diff --git "a/qwen3-vl-4b-agentnet-50pect-bsz384-step1824/trainer_state.json" "b/qwen3-vl-4b-agentnet-50pect-bsz384-step1824/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/qwen3-vl-4b-agentnet-50pect-bsz384-step1824/trainer_state.json"
@@ -0,0 +1,12802 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 10000.0,
+  "global_step": 1824,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005482456140350877,
+      "grad_norm": 14.492247581481934,
+      "learning_rate": 2.173913043478261e-07,
+      "loss": 1.176474928855896,
+      "step": 1
+    },
+    {
+      "epoch": 0.0010964912280701754,
+      "grad_norm": 14.619416236877441,
+      "learning_rate": 4.347826086956522e-07,
+      "loss": 1.1729357242584229,
+      "step": 2
+    },
+    {
+      "epoch": 0.001644736842105263,
+      "grad_norm": 14.84022045135498,
+      "learning_rate": 6.521739130434783e-07,
+      "loss": 1.2002801895141602,
+      "step": 3
+    },
+    {
+      "epoch": 0.0021929824561403508,
+      "grad_norm": 14.705394744873047,
+      "learning_rate": 8.695652173913044e-07,
+      "loss": 1.1794195175170898,
+      "step": 4
+    },
+    {
+      "epoch": 0.0027412280701754384,
+      "grad_norm": 14.440802574157715,
+      "learning_rate": 1.0869565217391306e-06,
+      "loss": 1.1603755950927734,
+      "step": 5
+    },
+    {
+      "epoch": 0.003289473684210526,
+      "grad_norm": 13.81436824798584,
+      "learning_rate": 1.3043478260869566e-06,
+      "loss": 1.129558801651001,
+      "step": 6
+    },
+    {
+      "epoch": 0.003837719298245614,
+      "grad_norm": 13.689131736755371,
+      "learning_rate": 1.521739130434783e-06,
+      "loss": 1.1161764860153198,
+      "step": 7
+    },
+    {
+      "epoch": 0.0043859649122807015,
+      "grad_norm": 12.181388854980469,
+      "learning_rate": 1.7391304347826088e-06,
+      "loss": 1.0747873783111572,
+      "step": 8
+    },
+    {
+      "epoch": 0.004934210526315789,
+      "grad_norm": 11.249547958374023,
+      "learning_rate": 1.956521739130435e-06,
+      "loss": 1.041851282119751,
+      "step": 9
+    },
+    {
+      "epoch": 0.005482456140350877,
+      "grad_norm": 7.409827709197998,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 0.9274959564208984,
+      "step": 10
+    },
+    {
+      "epoch": 0.006030701754385965,
+      "grad_norm": 6.691342353820801,
+      "learning_rate": 2.391304347826087e-06,
+      "loss": 0.8955242037773132,
+      "step": 11
+    },
+    {
+      "epoch": 0.006578947368421052,
+      "grad_norm": 5.925269603729248,
+      "learning_rate": 2.6086956521739132e-06,
+      "loss": 0.8827022314071655,
+      "step": 12
+    },
+    {
+      "epoch": 0.00712719298245614,
+      "grad_norm": 5.35697078704834,
+      "learning_rate": 2.8260869565217393e-06,
+      "loss": 0.8560182452201843,
+      "step": 13
+    },
+    {
+      "epoch": 0.007675438596491228,
+      "grad_norm": 2.069866895675659,
+      "learning_rate": 3.043478260869566e-06,
+      "loss": 0.7693394422531128,
+      "step": 14
+    },
+    {
+      "epoch": 0.008223684210526315,
+      "grad_norm": 1.9829829931259155,
+      "learning_rate": 3.2608695652173914e-06,
+      "loss": 0.7666728496551514,
+      "step": 15
+    },
+    {
+      "epoch": 0.008771929824561403,
+      "grad_norm": 2.0215251445770264,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 0.7644944190979004,
+      "step": 16
+    },
+    {
+      "epoch": 0.00932017543859649,
+      "grad_norm": 1.8338474035263062,
+      "learning_rate": 3.6956521739130436e-06,
+      "loss": 0.764083743095398,
+      "step": 17
+    },
+    {
+      "epoch": 0.009868421052631578,
+      "grad_norm": 1.620351791381836,
+      "learning_rate": 3.91304347826087e-06,
+      "loss": 0.7492412328720093,
+      "step": 18
+    },
+    {
+      "epoch": 0.010416666666666666,
+      "grad_norm": 2.259765863418579,
+      "learning_rate": 4.130434782608696e-06,
+      "loss": 0.7317222952842712,
+      "step": 19
+    },
+    {
+      "epoch": 0.010964912280701754,
+      "grad_norm": 2.3859360218048096,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 0.743137776851654,
+      "step": 20
+    },
+    {
+      "epoch": 0.011513157894736841,
+      "grad_norm": 1.8937352895736694,
+      "learning_rate": 4.565217391304348e-06,
+      "loss": 0.7197248935699463,
+      "step": 21
+    },
+    {
+      "epoch": 0.01206140350877193,
+      "grad_norm": 1.3714830875396729,
+      "learning_rate": 4.782608695652174e-06,
+      "loss": 0.7112306356430054,
+      "step": 22
+    },
+    {
+      "epoch": 0.012609649122807017,
+      "grad_norm": 1.0794036388397217,
+      "learning_rate": 5e-06,
+      "loss": 0.7069730758666992,
+      "step": 23
+    },
+    {
+      "epoch": 0.013157894736842105,
+      "grad_norm": 1.132503628730774,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 0.7090061902999878,
+      "step": 24
+    },
+    {
+      "epoch": 0.013706140350877192,
+      "grad_norm": 1.0365830659866333,
+      "learning_rate": 5.4347826086956525e-06,
+      "loss": 0.6998612284660339,
+      "step": 25
+    },
+    {
+      "epoch": 0.01425438596491228,
+      "grad_norm": 0.9834604859352112,
+      "learning_rate": 5.652173913043479e-06,
+      "loss": 0.6926441192626953,
+      "step": 26
+    },
+    {
+      "epoch": 0.014802631578947368,
+      "grad_norm": 1.0544486045837402,
+      "learning_rate": 5.8695652173913055e-06,
+      "loss": 0.665709376335144,
+      "step": 27
+    },
+    {
+      "epoch": 0.015350877192982455,
+      "grad_norm": 0.98713219165802,
+      "learning_rate": 6.086956521739132e-06,
+      "loss": 0.6699036955833435,
+      "step": 28
+    },
+    {
+      "epoch": 0.015899122807017545,
+      "grad_norm": 0.9099668860435486,
+      "learning_rate": 6.304347826086958e-06,
+      "loss": 0.6747962236404419,
+      "step": 29
+    },
+    {
+      "epoch": 0.01644736842105263,
+      "grad_norm": 0.9079251885414124,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 0.6670761108398438,
+      "step": 30
+    },
+    {
+      "epoch": 0.01699561403508772,
+      "grad_norm": 0.9192748069763184,
+      "learning_rate": 6.739130434782609e-06,
+      "loss": 0.6650468707084656,
+      "step": 31
+    },
+    {
+      "epoch": 0.017543859649122806,
+      "grad_norm": 0.8508347868919373,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 0.6534518003463745,
+      "step": 32
+    },
+    {
+      "epoch": 0.018092105263157895,
+      "grad_norm": 0.7839338183403015,
+      "learning_rate": 7.173913043478261e-06,
+      "loss": 0.6688723564147949,
+      "step": 33
+    },
+    {
+      "epoch": 0.01864035087719298,
+      "grad_norm": 0.8197664022445679,
+      "learning_rate": 7.391304347826087e-06,
+      "loss": 0.6579426527023315,
+      "step": 34
+    },
+    {
+      "epoch": 0.01918859649122807,
+      "grad_norm": 0.7888069748878479,
+      "learning_rate": 7.608695652173914e-06,
+      "loss": 0.6640398502349854,
+      "step": 35
+    },
+    {
+      "epoch": 0.019736842105263157,
+      "grad_norm": 0.7756356596946716,
+      "learning_rate": 7.82608695652174e-06,
+      "loss": 0.6506577730178833,
+      "step": 36
+    },
+    {
+      "epoch": 0.020285087719298246,
+      "grad_norm": 0.7720519304275513,
+      "learning_rate": 8.043478260869566e-06,
+      "loss": 0.6552173495292664,
+      "step": 37
+    },
+    {
+      "epoch": 0.020833333333333332,
+      "grad_norm": 0.742103099822998,
+      "learning_rate": 8.260869565217392e-06,
+      "loss": 0.6349525451660156,
+      "step": 38
+    },
+    {
+      "epoch": 0.02138157894736842,
+      "grad_norm": 0.6862183213233948,
+      "learning_rate": 8.478260869565218e-06,
+      "loss": 0.6491415500640869,
+      "step": 39
+    },
+    {
+      "epoch": 0.021929824561403508,
+      "grad_norm": 0.6290478110313416,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 0.6458108425140381,
+      "step": 40
+    },
+    {
+      "epoch": 0.022478070175438597,
+      "grad_norm": 0.6633671522140503,
+      "learning_rate": 8.91304347826087e-06,
+      "loss": 0.6461598873138428,
+      "step": 41
+    },
+    {
+      "epoch": 0.023026315789473683,
+      "grad_norm": 0.846064567565918,
+      "learning_rate": 9.130434782608697e-06,
+      "loss": 0.6433469653129578,
+      "step": 42
+    },
+    {
+      "epoch": 0.023574561403508772,
+      "grad_norm": 0.6658814549446106,
+      "learning_rate": 9.347826086956523e-06,
+      "loss": 0.6385210752487183,
+      "step": 43
+    },
+    {
+      "epoch": 0.02412280701754386,
+      "grad_norm": 0.6573966145515442,
+      "learning_rate": 9.565217391304349e-06,
+      "loss": 0.6399390697479248,
+      "step": 44
+    },
+    {
+      "epoch": 0.024671052631578948,
+      "grad_norm": 0.6428568959236145,
+      "learning_rate": 9.782608695652175e-06,
+      "loss": 0.6466711163520813,
+      "step": 45
+    },
+    {
+      "epoch": 0.025219298245614034,
+      "grad_norm": 0.6197839975357056,
+      "learning_rate": 1e-05,
+      "loss": 0.6346309781074524,
+      "step": 46
+    },
+    {
+      "epoch": 0.025767543859649123,
+      "grad_norm": 0.6529486179351807,
+      "learning_rate": 1.0217391304347829e-05,
+      "loss": 0.6368366479873657,
+      "step": 47
+    },
+    {
+      "epoch": 0.02631578947368421,
+      "grad_norm": 0.6419752836227417,
+      "learning_rate": 1.0434782608695653e-05,
+      "loss": 0.6297695636749268,
+      "step": 48
+    },
+    {
+      "epoch": 0.0268640350877193,
+      "grad_norm": 0.6125337481498718,
+      "learning_rate": 1.0652173913043479e-05,
+      "loss": 0.6156260967254639,
+      "step": 49
+    },
+    {
+      "epoch": 0.027412280701754384,
+      "grad_norm": 0.6913661360740662,
+      "learning_rate": 1.0869565217391305e-05,
+      "loss": 0.6182829737663269,
+      "step": 50
+    },
+    {
+      "epoch": 0.027960526315789474,
+      "grad_norm": 0.6140983700752258,
+      "learning_rate": 1.1086956521739131e-05,
+      "loss": 0.6342036724090576,
+      "step": 51
+    },
+    {
+      "epoch": 0.02850877192982456,
+      "grad_norm": 0.6957651376724243,
+      "learning_rate": 1.1304347826086957e-05,
+      "loss": 0.6274335384368896,
+      "step": 52
+    },
+    {
+      "epoch": 0.02905701754385965,
+      "grad_norm": 0.6550323367118835,
+      "learning_rate": 1.1521739130434783e-05,
+      "loss": 0.6252555847167969,
+      "step": 53
+    },
+    {
+      "epoch": 0.029605263157894735,
+      "grad_norm": 0.6278795003890991,
+      "learning_rate": 1.1739130434782611e-05,
+      "loss": 0.6444627642631531,
+      "step": 54
+    },
+    {
+      "epoch": 0.030153508771929825,
+      "grad_norm": 0.6082501411437988,
+      "learning_rate": 1.1956521739130435e-05,
+      "loss": 0.6297181844711304,
+      "step": 55
+    },
+    {
+      "epoch": 0.03070175438596491,
+      "grad_norm": 0.6487641930580139,
+      "learning_rate": 1.2173913043478263e-05,
+      "loss": 0.6289809346199036,
+      "step": 56
+    },
+    {
+      "epoch": 0.03125,
+      "grad_norm": 0.6447638869285583,
+      "learning_rate": 1.2391304347826088e-05,
+      "loss": 0.6308567523956299,
+      "step": 57
+    },
+    {
+      "epoch": 0.03179824561403509,
+      "grad_norm": 0.7155942916870117,
+      "learning_rate": 1.2608695652173915e-05,
+      "loss": 0.6242796182632446,
+      "step": 58
+    },
+    {
+      "epoch": 0.03234649122807018,
+      "grad_norm": 0.7464308738708496,
+      "learning_rate": 1.282608695652174e-05,
+      "loss": 0.631054699420929,
+      "step": 59
+    },
+    {
+      "epoch": 0.03289473684210526,
+      "grad_norm": 0.7163922786712646,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 0.6311353445053101,
+      "step": 60
+    },
+    {
+      "epoch": 0.03344298245614035,
+      "grad_norm": 0.7227900624275208,
+      "learning_rate": 1.3260869565217392e-05,
+      "loss": 0.6270331144332886,
+      "step": 61
+    },
+    {
+      "epoch": 0.03399122807017544,
+      "grad_norm": 0.6737267971038818,
+      "learning_rate": 1.3478260869565218e-05,
+      "loss": 0.6225444674491882,
+      "step": 62
+    },
+    {
+      "epoch": 0.03453947368421053,
+      "grad_norm": 0.8492542505264282,
+      "learning_rate": 1.3695652173913046e-05,
+      "loss": 0.611880898475647,
+      "step": 63
+    },
+    {
+      "epoch": 0.03508771929824561,
+      "grad_norm": 0.6433826088905334,
+      "learning_rate": 1.391304347826087e-05,
+      "loss": 0.6079282760620117,
+      "step": 64
+    },
+    {
+      "epoch": 0.0356359649122807,
+      "grad_norm": 0.7155340909957886,
+      "learning_rate": 1.4130434782608698e-05,
+      "loss": 0.6314013004302979,
+      "step": 65
+    },
+    {
+      "epoch": 0.03618421052631579,
+      "grad_norm": 0.6798359155654907,
+      "learning_rate": 1.4347826086956522e-05,
+      "loss": 0.6237474679946899,
+      "step": 66
+    },
+    {
+      "epoch": 0.03673245614035088,
+      "grad_norm": 0.6824196577072144,
+      "learning_rate": 1.456521739130435e-05,
+      "loss": 0.6280307173728943,
+      "step": 67
+    },
+    {
+      "epoch": 0.03728070175438596,
+      "grad_norm": 0.6930661201477051,
+      "learning_rate": 1.4782608695652174e-05,
+      "loss": 0.6118326783180237,
+      "step": 68
+    },
+    {
+      "epoch": 0.03782894736842105,
+      "grad_norm": 0.6970058083534241,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.6004877090454102,
+      "step": 69
+    },
+    {
+      "epoch": 0.03837719298245614,
+      "grad_norm": 0.6858695149421692,
+      "learning_rate": 1.5217391304347828e-05,
+      "loss": 0.6087049245834351,
+      "step": 70
+    },
+    {
+      "epoch": 0.03892543859649123,
+      "grad_norm": 0.6748414039611816,
+      "learning_rate": 1.5434782608695654e-05,
+      "loss": 0.6046065092086792,
+      "step": 71
+    },
+    {
+      "epoch": 0.039473684210526314,
+      "grad_norm": 0.6753906011581421,
+      "learning_rate": 1.565217391304348e-05,
+      "loss": 0.6173025965690613,
+      "step": 72
+    },
+    {
+      "epoch": 0.0400219298245614,
+      "grad_norm": 0.6844945549964905,
+      "learning_rate": 1.5869565217391306e-05,
+      "loss": 0.6172990202903748,
+      "step": 73
+    },
+    {
+      "epoch": 0.04057017543859649,
+      "grad_norm": 0.7520304918289185,
+      "learning_rate": 1.6086956521739132e-05,
+      "loss": 0.6136633157730103,
+      "step": 74
+    },
+    {
+      "epoch": 0.04111842105263158,
+      "grad_norm": 0.6270356774330139,
+      "learning_rate": 1.630434782608696e-05,
+      "loss": 0.6032594442367554,
+      "step": 75
+    },
+    {
+      "epoch": 0.041666666666666664,
+      "grad_norm": 0.7684335112571716,
+      "learning_rate": 1.6521739130434785e-05,
+      "loss": 0.6158262491226196,
+      "step": 76
+    },
+    {
+      "epoch": 0.042214912280701754,
+      "grad_norm": 0.5890095829963684,
+      "learning_rate": 1.673913043478261e-05,
+      "loss": 0.602159857749939,
+      "step": 77
+    },
+    {
+      "epoch": 0.04276315789473684,
+      "grad_norm": 0.6600304245948792,
+      "learning_rate": 1.6956521739130437e-05,
+      "loss": 0.613303542137146,
+      "step": 78
+    },
+    {
+      "epoch": 0.04331140350877193,
+      "grad_norm": 0.6526859402656555,
+      "learning_rate": 1.7173913043478263e-05,
+      "loss": 0.5942760705947876,
+      "step": 79
+    },
+    {
+      "epoch": 0.043859649122807015,
+      "grad_norm": 0.6946115493774414,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 0.611393928527832,
+      "step": 80
+    },
+    {
+      "epoch": 0.044407894736842105,
+      "grad_norm": 0.753236711025238,
+      "learning_rate": 1.7608695652173915e-05,
+      "loss": 0.6026621460914612,
+      "step": 81
+    },
+    {
+      "epoch": 0.044956140350877194,
+      "grad_norm": 0.6258090734481812,
+      "learning_rate": 1.782608695652174e-05,
+      "loss": 0.6033834218978882,
+      "step": 82
+    },
+    {
+      "epoch": 0.04550438596491228,
+      "grad_norm": 0.7388920783996582,
+      "learning_rate": 1.8043478260869567e-05,
+      "loss": 0.6103101968765259,
+      "step": 83
+    },
+    {
+      "epoch": 0.046052631578947366,
+      "grad_norm": 0.6599609851837158,
+      "learning_rate": 1.8260869565217393e-05,
+      "loss": 0.600335955619812,
+      "step": 84
+    },
+    {
+      "epoch": 0.046600877192982455,
+      "grad_norm": 0.7393732070922852,
+      "learning_rate": 1.847826086956522e-05,
+      "loss": 0.6183399558067322,
+      "step": 85
+    },
+    {
+      "epoch": 0.047149122807017545,
+      "grad_norm": 0.744738757610321,
+      "learning_rate": 1.8695652173913045e-05,
+      "loss": 0.6110402345657349,
+      "step": 86
+    },
+    {
+      "epoch": 0.047697368421052634,
+      "grad_norm": 0.7020261883735657,
+      "learning_rate": 1.891304347826087e-05,
+      "loss": 0.6024118661880493,
+      "step": 87
+    },
+    {
+      "epoch": 0.04824561403508772,
+      "grad_norm": 0.725536584854126,
+      "learning_rate": 1.9130434782608697e-05,
+      "loss": 0.5905160903930664,
+      "step": 88
+    },
+    {
+      "epoch": 0.048793859649122806,
+      "grad_norm": 0.7676002383232117,
+      "learning_rate": 1.9347826086956523e-05,
+      "loss": 0.5904608368873596,
+      "step": 89
+    },
+    {
+      "epoch": 0.049342105263157895,
+      "grad_norm": 0.7609053254127502,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 0.6026548147201538,
+      "step": 90
+    },
+    {
+      "epoch": 0.049890350877192985,
+      "grad_norm": 0.7540615200996399,
+      "learning_rate": 1.9782608695652176e-05,
+      "loss": 0.5942332744598389,
+      "step": 91
+    },
+    {
+      "epoch": 0.05043859649122807,
+      "grad_norm": 0.7101155519485474,
+      "learning_rate": 2e-05,
+      "loss": 0.5820369720458984,
+      "step": 92
+    },
+    {
+      "epoch": 0.05098684210526316,
+      "grad_norm": 0.7118958830833435,
+      "learning_rate": 1.9999983549698757e-05,
+      "loss": 0.5976592898368835,
+      "step": 93
+    },
+    {
+      "epoch": 0.051535087719298246,
+      "grad_norm": 0.7405442595481873,
+      "learning_rate": 1.9999934198849154e-05,
+      "loss": 0.5931892395019531,
+      "step": 94
+    },
+    {
+      "epoch": 0.052083333333333336,
+      "grad_norm": 0.6989566683769226,
+      "learning_rate": 1.999985194761355e-05,
+      "loss": 0.5946756601333618,
+      "step": 95
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 0.7052030563354492,
+      "learning_rate": 1.9999736796262564e-05,
+      "loss": 0.5996489524841309,
+      "step": 96
+    },
+    {
+      "epoch": 0.05317982456140351,
+      "grad_norm": 0.7175880074501038,
+      "learning_rate": 1.9999588745175048e-05,
+      "loss": 0.6101623773574829,
+      "step": 97
+    },
+    {
+      "epoch": 0.0537280701754386,
+      "grad_norm": 0.727748692035675,
+      "learning_rate": 1.9999407794838095e-05,
+      "loss": 0.6104993224143982,
+      "step": 98
+    },
+    {
+      "epoch": 0.054276315789473686,
+      "grad_norm": 0.6567267775535583,
+      "learning_rate": 1.9999193945847056e-05,
+      "loss": 0.5962132215499878,
+      "step": 99
+    },
+    {
+      "epoch": 0.05482456140350877,
+      "grad_norm": 0.7101923227310181,
+      "learning_rate": 1.9998947198905492e-05,
+      "loss": 0.5943471193313599,
+      "step": 100
+    },
+    {
+      "epoch": 0.05537280701754386,
+      "grad_norm": 0.6486233472824097,
+      "learning_rate": 1.999866755482522e-05,
+      "loss": 0.5930988788604736,
+      "step": 101
+    },
+    {
+      "epoch": 0.05592105263157895,
+      "grad_norm": 0.6858486533164978,
+      "learning_rate": 1.999835501452629e-05,
+      "loss": 0.5910634994506836,
+      "step": 102
+    },
+    {
+      "epoch": 0.05646929824561404,
+      "grad_norm": 0.7296199798583984,
+      "learning_rate": 1.9998009579036976e-05,
+      "loss": 0.6147328019142151,
+      "step": 103
+    },
+    {
+      "epoch": 0.05701754385964912,
+      "grad_norm": 0.6833330988883972,
+      "learning_rate": 1.999763124949378e-05,
+      "loss": 0.5927976965904236,
+      "step": 104
+    },
+    {
+      "epoch": 0.05756578947368421,
+      "grad_norm": 0.7425616383552551,
+      "learning_rate": 1.999722002714143e-05,
+      "loss": 0.5903884172439575,
+      "step": 105
+    },
+    {
+      "epoch": 0.0581140350877193,
+      "grad_norm": 0.6378883719444275,
+      "learning_rate": 1.9996775913332875e-05,
+      "loss": 0.5842282772064209,
+      "step": 106
+    },
+    {
+      "epoch": 0.05866228070175439,
+      "grad_norm": 0.7103923559188843,
+      "learning_rate": 1.999629890952927e-05,
+      "loss": 0.6085933446884155,
+      "step": 107
+    },
+    {
+      "epoch": 0.05921052631578947,
+      "grad_norm": 0.660241425037384,
+      "learning_rate": 1.999578901729999e-05,
+      "loss": 0.6022235155105591,
+      "step": 108
+    },
+    {
+      "epoch": 0.05975877192982456,
+      "grad_norm": 0.7913385629653931,
+      "learning_rate": 1.9995246238322613e-05,
+      "loss": 0.5908679962158203,
+      "step": 109
+    },
+    {
+      "epoch": 0.06030701754385965,
+      "grad_norm": 0.6590815186500549,
+      "learning_rate": 1.9994670574382913e-05,
+      "loss": 0.6054195761680603,
+      "step": 110
+    },
+    {
+      "epoch": 0.06085526315789474,
+      "grad_norm": 0.6430388689041138,
+      "learning_rate": 1.9994062027374856e-05,
+      "loss": 0.5946575999259949,
+      "step": 111
+    },
+    {
+      "epoch": 0.06140350877192982,
+      "grad_norm": 0.6510061621665955,
+      "learning_rate": 1.9993420599300603e-05,
+      "loss": 0.5910765528678894,
+      "step": 112
+    },
+    {
+      "epoch": 0.06195175438596491,
+      "grad_norm": 0.6270714998245239,
+      "learning_rate": 1.999274629227049e-05,
+      "loss": 0.5879694223403931,
+      "step": 113
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 0.6500689387321472,
+      "learning_rate": 1.9992039108503024e-05,
+      "loss": 0.5984656810760498,
+      "step": 114
+    },
+    {
+      "epoch": 0.06304824561403509,
+      "grad_norm": 0.6544696092605591,
+      "learning_rate": 1.9991299050324887e-05,
+      "loss": 0.5880492925643921,
+      "step": 115
+    },
+    {
+      "epoch": 0.06359649122807018,
+      "grad_norm": 0.6615366339683533,
+      "learning_rate": 1.999052612017091e-05,
+      "loss": 0.589603066444397,
+      "step": 116
+    },
+    {
+      "epoch": 0.06414473684210527,
+      "grad_norm": 0.6696038842201233,
+      "learning_rate": 1.9989720320584085e-05,
+      "loss": 0.5883747339248657,
+      "step": 117
+    },
+    {
+      "epoch": 0.06469298245614036,
+      "grad_norm": 0.7159208059310913,
+      "learning_rate": 1.998888165421554e-05,
+      "loss": 0.5921615958213806,
+      "step": 118
+    },
+    {
+      "epoch": 0.06524122807017543,
+      "grad_norm": 0.6331257820129395,
+      "learning_rate": 1.9988010123824534e-05,
+      "loss": 0.5725486278533936,
+      "step": 119
+    },
+    {
+      "epoch": 0.06578947368421052,
+      "grad_norm": 0.7334661483764648,
+      "learning_rate": 1.998710573227846e-05,
+      "loss": 0.5945963859558105,
+      "step": 120
+    },
+    {
+      "epoch": 0.06633771929824561,
+      "grad_norm": 0.6409698128700256,
+      "learning_rate": 1.9986168482552814e-05,
+      "loss": 0.5865694880485535,
+      "step": 121
+    },
+    {
+      "epoch": 0.0668859649122807,
+      "grad_norm": 0.7104285955429077,
+      "learning_rate": 1.998519837773121e-05,
+      "loss": 0.5784432888031006,
+      "step": 122
+    },
+    {
+      "epoch": 0.06743421052631579,
+      "grad_norm": 0.6517587900161743,
+      "learning_rate": 1.998419542100535e-05,
+      "loss": 0.5863971710205078,
+      "step": 123
+    },
+    {
+      "epoch": 0.06798245614035088,
+      "grad_norm": 0.6840674877166748,
+      "learning_rate": 1.9983159615675022e-05,
+      "loss": 0.5794469118118286,
+      "step": 124
+    },
+    {
+      "epoch": 0.06853070175438597,
+      "grad_norm": 0.6648241877555847,
+      "learning_rate": 1.9982090965148086e-05,
+      "loss": 0.5815461874008179,
+      "step": 125
+    },
+    {
+      "epoch": 0.06907894736842106,
+      "grad_norm": 0.6593918204307556,
+      "learning_rate": 1.9980989472940467e-05,
+      "loss": 0.5929619073867798,
+      "step": 126
+    },
+    {
+      "epoch": 0.06962719298245613,
+      "grad_norm": 0.6910731792449951,
+      "learning_rate": 1.9979855142676142e-05,
+      "loss": 0.5822854042053223,
+      "step": 127
+    },
+    {
+      "epoch": 0.07017543859649122,
+      "grad_norm": 0.6526376605033875,
+      "learning_rate": 1.9978687978087125e-05,
+      "loss": 0.5812543630599976,
+      "step": 128
+    },
+    {
+      "epoch": 0.07072368421052631,
+      "grad_norm": 0.7658113241195679,
+      "learning_rate": 1.9977487983013463e-05,
+      "loss": 0.5871016383171082,
+      "step": 129
+    },
+    {
+      "epoch": 0.0712719298245614,
+      "grad_norm": 0.6897526979446411,
+      "learning_rate": 1.9976255161403203e-05,
+      "loss": 0.575676679611206,
+      "step": 130
+    },
+    {
+      "epoch": 0.07182017543859649,
+      "grad_norm": 0.7414442896842957,
+      "learning_rate": 1.9974989517312407e-05,
+      "loss": 0.5839799642562866,
+      "step": 131
+    },
+    {
+      "epoch": 0.07236842105263158,
+      "grad_norm": 0.7151236534118652,
+      "learning_rate": 1.997369105490512e-05,
+      "loss": 0.5924729704856873,
+      "step": 132
+    },
+    {
+      "epoch": 0.07291666666666667,
+      "grad_norm": 0.7733912467956543,
+      "learning_rate": 1.997235977845336e-05,
+      "loss": 0.5748937129974365,
+      "step": 133
+    },
+    {
+      "epoch": 0.07346491228070176,
+      "grad_norm": 0.7178257703781128,
+      "learning_rate": 1.9970995692337113e-05,
+      "loss": 0.5769957900047302,
+      "step": 134
+    },
+    {
+      "epoch": 0.07401315789473684,
+      "grad_norm": 0.7747276425361633,
+      "learning_rate": 1.99695988010443e-05,
+      "loss": 0.5812219381332397,
+      "step": 135
+    },
+    {
+      "epoch": 0.07456140350877193,
+      "grad_norm": 0.7646524310112,
+      "learning_rate": 1.9968169109170775e-05,
+      "loss": 0.5893447399139404,
+      "step": 136
+    },
+    {
+      "epoch": 0.07510964912280702,
+      "grad_norm": 0.7229993939399719,
+      "learning_rate": 1.9966706621420314e-05,
+      "loss": 0.574189305305481,
+      "step": 137
+    },
+    {
+      "epoch": 0.0756578947368421,
+      "grad_norm": 0.7084233164787292,
+      "learning_rate": 1.9965211342604586e-05,
+      "loss": 0.5811874866485596,
+      "step": 138
+    },
+    {
+      "epoch": 0.0762061403508772,
+      "grad_norm": 0.6773183941841125,
+      "learning_rate": 1.9963683277643153e-05,
+      "loss": 0.5971899628639221,
+      "step": 139
+    },
+    {
+      "epoch": 0.07675438596491228,
+      "grad_norm": 0.7109978199005127,
+      "learning_rate": 1.996212243156344e-05,
+      "loss": 0.5952876806259155,
+      "step": 140
+    },
+    {
+      "epoch": 0.07730263157894737,
+      "grad_norm": 0.6656246781349182,
+      "learning_rate": 1.9960528809500724e-05,
+      "loss": 0.5804654955863953,
+      "step": 141
+    },
+    {
+      "epoch": 0.07785087719298246,
+      "grad_norm": 0.7269256114959717,
+      "learning_rate": 1.995890241669811e-05,
+      "loss": 0.5656623244285583,
+      "step": 142
+    },
+    {
+      "epoch": 0.07839912280701754,
+      "grad_norm": 0.6435320377349854,
+      "learning_rate": 1.995724325850654e-05,
+      "loss": 0.569736897945404,
+      "step": 143
+    },
+    {
+      "epoch": 0.07894736842105263,
+      "grad_norm": 0.6682058572769165,
+      "learning_rate": 1.9955551340384745e-05,
+      "loss": 0.5740841627120972,
+      "step": 144
+    },
+    {
+      "epoch": 0.07949561403508772,
+      "grad_norm": 0.7501179575920105,
+      "learning_rate": 1.9953826667899228e-05,
+      "loss": 0.5707544088363647,
+      "step": 145
+    },
+    {
+      "epoch": 0.0800438596491228,
+      "grad_norm": 0.6460590362548828,
+      "learning_rate": 1.995206924672427e-05,
+      "loss": 0.5712607502937317,
+      "step": 146
+    },
+    {
+      "epoch": 0.0805921052631579,
+      "grad_norm": 0.656384289264679,
+      "learning_rate": 1.995027908264189e-05,
+      "loss": 0.5888504981994629,
+      "step": 147
+    },
+    {
+      "epoch": 0.08114035087719298,
+      "grad_norm": 0.7195465564727783,
+      "learning_rate": 1.994845618154184e-05,
+      "loss": 0.5889352560043335,
+      "step": 148
+    },
+    {
+      "epoch": 0.08168859649122807,
+      "grad_norm": 0.6684480905532837,
+      "learning_rate": 1.994660054942157e-05,
+      "loss": 0.590423047542572,
+      "step": 149
+    },
+    {
+      "epoch": 0.08223684210526316,
+      "grad_norm": 0.6501762270927429,
+      "learning_rate": 1.994471219238623e-05,
+      "loss": 0.5742418766021729,
+      "step": 150
+    },
+    {
+      "epoch": 0.08278508771929824,
+      "grad_norm": 0.6628527045249939,
+      "learning_rate": 1.9942791116648617e-05,
+      "loss": 0.5750651359558105,
+      "step": 151
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 0.6320564150810242,
+      "learning_rate": 1.994083732852919e-05,
+      "loss": 0.5674133896827698,
+      "step": 152
+    },
+    {
+      "epoch": 0.08388157894736842,
+      "grad_norm": 0.7114735841751099,
+      "learning_rate": 1.9938850834456036e-05,
+      "loss": 0.5808606743812561,
+      "step": 153
+    },
+    {
+      "epoch": 0.08442982456140351,
+      "grad_norm": 0.6949970126152039,
+      "learning_rate": 1.993683164096483e-05,
+      "loss": 0.5758386254310608,
+      "step": 154
+    },
+    {
+      "epoch": 0.0849780701754386,
+      "grad_norm": 0.5890199542045593,
+      "learning_rate": 1.9934779754698848e-05,
+      "loss": 0.5657817721366882,
+      "step": 155
+    },
+    {
+      "epoch": 0.08552631578947369,
+      "grad_norm": 0.632293164730072,
+      "learning_rate": 1.9932695182408917e-05,
+      "loss": 0.5779306292533875,
+      "step": 156
+    },
+    {
+      "epoch": 0.08607456140350878,
+      "grad_norm": 0.6128914952278137,
+      "learning_rate": 1.9930577930953406e-05,
+      "loss": 0.570749044418335,
+      "step": 157
+    },
+    {
+      "epoch": 0.08662280701754387,
+      "grad_norm": 0.6463309526443481,
+      "learning_rate": 1.99284280072982e-05,
+      "loss": 0.5824733972549438,
+      "step": 158
+    },
+    {
+      "epoch": 0.08717105263157894,
+      "grad_norm": 0.5726247429847717,
+      "learning_rate": 1.992624541851667e-05,
+      "loss": 0.5759953856468201,
+      "step": 159
+    },
+    {
+      "epoch": 0.08771929824561403,
+      "grad_norm": 0.6129015684127808,
+      "learning_rate": 1.9924030171789676e-05,
+      "loss": 0.5806936621665955,
+      "step": 160
+    },
+    {
+      "epoch": 0.08826754385964912,
+      "grad_norm": 0.5705315470695496,
+      "learning_rate": 1.992178227440551e-05,
+      "loss": 0.5579361915588379,
+      "step": 161
+    },
+    {
+      "epoch": 0.08881578947368421,
+      "grad_norm": 0.6253746151924133,
+      "learning_rate": 1.9919501733759882e-05,
+      "loss": 0.5733797550201416,
+      "step": 162
+    },
+    {
+      "epoch": 0.0893640350877193,
+      "grad_norm": 0.5756648182868958,
+      "learning_rate": 1.991718855735592e-05,
+      "loss": 0.5644710063934326,
+      "step": 163
+    },
+    {
+      "epoch": 0.08991228070175439,
+      "grad_norm": 0.6245459318161011,
+      "learning_rate": 1.9914842752804106e-05,
+      "loss": 0.5706383585929871,
+      "step": 164
+    },
+    {
+      "epoch": 0.09046052631578948,
+      "grad_norm": 0.5780031681060791,
+      "learning_rate": 1.9912464327822277e-05,
+      "loss": 0.5664913058280945,
+      "step": 165
+    },
+    {
+      "epoch": 0.09100877192982457,
+      "grad_norm": 0.5702175498008728,
+      "learning_rate": 1.99100532902356e-05,
+      "loss": 0.5698821544647217,
+      "step": 166
+    },
+    {
+      "epoch": 0.09155701754385964,
+      "grad_norm": 0.6294623613357544,
+      "learning_rate": 1.990760964797653e-05,
+      "loss": 0.5785145163536072,
+      "step": 167
+    },
+    {
+      "epoch": 0.09210526315789473,
+      "grad_norm": 0.5997359752655029,
+      "learning_rate": 1.99051334090848e-05,
+      "loss": 0.5714824199676514,
+      "step": 168
+    },
+    {
+      "epoch": 0.09265350877192982,
+      "grad_norm": 0.726883053779602,
+      "learning_rate": 1.9902624581707385e-05,
+      "loss": 0.5575656890869141,
+      "step": 169
+    },
+    {
+      "epoch": 0.09320175438596491,
+      "grad_norm": 0.5980774164199829,
+      "learning_rate": 1.9900083174098474e-05,
+      "loss": 0.5692469477653503,
