-05,
-      "loss": 0.8553,
-      "step": 367
-    },
-    {
-      "epoch": 1.8969072164948453,
-      "grad_norm": 0.4746989905834198,
-      "learning_rate": 7.330590585554428e-05,
-      "loss": 0.7245,
-      "step": 368
-    },
-    {
-      "epoch": 1.902061855670103,
-      "grad_norm": 0.3898842930793762,
-      "learning_rate": 7.315471387084056e-05,
-      "loss": 0.5421,
-      "step": 369
-    },
-    {
-      "epoch": 1.9072164948453607,
-      "grad_norm": 0.5751209855079651,
-      "learning_rate": 7.300325188655761e-05,
-      "loss": 0.815,
-      "step": 370
-    },
-    {
-      "epoch": 1.9123711340206184,
-      "grad_norm": 0.5123909115791321,
-      "learning_rate": 7.285152166884432e-05,
-      "loss": 0.7358,
-      "step": 371
-    },
-    {
-      "epoch": 1.9175257731958761,
-      "grad_norm": 0.47217753529548645,
-      "learning_rate": 7.269952498697734e-05,
-      "loss": 0.7771,
-      "step": 372
-    },
-    {
-      "epoch": 1.922680412371134,
-      "grad_norm": 0.5343344807624817,
-      "learning_rate": 7.25472636133405e-05,
-      "loss": 0.6233,
-      "step": 373
-    },
-    {
-      "epoch": 1.9278350515463918,
-      "grad_norm": 0.5458410978317261,
-      "learning_rate": 7.23947393234041e-05,
-      "loss": 0.8072,
-      "step": 374
-    },
-    {
-      "epoch": 1.9329896907216495,
-      "grad_norm": 0.6140435338020325,
-      "learning_rate": 7.224195389570422e-05,
-      "loss": 0.7253,
-      "step": 375
-    },
-    {
-      "epoch": 1.9381443298969072,
-      "grad_norm": 0.6178478598594666,
-      "learning_rate": 7.208890911182197e-05,
-      "loss": 0.9358,
-      "step": 376
-    },
-    {
-      "epoch": 1.943298969072165,
-      "grad_norm": 0.5554368495941162,
-      "learning_rate": 7.193560675636277e-05,
-      "loss": 0.8622,
-      "step": 377
-    },
-    {
-      "epoch": 1.9484536082474226,
-      "grad_norm": 0.5233482718467712,
-      "learning_rate": 7.178204861693545e-05,
-      "loss": 0.768,
-      "step": 378
-    },
-    {
-      "epoch": 1.9536082474226806,
-      "grad_norm": 0.4429621994495392,
-      "learning_rate": 7.162823648413151e-05,
-      "loss": 0.6282,
-      "step": 379
-    },
-    {
-      "epoch": 1.9587628865979383,
-      "grad_norm": 0.44777798652648926,
-      "learning_rate": 7.14741721515041e-05,
-      "loss": 0.6284,
-      "step": 380
-    },
-    {
-      "epoch": 1.963917525773196,
-      "grad_norm": 0.4569181501865387,
-      "learning_rate": 7.131985741554728e-05,
-      "loss": 0.6296,
-      "step": 381
-    },
-    {
-      "epoch": 1.9690721649484537,
-      "grad_norm": 0.5347330570220947,
-      "learning_rate": 7.116529407567489e-05,
-      "loss": 0.7547,
-      "step": 382
-    },
-    {
-      "epoch": 1.9742268041237114,
-      "grad_norm": 0.5041331052780151,
-      "learning_rate": 7.101048393419977e-05,
-      "loss": 0.6748,
-      "step": 383
-    },
-    {
-      "epoch": 1.9793814432989691,
-      "grad_norm": 0.487981915473938,
-      "learning_rate": 7.085542879631253e-05,
-      "loss": 0.7449,
-      "step": 384
-    },
-    {
-      "epoch": 1.9845360824742269,
-      "grad_norm": 0.4471663236618042,
-      "learning_rate": 7.070013047006068e-05,
-      "loss": 0.6517,
-      "step": 385
-    },
-    {
-      "epoch": 1.9896907216494846,
-      "grad_norm": 0.4358983039855957,
-      "learning_rate": 7.054459076632743e-05,
-      "loss": 0.7218,
-      "step": 386
-    },
-    {
-      "epoch": 1.9948453608247423,
-      "grad_norm": 0.4451632797718048,
-      "learning_rate": 7.038881149881058e-05,
-      "loss": 0.6756,
-      "step": 387
-    },
-    {
-      "epoch": 2.0,
-      "grad_norm": 0.6731226444244385,
-      "learning_rate": 7.02327944840015e-05,
-      "loss": 0.7974,
-      "step": 388
-    },
-    {
-      "epoch": 2.0,
-      "eval_loss": 0.7645891904830933,
-      "eval_runtime": 23.115,
-      "eval_samples_per_second": 7.095,
-      "eval_steps_per_second": 1.774,
-      "step": 388
-    },
-    {
-      "epoch": 2.0051546391752577,
-      "grad_norm": 0.46283644437789917,
-      "learning_rate": 7.007654154116377e-05,
-      "loss": 0.6405,
-      "step": 389
-    },
-    {
-      "epoch": 2.0103092783505154,
-      "grad_norm": 0.4309590756893158,
-      "learning_rate": 6.992005449231208e-05,
-      "loss": 0.6401,
-      "step": 390
-    },
-    {
-      "epoch": 2.015463917525773,
-      "grad_norm": 0.5006699562072754,
-      "learning_rate": 6.976333516219096e-05,
-      "loss": 0.6664,
-      "step": 391
-    },
-    {
-      "epoch": 2.020618556701031,
-      "grad_norm": 0.5041174292564392,
-      "learning_rate": 6.960638537825352e-05,
-      "loss": 0.587,
-      "step": 392
-    },
-    {
-      "epoch": 2.0257731958762886,
-      "grad_norm": 0.4952574670314789,
-      "learning_rate": 6.944920697064004e-05,
-      "loss": 0.7003,
-      "step": 393
-    },
-    {
-      "epoch": 2.0309278350515463,
-      "grad_norm": 0.5484703183174133,
-      "learning_rate": 6.929180177215678e-05,
-      "loss": 0.6198,
-      "step": 394
-    },
-    {
-      "epoch": 2.036082474226804,
-      "grad_norm": 0.4701229929924011,
-      "learning_rate": 6.91341716182545e-05,
-      "loss": 0.6247,
-      "step": 395
-    },
-    {
-      "epoch": 2.0412371134020617,
-      "grad_norm": 0.5044362545013428,
-      "learning_rate": 6.897631834700709e-05,
-      "loss": 0.5468,
-      "step": 396
-    },
-    {
-      "epoch": 2.0463917525773194,
-      "grad_norm": 0.5583397746086121,
-      "learning_rate": 6.881824379909017e-05,
-      "loss": 0.6784,
-      "step": 397
-    },
-    {
-      "epoch": 2.051546391752577,
-      "grad_norm": 0.43878433108329773,
-      "learning_rate": 6.865994981775957e-05,
-      "loss": 0.5321,
-      "step": 398
-    },
-    {
-      "epoch": 2.056701030927835,
-      "grad_norm": 0.5170990228652954,
-      "learning_rate": 6.850143824882986e-05,
-      "loss": 0.5962,
-      "step": 399
-    },
-    {
-      "epoch": 2.0618556701030926,
-      "grad_norm": 0.5732204914093018,
-      "learning_rate": 6.834271094065283e-05,
-      "loss": 0.7051,
-      "step": 400
-    },
-    {
-      "epoch": 2.0670103092783507,
-      "grad_norm": 0.5963968634605408,
-      "learning_rate": 6.818376974409593e-05,
-      "loss": 0.6314,
-      "step": 401
-    },
-    {
-      "epoch": 2.0721649484536084,
-      "grad_norm": 0.5836310386657715,
-      "learning_rate": 6.802461651252073e-05,
-      "loss": 0.729,
-      "step": 402
-    },
-    {
-      "epoch": 2.077319587628866,
-      "grad_norm": 0.49259456992149353,
-      "learning_rate": 6.786525310176123e-05,
-      "loss": 0.6406,
-      "step": 403
-    },
-    {
-      "epoch": 2.082474226804124,
-      "grad_norm": 0.5154966115951538,
-      "learning_rate": 6.770568137010226e-05,
-      "loss": 0.6839,
-      "step": 404
-    },
-    {
-      "epoch": 2.0876288659793816,
-      "grad_norm": 0.5059208869934082,
-      "learning_rate": 6.754590317825785e-05,
-      "loss": 0.6115,
-      "step": 405
-    },
-    {
-      "epoch": 2.0927835051546393,
-      "grad_norm": 0.6629896759986877,
-      "learning_rate": 6.738592038934946e-05,
-      "loss": 0.7747,
-      "step": 406
-    },
-    {
-      "epoch": 2.097938144329897,
-      "grad_norm": 0.5021480321884155,
-      "learning_rate": 6.722573486888427e-05,
-      "loss": 0.6272,
-      "step": 407
-    },
-    {
-      "epoch": 2.1030927835051547,
-      "grad_norm": 0.5166428089141846,
-      "learning_rate": 6.706534848473352e-05,
-      "loss": 0.7801,
-      "step": 408
-    },
-    {
-      "epoch": 2.1082474226804124,
-      "grad_norm": 0.6196724772453308,
-      "learning_rate": 6.69047631071106e-05,
-      "loss": 0.7115,
-      "step": 409
-    },
-    {
-      "epoch": 2.11340206185567,
-      "grad_norm": 0.5208120346069336,
-      "learning_rate": 6.674398060854931e-05,
-      "loss": 0.7663,
-      "step": 410
-    },
-    {
-      "epoch": 2.118556701030928,
-      "grad_norm": 0.5327072739601135,
-      "learning_rate": 6.658300286388203e-05,
-      "loss": 0.6416,
-      "step": 411
-    },
-    {
-      "epoch": 2.1237113402061856,
-      "grad_norm": 0.3883470892906189,
-      "learning_rate": 6.642183175021779e-05,
-      "loss": 0.5093,
-      "step": 412
-    },
-    {
-      "epoch": 2.1288659793814433,
-      "grad_norm": 0.6093575358390808,
-      "learning_rate": 6.62604691469205e-05,
-      "loss": 0.7734,
-      "step": 413
-    },
-    {
-      "epoch": 2.134020618556701,
-      "grad_norm": 0.7094951868057251,
-      "learning_rate": 6.609891693558692e-05,
-      "loss": 0.536,
-      "step": 414
-    },
-    {
-      "epoch": 2.1391752577319587,
-      "grad_norm": 0.6357293725013733,
-      "learning_rate": 6.59371770000248e-05,
-      "loss": 0.7157,
-      "step": 415
-    },
-    {
-      "epoch": 2.1443298969072164,
-      "grad_norm": 0.5292159914970398,
-      "learning_rate": 6.577525122623084e-05,
-      "loss": 0.6215,
-      "step": 416
-    },
-    {
-      "epoch": 2.149484536082474,
-      "grad_norm": 0.5167898535728455,
-      "learning_rate": 6.561314150236882e-05,
-      "loss": 0.5474,
-      "step": 417
-    },
-    {
-      "epoch": 2.154639175257732,
-      "grad_norm": 0.6716488599777222,
-      "learning_rate": 6.545084971874738e-05,
-      "loss": 0.6841,
-      "step": 418
-    },
-    {
-      "epoch": 2.1597938144329896,
-      "grad_norm": 0.6054789423942566,
-      "learning_rate": 6.528837776779819e-05,
-      "loss": 0.5964,
-      "step": 419
-    },
-    {
-      "epoch": 2.1649484536082473,
-      "grad_norm": 0.5557959675788879,
-      "learning_rate": 6.51257275440538e-05,
-      "loss": 0.6791,
-      "step": 420
-    },
-    {
-      "epoch": 2.170103092783505,
-      "grad_norm": 0.4804280400276184,
-      "learning_rate": 6.496290094412546e-05,
-      "loss": 0.5954,
-      "step": 421
-    },
-    {
-      "epoch": 2.1752577319587627,
-      "grad_norm": 0.6319596767425537,
-      "learning_rate": 6.479989986668118e-05,
-      "loss": 0.6853,
-      "step": 422
-    },
-    {
-      "epoch": 2.1804123711340204,
-      "grad_norm": 0.5729424953460693,
-      "learning_rate": 6.463672621242342e-05,
-      "loss": 0.5408,
-      "step": 423
-    },
-    {
-      "epoch": 2.1855670103092786,
-      "grad_norm": 0.5444086790084839,
-      "learning_rate": 6.447338188406704e-05,
-      "loss": 0.7406,
-      "step": 424
-    },
-    {
-      "epoch": 2.1907216494845363,
-      "grad_norm": 0.5574979782104492,
-      "learning_rate": 6.430986878631707e-05,
-      "loss": 0.521,
-      "step": 425
-    },
-    {
-      "epoch": 2.195876288659794,
-      "grad_norm": 0.569279134273529,
-      "learning_rate": 6.41461888258465e-05,
-      "loss": 0.5863,
-      "step": 426
-    },
-    {
-      "epoch": 2.2010309278350517,
-      "grad_norm": 0.6239138245582581,
-      "learning_rate": 6.398234391127406e-05,
-      "loss": 0.8244,
-      "step": 427
-    },
-    {
-      "epoch": 2.2061855670103094,
-      "grad_norm": 0.5632001757621765,
-      "learning_rate": 6.381833595314195e-05,
-      "loss": 0.6423,
-      "step": 428
-    },
-    {
-      "epoch": 2.211340206185567,
-      "grad_norm": 0.6038775444030762,
-      "learning_rate": 6.365416686389358e-05,
-      "loss": 0.7034,
-      "step": 429
-    },
-    {
-      "epoch": 2.216494845360825,
-      "grad_norm": 0.45070335268974304,
-      "learning_rate": 6.348983855785121e-05,
-      "loss": 0.5302,
-      "step": 430
-    },
-    {
-      "epoch": 2.2216494845360826,
-      "grad_norm": 0.5909480452537537,
-      "learning_rate": 6.332535295119377e-05,
-      "loss": 0.6037,
-      "step": 431
-    },
-    {
-      "epoch": 2.2268041237113403,
-      "grad_norm": 0.5120466351509094,
-      "learning_rate": 6.31607119619343e-05,
-      "loss": 0.5515,
-      "step": 432
-    },
-    {
-      "epoch": 2.231958762886598,
-      "grad_norm": 0.5868075489997864,
-      "learning_rate": 6.299591750989779e-05,
-      "loss": 0.6891,
-      "step": 433
-    },
-    {
-      "epoch": 2.2371134020618557,
-      "grad_norm": 0.5526292324066162,
-      "learning_rate": 6.283097151669869e-05,
-      "loss": 0.6259,
-      "step": 434
-    },
-    {
-      "epoch": 2.2422680412371134,
-      "grad_norm": 0.6284964084625244,
-      "learning_rate": 6.266587590571852e-05,
-      "loss": 0.8128,
-      "step": 435
-    },
-    {
-      "epoch": 2.247422680412371,
-      "grad_norm": 0.6394625306129456,
-      "learning_rate": 6.250063260208346e-05,
-      "loss": 0.7159,
-      "step": 436
-    },
-    {
-      "epoch": 2.252577319587629,
-      "grad_norm": 0.4809950590133667,
-      "learning_rate": 6.233524353264187e-05,
-      "loss": 0.5191,
-      "step": 437
-    },
-    {
-      "epoch": 2.2577319587628866,
-      "grad_norm": 0.5848096609115601,
-      "learning_rate": 6.216971062594179e-05,
-      "loss": 0.6013,
-      "step": 438
-    },
-    {
-      "epoch": 2.2628865979381443,
-      "grad_norm": 0.6152583360671997,
-      "learning_rate": 6.200403581220861e-05,
-      "loss": 0.5314,
-      "step": 439
-    },
-    {
-      "epoch": 2.268041237113402,
-      "grad_norm": 0.5509049296379089,
-      "learning_rate": 6.183822102332234e-05,
-      "loss": 0.6,
-      "step": 440
-    },
-    {
-      "epoch": 2.2731958762886597,
-      "grad_norm": 0.611883282661438,
-      "learning_rate": 6.167226819279528e-05,
-      "loss": 0.7169,
-      "step": 441
-    },
-    {
-      "epoch": 2.2783505154639174,
-      "grad_norm": 0.6232669353485107,
-      "learning_rate": 6.150617925574933e-05,
-      "loss": 0.7636,
-      "step": 442
-    },
-    {
-      "epoch": 2.283505154639175,
-      "grad_norm": 0.5470356345176697,
-      "learning_rate": 6.13399561488935e-05,
-      "loss": 0.6322,
-      "step": 443
-    },
-    {
-      "epoch": 2.288659793814433,
-      "grad_norm": 0.6054553389549255,
-      "learning_rate": 6.117360081050136e-05,
-      "loss": 0.6398,
-      "step": 444
-    },
-    {
-      "epoch": 2.2938144329896906,
-      "grad_norm": 0.7440930604934692,
-      "learning_rate": 6.1007115180388285e-05,
-      "loss": 0.7202,
-      "step": 445
-    },
-    {
-      "epoch": 2.2989690721649483,
-      "grad_norm": 0.6226584911346436,
-      "learning_rate": 6.0840501199889046e-05,
-      "loss": 0.573,
-      "step": 446
-    },
-    {
-      "epoch": 2.304123711340206,
-      "grad_norm": 0.6319738030433655,
-      "learning_rate": 6.067376081183499e-05,
-      "loss": 0.6071,
-      "step": 447
-    },
-    {
-      "epoch": 2.3092783505154637,
-      "grad_norm": 0.5800845623016357,
-      "learning_rate": 6.050689596053151e-05,
-      "loss": 0.6816,
-      "step": 448
-    },
-    {
-      "epoch": 2.3144329896907214,
-      "grad_norm": 0.5708872079849243,
-      "learning_rate": 6.0339908591735296e-05,
-      "loss": 0.6731,
-      "step": 449
-    },
-    {
-      "epoch": 2.319587628865979,
-      "grad_norm": 0.5780487060546875,
-      "learning_rate": 6.01728006526317e-05,
-      "loss": 0.5966,
-      "step": 450
-    },
-    {
-      "epoch": 2.3247422680412373,
-      "grad_norm": 0.6469062566757202,
-      "learning_rate": 6.0005574091811964e-05,
-      "loss": 0.5754,
-      "step": 451
-    },
-    {
-      "epoch": 2.329896907216495,
-      "grad_norm": 0.5494647026062012,
-      "learning_rate": 5.9838230859250586e-05,
-      "loss": 0.5701,
-      "step": 452
-    },
-    {
-      "epoch": 2.3350515463917527,
-      "grad_norm": 0.5937216877937317,
-      "learning_rate": 5.967077290628249e-05,
-      "loss": 0.6324,
-      "step": 453
-    },
-    {
-      "epoch": 2.3402061855670104,
-      "grad_norm": 0.5645222663879395,
-      "learning_rate": 5.950320218558037e-05,
-      "loss": 0.639,
-      "step": 454
-    },
-    {
-      "epoch": 2.345360824742268,
-      "grad_norm": 0.7451580762863159,
-      "learning_rate": 5.9335520651131814e-05,
-      "loss": 0.607,
-      "step": 455
-    },
-    {
-      "epoch": 2.350515463917526,
-      "grad_norm": 0.5033421516418457,
-      "learning_rate": 5.9167730258216627e-05,
-      "loss": 0.5358,
-      "step": 456
-    },
-    {
-      "epoch": 2.3556701030927836,
-      "grad_norm": 0.6781536936759949,
-      "learning_rate": 5.899983296338392e-05,
-      "loss": 0.6033,
-      "step": 457
-    },
-    {
-      "epoch": 2.3608247422680413,
-      "grad_norm": 0.6157387495040894,
-      "learning_rate": 5.8831830724429384e-05,
-      "loss": 0.5855,
-      "step": 458
-    },
-    {
-      "epoch": 2.365979381443299,
-      "grad_norm": 0.6350586414337158,
-      "learning_rate": 5.866372550037242e-05,
-      "loss": 0.5524,
-      "step": 459
-    },
-    {
-      "epoch": 2.3711340206185567,
-      "grad_norm": 0.5306968688964844,
-      "learning_rate": 5.849551925143334e-05,
-      "loss": 0.6551,
-      "step": 460
-    },
-    {
-      "epoch": 2.3762886597938144,
-      "grad_norm": 0.6288350820541382,
-      "learning_rate": 5.8327213939010414e-05,
-      "loss": 0.6348,
-      "step": 461
-    },
-    {
-      "epoch": 2.381443298969072,
-      "grad_norm": 0.530822217464447,
-      "learning_rate": 5.815881152565712e-05,
-      "loss": 0.6298,
-      "step": 462
-    },
-    {
-      "epoch": 2.38659793814433,
-      "grad_norm": 0.5995349287986755,
-      "learning_rate": 5.799031397505913e-05,
-      "loss": 0.6262,
-      "step": 463
-    },
-    {
-      "epoch": 2.3917525773195876,
-      "grad_norm": 0.512988805770874,
-      "learning_rate": 5.782172325201155e-05,
-      "loss": 0.615,
-      "step": 464
-    },
-    {
-      "epoch": 2.3969072164948453,
-      "grad_norm": 0.5762256979942322,
-      "learning_rate": 5.7653041322395895e-05,
-      "loss": 0.5785,
-      "step": 465
-    },
-    {
-      "epoch": 2.402061855670103,
-      "grad_norm": 0.6673002243041992,
-      "learning_rate": 5.748427015315722e-05,
-      "loss": 0.777,
-      "step": 466
-    },
-    {
-      "epoch": 2.4072164948453607,
-      "grad_norm": 0.5349386930465698,
-      "learning_rate": 5.7315411712281186e-05,
-      "loss": 0.5766,
-      "step": 467
-    },
-    {
-      "epoch": 2.4123711340206184,
-      "grad_norm": 0.5195199847221375,
-      "learning_rate": 5.714646796877108e-05,
-      "loss": 0.5243,
-      "step": 468
-    },
-    {
-      "epoch": 2.417525773195876,
-      "grad_norm": 0.5453920960426331,
-      "learning_rate": 5.697744089262491e-05,
-      "loss": 0.5514,
-      "step": 469
-    },
-    {
-      "epoch": 2.422680412371134,
-      "grad_norm": 0.669133722782135,
-      "learning_rate": 5.680833245481234e-05,
-      "loss": 0.5849,
-      "step": 470
-    },
-    {
-      "epoch": 2.4278350515463916,
-      "grad_norm": 0.6312981843948364,
-      "learning_rate": 5.6639144627251816e-05,
-      "loss": 0.7814,
-      "step": 471
-    },
-    {
-      "epoch": 2.4329896907216497,
-      "grad_norm": 0.5807788968086243,
-      "learning_rate": 5.646987938278753e-05,
-      "loss": 0.561,
-      "step": 472
-    },
-    {
-      "epoch": 2.4381443298969074,
-      "grad_norm": 0.5659583210945129,
-      "learning_rate": 5.630053869516635e-05,
-      "loss": 0.5743,
-      "step": 473
-    },
-    {
-      "epoch": 2.443298969072165,
-      "grad_norm": 0.5923808217048645,
-      "learning_rate": 5.6131124539014926e-05,
-      "loss": 0.5651,
-      "step": 474
-    },
-    {
-      "epoch": 2.448453608247423,
-      "grad_norm": 0.5530904531478882,
-      "learning_rate": 5.596163888981656e-05,
-      "loss": 0.6827,
-      "step": 475
-    },
-    {
-      "epoch": 2.4536082474226806,
-      "grad_norm": 0.5958693623542786,
-      "learning_rate": 5.5792083723888225e-05,
-      "loss": 0.6882,
-      "step": 476
-    },
-    {
-      "epoch": 2.4587628865979383,
-      "grad_norm": 0.5510756373405457,
-      "learning_rate": 5.5622461018357486e-05,
-      "loss": 0.6592,
-      "step": 477
-    },
-    {
-      "epoch": 2.463917525773196,
-      "grad_norm": 0.4968793988227844,
-      "learning_rate": 5.5452772751139496e-05,
-      "loss": 0.4862,
-      "step": 478
-    },
-    {
-      "epoch": 2.4690721649484537,
-      "grad_norm": 0.7546355724334717,
-      "learning_rate": 5.5283020900913886e-05,
-      "loss": 0.6766,
-      "step": 479
-    },
-    {
-      "epoch": 2.4742268041237114,
-      "grad_norm": 0.7161941528320312,
-      "learning_rate": 5.511320744710171e-05,
-      "loss": 0.9015,
-      "step": 480
-    },
-    {
-      "epoch": 2.479381443298969,
-      "grad_norm": 0.690072238445282,
-      "learning_rate": 5.494333436984238e-05,
-      "loss": 0.6505,
-      "step": 481
-    },
-    {
-      "epoch": 2.484536082474227,
-      "grad_norm": 0.5719749927520752,
-      "learning_rate": 5.477340364997051e-05,
-      "loss": 0.6629,