+      "step": 170
+    },
+    {
+      "epoch": 0.09375,
+      "grad_norm": 0.6684207916259766,
+      "learning_rate": 1.989750919461946e-05,
+      "loss": 0.5694268941879272,
+      "step": 171
+    },
+    {
+      "epoch": 0.09429824561403509,
+      "grad_norm": 0.6513474583625793,
+      "learning_rate": 1.989490265173888e-05,
+      "loss": 0.5756077170372009,
+      "step": 172
+    },
+    {
+      "epoch": 0.09484649122807018,
+      "grad_norm": 0.61539626121521,
+      "learning_rate": 1.989226355403242e-05,
+      "loss": 0.571203887462616,
+      "step": 173
+    },
+    {
+      "epoch": 0.09539473684210527,
+      "grad_norm": 0.6571477055549622,
+      "learning_rate": 1.9889591910182878e-05,
+      "loss": 0.5581786632537842,
+      "step": 174
+    },
+    {
+      "epoch": 0.09594298245614036,
+      "grad_norm": 0.644450306892395,
+      "learning_rate": 1.9886887728980114e-05,
+      "loss": 0.5708379149436951,
+      "step": 175
+    },
+    {
+      "epoch": 0.09649122807017543,
+      "grad_norm": 0.6149620413780212,
+      "learning_rate": 1.9884151019321053e-05,
+      "loss": 0.5719337463378906,
+      "step": 176
+    },
+    {
+      "epoch": 0.09703947368421052,
+      "grad_norm": 0.6498218178749084,
+      "learning_rate": 1.9881381790209634e-05,
+      "loss": 0.5746930241584778,
+      "step": 177
+    },
+    {
+      "epoch": 0.09758771929824561,
+      "grad_norm": 0.6144945621490479,
+      "learning_rate": 1.987858005075678e-05,
+      "loss": 0.5792109966278076,
+      "step": 178
+    },
+    {
+      "epoch": 0.0981359649122807,
+      "grad_norm": 0.5841969847679138,
+      "learning_rate": 1.9875745810180394e-05,
+      "loss": 0.5848267078399658,
+      "step": 179
+    },
+    {
+      "epoch": 0.09868421052631579,
+      "grad_norm": 0.6415631175041199,
+      "learning_rate": 1.987287907780529e-05,
+      "loss": 0.5754548907279968,
+      "step": 180
+    },
+    {
+      "epoch": 0.09923245614035088,
+      "grad_norm": 0.5965489149093628,
+      "learning_rate": 1.9869979863063195e-05,
+      "loss": 0.5726244449615479,
+      "step": 181
+    },
+    {
+      "epoch": 0.09978070175438597,
+      "grad_norm": 0.6180820465087891,
+      "learning_rate": 1.9867048175492697e-05,
+      "loss": 0.5588217377662659,
+      "step": 182
+    },
+    {
+      "epoch": 0.10032894736842106,
+      "grad_norm": 0.6616155505180359,
+      "learning_rate": 1.9864084024739228e-05,
+      "loss": 0.5637444853782654,
+      "step": 183
+    },
+    {
+      "epoch": 0.10087719298245613,
+      "grad_norm": 0.582970142364502,
+      "learning_rate": 1.986108742055502e-05,
+      "loss": 0.5720033645629883,
+      "step": 184
+    },
+    {
+      "epoch": 0.10142543859649122,
+      "grad_norm": 0.604560136795044,
+      "learning_rate": 1.985805837279908e-05,
+      "loss": 0.5666155815124512,
+      "step": 185
+    },
+    {
+      "epoch": 0.10197368421052631,
+      "grad_norm": 0.6076090931892395,
+      "learning_rate": 1.985499689143716e-05,
+      "loss": 0.5611607432365417,
+      "step": 186
+    },
+    {
+      "epoch": 0.1025219298245614,
+      "grad_norm": 0.6265007853507996,
+      "learning_rate": 1.9851902986541714e-05,
+      "loss": 0.5762375593185425,
+      "step": 187
+    },
+    {
+      "epoch": 0.10307017543859649,
+      "grad_norm": 0.6510225534439087,
+      "learning_rate": 1.9848776668291885e-05,
+      "loss": 0.5727993845939636,
+      "step": 188
+    },
+    {
+      "epoch": 0.10361842105263158,
+      "grad_norm": 0.6026071906089783,
+      "learning_rate": 1.9845617946973445e-05,
+      "loss": 0.5830463171005249,
+      "step": 189
+    },
+    {
+      "epoch": 0.10416666666666667,
+      "grad_norm": 0.6423436999320984,
+      "learning_rate": 1.9842426832978766e-05,
+      "loss": 0.5695741176605225,
+      "step": 190
+    },
+    {
+      "epoch": 0.10471491228070176,
+      "grad_norm": 0.6096088290214539,
+      "learning_rate": 1.9839203336806822e-05,
+      "loss": 0.5536958575248718,
+      "step": 191
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.6307644248008728,
+      "learning_rate": 1.9835947469063098e-05,
+      "loss": 0.565772294998169,
+      "step": 192
+    },
+    {
+      "epoch": 0.10581140350877193,
+      "grad_norm": 0.6427217125892639,
+      "learning_rate": 1.98326592404596e-05,
+      "loss": 0.5632326006889343,
+      "step": 193
+    },
+    {
+      "epoch": 0.10635964912280702,
+      "grad_norm": 0.6283311247825623,
+      "learning_rate": 1.9829338661814798e-05,
+      "loss": 0.5652562975883484,
+      "step": 194
+    },
+    {
+      "epoch": 0.1069078947368421,
+      "grad_norm": 0.6760656833648682,
+      "learning_rate": 1.9825985744053594e-05,
+      "loss": 0.5492938160896301,
+      "step": 195
+    },
+    {
+      "epoch": 0.1074561403508772,
+      "grad_norm": 0.564399778842926,
+      "learning_rate": 1.9822600498207292e-05,
+      "loss": 0.5632847547531128,
+      "step": 196
+    },
+    {
+      "epoch": 0.10800438596491228,
+      "grad_norm": 0.6520889401435852,
+      "learning_rate": 1.981918293541355e-05,
+      "loss": 0.5750704407691956,
+      "step": 197
+    },
+    {
+      "epoch": 0.10855263157894737,
+      "grad_norm": 0.6204854249954224,
+      "learning_rate": 1.981573306691636e-05,
+      "loss": 0.5595630407333374,
+      "step": 198
+    },
+    {
+      "epoch": 0.10910087719298246,
+      "grad_norm": 0.6573178768157959,
+      "learning_rate": 1.9812250904065996e-05,
+      "loss": 0.5650523900985718,
+      "step": 199
+    },
+    {
+      "epoch": 0.10964912280701754,
+      "grad_norm": 0.6327071189880371,
+      "learning_rate": 1.9808736458318988e-05,
+      "loss": 0.5691615343093872,
+      "step": 200
+    },
+    {
+      "epoch": 0.11019736842105263,
+      "grad_norm": 0.5889390707015991,
+      "learning_rate": 1.9805189741238063e-05,
+      "loss": 0.5566825866699219,
+      "step": 201
+    },
+    {
+      "epoch": 0.11074561403508772,
+      "grad_norm": 0.6135177612304688,
+      "learning_rate": 1.9801610764492142e-05,
+      "loss": 0.5492175221443176,
+      "step": 202
+    },
+    {
+      "epoch": 0.1112938596491228,
+      "grad_norm": 0.6119044423103333,
+      "learning_rate": 1.9797999539856273e-05,
+      "loss": 0.55916827917099,
+      "step": 203
+    },
+    {
+      "epoch": 0.1118421052631579,
+      "grad_norm": 0.6478143930435181,
+      "learning_rate": 1.9794356079211605e-05,
+      "loss": 0.5680371522903442,
+      "step": 204
+    },
+    {
+      "epoch": 0.11239035087719298,
+      "grad_norm": 0.604614794254303,
+      "learning_rate": 1.979068039454534e-05,
+      "loss": 0.5741668939590454,
+      "step": 205
+    },
+    {
+      "epoch": 0.11293859649122807,
+      "grad_norm": 0.6160794496536255,
+      "learning_rate": 1.97869724979507e-05,
+      "loss": 0.5514141321182251,
+      "step": 206
+    },
+    {
+      "epoch": 0.11348684210526316,
+      "grad_norm": 0.6878485679626465,
+      "learning_rate": 1.9783232401626894e-05,
+      "loss": 0.5740391612052917,
+      "step": 207
+    },
+    {
+      "epoch": 0.11403508771929824,
+      "grad_norm": 0.6564989686012268,
+      "learning_rate": 1.9779460117879058e-05,
+      "loss": 0.5653142929077148,
+      "step": 208
+    },
+    {
+      "epoch": 0.11458333333333333,
+      "grad_norm": 0.6507999300956726,
+      "learning_rate": 1.9775655659118235e-05,
+      "loss": 0.5714613199234009,
+      "step": 209
+    },
+    {
+      "epoch": 0.11513157894736842,
+      "grad_norm": 0.5813721418380737,
+      "learning_rate": 1.9771819037861327e-05,
+      "loss": 0.5710395574569702,
+      "step": 210
+    },
+    {
+      "epoch": 0.11567982456140351,
+      "grad_norm": 0.6418222784996033,
+      "learning_rate": 1.9767950266731045e-05,
+      "loss": 0.5667479038238525,
+      "step": 211
+    },
+    {
+      "epoch": 0.1162280701754386,
+      "grad_norm": 0.6286323666572571,
+      "learning_rate": 1.976404935845588e-05,
+      "loss": 0.5599182844161987,
+      "step": 212
+    },
+    {
+      "epoch": 0.11677631578947369,
+      "grad_norm": 0.6144490242004395,
+      "learning_rate": 1.9760116325870053e-05,
+      "loss": 0.5594089031219482,
+      "step": 213
+    },
+    {
+      "epoch": 0.11732456140350878,
+      "grad_norm": 0.5968447327613831,
+      "learning_rate": 1.9756151181913483e-05,
+      "loss": 0.5569183826446533,
+      "step": 214
+    },
+    {
+      "epoch": 0.11787280701754387,
+      "grad_norm": 0.5543148517608643,
+      "learning_rate": 1.9752153939631733e-05,
+      "loss": 0.5508428812026978,
+      "step": 215
+    },
+    {
+      "epoch": 0.11842105263157894,
+      "grad_norm": 0.5568669438362122,
+      "learning_rate": 1.9748124612175967e-05,
+      "loss": 0.5641998648643494,
+      "step": 216
+    },
+    {
+      "epoch": 0.11896929824561403,
+      "grad_norm": 0.5984328985214233,
+      "learning_rate": 1.974406321280291e-05,
+      "loss": 0.5496900081634521,
+      "step": 217
+    },
+    {
+      "epoch": 0.11951754385964912,
+      "grad_norm": 0.7141798138618469,
+      "learning_rate": 1.9739969754874827e-05,
+      "loss": 0.54876708984375,
+      "step": 218
+    },
+    {
+      "epoch": 0.12006578947368421,
+      "grad_norm": 0.5774359703063965,
+      "learning_rate": 1.9735844251859425e-05,
+      "loss": 0.5808881521224976,
+      "step": 219
+    },
+    {
+      "epoch": 0.1206140350877193,
+      "grad_norm": 0.575563371181488,
+      "learning_rate": 1.9731686717329866e-05,
+      "loss": 0.5554652810096741,
+      "step": 220
+    },
+    {
+      "epoch": 0.12116228070175439,
+      "grad_norm": 0.5829038619995117,
+      "learning_rate": 1.9727497164964683e-05,
+      "loss": 0.5668386220932007,
+      "step": 221
+    },
+    {
+      "epoch": 0.12171052631578948,
+      "grad_norm": 0.6911284923553467,
+      "learning_rate": 1.9723275608547764e-05,
+      "loss": 0.5543766617774963,
+      "step": 222
+    },
+    {
+      "epoch": 0.12225877192982457,
+      "grad_norm": 0.5740495324134827,
+      "learning_rate": 1.971902206196828e-05,
+      "loss": 0.5608670115470886,
+      "step": 223
+    },
+    {
+      "epoch": 0.12280701754385964,
+      "grad_norm": 0.5948644876480103,
+      "learning_rate": 1.971473653922065e-05,
+      "loss": 0.5639830231666565,
+      "step": 224
+    },
+    {
+      "epoch": 0.12335526315789473,
+      "grad_norm": 0.5562756061553955,
+      "learning_rate": 1.9710419054404508e-05,
+      "loss": 0.5558134913444519,
+      "step": 225
+    },
+    {
+      "epoch": 0.12390350877192982,
+      "grad_norm": 0.589013934135437,
+      "learning_rate": 1.9706069621724643e-05,
+      "loss": 0.553057849407196,
+      "step": 226
+    },
+    {
+      "epoch": 0.12445175438596491,
+      "grad_norm": 0.5930812358856201,
+      "learning_rate": 1.970168825549094e-05,
+      "loss": 0.5508049726486206,
+      "step": 227
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.5706969499588013,
+      "learning_rate": 1.9697274970118365e-05,
+      "loss": 0.5701100826263428,
+      "step": 228
+    },
+    {
+      "epoch": 0.12554824561403508,
+      "grad_norm": 0.5633387565612793,
+      "learning_rate": 1.969282978012689e-05,
+      "loss": 0.5619994401931763,
+      "step": 229
+    },
+    {
+      "epoch": 0.12609649122807018,
+      "grad_norm": 0.6061105728149414,
+      "learning_rate": 1.968835270014146e-05,
+      "loss": 0.5603735446929932,
+      "step": 230
+    },
+    {
+      "epoch": 0.12664473684210525,
+      "grad_norm": 0.6145228743553162,
+      "learning_rate": 1.968384374489194e-05,
+      "loss": 0.5546954274177551,
+      "step": 231
+    },
+    {
+      "epoch": 0.12719298245614036,
+      "grad_norm": 0.5804170370101929,
+      "learning_rate": 1.967930292921306e-05,
+      "loss": 0.5411268472671509,
+      "step": 232
+    },
+    {
+      "epoch": 0.12774122807017543,
+      "grad_norm": 0.618490993976593,
+      "learning_rate": 1.967473026804438e-05,
+      "loss": 0.5622624158859253,
+      "step": 233
+    },
+    {
+      "epoch": 0.12828947368421054,
+      "grad_norm": 0.5959983468055725,
+      "learning_rate": 1.967012577643023e-05,
+      "loss": 0.5682417154312134,
+      "step": 234
+    },
+    {
+      "epoch": 0.1288377192982456,
+      "grad_norm": 0.7653847336769104,
+      "learning_rate": 1.966548946951966e-05,
+      "loss": 0.563367486000061,
+      "step": 235
+    },
+    {
+      "epoch": 0.12938596491228072,
+      "grad_norm": 0.5803858637809753,
+      "learning_rate": 1.9660821362566406e-05,
+      "loss": 0.5565321445465088,
+      "step": 236
+    },
+    {
+      "epoch": 0.1299342105263158,
+      "grad_norm": 0.5829041600227356,
+      "learning_rate": 1.9656121470928818e-05,
+      "loss": 0.568941593170166,
+      "step": 237
+    },
+    {
+      "epoch": 0.13048245614035087,
+      "grad_norm": 0.5557065010070801,
+      "learning_rate": 1.965138981006982e-05,
+      "loss": 0.5454533100128174,
+      "step": 238
+    },
+    {
+      "epoch": 0.13103070175438597,
+      "grad_norm": 0.6172096729278564,
+      "learning_rate": 1.964662639555687e-05,
+      "loss": 0.5447598695755005,
+      "step": 239
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 0.5418916344642639,
+      "learning_rate": 1.9641831243061878e-05,
+      "loss": 0.5526988506317139,
+      "step": 240
+    },
+    {
+      "epoch": 0.13212719298245615,
+      "grad_norm": 0.7085768580436707,
+      "learning_rate": 1.963700436836119e-05,
+      "loss": 0.5584993958473206,
+      "step": 241
+    },
+    {
+      "epoch": 0.13267543859649122,
+      "grad_norm": 0.6074377298355103,
+      "learning_rate": 1.963214578733552e-05,
+      "loss": 0.564329981803894,
+      "step": 242
+    },
+    {
+      "epoch": 0.13322368421052633,
+      "grad_norm": 0.6210350394248962,
+      "learning_rate": 1.9627255515969884e-05,
+      "loss": 0.5371741056442261,
+      "step": 243
+    },
+    {
+      "epoch": 0.1337719298245614,
+      "grad_norm": 0.5757085084915161,
+      "learning_rate": 1.962233357035357e-05,
+      "loss": 0.5590758323669434,
+      "step": 244
+    },
+    {
+      "epoch": 0.13432017543859648,
+      "grad_norm": 0.5684416890144348,
+      "learning_rate": 1.9617379966680077e-05,
+      "loss": 0.5677273273468018,
+      "step": 245
+    },
+    {
+      "epoch": 0.13486842105263158,
+      "grad_norm": 0.6423379182815552,
+      "learning_rate": 1.9612394721247063e-05,
+      "loss": 0.555820882320404,
+      "step": 246
+    },
+    {
+      "epoch": 0.13541666666666666,
+      "grad_norm": 0.5935850143432617,
+      "learning_rate": 1.960737785045628e-05,
+      "loss": 0.5559649467468262,
+      "step": 247
+    },
+    {
+      "epoch": 0.13596491228070176,
+      "grad_norm": 0.5883272290229797,
+      "learning_rate": 1.960232937081354e-05,
+      "loss": 0.5713253021240234,
+      "step": 248
+    },
+    {
+      "epoch": 0.13651315789473684,
+      "grad_norm": 0.6154787540435791,
+      "learning_rate": 1.9597249298928647e-05,
+      "loss": 0.5542075634002686,
+      "step": 249
+    },
+    {
+      "epoch": 0.13706140350877194,
+      "grad_norm": 0.5839691162109375,
+      "learning_rate": 1.9592137651515333e-05,
+      "loss": 0.5596610307693481,
+      "step": 250
+    },
+    {
+      "epoch": 0.13760964912280702,
+      "grad_norm": 0.5880065560340881,
+      "learning_rate": 1.9586994445391236e-05,
+      "loss": 0.5572192668914795,
+      "step": 251
+    },
+    {
+      "epoch": 0.13815789473684212,
+      "grad_norm": 0.5767812728881836,
+      "learning_rate": 1.9581819697477813e-05,
+      "loss": 0.5562660098075867,
+      "step": 252
+    },
+    {
+      "epoch": 0.1387061403508772,
+      "grad_norm": 0.57844477891922,
+      "learning_rate": 1.957661342480029e-05,
+      "loss": 0.5570098161697388,
+      "step": 253
+    },
+    {
+      "epoch": 0.13925438596491227,
+      "grad_norm": 0.5954297184944153,
+      "learning_rate": 1.9571375644487626e-05,
+      "loss": 0.5651256442070007,
+      "step": 254
+    },
+    {
+      "epoch": 0.13980263157894737,
+      "grad_norm": 0.6308397054672241,
+      "learning_rate": 1.956610637377243e-05,
+      "loss": 0.5574840903282166,
+      "step": 255
+    },
+    {
+      "epoch": 0.14035087719298245,
+      "grad_norm": 0.5578974485397339,
+      "learning_rate": 1.9560805629990917e-05,
+      "loss": 0.5468358993530273,
+      "step": 256
+    },
+    {
+      "epoch": 0.14089912280701755,
+      "grad_norm": 0.5789008140563965,
+      "learning_rate": 1.955547343058286e-05,
+      "loss": 0.5454850196838379,
+      "step": 257
+    },
+    {
+      "epoch": 0.14144736842105263,
+      "grad_norm": 0.5896825194358826,
+      "learning_rate": 1.955010979309151e-05,
+      "loss": 0.5465385317802429,
+      "step": 258
+    },
+    {
+      "epoch": 0.14199561403508773,
+      "grad_norm": 0.5821171402931213,
+      "learning_rate": 1.954471473516356e-05,
+      "loss": 0.5568474531173706,
+      "step": 259
+    },
+    {
+      "epoch": 0.1425438596491228,
+      "grad_norm": 0.609752893447876,
+      "learning_rate": 1.953928827454908e-05,
+      "loss": 0.5320336222648621,
+      "step": 260
+    },
+    {
+      "epoch": 0.14309210526315788,
+      "grad_norm": 0.5710572600364685,
+      "learning_rate": 1.953383042910144e-05,
+      "loss": 0.5608735680580139,
+      "step": 261
+    },
+    {
+      "epoch": 0.14364035087719298,
+      "grad_norm": 0.5719582438468933,
+      "learning_rate": 1.9528341216777296e-05,
+      "loss": 0.5430909991264343,
+      "step": 262
+    },
+    {
+      "epoch": 0.14418859649122806,
+      "grad_norm": 0.5792491436004639,
+      "learning_rate": 1.9522820655636475e-05,
+      "loss": 0.5550388693809509,
+      "step": 263
+    },
+    {
+      "epoch": 0.14473684210526316,
+      "grad_norm": 0.5587615370750427,
+      "learning_rate": 1.9517268763841965e-05,
+      "loss": 0.5363634824752808,
+      "step": 264
+    },
+    {
+      "epoch": 0.14528508771929824,
+      "grad_norm": 0.6018436551094055,
+      "learning_rate": 1.9511685559659816e-05,
+      "loss": 0.542389988899231,
+      "step": 265
+    },
+    {
+      "epoch": 0.14583333333333334,
+      "grad_norm": 0.5717230439186096,
+      "learning_rate": 1.950607106145911e-05,
+      "loss": 0.5556778907775879,
+      "step": 266
+    },
+    {
+      "epoch": 0.14638157894736842,
+      "grad_norm": 0.5293515920639038,
+      "learning_rate": 1.950042528771189e-05,
+      "loss": 0.5441399812698364,
+      "step": 267
+    },
+    {
+      "epoch": 0.14692982456140352,
+      "grad_norm": 0.5560038089752197,
+      "learning_rate": 1.9494748256993082e-05,
+      "loss": 0.5428458452224731,
+      "step": 268
+    },
+    {
+      "epoch": 0.1474780701754386,
+      "grad_norm": 0.5864049196243286,
+      "learning_rate": 1.9489039987980466e-05,
+      "loss": 0.5442084074020386,
+      "step": 269
+    },
+    {
+      "epoch": 0.14802631578947367,
+      "grad_norm": 0.5393058657646179,
+      "learning_rate": 1.9483300499454584e-05,
+      "loss": 0.5405887961387634,
+      "step": 270
+    },
+    {
+      "epoch": 0.14857456140350878,
+      "grad_norm": 0.5462262630462646,
+      "learning_rate": 1.9477529810298708e-05,
+      "loss": 0.5411845445632935,
+      "step": 271
+    },
+    {
+      "epoch": 0.14912280701754385,
+      "grad_norm": 0.570056676864624,
+      "learning_rate": 1.9471727939498746e-05,
+      "loss": 0.5517827868461609,
+      "step": 272
+    },
+    {
+      "epoch": 0.14967105263157895,
+      "grad_norm": 0.5791317820549011,
+      "learning_rate": 1.9465894906143206e-05,
+      "loss": 0.5526028871536255,
+      "step": 273
+    },
+    {
+      "epoch": 0.15021929824561403,
+      "grad_norm": 0.568991482257843,
+      "learning_rate": 1.9460030729423116e-05,
+      "loss": 0.5334166884422302,
+      "step": 274
+    },
+    {
+      "epoch": 0.15076754385964913,
+      "grad_norm": 0.5713422298431396,
+      "learning_rate": 1.9454135428631976e-05,
+      "loss": 0.5433835983276367,
+      "step": 275
+    },
+    {
+      "epoch": 0.1513157894736842,
+      "grad_norm": 0.5832074880599976,
+      "learning_rate": 1.9448209023165676e-05,
+      "loss": 0.5358787178993225,
+      "step": 276
+    },
+    {
+      "epoch": 0.15186403508771928,
+      "grad_norm": 0.562080979347229,
+      "learning_rate": 1.944225153252245e-05,
+      "loss": 0.5675616264343262,
+      "step": 277
+    },
+    {
+      "epoch": 0.1524122807017544,
+      "grad_norm": 0.5883001089096069,
+      "learning_rate": 1.9436262976302797e-05,
+      "loss": 0.5515336990356445,
+      "step": 278
+    },
+    {
+      "epoch": 0.15296052631578946,
+      "grad_norm": 0.5644806027412415,
+      "learning_rate": 1.9430243374209432e-05,
+      "loss": 0.5458020567893982,
+      "step": 279
+    },
+    {
+      "epoch": 0.15350877192982457,
+      "grad_norm": 0.5437331199645996,
+      "learning_rate": 1.9424192746047207e-05,
+      "loss": 0.5555316805839539,
+      "step": 280
+    },
+    {
+      "epoch": 0.15405701754385964,
+      "grad_norm": 0.571792483329773,
+      "learning_rate": 1.9418111111723056e-05,
+      "loss": 0.5325362682342529,
+      "step": 281
+    },
+    {
+      "epoch": 0.15460526315789475,
+      "grad_norm": 0.5624408721923828,
+      "learning_rate": 1.9411998491245917e-05,
+      "loss": 0.5465089082717896,
+      "step": 282
+    },
+    {
+      "epoch": 0.15515350877192982,
+      "grad_norm": 0.5995544791221619,
+      "learning_rate": 1.940585490472668e-05,
+      "loss": 0.5414011478424072,
+      "step": 283
+    },
+    {
+      "epoch": 0.15570175438596492,
+      "grad_norm": 0.6020653247833252,
+      "learning_rate": 1.9399680372378122e-05,
+      "loss": 0.5542804002761841,
+      "step": 284
+    },
+    {
+      "epoch": 0.15625,
+      "grad_norm": 0.5900858044624329,
+      "learning_rate": 1.9393474914514817e-05,
+      "loss": 0.5281364917755127,
+      "step": 285
+    },
+    {
+      "epoch": 0.15679824561403508,
+      "grad_norm": 0.5869341492652893,
+      "learning_rate": 1.9387238551553103e-05,
+      "loss": 0.5379394292831421,
+      "step": 286
+    },
+    {
+      "epoch": 0.15734649122807018,
+      "grad_norm": 0.6075244545936584,
+      "learning_rate": 1.9380971304010984e-05,
+      "loss": 0.5490105152130127,
+      "step": 287
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.5797610282897949,
+      "learning_rate": 1.937467319250809e-05,
+      "loss": 0.5395088195800781,
+      "step": 288
+    },
+    {
+      "epoch": 0.15844298245614036,
+      "grad_norm": 0.63418048620224,
+      "learning_rate": 1.9368344237765576e-05,
+      "loss": 0.5533697605133057,
+      "step": 289
+    },
+    {
+      "epoch": 0.15899122807017543,
+      "grad_norm": 0.5689693689346313,
+      "learning_rate": 1.9361984460606093e-05,
+      "loss": 0.5429133772850037,
+      "step": 290
+    },
+    {
+      "epoch": 0.15953947368421054,
+      "grad_norm": 0.5940237641334534,
+      "learning_rate": 1.9355593881953685e-05,
+      "loss": 0.5357928276062012,
+      "step": 291
+    },
+    {
+      "epoch": 0.1600877192982456,
+      "grad_norm": 0.5879078507423401,
+      "learning_rate": 1.9349172522833747e-05,
+      "loss": 0.5496682524681091,
+      "step": 292
+    },
+    {
+      "epoch": 0.16063596491228072,
+      "grad_norm": 0.6074959635734558,
+      "learning_rate": 1.9342720404372938e-05,
+      "loss": 0.5387532114982605,
+      "step": 293
+    },
+    {
+      "epoch": 0.1611842105263158,
+      "grad_norm": 0.6277965307235718,
+      "learning_rate": 1.9336237547799108e-05,
+      "loss": 0.5568438172340393,
+      "step": 294
+    },
+    {
+      "epoch": 0.16173245614035087,
+      "grad_norm": 0.5389106869697571,
+      "learning_rate": 1.9329723974441253e-05,
+      "loss": 0.5376222133636475,
+      "step": 295
+    },
+    {
+      "epoch": 0.16228070175438597,
+      "grad_norm": 0.544335126876831,
+      "learning_rate": 1.932317970572942e-05,
+      "loss": 0.5440766215324402,
+      "step": 296
+    },
+    {
+      "epoch": 0.16282894736842105,
+      "grad_norm": 0.6015032529830933,
+      "learning_rate": 1.9316604763194644e-05,
+      "loss": 0.545868456363678,
+      "step": 297
+    },
+    {
+      "epoch": 0.16337719298245615,
+      "grad_norm": 0.5448634624481201,
+      "learning_rate": 1.9309999168468887e-05,
+      "loss": 0.5508019328117371,
+      "step": 298
+    },
+    {
+      "epoch": 0.16392543859649122,
+      "grad_norm": 0.5765957832336426,
+      "learning_rate": 1.930336294328495e-05,
+      "loss": 0.5424585342407227,
+      "step": 299
+    },
+    {
+      "epoch": 0.16447368421052633,
+      "grad_norm": 0.5915053486824036,
+      "learning_rate": 1.9296696109476417e-05,
+      "loss": 0.5394760966300964,
+      "step": 300
+    },
+    {
+      "epoch": 0.1650219298245614,
+      "grad_norm": 0.579708456993103,
+      "learning_rate": 1.9289998688977573e-05,
+      "loss": 0.5493593215942383,
+      "step": 301
+    },
+    {
+      "epoch": 0.16557017543859648,
+      "grad_norm": 0.5751671195030212,
+      "learning_rate": 1.928327070382333e-05,
+      "loss": 0.5433315634727478,
+      "step": 302
+    },
+    {
+      "epoch": 0.16611842105263158,
+      "grad_norm": 0.5499085783958435,
+      "learning_rate": 1.927651217614917e-05,
+      "loss": 0.5484105348587036,
+      "step": 303
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.5443341732025146,
+      "learning_rate": 1.926972312819105e-05,
+      "loss": 0.5325586199760437,
+      "step": 304
+    },
+    {
+      "epoch": 0.16721491228070176,
+      "grad_norm": 0.5641672015190125,
+      "learning_rate": 1.9262903582285352e-05,
+      "loss": 0.5600139498710632,
+      "step": 305
+    },
+    {
+      "epoch": 0.16776315789473684,
+      "grad_norm": 0.5435248613357544,
+      "learning_rate": 1.9256053560868793e-05,
+      "loss": 0.5464034080505371,
+      "step": 306
+    },
+    {
+      "epoch": 0.16831140350877194,
+      "grad_norm": 0.5279967784881592,
+      "learning_rate": 1.9249173086478352e-05,
+      "loss": 0.5260881781578064,
+      "step": 307
+    },
+    {
+      "epoch": 0.16885964912280702,
+      "grad_norm": 0.533305287361145,
+      "learning_rate": 1.9242262181751207e-05,
+      "loss": 0.5478596091270447,
+      "step": 308
+    },
+    {
+      "epoch": 0.16940789473684212,
+      "grad_norm": 0.543569028377533,
+      "learning_rate": 1.9235320869424656e-05,
+      "loss": 0.5306735038757324,
+      "step": 309
+    },
+    {
+      "epoch": 0.1699561403508772,
+      "grad_norm": 0.5425512790679932,
+      "learning_rate": 1.9228349172336023e-05,
+      "loss": 0.5532375574111938,
+      "step": 310
+    },
+    {
+      "epoch": 0.17050438596491227,
+      "grad_norm": 0.5322792530059814,
+      "learning_rate": 1.9221347113422625e-05,
+      "loss": 0.5421377420425415,
+      "step": 311
+    },
+    {
+      "epoch": 0.17105263157894737,
+      "grad_norm": 0.5279988646507263,
+      "learning_rate": 1.9214314715721646e-05,
+      "loss": 0.544383704662323,
+      "step": 312
+    },
+    {
+      "epoch": 0.17160087719298245,
+      "grad_norm": 0.5296905040740967,
+      "learning_rate": 1.9207252002370103e-05,
+      "loss": 0.5539359450340271,
+      "step": 313
+    },
+    {
+      "epoch": 0.17214912280701755,
+      "grad_norm": 0.5464092493057251,
+      "learning_rate": 1.920015899660475e-05,
+      "loss": 0.5502333641052246,
+      "step": 314
+    },
+    {
+      "epoch": 0.17269736842105263,
+      "grad_norm": 0.5278061628341675,
+      "learning_rate": 1.9193035721762002e-05,
+      "loss": 0.541316032409668,
+      "step": 315
+    },
+    {
+      "epoch": 0.17324561403508773,
+      "grad_norm": 0.5732148289680481,
+      "learning_rate": 1.918588220127786e-05,
+      "loss": 0.5292676687240601,
+      "step": 316
+    },
+    {
+      "epoch": 0.1737938596491228,
+      "grad_norm": 0.5602945685386658,
+      "learning_rate": 1.9178698458687844e-05,
+      "loss": 0.5401516556739807,
+      "step": 317
+    },
+    {
+      "epoch": 0.17434210526315788,
+      "grad_norm": 0.581879198551178,
+      "learning_rate": 1.9171484517626893e-05,
+      "loss": 0.5478727221488953,
+      "step": 318
+    },
+    {
+      "epoch": 0.17489035087719298,
+      "grad_norm": 0.49734967947006226,
+      "learning_rate": 1.9164240401829308e-05,
+      "loss": 0.5426466464996338,
+      "step": 319
+    },
+    {
+      "epoch": 0.17543859649122806,
+      "grad_norm": 0.6240032315254211,
+      "learning_rate": 1.915696613512867e-05,
+      "loss": 0.5309844017028809,
+      "step": 320
+    },
+    {
+      "epoch": 0.17598684210526316,
+      "grad_norm": 0.543289303779602,
+      "learning_rate": 1.9149661741457755e-05,
+      "loss": 0.52463698387146,
+      "step": 321
+    },
+    {
+      "epoch": 0.17653508771929824,
+      "grad_norm": 0.5531445145606995,
+      "learning_rate": 1.914232724484845e-05,
+      "loss": 0.5420910120010376,
+      "step": 322
+    },
+    {
+      "epoch": 0.17708333333333334,
+      "grad_norm": 0.5311645865440369,
+      "learning_rate": 1.9134962669431705e-05,
+      "loss": 0.5336942672729492,
+      "step": 323
+    },
+    {
+      "epoch": 0.17763157894736842,
+      "grad_norm": 0.5692971348762512,
+      "learning_rate": 1.912756803943741e-05,
+      "loss": 0.5401374697685242,
+      "step": 324
+    },
+    {
+      "epoch": 0.17817982456140352,
+      "grad_norm": 0.5489525198936462,
+      "learning_rate": 1.9120143379194337e-05,
+      "loss": 0.5350661277770996,
+      "step": 325
+    },
+    {
+      "epoch": 0.1787280701754386,
+      "grad_norm": 0.5243253707885742,
+      "learning_rate": 1.911268871313007e-05,
+      "loss": 0.5375982522964478,
+      "step": 326
+    },
+    {
+      "epoch": 0.17927631578947367,
+      "grad_norm": 0.5977224111557007,
+      "learning_rate": 1.910520406577091e-05,
+      "loss": 0.5465099215507507,
+      "step": 327
+    },
+    {
+      "epoch": 0.17982456140350878,
+      "grad_norm": 0.5903772711753845,
+      "learning_rate": 1.9097689461741805e-05,
+      "loss": 0.5511510372161865,
+      "step": 328
+    },
+    {
+      "epoch": 0.18037280701754385,
+      "grad_norm": 0.678267240524292,
+      "learning_rate": 1.909014492576624e-05,
+      "loss": 0.528895378112793,
+      "step": 329
+    },
+    {
+      "epoch": 0.18092105263157895,
+      "grad_norm": 0.5741175413131714,
+      "learning_rate": 1.9082570482666208e-05,
+      "loss": 0.5306143760681152,
+      "step": 330
+    },
+    {
+      "epoch": 0.18146929824561403,
+      "grad_norm": 0.5529358983039856,
+      "learning_rate": 1.9074966157362072e-05,
+      "loss": 0.5563026666641235,
+      "step": 331
+    },
+    {
+      "epoch": 0.18201754385964913,
+      "grad_norm": 0.5264307856559753,
+      "learning_rate": 1.9067331974872524e-05,
+      "loss": 0.5446305871009827,
+      "step": 332
+    },
+    {
+      "epoch": 0.1825657894736842,
+      "grad_norm": 0.5596775412559509,
+      "learning_rate": 1.9059667960314486e-05,
+      "loss": 0.5350501537322998,
+      "step": 333
+    },
+    {
+      "epoch": 0.18311403508771928,
+      "grad_norm": 0.5849246978759766,
+      "learning_rate": 1.9051974138903028e-05,
+      "loss": 0.5333458781242371,
+      "step": 334
+    },
+    {
+      "epoch": 0.1836622807017544,
+      "grad_norm": 0.5496376752853394,
+      "learning_rate": 1.9044250535951285e-05,
+      "loss": 0.5393158793449402,
+      "step": 335
+    },
+    {
+      "epoch": 0.18421052631578946,
+      "grad_norm": 0.5456401109695435,
+      "learning_rate": 1.903649717687037e-05,
+      "loss": 0.5372017025947571,
+      "step": 336
+    },
+    {
+      "epoch": 0.18475877192982457,
+      "grad_norm": 0.5497374534606934,
+      "learning_rate": 1.902871408716931e-05,
+      "loss": 0.5300045609474182,
+      "step": 337
+    },
+    {
+      "epoch": 0.18530701754385964,
+      "grad_norm": 0.5420397520065308,
+      "learning_rate": 1.902090129245494e-05,
+      "loss": 0.5352155566215515,
+      "step": 338
+    },
+    {
+      "epoch": 0.18585526315789475,
+      "grad_norm": 0.5222429037094116,
+      "learning_rate": 1.9013058818431817e-05,
+      "loss": 0.5446464419364929,
+      "step": 339
+    },
+    {
+      "epoch": 0.18640350877192982,
+      "grad_norm": 0.558821976184845,
+      "learning_rate": 1.9005186690902157e-05,
+      "loss": 0.5501251220703125,
+      "step": 340
+    },
+    {
+      "epoch": 0.18695175438596492,
+      "grad_norm": 0.5371629595756531,
+      "learning_rate": 1.8997284935765735e-05,
+      "loss": 0.5312443971633911,