-      "step": 482
-    },
-    {
-      "epoch": 2.4896907216494846,
-      "grad_norm": 0.5597423911094666,
-      "learning_rate": 5.460341726899291e-05,
-      "loss": 0.6062,
-      "step": 483
-    },
-    {
-      "epoch": 2.4948453608247423,
-      "grad_norm": 0.6454108953475952,
-      "learning_rate": 5.4433377209065414e-05,
-      "loss": 0.6228,
-      "step": 484
-    },
-    {
-      "epoch": 2.5,
-      "grad_norm": 0.5420824289321899,
-      "learning_rate": 5.4263285452969806e-05,
-      "loss": 0.5897,
-      "step": 485
-    },
-    {
-      "epoch": 2.5,
-      "eval_loss": 0.7722232341766357,
-      "eval_runtime": 23.0614,
-      "eval_samples_per_second": 7.111,
-      "eval_steps_per_second": 1.778,
-      "step": 485
-    },
-    {
-      "epoch": 2.5051546391752577,
-      "grad_norm": 0.6164212226867676,
-      "learning_rate": 5.409314398409067e-05,
-      "loss": 0.562,
-      "step": 486
-    },
-    {
-      "epoch": 2.5103092783505154,
-      "grad_norm": 0.5126146078109741,
-      "learning_rate": 5.392295478639225e-05,
-      "loss": 0.697,
-      "step": 487
-    },
-    {
-      "epoch": 2.515463917525773,
-      "grad_norm": 0.6363269686698914,
-      "learning_rate": 5.3752719844395405e-05,
-      "loss": 0.7262,
-      "step": 488
-    },
-    {
-      "epoch": 2.520618556701031,
-      "grad_norm": 0.5842899084091187,
-      "learning_rate": 5.358244114315434e-05,
-      "loss": 0.6244,
-      "step": 489
-    },
-    {
-      "epoch": 2.5257731958762886,
-      "grad_norm": 0.5761100649833679,
-      "learning_rate": 5.341212066823355e-05,
-      "loss": 0.6786,
-      "step": 490
-    },
-    {
-      "epoch": 2.5309278350515463,
-      "grad_norm": 0.6791864037513733,
-      "learning_rate": 5.324176040568465e-05,
-      "loss": 0.6026,
-      "step": 491
-    },
-    {
-      "epoch": 2.536082474226804,
-      "grad_norm": 0.6826097965240479,
-      "learning_rate": 5.307136234202318e-05,
-      "loss": 0.6967,
-      "step": 492
-    },
-    {
-      "epoch": 2.5412371134020617,
-      "grad_norm": 0.5074894428253174,
-      "learning_rate": 5.290092846420548e-05,
-      "loss": 0.6554,
-      "step": 493
-    },
-    {
-      "epoch": 2.5463917525773194,
-      "grad_norm": 0.5714629292488098,
-      "learning_rate": 5.27304607596055e-05,
-      "loss": 0.7671,
-      "step": 494
-    },
-    {
-      "epoch": 2.551546391752577,
-      "grad_norm": 0.5728024244308472,
-      "learning_rate": 5.255996121599167e-05,
-      "loss": 0.6811,
-      "step": 495
-    },
-    {
-      "epoch": 2.556701030927835,
-      "grad_norm": 0.626042902469635,
-      "learning_rate": 5.2389431821503606e-05,
-      "loss": 0.6168,
-      "step": 496
-    },
-    {
-      "epoch": 2.5618556701030926,
-      "grad_norm": 0.6739702224731445,
-      "learning_rate": 5.221887456462907e-05,
-      "loss": 0.7175,
-      "step": 497
-    },
-    {
-      "epoch": 2.5670103092783503,
-      "grad_norm": 0.6577808856964111,
-      "learning_rate": 5.2048291434180716e-05,
-      "loss": 0.6459,
-      "step": 498
-    },
-    {
-      "epoch": 2.572164948453608,
-      "grad_norm": 0.6304759383201599,
-      "learning_rate": 5.1877684419272875e-05,
-      "loss": 0.687,
-      "step": 499
-    },
-    {
-      "epoch": 2.5773195876288657,
-      "grad_norm": 0.6931384801864624,
-      "learning_rate": 5.1707055509298396e-05,
-      "loss": 0.6961,
-      "step": 500
-    },
-    {
-      "epoch": 2.582474226804124,
-      "grad_norm": 0.6150075793266296,
-      "learning_rate": 5.153640669390546e-05,
-      "loss": 0.7175,
-      "step": 501
-    },
-    {
-      "epoch": 2.5876288659793816,
-      "grad_norm": 0.6940797567367554,
-      "learning_rate": 5.1365739962974304e-05,
-      "loss": 0.6951,
-      "step": 502
-    },
-    {
-      "epoch": 2.5927835051546393,
-      "grad_norm": 0.6109462976455688,
-      "learning_rate": 5.119505730659413e-05,
-      "loss": 0.5898,
-      "step": 503
-    },
-    {
-      "epoch": 2.597938144329897,
-      "grad_norm": 0.5795748829841614,
-      "learning_rate": 5.102436071503982e-05,
-      "loss": 0.5456,
-      "step": 504
-    },
-    {
-      "epoch": 2.6030927835051547,
-      "grad_norm": 0.6374852657318115,
-      "learning_rate": 5.0853652178748746e-05,
-      "loss": 0.5659,
-      "step": 505
-    },
-    {
-      "epoch": 2.6082474226804124,
-      "grad_norm": 0.6344237923622131,
-      "learning_rate": 5.068293368829755e-05,
-      "loss": 0.6678,
-      "step": 506
-    },
-    {
-      "epoch": 2.61340206185567,
-      "grad_norm": 0.6286026835441589,
-      "learning_rate": 5.0512207234379004e-05,
-      "loss": 0.5786,
-      "step": 507
-    },
-    {
-      "epoch": 2.618556701030928,
-      "grad_norm": 0.6388340592384338,
-      "learning_rate": 5.0341474807778663e-05,
-      "loss": 0.7377,
-      "step": 508
-    },
-    {
-      "epoch": 2.6237113402061856,
-      "grad_norm": 0.5540049076080322,
-      "learning_rate": 5.017073839935178e-05,
-      "loss": 0.6299,
-      "step": 509
-    },
-    {
-      "epoch": 2.6288659793814433,
-      "grad_norm": 0.7146233320236206,
-      "learning_rate": 5e-05,
-      "loss": 0.64,
-      "step": 510
-    },
-    {
-      "epoch": 2.634020618556701,
-      "grad_norm": 0.5796692967414856,
-      "learning_rate": 4.982926160064823e-05,
-      "loss": 0.5692,
-      "step": 511
-    },
-    {
-      "epoch": 2.6391752577319587,
-      "grad_norm": 0.6053364872932434,
-      "learning_rate": 4.965852519222134e-05,
-      "loss": 0.5449,
-      "step": 512
-    },
-    {
-      "epoch": 2.6443298969072164,
-      "grad_norm": 0.6368183493614197,
-      "learning_rate": 4.948779276562101e-05,
-      "loss": 0.6213,
-      "step": 513
-    },
-    {
-      "epoch": 2.649484536082474,
-      "grad_norm": 0.6399299502372742,
-      "learning_rate": 4.9317066311702456e-05,
-      "loss": 0.7879,
-      "step": 514
-    },
-    {
-      "epoch": 2.654639175257732,
-      "grad_norm": 0.8016248941421509,
-      "learning_rate": 4.9146347821251266e-05,
-      "loss": 0.6358,
-      "step": 515
-    },
-    {
-      "epoch": 2.6597938144329896,
-      "grad_norm": 0.6955182552337646,
-      "learning_rate": 4.89756392849602e-05,
-      "loss": 0.7572,
-      "step": 516
-    },
-    {
-      "epoch": 2.6649484536082473,
-      "grad_norm": 0.7389543056488037,
-      "learning_rate": 4.880494269340588e-05,
-      "loss": 0.6965,
-      "step": 517
-    },
-    {
-      "epoch": 2.670103092783505,
-      "grad_norm": 0.5461128950119019,
-      "learning_rate": 4.863426003702572e-05,
-      "loss": 0.6847,
-      "step": 518
-    },
-    {
-      "epoch": 2.675257731958763,
-      "grad_norm": 0.578420877456665,
-      "learning_rate": 4.8463593306094555e-05,
-      "loss": 0.5675,
-      "step": 519
-    },
-    {
-      "epoch": 2.680412371134021,
-      "grad_norm": 0.4487128257751465,
-      "learning_rate": 4.829294449070161e-05,
-      "loss": 0.5949,
-      "step": 520
-    },
-    {
-      "epoch": 2.6855670103092786,
-      "grad_norm": 0.5900899171829224,
-      "learning_rate": 4.8122315580727136e-05,
-      "loss": 0.5375,
-      "step": 521
-    },
-    {
-      "epoch": 2.6907216494845363,
-      "grad_norm": 0.6414706110954285,
-      "learning_rate": 4.795170856581929e-05,
-      "loss": 0.7276,
-      "step": 522
-    },
-    {
-      "epoch": 2.695876288659794,
-      "grad_norm": 0.6505663394927979,
-      "learning_rate": 4.778112543537094e-05,
-      "loss": 0.7086,
-      "step": 523
-    },
-    {
-      "epoch": 2.7010309278350517,
-      "grad_norm": 0.5941667556762695,
-      "learning_rate": 4.7610568178496405e-05,
-      "loss": 0.5269,
-      "step": 524
-    },
-    {
-      "epoch": 2.7061855670103094,
-      "grad_norm": 0.6392844319343567,
-      "learning_rate": 4.744003878400835e-05,
-      "loss": 0.6261,
-      "step": 525
-    },
-    {
-      "epoch": 2.711340206185567,
-      "grad_norm": 0.608360767364502,
-      "learning_rate": 4.726953924039451e-05,
-      "loss": 0.6867,
-      "step": 526
-    },
-    {
-      "epoch": 2.716494845360825,
-      "grad_norm": 0.5300198197364807,
-      "learning_rate": 4.709907153579454e-05,
-      "loss": 0.5922,
-      "step": 527
-    },
-    {
-      "epoch": 2.7216494845360826,
-      "grad_norm": 0.6723424792289734,
-      "learning_rate": 4.692863765797683e-05,
-      "loss": 0.6799,
-      "step": 528
-    },
-    {
-      "epoch": 2.7268041237113403,
-      "grad_norm": 0.5378847122192383,
-      "learning_rate": 4.675823959431535e-05,
-      "loss": 0.5339,
-      "step": 529
-    },
-    {
-      "epoch": 2.731958762886598,
-      "grad_norm": 0.6574480533599854,
-      "learning_rate": 4.658787933176646e-05,
-      "loss": 0.5919,
-      "step": 530
-    },
-    {
-      "epoch": 2.7371134020618557,
-      "grad_norm": 0.633269727230072,
-      "learning_rate": 4.641755885684566e-05,
-      "loss": 0.8605,
-      "step": 531
-    },
-    {
-      "epoch": 2.7422680412371134,
-      "grad_norm": 0.6804322004318237,
-      "learning_rate": 4.624728015560461e-05,
-      "loss": 0.7026,
-      "step": 532
-    },
-    {
-      "epoch": 2.747422680412371,
-      "grad_norm": 0.6961231231689453,
-      "learning_rate": 4.607704521360776e-05,
-      "loss": 0.7013,
-      "step": 533
-    },
-    {
-      "epoch": 2.752577319587629,
-      "grad_norm": 0.5691372752189636,
-      "learning_rate": 4.590685601590936e-05,
-      "loss": 0.5371,
-      "step": 534
-    },
-    {
-      "epoch": 2.7577319587628866,
-      "grad_norm": 0.7646403908729553,
-      "learning_rate": 4.57367145470302e-05,
-      "loss": 0.6463,
-      "step": 535
-    },
-    {
-      "epoch": 2.7628865979381443,
-      "grad_norm": 0.6509224772453308,
-      "learning_rate": 4.5566622790934604e-05,
-      "loss": 0.66,
-      "step": 536
-    },
-    {
-      "epoch": 2.768041237113402,
-      "grad_norm": 0.6796517968177795,
-      "learning_rate": 4.5396582731007095e-05,
-      "loss": 0.7068,
-      "step": 537
-    },
-    {
-      "epoch": 2.7731958762886597,
-      "grad_norm": 0.6646586060523987,
-      "learning_rate": 4.52265963500295e-05,
-      "loss": 0.6509,
-      "step": 538
-    },
-    {
-      "epoch": 2.7783505154639174,
-      "grad_norm": 0.713420569896698,
-      "learning_rate": 4.505666563015763e-05,
-      "loss": 0.5404,
-      "step": 539
-    },