+      "step": 341
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 0.6179410815238953,
+      "learning_rate": 1.89893535790198e-05,
+      "loss": 0.5460267066955566,
+      "step": 342
+    },
+    {
+      "epoch": 0.18804824561403508,
+      "grad_norm": 0.5250634551048279,
+      "learning_rate": 1.8981392646758996e-05,
+      "loss": 0.5355827808380127,
+      "step": 343
+    },
+    {
+      "epoch": 0.18859649122807018,
+      "grad_norm": 0.5826337933540344,
+      "learning_rate": 1.897340216517527e-05,
+      "loss": 0.5478694438934326,
+      "step": 344
+    },
+    {
+      "epoch": 0.18914473684210525,
+      "grad_norm": 0.5900570750236511,
+      "learning_rate": 1.8965382160557782e-05,
+      "loss": 0.5379772186279297,
+      "step": 345
+    },
+    {
+      "epoch": 0.18969298245614036,
+      "grad_norm": 0.5510210990905762,
+      "learning_rate": 1.8957332659292834e-05,
+      "loss": 0.5222848653793335,
+      "step": 346
+    },
+    {
+      "epoch": 0.19024122807017543,
+      "grad_norm": 0.6057240962982178,
+      "learning_rate": 1.8949253687863774e-05,
+      "loss": 0.5323337316513062,
+      "step": 347
+    },
+    {
+      "epoch": 0.19078947368421054,
+      "grad_norm": 0.5485894680023193,
+      "learning_rate": 1.8941145272850902e-05,
+      "loss": 0.5386132001876831,
+      "step": 348
+    },
+    {
+      "epoch": 0.1913377192982456,
+      "grad_norm": 0.585307776927948,
+      "learning_rate": 1.8933007440931385e-05,
+      "loss": 0.5313419103622437,
+      "step": 349
+    },
+    {
+      "epoch": 0.19188596491228072,
+      "grad_norm": 0.551096498966217,
+      "learning_rate": 1.892484021887919e-05,
+      "loss": 0.5472394227981567,
+      "step": 350
+    },
+    {
+      "epoch": 0.1924342105263158,
+      "grad_norm": 0.5347589254379272,
+      "learning_rate": 1.8916643633564966e-05,
+      "loss": 0.5171635746955872,
+      "step": 351
+    },
+    {
+      "epoch": 0.19298245614035087,
+      "grad_norm": 0.5558242201805115,
+      "learning_rate": 1.8908417711955973e-05,
+      "loss": 0.529994547367096,
+      "step": 352
+    },
+    {
+      "epoch": 0.19353070175438597,
+      "grad_norm": 0.5679911971092224,
+      "learning_rate": 1.8900162481115986e-05,
+      "loss": 0.5229799151420593,
+      "step": 353
+    },
+    {
+      "epoch": 0.19407894736842105,
+      "grad_norm": 0.5330626964569092,
+      "learning_rate": 1.8891877968205213e-05,
+      "loss": 0.54994136095047,
+      "step": 354
+    },
+    {
+      "epoch": 0.19462719298245615,
+      "grad_norm": 0.5267235040664673,
+      "learning_rate": 1.8883564200480208e-05,
+      "loss": 0.5337533950805664,
+      "step": 355
+    },
+    {
+      "epoch": 0.19517543859649122,
+      "grad_norm": 0.5202456712722778,
+      "learning_rate": 1.8875221205293755e-05,
+      "loss": 0.5240612626075745,
+      "step": 356
+    },
+    {
+      "epoch": 0.19572368421052633,
+      "grad_norm": 0.550948977470398,
+      "learning_rate": 1.886684901009482e-05,
+      "loss": 0.525791585445404,
+      "step": 357
+    },
+    {
+      "epoch": 0.1962719298245614,
+      "grad_norm": 0.508272647857666,
+      "learning_rate": 1.8858447642428426e-05,
+      "loss": 0.5174123048782349,
+      "step": 358
+    },
+    {
+      "epoch": 0.19682017543859648,
+      "grad_norm": 0.565328061580658,
+      "learning_rate": 1.885001712993558e-05,
+      "loss": 0.5382243394851685,
+      "step": 359
+    },
+    {
+      "epoch": 0.19736842105263158,
+      "grad_norm": 0.5756270885467529,
+      "learning_rate": 1.8841557500353175e-05,
+      "loss": 0.5362300276756287,
+      "step": 360
+    },
+    {
+      "epoch": 0.19791666666666666,
+      "grad_norm": 0.54173344373703,
+      "learning_rate": 1.8833068781513902e-05,
+      "loss": 0.5481382012367249,
+      "step": 361
+    },
+    {
+      "epoch": 0.19846491228070176,
+      "grad_norm": 0.5440255999565125,
+      "learning_rate": 1.882455100134616e-05,
+      "loss": 0.5397817492485046,
+      "step": 362
+    },
+    {
+      "epoch": 0.19901315789473684,
+      "grad_norm": 0.5340250134468079,
+      "learning_rate": 1.8816004187873956e-05,
+      "loss": 0.5356112718582153,
+      "step": 363
+    },
+    {
+      "epoch": 0.19956140350877194,
+      "grad_norm": 0.5714414119720459,
+      "learning_rate": 1.8807428369216822e-05,
+      "loss": 0.5300136804580688,
+      "step": 364
+    },
+    {
+      "epoch": 0.20010964912280702,
+      "grad_norm": 0.5363687872886658,
+      "learning_rate": 1.879882357358972e-05,
+      "loss": 0.5302585363388062,
+      "step": 365
+    },
+    {
+      "epoch": 0.20065789473684212,
+      "grad_norm": 0.5832900404930115,
+      "learning_rate": 1.8790189829302943e-05,
+      "loss": 0.5480587482452393,
+      "step": 366
+    },
+    {
+      "epoch": 0.2012061403508772,
+      "grad_norm": 0.557942807674408,
+      "learning_rate": 1.878152716476203e-05,
+      "loss": 0.5448884963989258,
+      "step": 367
+    },
+    {
+      "epoch": 0.20175438596491227,
+      "grad_norm": 0.5794870853424072,
+      "learning_rate": 1.8772835608467673e-05,
+      "loss": 0.534662663936615,
+      "step": 368
+    },
+    {
+      "epoch": 0.20230263157894737,
+      "grad_norm": 0.5900275707244873,
+      "learning_rate": 1.876411518901561e-05,
+      "loss": 0.5280211567878723,
+      "step": 369
+    },
+    {
+      "epoch": 0.20285087719298245,
+      "grad_norm": 0.5694456696510315,
+      "learning_rate": 1.875536593509655e-05,
+      "loss": 0.5381162166595459,
+      "step": 370
+    },
+    {
+      "epoch": 0.20339912280701755,
+      "grad_norm": 0.5804063677787781,
+      "learning_rate": 1.874658787549607e-05,
+      "loss": 0.5450135469436646,
+      "step": 371
+    },
+    {
+      "epoch": 0.20394736842105263,
+      "grad_norm": 0.5541572570800781,
+      "learning_rate": 1.87377810390945e-05,
+      "loss": 0.5326411724090576,
+      "step": 372
+    },
+    {
+      "epoch": 0.20449561403508773,
+      "grad_norm": 0.5617303252220154,
+      "learning_rate": 1.8728945454866883e-05,
+      "loss": 0.5263036489486694,
+      "step": 373
+    },
+    {
+      "epoch": 0.2050438596491228,
+      "grad_norm": 0.5810173153877258,
+      "learning_rate": 1.872008115188281e-05,
+      "loss": 0.5250884890556335,
+      "step": 374
+    },
+    {
+      "epoch": 0.20559210526315788,
+      "grad_norm": 0.5345124006271362,
+      "learning_rate": 1.8711188159306374e-05,
+      "loss": 0.5363500714302063,
+      "step": 375
+    },
+    {
+      "epoch": 0.20614035087719298,
+      "grad_norm": 0.5278041958808899,
+      "learning_rate": 1.8702266506396055e-05,
+      "loss": 0.5263749361038208,
+      "step": 376
+    },
+    {
+      "epoch": 0.20668859649122806,
+      "grad_norm": 0.5594993233680725,
+      "learning_rate": 1.8693316222504635e-05,
+      "loss": 0.5244511961936951,
+      "step": 377
+    },
+    {
+      "epoch": 0.20723684210526316,
+      "grad_norm": 0.4992246925830841,
+      "learning_rate": 1.868433733707908e-05,
+      "loss": 0.5366593599319458,
+      "step": 378
+    },
+    {
+      "epoch": 0.20778508771929824,
+      "grad_norm": 0.59859699010849,
+      "learning_rate": 1.867532987966047e-05,
+      "loss": 0.5325961112976074,
+      "step": 379
+    },
+    {
+      "epoch": 0.20833333333333334,
+      "grad_norm": 0.5476097464561462,
+      "learning_rate": 1.8666293879883875e-05,
+      "loss": 0.5390352606773376,
+      "step": 380
+    },
+    {
+      "epoch": 0.20888157894736842,
+      "grad_norm": 0.5338640213012695,
+      "learning_rate": 1.8657229367478286e-05,
+      "loss": 0.5377339720726013,
+      "step": 381
+    },
+    {
+      "epoch": 0.20942982456140352,
+      "grad_norm": 0.5540474057197571,
+      "learning_rate": 1.864813637226649e-05,
+      "loss": 0.5436443090438843,
+      "step": 382
+    },
+    {
+      "epoch": 0.2099780701754386,
+      "grad_norm": 0.5551073551177979,
+      "learning_rate": 1.8639014924164998e-05,
+      "loss": 0.5364835858345032,
+      "step": 383
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.5687278509140015,
+      "learning_rate": 1.862986505318391e-05,
+      "loss": 0.5310265421867371,
+      "step": 384
+    },
+    {
+      "epoch": 0.21107456140350878,
+      "grad_norm": 0.5749037265777588,
+      "learning_rate": 1.8620686789426864e-05,
+      "loss": 0.534489631652832,
+      "step": 385
+    },
+    {
+      "epoch": 0.21162280701754385,
+      "grad_norm": 0.5327162742614746,
+      "learning_rate": 1.8611480163090897e-05,
+      "loss": 0.5294643640518188,
+      "step": 386
+    },
+    {
+      "epoch": 0.21217105263157895,
+      "grad_norm": 0.5638524889945984,
+      "learning_rate": 1.8602245204466363e-05,
+      "loss": 0.5338449478149414,
+      "step": 387
+    },
+    {
+      "epoch": 0.21271929824561403,
+      "grad_norm": 0.5824586749076843,
+      "learning_rate": 1.859298194393683e-05,
+      "loss": 0.5351516008377075,
+      "step": 388
+    },
+    {
+      "epoch": 0.21326754385964913,
+      "grad_norm": 0.5394695401191711,
+      "learning_rate": 1.8583690411978993e-05,
+      "loss": 0.5316178202629089,
+      "step": 389
+    },
+    {
+      "epoch": 0.2138157894736842,
+      "grad_norm": 0.5534507632255554,
+      "learning_rate": 1.857437063916254e-05,
+      "loss": 0.5347613096237183,
+      "step": 390
+    },
+    {
+      "epoch": 0.21436403508771928,
+      "grad_norm": 1.921692132949829,
+      "learning_rate": 1.856502265615009e-05,
+      "loss": 0.5341858863830566,
+      "step": 391
+    },
+    {
+      "epoch": 0.2149122807017544,
+      "grad_norm": 0.5521547198295593,
+      "learning_rate": 1.8555646493697075e-05,
+      "loss": 0.5187911987304688,
+      "step": 392
+    },
+    {
+      "epoch": 0.21546052631578946,
+      "grad_norm": 0.5613106489181519,
+      "learning_rate": 1.8546242182651627e-05,
+      "loss": 0.5340178608894348,
+      "step": 393
+    },
+    {
+      "epoch": 0.21600877192982457,
+      "grad_norm": 0.5031777024269104,
+      "learning_rate": 1.8536809753954502e-05,
+      "loss": 0.5230652093887329,
+      "step": 394
+    },
+    {
+      "epoch": 0.21655701754385964,
+      "grad_norm": 0.530637264251709,
+      "learning_rate": 1.852734923863895e-05,
+      "loss": 0.5346648693084717,
+      "step": 395
+    },
+    {
+      "epoch": 0.21710526315789475,
+      "grad_norm": 0.543538510799408,
+      "learning_rate": 1.851786066783065e-05,
+      "loss": 0.539505124092102,
+      "step": 396
+    },
+    {
+      "epoch": 0.21765350877192982,
+      "grad_norm": 0.5140752196311951,
+      "learning_rate": 1.850834407274756e-05,
+      "loss": 0.5221940279006958,
+      "step": 397
+    },
+    {
+      "epoch": 0.21820175438596492,
+      "grad_norm": 0.5462019443511963,
+      "learning_rate": 1.8498799484699856e-05,
+      "loss": 0.5088600516319275,
+      "step": 398
+    },
+    {
+      "epoch": 0.21875,
+      "grad_norm": 0.5661113858222961,
+      "learning_rate": 1.8489226935089805e-05,
+      "loss": 0.5322563648223877,
+      "step": 399
+    },
+    {
+      "epoch": 0.21929824561403508,
+      "grad_norm": 0.5222601890563965,
+      "learning_rate": 1.847962645541168e-05,
+      "loss": 0.5383967161178589,
+      "step": 400
+    },
+    {
+      "epoch": 0.21984649122807018,
+      "grad_norm": 0.5328034162521362,
+      "learning_rate": 1.8469998077251623e-05,
+      "loss": 0.5342406630516052,
+      "step": 401
+    },
+    {
+      "epoch": 0.22039473684210525,
+      "grad_norm": 0.5401366949081421,
+      "learning_rate": 1.846034183228759e-05,
+      "loss": 0.5323405861854553,
+      "step": 402
+    },
+    {
+      "epoch": 0.22094298245614036,
+      "grad_norm": 0.5516491532325745,
+      "learning_rate": 1.8450657752289204e-05,
+      "loss": 0.5423470735549927,
+      "step": 403
+    },
+    {
+      "epoch": 0.22149122807017543,
+      "grad_norm": 0.5484375953674316,
+      "learning_rate": 1.8440945869117676e-05,
+      "loss": 0.5268591642379761,
+      "step": 404
+    },
+    {
+      "epoch": 0.22203947368421054,
+      "grad_norm": 0.542039155960083,
+      "learning_rate": 1.843120621472568e-05,
+      "loss": 0.5249344706535339,
+      "step": 405
+    },
+    {
+      "epoch": 0.2225877192982456,
+      "grad_norm": 0.5442968010902405,
+      "learning_rate": 1.842143882115727e-05,
+      "loss": 0.5164295434951782,
+      "step": 406
+    },
+    {
+      "epoch": 0.22313596491228072,
+      "grad_norm": 0.5494210124015808,
+      "learning_rate": 1.841164372054776e-05,
+      "loss": 0.5340598821640015,
+      "step": 407
+    },
+    {
+      "epoch": 0.2236842105263158,
+      "grad_norm": 0.5597900748252869,
+      "learning_rate": 1.840182094512362e-05,
+      "loss": 0.5296148061752319,
+      "step": 408
+    },
+    {
+      "epoch": 0.22423245614035087,
+      "grad_norm": 0.5344291925430298,
+      "learning_rate": 1.839197052720237e-05,
+      "loss": 0.5298682451248169,
+      "step": 409
+    },
+    {
+      "epoch": 0.22478070175438597,
+      "grad_norm": 0.5179139375686646,
+      "learning_rate": 1.8382092499192484e-05,
+      "loss": 0.5287031531333923,
+      "step": 410
+    },
+    {
+      "epoch": 0.22532894736842105,
+      "grad_norm": 0.5144736170768738,
+      "learning_rate": 1.8372186893593266e-05,
+      "loss": 0.5490144491195679,
+      "step": 411
+    },
+    {
+      "epoch": 0.22587719298245615,
+      "grad_norm": 0.543007493019104,
+      "learning_rate": 1.8362253742994757e-05,
+      "loss": 0.5291949510574341,
+      "step": 412
+    },
+    {
+      "epoch": 0.22642543859649122,
+      "grad_norm": 0.5593482851982117,
+      "learning_rate": 1.835229308007762e-05,
+      "loss": 0.5276323556900024,
+      "step": 413
+    },
+    {
+      "epoch": 0.22697368421052633,
+      "grad_norm": 0.5211661458015442,
+      "learning_rate": 1.8342304937613034e-05,
+      "loss": 0.5092172026634216,
+      "step": 414
+    },
+    {
+      "epoch": 0.2275219298245614,
+      "grad_norm": 0.5620855093002319,
+      "learning_rate": 1.8332289348462592e-05,
+      "loss": 0.5306937098503113,
+      "step": 415
+    },
+    {
+      "epoch": 0.22807017543859648,
+      "grad_norm": 0.4984802007675171,
+      "learning_rate": 1.8322246345578182e-05,
+      "loss": 0.5322443246841431,
+      "step": 416
+    },
+    {
+      "epoch": 0.22861842105263158,
+      "grad_norm": 0.5562717914581299,
+      "learning_rate": 1.8312175962001894e-05,
+      "loss": 0.5333069562911987,
+      "step": 417
+    },
+    {
+      "epoch": 0.22916666666666666,
+      "grad_norm": 0.5334926247596741,
+      "learning_rate": 1.8302078230865893e-05,
+      "loss": 0.5284031629562378,
+      "step": 418
+    },
+    {
+      "epoch": 0.22971491228070176,
+      "grad_norm": 0.5168144702911377,
+      "learning_rate": 1.829195318539233e-05,
+      "loss": 0.5266040563583374,
+      "step": 419
+    },
+    {
+      "epoch": 0.23026315789473684,
+      "grad_norm": 0.5674131512641907,
+      "learning_rate": 1.8281800858893203e-05,
+      "loss": 0.5325175523757935,
+      "step": 420
+    },
+    {
+      "epoch": 0.23081140350877194,
+      "grad_norm": 0.5394864082336426,
+      "learning_rate": 1.8271621284770282e-05,
+      "loss": 0.5242091417312622,
+      "step": 421
+    },
+    {
+      "epoch": 0.23135964912280702,
+      "grad_norm": 0.5389179587364197,
+      "learning_rate": 1.8261414496514985e-05,
+      "loss": 0.5218663215637207,
+      "step": 422
+    },
+    {
+      "epoch": 0.23190789473684212,
+      "grad_norm": 0.5189980268478394,
+      "learning_rate": 1.8251180527708256e-05,
+      "loss": 0.5272551774978638,
+      "step": 423
+    },
+    {
+      "epoch": 0.2324561403508772,
+      "grad_norm": 0.5364722013473511,
+      "learning_rate": 1.8240919412020467e-05,
+      "loss": 0.5427874326705933,
+      "step": 424
+    },
+    {
+      "epoch": 0.23300438596491227,
+      "grad_norm": 0.5321043729782104,
+      "learning_rate": 1.8230631183211308e-05,
+      "loss": 0.5246886014938354,
+      "step": 425
+    },
+    {
+      "epoch": 0.23355263157894737,
+      "grad_norm": 0.5318389534950256,
+      "learning_rate": 1.8220315875129674e-05,
+      "loss": 0.5147177577018738,
+      "step": 426
+    },
+    {
+      "epoch": 0.23410087719298245,
+      "grad_norm": 0.5194035172462463,
+      "learning_rate": 1.8209973521713545e-05,
+      "loss": 0.5231367349624634,
+      "step": 427
+    },
+    {
+      "epoch": 0.23464912280701755,
+      "grad_norm": 0.5205478668212891,
+      "learning_rate": 1.8199604156989895e-05,
+      "loss": 0.5181112885475159,
+      "step": 428
+    },
+    {
+      "epoch": 0.23519736842105263,
+      "grad_norm": 0.5329371094703674,
+      "learning_rate": 1.818920781507455e-05,
+      "loss": 0.526331901550293,
+      "step": 429
+    },
+    {
+      "epoch": 0.23574561403508773,
+      "grad_norm": 0.5039587020874023,
+      "learning_rate": 1.817878453017211e-05,
+      "loss": 0.5373213887214661,
+      "step": 430
+    },
+    {
+      "epoch": 0.2362938596491228,
+      "grad_norm": 0.5267012715339661,
+      "learning_rate": 1.8168334336575802e-05,
+      "loss": 0.5282840728759766,
+      "step": 431
+    },
+    {
+      "epoch": 0.23684210526315788,
+      "grad_norm": 0.5210471749305725,
+      "learning_rate": 1.81578572686674e-05,
+      "loss": 0.5187563896179199,
+      "step": 432
+    },
+    {
+      "epoch": 0.23739035087719298,
+      "grad_norm": 0.5142402648925781,
+      "learning_rate": 1.8147353360917083e-05,
+      "loss": 0.5163940787315369,
+      "step": 433
+    },
+    {
+      "epoch": 0.23793859649122806,
+      "grad_norm": 0.5326563119888306,
+      "learning_rate": 1.813682264788334e-05,
+      "loss": 0.5173913836479187,
+      "step": 434
+    },
+    {
+      "epoch": 0.23848684210526316,
+      "grad_norm": 0.5267801880836487,
+      "learning_rate": 1.812626516421285e-05,
+      "loss": 0.5498681664466858,
+      "step": 435
+    },
+    {
+      "epoch": 0.23903508771929824,
+      "grad_norm": 0.52644944190979,
+      "learning_rate": 1.8115680944640384e-05,
+      "loss": 0.5258373022079468,
+      "step": 436
+    },
+    {
+      "epoch": 0.23958333333333334,
+      "grad_norm": 0.5588622093200684,
+      "learning_rate": 1.8105070023988646e-05,
+      "loss": 0.5311025977134705,
+      "step": 437
+    },
+    {
+      "epoch": 0.24013157894736842,
+      "grad_norm": 0.585602879524231,
+      "learning_rate": 1.809443243716821e-05,
+      "loss": 0.521843671798706,
+      "step": 438
+    },
+    {
+      "epoch": 0.24067982456140352,
+      "grad_norm": 0.5403777956962585,
+      "learning_rate": 1.808376821917738e-05,
+      "loss": 0.527265727519989,
+      "step": 439
+    },
+    {
+      "epoch": 0.2412280701754386,
+      "grad_norm": 0.5174070596694946,
+      "learning_rate": 1.8073077405102074e-05,
+      "loss": 0.5219606161117554,
+      "step": 440
+    },
+    {
+      "epoch": 0.24177631578947367,
+      "grad_norm": 0.5555534362792969,
+      "learning_rate": 1.806236003011571e-05,
+      "loss": 0.5224829912185669,
+      "step": 441
+    },
+    {
+      "epoch": 0.24232456140350878,
+      "grad_norm": 0.5561994910240173,
+      "learning_rate": 1.8051616129479102e-05,
+      "loss": 0.5335409641265869,
+      "step": 442
+    },
+    {
+      "epoch": 0.24287280701754385,
+      "grad_norm": 0.5374264717102051,
+      "learning_rate": 1.804084573854033e-05,
+      "loss": 0.5216151475906372,
+      "step": 443
+    },
+    {
+      "epoch": 0.24342105263157895,
+      "grad_norm": 0.5270918607711792,
+      "learning_rate": 1.803004889273463e-05,
+      "loss": 0.5328495502471924,
+      "step": 444
+    },
+    {
+      "epoch": 0.24396929824561403,
+      "grad_norm": 0.5628570914268494,
+      "learning_rate": 1.8019225627584275e-05,
+      "loss": 0.5127445459365845,
+      "step": 445
+    },
+    {
+      "epoch": 0.24451754385964913,
+      "grad_norm": 0.5093899965286255,
+      "learning_rate": 1.8008375978698452e-05,
+      "loss": 0.5272922515869141,
+      "step": 446
+    },
+    {
+      "epoch": 0.2450657894736842,
+      "grad_norm": 0.5224986672401428,
+      "learning_rate": 1.799749998177317e-05,
+      "loss": 0.5233873128890991,
+      "step": 447
+    },
+    {
+      "epoch": 0.24561403508771928,
+      "grad_norm": 0.6797884106636047,
+      "learning_rate": 1.7986597672591113e-05,
+      "loss": 0.5107884407043457,
+      "step": 448
+    },
+    {
+      "epoch": 0.2461622807017544,
+      "grad_norm": 0.5089065432548523,
+      "learning_rate": 1.797566908702153e-05,
+      "loss": 0.521450936794281,
+      "step": 449
+    },
+    {
+      "epoch": 0.24671052631578946,
+      "grad_norm": 0.5519887208938599,
+      "learning_rate": 1.7964714261020127e-05,
+      "loss": 0.5311073660850525,
+      "step": 450
+    },
+    {
+      "epoch": 0.24725877192982457,
+      "grad_norm": 0.5152913928031921,
+      "learning_rate": 1.7953733230628942e-05,
+      "loss": 0.5231930613517761,
+      "step": 451
+    },
+    {
+      "epoch": 0.24780701754385964,
+      "grad_norm": 0.533416211605072,
+      "learning_rate": 1.794272603197623e-05,
+      "loss": 0.5366246700286865,
+      "step": 452
+    },
+    {
+      "epoch": 0.24835526315789475,
+      "grad_norm": 0.5098294019699097,
+      "learning_rate": 1.7931692701276332e-05,
+      "loss": 0.5331299304962158,
+      "step": 453
+    },
+    {
+      "epoch": 0.24890350877192982,
+      "grad_norm": 0.552507758140564,
+      "learning_rate": 1.7920633274829577e-05,
+      "loss": 0.5277151465415955,
+      "step": 454
+    },
+    {
+      "epoch": 0.24945175438596492,
+      "grad_norm": 0.5532183647155762,
+      "learning_rate": 1.7909547789022134e-05,
+      "loss": 0.540197491645813,
+      "step": 455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5785213112831116,
+      "learning_rate": 1.789843628032593e-05,
+      "loss": 0.5247042179107666,
+      "step": 456
+    },
+    {
+      "epoch": 0.2505482456140351,
+      "grad_norm": 0.5596621632575989,
+      "learning_rate": 1.7887298785298498e-05,
+      "loss": 0.5302076935768127,
+      "step": 457
+    },
+    {
+      "epoch": 0.25109649122807015,
+      "grad_norm": 0.5846956372261047,
+      "learning_rate": 1.787613534058286e-05,
+      "loss": 0.5140069723129272,
+      "step": 458
+    },
+    {
+      "epoch": 0.25164473684210525,
+      "grad_norm": 0.5351735949516296,
+      "learning_rate": 1.7864945982907426e-05,
+      "loss": 0.5310229063034058,
+      "step": 459
+    },
+    {
+      "epoch": 0.25219298245614036,
+      "grad_norm": 0.5385695099830627,
+      "learning_rate": 1.7853730749085856e-05,
+      "loss": 0.5263150930404663,
+      "step": 460
+    },
+    {
+      "epoch": 0.25274122807017546,
+      "grad_norm": 0.5790038704872131,
+      "learning_rate": 1.784248967601695e-05,
+      "loss": 0.5242854952812195,
+      "step": 461
+    },
+    {
+      "epoch": 0.2532894736842105,
+      "grad_norm": 0.5039320588111877,
+      "learning_rate": 1.7831222800684504e-05,
+      "loss": 0.5093971490859985,
+      "step": 462
+    },
+    {
+      "epoch": 0.2538377192982456,
+      "grad_norm": 0.5355396866798401,
+      "learning_rate": 1.7819930160157228e-05,
+      "loss": 0.5264978408813477,
+      "step": 463
+    },
+    {
+      "epoch": 0.2543859649122807,
+      "grad_norm": 0.5939465761184692,
+      "learning_rate": 1.7808611791588584e-05,
+      "loss": 0.5278671383857727,
+      "step": 464
+    },
+    {
+      "epoch": 0.25493421052631576,
+      "grad_norm": 0.5435965061187744,
+      "learning_rate": 1.779726773221669e-05,
+      "loss": 0.5312660932540894,
+      "step": 465
+    },
+    {
+      "epoch": 0.25548245614035087,
+      "grad_norm": 0.5324194431304932,
+      "learning_rate": 1.7785898019364176e-05,
+      "loss": 0.5131596922874451,
+      "step": 466
+    },
+    {
+      "epoch": 0.25603070175438597,
+      "grad_norm": 0.5465553402900696,
+      "learning_rate": 1.7774502690438097e-05,
+      "loss": 0.5083457231521606,
+      "step": 467
+    },
+    {
+      "epoch": 0.2565789473684211,
+      "grad_norm": 0.5563082695007324,
+      "learning_rate": 1.776308178292976e-05,
+      "loss": 0.5215815901756287,
+      "step": 468
+    },
+    {
+      "epoch": 0.2571271929824561,
+      "grad_norm": 0.5333871245384216,
+      "learning_rate": 1.7751635334414642e-05,
+      "loss": 0.5281736254692078,
+      "step": 469
+    },
+    {
+      "epoch": 0.2576754385964912,
+      "grad_norm": 0.5244386196136475,
+      "learning_rate": 1.7740163382552248e-05,
+      "loss": 0.5255697965621948,
+      "step": 470
+    },
+    {
+      "epoch": 0.2582236842105263,
+      "grad_norm": 0.5395222306251526,
+      "learning_rate": 1.7728665965085995e-05,
+      "loss": 0.5342221856117249,
+      "step": 471
+    },
+    {
+      "epoch": 0.25877192982456143,
+      "grad_norm": 0.510644793510437,
+      "learning_rate": 1.7717143119843078e-05,
+      "loss": 0.5168928503990173,
+      "step": 472
+    },
+    {
+      "epoch": 0.2593201754385965,
+      "grad_norm": 0.6781021952629089,
+      "learning_rate": 1.7705594884734342e-05,
+      "loss": 0.5226818323135376,
+      "step": 473
+    },
+    {
+      "epoch": 0.2598684210526316,
+      "grad_norm": 0.5210604071617126,
+      "learning_rate": 1.769402129775419e-05,
+      "loss": 0.5066401362419128,
+      "step": 474
+    },
+    {
+      "epoch": 0.2604166666666667,
+      "grad_norm": 0.5117266774177551,
+      "learning_rate": 1.768242239698041e-05,
+      "loss": 0.514409065246582,
+      "step": 475
+    },
+    {
+      "epoch": 0.26096491228070173,
+      "grad_norm": 0.9269617795944214,
+      "learning_rate": 1.7670798220574093e-05,
+      "loss": 0.5230674147605896,
+      "step": 476
+    },
+    {
+      "epoch": 0.26151315789473684,
+      "grad_norm": 0.620804488658905,
+      "learning_rate": 1.7659148806779475e-05,
+      "loss": 0.5233884453773499,
+      "step": 477
+    },
+    {
+      "epoch": 0.26206140350877194,
+      "grad_norm": 0.5338650941848755,
+      "learning_rate": 1.764747419392383e-05,
+      "loss": 0.5138469934463501,
+      "step": 478
+    },
+    {
+      "epoch": 0.26260964912280704,
+      "grad_norm": 0.548736035823822,
+      "learning_rate": 1.7635774420417336e-05,
+      "loss": 0.5272300243377686,
+      "step": 479
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.5450613498687744,
+      "learning_rate": 1.7624049524752954e-05,
+      "loss": 0.5213850736618042,
+      "step": 480
+    },
+    {
+      "epoch": 0.2637061403508772,
+      "grad_norm": 0.5711430907249451,
+      "learning_rate": 1.76122995455063e-05,
+      "loss": 0.5210334062576294,
+      "step": 481
+    },
+    {
+      "epoch": 0.2642543859649123,
+      "grad_norm": 0.5337658524513245,
+      "learning_rate": 1.7600524521335512e-05,
+      "loss": 0.5187428593635559,
+      "step": 482
+    },
+    {
+      "epoch": 0.26480263157894735,
+      "grad_norm": 0.5508308410644531,
+      "learning_rate": 1.7588724490981125e-05,
+      "loss": 0.5229613780975342,
+      "step": 483
+    },
+    {
+      "epoch": 0.26535087719298245,
+      "grad_norm": 0.515425443649292,
+      "learning_rate": 1.7576899493265952e-05,
+      "loss": 0.5202820897102356,
+      "step": 484
+    },
+    {
+      "epoch": 0.26589912280701755,
+      "grad_norm": 0.5129563808441162,
+      "learning_rate": 1.756504956709495e-05,
+      "loss": 0.5205594897270203,
+      "step": 485
+    },
+    {
+      "epoch": 0.26644736842105265,
+      "grad_norm": 0.5160803198814392,
+      "learning_rate": 1.755317475145509e-05,
+      "loss": 0.5141991376876831,
+      "step": 486
+    },
+    {
+      "epoch": 0.2669956140350877,
+      "grad_norm": 0.5115503072738647,
+      "learning_rate": 1.754127508541523e-05,
+      "loss": 0.5174502730369568,
+      "step": 487
+    },
+    {
+      "epoch": 0.2675438596491228,
+      "grad_norm": 0.5334891080856323,
+      "learning_rate": 1.7529350608125985e-05,
+      "loss": 0.5169588327407837,
+      "step": 488
+    },
+    {
+      "epoch": 0.2680921052631579,
+      "grad_norm": 0.5355765223503113,
+      "learning_rate": 1.7517401358819608e-05,
+      "loss": 0.5175495147705078,
+      "step": 489
+    },
+    {
+      "epoch": 0.26864035087719296,
+      "grad_norm": 0.48000240325927734,
+      "learning_rate": 1.7505427376809848e-05,
+      "loss": 0.5160954594612122,
+      "step": 490
+    },
+    {
+      "epoch": 0.26918859649122806,
+      "grad_norm": 0.6343967914581299,
+      "learning_rate": 1.7493428701491827e-05,
+      "loss": 0.510797381401062,
+      "step": 491
+    },
+    {
+      "epoch": 0.26973684210526316,
+      "grad_norm": 0.8650885820388794,
+      "learning_rate": 1.748140537234191e-05,
+      "loss": 0.5214411616325378,
+      "step": 492
+    },
+    {
+      "epoch": 0.27028508771929827,
+      "grad_norm": 0.525759220123291,
+      "learning_rate": 1.7469357428917574e-05,
+      "loss": 0.5195897817611694,
+      "step": 493
+    },
+    {
+      "epoch": 0.2708333333333333,
+      "grad_norm": 0.4793441891670227,
+      "learning_rate": 1.745728491085728e-05,
+      "loss": 0.5165450572967529,
+      "step": 494
+    },
+    {
+      "epoch": 0.2713815789473684,
+      "grad_norm": 0.6296154856681824,
+      "learning_rate": 1.7445187857880334e-05,
+      "loss": 0.5223666429519653,
+      "step": 495
+    },
+    {
+      "epoch": 0.2719298245614035,
+      "grad_norm": 0.49726709723472595,
+      "learning_rate": 1.743306630978678e-05,
+      "loss": 0.522741436958313,
+      "step": 496
+    },
+    {
+      "epoch": 0.27247807017543857,
+      "grad_norm": 0.5264485478401184,
+      "learning_rate": 1.742092030645723e-05,
+      "loss": 0.5052363276481628,
+      "step": 497
+    },
+    {
+      "epoch": 0.2730263157894737,
+      "grad_norm": 0.5733623504638672,
+      "learning_rate": 1.740874988785277e-05,
+      "loss": 0.5072427988052368,
+      "step": 498
+    },
+    {
+      "epoch": 0.2735745614035088,
+      "grad_norm": 0.5145069360733032,
+      "learning_rate": 1.739655509401482e-05,
+      "loss": 0.5078324675559998,
+      "step": 499
+    },
+    {
+      "epoch": 0.2741228070175439,
+      "grad_norm": 0.5487716197967529,
+      "learning_rate": 1.7384335965064974e-05,
+      "loss": 0.5313689112663269,
+      "step": 500
+    },
+    {
+      "epoch": 0.2746710526315789,
+      "grad_norm": 0.5024135708808899,
+      "learning_rate": 1.737209254120491e-05,
+      "loss": 0.5236111879348755,
+      "step": 501
+    },
+    {
+      "epoch": 0.27521929824561403,
+      "grad_norm": 0.5059981942176819,
+      "learning_rate": 1.735982486271622e-05,
+      "loss": 0.5159262418746948,
+      "step": 502
+    },
+    {
+      "epoch": 0.27576754385964913,
+      "grad_norm": 0.5166164636611938,
+      "learning_rate": 1.7347532969960322e-05,
+      "loss": 0.5162345767021179,
+      "step": 503
+    },
+    {
+      "epoch": 0.27631578947368424,
+      "grad_norm": 0.5122005343437195,
+      "learning_rate": 1.733521690337827e-05,
+      "loss": 0.49863386154174805,
+      "step": 504
+    },
+    {
+      "epoch": 0.2768640350877193,
+      "grad_norm": 0.5083131790161133,
+      "learning_rate": 1.732287670349067e-05,
+      "loss": 0.5120081901550293,
+      "step": 505
+    },
+    {
+      "epoch": 0.2774122807017544,
+      "grad_norm": 0.5615320801734924,
+      "learning_rate": 1.731051241089752e-05,
+      "loss": 0.5304388403892517,
+      "step": 506
+    },
+    {
+      "epoch": 0.2779605263157895,
+      "grad_norm": 0.5108237266540527,
+      "learning_rate": 1.7298124066278094e-05,
+      "loss": 0.5019451975822449,
+      "step": 507
+    },
+    {
+      "epoch": 0.27850877192982454,
+      "grad_norm": 0.5400595664978027,
+      "learning_rate": 1.7285711710390788e-05,
+      "loss": 0.5011708736419678,
+      "step": 508
+    },
+    {
+      "epoch": 0.27905701754385964,
+      "grad_norm": 0.5205352902412415,
+      "learning_rate": 1.7273275384073e-05,
+      "loss": 0.5174415111541748,
+      "step": 509
+    },
+    {
+      "epoch": 0.27960526315789475,
+      "grad_norm": 0.5404713153839111,
+      "learning_rate": 1.726081512824099e-05,
+      "loss": 0.5089818239212036,
+      "step": 510
+    },
+    {
+      "epoch": 0.28015350877192985,
+      "grad_norm": 0.5409142374992371,
+      "learning_rate": 1.724833098388976e-05,
+      "loss": 0.5215460062026978,
+      "step": 511
+    },
+    {
+      "epoch": 0.2807017543859649,
+      "grad_norm": 0.5232543349266052,
+      "learning_rate": 1.723582299209289e-05,
+      "loss": 0.5150152444839478,
+      "step": 512
+    },
+    {
+      "epoch": 0.28125,
+      "grad_norm": 0.5587122440338135,
+      "learning_rate": 1.722329119400243e-05,
+      "loss": 0.5138131380081177,
+      "step": 513
+    },
+    {
+      "epoch": 0.2817982456140351,
+      "grad_norm": 0.524645209312439,
+      "learning_rate": 1.7210735630848747e-05,
+      "loss": 0.523358941078186,
+      "step": 514
+    },
+    {
+      "epoch": 0.28234649122807015,
+      "grad_norm": 0.5060891509056091,
+      "learning_rate": 1.71981563439404e-05,
+      "loss": 0.518589973449707,
+      "step": 515
+    },
+    {
+      "epoch": 0.28289473684210525,
+      "grad_norm": 0.5494999885559082,
+      "learning_rate": 1.7185553374664006e-05,
+      "loss": 0.5273317098617554,
+      "step": 516
+    },
+    {
+      "epoch": 0.28344298245614036,
+      "grad_norm": 0.5179638266563416,
+      "learning_rate": 1.7172926764484085e-05,
+      "loss": 0.5220479965209961,
+      "step": 517
+    },
+    {
+      "epoch": 0.28399122807017546,
+      "grad_norm": 0.497627317905426,
+      "learning_rate": 1.7160276554942953e-05,
+      "loss": 0.517738401889801,
+      "step": 518
+    },
+    {
+      "epoch": 0.2845394736842105,
+      "grad_norm": 0.4714062511920929,
+      "learning_rate": 1.7147602787660563e-05,
+      "loss": 0.5142806768417358,
+      "step": 519
+    },
+    {
+      "epoch": 0.2850877192982456,
+      "grad_norm": 0.5007819533348083,
+      "learning_rate": 1.7134905504334365e-05,
+      "loss": 0.51795893907547,
+      "step": 520
+    },
+    {
+      "epoch": 0.2856359649122807,
+      "grad_norm": 0.49607014656066895,
+      "learning_rate": 1.7122184746739188e-05,
+      "loss": 0.5078467726707458,
+      "step": 521
+    },
+    {
+      "epoch": 0.28618421052631576,
+      "grad_norm": 0.49878785014152527,
+      "learning_rate": 1.7109440556727094e-05,
+      "loss": 0.5141764283180237,
+      "step": 522
+    },
+    {
+      "epoch": 0.28673245614035087,
+      "grad_norm": 0.5011985301971436,
+      "learning_rate": 1.7096672976227236e-05,
+      "loss": 0.5168810486793518,
+      "step": 523
+    },
+    {
+      "epoch": 0.28728070175438597,
+      "grad_norm": 0.5092190504074097,
+      "learning_rate": 1.708388204724572e-05,
+      "loss": 0.5183935761451721,
+      "step": 524
+    },
+    {
+      "epoch": 0.2878289473684211,
+      "grad_norm": 0.5125263929367065,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.509937047958374,
+      "step": 525
+    },
+    {
+      "epoch": 0.2883771929824561,
+      "grad_norm": 0.5333483219146729,
+      "learning_rate": 1.705823031224611e-05,
+      "loss": 0.5176279544830322,
+      "step": 526
+    },
+    {
+      "epoch": 0.2889254385964912,
+      "grad_norm": 0.5202133655548096,
+      "learning_rate": 1.7045369590623768e-05,
+      "loss": 0.5100720524787903,
+      "step": 527
+    },
+    {
+      "epoch": 0.2894736842105263,
+      "grad_norm": 0.5099378228187561,
+      "learning_rate": 1.7032485689310997e-05,
+      "loss": 0.5064803957939148,
+      "step": 528
+    },
+    {
+      "epoch": 0.29002192982456143,
+      "grad_norm": 0.5217911005020142,
+      "learning_rate": 1.7019578650696614e-05,
+      "loss": 0.514005184173584,
+      "step": 529
+    },
+    {
+      "epoch": 0.2905701754385965,
+      "grad_norm": 0.5043156147003174,
+      "learning_rate": 1.7006648517245547e-05,
+      "loss": 0.52207350730896,
+      "step": 530
+    },
+    {
+      "epoch": 0.2911184210526316,
+      "grad_norm": 0.5025941133499146,
+      "learning_rate": 1.6993695331498723e-05,
+      "loss": 0.514879584312439,
+      "step": 531
+    },
+    {
+      "epoch": 0.2916666666666667,
+      "grad_norm": 0.5254765748977661,
+      "learning_rate": 1.6980719136072892e-05,
+      "loss": 0.5103027820587158,
+      "step": 532
+    },
+    {
+      "epoch": 0.29221491228070173,
+      "grad_norm": 0.5231887698173523,
+      "learning_rate": 1.696771997366053e-05,
+      "loss": 0.5095311403274536,
+      "step": 533
+    },
+    {
+      "epoch": 0.29276315789473684,
+      "grad_norm": 0.49643293023109436,
+      "learning_rate": 1.6954697887029657e-05,
+      "loss": 0.5114062428474426,
+      "step": 534
+    },
+    {
+      "epoch": 0.29331140350877194,
+      "grad_norm": 0.5372974276542664,
+      "learning_rate": 1.6941652919023727e-05,
+      "loss": 0.5112413167953491,
+      "step": 535
+    },
+    {
+      "epoch": 0.29385964912280704,
+      "grad_norm": 0.5067369341850281,
+      "learning_rate": 1.6928585112561468e-05,
+      "loss": 0.5218123197555542,
+      "step": 536
+    },
+    {
+      "epoch": 0.2944078947368421,
+      "grad_norm": 0.5153988599777222,
+      "learning_rate": 1.6915494510636748e-05,
+      "loss": 0.510006308555603,
+      "step": 537
+    },
+    {
+      "epoch": 0.2949561403508772,
+      "grad_norm": 0.5265623927116394,
+      "learning_rate": 1.690238115631844e-05,
+      "loss": 0.5182325839996338,
+      "step": 538
+    },
+    {
+      "epoch": 0.2955043859649123,
+      "grad_norm": 0.5706140398979187,
+      "learning_rate": 1.6889245092750274e-05,
+      "loss": 0.5318684577941895,
+      "step": 539
+    },
+    {
+      "epoch": 0.29605263157894735,
+      "grad_norm": 0.48171648383140564,
+      "learning_rate": 1.687608636315068e-05,
+      "loss": 0.5149196982383728,
+      "step": 540
+    },
+    {
+      "epoch": 0.29660087719298245,
+      "grad_norm": 0.5281539559364319,
+      "learning_rate": 1.686290501081268e-05,
+      "loss": 0.5077455043792725,
+      "step": 541
+    },
+    {
+      "epoch": 0.29714912280701755,
+      "grad_norm": 0.5157037973403931,
+      "learning_rate": 1.6849701079103714e-05,
+      "loss": 0.49477753043174744,
+      "step": 542
+    },
+    {
+      "epoch": 0.29769736842105265,
+      "grad_norm": 0.5387939214706421,
+      "learning_rate": 1.6836474611465515e-05,
+      "loss": 0.5263862609863281,
+      "step": 543
+    },
+    {
+      "epoch": 0.2982456140350877,
+      "grad_norm": 0.5387179255485535,
+      "learning_rate": 1.682322565141395e-05,
+      "loss": 0.5146876573562622,
+      "step": 544
+    },
+    {
+      "epoch": 0.2987938596491228,
+      "grad_norm": 0.5182175040245056,
+      "learning_rate": 1.6809954242538907e-05,
+      "loss": 0.5151358246803284,
+      "step": 545
+    },
+    {
+      "epoch": 0.2993421052631579,
+      "grad_norm": 0.5380533337593079,
+      "learning_rate": 1.6796660428504114e-05,
+      "loss": 0.5249975919723511,
+      "step": 546
+    },
+    {
+      "epoch": 0.29989035087719296,
+      "grad_norm": 0.5614789724349976,
+      "learning_rate": 1.6783344253047027e-05,
+      "loss": 0.5180106163024902,
+      "step": 547
+    },
+    {
+      "epoch": 0.30043859649122806,
+      "grad_norm": 0.5239329934120178,
+      "learning_rate": 1.6770005759978656e-05,
+      "loss": 0.5265669822692871,
+      "step": 548
+    },
+    {
+      "epoch": 0.30098684210526316,
+      "grad_norm": 0.540311872959137,
+      "learning_rate": 1.6756644993183453e-05,
+      "loss": 0.5158952474594116,
+      "step": 549
+    },
+    {
+      "epoch": 0.30153508771929827,
+      "grad_norm": 0.5373164415359497,
+      "learning_rate": 1.6743261996619145e-05,
+      "loss": 0.4985647201538086,
+      "step": 550
+    },
+    {
+      "epoch": 0.3020833333333333,
+      "grad_norm": 0.5068860650062561,
+      "learning_rate": 1.6729856814316595e-05,
+      "loss": 0.5132797360420227,
+      "step": 551
+    },
+    {
+      "epoch": 0.3026315789473684,
+      "grad_norm": 0.5771002173423767,
+      "learning_rate": 1.671642949037966e-05,
+      "loss": 0.5127143859863281,
+      "step": 552
+    },
+    {
+      "epoch": 0.3031798245614035,
+      "grad_norm": 0.5291208624839783,
+      "learning_rate": 1.6702980068985048e-05,
+      "loss": 0.5070824027061462,
+      "step": 553
+    },
+    {
+      "epoch": 0.30372807017543857,
+      "grad_norm": 0.5508853793144226,
+      "learning_rate": 1.668950859438216e-05,
+      "loss": 0.5169316530227661,
+      "step": 554
+    },
+    {
+      "epoch": 0.3042763157894737,
+      "grad_norm": 0.5345003604888916,
+      "learning_rate": 1.6676015110892973e-05,
+      "loss": 0.5074903964996338,
+      "step": 555
+    },
+    {
+      "epoch": 0.3048245614035088,
+      "grad_norm": 0.5480985641479492,
+      "learning_rate": 1.6662499662911843e-05,
+      "loss": 0.5174945592880249,
+      "step": 556
+    },
+    {
+      "epoch": 0.3053728070175439,
+      "grad_norm": 0.531358003616333,
+      "learning_rate": 1.6648962294905416e-05,
+      "loss": 0.5207014679908752,
+      "step": 557
+    },
+    {
+      "epoch": 0.3059210526315789,
+      "grad_norm": 0.49597254395484924,
+      "learning_rate": 1.6635403051412452e-05,
+      "loss": 0.5044191479682922,
+      "step": 558
+    },
+    {
+      "epoch": 0.30646929824561403,
+      "grad_norm": 0.5143059492111206,
+      "learning_rate": 1.6621821977043673e-05,
+      "loss": 0.5156230926513672,
+      "step": 559
+    },
+    {
+      "epoch": 0.30701754385964913,
+      "grad_norm": 0.5195500254631042,
+      "learning_rate": 1.6608219116481633e-05,
+      "loss": 0.5130860209465027,
+      "step": 560
+    },
+    {
+      "epoch": 0.30756578947368424,
+      "grad_norm": 0.5337082147598267,
+      "learning_rate": 1.6594594514480564e-05,
+      "loss": 0.49173516035079956,
+      "step": 561
+    },
+    {
+      "epoch": 0.3081140350877193,
+      "grad_norm": 0.5399724841117859,
+      "learning_rate": 1.6580948215866232e-05,
+      "loss": 0.5126796960830688,
+      "step": 562
+    },
+    {
+      "epoch": 0.3086622807017544,
+      "grad_norm": 0.5642949938774109,
+      "learning_rate": 1.656728026553577e-05,
+      "loss": 0.507789671421051,
+      "step": 563
+    },
+    {
+      "epoch": 0.3092105263157895,
+      "grad_norm": 0.5704178214073181,
+      "learning_rate": 1.655359070845757e-05,
+      "loss": 0.5131293535232544,
+      "step": 564
+    },
+    {
+      "epoch": 0.30975877192982454,
+      "grad_norm": 0.5212867856025696,
+      "learning_rate": 1.6539879589671093e-05,
+      "loss": 0.5129238367080688,
+      "step": 565
+    },
+    {
+      "epoch": 0.31030701754385964,
+      "grad_norm": 0.5646382570266724,
+      "learning_rate": 1.652614695428675e-05,
+      "loss": 0.49918338656425476,
+      "step": 566
+    },
+    {
+      "epoch": 0.31085526315789475,
+      "grad_norm": 0.5082690715789795,
+      "learning_rate": 1.6512392847485733e-05,
+      "loss": 0.5041823983192444,
+      "step": 567
+    },
+    {
+      "epoch": 0.31140350877192985,
+      "grad_norm": 0.5300014615058899,
+      "learning_rate": 1.6498617314519886e-05,
+      "loss": 0.5186752676963806,
+      "step": 568
+    },
+    {
+      "epoch": 0.3119517543859649,
+      "grad_norm": 0.554815948009491,
+      "learning_rate": 1.6484820400711545e-05,
+      "loss": 0.5164564251899719,
+      "step": 569
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.5678638815879822,
+      "learning_rate": 1.647100215145338e-05,
+      "loss": 0.500181257724762,
+      "step": 570
+    },
+    {
+      "epoch": 0.3130482456140351,
+      "grad_norm": 0.5426273941993713,
+      "learning_rate": 1.645716261220827e-05,
+      "loss": 0.5085293054580688,
+      "step": 571
+    },
+    {
+      "epoch": 0.31359649122807015,
+      "grad_norm": 0.5601498484611511,
+      "learning_rate": 1.644330182850913e-05,
+      "loss": 0.5364582538604736,
+      "step": 572
+    },
+    {
+      "epoch": 0.31414473684210525,
+      "grad_norm": 0.554161012172699,
+      "learning_rate": 1.6429419845958778e-05,
+      "loss": 0.5135458707809448,
+      "step": 573
+    },
+    {
+      "epoch": 0.31469298245614036,
+      "grad_norm": 0.5089998245239258,
+      "learning_rate": 1.6415516710229767e-05,
+      "loss": 0.5096344351768494,
+      "step": 574
+    },
+    {
+      "epoch": 0.31524122807017546,
+      "grad_norm": 0.5251020789146423,
+      "learning_rate": 1.6401592467064254e-05,
+      "loss": 0.4982328414916992,
+      "step": 575
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.5574110150337219,
+      "learning_rate": 1.6387647162273837e-05,
+      "loss": 0.49430570006370544,
+      "step": 576
+    },
+    {
+      "epoch": 0.3163377192982456,
+      "grad_norm": 0.5299574732780457,
+      "learning_rate": 1.637368084173941e-05,
+      "loss": 0.5284730792045593,
+      "step": 577
+    },
+    {
+      "epoch": 0.3168859649122807,
+      "grad_norm": 0.4945147931575775,
+      "learning_rate": 1.6359693551411008e-05,
+      "loss": 0.5055221319198608,
+      "step": 578
+    },
+    {
+      "epoch": 0.31743421052631576,
+      "grad_norm": 1.933510661125183,
+      "learning_rate": 1.6345685337307656e-05,
+      "loss": 0.5102857947349548,
+      "step": 579
+    },
+    {
+      "epoch": 0.31798245614035087,
+      "grad_norm": 0.5664254426956177,
+      "learning_rate": 1.633165624551723e-05,
+      "loss": 0.5002626180648804,
+      "step": 580
+    },
+    {
+      "epoch": 0.31853070175438597,
+      "grad_norm": 0.5354625582695007,
+      "learning_rate": 1.6317606322196284e-05,
+      "loss": 0.503082275390625,
+      "step": 581
+    },
+    {
+      "epoch": 0.3190789473684211,
+      "grad_norm": 0.5543798804283142,
+      "learning_rate": 1.6303535613569908e-05,
+      "loss": 0.5081637501716614,
+      "step": 582
+    },
+    {
+      "epoch": 0.3196271929824561,
+      "grad_norm": 0.5393402576446533,
+      "learning_rate": 1.6289444165931587e-05,
+      "loss": 0.4948890209197998,
+      "step": 583
+    },
+    {
+      "epoch": 0.3201754385964912,
+      "grad_norm": 0.5106928944587708,
+      "learning_rate": 1.627533202564303e-05,
+      "loss": 0.5148372650146484,
+      "step": 584
+    },
+    {
+      "epoch": 0.3207236842105263,
+      "grad_norm": 0.5648923516273499,
+      "learning_rate": 1.6261199239134023e-05,
+      "loss": 0.5127902626991272,
+      "step": 585
+    },
+    {
+      "epoch": 0.32127192982456143,
+      "grad_norm": 0.5218735933303833,
+      "learning_rate": 1.6247045852902298e-05,
+      "loss": 0.5089883804321289,
+      "step": 586
+    },
+    {
+      "epoch": 0.3218201754385965,
+      "grad_norm": 0.5288784503936768,
+      "learning_rate": 1.623287191351334e-05,
+      "loss": 0.5053139925003052,
+      "step": 587
+    },
+    {
+      "epoch": 0.3223684210526316,
+      "grad_norm": 0.5042205452919006,
+      "learning_rate": 1.6218677467600264e-05,
+      "loss": 0.49913498759269714,
+      "step": 588
+    },
+    {
+      "epoch": 0.3229166666666667,
+      "grad_norm": 0.5079101324081421,
+      "learning_rate": 1.6204462561863652e-05,
+      "loss": 0.5198526978492737,
+      "step": 589
+    },
+    {
+      "epoch": 0.32346491228070173,
+      "grad_norm": 0.5735262632369995,
+      "learning_rate": 1.6190227243071402e-05,
+      "loss": 0.5066660642623901,
+      "step": 590
+    },
+    {
+      "epoch": 0.32401315789473684,
+      "grad_norm": 0.5124266743659973,
+      "learning_rate": 1.617597155805857e-05,
+      "loss": 0.4869317412376404,
+      "step": 591
+    },
+    {
+      "epoch": 0.32456140350877194,
+      "grad_norm": 0.5241952538490295,
+      "learning_rate": 1.6161695553727218e-05,
+      "loss": 0.500100314617157,
+      "step": 592
+    },
+    {
+      "epoch": 0.32510964912280704,
+      "grad_norm": 0.5251774787902832,
+      "learning_rate": 1.6147399277046264e-05,
+      "loss": 0.5080313086509705,
+      "step": 593
+    },
+    {
+      "epoch": 0.3256578947368421,
+      "grad_norm": 0.4947800934314728,
+      "learning_rate": 1.6133082775051312e-05,
+      "loss": 0.5108361840248108,
+      "step": 594
+    },
+    {
+      "epoch": 0.3262061403508772,
+      "grad_norm": 0.4975525438785553,
+      "learning_rate": 1.6118746094844523e-05,
+      "loss": 0.5160967111587524,
+      "step": 595
+    },
+    {
+      "epoch": 0.3267543859649123,
+      "grad_norm": 0.5103340744972229,
+      "learning_rate": 1.6104389283594435e-05,
+      "loss": 0.5133464336395264,
+      "step": 596
+    },
+    {
+      "epoch": 0.32730263157894735,
+      "grad_norm": 0.5223587155342102,
+      "learning_rate": 1.6090012388535825e-05,
+      "loss": 0.5166263580322266,
+      "step": 597
+    },
+    {
+      "epoch": 0.32785087719298245,
+      "grad_norm": 0.5182468295097351,
+      "learning_rate": 1.607561545696954e-05,
+      "loss": 0.5085626840591431,
+      "step": 598
+    },
+    {
+      "epoch": 0.32839912280701755,
+      "grad_norm": 0.5165911912918091,
+      "learning_rate": 1.6061198536262355e-05,
+      "loss": 0.5134096741676331,
+      "step": 599
+    },
+    {
+      "epoch": 0.32894736842105265,
+      "grad_norm": 0.5011234283447266,
+      "learning_rate": 1.604676167384681e-05,
+      "loss": 0.5025832056999207,
+      "step": 600
+    },
+    {
+      "epoch": 0.3294956140350877,
+      "grad_norm": 0.5092823505401611,
+      "learning_rate": 1.6032304917221045e-05,
+      "loss": 0.5118825435638428,
+      "step": 601
+    },
+    {
+      "epoch": 0.3300438596491228,
+      "grad_norm": 0.5826072692871094,
+      "learning_rate": 1.6017828313948667e-05,
+      "loss": 0.503429651260376,
+      "step": 602
+    },
+    {
+      "epoch": 0.3305921052631579,
+      "grad_norm": 0.5502949953079224,
+      "learning_rate": 1.600333191165857e-05,
+      "loss": 0.5013739466667175,
+      "step": 603
+    },
+    {
+      "epoch": 0.33114035087719296,
+      "grad_norm": 0.4851742386817932,
+      "learning_rate": 1.5988815758044794e-05,
+      "loss": 0.498490571975708,
+      "step": 604
+    },
+    {
+      "epoch": 0.33168859649122806,
+      "grad_norm": 0.5124604105949402,
+      "learning_rate": 1.5974279900866353e-05,
+      "loss": 0.5156079530715942,
+      "step": 605
+    },
+    {
+      "epoch": 0.33223684210526316,
+      "grad_norm": 0.5280570387840271,
+      "learning_rate": 1.59597243879471e-05,
+      "loss": 0.5016382932662964,
+      "step": 606
+    },
+    {
+      "epoch": 0.33278508771929827,
+      "grad_norm": 0.5246084928512573,
+      "learning_rate": 1.594514926717554e-05,
+      "loss": 0.5041351318359375,
+      "step": 607
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.5151121020317078,
+      "learning_rate": 1.593055458650471e-05,
+      "loss": 0.4976646900177002,
+      "step": 608
+    },
+    {
+      "epoch": 0.3338815789473684,
+      "grad_norm": 0.5183243751525879,
+      "learning_rate": 1.5915940393951975e-05,
+      "loss": 0.4976310133934021,
+      "step": 609
+    },
+    {
+      "epoch": 0.3344298245614035,
+      "grad_norm": 0.5183877944946289,
+      "learning_rate": 1.5901306737598922e-05,
+      "loss": 0.49520882964134216,
+      "step": 610
+    },
+    {
+      "epoch": 0.33497807017543857,
+      "grad_norm": 0.5211166739463806,
+      "learning_rate": 1.5886653665591158e-05,
+      "loss": 0.5143977403640747,
+      "step": 611
+    },
+    {
+      "epoch": 0.3355263157894737,
+      "grad_norm": 0.48644551634788513,
+      "learning_rate": 1.5871981226138174e-05,
+      "loss": 0.49930331110954285,
+      "step": 612
+    },
+    {
+      "epoch": 0.3360745614035088,
+      "grad_norm": 0.5271084308624268,
+      "learning_rate": 1.5857289467513172e-05,
+      "loss": 0.48993366956710815,
+      "step": 613
+    },
+    {
+      "epoch": 0.3366228070175439,
+      "grad_norm": 0.5187105536460876,
+      "learning_rate": 1.584257843805293e-05,
+      "loss": 0.5129623413085938,
+      "step": 614
+    },
+    {
+      "epoch": 0.3371710526315789,
+      "grad_norm": 0.5101479887962341,
+      "learning_rate": 1.582784818615762e-05,
+      "loss": 0.5065633654594421,
+      "step": 615
+    },
+    {
+      "epoch": 0.33771929824561403,
+      "grad_norm": 0.4912799298763275,
+      "learning_rate": 1.581309876029066e-05,
+      "loss": 0.48601680994033813,
+      "step": 616
+    },
+    {
+      "epoch": 0.33826754385964913,
+      "grad_norm": 0.4957771301269531,
+      "learning_rate": 1.5798330208978545e-05,
+      "loss": 0.4986298680305481,
+      "step": 617
+    },
+    {
+      "epoch": 0.33881578947368424,
+      "grad_norm": 0.5334628224372864,
+      "learning_rate": 1.5783542580810702e-05,
+      "loss": 0.5085287690162659,
+      "step": 618
+    },
+    {
+      "epoch": 0.3393640350877193,
+      "grad_norm": 0.5106218457221985,
+      "learning_rate": 1.5768735924439317e-05,
+      "loss": 0.4843800961971283,
+      "step": 619
+    },
+    {
+      "epoch": 0.3399122807017544,
+      "grad_norm": 0.5028548836708069,
+      "learning_rate": 1.5753910288579188e-05,
+      "loss": 0.4937993884086609,
+      "step": 620
+    },
+    {
+      "epoch": 0.3404605263157895,
+      "grad_norm": 0.521271288394928,
+      "learning_rate": 1.573906572200754e-05,
+      "loss": 0.5151383280754089,
+      "step": 621
+    },
+    {
+      "epoch": 0.34100877192982454,
+      "grad_norm": 0.5260574817657471,
+      "learning_rate": 1.5724202273563896e-05,
+      "loss": 0.5101398229598999,
+      "step": 622
+    },
+    {
+      "epoch": 0.34155701754385964,
+      "grad_norm": 0.5086724162101746,
+      "learning_rate": 1.57093199921499e-05,
+      "loss": 0.5031939744949341,
+      "step": 623
+    },
+    {
+      "epoch": 0.34210526315789475,
+      "grad_norm": 0.5297996401786804,
+      "learning_rate": 1.5694418926729148e-05,
+      "loss": 0.4935210943222046,
+      "step": 624
+    },
+    {
+      "epoch": 0.34265350877192985,
+      "grad_norm": 0.5105075240135193,
+      "learning_rate": 1.5679499126327044e-05,
+      "loss": 0.4937981963157654,
+      "step": 625
+    },
+    {
+      "epoch": 0.3432017543859649,
+      "grad_norm": 0.4859083592891693,
+      "learning_rate": 1.566456064003064e-05,
+      "loss": 0.4915929436683655,
+      "step": 626
+    },
+    {
+      "epoch": 0.34375,
+      "grad_norm": 0.5445349812507629,
+      "learning_rate": 1.564960351698844e-05,
+      "loss": 0.5122596025466919,
+      "step": 627
+    },
+    {
+      "epoch": 0.3442982456140351,
+      "grad_norm": 0.49146509170532227,
+      "learning_rate": 1.5634627806410296e-05,
+      "loss": 0.4980570673942566,
+      "step": 628
+    },
+    {
+      "epoch": 0.34484649122807015,
+      "grad_norm": 0.5087494254112244,
+      "learning_rate": 1.561963355756719e-05,
+      "loss": 0.4985271394252777,
+      "step": 629
+    },
+    {
+      "epoch": 0.34539473684210525,
+      "grad_norm": 0.5231812000274658,
+      "learning_rate": 1.56046208197911e-05,
+      "loss": 0.5052623152732849,
+      "step": 630
+    },
+    {
+      "epoch": 0.34594298245614036,
+      "grad_norm": 0.5085192322731018,
+      "learning_rate": 1.5589589642474843e-05,
+      "loss": 0.5217705965042114,
+      "step": 631
+    },
+    {
+      "epoch": 0.34649122807017546,
+      "grad_norm": 0.5085150599479675,
+      "learning_rate": 1.55745400750719e-05,
+      "loss": 0.5063439607620239,
+      "step": 632
+    },
+    {
+      "epoch": 0.3470394736842105,
+      "grad_norm": 0.4942718744277954,
+      "learning_rate": 1.5559472167096252e-05,
+      "loss": 0.508962869644165,
+      "step": 633
+    },
+    {
+      "epoch": 0.3475877192982456,
+      "grad_norm": 0.5233277082443237,
+      "learning_rate": 1.5544385968122225e-05,
+      "loss": 0.4943055808544159,
+      "step": 634
+    },
+    {
+      "epoch": 0.3481359649122807,
+      "grad_norm": 0.5197892189025879,
+      "learning_rate": 1.552928152778432e-05,
+      "loss": 0.5063210725784302,
+      "step": 635
+    },
+    {
+      "epoch": 0.34868421052631576,
+      "grad_norm": 0.5159553289413452,
+      "learning_rate": 1.5514158895777057e-05,
+      "loss": 0.5054895877838135,
+      "step": 636
+    },
+    {
+      "epoch": 0.34923245614035087,
+      "grad_norm": 0.49733617901802063,
+      "learning_rate": 1.5499018121854808e-05,
+      "loss": 0.49076005816459656,
+      "step": 637
+    },
+    {
+      "epoch": 0.34978070175438597,
+      "grad_norm": 0.5775347948074341,
+      "learning_rate": 1.5483859255831628e-05,
+      "loss": 0.4983357787132263,
+      "step": 638
+    },
+    {
+      "epoch": 0.3503289473684211,
+      "grad_norm": 0.5254983305931091,
+      "learning_rate": 1.5468682347581102e-05,
+      "loss": 0.5100836753845215,
+      "step": 639
+    },
+    {
+      "epoch": 0.3508771929824561,
+      "grad_norm": 0.5384511947631836,
+      "learning_rate": 1.5453487447036174e-05,
+      "loss": 0.517541766166687,
+      "step": 640
+    },
+    {
+      "epoch": 0.3514254385964912,
+      "grad_norm": 0.5437305569648743,
+      "learning_rate": 1.5438274604188977e-05,
+      "loss": 0.5047601461410522,
+      "step": 641
+    },
+    {
+      "epoch": 0.3519736842105263,
+      "grad_norm": 0.5298881530761719,
+      "learning_rate": 1.5423043869090687e-05,
+      "loss": 0.49070271849632263,
+      "step": 642
+    },
+    {
+      "epoch": 0.35252192982456143,
+      "grad_norm": 0.5131424069404602,
+      "learning_rate": 1.5407795291851336e-05,
+      "loss": 0.5067086219787598,
+      "step": 643
+    },
+    {
+      "epoch": 0.3530701754385965,
+      "grad_norm": 0.538144588470459,
+      "learning_rate": 1.5392528922639662e-05,
+      "loss": 0.4947914481163025,
+      "step": 644
+    },
+    {
+      "epoch": 0.3536184210526316,
+      "grad_norm": 0.5141587257385254,
+      "learning_rate": 1.537724481168294e-05,
+      "loss": 0.492933452129364,
+      "step": 645
+    },
+    {
+      "epoch": 0.3541666666666667,
+      "grad_norm": 0.5233535170555115,
+      "learning_rate": 1.536194300926682e-05,
+      "loss": 0.5006459951400757,
+      "step": 646
+    },
+    {
+      "epoch": 0.35471491228070173,
+      "grad_norm": 0.4955417215824127,
+      "learning_rate": 1.5346623565735145e-05,
+      "loss": 0.5072038769721985,
+      "step": 647
+    },
+    {
+      "epoch": 0.35526315789473684,
+      "grad_norm": 0.4693199098110199,
+      "learning_rate": 1.533128653148982e-05,
+      "loss": 0.49651047587394714,
+      "step": 648
+    },
+    {
+      "epoch": 0.35581140350877194,
+      "grad_norm": 0.5177298784255981,
+      "learning_rate": 1.53159319569906e-05,
+      "loss": 0.5096155405044556,
+      "step": 649
+    },
+    {
+      "epoch": 0.35635964912280704,
+      "grad_norm": 0.5220261216163635,
+      "learning_rate": 1.5300559892754957e-05,
+      "loss": 0.503006637096405,
+      "step": 650
+    },
+    {
+      "epoch": 0.3569078947368421,
+      "grad_norm": 0.4956647753715515,
+      "learning_rate": 1.5285170389357924e-05,
+      "loss": 0.492157906293869,
+      "step": 651
+    },
+    {
+      "epoch": 0.3574561403508772,
+      "grad_norm": 0.5249745845794678,
+      "learning_rate": 1.5269763497431882e-05,
+      "loss": 0.5008402466773987,
+      "step": 652
+    },
+    {
+      "epoch": 0.3580043859649123,
+      "grad_norm": 0.5125779509544373,
+      "learning_rate": 1.525433926766644e-05,
+      "loss": 0.49728164076805115,
+      "step": 653
+    },
+    {
+      "epoch": 0.35855263157894735,
+      "grad_norm": 0.4863375425338745,
+      "learning_rate": 1.5238897750808242e-05,
+      "loss": 0.5007228255271912,
+      "step": 654
+    },
+    {
+      "epoch": 0.35910087719298245,
+      "grad_norm": 0.5670150518417358,
+      "learning_rate": 1.5223438997660802e-05,
+      "loss": 0.4918896555900574,
+      "step": 655
+    },
+    {
+      "epoch": 0.35964912280701755,
+      "grad_norm": 0.5108444690704346,
+      "learning_rate": 1.520796305908436e-05,
+      "loss": 0.4987252354621887,
+      "step": 656
+    },
+    {
+      "epoch": 0.36019736842105265,
+      "grad_norm": 0.48743292689323425,
+      "learning_rate": 1.5192469985995684e-05,
+      "loss": 0.5023507475852966,
+      "step": 657
+    },
+    {
+      "epoch": 0.3607456140350877,
+      "grad_norm": 0.4842602610588074,
+      "learning_rate": 1.5176959829367907e-05,
+      "loss": 0.5085281133651733,
+      "step": 658
+    },
+    {
+      "epoch": 0.3612938596491228,
+      "grad_norm": 0.4879213869571686,
+      "learning_rate": 1.5161432640230391e-05,
+      "loss": 0.4956406056880951,
+      "step": 659
+    },
+    {
+      "epoch": 0.3618421052631579,
+      "grad_norm": 0.4824928343296051,
+      "learning_rate": 1.5145888469668521e-05,
+      "loss": 0.5034121870994568,
+      "step": 660
+    },
+    {
+      "epoch": 0.36239035087719296,
+      "grad_norm": 0.4922201931476593,
+      "learning_rate": 1.5130327368823551e-05,
+      "loss": 0.5032166838645935,
+      "step": 661
+    },
+    {
+      "epoch": 0.36293859649122806,
+      "grad_norm": 0.4866214096546173,
+      "learning_rate": 1.5114749388892445e-05,
+      "loss": 0.5006184577941895,
+      "step": 662
+    },
+    {
+      "epoch": 0.36348684210526316,
+      "grad_norm": 0.48263096809387207,
+      "learning_rate": 1.509915458112769e-05,
+      "loss": 0.4939555525779724,
+      "step": 663
+    },
+    {
+      "epoch": 0.36403508771929827,
+      "grad_norm": 0.47373154759407043,
+      "learning_rate": 1.508354299683715e-05,
+      "loss": 0.49878406524658203,
+      "step": 664
+    },
+    {
+      "epoch": 0.3645833333333333,
+      "grad_norm": 0.48171475529670715,
+      "learning_rate": 1.5067914687383873e-05,
+      "loss": 0.5021742582321167,
+      "step": 665
+    },
+    {
+      "epoch": 0.3651315789473684,
+      "grad_norm": 0.4749904274940491,
+      "learning_rate": 1.505226970418594e-05,
+      "loss": 0.4902474880218506,
+      "step": 666
+    },
+    {
+      "epoch": 0.3656798245614035,
+      "grad_norm": 0.5045875310897827,
+      "learning_rate": 1.5036608098716287e-05,
+      "loss": 0.5088133811950684,
+      "step": 667
+    },
+    {
+      "epoch": 0.36622807017543857,
+      "grad_norm": 0.5372963547706604,
+      "learning_rate": 1.5020929922502542e-05,
+      "loss": 0.5023071765899658,
+      "step": 668
+    },
+    {
+      "epoch": 0.3667763157894737,
+      "grad_norm": 0.4800300598144531,
+      "learning_rate": 1.5005235227126846e-05,
+      "loss": 0.505648136138916,
+      "step": 669
+    },
+    {
+      "epoch": 0.3673245614035088,
+      "grad_norm": 0.4792078137397766,
+      "learning_rate": 1.4989524064225696e-05,
+      "loss": 0.5006028413772583,
+      "step": 670
+    },
+    {
+      "epoch": 0.3678728070175439,
+      "grad_norm": 0.48742419481277466,
+      "learning_rate": 1.4973796485489764e-05,
+      "loss": 0.515373170375824,
+      "step": 671
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.5054008960723877,
+      "learning_rate": 1.4958052542663728e-05,
+      "loss": 0.4846934676170349,
+      "step": 672
+    },
+    {
+      "epoch": 0.36896929824561403,
+      "grad_norm": 0.49790146946907043,
+      "learning_rate": 1.4942292287546114e-05,
+      "loss": 0.49283909797668457,
+      "step": 673
+    },
+    {
+      "epoch": 0.36951754385964913,
+      "grad_norm": 0.5341838002204895,
+      "learning_rate": 1.4926515771989106e-05,
+      "loss": 0.5021057724952698,
+      "step": 674
+    },
+    {
+      "epoch": 0.37006578947368424,
+      "grad_norm": 0.5239306092262268,
+      "learning_rate": 1.4910723047898392e-05,
+      "loss": 0.49403518438339233,
+      "step": 675
+    },
+    {
+      "epoch": 0.3706140350877193,
+      "grad_norm": 0.4920117259025574,
+      "learning_rate": 1.4894914167232988e-05,
+      "loss": 0.49569496512413025,
+      "step": 676
+    },
+    {
+      "epoch": 0.3711622807017544,
+      "grad_norm": 0.49796634912490845,
+      "learning_rate": 1.4879089182005062e-05,
+      "loss": 0.5069082975387573,
+      "step": 677
+    },
+    {
+      "epoch": 0.3717105263157895,
+      "grad_norm": 0.5147514343261719,
+      "learning_rate": 1.4863248144279769e-05,
+      "loss": 0.4919740557670593,
+      "step": 678
+    },
+    {
+      "epoch": 0.37225877192982454,
+      "grad_norm": 0.5094771385192871,
+      "learning_rate": 1.4847391106175075e-05,
+      "loss": 0.5015633702278137,
+      "step": 679
+    },
+    {
+      "epoch": 0.37280701754385964,
+      "grad_norm": 0.5420538783073425,
+      "learning_rate": 1.4831518119861597e-05,
+      "loss": 0.48816415667533875,
+      "step": 680
+    },
+    {
+      "epoch": 0.37335526315789475,
+      "grad_norm": 0.5352575778961182,
+      "learning_rate": 1.4815629237562411e-05,
+      "loss": 0.49396681785583496,
+      "step": 681
+    },
+    {
+      "epoch": 0.37390350877192985,
+      "grad_norm": 0.5042607188224792,
+      "learning_rate": 1.4799724511552898e-05,
+      "loss": 0.48261889815330505,
+      "step": 682
+    },
+    {
+      "epoch": 0.3744517543859649,
+      "grad_norm": 0.5831988453865051,
+      "learning_rate": 1.4783803994160568e-05,