-    {
-      "epoch": 2.783505154639175,
-      "grad_norm": 0.6599990129470825,
-      "learning_rate": 4.4886792552898286e-05,
-      "loss": 0.8083,
-      "step": 540
-    },
-    {
-      "epoch": 2.788659793814433,
-      "grad_norm": 0.7151883244514465,
-      "learning_rate": 4.471697909908613e-05,
-      "loss": 0.6625,
-      "step": 541
-    },
-    {
-      "epoch": 2.7938144329896906,
-      "grad_norm": 0.6712698936462402,
-      "learning_rate": 4.454722724886051e-05,
-      "loss": 0.773,
-      "step": 542
-    },
-    {
-      "epoch": 2.7989690721649483,
-      "grad_norm": 0.7360748648643494,
-      "learning_rate": 4.437753898164254e-05,
-      "loss": 0.7999,
-      "step": 543
-    },
-    {
-      "epoch": 2.804123711340206,
-      "grad_norm": 0.6323828101158142,
-      "learning_rate": 4.420791627611179e-05,
-      "loss": 0.6039,
-      "step": 544
-    },
-    {
-      "epoch": 2.8092783505154637,
-      "grad_norm": 0.6517678499221802,
-      "learning_rate": 4.403836111018346e-05,
-      "loss": 0.7171,
-      "step": 545
-    },
-    {
-      "epoch": 2.8144329896907214,
-      "grad_norm": 0.5681766271591187,
-      "learning_rate": 4.3868875460985085e-05,
-      "loss": 0.51,
-      "step": 546
-    },
-    {
-      "epoch": 2.819587628865979,
-      "grad_norm": 0.6580605506896973,
-      "learning_rate": 4.369946130483364e-05,
-      "loss": 0.6876,
-      "step": 547
-    },
-    {
-      "epoch": 2.824742268041237,
-      "grad_norm": 0.6996796131134033,
-      "learning_rate": 4.353012061721249e-05,
-      "loss": 0.6901,
-      "step": 548
-    },
-    {
-      "epoch": 2.829896907216495,
-      "grad_norm": 0.8233026266098022,
-      "learning_rate": 4.336085537274818e-05,
-      "loss": 0.7917,
-      "step": 549
-    },
-    {
-      "epoch": 2.8350515463917527,
-      "grad_norm": 0.5782327651977539,
-      "learning_rate": 4.319166754518768e-05,
-      "loss": 0.5233,
-      "step": 550
-    },
-    {
-      "epoch": 2.8402061855670104,
-      "grad_norm": 0.6856580376625061,
-      "learning_rate": 4.3022559107375106e-05,
-      "loss": 0.7887,
-      "step": 551
-    },
-    {
-      "epoch": 2.845360824742268,
-      "grad_norm": 0.5951405167579651,
-      "learning_rate": 4.285353203122893e-05,
-      "loss": 0.6688,
-      "step": 552
-    },
-    {
-      "epoch": 2.850515463917526,
-      "grad_norm": 0.604046642780304,
-      "learning_rate": 4.268458828771883e-05,
-      "loss": 0.6287,
-      "step": 553
-    },
-    {
-      "epoch": 2.8556701030927836,
-      "grad_norm": 0.6061568856239319,
-      "learning_rate": 4.251572984684281e-05,
-      "loss": 0.6911,
-      "step": 554
-    },
-    {
-      "epoch": 2.8608247422680413,
-      "grad_norm": 0.682518720626831,
-      "learning_rate": 4.234695867760412e-05,
-      "loss": 0.604,
-      "step": 555
-    },
-    {
-      "epoch": 2.865979381443299,
-      "grad_norm": 0.6043644547462463,
-      "learning_rate": 4.2178276747988446e-05,
-      "loss": 0.7149,
-      "step": 556
-    },
-    {
-      "epoch": 2.8711340206185567,
-      "grad_norm": 0.6495589017868042,
-      "learning_rate": 4.200968602494087e-05,
-      "loss": 0.6281,
-      "step": 557
-    },
-    {
-      "epoch": 2.8762886597938144,
-      "grad_norm": 0.5073755383491516,
-      "learning_rate": 4.18411884743429e-05,
-      "loss": 0.5349,
-      "step": 558
-    },
-    {
-      "epoch": 2.881443298969072,
-      "grad_norm": 0.7583365440368652,
-      "learning_rate": 4.16727860609896e-05,
-      "loss": 0.6454,
-      "step": 559
-    },
-    {
-      "epoch": 2.88659793814433,
-      "grad_norm": 0.6484652757644653,
-      "learning_rate": 4.150448074856667e-05,
-      "loss": 0.7474,
-      "step": 560
-    },
-    {
-      "epoch": 2.8917525773195876,
-      "grad_norm": 0.7517526149749756,
-      "learning_rate": 4.1336274499627596e-05,
-      "loss": 0.7225,
-      "step": 561
-    },
-    {
-      "epoch": 2.8969072164948453,
-      "grad_norm": 0.7365071773529053,
-      "learning_rate": 4.1168169275570635e-05,
-      "loss": 0.6508,
-      "step": 562
-    },
-    {
-      "epoch": 2.902061855670103,
-      "grad_norm": 0.5838257670402527,
-      "learning_rate": 4.1000167036616113e-05,
-      "loss": 0.6031,
-      "step": 563
-    },
-    {
-      "epoch": 2.9072164948453607,
-      "grad_norm": 0.6092349886894226,
-      "learning_rate": 4.083226974178339e-05,
-      "loss": 0.6745,
-      "step": 564
-    },
-    {
-      "epoch": 2.9123711340206184,
-      "grad_norm": 0.7184647917747498,
-      "learning_rate": 4.066447934886819e-05,
-      "loss": 0.6873,
-      "step": 565
-    },
-    {
-      "epoch": 2.917525773195876,
-      "grad_norm": 0.7946974039077759,
-      "learning_rate": 4.049679781441965e-05,
-      "loss": 0.6382,
-      "step": 566
-    },
-    {
-      "epoch": 2.9226804123711343,
-      "grad_norm": 0.5720714926719666,
-      "learning_rate": 4.0329227093717515e-05,
-      "loss": 0.6175,
-      "step": 567
-    },
-    {
-      "epoch": 2.927835051546392,
-      "grad_norm": 0.5512979626655579,
-      "learning_rate": 4.016176914074944e-05,
-      "loss": 0.6814,
-      "step": 568
-    },
-    {
-      "epoch": 2.9329896907216497,
-      "grad_norm": 0.6203610301017761,
-      "learning_rate": 3.999442590818804e-05,
-      "loss": 0.7352,
-      "step": 569
-    },
-    {
-      "epoch": 2.9381443298969074,
-      "grad_norm": 0.7232295870780945,
-      "learning_rate": 3.982719934736832e-05,
-      "loss": 0.7115,
-      "step": 570
-    },
-    {
-      "epoch": 2.943298969072165,
-      "grad_norm": 0.8409321904182434,
-      "learning_rate": 3.9660091408264716e-05,
-      "loss": 0.7007,
-      "step": 571
-    },
-    {
-      "epoch": 2.948453608247423,
-      "grad_norm": 0.5926352143287659,
-      "learning_rate": 3.949310403946849e-05,
-      "loss": 0.6221,
-      "step": 572
-    },
-    {
-      "epoch": 2.9536082474226806,
-      "grad_norm": 0.731905996799469,
-      "learning_rate": 3.9326239188165025e-05,
-      "loss": 0.7804,
-      "step": 573
-    },
-    {
-      "epoch": 2.9587628865979383,
-      "grad_norm": 0.6734544038772583,
-      "learning_rate": 3.915949880011096e-05,
-      "loss": 0.7285,
-      "step": 574
-    },
-    {
-      "epoch": 2.963917525773196,
-      "grad_norm": 0.5474951863288879,
-      "learning_rate": 3.899288481961173e-05,
-      "loss": 0.5515,
-      "step": 575
-    },
-    {
-      "epoch": 2.9690721649484537,
-      "grad_norm": 0.8142836093902588,
-      "learning_rate": 3.8826399189498654e-05,
-      "loss": 0.7919,
-      "step": 576
-    },
-    {
-      "epoch": 2.9742268041237114,
-      "grad_norm": 0.7191728949546814,
-      "learning_rate": 3.86600438511065e-05,
-      "loss": 0.7446,
-      "step": 577
-    },
-    {
-      "epoch": 2.979381443298969,
-      "grad_norm": 0.6951670050621033,
-      "learning_rate": 3.8493820744250685e-05,
-      "loss": 0.7408,
-      "step": 578
-    },
-    {
-      "epoch": 2.984536082474227,
-      "grad_norm": 0.48555153608322144,
-      "learning_rate": 3.832773180720475e-05,
-      "loss": 0.6051,
-      "step": 579
-    },
-    {
-      "epoch": 2.9896907216494846,
-      "grad_norm": 0.5975886583328247,
-      "learning_rate": 3.8161778976677666e-05,
-      "loss": 0.5546,
-      "step": 580
-    },
-    {
-      "epoch": 2.9948453608247423,
-      "grad_norm": 0.6971765756607056,
-      "learning_rate": 3.79959641877914e-05,
-      "loss": 0.5505,
-      "step": 581
-    },
-    {
-      "epoch": 3.0,
-      "grad_norm": 0.6793974041938782,
-      "learning_rate": 3.783028937405821e-05,
-      "loss": 0.5491,
-      "step": 582
-    },
-    {
-      "epoch": 3.0,
-      "eval_loss": 0.7669576406478882,
-      "eval_runtime": 23.0068,
-      "eval_samples_per_second": 7.128,
-      "eval_steps_per_second": 1.782,
-      "step": 582
-    },
-    {
-      "epoch": 3.0051546391752577,
-      "grad_norm": 0.589939296245575,
-      "learning_rate": 3.766475646735815e-05,
-      "loss": 0.6386,
-      "step": 583
-    },
-    {
-      "epoch": 3.0103092783505154,
-      "grad_norm": 0.7189310193061829,
-      "learning_rate": 3.7499367397916555e-05,
-      "loss": 0.6847,
-      "step": 584
-    },
-    {
-      "epoch": 3.015463917525773,
-      "grad_norm": 0.5985479950904846,
-      "learning_rate": 3.733412409428148e-05,
-      "loss": 0.5704,
-      "step": 585
-    },
-    {
-      "epoch": 3.020618556701031,
-      "grad_norm": 0.6598166227340698,
-      "learning_rate": 3.716902848330133e-05,
-      "loss": 0.5275,
-      "step": 586
-    },
-    {
-      "epoch": 3.0257731958762886,
-      "grad_norm": 0.6075910329818726,
-      "learning_rate": 3.7004082490102226e-05,
-      "loss": 0.6668,
-      "step": 587
-    },
-    {
-      "epoch": 3.0309278350515463,
-      "grad_norm": 0.5255468487739563,
-      "learning_rate": 3.6839288038065734e-05,
-      "loss": 0.5409,
-      "step": 588
-    },
-    {
-      "epoch": 3.036082474226804,
-      "grad_norm": 0.6277021765708923,
-      "learning_rate": 3.667464704880625e-05,
-      "loss": 0.641,
-      "step": 589
-    },
-    {
-      "epoch": 3.0412371134020617,
-      "grad_norm": 0.6998080611228943,
-      "learning_rate": 3.651016144214878e-05,
-      "loss": 0.5058,
-      "step": 590
-    },
-    {
-      "epoch": 3.0463917525773194,
-      "grad_norm": 0.7781403064727783,
-      "learning_rate": 3.634583313610644e-05,
-      "loss": 0.8171,
-      "step": 591
-    },
-    {
-      "epoch": 3.051546391752577,
-      "grad_norm": 0.740594208240509,
-      "learning_rate": 3.618166404685805e-05,
-      "loss": 0.4836,
-      "step": 592
-    },
-    {
-      "epoch": 3.056701030927835,
-      "grad_norm": 0.5763426423072815,
-      "learning_rate": 3.601765608872595e-05,
-      "loss": 0.636,
-      "step": 593
-    },
-    {
-      "epoch": 3.0618556701030926,
-      "grad_norm": 0.8265470862388611,
-      "learning_rate": 3.585381117415349e-05,
-      "loss": 0.6,
-      "step": 594
-    },
-    {
-      "epoch": 3.0670103092783507,
-      "grad_norm": 0.6171466112136841,
-      "learning_rate": 3.5690131213682944e-05,
-      "loss": 0.4655,
-      "step": 595
-    },
-    {
-      "epoch": 3.0721649484536084,
-      "grad_norm": 0.6962155103683472,
-      "learning_rate": 3.5526618115932975e-05,
-      "loss": 0.811,
-      "step": 596
-    },
-    {
-      "epoch": 3.077319587628866,
-      "grad_norm": 0.6894726157188416,
-      "learning_rate": 3.53632737875766e-05,
-      "loss": 0.4777,
-      "step": 597
-    },
-    {
-      "epoch": 3.082474226804124,
-      "grad_norm": 0.7357553839683533,
-      "learning_rate": 3.5200100133318834e-05,
-      "loss": 0.5643,
-      "step": 598
-    },
-    {
-      "epoch": 3.0876288659793816,
-      "grad_norm": 0.5785653591156006,
-      "learning_rate": 3.5037099055874536e-05,
-      "loss": 0.4896,
-      "step": 599
-    },
-    {
-      "epoch": 3.0927835051546393,
-      "grad_norm": 0.7625001668930054,
-      "learning_rate": 3.487427245594622e-05,
-      "loss": 0.5622,
-      "step": 600
-    },
-    {
-      "epoch": 3.097938144329897,
-      "grad_norm": 0.692327082157135,
-      "learning_rate": 3.47116222322018e-05,
-      "loss": 0.5542,
-      "step": 601
-    },
-    {
-      "epoch": 3.1030927835051547,
-      "grad_norm": 0.6226824522018433,
-      "learning_rate": 3.4549150281252636e-05,
-      "loss": 0.5613,
-      "step": 602
-    },
-    {
-      "epoch": 3.1082474226804124,
-      "grad_norm": 0.6324117183685303,
-      "learning_rate": 3.4386858497631205e-05,
-      "loss": 0.5095,
-      "step": 603
-    },
-    {
-      "epoch": 3.11340206185567,
-      "grad_norm": 0.6947232484817505,
-      "learning_rate": 3.422474877376917e-05,
-      "loss": 0.5573,
-      "step": 604
-    },
-    {
-      "epoch": 3.118556701030928,
-      "grad_norm": 0.6270745992660522,
-      "learning_rate": 3.406282299997521e-05,
-      "loss": 0.6132,
-      "step": 605
-    },
-    {
-      "epoch": 3.1237113402061856,
-      "grad_norm": 0.7347365021705627,
-      "learning_rate": 3.3901083064413095e-05,
-      "loss": 0.6026,
-      "step": 606
-    },
-    {
-      "epoch": 3.1288659793814433,
-      "grad_norm": 0.5122291445732117,
-      "learning_rate": 3.3739530853079516e-05,
-      "loss": 0.4393,
-      "step": 607
-    },
-    {
-      "epoch": 3.134020618556701,
-      "grad_norm": 0.6253957152366638,
-      "learning_rate": 3.357816824978222e-05,
-      "loss": 0.5027,
-      "step": 608
-    },
-    {
-      "epoch": 3.1391752577319587,
-      "grad_norm": 0.6118744015693665,
-      "learning_rate": 3.341699713611799e-05,
-      "loss": 0.5994,
-      "step": 609
-    },
-    {
-      "epoch": 3.1443298969072164,
-      "grad_norm": 0.7655370235443115,
-      "learning_rate": 3.325601939145069e-05,
-      "loss": 0.5523,
-      "step": 610
-    },
-    {
-      "epoch": 3.149484536082474,
-      "grad_norm": 0.7117406725883484,
-      "learning_rate": 3.309523689288941e-05,
-      "loss": 0.5523,
-      "step": 611
-    },
-    {
-      "epoch": 3.154639175257732,
-      "grad_norm": 0.6880069375038147,
-      "learning_rate": 3.293465151526649e-05,
-      "loss": 0.6398,
-      "step": 612
-    },
-    {
-      "epoch": 3.1597938144329896,
-      "grad_norm": 0.6741546988487244,
-      "learning_rate": 3.277426513111575e-05,
-      "loss": 0.6644,
-      "step": 613
-    },
-    {
-      "epoch": 3.1649484536082473,
-      "grad_norm": 0.6650920510292053,
-      "learning_rate": 3.261407961065056e-05,
-      "loss": 0.6558,
-      "step": 614
-    },
-    {
-      "epoch": 3.170103092783505,
-      "grad_norm": 0.6102427840232849,
-      "learning_rate": 3.245409682174217e-05,
-      "loss": 0.4753,
-      "step": 615
-    },
-    {
-      "epoch": 3.1752577319587627,
-      "grad_norm": 0.6675355434417725,
-      "learning_rate": 3.229431862989775e-05,
-      "loss": 0.6975,
-      "step": 616
-    },
-    {
-      "epoch": 3.1804123711340204,
-      "grad_norm": 0.7167260646820068,
-      "learning_rate": 3.2134746898238774e-05,
-      "loss": 0.5593,
-      "step": 617
-    },
-    {
-      "epoch": 3.1855670103092786,
-      "grad_norm": 0.7053158283233643,
-      "learning_rate": 3.197538348747927e-05,
-      "loss": 0.5882,
-      "step": 618
-    },
-    {
-      "epoch": 3.1907216494845363,
-      "grad_norm": 0.5679628849029541,
-      "learning_rate": 3.181623025590405e-05,
-      "loss": 0.4235,
-      "step": 619
-    },
-    {
-      "epoch": 3.195876288659794,
-      "grad_norm": 0.4699335992336273,
-      "learning_rate": 3.165728905934718e-05,
-      "loss": 0.3866,
-      "step": 620
-    },
-    {
-      "epoch": 3.2010309278350517,
-      "grad_norm": 0.7112507224082947,
-      "learning_rate": 3.149856175117014e-05,
-      "loss": 0.4572,
-      "step": 621
-    },
-    {
-      "epoch": 3.2061855670103094,
-      "grad_norm": 0.7620934247970581,
-      "learning_rate": 3.134005018224044e-05,
-      "loss": 0.6401,
-      "step": 622
-    },
-    {
-      "epoch": 3.211340206185567,
-      "grad_norm": 0.8001824021339417,
-      "learning_rate": 3.118175620090983e-05,
-      "loss": 0.5794,
-      "step": 623
-    },
-    {
-      "epoch": 3.216494845360825,
-      "grad_norm": 0.6476466655731201,
-      "learning_rate": 3.1023681652992926e-05,
-      "loss": 0.6406,
-      "step": 624
-    },
-    {
-      "epoch": 3.2216494845360826,
-      "grad_norm": 0.7699551582336426,
-      "learning_rate": 3.086582838174551e-05,
-      "loss": 0.6377,
-      "step": 625
-    },
-    {
-      "epoch": 3.2268041237113403,
-      "grad_norm": 0.6394461393356323,
-      "learning_rate": 3.070819822784323e-05,
-      "loss": 0.562,
-      "step": 626
-    },
-    {
-      "epoch": 3.231958762886598,
-      "grad_norm": 0.7696752548217773,
-      "learning_rate": 3.055079302935997e-05,
-      "loss": 0.541,
-      "step": 627
-    },
-    {
-      "epoch": 3.2371134020618557,
-      "grad_norm": 0.7611710429191589,
-      "learning_rate": 3.0393614621746498e-05,
-      "loss": 0.552,
-      "step": 628
-    },
-    {
-      "epoch": 3.2422680412371134,
-      "grad_norm": 0.7990087270736694,
-      "learning_rate": 3.023666483780905e-05,
-      "loss": 0.6217,
-      "step": 629
-    },
-    {
-      "epoch": 3.247422680412371,
-      "grad_norm": 0.6985018849372864,
-      "learning_rate": 3.007994550768793e-05,
-      "loss": 0.554,
-      "step": 630
-    },
-    {
-      "epoch": 3.252577319587629,
-      "grad_norm": 0.7422522902488708,
-      "learning_rate": 2.9923458458836258e-05,
-      "loss": 0.6494,
-      "step": 631
-    },
-    {
-      "epoch": 3.2577319587628866,
-      "grad_norm": 0.7043977975845337,
-      "learning_rate": 2.9767205515998518e-05,
-      "loss": 0.483,
-      "step": 632
-    },
-    {
-      "epoch": 3.2628865979381443,
-      "grad_norm": 0.9038345813751221,
-      "learning_rate": 2.9611188501189435e-05,
-      "loss": 0.6607,
-      "step": 633
-    },
-    {
-      "epoch": 3.268041237113402,
-      "grad_norm": 0.7127012014389038,
-      "learning_rate": 2.9455409233672592e-05,
-      "loss": 0.5094,
-      "step": 634
-    },
-    {
-      "epoch": 3.2731958762886597,
-      "grad_norm": 0.6978604197502136,
-      "learning_rate": 2.929986952993933e-05,
-      "loss": 0.6598,
-      "step": 635
-    },
-    {
-      "epoch": 3.2783505154639174,
-      "grad_norm": 0.880826473236084,
-      "learning_rate": 2.9144571203687476e-05,
-      "loss": 0.5985,
-      "step": 636
-    },
-    {
-      "epoch": 3.283505154639175,
-      "grad_norm": 0.7255879044532776,
-      "learning_rate": 2.8989516065800238e-05,
-      "loss": 0.5601,
-      "step": 637
-    },
-    {
-      "epoch": 3.288659793814433,
-      "grad_norm": 0.8017705082893372,
-      "learning_rate": 2.8834705924325118e-05,
-      "loss": 0.6538,
-      "step": 638
-    },
-    {
-      "epoch": 3.2938144329896906,
-      "grad_norm": 0.6248070001602173,
-      "learning_rate": 2.8680142584452742e-05,
-      "loss": 0.5287,
-      "step": 639
-    },
-    {
-      "epoch": 3.2989690721649483,
-      "grad_norm": 0.778263509273529,
-      "learning_rate": 2.8525827848495913e-05,
-      "loss": 0.6826,
-      "step": 640
-    },
-    {
-      "epoch": 3.304123711340206,
-      "grad_norm": 0.7028930187225342,
-      "learning_rate": 2.83717635158685e-05,
-      "loss": 0.6151,
-      "step": 641
-    },
-    {
-      "epoch": 3.3092783505154637,
-      "grad_norm": 0.785013735294342,
-      "learning_rate": 2.8217951383064544e-05,
-      "loss": 0.7159,
-      "step": 642
-    },
-    {
-      "epoch": 3.3144329896907214,
-      "grad_norm": 0.686386227607727,
-      "learning_rate": 2.8064393243637222e-05,
-      "loss": 0.5513,
-      "step": 643
-    },
-    {
-      "epoch": 3.319587628865979,
-      "grad_norm": 0.6582889556884766,
-      "learning_rate": 2.791109088817803e-05,
-      "loss": 0.4522,
-      "step": 644
-    },
-    {
-      "epoch": 3.3247422680412373,
-      "grad_norm": 0.6924684047698975,
-      "learning_rate": 2.7758046104295797e-05,
-      "loss": 0.6779,
-      "step": 645
-    },
-    {
-      "epoch": 3.329896907216495,
-      "grad_norm": 0.8219475746154785,
-      "learning_rate": 2.760526067659591e-05,
-      "loss": 0.5771,
-      "step": 646
-    },
-    {
-      "epoch": 3.3350515463917527,
-      "grad_norm": 0.8043853640556335,
-      "learning_rate": 2.7452736386659516e-05,
-      "loss": 0.7175,
-      "step": 647
-    },
-    {
-      "epoch": 3.3402061855670104,
-      "grad_norm": 0.6423444151878357,
-      "learning_rate": 2.7300475013022663e-05,
-      "loss": 0.5696,
-      "step": 648
-    },
-    {
-      "epoch": 3.345360824742268,
-      "grad_norm": 0.6770009994506836,
-      "learning_rate": 2.7148478331155702e-05,
-      "loss": 0.501,
-      "step": 649
-    },
-    {
-      "epoch": 3.350515463917526,
-      "grad_norm": 0.7062271237373352,
-      "learning_rate": 2.6996748113442394e-05,
-      "loss": 0.4872,
-      "step": 650
-    },
-    {
-      "epoch": 3.3556701030927836,
-      "grad_norm": 0.6703154444694519,
-      "learning_rate": 2.6845286129159464e-05,
-      "loss": 0.4888,
-      "step": 651
-    },
-    {
-      "epoch": 3.3608247422680413,
-      "grad_norm": 0.6826891303062439,
-      "learning_rate": 2.669409414445574e-05,
-      "loss": 0.5669,
-      "step": 652
-    },
-    {
-      "epoch": 3.365979381443299,
-      "grad_norm": 0.8772823810577393,
-      "learning_rate": 2.6543173922331743e-05,
-      "loss": 0.7018,
-      "step": 653
-    },
-    {
-      "epoch": 3.3711340206185567,
-      "grad_norm": 0.6661205887794495,
-      "learning_rate": 2.639252722261908e-05,
-      "loss": 0.5039,
-      "step": 654
-    },
-    {