+      "loss": 0.4972837567329407,
+      "step": 683
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.48803001642227173,
+      "learning_rate": 1.4767867737764877e-05,
+      "loss": 0.49230509996414185,
+      "step": 684
+    },
+    {
+      "epoch": 0.3755482456140351,
+      "grad_norm": 0.5121214985847473,
+      "learning_rate": 1.4751915794797072e-05,
+      "loss": 0.5085790157318115,
+      "step": 685
+    },
+    {
+      "epoch": 0.37609649122807015,
+      "grad_norm": 0.5156803727149963,
+      "learning_rate": 1.4735948217740006e-05,
+      "loss": 0.4978635013103485,
+      "step": 686
+    },
+    {
+      "epoch": 0.37664473684210525,
+      "grad_norm": 0.46112072467803955,
+      "learning_rate": 1.4719965059127969e-05,
+      "loss": 0.49163517355918884,
+      "step": 687
+    },
+    {
+      "epoch": 0.37719298245614036,
+      "grad_norm": 0.5171255469322205,
+      "learning_rate": 1.4703966371546519e-05,
+      "loss": 0.5190042853355408,
+      "step": 688
+    },
+    {
+      "epoch": 0.37774122807017546,
+      "grad_norm": 0.5324544310569763,
+      "learning_rate": 1.4687952207632296e-05,
+      "loss": 0.5000194907188416,
+      "step": 689
+    },
+    {
+      "epoch": 0.3782894736842105,
+      "grad_norm": 0.5022001266479492,
+      "learning_rate": 1.4671922620072869e-05,
+      "loss": 0.5025606751441956,
+      "step": 690
+    },
+    {
+      "epoch": 0.3788377192982456,
+      "grad_norm": 0.5057377219200134,
+      "learning_rate": 1.4655877661606546e-05,
+      "loss": 0.49487099051475525,
+      "step": 691
+    },
+    {
+      "epoch": 0.3793859649122807,
+      "grad_norm": 0.5015015006065369,
+      "learning_rate": 1.4639817385022206e-05,
+      "loss": 0.496290385723114,
+      "step": 692
+    },
+    {
+      "epoch": 0.37993421052631576,
+      "grad_norm": 0.49644434452056885,
+      "learning_rate": 1.4623741843159124e-05,
+      "loss": 0.49527817964553833,
+      "step": 693
+    },
+    {
+      "epoch": 0.38048245614035087,
+      "grad_norm": 0.5267723202705383,
+      "learning_rate": 1.460765108890681e-05,
+      "loss": 0.5009031295776367,
+      "step": 694
+    },
+    {
+      "epoch": 0.38103070175438597,
+      "grad_norm": 0.5172473192214966,
+      "learning_rate": 1.4591545175204802e-05,
+      "loss": 0.511361837387085,
+      "step": 695
+    },
+    {
+      "epoch": 0.3815789473684211,
+      "grad_norm": 0.5121990442276001,
+      "learning_rate": 1.4575424155042537e-05,
+      "loss": 0.4846411943435669,
+      "step": 696
+    },
+    {
+      "epoch": 0.3821271929824561,
+      "grad_norm": 0.508580207824707,
+      "learning_rate": 1.4559288081459142e-05,
+      "loss": 0.49809813499450684,
+      "step": 697
+    },
+    {
+      "epoch": 0.3826754385964912,
+      "grad_norm": 0.5329216718673706,
+      "learning_rate": 1.4543137007543265e-05,
+      "loss": 0.5044119358062744,
+      "step": 698
+    },
+    {
+      "epoch": 0.3832236842105263,
+      "grad_norm": 0.524181604385376,
+      "learning_rate": 1.4526970986432916e-05,
+      "loss": 0.4943583011627197,
+      "step": 699
+    },
+    {
+      "epoch": 0.38377192982456143,
+      "grad_norm": 0.5000477433204651,
+      "learning_rate": 1.4510790071315278e-05,
+      "loss": 0.4813290536403656,
+      "step": 700
+    },
+    {
+      "epoch": 0.3843201754385965,
+      "grad_norm": 0.50678551197052,
+      "learning_rate": 1.4494594315426535e-05,
+      "loss": 0.5046131610870361,
+      "step": 701
+    },
+    {
+      "epoch": 0.3848684210526316,
+      "grad_norm": 0.5136652588844299,
+      "learning_rate": 1.4478383772051707e-05,
+      "loss": 0.5002473592758179,
+      "step": 702
+    },
+    {
+      "epoch": 0.3854166666666667,
+      "grad_norm": 0.5075159668922424,
+      "learning_rate": 1.4462158494524447e-05,
+      "loss": 0.4981658458709717,
+      "step": 703
+    },
+    {
+      "epoch": 0.38596491228070173,
+      "grad_norm": 0.48104289174079895,
+      "learning_rate": 1.4445918536226904e-05,
+      "loss": 0.49785614013671875,
+      "step": 704
+    },
+    {
+      "epoch": 0.38651315789473684,
+      "grad_norm": 0.49635618925094604,
+      "learning_rate": 1.4429663950589517e-05,
+      "loss": 0.5005432367324829,
+      "step": 705
+    },
+    {
+      "epoch": 0.38706140350877194,
+      "grad_norm": 0.5516581535339355,
+      "learning_rate": 1.441339479109085e-05,
+      "loss": 0.4959232211112976,
+      "step": 706
+    },
+    {
+      "epoch": 0.38760964912280704,
+      "grad_norm": 0.5026279091835022,
+      "learning_rate": 1.439711111125742e-05,
+      "loss": 0.4876520037651062,
+      "step": 707
+    },
+    {
+      "epoch": 0.3881578947368421,
+      "grad_norm": 0.5157284736633301,
+      "learning_rate": 1.4380812964663514e-05,
+      "loss": 0.486855149269104,
+      "step": 708
+    },
+    {
+      "epoch": 0.3887061403508772,
+      "grad_norm": 0.5165179967880249,
+      "learning_rate": 1.4364500404931017e-05,
+      "loss": 0.4917711615562439,
+      "step": 709
+    },
+    {
+      "epoch": 0.3892543859649123,
+      "grad_norm": 0.5212239027023315,
+      "learning_rate": 1.4348173485729233e-05,
+      "loss": 0.4904009699821472,
+      "step": 710
+    },
+    {
+      "epoch": 0.38980263157894735,
+      "grad_norm": 0.5552125573158264,
+      "learning_rate": 1.433183226077471e-05,
+      "loss": 0.4961950480937958,
+      "step": 711
+    },
+    {
+      "epoch": 0.39035087719298245,
+      "grad_norm": 0.5230804681777954,
+      "learning_rate": 1.4315476783831062e-05,
+      "loss": 0.4872981309890747,
+      "step": 712
+    },
+    {
+      "epoch": 0.39089912280701755,
+      "grad_norm": 0.5566909313201904,
+      "learning_rate": 1.4299107108708792e-05,
+      "loss": 0.4958663880825043,
+      "step": 713
+    },
+    {
+      "epoch": 0.39144736842105265,
+      "grad_norm": 0.5392793416976929,
+      "learning_rate": 1.4282723289265122e-05,
+      "loss": 0.49213212728500366,
+      "step": 714
+    },
+    {
+      "epoch": 0.3919956140350877,
+      "grad_norm": 0.4880681335926056,
+      "learning_rate": 1.42663253794038e-05,
+      "loss": 0.4859275221824646,
+      "step": 715
+    },
+    {
+      "epoch": 0.3925438596491228,
+      "grad_norm": 0.5517839789390564,
+      "learning_rate": 1.4249913433074942e-05,
+      "loss": 0.5010792016983032,
+      "step": 716
+    },
+    {
+      "epoch": 0.3930921052631579,
+      "grad_norm": 0.49621888995170593,
+      "learning_rate": 1.4233487504274837e-05,
+      "loss": 0.49741482734680176,
+      "step": 717
+    },
+    {
+      "epoch": 0.39364035087719296,
+      "grad_norm": 0.505009651184082,
+      "learning_rate": 1.4217047647045781e-05,
+      "loss": 0.49573537707328796,
+      "step": 718
+    },
+    {
+      "epoch": 0.39418859649122806,
+      "grad_norm": 0.49585750699043274,
+      "learning_rate": 1.4200593915475895e-05,
+      "loss": 0.4880044460296631,
+      "step": 719
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 0.4711382985115051,
+      "learning_rate": 1.4184126363698949e-05,
+      "loss": 0.4827781319618225,
+      "step": 720
+    },
+    {
+      "epoch": 0.39528508771929827,
+      "grad_norm": 0.500230073928833,
+      "learning_rate": 1.4167645045894176e-05,
+      "loss": 0.48407694697380066,
+      "step": 721
+    },
+    {
+      "epoch": 0.3958333333333333,
+      "grad_norm": 0.5288121700286865,
+      "learning_rate": 1.415115001628611e-05,
+      "loss": 0.48783645033836365,
+      "step": 722
+    },
+    {
+      "epoch": 0.3963815789473684,
+      "grad_norm": 0.47950515151023865,
+      "learning_rate": 1.413464132914439e-05,
+      "loss": 0.49616605043411255,
+      "step": 723
+    },
+    {
+      "epoch": 0.3969298245614035,
+      "grad_norm": 0.5053483247756958,
+      "learning_rate": 1.4118119038783589e-05,
+      "loss": 0.484591007232666,
+      "step": 724
+    },
+    {
+      "epoch": 0.39747807017543857,
+      "grad_norm": 0.4833005964756012,
+      "learning_rate": 1.4101583199563038e-05,
+      "loss": 0.4897410273551941,
+      "step": 725
+    },
+    {
+      "epoch": 0.3980263157894737,
+      "grad_norm": 0.5118196606636047,
+      "learning_rate": 1.4085033865886649e-05,
+      "loss": 0.4865984320640564,
+      "step": 726
+    },
+    {
+      "epoch": 0.3985745614035088,
+      "grad_norm": 0.5035130977630615,
+      "learning_rate": 1.4068471092202723e-05,
+      "loss": 0.5033597946166992,
+      "step": 727
+    },
+    {
+      "epoch": 0.3991228070175439,
+      "grad_norm": 0.5169758200645447,
+      "learning_rate": 1.4051894933003783e-05,
+      "loss": 0.49889934062957764,
+      "step": 728
+    },
+    {
+      "epoch": 0.3996710526315789,
+      "grad_norm": 0.5340895652770996,
+      "learning_rate": 1.4035305442826392e-05,
+      "loss": 0.5121870040893555,
+      "step": 729
+    },
+    {
+      "epoch": 0.40021929824561403,
+      "grad_norm": 0.513810932636261,
+      "learning_rate": 1.4018702676250973e-05,
+      "loss": 0.46407610177993774,
+      "step": 730
+    },
+    {
+      "epoch": 0.40076754385964913,
+      "grad_norm": 0.47475552558898926,
+      "learning_rate": 1.400208668790163e-05,
+      "loss": 0.5128594636917114,
+      "step": 731
+    },
+    {
+      "epoch": 0.40131578947368424,
+      "grad_norm": 0.532416045665741,
+      "learning_rate": 1.3985457532445964e-05,
+      "loss": 0.4869621694087982,
+      "step": 732
+    },
+    {
+      "epoch": 0.4018640350877193,
+      "grad_norm": 0.5194771885871887,
+      "learning_rate": 1.3968815264594895e-05,
+      "loss": 0.4942741096019745,
+      "step": 733
+    },
+    {
+      "epoch": 0.4024122807017544,
+      "grad_norm": 0.5377725958824158,
+      "learning_rate": 1.3952159939102492e-05,
+      "loss": 0.499100923538208,
+      "step": 734
+    },
+    {
+      "epoch": 0.4029605263157895,
+      "grad_norm": 0.5000717639923096,
+      "learning_rate": 1.3935491610765776e-05,
+      "loss": 0.4920429587364197,
+      "step": 735
+    },
+    {
+      "epoch": 0.40350877192982454,
+      "grad_norm": 0.5088208913803101,
+      "learning_rate": 1.3918810334424556e-05,
+      "loss": 0.49835777282714844,
+      "step": 736
+    },
+    {
+      "epoch": 0.40405701754385964,
+      "grad_norm": 0.4984104335308075,
+      "learning_rate": 1.390211616496123e-05,
+      "loss": 0.492436945438385,
+      "step": 737
+    },
+    {
+      "epoch": 0.40460526315789475,
+      "grad_norm": 0.531671941280365,
+      "learning_rate": 1.3885409157300624e-05,
+      "loss": 0.5021324157714844,
+      "step": 738
+    },
+    {
+      "epoch": 0.40515350877192985,
+      "grad_norm": 0.521382212638855,
+      "learning_rate": 1.3868689366409802e-05,
+      "loss": 0.48615261912345886,
+      "step": 739
+    },
+    {
+      "epoch": 0.4057017543859649,
+      "grad_norm": 0.4979918301105499,
+      "learning_rate": 1.3851956847297882e-05,
+      "loss": 0.48778805136680603,
+      "step": 740
+    },
+    {
+      "epoch": 0.40625,
+      "grad_norm": 0.5014253258705139,
+      "learning_rate": 1.3835211655015856e-05,
+      "loss": 0.5012838840484619,
+      "step": 741
+    },
+    {
+      "epoch": 0.4067982456140351,
+      "grad_norm": 0.5053998827934265,
+      "learning_rate": 1.3818453844656423e-05,
+      "loss": 0.49050578474998474,
+      "step": 742
+    },
+    {
+      "epoch": 0.40734649122807015,
+      "grad_norm": 0.5173316597938538,
+      "learning_rate": 1.3801683471353781e-05,
+      "loss": 0.49842238426208496,
+      "step": 743
+    },
+    {
+      "epoch": 0.40789473684210525,
+      "grad_norm": 0.5086451172828674,
+      "learning_rate": 1.3784900590283474e-05,
+      "loss": 0.5004265308380127,
+      "step": 744
+    },
+    {
+      "epoch": 0.40844298245614036,
+      "grad_norm": 0.5439842939376831,
+      "learning_rate": 1.3768105256662194e-05,
+      "loss": 0.4831674098968506,
+      "step": 745
+    },
+    {
+      "epoch": 0.40899122807017546,
+      "grad_norm": 0.4946817457675934,
+      "learning_rate": 1.3751297525747589e-05,
+      "loss": 0.49196428060531616,
+      "step": 746
+    },
+    {
+      "epoch": 0.4095394736842105,
+      "grad_norm": 0.5227308869361877,
+      "learning_rate": 1.3734477452838117e-05,
+      "loss": 0.490816205739975,
+      "step": 747
+    },
+    {
+      "epoch": 0.4100877192982456,
+      "grad_norm": 0.560154914855957,
+      "learning_rate": 1.3717645093272834e-05,
+      "loss": 0.4868409037590027,
+      "step": 748
+    },
+    {
+      "epoch": 0.4106359649122807,
+      "grad_norm": 0.5022367835044861,
+      "learning_rate": 1.3700800502431202e-05,
+      "loss": 0.4902645945549011,
+      "step": 749
+    },
+    {
+      "epoch": 0.41118421052631576,
+      "grad_norm": 0.5088184475898743,
+      "learning_rate": 1.3683943735732956e-05,
+      "loss": 0.4881845712661743,
+      "step": 750
+    },
+    {
+      "epoch": 0.41173245614035087,
+      "grad_norm": 0.5308793187141418,
+      "learning_rate": 1.3667074848637862e-05,
+      "loss": 0.4919361472129822,
+      "step": 751
+    },
+    {
+      "epoch": 0.41228070175438597,
+      "grad_norm": 0.48994314670562744,
+      "learning_rate": 1.3650193896645582e-05,
+      "loss": 0.4936677813529968,
+      "step": 752
+    },
+    {
+      "epoch": 0.4128289473684211,
+      "grad_norm": 0.5036956667900085,
+      "learning_rate": 1.3633300935295468e-05,
+      "loss": 0.4867483079433441,
+      "step": 753
+    },
+    {
+      "epoch": 0.4133771929824561,
+      "grad_norm": 0.4999926686286926,
+      "learning_rate": 1.3616396020166368e-05,
+      "loss": 0.4871666133403778,
+      "step": 754
+    },
+    {
+      "epoch": 0.4139254385964912,
+      "grad_norm": 0.4995878338813782,
+      "learning_rate": 1.3599479206876483e-05,
+      "loss": 0.49972087144851685,
+      "step": 755
+    },
+    {
+      "epoch": 0.4144736842105263,
+      "grad_norm": 0.481976717710495,
+      "learning_rate": 1.3582550551083143e-05,
+      "loss": 0.49500545859336853,
+      "step": 756
+    },
+    {
+      "epoch": 0.41502192982456143,
+      "grad_norm": 0.48889246582984924,
+      "learning_rate": 1.3565610108482645e-05,
+      "loss": 0.49687570333480835,
+      "step": 757
+    },
+    {
+      "epoch": 0.4155701754385965,
+      "grad_norm": 0.5260644555091858,
+      "learning_rate": 1.354865793481007e-05,
+      "loss": 0.4825992286205292,
+      "step": 758
+    },
+    {
+      "epoch": 0.4161184210526316,
+      "grad_norm": 0.4933117926120758,
+      "learning_rate": 1.3531694085839084e-05,
+      "loss": 0.48860979080200195,
+      "step": 759
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 0.5438188314437866,
+      "learning_rate": 1.3514718617381779e-05,
+      "loss": 0.49466273188591003,
+      "step": 760
+    },
+    {
+      "epoch": 0.41721491228070173,
+      "grad_norm": 0.5429419875144958,
+      "learning_rate": 1.3497731585288466e-05,
+      "loss": 0.496579647064209,
+      "step": 761
+    },
+    {
+      "epoch": 0.41776315789473684,
+      "grad_norm": 0.4868009090423584,
+      "learning_rate": 1.3480733045447505e-05,
+      "loss": 0.4871636629104614,
+      "step": 762
+    },
+    {
+      "epoch": 0.41831140350877194,
+      "grad_norm": 0.5444782376289368,
+      "learning_rate": 1.3463723053785114e-05,
+      "loss": 0.4599669873714447,
+      "step": 763
+    },
+    {
+      "epoch": 0.41885964912280704,
+      "grad_norm": 0.503890335559845,
+      "learning_rate": 1.3446701666265192e-05,
+      "loss": 0.48242682218551636,
+      "step": 764
+    },
+    {
+      "epoch": 0.4194078947368421,
+      "grad_norm": 0.6804667711257935,
+      "learning_rate": 1.3429668938889127e-05,
+      "loss": 0.49555811285972595,
+      "step": 765
+    },
+    {
+      "epoch": 0.4199561403508772,
+      "grad_norm": 0.5674505829811096,
+      "learning_rate": 1.341262492769562e-05,
+      "loss": 0.494424045085907,
+      "step": 766
+    },
+    {
+      "epoch": 0.4205043859649123,
+      "grad_norm": 0.49492350220680237,
+      "learning_rate": 1.3395569688760499e-05,
+      "loss": 0.4926236569881439,
+      "step": 767
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.48057791590690613,
+      "learning_rate": 1.3378503278196522e-05,
+      "loss": 0.48030886054039,
+      "step": 768
+    },
+    {
+      "epoch": 0.42160087719298245,
+      "grad_norm": 0.49380749464035034,
+      "learning_rate": 1.336142575215321e-05,
+      "loss": 0.48291677236557007,
+      "step": 769
+    },
+    {
+      "epoch": 0.42214912280701755,
+      "grad_norm": 0.5208574533462524,
+      "learning_rate": 1.3344337166816652e-05,
+      "loss": 0.482435405254364,
+      "step": 770
+    },
+    {
+      "epoch": 0.42269736842105265,
+      "grad_norm": 0.48892274498939514,
+      "learning_rate": 1.3327237578409323e-05,
+      "loss": 0.48565933108329773,
+      "step": 771
+    },
+    {
+      "epoch": 0.4232456140350877,
+      "grad_norm": 0.4815227687358856,
+      "learning_rate": 1.33101270431899e-05,
+      "loss": 0.48251694440841675,
+      "step": 772
+    },
+    {
+      "epoch": 0.4237938596491228,
+      "grad_norm": 0.4896012544631958,
+      "learning_rate": 1.3293005617453075e-05,
+      "loss": 0.47662657499313354,
+      "step": 773
+    },
+    {
+      "epoch": 0.4243421052631579,
+      "grad_norm": 0.4705606698989868,
+      "learning_rate": 1.327587335752937e-05,
+      "loss": 0.4908991754055023,
+      "step": 774
+    },
+    {
+      "epoch": 0.42489035087719296,
+      "grad_norm": 0.46313512325286865,
+      "learning_rate": 1.325873031978495e-05,
+      "loss": 0.4786376357078552,
+      "step": 775
+    },
+    {
+      "epoch": 0.42543859649122806,
+      "grad_norm": 0.4871222972869873,
+      "learning_rate": 1.3241576560621444e-05,
+      "loss": 0.4758555293083191,
+      "step": 776
+    },
+    {
+      "epoch": 0.42598684210526316,
+      "grad_norm": 0.4975806474685669,
+      "learning_rate": 1.3224412136475757e-05,
+      "loss": 0.4800441265106201,
+      "step": 777
+    },
+    {
+      "epoch": 0.42653508771929827,
+      "grad_norm": 0.4865629971027374,
+      "learning_rate": 1.320723710381987e-05,
+      "loss": 0.4875626564025879,
+      "step": 778
+    },
+    {
+      "epoch": 0.4270833333333333,
+      "grad_norm": 0.5996370315551758,
+      "learning_rate": 1.3190051519160685e-05,
+      "loss": 0.4718167781829834,
+      "step": 779
+    },
+    {
+      "epoch": 0.4276315789473684,
+      "grad_norm": 0.47141125798225403,
+      "learning_rate": 1.3172855439039802e-05,
+      "loss": 0.47652530670166016,
+      "step": 780
+    },
+    {
+      "epoch": 0.4281798245614035,
+      "grad_norm": 0.48037710785865784,
+      "learning_rate": 1.3155648920033363e-05,
+      "loss": 0.4848669469356537,
+      "step": 781
+    },
+    {
+      "epoch": 0.42872807017543857,
+      "grad_norm": 0.49942100048065186,
+      "learning_rate": 1.313843201875186e-05,
+      "loss": 0.48465991020202637,
+      "step": 782
+    },
+    {
+      "epoch": 0.4292763157894737,
+      "grad_norm": 0.4947258234024048,
+      "learning_rate": 1.3121204791839926e-05,
+      "loss": 0.48432278633117676,
+      "step": 783
+    },
+    {
+      "epoch": 0.4298245614035088,
+      "grad_norm": 0.5264876484870911,
+      "learning_rate": 1.3103967295976178e-05,
+      "loss": 0.48953860998153687,
+      "step": 784
+    },
+    {
+      "epoch": 0.4303728070175439,
+      "grad_norm": 0.5139599442481995,
+      "learning_rate": 1.3086719587873015e-05,
+      "loss": 0.48061832785606384,
+      "step": 785
+    },
+    {
+      "epoch": 0.4309210526315789,
+      "grad_norm": 0.48673495650291443,
+      "learning_rate": 1.306946172427644e-05,
+      "loss": 0.47429126501083374,
+      "step": 786
+    },
+    {
+      "epoch": 0.43146929824561403,
+      "grad_norm": 0.5102007389068604,
+      "learning_rate": 1.3052193761965861e-05,
+      "loss": 0.4935256838798523,
+      "step": 787
+    },
+    {
+      "epoch": 0.43201754385964913,
+      "grad_norm": 0.5277912020683289,
+      "learning_rate": 1.3034915757753917e-05,
+      "loss": 0.4868704676628113,
+      "step": 788
+    },
+    {
+      "epoch": 0.43256578947368424,
+      "grad_norm": 0.49750053882598877,
+      "learning_rate": 1.3017627768486277e-05,
+      "loss": 0.4760574698448181,
+      "step": 789
+    },
+    {
+      "epoch": 0.4331140350877193,
+      "grad_norm": 0.6148756146430969,
+      "learning_rate": 1.3000329851041474e-05,
+      "loss": 0.48425477743148804,
+      "step": 790
+    },
+    {
+      "epoch": 0.4336622807017544,
+      "grad_norm": 0.487529993057251,
+      "learning_rate": 1.298302206233069e-05,
+      "loss": 0.49096059799194336,
+      "step": 791
+    },
+    {
+      "epoch": 0.4342105263157895,
+      "grad_norm": 0.5019875168800354,
+      "learning_rate": 1.2965704459297603e-05,
+      "loss": 0.48441147804260254,
+      "step": 792
+    },
+    {
+      "epoch": 0.43475877192982454,
+      "grad_norm": 0.48055264353752136,
+      "learning_rate": 1.2948377098918166e-05,
+      "loss": 0.49248749017715454,
+      "step": 793
+    },
+    {
+      "epoch": 0.43530701754385964,
+      "grad_norm": 0.4806983470916748,
+      "learning_rate": 1.2931040038200435e-05,
+      "loss": 0.46774157881736755,
+      "step": 794
+    },
+    {
+      "epoch": 0.43585526315789475,
+      "grad_norm": 0.5055975914001465,
+      "learning_rate": 1.2913693334184387e-05,
+      "loss": 0.46599435806274414,
+      "step": 795
+    },
+    {
+      "epoch": 0.43640350877192985,
+      "grad_norm": 0.47909224033355713,
+      "learning_rate": 1.289633704394173e-05,
+      "loss": 0.49138325452804565,
+      "step": 796
+    },
+    {
+      "epoch": 0.4369517543859649,
+      "grad_norm": 0.5127816796302795,
+      "learning_rate": 1.2878971224575689e-05,
+      "loss": 0.49381622672080994,
+      "step": 797
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 0.5159894824028015,
+      "learning_rate": 1.2861595933220873e-05,
+      "loss": 0.4878365397453308,
+      "step": 798
+    },
+    {
+      "epoch": 0.4380482456140351,
+      "grad_norm": 0.48489898443222046,
+      "learning_rate": 1.2844211227043024e-05,
+      "loss": 0.4896066188812256,
+      "step": 799
+    },
+    {
+      "epoch": 0.43859649122807015,
+      "grad_norm": 0.5021419525146484,
+      "learning_rate": 1.282681716323888e-05,
+      "loss": 0.48621922731399536,
+      "step": 800
+    },
+    {
+      "epoch": 0.43914473684210525,
+      "grad_norm": 0.48859429359436035,
+      "learning_rate": 1.2809413799035961e-05,
+      "loss": 0.48754948377609253,
+      "step": 801
+    },
+    {
+      "epoch": 0.43969298245614036,
+      "grad_norm": 0.47973889112472534,
+      "learning_rate": 1.2792001191692375e-05,
+      "loss": 0.47506245970726013,
+      "step": 802
+    },
+    {
+      "epoch": 0.44024122807017546,
+      "grad_norm": 0.4873831272125244,
+      "learning_rate": 1.2774579398496656e-05,
+      "loss": 0.48494696617126465,
+      "step": 803
+    },
+    {
+      "epoch": 0.4407894736842105,
+      "grad_norm": 0.5331482291221619,
+      "learning_rate": 1.2757148476767553e-05,
+      "loss": 0.49041277170181274,
+      "step": 804
+    },
+    {
+      "epoch": 0.4413377192982456,
+      "grad_norm": 0.4975897967815399,
+      "learning_rate": 1.2739708483853845e-05,
+      "loss": 0.49118441343307495,
+      "step": 805
+    },
+    {
+      "epoch": 0.4418859649122807,
+      "grad_norm": 0.49233025312423706,
+      "learning_rate": 1.2722259477134162e-05,
+      "loss": 0.4929482936859131,
+      "step": 806
+    },
+    {
+      "epoch": 0.44243421052631576,
+      "grad_norm": 0.531247079372406,
+      "learning_rate": 1.2704801514016789e-05,
+      "loss": 0.4968828856945038,
+      "step": 807
+    },
+    {
+      "epoch": 0.44298245614035087,
+      "grad_norm": 0.5535720586776733,
+      "learning_rate": 1.2687334651939471e-05,
+      "loss": 0.48034191131591797,
+      "step": 808
+    },
+    {
+      "epoch": 0.44353070175438597,
+      "grad_norm": 0.5043017268180847,
+      "learning_rate": 1.2669858948369242e-05,
+      "loss": 0.4804898500442505,
+      "step": 809
+    },
+    {
+      "epoch": 0.4440789473684211,
+      "grad_norm": 0.5200436115264893,
+      "learning_rate": 1.2652374460802218e-05,
+      "loss": 0.4860343933105469,
+      "step": 810
+    },
+    {
+      "epoch": 0.4446271929824561,
+      "grad_norm": 0.5313208103179932,
+      "learning_rate": 1.2634881246763415e-05,
+      "loss": 0.49038752913475037,
+      "step": 811
+    },
+    {
+      "epoch": 0.4451754385964912,
+      "grad_norm": 0.4981192946434021,
+      "learning_rate": 1.2617379363806563e-05,
+      "loss": 0.47683286666870117,
+      "step": 812
+    },
+    {
+      "epoch": 0.4457236842105263,
+      "grad_norm": 0.4823460280895233,
+      "learning_rate": 1.2599868869513912e-05,
+      "loss": 0.4903796911239624,
+      "step": 813
+    },
+    {
+      "epoch": 0.44627192982456143,
+      "grad_norm": 0.48653507232666016,
+      "learning_rate": 1.2582349821496041e-05,
+      "loss": 0.48963838815689087,
+      "step": 814
+    },
+    {
+      "epoch": 0.4468201754385965,
+      "grad_norm": 0.5077612400054932,
+      "learning_rate": 1.2564822277391673e-05,
+      "loss": 0.48124510049819946,
+      "step": 815
+    },
+    {
+      "epoch": 0.4473684210526316,
+      "grad_norm": 0.49369195103645325,
+      "learning_rate": 1.2547286294867486e-05,
+      "loss": 0.4819800555706024,
+      "step": 816
+    },
+    {
+      "epoch": 0.4479166666666667,
+      "grad_norm": 0.506385087966919,
+      "learning_rate": 1.252974193161792e-05,
+      "loss": 0.48602816462516785,
+      "step": 817
+    },
+    {
+      "epoch": 0.44846491228070173,
+      "grad_norm": 0.49783554673194885,
+      "learning_rate": 1.2512189245364986e-05,
+      "loss": 0.5015057921409607,
+      "step": 818
+    },
+    {
+      "epoch": 0.44901315789473684,
+      "grad_norm": 0.49489626288414,
+      "learning_rate": 1.2494628293858078e-05,
+      "loss": 0.4883100986480713,
+      "step": 819
+    },
+    {
+      "epoch": 0.44956140350877194,
+      "grad_norm": 0.48849615454673767,
+      "learning_rate": 1.2477059134873784e-05,
+      "loss": 0.4841647148132324,
+      "step": 820
+    },
+    {
+      "epoch": 0.45010964912280704,
+      "grad_norm": 0.4946068227291107,
+      "learning_rate": 1.2459481826215698e-05,
+      "loss": 0.47616252303123474,
+      "step": 821
+    },
+    {
+      "epoch": 0.4506578947368421,
+      "grad_norm": 0.4865962862968445,
+      "learning_rate": 1.2441896425714224e-05,
+      "loss": 0.4910244643688202,
+      "step": 822
+    },
+    {
+      "epoch": 0.4512061403508772,
+      "grad_norm": 0.47502800822257996,
+      "learning_rate": 1.242430299122639e-05,
+      "loss": 0.4838365912437439,
+      "step": 823
+    },
+    {
+      "epoch": 0.4517543859649123,
+      "grad_norm": 0.47981613874435425,
+      "learning_rate": 1.240670158063565e-05,
+      "loss": 0.4831865429878235,
+      "step": 824
+    },
+    {
+      "epoch": 0.45230263157894735,
+      "grad_norm": 0.531018853187561,
+      "learning_rate": 1.238909225185171e-05,
+      "loss": 0.4798121452331543,
+      "step": 825
+    },
+    {
+      "epoch": 0.45285087719298245,
+      "grad_norm": 0.4660738408565521,
+      "learning_rate": 1.2371475062810324e-05,
+      "loss": 0.47949931025505066,
+      "step": 826
+    },
+    {
+      "epoch": 0.45339912280701755,
+      "grad_norm": 0.48740774393081665,
+      "learning_rate": 1.2353850071473103e-05,
+      "loss": 0.4878680408000946,
+      "step": 827
+    },
+    {
+      "epoch": 0.45394736842105265,
+      "grad_norm": 0.49357178807258606,
+      "learning_rate": 1.2336217335827332e-05,
+      "loss": 0.4781498312950134,
+      "step": 828
+    },
+    {
+      "epoch": 0.4544956140350877,
+      "grad_norm": 0.5055526494979858,
+      "learning_rate": 1.231857691388577e-05,
+      "loss": 0.48239409923553467,
+      "step": 829
+    },
+    {
+      "epoch": 0.4550438596491228,
+      "grad_norm": 0.5338366031646729,
+      "learning_rate": 1.2300928863686472e-05,
+      "loss": 0.4846789240837097,
+      "step": 830
+    },
+    {
+      "epoch": 0.4555921052631579,
+      "grad_norm": 0.4883231818675995,
+      "learning_rate": 1.2283273243292583e-05,
+      "loss": 0.47418880462646484,
+      "step": 831
+    },
+    {
+      "epoch": 0.45614035087719296,
+      "grad_norm": 0.48572519421577454,
+      "learning_rate": 1.2265610110792161e-05,
+      "loss": 0.49370428919792175,
+      "step": 832
+    },
+    {
+      "epoch": 0.45668859649122806,
+      "grad_norm": 0.5001595616340637,
+      "learning_rate": 1.2247939524297977e-05,
+      "loss": 0.48919588327407837,
+      "step": 833
+    },
+    {
+      "epoch": 0.45723684210526316,
+      "grad_norm": 0.5046356320381165,
+      "learning_rate": 1.2230261541947316e-05,
+      "loss": 0.4786142110824585,
+      "step": 834
+    },
+    {
+      "epoch": 0.45778508771929827,
+      "grad_norm": 0.49759823083877563,
+      "learning_rate": 1.2212576221901819e-05,
+      "loss": 0.48940107226371765,
+      "step": 835
+    },
+    {
+      "epoch": 0.4583333333333333,
+      "grad_norm": 0.4834117293357849,
+      "learning_rate": 1.2194883622347247e-05,
+      "loss": 0.4996738135814667,
+      "step": 836
+    },
+    {
+      "epoch": 0.4588815789473684,
+      "grad_norm": 0.46439656615257263,
+      "learning_rate": 1.2177183801493313e-05,
+      "loss": 0.48260292410850525,
+      "step": 837
+    },
+    {
+      "epoch": 0.4594298245614035,
+      "grad_norm": 0.4949783384799957,
+      "learning_rate": 1.2159476817573506e-05,
+      "loss": 0.4749537706375122,
+      "step": 838
+    },
+    {
+      "epoch": 0.45997807017543857,
+      "grad_norm": 0.4927411377429962,
+      "learning_rate": 1.2141762728844862e-05,
+      "loss": 0.484489768743515,
+      "step": 839
+    },
+    {
+      "epoch": 0.4605263157894737,
+      "grad_norm": 0.4821980595588684,
+      "learning_rate": 1.21240415935878e-05,
+      "loss": 0.489577054977417,
+      "step": 840
+    },
+    {
+      "epoch": 0.4610745614035088,
+      "grad_norm": 0.47517091035842896,
+      "learning_rate": 1.2106313470105925e-05,
+      "loss": 0.47467929124832153,
+      "step": 841
+    },
+    {
+      "epoch": 0.4616228070175439,
+      "grad_norm": 0.5108216404914856,
+      "learning_rate": 1.2088578416725829e-05,
+      "loss": 0.4935423731803894,
+      "step": 842
+    },
+    {
+      "epoch": 0.4621710526315789,
+      "grad_norm": 0.5016485452651978,
+      "learning_rate": 1.207083649179691e-05,
+      "loss": 0.4873450696468353,
+      "step": 843
+    },
+    {
+      "epoch": 0.46271929824561403,
+      "grad_norm": 0.49393925070762634,
+      "learning_rate": 1.2053087753691173e-05,
+      "loss": 0.47483953833580017,
+      "step": 844
+    },
+    {
+      "epoch": 0.46326754385964913,
+      "grad_norm": 0.524784505367279,
+      "learning_rate": 1.2035332260803026e-05,
+      "loss": 0.4782135486602783,
+      "step": 845
+    },
+    {
+      "epoch": 0.46381578947368424,
+      "grad_norm": 0.5134442448616028,
+      "learning_rate": 1.2017570071549113e-05,
+      "loss": 0.4677678346633911,
+      "step": 846
+    },
+    {
+      "epoch": 0.4643640350877193,
+      "grad_norm": 0.5348238348960876,
+      "learning_rate": 1.1999801244368112e-05,
+      "loss": 0.47131556272506714,
+      "step": 847
+    },
+    {
+      "epoch": 0.4649122807017544,
+      "grad_norm": 0.47679224610328674,
+      "learning_rate": 1.1982025837720533e-05,
+      "loss": 0.4836603105068207,
+      "step": 848
+    },
+    {
+      "epoch": 0.4654605263157895,
+      "grad_norm": 0.509976863861084,
+      "learning_rate": 1.1964243910088532e-05,
+      "loss": 0.4757232666015625,
+      "step": 849
+    },
+    {
+      "epoch": 0.46600877192982454,
+      "grad_norm": 0.48536261916160583,
+      "learning_rate": 1.1946455519975724e-05,
+      "loss": 0.4884200692176819,
+      "step": 850
+    },
+    {
+      "epoch": 0.46655701754385964,
+      "grad_norm": 0.49048852920532227,
+      "learning_rate": 1.1928660725906984e-05,
+      "loss": 0.4773423671722412,
+      "step": 851
+    },
+    {
+      "epoch": 0.46710526315789475,
+      "grad_norm": 0.6477540135383606,
+      "learning_rate": 1.1910859586428258e-05,
+      "loss": 0.47236260771751404,
+      "step": 852
+    },
+    {
+      "epoch": 0.46765350877192985,
+      "grad_norm": 0.492621511220932,
+      "learning_rate": 1.1893052160106364e-05,
+      "loss": 0.47727149724960327,
+      "step": 853
+    },
+    {
+      "epoch": 0.4682017543859649,
+      "grad_norm": 0.49619951844215393,
+      "learning_rate": 1.187523850552881e-05,
+      "loss": 0.4871925115585327,
+      "step": 854
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 0.4711047112941742,
+      "learning_rate": 1.1857418681303592e-05,
+      "loss": 0.49179786443710327,
+      "step": 855
+    },
+    {
+      "epoch": 0.4692982456140351,
+      "grad_norm": 0.5045946836471558,
+      "learning_rate": 1.1839592746059008e-05,
+      "loss": 0.47911369800567627,
+      "step": 856
+    },
+    {
+      "epoch": 0.46984649122807015,
+      "grad_norm": 0.532071053981781,
+      "learning_rate": 1.1821760758443455e-05,
+      "loss": 0.48660939931869507,
+      "step": 857
+    },
+    {
+      "epoch": 0.47039473684210525,
+      "grad_norm": 0.49183908104896545,
+      "learning_rate": 1.1803922777125247e-05,
+      "loss": 0.49535202980041504,
+      "step": 858
+    },
+    {
+      "epoch": 0.47094298245614036,
+      "grad_norm": 0.475089967250824,
+      "learning_rate": 1.178607886079242e-05,
+      "loss": 0.48570260405540466,
+      "step": 859
+    },
+    {
+      "epoch": 0.47149122807017546,
+      "grad_norm": 0.4888152778148651,
+      "learning_rate": 1.1768229068152533e-05,
+      "loss": 0.4838455319404602,
+      "step": 860
+    },
+    {
+      "epoch": 0.4720394736842105,
+      "grad_norm": 0.47159093618392944,
+      "learning_rate": 1.1750373457932477e-05,
+      "loss": 0.47372183203697205,
+      "step": 861
+    },
+    {
+      "epoch": 0.4725877192982456,
+      "grad_norm": 0.5381128191947937,
+      "learning_rate": 1.1732512088878287e-05,
+      "loss": 0.4868219792842865,
+      "step": 862
+    },
+    {
+      "epoch": 0.4731359649122807,
+      "grad_norm": 0.47580257058143616,
+      "learning_rate": 1.1714645019754944e-05,
+      "loss": 0.4665176272392273,
+      "step": 863
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.5036393404006958,
+      "learning_rate": 1.1696772309346183e-05,
+      "loss": 0.4782337546348572,
+      "step": 864
+    },
+    {
+      "epoch": 0.47423245614035087,
+      "grad_norm": 0.5104615688323975,
+      "learning_rate": 1.1678894016454292e-05,
+      "loss": 0.47208207845687866,
+      "step": 865
+    },
+    {
+      "epoch": 0.47478070175438597,
+      "grad_norm": 0.509761393070221,
+      "learning_rate": 1.1661010199899938e-05,
+      "loss": 0.4760264754295349,
+      "step": 866
+    },
+    {
+      "epoch": 0.4753289473684211,
+      "grad_norm": 0.4815259277820587,
+      "learning_rate": 1.1643120918521952e-05,
+      "loss": 0.4828665256500244,
+      "step": 867
+    },
+    {
+      "epoch": 0.4758771929824561,
+      "grad_norm": 0.48106735944747925,
+      "learning_rate": 1.162522623117715e-05,
+      "loss": 0.47156989574432373,
+      "step": 868
+    },
+    {
+      "epoch": 0.4764254385964912,
+      "grad_norm": 0.5169423818588257,
+      "learning_rate": 1.160732619674013e-05,
+      "loss": 0.4661808907985687,
+      "step": 869
+    },
+    {
+      "epoch": 0.4769736842105263,
+      "grad_norm": 0.5183056592941284,
+      "learning_rate": 1.1589420874103081e-05,
+      "loss": 0.49258583784103394,
+      "step": 870
+    },
+    {
+      "epoch": 0.47752192982456143,
+      "grad_norm": 0.4648781418800354,
+      "learning_rate": 1.1571510322175598e-05,
+      "loss": 0.4734388291835785,
+      "step": 871
+    },
+    {
+      "epoch": 0.4780701754385965,
+      "grad_norm": 0.5181176662445068,
+      "learning_rate": 1.1553594599884472e-05,
+      "loss": 0.4744289517402649,
+      "step": 872
+    },
+    {
+      "epoch": 0.4786184210526316,
+      "grad_norm": 0.4918321967124939,
+      "learning_rate": 1.1535673766173512e-05,
+      "loss": 0.48278099298477173,
+      "step": 873
+    },
+    {
+      "epoch": 0.4791666666666667,
+      "grad_norm": 0.47864681482315063,
+      "learning_rate": 1.1517747880003336e-05,
+      "loss": 0.4869686961174011,
+      "step": 874
+    },
+    {
+      "epoch": 0.47971491228070173,
+      "grad_norm": 0.5176661610603333,
+      "learning_rate": 1.14998170003512e-05,
+      "loss": 0.4895298480987549,
+      "step": 875
+    },
+    {
+      "epoch": 0.48026315789473684,
+      "grad_norm": 0.5217961072921753,
+      "learning_rate": 1.1481881186210767e-05,
+      "loss": 0.4778291881084442,
+      "step": 876
+    },
+    {
+      "epoch": 0.48081140350877194,
+      "grad_norm": 0.4688926041126251,
+      "learning_rate": 1.1463940496591946e-05,
+      "loss": 0.48186585307121277,
+      "step": 877
+    },
+    {
+      "epoch": 0.48135964912280704,
+      "grad_norm": 0.5088805556297302,
+      "learning_rate": 1.1445994990520697e-05,
+      "loss": 0.485612154006958,
+      "step": 878
+    },
+    {
+      "epoch": 0.4819078947368421,
+      "grad_norm": 0.5077580809593201,
+      "learning_rate": 1.1428044727038807e-05,
+      "loss": 0.48806190490722656,
+      "step": 879
+    },
+    {
+      "epoch": 0.4824561403508772,
+      "grad_norm": 0.4916880428791046,
+      "learning_rate": 1.1410089765203724e-05,
+      "loss": 0.49692898988723755,
+      "step": 880
+    },
+    {
+      "epoch": 0.4830043859649123,
+      "grad_norm": 0.49453645944595337,
+      "learning_rate": 1.1392130164088364e-05,
+      "loss": 0.48662543296813965,
+      "step": 881
+    },
+    {
+      "epoch": 0.48355263157894735,
+      "grad_norm": 0.5179969072341919,
+      "learning_rate": 1.1374165982780886e-05,
+      "loss": 0.4894055724143982,
+      "step": 882
+    },
+    {
+      "epoch": 0.48410087719298245,
+      "grad_norm": 0.5080739855766296,
+      "learning_rate": 1.1356197280384536e-05,
+      "loss": 0.4840007424354553,
+      "step": 883
+    },
+    {
+      "epoch": 0.48464912280701755,
+      "grad_norm": 0.4707460403442383,
+      "learning_rate": 1.1338224116017425e-05,
+      "loss": 0.4837796092033386,
+      "step": 884
+    },
+    {
+      "epoch": 0.48519736842105265,
+      "grad_norm": 0.5029415488243103,
+      "learning_rate": 1.1320246548812343e-05,
+      "loss": 0.4835893511772156,
+      "step": 885
+    },
+    {
+      "epoch": 0.4857456140350877,
+      "grad_norm": 0.5508713126182556,
+      "learning_rate": 1.1302264637916576e-05,
+      "loss": 0.49281996488571167,
+      "step": 886
+    },
+    {
+      "epoch": 0.4862938596491228,
+      "grad_norm": 0.6094909310340881,
+      "learning_rate": 1.128427844249169e-05,
+      "loss": 0.483834832906723,
+      "step": 887
+    },
+    {
+      "epoch": 0.4868421052631579,
+      "grad_norm": 0.45924264192581177,
+      "learning_rate": 1.1266288021713347e-05,
+      "loss": 0.4871005415916443,
+      "step": 888
+    },
+    {
+      "epoch": 0.48739035087719296,
+      "grad_norm": 0.4807246923446655,
+      "learning_rate": 1.1248293434771127e-05,
+      "loss": 0.47325021028518677,
+      "step": 889
+    },
+    {
+      "epoch": 0.48793859649122806,
+      "grad_norm": 0.5099015831947327,
+      "learning_rate": 1.1230294740868294e-05,
+      "loss": 0.4735136032104492,
+      "step": 890
+    },
+    {
+      "epoch": 0.48848684210526316,
+      "grad_norm": 0.4942789673805237,
+      "learning_rate": 1.1212291999221644e-05,
+      "loss": 0.4851912260055542,
+      "step": 891
+    },
+    {
+      "epoch": 0.48903508771929827,
+      "grad_norm": 0.5109583735466003,
+      "learning_rate": 1.1194285269061277e-05,
+      "loss": 0.4773210287094116,
+      "step": 892
+    },
+    {
+      "epoch": 0.4895833333333333,
+      "grad_norm": 0.4867546856403351,
+      "learning_rate": 1.1176274609630418e-05,
+      "loss": 0.48137515783309937,
+      "step": 893
+    },
+    {
+      "epoch": 0.4901315789473684,
+      "grad_norm": 0.48830315470695496,
+      "learning_rate": 1.1158260080185226e-05,
+      "loss": 0.4856005609035492,
+      "step": 894
+    },
+    {
+      "epoch": 0.4906798245614035,
+      "grad_norm": 0.46730145812034607,
+      "learning_rate": 1.1140241739994589e-05,
+      "loss": 0.46794044971466064,
+      "step": 895
+    },
+    {
+      "epoch": 0.49122807017543857,
+      "grad_norm": 0.484876811504364,
+      "learning_rate": 1.1122219648339925e-05,
+      "loss": 0.47470468282699585,
+      "step": 896
+    },
+    {
+      "epoch": 0.4917763157894737,
+      "grad_norm": 0.48526665568351746,
+      "learning_rate": 1.110419386451501e-05,
+      "loss": 0.46910759806632996,
+      "step": 897
+    },
+    {
+      "epoch": 0.4923245614035088,
+      "grad_norm": 0.48200756311416626,
+      "learning_rate": 1.108616444782575e-05,
+      "loss": 0.4611569344997406,
+      "step": 898
+    },
+    {
+      "epoch": 0.4928728070175439,
+      "grad_norm": 0.4781787693500519,
+      "learning_rate": 1.1068131457590022e-05,
+      "loss": 0.4725024104118347,
+      "step": 899
+    },
+    {
+      "epoch": 0.4934210526315789,
+      "grad_norm": 0.4791221618652344,
+      "learning_rate": 1.1050094953137444e-05,
+      "loss": 0.4657338857650757,
+      "step": 900
+    },
+    {
+      "epoch": 0.49396929824561403,
+      "grad_norm": 0.4847608804702759,
+      "learning_rate": 1.10320549938092e-05,
+      "loss": 0.470599889755249,
+      "step": 901
+    },
+    {
+      "epoch": 0.49451754385964913,
+      "grad_norm": 0.49423038959503174,
+      "learning_rate": 1.1014011638957849e-05,
+      "loss": 0.47278890013694763,
+      "step": 902
+    },
+    {
+      "epoch": 0.49506578947368424,
+      "grad_norm": 0.49379685521125793,
+      "learning_rate": 1.0995964947947114e-05,
+      "loss": 0.48167523741722107,
+      "step": 903
+    },
+    {
+      "epoch": 0.4956140350877193,
+      "grad_norm": 0.48116910457611084,
+      "learning_rate": 1.0977914980151692e-05,
+      "loss": 0.48413026332855225,
+      "step": 904
+    },
+    {
+      "epoch": 0.4961622807017544,
+      "grad_norm": 0.47256386280059814,
+      "learning_rate": 1.095986179495707e-05,
+      "loss": 0.4625633656978607,
+      "step": 905
+    },
+    {
+      "epoch": 0.4967105263157895,
+      "grad_norm": 0.45926183462142944,
+      "learning_rate": 1.0941805451759311e-05,
+      "loss": 0.47543013095855713,
+      "step": 906
+    },
+    {
+      "epoch": 0.49725877192982454,
+      "grad_norm": 0.46669498085975647,
+      "learning_rate": 1.0923746009964873e-05,
+      "loss": 0.4802945554256439,
+      "step": 907
+    },
+    {
+      "epoch": 0.49780701754385964,
+      "grad_norm": 0.4964379072189331,
+      "learning_rate": 1.0905683528990407e-05,
+      "loss": 0.47815272212028503,
+      "step": 908
+    },
+    {
+      "epoch": 0.49835526315789475,
+      "grad_norm": 0.48577016592025757,
+      "learning_rate": 1.0887618068262566e-05,
+      "loss": 0.4769788980484009,
+      "step": 909
+    },
+    {
+      "epoch": 0.49890350877192985,
+      "grad_norm": 0.48084649443626404,
+      "learning_rate": 1.0869549687217802e-05,
+      "loss": 0.47855353355407715,
+      "step": 910
+    },
+    {
+      "epoch": 0.4994517543859649,
+      "grad_norm": 0.5205073952674866,
+      "learning_rate": 1.0851478445302179e-05,
+      "loss": 0.4734543263912201,
+      "step": 911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47468581795692444,
+      "learning_rate": 1.083340440197117e-05,
+      "loss": 0.4848591089248657,
+      "step": 912
+    },
+    {
+      "epoch": 0.5005482456140351,
+      "grad_norm": 0.4783199727535248,
+      "learning_rate": 1.0815327616689468e-05,
+      "loss": 0.4678228795528412,
+      "step": 913
+    },
+    {
+      "epoch": 0.5010964912280702,
+      "grad_norm": 0.49609270691871643,
+      "learning_rate": 1.0797248148930783e-05,
+      "loss": 0.47647619247436523,
+      "step": 914
+    },
+    {
+      "epoch": 0.5016447368421053,
+      "grad_norm": 0.4896128475666046,
+      "learning_rate": 1.0779166058177657e-05,
+      "loss": 0.485650897026062,
+      "step": 915
+    },
+    {
+      "epoch": 0.5021929824561403,
+      "grad_norm": 0.47610363364219666,
+      "learning_rate": 1.0761081403921255e-05,
+      "loss": 0.4849432110786438,
+      "step": 916
+    },
+    {
+      "epoch": 0.5027412280701754,
+      "grad_norm": 0.4984195828437805,
+      "learning_rate": 1.0742994245661178e-05,
+      "loss": 0.47665756940841675,
+      "step": 917
+    },
+    {
+      "epoch": 0.5032894736842105,
+      "grad_norm": 0.5013912916183472,
+      "learning_rate": 1.0724904642905272e-05,
+      "loss": 0.46333345770835876,
+      "step": 918
+    },
+    {
+      "epoch": 0.5038377192982456,
+      "grad_norm": 0.5182805061340332,
+      "learning_rate": 1.0706812655169414e-05,
+      "loss": 0.4818127155303955,
+      "step": 919
+    },
+    {
+      "epoch": 0.5043859649122807,
+      "grad_norm": 0.5016067028045654,
+      "learning_rate": 1.0688718341977336e-05,
+      "loss": 0.4823892116546631,
+      "step": 920
+    },
+    {
+      "epoch": 0.5049342105263158,
+      "grad_norm": 0.4810698330402374,
+      "learning_rate": 1.0670621762860419e-05,
+      "loss": 0.4859499931335449,
+      "step": 921
+    },
+    {
+      "epoch": 0.5054824561403509,
+      "grad_norm": 0.5582836866378784,
+      "learning_rate": 1.0652522977357498e-05,
+      "loss": 0.47283801436424255,
+      "step": 922
+    },
+    {
+      "epoch": 0.5060307017543859,
+      "grad_norm": 0.48066940903663635,
+      "learning_rate": 1.063442204501467e-05,
+      "loss": 0.4813811182975769,
+      "step": 923
+    },
+    {
+      "epoch": 0.506578947368421,
+      "grad_norm": 0.47318175435066223,
+      "learning_rate": 1.0616319025385088e-05,
+      "loss": 0.47437286376953125,
+      "step": 924
+    },
+    {
+      "epoch": 0.5071271929824561,
+      "grad_norm": 0.527256429195404,
+      "learning_rate": 1.059821397802878e-05,
+      "loss": 0.5002198219299316,
+      "step": 925
+    },
+    {
+      "epoch": 0.5076754385964912,
+      "grad_norm": 0.49084538221359253,
+      "learning_rate": 1.0580106962512446e-05,
+      "loss": 0.4681828022003174,
+      "step": 926
+    },
+    {
+      "epoch": 0.5082236842105263,
+      "grad_norm": 0.49694955348968506,
+      "learning_rate": 1.0561998038409252e-05,
+      "loss": 0.4772576689720154,
+      "step": 927
+    },
+    {
+      "epoch": 0.5087719298245614,
+      "grad_norm": 0.49907466769218445,
+      "learning_rate": 1.0543887265298651e-05,
+      "loss": 0.4743449091911316,
+      "step": 928
+    },
+    {
+      "epoch": 0.5093201754385965,
+      "grad_norm": 0.47016963362693787,
+      "learning_rate": 1.0525774702766183e-05,
+      "loss": 0.4842929244041443,
+      "step": 929
+    },
+    {
+      "epoch": 0.5098684210526315,
+      "grad_norm": 0.48708009719848633,
+      "learning_rate": 1.050766041040326e-05,
+      "loss": 0.49047666788101196,
+      "step": 930
+    },
+    {
+      "epoch": 0.5104166666666666,
+      "grad_norm": 0.48263296484947205,
+      "learning_rate": 1.0489544447807006e-05,
+      "loss": 0.4745303988456726,
+      "step": 931
+    },
+    {
+      "epoch": 0.5109649122807017,
+      "grad_norm": 0.4735895097255707,
+      "learning_rate": 1.0471426874580025e-05,
+      "loss": 0.47902441024780273,
+      "step": 932
+    },
+    {
+      "epoch": 0.5115131578947368,
+      "grad_norm": 0.477176308631897,
+      "learning_rate": 1.0453307750330217e-05,
+      "loss": 0.45984840393066406,
+      "step": 933
+    },
+    {
+      "epoch": 0.5120614035087719,
+      "grad_norm": 0.5000417828559875,
+      "learning_rate": 1.0435187134670608e-05,
+      "loss": 0.4723761975765228,
+      "step": 934
+    },
+    {
+      "epoch": 0.512609649122807,
+      "grad_norm": 0.4939057230949402,
+      "learning_rate": 1.0417065087219101e-05,
+      "loss": 0.4698931574821472,
+      "step": 935
+    },
+    {
+      "epoch": 0.5131578947368421,
+      "grad_norm": 0.4751085937023163,
+      "learning_rate": 1.0398941667598328e-05,
+      "loss": 0.47515150904655457,
+      "step": 936
+    },
+    {
+      "epoch": 0.5137061403508771,
+      "grad_norm": 0.5196514129638672,
+      "learning_rate": 1.0380816935435438e-05,
+      "loss": 0.4751243591308594,
+      "step": 937
+    },
+    {
+      "epoch": 0.5142543859649122,
+      "grad_norm": 0.46978095173835754,
+      "learning_rate": 1.0362690950361882e-05,
+      "loss": 0.4780597686767578,
+      "step": 938
+    },
+    {
+      "epoch": 0.5148026315789473,
+      "grad_norm": 0.4803449511528015,
+      "learning_rate": 1.034456377201325e-05,
+      "loss": 0.472648948431015,
+      "step": 939
+    },
+    {
+      "epoch": 0.5153508771929824,
+      "grad_norm": 0.47589024901390076,
+      "learning_rate": 1.0326435460029051e-05,
+      "loss": 0.4832885265350342,
+      "step": 940
+    },
+    {
+      "epoch": 0.5158991228070176,
+      "grad_norm": 0.47134554386138916,
+      "learning_rate": 1.0308306074052517e-05,
+      "loss": 0.47504305839538574,
+      "step": 941
+    },
+    {
+      "epoch": 0.5164473684210527,
+      "grad_norm": 0.4921835958957672,
+      "learning_rate": 1.0290175673730426e-05,
+      "loss": 0.4666009843349457,
+      "step": 942
+    },
+    {
+      "epoch": 0.5169956140350878,
+      "grad_norm": 0.4763789772987366,
+      "learning_rate": 1.0272044318712888e-05,
+      "loss": 0.4826936721801758,
+      "step": 943
+    },
+    {
+      "epoch": 0.5175438596491229,
+      "grad_norm": 0.4789031744003296,
+      "learning_rate": 1.0253912068653147e-05,
+      "loss": 0.45768702030181885,
+      "step": 944
+    },
+    {
+      "epoch": 0.5180921052631579,
+      "grad_norm": 0.5070001482963562,
+      "learning_rate": 1.0235778983207404e-05,
+      "loss": 0.45878398418426514,
+      "step": 945
+    },
+    {
+      "epoch": 0.518640350877193,
+      "grad_norm": 0.4759475588798523,
+      "learning_rate": 1.0217645122034603e-05,
+      "loss": 0.46790462732315063,
+      "step": 946
+    },
+    {
+      "epoch": 0.5191885964912281,
+      "grad_norm": 0.507018506526947,
+      "learning_rate": 1.0199510544796237e-05,
+      "loss": 0.4877185523509979,
+      "step": 947
+    },
+    {
+      "epoch": 0.5197368421052632,
+      "grad_norm": 0.4932757616043091,
+      "learning_rate": 1.0181375311156156e-05,
+      "loss": 0.4681272506713867,
+      "step": 948
+    },
+    {
+      "epoch": 0.5202850877192983,
+      "grad_norm": 0.5080598592758179,
+      "learning_rate": 1.0163239480780376e-05,
+      "loss": 0.47993144392967224,
+      "step": 949
+    },
+    {
+      "epoch": 0.5208333333333334,
+      "grad_norm": 0.4807532727718353,
+      "learning_rate": 1.0145103113336868e-05,
+      "loss": 0.46426552534103394,
+      "step": 950
+    },
+    {
+      "epoch": 0.5213815789473685,
+      "grad_norm": 0.4675240218639374,
+      "learning_rate": 1.0126966268495377e-05,
+      "loss": 0.4759901762008667,
+      "step": 951
+    },
+    {
+      "epoch": 0.5219298245614035,
+      "grad_norm": 0.47961559891700745,
+      "learning_rate": 1.010882900592721e-05,
+      "loss": 0.46679529547691345,
+      "step": 952
+    },
+    {
+      "epoch": 0.5224780701754386,
+      "grad_norm": 0.4580441117286682,
+      "learning_rate": 1.0090691385305057e-05,
+      "loss": 0.4677969217300415,
+      "step": 953
+    },
+    {
+      "epoch": 0.5230263157894737,
+      "grad_norm": 0.48743894696235657,
+      "learning_rate": 1.0072553466302783e-05,
+      "loss": 0.45729967951774597,
+      "step": 954
+    },
+    {
+      "epoch": 0.5235745614035088,
+      "grad_norm": 0.480928510427475,
+      "learning_rate": 1.0054415308595235e-05,
+      "loss": 0.46743547916412354,
+      "step": 955
+    },
+    {
+      "epoch": 0.5241228070175439,
+      "grad_norm": 0.4802962839603424,
+      "learning_rate": 1.0036276971858043e-05,
+      "loss": 0.4764847755432129,
+      "step": 956
+    },
+    {
+      "epoch": 0.524671052631579,
+      "grad_norm": 0.4960572123527527,
+      "learning_rate": 1.0018138515767425e-05,
+      "loss": 0.4603794515132904,
+      "step": 957
+    },
+    {
+      "epoch": 0.5252192982456141,
+      "grad_norm": 0.5283850431442261,
+      "learning_rate": 1e-05,
+      "loss": 0.46352872252464294,
+      "step": 958
+    },
+    {
+      "epoch": 0.5257675438596491,
+      "grad_norm": 0.5403750538825989,
+      "learning_rate": 9.981861484232576e-06,
+      "loss": 0.4582330584526062,
+      "step": 959
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.4815463721752167,
+      "learning_rate": 9.963723028141959e-06,
+      "loss": 0.4711856544017792,
+      "step": 960
+    },
+    {
+      "epoch": 0.5268640350877193,
+      "grad_norm": 0.4983313977718353,
+      "learning_rate": 9.945584691404768e-06,
+      "loss": 0.4663437306880951,
+      "step": 961
+    },
+    {
+      "epoch": 0.5274122807017544,
+      "grad_norm": 0.4963984787464142,
+      "learning_rate": 9.927446533697218e-06,
+      "loss": 0.4564981460571289,
+      "step": 962
+    },
+    {
+      "epoch": 0.5279605263157895,
+      "grad_norm": 0.4613352119922638,
+      "learning_rate": 9.909308614694944e-06,
+      "loss": 0.4883826673030853,
+      "step": 963
+    },
+    {
+      "epoch": 0.5285087719298246,
+      "grad_norm": 0.47054144740104675,
+      "learning_rate": 9.891170994072792e-06,
+      "loss": 0.4734528660774231,
+      "step": 964
+    },
+    {
+      "epoch": 0.5290570175438597,
+      "grad_norm": 0.49731600284576416,
+      "learning_rate": 9.873033731504628e-06,
+      "loss": 0.4818234443664551,
+      "step": 965
+    },
+    {
+      "epoch": 0.5296052631578947,
+      "grad_norm": 0.4624738395214081,
+      "learning_rate": 9.854896886663134e-06,
+      "loss": 0.47068604826927185,
+      "step": 966
+    },
+    {
+      "epoch": 0.5301535087719298,
+      "grad_norm": 0.4743126630783081,
+      "learning_rate": 9.836760519219627e-06,
+      "loss": 0.48319944739341736,
+      "step": 967
+    },
+    {
+      "epoch": 0.5307017543859649,
+      "grad_norm": 0.4714062511920929,
+      "learning_rate": 9.818624688843846e-06,
+      "loss": 0.4893440902233124,
+      "step": 968
+    },
+    {
+      "epoch": 0.53125,
+      "grad_norm": 0.5396475791931152,
+      "learning_rate": 9.800489455203765e-06,
+      "loss": 0.47514447569847107,
+      "step": 969
+    },
+    {
+      "epoch": 0.5317982456140351,
+      "grad_norm": 0.46770763397216797,
+      "learning_rate": 9.7823548779654e-06,
+      "loss": 0.47769927978515625,
+      "step": 970
+    },
+    {
+      "epoch": 0.5323464912280702,
+      "grad_norm": 0.5117153525352478,
+      "learning_rate": 9.764221016792597e-06,
+      "loss": 0.46777576208114624,
+      "step": 971
+    },
+    {
+      "epoch": 0.5328947368421053,
+      "grad_norm": 0.4876892864704132,
+      "learning_rate": 9.746087931346853e-06,
+      "loss": 0.4674271047115326,
+      "step": 972
+    },
+    {
+      "epoch": 0.5334429824561403,
+      "grad_norm": 0.5015421509742737,
+      "learning_rate": 9.727955681287115e-06,
+      "loss": 0.4751541018486023,
+      "step": 973
+    },
+    {
+      "epoch": 0.5339912280701754,
+      "grad_norm": 0.5426156520843506,
+      "learning_rate": 9.709824326269577e-06,
+      "loss": 0.4698949456214905,
+      "step": 974
+    },
+    {
+      "epoch": 0.5345394736842105,
+      "grad_norm": 0.5053887963294983,
+      "learning_rate": 9.691693925947485e-06,
+      "loss": 0.4778580069541931,
+      "step": 975
+    },
+    {
+      "epoch": 0.5350877192982456,
+      "grad_norm": 0.476938933134079,
+      "learning_rate": 9.67356453997095e-06,
+      "loss": 0.47415146231651306,
+      "step": 976
+    },
+    {
+      "epoch": 0.5356359649122807,
+      "grad_norm": 0.4961269795894623,
+      "learning_rate": 9.655436227986752e-06,
+      "loss": 0.4764622449874878,
+      "step": 977
+    },
+    {
+      "epoch": 0.5361842105263158,
+      "grad_norm": 0.4789068400859833,
+      "learning_rate": 9.637309049638121e-06,
+      "loss": 0.4756825864315033,
+      "step": 978
+    },
+    {
+      "epoch": 0.5367324561403509,
+      "grad_norm": 0.49088576436042786,
+      "learning_rate": 9.619183064564564e-06,
+      "loss": 0.46228572726249695,
+      "step": 979
+    },
+    {
+      "epoch": 0.5372807017543859,
+      "grad_norm": 0.49654847383499146,
+      "learning_rate": 9.601058332401674e-06,
+      "loss": 0.4739253520965576,
+      "step": 980
+    },
+    {
+      "epoch": 0.537828947368421,
+      "grad_norm": 0.47030875086784363,
+      "learning_rate": 9.582934912780904e-06,
+      "loss": 0.48689985275268555,
+      "step": 981
+    },
+    {
+      "epoch": 0.5383771929824561,
+      "grad_norm": 0.4943940043449402,
+      "learning_rate": 9.564812865329396e-06,
+      "loss": 0.4729098677635193,
+      "step": 982
+    },
+    {
+      "epoch": 0.5389254385964912,
+      "grad_norm": 0.4732706546783447,
+      "learning_rate": 9.546692249669784e-06,
+      "loss": 0.4672876298427582,
+      "step": 983
+    },
+    {
+      "epoch": 0.5394736842105263,
+      "grad_norm": 0.495355486869812,
+      "learning_rate": 9.52857312541998e-06,
+      "loss": 0.4782235026359558,
+      "step": 984
+    },
+    {
+      "epoch": 0.5400219298245614,
+      "grad_norm": 0.4918850064277649,
+      "learning_rate": 9.510455552192996e-06,
+      "loss": 0.48038250207901,
+      "step": 985
+    },
+    {
+      "epoch": 0.5405701754385965,
+      "grad_norm": 0.484025239944458,
+      "learning_rate": 9.492339589596742e-06,
+      "loss": 0.4546385705471039,
+      "step": 986
+    },
+    {
+      "epoch": 0.5411184210526315,
+      "grad_norm": 0.5004245638847351,
+      "learning_rate": 9.474225297233822e-06,
+      "loss": 0.4660744071006775,
+      "step": 987
+    },
+    {
+      "epoch": 0.5416666666666666,
+      "grad_norm": 0.4947008788585663,
+      "learning_rate": 9.456112734701349e-06,
+      "loss": 0.4723333716392517,
+      "step": 988
+    },
+    {
+      "epoch": 0.5422149122807017,
+      "grad_norm": 0.4967457056045532,
+      "learning_rate": 9.43800196159075e-06,
+      "loss": 0.4838150441646576,
+      "step": 989
+    },
+    {
+      "epoch": 0.5427631578947368,
+      "grad_norm": 0.5091995596885681,
+      "learning_rate": 9.41989303748756e-06,
+      "loss": 0.4611927270889282,
+      "step": 990
+    },
+    {
+      "epoch": 0.5433114035087719,
+      "grad_norm": 0.5165793895721436,
+      "learning_rate": 9.401786021971222e-06,
+      "loss": 0.4575102925300598,
+      "step": 991
+    },
+    {
+      "epoch": 0.543859649122807,
+      "grad_norm": 0.4805612862110138,
+      "learning_rate": 9.383680974614915e-06,
+      "loss": 0.4832637310028076,
+      "step": 992
+    },
+    {
+      "epoch": 0.5444078947368421,
+      "grad_norm": 0.5379158854484558,
+      "learning_rate": 9.365577954985336e-06,
+      "loss": 0.4655495584011078,
+      "step": 993
+    },
+    {
+      "epoch": 0.5449561403508771,
+      "grad_norm": 0.4972173571586609,
+      "learning_rate": 9.347477022642504e-06,
+      "loss": 0.4600108861923218,
+      "step": 994
+    },
+    {
+      "epoch": 0.5455043859649122,
+      "grad_norm": 0.48388564586639404,
+      "learning_rate": 9.329378237139583e-06,
+      "loss": 0.477608859539032,
+      "step": 995
+    },
+    {
+      "epoch": 0.5460526315789473,
+      "grad_norm": 0.4924751818180084,
+      "learning_rate": 9.31128165802267e-06,
+      "loss": 0.4669070243835449,
+      "step": 996
+    },
+    {
+      "epoch": 0.5466008771929824,
+      "grad_norm": 0.5232599377632141,
+      "learning_rate": 9.29318734483059e-06,
+      "loss": 0.482491672039032,
+      "step": 997
+    },
+    {
+      "epoch": 0.5471491228070176,
+      "grad_norm": 0.4885530173778534,
+      "learning_rate": 9.275095357094732e-06,
+      "loss": 0.483872652053833,
+      "step": 998
+    },
+    {
+      "epoch": 0.5476973684210527,
+      "grad_norm": 0.4905492961406708,
+      "learning_rate": 9.257005754338827e-06,
+      "loss": 0.470367968082428,
+      "step": 999
+    },
+    {
+      "epoch": 0.5482456140350878,
+      "grad_norm": 0.47469961643218994,
+      "learning_rate": 9.238918596078747e-06,
+      "loss": 0.4565841853618622,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5487938596491229,
+      "grad_norm": 0.5193788409233093,
+      "learning_rate": 9.220833941822346e-06,
+      "loss": 0.47289925813674927,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5493421052631579,
+      "grad_norm": 0.4825241267681122,
+      "learning_rate": 9.202751851069219e-06,
+      "loss": 0.4673827588558197,
+      "step": 1002
+    },
+    {
+      "epoch": 0.549890350877193,
+      "grad_norm": 0.4646058678627014,
+      "learning_rate": 9.184672383310534e-06,
+      "loss": 0.4733600616455078,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5504385964912281,
+      "grad_norm": 0.4764874577522278,
+      "learning_rate": 9.166595598028833e-06,
+      "loss": 0.47691139578819275,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5509868421052632,
+      "grad_norm": 0.5994452834129333,
+      "learning_rate": 9.148521554697823e-06,
+      "loss": 0.47110623121261597,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5515350877192983,
+      "grad_norm": 0.49562346935272217,
+      "learning_rate": 9.130450312782198e-06,
+      "loss": 0.44304126501083374,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5520833333333334,
+      "grad_norm": 0.49325883388519287,
+      "learning_rate": 9.112381931737437e-06,
+      "loss": 0.46718454360961914,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5526315789473685,
+      "grad_norm": 0.48549625277519226,
+      "learning_rate": 9.094316471009595e-06,
+      "loss": 0.470458984375,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5531798245614035,
+      "grad_norm": 0.5088659524917603,
+      "learning_rate": 9.076253990035132e-06,
+      "loss": 0.4785817265510559,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5537280701754386,
+      "grad_norm": 0.4731486439704895,
+      "learning_rate": 9.058194548240694e-06,
+      "loss": 0.4724987745285034,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5542763157894737,
+      "grad_norm": 0.4680763781070709,
+      "learning_rate": 9.040138205042935e-06,
+      "loss": 0.4792757034301758,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5548245614035088,
+      "grad_norm": 0.4451364576816559,
+      "learning_rate": 9.022085019848313e-06,
+      "loss": 0.47036516666412354,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5553728070175439,
+      "grad_norm": 0.508719801902771,
+      "learning_rate": 9.004035052052891e-06,
+      "loss": 0.4573447108268738,
+      "step": 1013
+    },
+    {
+      "epoch": 0.555921052631579,
+      "grad_norm": 0.5096684098243713,
+      "learning_rate": 8.985988361042153e-06,
+      "loss": 0.4640441834926605,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5564692982456141,
+      "grad_norm": 0.47320401668548584,
+      "learning_rate": 8.967945006190804e-06,
+      "loss": 0.45948725938796997,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5570175438596491,
+      "grad_norm": 0.4911516606807709,
+      "learning_rate": 8.94990504686256e-06,
+      "loss": 0.4813142716884613,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5575657894736842,
+      "grad_norm": 0.5072831511497498,
+      "learning_rate": 8.931868542409983e-06,
+      "loss": 0.470254510641098,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5581140350877193,
+      "grad_norm": 0.5038588047027588,
+      "learning_rate": 8.913835552174251e-06,
+      "loss": 0.4722591042518616,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5586622807017544,
+      "grad_norm": 0.45923951268196106,
+      "learning_rate": 8.895806135484992e-06,
+      "loss": 0.4707237184047699,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5592105263157895,
+      "grad_norm": 0.48106175661087036,
+      "learning_rate": 8.877780351660079e-06,
+      "loss": 0.46508774161338806,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5597587719298246,
+      "grad_norm": 0.5078846216201782,
+      "learning_rate": 8.859758260005416e-06,
+      "loss": 0.4740423858165741,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5603070175438597,
+      "grad_norm": 0.440356969833374,
+      "learning_rate": 8.841739919814775e-06,
+      "loss": 0.4663727879524231,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5608552631578947,
+      "grad_norm": 0.4811111092567444,
+      "learning_rate": 8.823725390369586e-06,
+      "loss": 0.45025596022605896,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5614035087719298,
+      "grad_norm": 0.5155603289604187,
+      "learning_rate": 8.805714730938728e-06,
+      "loss": 0.47258269786834717,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5619517543859649,