-      "epoch": 3.3762886597938144,
-      "grad_norm": 0.7768886685371399,
-      "learning_rate": 2.624215580195981e-05,
-      "loss": 0.5708,
-      "step": 655
-    },
-    {
-      "epoch": 3.381443298969072,
-      "grad_norm": 0.7129529118537903,
-      "learning_rate": 2.6092061413786156e-05,
-      "loss": 0.5247,
-      "step": 656
-    },
-    {
-      "epoch": 3.38659793814433,
-      "grad_norm": 0.7244031429290771,
-      "learning_rate": 2.5942245808299886e-05,
-      "loss": 0.5497,
-      "step": 657
-    },
-    {
-      "epoch": 3.3917525773195876,
-      "grad_norm": 0.7318927049636841,
-      "learning_rate": 2.5792710732451997e-05,
-      "loss": 0.6757,
-      "step": 658
-    },
-    {
-      "epoch": 3.3969072164948453,
-      "grad_norm": 0.6742333173751831,
-      "learning_rate": 2.56434579299223e-05,
-      "loss": 0.6005,
-      "step": 659
-    },
-    {
-      "epoch": 3.402061855670103,
-      "grad_norm": 0.82306969165802,
-      "learning_rate": 2.5494489141099153e-05,
-      "loss": 0.6058,
-      "step": 660
-    },
-    {
-      "epoch": 3.4072164948453607,
-      "grad_norm": 0.5822560787200928,
-      "learning_rate": 2.534580610305909e-05,
-      "loss": 0.4509,
-      "step": 661
-    },
-    {
-      "epoch": 3.4123711340206184,
-      "grad_norm": 0.6220210790634155,
-      "learning_rate": 2.5197410549546595e-05,
-      "loss": 0.3713,
-      "step": 662
-    },
-    {
-      "epoch": 3.417525773195876,
-      "grad_norm": 0.6950068473815918,
-      "learning_rate": 2.5049304210953933e-05,
-      "loss": 0.532,
-      "step": 663
-    },
-    {
-      "epoch": 3.422680412371134,
-      "grad_norm": 0.735115647315979,
-      "learning_rate": 2.4901488814300856e-05,
-      "loss": 0.5285,
-      "step": 664
-    },
-    {
-      "epoch": 3.4278350515463916,
-      "grad_norm": 0.7899560928344727,
-      "learning_rate": 2.4753966083214615e-05,
-      "loss": 0.5996,
-      "step": 665
-    },
-    {
-      "epoch": 3.4329896907216497,
-      "grad_norm": 0.724655032157898,
-      "learning_rate": 2.4606737737909697e-05,
-      "loss": 0.5587,
-      "step": 666
-    },
-    {
-      "epoch": 3.4381443298969074,
-      "grad_norm": 0.8545822501182556,
-      "learning_rate": 2.4459805495167942e-05,
-      "loss": 0.6505,
-      "step": 667
-    },
-    {
-      "epoch": 3.443298969072165,
-      "grad_norm": 0.7957090735435486,
-      "learning_rate": 2.4313171068318357e-05,
-      "loss": 0.6138,
-      "step": 668
-    },
-    {
-      "epoch": 3.448453608247423,
-      "grad_norm": 0.9017428159713745,
-      "learning_rate": 2.4166836167217283e-05,
-      "loss": 0.6846,
-      "step": 669
-    },
-    {
-      "epoch": 3.4536082474226806,
-      "grad_norm": 0.6370732188224792,
-      "learning_rate": 2.4020802498228335e-05,
-      "loss": 0.5674,
-      "step": 670
-    },
-    {
-      "epoch": 3.4587628865979383,
-      "grad_norm": 0.8172600269317627,
-      "learning_rate": 2.3875071764202563e-05,
-      "loss": 0.606,
-      "step": 671
-    },
-    {
-      "epoch": 3.463917525773196,
-      "grad_norm": 0.8876442909240723,
-      "learning_rate": 2.3729645664458638e-05,
-      "loss": 0.6484,
-      "step": 672
-    },
-    {
-      "epoch": 3.4690721649484537,
-      "grad_norm": 0.6860740184783936,
-      "learning_rate": 2.3584525894762928e-05,
-      "loss": 0.4335,
-      "step": 673
-    },
-    {
-      "epoch": 3.4742268041237114,
-      "grad_norm": 0.6609134674072266,
-      "learning_rate": 2.3439714147309845e-05,
-      "loss": 0.5055,
-      "step": 674
-    },
-    {
-      "epoch": 3.479381443298969,
-      "grad_norm": 0.768539309501648,
-      "learning_rate": 2.329521211070199e-05,
-      "loss": 0.5386,
-      "step": 675
-    },
-    {
-      "epoch": 3.484536082474227,
-      "grad_norm": 0.7136728763580322,
-      "learning_rate": 2.3151021469930613e-05,
-      "loss": 0.6505,
-      "step": 676
-    },
-    {
-      "epoch": 3.4896907216494846,
-      "grad_norm": 0.7553372979164124,
-      "learning_rate": 2.3007143906355767e-05,
-      "loss": 0.5787,
-      "step": 677
-    },
-    {
-      "epoch": 3.4948453608247423,
-      "grad_norm": 0.9047206044197083,
-      "learning_rate": 2.2863581097686925e-05,
-      "loss": 0.8126,
-      "step": 678
-    },
-    {
-      "epoch": 3.5,
-      "grad_norm": 0.665693461894989,
-      "learning_rate": 2.2720334717963222e-05,
-      "loss": 0.5814,
-      "step": 679
-    },
-    {
-      "epoch": 3.5,
-      "eval_loss": 0.7873027920722961,
-      "eval_runtime": 23.1508,
-      "eval_samples_per_second": 7.084,
-      "eval_steps_per_second": 1.771,
-      "step": 679
-    },
-    {
-      "epoch": 3.5051546391752577,
-      "grad_norm": 0.6601768136024475,
-      "learning_rate": 2.2577406437534054e-05,
-      "loss": 0.5339,
-      "step": 680
-    },
-    {
-      "epoch": 3.5103092783505154,
-      "grad_norm": 0.7132072448730469,
-      "learning_rate": 2.2434797923039598e-05,
-      "loss": 0.5322,
-      "step": 681
-    },
-    {
-      "epoch": 3.515463917525773,
-      "grad_norm": 0.7644983530044556,
-      "learning_rate": 2.2292510837391267e-05,
-      "loss": 0.6137,
-      "step": 682
-    },
-    {
-      "epoch": 3.520618556701031,
-      "grad_norm": 0.6903513669967651,
-      "learning_rate": 2.2150546839752438e-05,
-      "loss": 0.6657,
-      "step": 683
-    },
-    {
-      "epoch": 3.5257731958762886,
-      "grad_norm": 0.840147852897644,
-      "learning_rate": 2.2008907585519095e-05,
-      "loss": 0.5655,
-      "step": 684
-    },
-    {
-      "epoch": 3.5309278350515463,
-      "grad_norm": 0.7994117736816406,
-      "learning_rate": 2.186759472630045e-05,
-      "loss": 0.8075,
-      "step": 685
-    },
-    {
-      "epoch": 3.536082474226804,
-      "grad_norm": 0.7331596612930298,
-      "learning_rate": 2.172660990989971e-05,
-      "loss": 0.5532,
-      "step": 686
-    },
-    {
-      "epoch": 3.5412371134020617,
-      "grad_norm": 0.7885189652442932,
-      "learning_rate": 2.1585954780294947e-05,
-      "loss": 0.6312,
-      "step": 687
-    },
-    {
-      "epoch": 3.5463917525773194,
-      "grad_norm": 0.8121082186698914,
-      "learning_rate": 2.144563097761984e-05,
-      "loss": 0.54,
-      "step": 688
-    },
-    {
-      "epoch": 3.551546391752577,
-      "grad_norm": 0.7569543123245239,
-      "learning_rate": 2.130564013814453e-05,
-      "loss": 0.6202,
-      "step": 689
-    },
-    {
-      "epoch": 3.556701030927835,
-      "grad_norm": 0.6454798579216003,
-      "learning_rate": 2.1165983894256647e-05,
-      "loss": 0.5141,
-      "step": 690
-    },
-    {
-      "epoch": 3.5618556701030926,
-      "grad_norm": 0.8562419414520264,
-      "learning_rate": 2.102666387444215e-05,
-      "loss": 0.6766,
-      "step": 691
-    },
-    {
-      "epoch": 3.5670103092783503,
-      "grad_norm": 0.6552653908729553,
-      "learning_rate": 2.0887681703266453e-05,
-      "loss": 0.4334,
-      "step": 692
-    },
-    {
-      "epoch": 3.572164948453608,
-      "grad_norm": 0.6466341614723206,
-      "learning_rate": 2.0749039001355375e-05,
-      "loss": 0.6025,
-      "step": 693
-    },
-    {
-      "epoch": 3.5773195876288657,
-      "grad_norm": 0.7250998616218567,
-      "learning_rate": 2.061073738537635e-05,
-      "loss": 0.5688,
-      "step": 694
-    },
-    {
-      "epoch": 3.582474226804124,
-      "grad_norm": 0.7872217297554016,
-      "learning_rate": 2.0472778468019454e-05,
-      "loss": 0.5565,
-      "step": 695
-    },
-    {
-      "epoch": 3.5876288659793816,
-      "grad_norm": 0.6609041690826416,
-      "learning_rate": 2.0335163857978744e-05,
-      "loss": 0.4888,
-      "step": 696
-    },
-    {
-      "epoch": 3.5927835051546393,
-      "grad_norm": 0.9081669449806213,
-      "learning_rate": 2.019789515993336e-05,
-      "loss": 0.6244,
-      "step": 697
-    },
-    {
-      "epoch": 3.597938144329897,
-      "grad_norm": 0.7092580199241638,
-      "learning_rate": 2.0060973974528874e-05,
-      "loss": 0.4923,
-      "step": 698
-    },
-    {
-      "epoch": 3.6030927835051547,
-      "grad_norm": 0.7079344987869263,
-      "learning_rate": 1.992440189835869e-05,
-      "loss": 0.7202,
-      "step": 699
-    },
-    {
-      "epoch": 3.6082474226804124,
-      "grad_norm": 0.813388466835022,
-      "learning_rate": 1.9788180523945277e-05,
-      "loss": 0.5914,
-      "step": 700
-    },
-    {
-      "epoch": 3.61340206185567,
-      "grad_norm": 0.8015619516372681,
-      "learning_rate": 1.9652311439721764e-05,
-      "loss": 0.6758,
-      "step": 701
-    },
-    {
-      "epoch": 3.618556701030928,
-      "grad_norm": 0.6869595050811768,
-      "learning_rate": 1.9516796230013272e-05,
-      "loss": 0.5988,
-      "step": 702
-    },
-    {
-      "epoch": 3.6237113402061856,
-      "grad_norm": 0.6902271509170532,
-      "learning_rate": 1.9381636475018577e-05,
-      "loss": 0.5267,
-      "step": 703
-    },
-    {
-      "epoch": 3.6288659793814433,
-      "grad_norm": 0.6685857176780701,
-      "learning_rate": 1.9246833750791526e-05,
-      "loss": 0.4817,
-      "step": 704
-    },
-    {
-      "epoch": 3.634020618556701,
-      "grad_norm": 0.6419634222984314,
-      "learning_rate": 1.9112389629222823e-05,
-      "loss": 0.6316,
-      "step": 705
-    },
-    {
-      "epoch": 3.6391752577319587,
-      "grad_norm": 0.8290701508522034,
-      "learning_rate": 1.8978305678021595e-05,
-      "loss": 0.6548,
-      "step": 706
-    },
-    {
-      "epoch": 3.6443298969072164,
-      "grad_norm": 0.6211696267127991,
-      "learning_rate": 1.884458346069713e-05,
-      "loss": 0.5943,
-      "step": 707
-    },
-    {
-      "epoch": 3.649484536082474,
-      "grad_norm": 0.7741536498069763,
-      "learning_rate": 1.8711224536540678e-05,
-      "loss": 0.5249,
-      "step": 708
-    },
-    {
-      "epoch": 3.654639175257732,
-      "grad_norm": 0.7374269962310791,
-      "learning_rate": 1.857823046060722e-05,
-      "loss": 0.6726,
-      "step": 709
-    },
-    {
-      "epoch": 3.6597938144329896,
-      "grad_norm": 0.6983458995819092,
-      "learning_rate": 1.8445602783697374e-05,
-      "loss": 0.5565,
-      "step": 710
-    },
-    {
-      "epoch": 3.6649484536082473,
-      "grad_norm": 0.7474353313446045,
-      "learning_rate": 1.831334305233928e-05,
-      "loss": 0.5496,
-      "step": 711
-    },
-    {
-      "epoch": 3.670103092783505,
-      "grad_norm": 0.6772493720054626,
-      "learning_rate": 1.8181452808770637e-05,
-      "loss": 0.5827,
-      "step": 712
-    },
-    {
-      "epoch": 3.675257731958763,
-      "grad_norm": 0.6143507957458496,
-      "learning_rate": 1.804993359092059e-05,
-      "loss": 0.5753,
-      "step": 713
-    },
-    {
-      "epoch": 3.680412371134021,
-      "grad_norm": 0.6349582076072693,
-      "learning_rate": 1.7918786932391944e-05,
-      "loss": 0.5396,
-      "step": 714
-    },
-    {
-      "epoch": 3.6855670103092786,
-      "grad_norm": 0.7443505525588989,
-      "learning_rate": 1.778801436244319e-05,
-      "loss": 0.6715,
-      "step": 715
-    },
-    {
-      "epoch": 3.6907216494845363,
-      "grad_norm": 0.7070426940917969,
-      "learning_rate": 1.765761740597065e-05,
-      "loss": 0.6967,
-      "step": 716
-    },
-    {
-      "epoch": 3.695876288659794,
-      "grad_norm": 0.7999926805496216,
-      "learning_rate": 1.7527597583490822e-05,
-      "loss": 0.6274,
-      "step": 717
-    },
-    {
-      "epoch": 3.7010309278350517,
-      "grad_norm": 0.571897566318512,
-      "learning_rate": 1.739795641112248e-05,
-      "loss": 0.4957,
-      "step": 718
-    },
-    {
-      "epoch": 3.7061855670103094,
-      "grad_norm": 0.8360907435417175,
-      "learning_rate": 1.726869540056915e-05,
-      "loss": 0.6236,
-      "step": 719
-    },
-    {
-      "epoch": 3.711340206185567,
-      "grad_norm": 0.689716100692749,
-      "learning_rate": 1.713981605910137e-05,
-      "loss": 0.6277,
-      "step": 720
-    },
-    {
-      "epoch": 3.716494845360825,
-      "grad_norm": 0.7487063407897949,
-      "learning_rate": 1.70113198895392e-05,
-      "loss": 0.528,
-      "step": 721
-    },
-    {
-      "epoch": 3.7216494845360826,
-      "grad_norm": 0.84012770652771,
-      "learning_rate": 1.6883208390234628e-05,
-      "loss": 0.5405,
-      "step": 722
-    },
-    {
-      "epoch": 3.7268041237113403,
-      "grad_norm": 0.6260896921157837,
-      "learning_rate": 1.6755483055054105e-05,
-      "loss": 0.4328,
-      "step": 723
-    },
-    {
-      "epoch": 3.731958762886598,
-      "grad_norm": 0.8109961152076721,
-      "learning_rate": 1.662814537336122e-05,
-      "loss": 0.5839,
-      "step": 724
-    },
-    {
-      "epoch": 3.7371134020618557,
-      "grad_norm": 0.9204162955284119,
-      "learning_rate": 1.650119682999918e-05,
-      "loss": 0.7058,
-      "step": 725
-    },
-    {
-      "epoch": 3.7422680412371134,
-      "grad_norm": 0.5918567776679993,
-      "learning_rate": 1.6374638905273643e-05,
-      "loss": 0.4917,
-      "step": 726
-    },
-    {
-      "epoch": 3.747422680412371,
-      "grad_norm": 0.7899376749992371,
-      "learning_rate": 1.624847307493534e-05,
-      "loss": 0.5678,
-      "step": 727
-    },
-    {
-      "epoch": 3.752577319587629,
-      "grad_norm": 0.8229061961174011,
-      "learning_rate": 1.6122700810162966e-05,
-      "loss": 0.6471,
-      "step": 728
-    },
-    {
-      "epoch": 3.7577319587628866,
-      "grad_norm": 0.737942636013031,
-      "learning_rate": 1.5997323577545915e-05,
-      "loss": 0.6076,
-      "step": 729
-    },
-    {
-      "epoch": 3.7628865979381443,
-      "grad_norm": 0.7289573550224304,
-      "learning_rate": 1.5872342839067306e-05,
-      "loss": 0.5477,
-      "step": 730
-    },
-    {
-      "epoch": 3.768041237113402,
-      "grad_norm": 0.7133212685585022,
-      "learning_rate": 1.5747760052086803e-05,
-      "loss": 0.7274,
-      "step": 731
-    },
-    {
-      "epoch": 3.7731958762886597,
-      "grad_norm": 0.7502733469009399,
-      "learning_rate": 1.5623576669323743e-05,
-      "loss": 0.6026,
-      "step": 732
-    },
-    {
-      "epoch": 3.7783505154639174,
-      "grad_norm": 0.620846152305603,
-      "learning_rate": 1.5499794138840122e-05,
-      "loss": 0.5394,
-      "step": 733
-    },
-    {
-      "epoch": 3.783505154639175,
-      "grad_norm": 0.6744952201843262,
-      "learning_rate": 1.5376413904023722e-05,
-      "loss": 0.4315,
-      "step": 734
-    },
-    {
-      "epoch": 3.788659793814433,
-      "grad_norm": 0.6577085256576538,
-      "learning_rate": 1.525343740357128e-05,
-      "loss": 0.5313,
-      "step": 735
-    },
-    {
-      "epoch": 3.7938144329896906,
-      "grad_norm": 0.8090968132019043,
-      "learning_rate": 1.5130866071471717e-05,
-      "loss": 0.573,
-      "step": 736
-    },
-    {
-      "epoch": 3.7989690721649483,
-      "grad_norm": 0.8299245834350586,
-      "learning_rate": 1.500870133698945e-05,
-      "loss": 0.624,
-      "step": 737
-    },
-    {
-      "epoch": 3.804123711340206,
-      "grad_norm": 0.7030582427978516,
-      "learning_rate": 1.4886944624647647e-05,
-      "loss": 0.5036,
-      "step": 738
-    },
-    {
-      "epoch": 3.8092783505154637,
-      "grad_norm": 0.6806750297546387,
-      "learning_rate": 1.4765597354211713e-05,
-      "loss": 0.497,
-      "step": 739
-    },
-    {
-      "epoch": 3.8144329896907214,
-      "grad_norm": 0.7690914273262024,
-      "learning_rate": 1.4644660940672627e-05,
-      "loss": 0.5217,
-      "step": 740
-    },
-    {
-      "epoch": 3.819587628865979,
-      "grad_norm": 0.6966343522071838,
-      "learning_rate": 1.4524136794230547e-05,
-      "loss": 0.5966,
-      "step": 741
-    },
-    {
-      "epoch": 3.824742268041237,
-      "grad_norm": 0.6488124132156372,
-      "learning_rate": 1.4404026320278318e-05,
-      "loss": 0.454,
-      "step": 742
-    },
-    {
-      "epoch": 3.829896907216495,
-      "grad_norm": 0.7203002572059631,
-      "learning_rate": 1.4284330919385036e-05,
-      "loss": 0.6012,
-      "step": 743
-    },
-    {
-      "epoch": 3.8350515463917527,
-      "grad_norm": 0.8698861598968506,
-      "learning_rate": 1.4165051987279831e-05,
-      "loss": 0.7224,
-      "step": 744
-    },
-    {
-      "epoch": 3.8402061855670104,
-      "grad_norm": 0.7975336909294128,
-      "learning_rate": 1.404619091483546e-05,
-      "loss": 0.7323,
-      "step": 745
-    },
-    {
-      "epoch": 3.845360824742268,
-      "grad_norm": 0.697154700756073,
-      "learning_rate": 1.3927749088052217e-05,
-      "loss": 0.5375,
-      "step": 746
-    },
-    {
-      "epoch": 3.850515463917526,
-      "grad_norm": 0.857922375202179,
-      "learning_rate": 1.3809727888041668e-05,
-      "loss": 0.6702,
-      "step": 747
-    },
-    {
-      "epoch": 3.8556701030927836,
-      "grad_norm": 0.7249358296394348,
-      "learning_rate": 1.3692128691010592e-05,
-      "loss": 0.578,
-      "step": 748
-    },
-    {
-      "epoch": 3.8608247422680413,
-      "grad_norm": 1.0248481035232544,
-      "learning_rate": 1.3574952868244922e-05,
-      "loss": 0.6964,
-      "step": 749
-    },
-    {
-      "epoch": 3.865979381443299,
-      "grad_norm": 0.8700373768806458,
-      "learning_rate": 1.3458201786093794e-05,
-      "loss": 0.686,
-      "step": 750
-    },
-    {
-      "epoch": 3.8711340206185567,
-      "grad_norm": 0.7813423871994019,
-      "learning_rate": 1.334187680595358e-05,
-      "loss": 0.5817,
-      "step": 751
-    },
-    {
-      "epoch": 3.8762886597938144,
-      "grad_norm": 0.6536322832107544,
-      "learning_rate": 1.3225979284251954e-05,
-      "loss": 0.5981,
-      "step": 752
-    },
-    {
-      "epoch": 3.881443298969072,
-      "grad_norm": 0.8255348205566406,
-      "learning_rate": 1.3110510572432221e-05,
-      "loss": 0.7288,
-      "step": 753
-    },
-    {
-      "epoch": 3.88659793814433,
-      "grad_norm": 0.7284218668937683,
-      "learning_rate": 1.2995472016937404e-05,
-      "loss": 0.5522,
-      "step": 754
-    },
-    {
-      "epoch": 3.8917525773195876,
-      "grad_norm": 0.7565793395042419,
-      "learning_rate": 1.2880864959194665e-05,
-      "loss": 0.6455,
-      "step": 755
-    },
-    {
-      "epoch": 3.8969072164948453,
-      "grad_norm": 0.7215232253074646,
-      "learning_rate": 1.2766690735599568e-05,
-      "loss": 0.5919,
-      "step": 756
-    },
-    {
-      "epoch": 3.902061855670103,
-      "grad_norm": 0.5804430246353149,
-      "learning_rate": 1.2652950677500574e-05,
-      "loss": 0.5657,
-      "step": 757
-    },
-    {
-      "epoch": 3.9072164948453607,
-      "grad_norm": 0.6634606122970581,
-      "learning_rate": 1.253964611118345e-05,
-      "loss": 0.4888,
-      "step": 758
-    },
-    {
-      "epoch": 3.9123711340206184,
-      "grad_norm": 0.5982066988945007,
-      "learning_rate": 1.2426778357855873e-05,
-      "loss": 0.539,
-      "step": 759
-    },
-    {
-      "epoch": 3.917525773195876,
-      "grad_norm": 0.7859188914299011,
-      "learning_rate": 1.2314348733631959e-05,
-      "loss": 0.6816,
-      "step": 760
-    },
-    {
-      "epoch": 3.9226804123711343,
-      "grad_norm": 0.7802892327308655,
-      "learning_rate": 1.2202358549516923e-05,
-      "loss": 0.6872,
-      "step": 761
-    },
-    {
-      "epoch": 3.927835051546392,
-      "grad_norm": 0.699567437171936,
-      "learning_rate": 1.209080911139187e-05,
-      "loss": 0.5798,
-      "step": 762
-    },
-    {
-      "epoch": 3.9329896907216497,
-      "grad_norm": 0.777010440826416,
-      "learning_rate": 1.1979701719998453e-05,
-      "loss": 0.6204,
-      "step": 763
-    },
-    {
-      "epoch": 3.9381443298969074,
-      "grad_norm": 0.7714626789093018,
-      "learning_rate": 1.1869037670923815e-05,
-      "loss": 0.7335,
-      "step": 764
-    },
-    {
-      "epoch": 3.943298969072165,
-      "grad_norm": 0.7658986449241638,
-      "learning_rate": 1.1758818254585369e-05,
-      "loss": 0.6169,
-      "step": 765
-    },
-    {
-      "epoch": 3.948453608247423,
-      "grad_norm": 0.7642006278038025,
-      "learning_rate": 1.164904475621587e-05,
-      "loss": 0.6777,
-      "step": 766
-    },
-    {
-      "epoch": 3.9536082474226806,
-      "grad_norm": 0.8413475751876831,
-      "learning_rate": 1.1539718455848309e-05,
-      "loss": 0.4705,
-      "step": 767
-    },
-    {
-      "epoch": 3.9587628865979383,
-      "grad_norm": 0.7557904720306396,
-      "learning_rate": 1.1430840628301093e-05,
-      "loss": 0.5588,
-      "step": 768
-    },
-    {
-      "epoch": 3.963917525773196,
-      "grad_norm": 0.7613516449928284,
-      "learning_rate": 1.1322412543163135e-05,
-      "loss": 0.6615,
-      "step": 769
-    },