+      "grad_norm": 0.5031796097755432,
+      "learning_rate": 8.787708000778358e-06,
+      "loss": 0.45634135603904724,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.46268266439437866,
+      "learning_rate": 8.769705259131707e-06,
+      "loss": 0.46244633197784424,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5630482456140351,
+      "grad_norm": 0.4950982928276062,
+      "learning_rate": 8.751706565228878e-06,
+      "loss": 0.4840429723262787,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5635964912280702,
+      "grad_norm": 0.5139880180358887,
+      "learning_rate": 8.733711978286653e-06,
+      "loss": 0.4671246409416199,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5641447368421053,
+      "grad_norm": 0.48547184467315674,
+      "learning_rate": 8.715721557508315e-06,
+      "loss": 0.4675450325012207,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5646929824561403,
+      "grad_norm": 0.4884546101093292,
+      "learning_rate": 8.697735362083428e-06,
+      "loss": 0.46358034014701843,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5652412280701754,
+      "grad_norm": 0.4742810130119324,
+      "learning_rate": 8.679753451187659e-06,
+      "loss": 0.47714877128601074,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5657894736842105,
+      "grad_norm": 0.5046728253364563,
+      "learning_rate": 8.661775883982577e-06,
+      "loss": 0.4886934161186218,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5663377192982456,
+      "grad_norm": 0.4937182664871216,
+      "learning_rate": 8.643802719615467e-06,
+      "loss": 0.4588352143764496,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5668859649122807,
+      "grad_norm": 0.4708634316921234,
+      "learning_rate": 8.625834017219114e-06,
+      "loss": 0.4672088027000427,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5674342105263158,
+      "grad_norm": 0.5143368244171143,
+      "learning_rate": 8.607869835911638e-06,
+      "loss": 0.45802921056747437,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5679824561403509,
+      "grad_norm": 0.4756060540676117,
+      "learning_rate": 8.589910234796277e-06,
+      "loss": 0.46481484174728394,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5685307017543859,
+      "grad_norm": 0.464942991733551,
+      "learning_rate": 8.571955272961196e-06,
+      "loss": 0.45955371856689453,
+      "step": 1037
+    },
+    {
+      "epoch": 0.569078947368421,
+      "grad_norm": 0.4996425211429596,
+      "learning_rate": 8.554005009479307e-06,
+      "loss": 0.46826648712158203,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5696271929824561,
+      "grad_norm": 0.458551824092865,
+      "learning_rate": 8.536059503408058e-06,
+      "loss": 0.4795677065849304,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5701754385964912,
+      "grad_norm": 0.48602548241615295,
+      "learning_rate": 8.518118813789238e-06,
+      "loss": 0.4701020419597626,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5707236842105263,
+      "grad_norm": 0.49667230248451233,
+      "learning_rate": 8.500182999648804e-06,
+      "loss": 0.47075098752975464,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5712719298245614,
+      "grad_norm": 0.46576860547065735,
+      "learning_rate": 8.482252119996666e-06,
+      "loss": 0.47423285245895386,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5718201754385965,
+      "grad_norm": 0.4754815697669983,
+      "learning_rate": 8.464326233826492e-06,
+      "loss": 0.46078214049339294,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5723684210526315,
+      "grad_norm": 0.47265875339508057,
+      "learning_rate": 8.44640540011553e-06,
+      "loss": 0.461892306804657,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5729166666666666,
+      "grad_norm": 0.499237060546875,
+      "learning_rate": 8.428489677824405e-06,
+      "loss": 0.4715636372566223,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5734649122807017,
+      "grad_norm": 0.5131467580795288,
+      "learning_rate": 8.410579125896924e-06,
+      "loss": 0.46874022483825684,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5740131578947368,
+      "grad_norm": 0.46456998586654663,
+      "learning_rate": 8.392673803259875e-06,
+      "loss": 0.4896584451198578,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5745614035087719,
+      "grad_norm": 0.4798382520675659,
+      "learning_rate": 8.374773768822852e-06,
+      "loss": 0.46989163756370544,
+      "step": 1048
+    },
+    {
+      "epoch": 0.575109649122807,
+      "grad_norm": 0.49859294295310974,
+      "learning_rate": 8.356879081478051e-06,
+      "loss": 0.46852609515190125,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5756578947368421,
+      "grad_norm": 0.4816218614578247,
+      "learning_rate": 8.338989800100065e-06,
+      "loss": 0.4753386378288269,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5762061403508771,
+      "grad_norm": 0.4893586039543152,
+      "learning_rate": 8.32110598354571e-06,
+      "loss": 0.46335873007774353,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5767543859649122,
+      "grad_norm": 0.4901929795742035,
+      "learning_rate": 8.303227690653824e-06,
+      "loss": 0.4654930830001831,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5773026315789473,
+      "grad_norm": 0.621925950050354,
+      "learning_rate": 8.285354980245058e-06,
+      "loss": 0.4590475559234619,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5778508771929824,
+      "grad_norm": 0.47231367230415344,
+      "learning_rate": 8.267487911121714e-06,
+      "loss": 0.4794731140136719,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5783991228070176,
+      "grad_norm": 0.47593390941619873,
+      "learning_rate": 8.249626542067528e-06,
+      "loss": 0.4682958722114563,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.501139760017395,
+      "learning_rate": 8.231770931847469e-06,
+      "loss": 0.4564785361289978,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5794956140350878,
+      "grad_norm": 0.5357538461685181,
+      "learning_rate": 8.213921139207583e-06,
+      "loss": 0.4616129994392395,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5800438596491229,
+      "grad_norm": 0.47568657994270325,
+      "learning_rate": 8.196077222874755e-06,
+      "loss": 0.4671221673488617,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5805921052631579,
+      "grad_norm": 0.4939679801464081,
+      "learning_rate": 8.178239241556547e-06,
+      "loss": 0.47691187262535095,
+      "step": 1059
+    },
+    {
+      "epoch": 0.581140350877193,
+      "grad_norm": 0.49383652210235596,
+      "learning_rate": 8.160407253940996e-06,
+      "loss": 0.4642741084098816,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5816885964912281,
+      "grad_norm": 0.4947352707386017,
+      "learning_rate": 8.14258131869641e-06,
+      "loss": 0.45315033197402954,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5822368421052632,
+      "grad_norm": 0.46344584226608276,
+      "learning_rate": 8.124761494471191e-06,
+      "loss": 0.4679107666015625,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5827850877192983,
+      "grad_norm": 0.4847925901412964,
+      "learning_rate": 8.106947839893639e-06,
+      "loss": 0.46107810735702515,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 0.45921164751052856,
+      "learning_rate": 8.089140413571747e-06,
+      "loss": 0.4712381958961487,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5838815789473685,
+      "grad_norm": 0.5184562802314758,
+      "learning_rate": 8.071339274093018e-06,
+      "loss": 0.44746991991996765,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5844298245614035,
+      "grad_norm": 0.48062455654144287,
+      "learning_rate": 8.05354448002428e-06,
+      "loss": 0.4749741554260254,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5849780701754386,
+      "grad_norm": 0.45290520787239075,
+      "learning_rate": 8.035756089911471e-06,
+      "loss": 0.4632267653942108,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5855263157894737,
+      "grad_norm": 0.4724346697330475,
+      "learning_rate": 8.017974162279469e-06,
+      "loss": 0.4638788402080536,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5860745614035088,
+      "grad_norm": 0.48686882853507996,
+      "learning_rate": 8.00019875563189e-06,
+      "loss": 0.45769965648651123,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5866228070175439,
+      "grad_norm": 0.46247026324272156,
+      "learning_rate": 7.982429928450888e-06,
+      "loss": 0.45463666319847107,
+      "step": 1070
+    },
+    {
+      "epoch": 0.587171052631579,
+      "grad_norm": 0.5366244912147522,
+      "learning_rate": 7.964667739196976e-06,
+      "loss": 0.47857174277305603,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5877192982456141,
+      "grad_norm": 0.49044543504714966,
+      "learning_rate": 7.94691224630883e-06,
+      "loss": 0.4604954123497009,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5882675438596491,
+      "grad_norm": 0.49987998604774475,
+      "learning_rate": 7.929163508203091e-06,
+      "loss": 0.469890832901001,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5888157894736842,
+      "grad_norm": 0.5040852427482605,
+      "learning_rate": 7.91142158327417e-06,
+      "loss": 0.47207868099212646,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5893640350877193,
+      "grad_norm": 0.4750802218914032,
+      "learning_rate": 7.893686529894076e-06,
+      "loss": 0.47198259830474854,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5899122807017544,
+      "grad_norm": 0.4758901298046112,
+      "learning_rate": 7.875958406412206e-06,
+      "loss": 0.46596193313598633,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5904605263157895,
+      "grad_norm": 0.4805660545825958,
+      "learning_rate": 7.858237271155145e-06,
+      "loss": 0.454732745885849,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5910087719298246,
+      "grad_norm": 0.4993060529232025,
+      "learning_rate": 7.840523182426497e-06,
+      "loss": 0.46462464332580566,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5915570175438597,
+      "grad_norm": 0.47512087225914,
+      "learning_rate": 7.82281619850669e-06,
+      "loss": 0.4709402024745941,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5921052631578947,
+      "grad_norm": 0.4666202962398529,
+      "learning_rate": 7.80511637765276e-06,
+      "loss": 0.451152503490448,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5926535087719298,
+      "grad_norm": 0.48495787382125854,
+      "learning_rate": 7.787423778098183e-06,
+      "loss": 0.4697365164756775,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5932017543859649,
+      "grad_norm": 0.464061975479126,
+      "learning_rate": 7.769738458052685e-06,
+      "loss": 0.46991023421287537,
+      "step": 1082
+    },
+    {
+      "epoch": 0.59375,
+      "grad_norm": 0.47889429330825806,
+      "learning_rate": 7.752060475702028e-06,
+      "loss": 0.4730076491832733,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5942982456140351,
+      "grad_norm": 0.5057309865951538,
+      "learning_rate": 7.73438988920784e-06,
+      "loss": 0.46246618032455444,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5948464912280702,
+      "grad_norm": 0.47753041982650757,
+      "learning_rate": 7.716726756707418e-06,
+      "loss": 0.46080321073532104,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5953947368421053,
+      "grad_norm": 0.49280911684036255,
+      "learning_rate": 7.699071136313532e-06,
+      "loss": 0.46324023604393005,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5959429824561403,
+      "grad_norm": 0.4621438682079315,
+      "learning_rate": 7.681423086114232e-06,
+      "loss": 0.4490150213241577,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5964912280701754,
+      "grad_norm": 0.4937477111816406,
+      "learning_rate": 7.663782664172671e-06,
+      "loss": 0.47565126419067383,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5970394736842105,
+      "grad_norm": 0.4713328182697296,
+      "learning_rate": 7.6461499285269e-06,
+      "loss": 0.4622294306755066,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5975877192982456,
+      "grad_norm": 0.4742901027202606,
+      "learning_rate": 7.628524937189678e-06,
+      "loss": 0.46156418323516846,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5981359649122807,
+      "grad_norm": 0.5026023983955383,
+      "learning_rate": 7.610907748148291e-06,
+      "loss": 0.46600472927093506,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5986842105263158,
+      "grad_norm": 0.477543443441391,
+      "learning_rate": 7.593298419364355e-06,
+      "loss": 0.47159242630004883,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5992324561403509,
+      "grad_norm": 0.48052772879600525,
+      "learning_rate": 7.575697008773615e-06,
+      "loss": 0.44670578837394714,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5997807017543859,
+      "grad_norm": 0.4850292503833771,
+      "learning_rate": 7.558103574285779e-06,
+      "loss": 0.4710618853569031,
+      "step": 1094
+    },
+    {
+      "epoch": 0.600328947368421,
+      "grad_norm": 0.4854431450366974,
+      "learning_rate": 7.540518173784306e-06,
+      "loss": 0.4577372670173645,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6008771929824561,
+      "grad_norm": 0.465509831905365,
+      "learning_rate": 7.522940865126218e-06,
+      "loss": 0.45736953616142273,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6014254385964912,
+      "grad_norm": 0.4870331585407257,
+      "learning_rate": 7.505371706141925e-06,
+      "loss": 0.4602835774421692,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6019736842105263,
+      "grad_norm": 0.487417608499527,
+      "learning_rate": 7.4878107546350184e-06,
+      "loss": 0.4558318257331848,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6025219298245614,
+      "grad_norm": 0.49326610565185547,
+      "learning_rate": 7.470258068382079e-06,
+      "loss": 0.46510666608810425,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6030701754385965,
+      "grad_norm": 0.49196192622184753,
+      "learning_rate": 7.452713705132515e-06,
+      "loss": 0.457396924495697,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6036184210526315,
+      "grad_norm": 0.46562331914901733,
+      "learning_rate": 7.435177722608329e-06,
+      "loss": 0.46910542249679565,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6041666666666666,
+      "grad_norm": 0.4977482259273529,
+      "learning_rate": 7.417650178503959e-06,
+      "loss": 0.4489624500274658,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6047149122807017,
+      "grad_norm": 0.46233120560646057,
+      "learning_rate": 7.40013113048609e-06,
+      "loss": 0.4486108720302582,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6052631578947368,
+      "grad_norm": 0.4903429448604584,
+      "learning_rate": 7.382620636193438e-06,
+      "loss": 0.453243613243103,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6058114035087719,
+      "grad_norm": 0.4849245250225067,
+      "learning_rate": 7.3651187532365845e-06,
+      "loss": 0.4521423280239105,
+      "step": 1105
+    },
+    {
+      "epoch": 0.606359649122807,
+      "grad_norm": 0.4903513491153717,
+      "learning_rate": 7.347625539197785e-06,
+      "loss": 0.4613727033138275,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6069078947368421,
+      "grad_norm": 0.476204514503479,
+      "learning_rate": 7.3301410516307595e-06,
+      "loss": 0.4579714238643646,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6074561403508771,
+      "grad_norm": 0.47110995650291443,
+      "learning_rate": 7.312665348060533e-06,
+      "loss": 0.4564129710197449,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6080043859649122,
+      "grad_norm": 0.4694063067436218,
+      "learning_rate": 7.295198485983215e-06,
+      "loss": 0.4591718316078186,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6085526315789473,
+      "grad_norm": 0.46599674224853516,
+      "learning_rate": 7.2777405228658405e-06,
+      "loss": 0.4712226092815399,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6091008771929824,
+      "grad_norm": 0.4798458218574524,
+      "learning_rate": 7.2602915161461605e-06,
+      "loss": 0.4686587452888489,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6096491228070176,
+      "grad_norm": 0.48706597089767456,
+      "learning_rate": 7.242851523232448e-06,
+      "loss": 0.4550248086452484,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6101973684210527,
+      "grad_norm": 0.4822862148284912,
+      "learning_rate": 7.225420601503347e-06,
+      "loss": 0.4606156349182129,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6107456140350878,
+      "grad_norm": 0.47604596614837646,
+      "learning_rate": 7.207998808307628e-06,
+      "loss": 0.4694164991378784,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6112938596491229,
+      "grad_norm": 0.4813266396522522,
+      "learning_rate": 7.190586200964041e-06,
+      "loss": 0.4651000499725342,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6118421052631579,
+      "grad_norm": 0.4579448401927948,
+      "learning_rate": 7.173182836761121e-06,
+      "loss": 0.4731736183166504,
+      "step": 1116
+    },
+    {
+      "epoch": 0.612390350877193,
+      "grad_norm": 0.4748271107673645,
+      "learning_rate": 7.155788772956978e-06,
+      "loss": 0.4703238010406494,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6129385964912281,
+      "grad_norm": 0.4719952940940857,
+      "learning_rate": 7.13840406677913e-06,
+      "loss": 0.45160242915153503,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6134868421052632,
+      "grad_norm": 0.4611450731754303,
+      "learning_rate": 7.121028775424313e-06,
+      "loss": 0.47287964820861816,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6140350877192983,
+      "grad_norm": 0.5155686736106873,
+      "learning_rate": 7.103662956058277e-06,
+      "loss": 0.476226806640625,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6145833333333334,
+      "grad_norm": 0.46889781951904297,
+      "learning_rate": 7.0863066658156124e-06,
+      "loss": 0.455394446849823,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6151315789473685,
+      "grad_norm": 0.49264058470726013,
+      "learning_rate": 7.068959961799569e-06,
+      "loss": 0.4657437801361084,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6156798245614035,
+      "grad_norm": 0.454782634973526,
+      "learning_rate": 7.051622901081838e-06,
+      "loss": 0.45418664813041687,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6162280701754386,
+      "grad_norm": 0.487419992685318,
+      "learning_rate": 7.034295540702398e-06,
+      "loss": 0.45389851927757263,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6167763157894737,
+      "grad_norm": 0.46922385692596436,
+      "learning_rate": 7.016977937669312e-06,
+      "loss": 0.45923101902008057,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6173245614035088,
+      "grad_norm": 0.4770352840423584,
+      "learning_rate": 6.999670148958533e-06,
+      "loss": 0.4591761827468872,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6178728070175439,
+      "grad_norm": 0.4865526258945465,
+      "learning_rate": 6.9823722315137255e-06,
+      "loss": 0.4650563597679138,
+      "step": 1127
+    },
+    {
+      "epoch": 0.618421052631579,
+      "grad_norm": 0.47865018248558044,
+      "learning_rate": 6.9650842422460875e-06,
+      "loss": 0.4477848410606384,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6189692982456141,
+      "grad_norm": 0.4948062300682068,
+      "learning_rate": 6.947806238034143e-06,
+      "loss": 0.450751394033432,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6195175438596491,
+      "grad_norm": 0.4848669767379761,
+      "learning_rate": 6.930538275723562e-06,
+      "loss": 0.45120009779930115,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6200657894736842,
+      "grad_norm": 0.5030041933059692,
+      "learning_rate": 6.9132804121269855e-06,
+      "loss": 0.468425989151001,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6206140350877193,
+      "grad_norm": 0.4994588792324066,
+      "learning_rate": 6.896032704023828e-06,
+      "loss": 0.44643884897232056,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6211622807017544,
+      "grad_norm": 0.4888147711753845,
+      "learning_rate": 6.878795208160077e-06,
+      "loss": 0.4606155753135681,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6217105263157895,
+      "grad_norm": 0.49748578667640686,
+      "learning_rate": 6.861567981248143e-06,
+      "loss": 0.4452589452266693,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6222587719298246,
+      "grad_norm": 0.5071046352386475,
+      "learning_rate": 6.8443510799666375e-06,
+      "loss": 0.4486559331417084,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6228070175438597,
+      "grad_norm": 0.4967880845069885,
+      "learning_rate": 6.827144560960201e-06,
+      "loss": 0.45914992690086365,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6233552631578947,
+      "grad_norm": 0.47261181473731995,
+      "learning_rate": 6.809948480839319e-06,
+      "loss": 0.4559020400047302,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6239035087719298,
+      "grad_norm": 0.4834285080432892,
+      "learning_rate": 6.792762896180133e-06,
+      "loss": 0.44998395442962646,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6244517543859649,
+      "grad_norm": 0.4783952236175537,
+      "learning_rate": 6.7755878635242445e-06,
+      "loss": 0.45917055010795593,
+      "step": 1139
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.4811253249645233,
+      "learning_rate": 6.758423439378556e-06,
+      "loss": 0.4707831144332886,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6255482456140351,
+      "grad_norm": 0.458852082490921,
+      "learning_rate": 6.741269680215053e-06,
+      "loss": 0.4482444226741791,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6260964912280702,
+      "grad_norm": 0.47917571663856506,
+      "learning_rate": 6.7241266424706365e-06,
+      "loss": 0.45533668994903564,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6266447368421053,
+      "grad_norm": 0.4818931221961975,
+      "learning_rate": 6.706994382546927e-06,
+      "loss": 0.4754176139831543,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6271929824561403,
+      "grad_norm": 0.48275595903396606,
+      "learning_rate": 6.689872956810103e-06,
+      "loss": 0.4723743200302124,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6277412280701754,
+      "grad_norm": 0.4792291224002838,
+      "learning_rate": 6.672762421590682e-06,
+      "loss": 0.4706377685070038,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6282894736842105,
+      "grad_norm": 0.49965038895606995,
+      "learning_rate": 6.655662833183352e-06,
+      "loss": 0.46659570932388306,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6288377192982456,
+      "grad_norm": 0.5105276703834534,
+      "learning_rate": 6.638574247846793e-06,
+      "loss": 0.46983322501182556,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6293859649122807,
+      "grad_norm": 0.4503490924835205,
+      "learning_rate": 6.621496721803482e-06,
+      "loss": 0.4630698561668396,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6299342105263158,
+      "grad_norm": 0.4773010313510895,
+      "learning_rate": 6.604430311239504e-06,
+      "loss": 0.45521196722984314,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6304824561403509,
+      "grad_norm": 0.4769289791584015,
+      "learning_rate": 6.5873750723043805e-06,
+      "loss": 0.4638155698776245,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6310307017543859,
+      "grad_norm": 0.4745919108390808,
+      "learning_rate": 6.570331061110877e-06,
+      "loss": 0.45071423053741455,
+      "step": 1151
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.470969557762146,
+      "learning_rate": 6.553298333734812e-06,
+      "loss": 0.4583868980407715,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6321271929824561,
+      "grad_norm": 0.4607287645339966,
+      "learning_rate": 6.53627694621489e-06,
+      "loss": 0.4590507745742798,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6326754385964912,
+      "grad_norm": 0.4686935842037201,
+      "learning_rate": 6.519266954552501e-06,
+      "loss": 0.4589221477508545,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6332236842105263,
+      "grad_norm": 0.4886135160923004,
+      "learning_rate": 6.502268414711534e-06,
+      "loss": 0.4689658582210541,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6337719298245614,
+      "grad_norm": 0.48805180191993713,
+      "learning_rate": 6.485281382618222e-06,
+      "loss": 0.4538452625274658,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6343201754385965,
+      "grad_norm": 0.47901463508605957,
+      "learning_rate": 6.468305914160917e-06,
+      "loss": 0.4631307125091553,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6348684210526315,
+      "grad_norm": 0.4804033637046814,
+      "learning_rate": 6.451342065189932e-06,
+      "loss": 0.47523754835128784,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6354166666666666,
+      "grad_norm": 0.4747447073459625,
+      "learning_rate": 6.434389891517356e-06,
+      "loss": 0.4520912766456604,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6359649122807017,
+      "grad_norm": 0.46864253282546997,
+      "learning_rate": 6.41744944891686e-06,
+      "loss": 0.44075965881347656,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6365131578947368,
+      "grad_norm": 0.46973252296447754,
+      "learning_rate": 6.400520793123519e-06,
+      "loss": 0.47107774019241333,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6370614035087719,
+      "grad_norm": 0.4863990843296051,
+      "learning_rate": 6.383603979833635e-06,
+      "loss": 0.46418333053588867,
+      "step": 1162
+    },
+    {
+      "epoch": 0.637609649122807,
+      "grad_norm": 0.45822757482528687,
+      "learning_rate": 6.366699064704538e-06,
+      "loss": 0.45333075523376465,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6381578947368421,
+      "grad_norm": 0.47059157490730286,
+      "learning_rate": 6.349806103354417e-06,
+      "loss": 0.4626568555831909,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6387061403508771,
+      "grad_norm": 0.48919686675071716,
+      "learning_rate": 6.332925151362139e-06,
+      "loss": 0.4648444652557373,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6392543859649122,
+      "grad_norm": 0.4936853349208832,
+      "learning_rate": 6.316056264267048e-06,
+      "loss": 0.45927223563194275,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6398026315789473,
+      "grad_norm": 0.49010518193244934,
+      "learning_rate": 6.299199497568798e-06,
+      "loss": 0.4588416516780853,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6403508771929824,
+      "grad_norm": 0.48247334361076355,
+      "learning_rate": 6.282354906727171e-06,
+      "loss": 0.44723185896873474,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6408991228070176,
+      "grad_norm": 0.49713844060897827,
+      "learning_rate": 6.265522547161884e-06,
+      "loss": 0.4614097476005554,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6414473684210527,
+      "grad_norm": 0.4781050682067871,
+      "learning_rate": 6.24870247425241e-06,
+      "loss": 0.45371532440185547,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6419956140350878,
+      "grad_norm": 0.5242354869842529,
+      "learning_rate": 6.23189474333781e-06,
+      "loss": 0.4608764350414276,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6425438596491229,
+      "grad_norm": 0.5145214200019836,
+      "learning_rate": 6.215099409716527e-06,
+      "loss": 0.4503445625305176,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6430921052631579,
+      "grad_norm": 0.47920140624046326,
+      "learning_rate": 6.19831652864622e-06,
+      "loss": 0.44728970527648926,
+      "step": 1173
+    },
+    {
+      "epoch": 0.643640350877193,
+      "grad_norm": 0.49056777358055115,
+      "learning_rate": 6.1815461553435784e-06,
+      "loss": 0.46014583110809326,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6441885964912281,
+      "grad_norm": 0.46811583638191223,
+      "learning_rate": 6.164788344984147e-06,
+      "loss": 0.45799893140792847,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6447368421052632,
+      "grad_norm": 0.4585595726966858,
+      "learning_rate": 6.1480431527021235e-06,
+      "loss": 0.4535917043685913,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6452850877192983,
+      "grad_norm": 0.4631287753582001,
+      "learning_rate": 6.1313106335902e-06,
+      "loss": 0.4646044671535492,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6458333333333334,
+      "grad_norm": 0.46791666746139526,
+      "learning_rate": 6.114590842699379e-06,
+      "loss": 0.45151329040527344,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6463815789473685,
+      "grad_norm": 0.5028589963912964,
+      "learning_rate": 6.097883835038773e-06,
+      "loss": 0.4618037939071655,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6469298245614035,
+      "grad_norm": 0.46903207898139954,
+      "learning_rate": 6.081189665575447e-06,
+      "loss": 0.4541777968406677,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6474780701754386,
+      "grad_norm": 0.4830973744392395,
+      "learning_rate": 6.0645083892342274e-06,
+      "loss": 0.45828375220298767,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6480263157894737,
+      "grad_norm": 0.46263155341148376,
+      "learning_rate": 6.047840060897512e-06,
+      "loss": 0.44502386450767517,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6485745614035088,
+      "grad_norm": 0.4844227731227875,
+      "learning_rate": 6.031184735405106e-06,
+      "loss": 0.4488610625267029,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6491228070175439,
+      "grad_norm": 1.6031123399734497,
+      "learning_rate": 6.01454246755404e-06,
+      "loss": 0.4552503824234009,
+      "step": 1184
+    },
+    {
+      "epoch": 0.649671052631579,
+      "grad_norm": 0.5204753875732422,
+      "learning_rate": 5.997913312098374e-06,
+      "loss": 0.4529512822628021,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6502192982456141,
+      "grad_norm": 0.46702829003334045,
+      "learning_rate": 5.981297323749028e-06,
+      "loss": 0.45574164390563965,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6507675438596491,
+      "grad_norm": 0.4901115894317627,
+      "learning_rate": 5.964694557173609e-06,
+      "loss": 0.46220001578330994,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6513157894736842,
+      "grad_norm": 0.48896780610084534,
+      "learning_rate": 5.948105066996221e-06,
+      "loss": 0.45643150806427,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6518640350877193,
+      "grad_norm": 0.4706425964832306,
+      "learning_rate": 5.931528907797281e-06,
+      "loss": 0.4490327537059784,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6524122807017544,
+      "grad_norm": 0.46755537390708923,
+      "learning_rate": 5.914966134113354e-06,
+      "loss": 0.46012648940086365,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6529605263157895,
+      "grad_norm": 0.471127986907959,
+      "learning_rate": 5.898416800436966e-06,
+      "loss": 0.4527897834777832,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6535087719298246,
+      "grad_norm": 0.48551979660987854,
+      "learning_rate": 5.8818809612164155e-06,
+      "loss": 0.4556393623352051,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6540570175438597,
+      "grad_norm": 0.47003045678138733,
+      "learning_rate": 5.865358670855614e-06,
+      "loss": 0.4481228291988373,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6546052631578947,
+      "grad_norm": 0.4837598204612732,
+      "learning_rate": 5.848849983713894e-06,
+      "loss": 0.44623738527297974,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6551535087719298,
+      "grad_norm": 0.5016481280326843,
+      "learning_rate": 5.832354954105826e-06,
+      "loss": 0.4614674746990204,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6557017543859649,
+      "grad_norm": 0.46373236179351807,
+      "learning_rate": 5.815873636301053e-06,
+      "loss": 0.45254772901535034,