-    {
-      "epoch": 3.9690721649484537,
-      "grad_norm": 0.7262148261070251,
-      "learning_rate": 1.1214435464779006e-05,
-      "loss": 0.6986,
-      "step": 770
-    },
-    {
-      "epoch": 3.9742268041237114,
-      "grad_norm": 0.6863129734992981,
-      "learning_rate": 1.1106910652234276e-05,
-      "loss": 0.5653,
-      "step": 771
-    },
-    {
-      "epoch": 3.979381443298969,
-      "grad_norm": 0.7940157055854797,
-      "learning_rate": 1.099983935934077e-05,
-      "loss": 0.6754,
-      "step": 772
-    },
-    {
-      "epoch": 3.984536082474227,
-      "grad_norm": 0.8549270629882812,
-      "learning_rate": 1.089322283462197e-05,
-      "loss": 0.6918,
-      "step": 773
-    },
-    {
-      "epoch": 3.9896907216494846,
-      "grad_norm": 0.6270641088485718,
-      "learning_rate": 1.0787062321298442e-05,
-      "loss": 0.5433,
-      "step": 774
-    },
-    {
-      "epoch": 3.9948453608247423,
-      "grad_norm": 0.758388340473175,
-      "learning_rate": 1.0681359057273388e-05,
-      "loss": 0.6541,
-      "step": 775
-    },
-    {
-      "epoch": 4.0,
-      "grad_norm": 1.1503102779388428,
-      "learning_rate": 1.0576114275118131e-05,
-      "loss": 0.6927,
-      "step": 776
-    },
-    {
-      "epoch": 4.0,
-      "eval_loss": 0.7827529907226562,
-      "eval_runtime": 23.077,
-      "eval_samples_per_second": 7.107,
-      "eval_steps_per_second": 1.777,
-      "step": 776
-    },
-    {
-      "epoch": 4.005154639175258,
-      "grad_norm": 0.6961793899536133,
-      "learning_rate": 1.0471329202057823e-05,
-      "loss": 0.4523,
-      "step": 777
-    },
-    {
-      "epoch": 4.010309278350515,
-      "grad_norm": 0.5918589234352112,
-      "learning_rate": 1.0367005059957096e-05,
-      "loss": 0.4715,
-      "step": 778
-    },
-    {
-      "epoch": 4.015463917525773,
-      "grad_norm": 0.7088159322738647,
-      "learning_rate": 1.0263143065305769e-05,
-      "loss": 0.4329,
-      "step": 779
-    },
-    {
-      "epoch": 4.020618556701031,
-      "grad_norm": 0.6903660893440247,
-      "learning_rate": 1.0159744429204777e-05,
-      "loss": 0.5347,
-      "step": 780
-    },
-    {
-      "epoch": 4.025773195876289,
-      "grad_norm": 0.6392103433609009,
-      "learning_rate": 1.005681035735192e-05,
-      "loss": 0.5343,
-      "step": 781
-    },
-    {
-      "epoch": 4.030927835051546,
-      "grad_norm": 0.6996196508407593,
-      "learning_rate": 9.954342050027921e-06,
-      "loss": 0.5407,
-      "step": 782
-    },
-    {
-      "epoch": 4.036082474226804,
-      "grad_norm": 0.6675930619239807,
-      "learning_rate": 9.852340702082318e-06,
-      "loss": 0.4305,
-      "step": 783
-    },
-    {
-      "epoch": 4.041237113402062,
-      "grad_norm": 0.6556795239448547,
-      "learning_rate": 9.750807502919652e-06,
-      "loss": 0.5781,
-      "step": 784
-    },
-    {
-      "epoch": 4.046391752577319,
-      "grad_norm": 0.7473887205123901,
-      "learning_rate": 9.64974363648548e-06,
-      "loss": 0.733,
-      "step": 785
-    },
-    {
-      "epoch": 4.051546391752577,
-      "grad_norm": 0.774716317653656,
-      "learning_rate": 9.549150281252633e-06,
-      "loss": 0.5335,
-      "step": 786
-    },
-    {
-      "epoch": 4.056701030927835,
-      "grad_norm": 0.6763889193534851,
-      "learning_rate": 9.449028610207494e-06,
-      "loss": 0.5087,
-      "step": 787
-    },
-    {
-      "epoch": 4.061855670103093,
-      "grad_norm": 0.8375513553619385,
-      "learning_rate": 9.349379790836243e-06,
-      "loss": 0.6384,
-      "step": 788
-    },
-    {
-      "epoch": 4.06701030927835,
-      "grad_norm": 0.7478683590888977,
-      "learning_rate": 9.25020498511135e-06,
-      "loss": 0.6164,
-      "step": 789
-    },
-    {
-      "epoch": 4.072164948453608,
-      "grad_norm": 0.7629753947257996,
-      "learning_rate": 9.151505349477902e-06,
-      "loss": 0.5738,
-      "step": 790
-    },
-    {
-      "epoch": 4.077319587628866,
-      "grad_norm": 0.6709279417991638,
-      "learning_rate": 9.053282034840238e-06,
-      "loss": 0.5869,
-      "step": 791
-    },
-    {
-      "epoch": 4.082474226804123,
-      "grad_norm": 0.6376342177391052,
-      "learning_rate": 8.955536186548425e-06,
-      "loss": 0.5019,
-      "step": 792
-    },
-    {
-      "epoch": 4.087628865979381,
-      "grad_norm": 0.7179467082023621,
-      "learning_rate": 8.858268944384995e-06,
-      "loss": 0.5092,
-      "step": 793
-    },
-    {
-      "epoch": 4.092783505154639,
-      "grad_norm": 0.789688229560852,
-      "learning_rate": 8.761481442551573e-06,
-      "loss": 0.6966,
-      "step": 794
-    },
-    {
-      "epoch": 4.097938144329897,
-      "grad_norm": 0.8050559163093567,
-      "learning_rate": 8.665174809655708e-06,
-      "loss": 0.6005,
-      "step": 795
-    },
-    {
-      "epoch": 4.103092783505154,
-      "grad_norm": 0.7821665406227112,
-      "learning_rate": 8.569350168697704e-06,
-      "loss": 0.6106,
-      "step": 796
-    },
-    {
-      "epoch": 4.108247422680412,
-      "grad_norm": 0.778394341468811,
-      "learning_rate": 8.474008637057478e-06,
-      "loss": 0.5891,
-      "step": 797
-    },
-    {
-      "epoch": 4.11340206185567,
-      "grad_norm": 0.9033699035644531,
-      "learning_rate": 8.379151326481587e-06,
-      "loss": 0.5541,
-      "step": 798
-    },
-    {
-      "epoch": 4.118556701030927,
-      "grad_norm": 0.8777570724487305,
-      "learning_rate": 8.284779343070265e-06,
-      "loss": 0.5675,
-      "step": 799
-    },
-    {
-      "epoch": 4.123711340206185,
-      "grad_norm": 0.6416880488395691,
-      "learning_rate": 8.19089378726447e-06,
-      "loss": 0.4717,
-      "step": 800
-    },
-    {
-      "epoch": 4.128865979381443,
-      "grad_norm": 0.7612876296043396,
-      "learning_rate": 8.097495753833078e-06,
-      "loss": 0.6074,
-      "step": 801
-    },
-    {
-      "epoch": 4.134020618556701,
-      "grad_norm": 0.8172696828842163,
-      "learning_rate": 8.004586331860175e-06,
-      "loss": 0.6051,
-      "step": 802
-    },
-    {
-      "epoch": 4.139175257731959,
-      "grad_norm": 0.7682762145996094,
-      "learning_rate": 7.91216660473228e-06,
-      "loss": 0.5949,
-      "step": 803
-    },
-    {
-      "epoch": 4.144329896907217,
-      "grad_norm": 0.7049791812896729,
-      "learning_rate": 7.820237650125712e-06,
-      "loss": 0.6494,
-      "step": 804
-    },
-    {
-      "epoch": 4.149484536082475,
-      "grad_norm": 0.9258434176445007,
-      "learning_rate": 7.728800539994113e-06,
-      "loss": 0.7203,
-      "step": 805
-    },
-    {
-      "epoch": 4.154639175257732,
-      "grad_norm": 0.7941154837608337,
-      "learning_rate": 7.637856340555822e-06,
-      "loss": 0.5589,
-      "step": 806
-    },
-    {
-      "epoch": 4.15979381443299,
-      "grad_norm": 0.8220565319061279,
-      "learning_rate": 7.547406112281557e-06,
-      "loss": 0.5536,
-      "step": 807
-    },
-    {
-      "epoch": 4.164948453608248,
-      "grad_norm": 0.7357336282730103,
-      "learning_rate": 7.457450909881969e-06,
-      "loss": 0.5021,
-      "step": 808
-    },
-    {
-      "epoch": 4.170103092783505,
-      "grad_norm": 0.8894550800323486,
-      "learning_rate": 7.367991782295391e-06,
-      "loss": 0.4791,
-      "step": 809
-    },
-    {
-      "epoch": 4.175257731958763,
-      "grad_norm": 0.8629480004310608,
-      "learning_rate": 7.2790297726755716e-06,
-      "loss": 0.4588,
-      "step": 810
-    },
-    {
-      "epoch": 4.180412371134021,
-      "grad_norm": 0.8441846370697021,
-      "learning_rate": 7.190565918379549e-06,
-      "loss": 0.6378,
-      "step": 811
-    },
-    {
-      "epoch": 4.185567010309279,
-      "grad_norm": 0.8104022145271301,
-      "learning_rate": 7.1026012509555265e-06,
-      "loss": 0.5777,
-      "step": 812
-    },
-    {
-      "epoch": 4.190721649484536,
-      "grad_norm": 0.855922281742096,
-      "learning_rate": 7.015136796130828e-06,
-      "loss": 0.6161,
-      "step": 813
-    },
-    {
-      "epoch": 4.195876288659794,
-      "grad_norm": 0.7620590329170227,
-      "learning_rate": 6.928173573800006e-06,
-      "loss": 0.5267,
-      "step": 814
-    },
-    {
-      "epoch": 4.201030927835052,
-      "grad_norm": 0.8558603525161743,
-      "learning_rate": 6.8417125980128675e-06,
-      "loss": 0.6221,
-      "step": 815
-    },
-    {
-      "epoch": 4.206185567010309,
-      "grad_norm": 0.7612592577934265,
-      "learning_rate": 6.755754876962711e-06,
-      "loss": 0.5259,
-      "step": 816
-    },
-    {
-      "epoch": 4.211340206185567,
-      "grad_norm": 0.6571881771087646,
-      "learning_rate": 6.670301412974511e-06,
-      "loss": 0.4846,
-      "step": 817
-    },
-    {
-      "epoch": 4.216494845360825,
-      "grad_norm": 0.8638062477111816,
-      "learning_rate": 6.585353202493322e-06,
-      "loss": 0.5727,
-      "step": 818
-    },
-    {
-      "epoch": 4.221649484536083,
-      "grad_norm": 0.7832995653152466,
-      "learning_rate": 6.500911236072532e-06,
-      "loss": 0.6475,
-      "step": 819
-    },
-    {
-      "epoch": 4.22680412371134,
-      "grad_norm": 0.7984924912452698,
-      "learning_rate": 6.416976498362432e-06,
-      "loss": 0.4848,
-      "step": 820
-    },
-    {
-      "epoch": 4.231958762886598,
-      "grad_norm": 0.743242084980011,
-      "learning_rate": 6.333549968098684e-06,
-      "loss": 0.5729,
-      "step": 821
-    },
-    {
-      "epoch": 4.237113402061856,
-      "grad_norm": 0.8120692372322083,
-      "learning_rate": 6.250632618090868e-06,
-      "loss": 0.4599,
-      "step": 822
-    },
-    {
-      "epoch": 4.242268041237113,
-      "grad_norm": 0.8082096576690674,
-      "learning_rate": 6.168225415211226e-06,
-      "loss": 0.6201,
-      "step": 823
-    },
-    {
-      "epoch": 4.247422680412371,
-      "grad_norm": 0.8432015180587769,
-      "learning_rate": 6.0863293203833105e-06,
-      "loss": 0.6398,
-      "step": 824
-    },
-    {
-      "epoch": 4.252577319587629,
-      "grad_norm": 0.7766333818435669,
-      "learning_rate": 6.004945288570813e-06,
-      "loss": 0.517,
-      "step": 825
     }
   ],
   "logging_steps": 1,
@@ -5873,7 +279,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.1905813118530355e+17,
+  "total_flos": 1.354811597389824e+16,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null