+      "step": 1196
+    },
+    {
+      "epoch": 0.65625,
+      "grad_norm": 0.48568040132522583,
+      "learning_rate": 5.799406084524106e-06,
+      "loss": 0.4651646018028259,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6567982456140351,
+      "grad_norm": 0.48499056696891785,
+      "learning_rate": 5.7829523529542185e-06,
+      "loss": 0.4580027759075165,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6573464912280702,
+      "grad_norm": 0.48886817693710327,
+      "learning_rate": 5.766512495725165e-06,
+      "loss": 0.4550062119960785,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 0.4718495011329651,
+      "learning_rate": 5.7500865669250625e-06,
+      "loss": 0.4428224563598633,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6584429824561403,
+      "grad_norm": 0.9899610280990601,
+      "learning_rate": 5.733674620596202e-06,
+      "loss": 0.46023988723754883,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6589912280701754,
+      "grad_norm": 0.4938037395477295,
+      "learning_rate": 5.717276710734879e-06,
+      "loss": 0.4411492943763733,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6595394736842105,
+      "grad_norm": 0.49084752798080444,
+      "learning_rate": 5.7008928912912096e-06,
+      "loss": 0.4550079107284546,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6600877192982456,
+      "grad_norm": 0.48547691106796265,
+      "learning_rate": 5.68452321616894e-06,
+      "loss": 0.4425356388092041,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6606359649122807,
+      "grad_norm": 0.7199835777282715,
+      "learning_rate": 5.668167739225294e-06,
+      "loss": 0.4713376462459564,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6611842105263158,
+      "grad_norm": 0.47304996848106384,
+      "learning_rate": 5.651826514270773e-06,
+      "loss": 0.45630842447280884,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6617324561403509,
+      "grad_norm": 0.45576387643814087,
+      "learning_rate": 5.635499595068986e-06,
+      "loss": 0.454340398311615,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6622807017543859,
+      "grad_norm": 0.47770845890045166,
+      "learning_rate": 5.619187035336486e-06,
+      "loss": 0.4564405679702759,
+      "step": 1208
+    },
+    {
+      "epoch": 0.662828947368421,
+      "grad_norm": 0.4835106134414673,
+      "learning_rate": 5.602888888742583e-06,
+      "loss": 0.46910983324050903,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6633771929824561,
+      "grad_norm": 0.500394344329834,
+      "learning_rate": 5.586605208909155e-06,
+      "loss": 0.4477820098400116,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6639254385964912,
+      "grad_norm": 0.4716354012489319,
+      "learning_rate": 5.570336049410487e-06,
+      "loss": 0.4602228105068207,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6644736842105263,
+      "grad_norm": 0.4881833791732788,
+      "learning_rate": 5.554081463773098e-06,
+      "loss": 0.43738436698913574,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6650219298245614,
+      "grad_norm": 0.6254478693008423,
+      "learning_rate": 5.537841505475556e-06,
+      "loss": 0.44146454334259033,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6655701754385965,
+      "grad_norm": 0.48677903413772583,
+      "learning_rate": 5.521616227948297e-06,
+      "loss": 0.44918176531791687,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6661184210526315,
+      "grad_norm": 0.4938996434211731,
+      "learning_rate": 5.505405684573466e-06,
+      "loss": 0.45282015204429626,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.47285711765289307,
+      "learning_rate": 5.489209928684728e-06,
+      "loss": 0.4627668857574463,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6672149122807017,
+      "grad_norm": 0.46757328510284424,
+      "learning_rate": 5.473029013567089e-06,
+      "loss": 0.4532211422920227,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6677631578947368,
+      "grad_norm": 0.49316030740737915,
+      "learning_rate": 5.456862992456738e-06,
+      "loss": 0.4442199170589447,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6683114035087719,
+      "grad_norm": 0.4487299919128418,
+      "learning_rate": 5.440711918540863e-06,
+      "loss": 0.4583362936973572,
+      "step": 1219
+    },
+    {
+      "epoch": 0.668859649122807,
+      "grad_norm": 0.47031185030937195,
+      "learning_rate": 5.424575844957463e-06,
+      "loss": 0.4489290714263916,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6694078947368421,
+      "grad_norm": 0.4853135645389557,
+      "learning_rate": 5.408454824795199e-06,
+      "loss": 0.4479921758174896,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6699561403508771,
+      "grad_norm": 0.4838583171367645,
+      "learning_rate": 5.392348911093198e-06,
+      "loss": 0.4525047242641449,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6705043859649122,
+      "grad_norm": 0.485013484954834,
+      "learning_rate": 5.376258156840874e-06,
+      "loss": 0.45637011528015137,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6710526315789473,
+      "grad_norm": 0.4787314832210541,
+      "learning_rate": 5.360182614977797e-06,
+      "loss": 0.457485169172287,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6716008771929824,
+      "grad_norm": 0.46060094237327576,
+      "learning_rate": 5.344122338393458e-06,
+      "loss": 0.4558663070201874,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6721491228070176,
+      "grad_norm": 0.49685534834861755,
+      "learning_rate": 5.328077379927133e-06,
+      "loss": 0.46039676666259766,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6726973684210527,
+      "grad_norm": 0.4821500778198242,
+      "learning_rate": 5.312047792367707e-06,
+      "loss": 0.46990689635276794,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6732456140350878,
+      "grad_norm": 0.49922671914100647,
+      "learning_rate": 5.296033628453484e-06,
+      "loss": 0.45971810817718506,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6737938596491229,
+      "grad_norm": 0.48737815022468567,
+      "learning_rate": 5.28003494087203e-06,
+      "loss": 0.4628313481807709,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6743421052631579,
+      "grad_norm": 0.4753425419330597,
+      "learning_rate": 5.264051782259996e-06,
+      "loss": 0.44580182433128357,
+      "step": 1230
+    },
+    {
+      "epoch": 0.674890350877193,
+      "grad_norm": 0.4703112244606018,
+      "learning_rate": 5.248084205202932e-06,
+      "loss": 0.4599781036376953,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6754385964912281,
+      "grad_norm": 0.5177963376045227,
+      "learning_rate": 5.232132262235126e-06,
+      "loss": 0.4688262641429901,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6759868421052632,
+      "grad_norm": 0.5090923309326172,
+      "learning_rate": 5.2161960058394375e-06,
+      "loss": 0.45433613657951355,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6765350877192983,
+      "grad_norm": 0.481566458940506,
+      "learning_rate": 5.200275488447104e-06,
+      "loss": 0.44734376668930054,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6770833333333334,
+      "grad_norm": 0.4732418954372406,
+      "learning_rate": 5.184370762437589e-06,
+      "loss": 0.458044171333313,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6776315789473685,
+      "grad_norm": 0.4842263162136078,
+      "learning_rate": 5.168481880138405e-06,
+      "loss": 0.4429929256439209,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6781798245614035,
+      "grad_norm": 0.46749964356422424,
+      "learning_rate": 5.1526088938249275e-06,
+      "loss": 0.4532344341278076,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6787280701754386,
+      "grad_norm": 0.48561593890190125,
+      "learning_rate": 5.136751855720235e-06,
+      "loss": 0.45579272508621216,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6792763157894737,
+      "grad_norm": 0.5013870596885681,
+      "learning_rate": 5.120910817994939e-06,
+      "loss": 0.45137521624565125,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6798245614035088,
+      "grad_norm": 0.4608660638332367,
+      "learning_rate": 5.105085832767014e-06,
+      "loss": 0.45345020294189453,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6803728070175439,
+      "grad_norm": 0.47580307722091675,
+      "learning_rate": 5.0892769521016075e-06,
+      "loss": 0.46015557646751404,
+      "step": 1241
+    },
+    {
+      "epoch": 0.680921052631579,
+      "grad_norm": 0.4806741774082184,
+      "learning_rate": 5.073484228010897e-06,
+      "loss": 0.436834454536438,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6814692982456141,
+      "grad_norm": 0.5022061467170715,
+      "learning_rate": 5.057707712453891e-06,
+      "loss": 0.45409783720970154,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6820175438596491,
+      "grad_norm": 0.4873558282852173,
+      "learning_rate": 5.041947457336273e-06,
+      "loss": 0.4772658944129944,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6825657894736842,
+      "grad_norm": 0.46109822392463684,
+      "learning_rate": 5.0262035145102375e-06,
+      "loss": 0.4520254135131836,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6831140350877193,
+      "grad_norm": 0.49353766441345215,
+      "learning_rate": 5.010475935774306e-06,
+      "loss": 0.4518451392650604,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6836622807017544,
+      "grad_norm": 0.483317494392395,
+      "learning_rate": 4.994764772873158e-06,
+      "loss": 0.45198535919189453,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.4909769594669342,
+      "learning_rate": 4.979070077497461e-06,
+      "loss": 0.4568019509315491,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6847587719298246,
+      "grad_norm": 0.4987059235572815,
+      "learning_rate": 4.963391901283717e-06,
+      "loss": 0.4692850708961487,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6853070175438597,
+      "grad_norm": 0.49539315700531006,
+      "learning_rate": 4.947730295814063e-06,
+      "loss": 0.4367455244064331,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6858552631578947,
+      "grad_norm": 0.4935382604598999,
+      "learning_rate": 4.932085312616127e-06,
+      "loss": 0.476955384016037,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6864035087719298,
+      "grad_norm": 0.4814411997795105,
+      "learning_rate": 4.9164570031628525e-06,
+      "loss": 0.459894597530365,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6869517543859649,
+      "grad_norm": 0.49503499269485474,
+      "learning_rate": 4.900845418872312e-06,
+      "loss": 0.44997191429138184,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.503138542175293,
+      "learning_rate": 4.885250611107558e-06,
+      "loss": 0.4328072667121887,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6880482456140351,
+      "grad_norm": 0.4843796491622925,
+      "learning_rate": 4.8696726311764485e-06,
+      "loss": 0.4629124104976654,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6885964912280702,
+      "grad_norm": 0.531264066696167,
+      "learning_rate": 4.854111530331482e-06,
+      "loss": 0.45157432556152344,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6891447368421053,
+      "grad_norm": 0.4729582667350769,
+      "learning_rate": 4.8385673597696095e-06,
+      "loss": 0.4563821256160736,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6896929824561403,
+      "grad_norm": 0.497354120016098,
+      "learning_rate": 4.823040170632095e-06,
+      "loss": 0.45511579513549805,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6902412280701754,
+      "grad_norm": 0.4853935241699219,
+      "learning_rate": 4.807530014004325e-06,
+      "loss": 0.45746946334838867,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6907894736842105,
+      "grad_norm": 0.48168012499809265,
+      "learning_rate": 4.792036940915642e-06,
+      "loss": 0.4588317573070526,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6913377192982456,
+      "grad_norm": 0.47273921966552734,
+      "learning_rate": 4.776561002339198e-06,
+      "loss": 0.4494730234146118,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6918859649122807,
+      "grad_norm": 0.4824991524219513,
+      "learning_rate": 4.761102249191763e-06,
+      "loss": 0.45408129692077637,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6924342105263158,
+      "grad_norm": 0.494086891412735,
+      "learning_rate": 4.745660732333561e-06,
+      "loss": 0.4487881064414978,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6929824561403509,
+      "grad_norm": 0.4966714680194855,
+      "learning_rate": 4.730236502568121e-06,
+      "loss": 0.4485003650188446,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6935307017543859,
+      "grad_norm": 0.49037015438079834,
+      "learning_rate": 4.714829610642082e-06,
+      "loss": 0.45986589789390564,
+      "step": 1265
+    },
+    {
+      "epoch": 0.694078947368421,
+      "grad_norm": 0.4807148873806,
+      "learning_rate": 4.699440107245041e-06,
+      "loss": 0.4428585469722748,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6946271929824561,
+      "grad_norm": 0.48849642276763916,
+      "learning_rate": 4.684068043009406e-06,
+      "loss": 0.4557689130306244,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6951754385964912,
+      "grad_norm": 0.4828411042690277,
+      "learning_rate": 4.668713468510187e-06,
+      "loss": 0.4486156404018402,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6957236842105263,
+      "grad_norm": 0.4808543026447296,
+      "learning_rate": 4.6533764342648546e-06,
+      "loss": 0.4532542824745178,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6962719298245614,
+      "grad_norm": 0.46090418100357056,
+      "learning_rate": 4.638056990733184e-06,
+      "loss": 0.466025710105896,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6968201754385965,
+      "grad_norm": 0.46584606170654297,
+      "learning_rate": 4.622755188317059e-06,
+      "loss": 0.4666459262371063,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6973684210526315,
+      "grad_norm": 0.48791778087615967,
+      "learning_rate": 4.607471077360337e-06,
+      "loss": 0.4741817116737366,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6979166666666666,
+      "grad_norm": 0.4724322259426117,
+      "learning_rate": 4.5922047081486665e-06,
+      "loss": 0.4573628306388855,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6984649122807017,
+      "grad_norm": 0.48471513390541077,
+      "learning_rate": 4.576956130909317e-06,
+      "loss": 0.4618206024169922,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6990131578947368,
+      "grad_norm": 0.47347649931907654,
+      "learning_rate": 4.561725395811027e-06,
+      "loss": 0.44821199774742126,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6995614035087719,
+      "grad_norm": 0.4988267421722412,
+      "learning_rate": 4.546512552963831e-06,
+      "loss": 0.4621202349662781,
+      "step": 1276
+    },
+    {
+      "epoch": 0.700109649122807,
+      "grad_norm": 0.511737585067749,
+      "learning_rate": 4.5313176524188995e-06,
+      "loss": 0.4463235139846802,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7006578947368421,
+      "grad_norm": 0.46017760038375854,
+      "learning_rate": 4.516140744168376e-06,
+      "loss": 0.45576271414756775,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7012061403508771,
+      "grad_norm": 0.472093790769577,
+      "learning_rate": 4.500981878145195e-06,
+      "loss": 0.45104914903640747,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7017543859649122,
+      "grad_norm": 0.4837212562561035,
+      "learning_rate": 4.485841104222946e-06,
+      "loss": 0.45531660318374634,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7023026315789473,
+      "grad_norm": 0.5072401165962219,
+      "learning_rate": 4.470718472215683e-06,
+      "loss": 0.4535329341888428,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7028508771929824,
+      "grad_norm": 0.5329047441482544,
+      "learning_rate": 4.455614031877775e-06,
+      "loss": 0.4547671973705292,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7033991228070176,
+      "grad_norm": 0.4614371061325073,
+      "learning_rate": 4.440527832903748e-06,
+      "loss": 0.4641307294368744,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7039473684210527,
+      "grad_norm": 0.4439223110675812,
+      "learning_rate": 4.425459924928102e-06,
+      "loss": 0.44951772689819336,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7044956140350878,
+      "grad_norm": 0.4774690568447113,
+      "learning_rate": 4.4104103575251565e-06,
+      "loss": 0.4569922685623169,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7050438596491229,
+      "grad_norm": 0.4845373332500458,
+      "learning_rate": 4.395379180208904e-06,
+      "loss": 0.45955735445022583,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7055921052631579,
+      "grad_norm": 0.45962685346603394,
+      "learning_rate": 4.380366442432814e-06,
+      "loss": 0.44685596227645874,
+      "step": 1287
+    },
+    {
+      "epoch": 0.706140350877193,
+      "grad_norm": 0.46158796548843384,
+      "learning_rate": 4.365372193589704e-06,
+      "loss": 0.462146520614624,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7066885964912281,
+      "grad_norm": 0.49494469165802,
+      "learning_rate": 4.350396483011559e-06,
+      "loss": 0.4598894715309143,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7072368421052632,
+      "grad_norm": 0.47989487648010254,
+      "learning_rate": 4.335439359969367e-06,
+      "loss": 0.4486043155193329,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7077850877192983,
+      "grad_norm": 0.46652752161026,
+      "learning_rate": 4.320500873672957e-06,
+      "loss": 0.4489085078239441,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7083333333333334,
+      "grad_norm": 0.46897727251052856,
+      "learning_rate": 4.3055810732708584e-06,
+      "loss": 0.4552903175354004,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7088815789473685,
+      "grad_norm": 0.461935430765152,
+      "learning_rate": 4.290680007850105e-06,
+      "loss": 0.4531635046005249,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7094298245614035,
+      "grad_norm": 0.4682982861995697,
+      "learning_rate": 4.275797726436105e-06,
+      "loss": 0.4542514979839325,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7099780701754386,
+      "grad_norm": 0.4622385501861572,
+      "learning_rate": 4.260934277992463e-06,
+      "loss": 0.45592039823532104,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7105263157894737,
+      "grad_norm": 0.46560776233673096,
+      "learning_rate": 4.246089711420817e-06,
+      "loss": 0.4482199549674988,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7110745614035088,
+      "grad_norm": 0.4624495208263397,
+      "learning_rate": 4.231264075560684e-06,
+      "loss": 0.44374164938926697,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7116228070175439,
+      "grad_norm": 0.4659866690635681,
+      "learning_rate": 4.216457419189298e-06,
+      "loss": 0.4579788148403168,
+      "step": 1298
+    },
+    {
+      "epoch": 0.712171052631579,
+      "grad_norm": 0.45782938599586487,
+      "learning_rate": 4.201669791021458e-06,
+      "loss": 0.4533209204673767,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7127192982456141,
+      "grad_norm": 0.4743126332759857,
+      "learning_rate": 4.186901239709342e-06,
+      "loss": 0.4503573775291443,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7132675438596491,
+      "grad_norm": 0.4711255729198456,
+      "learning_rate": 4.172151813842382e-06,
+      "loss": 0.43793973326683044,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7138157894736842,
+      "grad_norm": 0.4743192791938782,
+      "learning_rate": 4.157421561947075e-06,
+      "loss": 0.4586021304130554,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7143640350877193,
+      "grad_norm": 0.47850310802459717,
+      "learning_rate": 4.142710532486831e-06,
+      "loss": 0.45045897364616394,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7149122807017544,
+      "grad_norm": 0.45966672897338867,
+      "learning_rate": 4.12801877386183e-06,
+      "loss": 0.46221137046813965,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7154605263157895,
+      "grad_norm": 0.48182302713394165,
+      "learning_rate": 4.113346334408844e-06,
+      "loss": 0.4529826045036316,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7160087719298246,
+      "grad_norm": 0.4576120376586914,
+      "learning_rate": 4.0986932624010775e-06,
+      "loss": 0.4573781490325928,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7165570175438597,
+      "grad_norm": 0.48668473958969116,
+      "learning_rate": 4.0840596060480266e-06,
+      "loss": 0.425226628780365,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7171052631578947,
+      "grad_norm": 0.5012518763542175,
+      "learning_rate": 4.0694454134952956e-06,
+      "loss": 0.45593565702438354,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7176535087719298,
+      "grad_norm": 0.4993784725666046,
+      "learning_rate": 4.054850732824463e-06,
+      "loss": 0.4585998058319092,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7182017543859649,
+      "grad_norm": 0.482063889503479,
+      "learning_rate": 4.040275612052905e-06,
+      "loss": 0.44233378767967224,
+      "step": 1310
+    },
+    {
+      "epoch": 0.71875,
+      "grad_norm": 0.4934483468532562,
+      "learning_rate": 4.025720099133651e-06,
+      "loss": 0.44768810272216797,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7192982456140351,
+      "grad_norm": 0.48552730679512024,
+      "learning_rate": 4.011184241955212e-06,
+      "loss": 0.45926231145858765,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7198464912280702,
+      "grad_norm": 0.5012895464897156,
+      "learning_rate": 3.996668088341432e-06,
+      "loss": 0.45643967390060425,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7203947368421053,
+      "grad_norm": 0.4595252275466919,
+      "learning_rate": 3.982171686051333e-06,
+      "loss": 0.4594549834728241,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7209429824561403,
+      "grad_norm": 0.47561824321746826,
+      "learning_rate": 3.967695082778958e-06,
+      "loss": 0.45136356353759766,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7214912280701754,
+      "grad_norm": 0.46955716609954834,
+      "learning_rate": 3.953238326153193e-06,
+      "loss": 0.45608794689178467,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7220394736842105,
+      "grad_norm": 0.4859420359134674,
+      "learning_rate": 3.938801463737647e-06,
+      "loss": 0.4654580056667328,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7225877192982456,
+      "grad_norm": 0.47013059258461,
+      "learning_rate": 3.924384543030464e-06,
+      "loss": 0.4543027877807617,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7231359649122807,
+      "grad_norm": 0.47090038657188416,
+      "learning_rate": 3.909987611464179e-06,
+      "loss": 0.4615466594696045,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7236842105263158,
+      "grad_norm": 0.4833797216415405,
+      "learning_rate": 3.895610716405565e-06,
+      "loss": 0.4522587060928345,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7242324561403509,
+      "grad_norm": 0.48517337441444397,
+      "learning_rate": 3.881253905155481e-06,
+      "loss": 0.44629907608032227,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7247807017543859,
+      "grad_norm": 0.48679855465888977,
+      "learning_rate": 3.86691722494869e-06,
+      "loss": 0.4545023441314697,
+      "step": 1322
+    },
+    {
+      "epoch": 0.725328947368421,
+      "grad_norm": 0.5102160573005676,
+      "learning_rate": 3.852600722953741e-06,
+      "loss": 0.4341837763786316,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7258771929824561,
+      "grad_norm": 0.4852652847766876,
+      "learning_rate": 3.838304446272782e-06,
+      "loss": 0.4571802616119385,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7264254385964912,
+      "grad_norm": 0.462271511554718,
+      "learning_rate": 3.82402844194143e-06,
+      "loss": 0.44814571738243103,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7269736842105263,
+      "grad_norm": 0.4705876111984253,
+      "learning_rate": 3.8097727569286003e-06,
+      "loss": 0.457721084356308,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7275219298245614,
+      "grad_norm": 0.4641494154930115,
+      "learning_rate": 3.795537438136352e-06,
+      "loss": 0.4574616849422455,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7280701754385965,
+      "grad_norm": 0.4461621046066284,
+      "learning_rate": 3.7813225323997395e-06,
+      "loss": 0.45125675201416016,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7286184210526315,
+      "grad_norm": 0.4990176260471344,
+      "learning_rate": 3.7671280864866644e-06,
+      "loss": 0.446943461894989,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7291666666666666,
+      "grad_norm": 0.4785372018814087,
+      "learning_rate": 3.7529541470977037e-06,
+      "loss": 0.4530734419822693,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7297149122807017,
+      "grad_norm": 0.47964751720428467,
+      "learning_rate": 3.738800760865975e-06,
+      "loss": 0.4506794810295105,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7302631578947368,
+      "grad_norm": 0.4757719039916992,
+      "learning_rate": 3.7246679743569737e-06,
+      "loss": 0.4483231008052826,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7308114035087719,
+      "grad_norm": 0.4769383370876312,
+      "learning_rate": 3.710555834068418e-06,
+      "loss": 0.4365041255950928,
+      "step": 1333
+    },
+    {
+      "epoch": 0.731359649122807,
+      "grad_norm": 0.49054670333862305,
+      "learning_rate": 3.696464386430093e-06,
+      "loss": 0.4484644830226898,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7319078947368421,
+      "grad_norm": 0.4718336760997772,
+      "learning_rate": 3.6823936778037173e-06,
+      "loss": 0.4425055980682373,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7324561403508771,
+      "grad_norm": 0.548764169216156,
+      "learning_rate": 3.6683437544827704e-06,
+      "loss": 0.45093807578086853,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7330043859649122,
+      "grad_norm": 0.47863689064979553,
+      "learning_rate": 3.6543146626923422e-06,
+      "loss": 0.4338614344596863,
+      "step": 1337
+    },
+    {
+      "epoch": 0.7335526315789473,
+      "grad_norm": 0.4642828702926636,
+      "learning_rate": 3.640306448588996e-06,
+      "loss": 0.46210381388664246,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7341008771929824,
+      "grad_norm": 0.47700735926628113,
+      "learning_rate": 3.626319158260595e-06,
+      "loss": 0.45029470324516296,
+      "step": 1339
+    },
+    {
+      "epoch": 0.7346491228070176,
+      "grad_norm": 0.46553367376327515,
+      "learning_rate": 3.6123528377261663e-06,
+      "loss": 0.44773992896080017,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7351973684210527,
+      "grad_norm": 0.47391775250434875,
+      "learning_rate": 3.598407532935748e-06,
+      "loss": 0.4471183717250824,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7357456140350878,
+      "grad_norm": 0.4671850800514221,
+      "learning_rate": 3.5844832897702363e-06,
+      "loss": 0.454129695892334,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7362938596491229,
+      "grad_norm": 0.47181615233421326,
+      "learning_rate": 3.5705801540412268e-06,
+      "loss": 0.44526728987693787,
+      "step": 1343
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.4609008729457855,
+      "learning_rate": 3.556698171490871e-06,
+      "loss": 0.45439064502716064,
+      "step": 1344
+    },
+    {
+      "epoch": 0.737390350877193,
+      "grad_norm": 0.4642679691314697,
+      "learning_rate": 3.542837387791733e-06,
+      "loss": 0.4484332501888275,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7379385964912281,
+      "grad_norm": 0.47382134199142456,
+      "learning_rate": 3.5289978485466224e-06,
+      "loss": 0.44568178057670593,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7384868421052632,
+      "grad_norm": 0.5014953017234802,
+      "learning_rate": 3.5151795992884575e-06,
+      "loss": 0.4595361649990082,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7390350877192983,
+      "grad_norm": 0.4758802354335785,
+      "learning_rate": 3.501382685480116e-06,
+      "loss": 0.4526955485343933,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7395833333333334,
+      "grad_norm": 0.47585156559944153,
+      "learning_rate": 3.487607152514272e-06,
+      "loss": 0.44488176703453064,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7401315789473685,
+      "grad_norm": 0.5094396471977234,
+      "learning_rate": 3.473853045713255e-06,
+      "loss": 0.44131070375442505,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7406798245614035,
+      "grad_norm": 0.47419655323028564,
+      "learning_rate": 3.460120410328908e-06,
+      "loss": 0.44449982047080994,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7412280701754386,
+      "grad_norm": 0.47309622168540955,
+      "learning_rate": 3.4464092915424328e-06,
+      "loss": 0.4419247508049011,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7417763157894737,
+      "grad_norm": 0.46527430415153503,
+      "learning_rate": 3.4327197344642304e-06,
+      "loss": 0.4458807408809662,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7423245614035088,
+      "grad_norm": 0.47822335362434387,
+      "learning_rate": 3.419051784133773e-06,
+      "loss": 0.4569171369075775,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7428728070175439,
+      "grad_norm": 0.49528107047080994,
+      "learning_rate": 3.4054054855194395e-06,
+      "loss": 0.44940003752708435,
+      "step": 1355
+    },
+    {
+      "epoch": 0.743421052631579,
+      "grad_norm": 0.463140070438385,
+      "learning_rate": 3.3917808835183707e-06,
+      "loss": 0.45283064246177673,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7439692982456141,
+      "grad_norm": 0.46392157673835754,
+      "learning_rate": 3.37817802295633e-06,
+      "loss": 0.45993560552597046,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7445175438596491,
+      "grad_norm": 0.6057497262954712,
+      "learning_rate": 3.3645969485875528e-06,
+      "loss": 0.4528937041759491,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7450657894736842,
+      "grad_norm": 0.49526041746139526,
+      "learning_rate": 3.3510377050945853e-06,
+      "loss": 0.443331241607666,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7456140350877193,
+      "grad_norm": 0.5202312469482422,
+      "learning_rate": 3.337500337088162e-06,
+      "loss": 0.44786232709884644,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7461622807017544,
+      "grad_norm": 0.4547555148601532,
+      "learning_rate": 3.323984889107035e-06,
+      "loss": 0.44628724455833435,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7467105263157895,
+      "grad_norm": 0.4743717312812805,
+      "learning_rate": 3.3104914056178406e-06,
+      "loss": 0.46816253662109375,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7472587719298246,
+      "grad_norm": 0.46588048338890076,
+      "learning_rate": 3.2970199310149543e-06,
+      "loss": 0.4552823007106781,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7478070175438597,
+      "grad_norm": 0.4879978895187378,
+      "learning_rate": 3.283570509620344e-06,
+      "loss": 0.4521896243095398,
+      "step": 1364
+    },
+    {
+      "epoch": 0.7483552631578947,
+      "grad_norm": 0.4443570077419281,
+      "learning_rate": 3.2701431856834087e-06,
+      "loss": 0.45289620757102966,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7489035087719298,
+      "grad_norm": 0.4682408571243286,
+      "learning_rate": 3.2567380033808603e-06,
+      "loss": 0.4340660572052002,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7494517543859649,
+      "grad_norm": 0.5005453824996948,
+      "learning_rate": 3.2433550068165497e-06,
+      "loss": 0.45293864607810974,
+      "step": 1367
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4666679799556732,
+      "learning_rate": 3.2299942400213447e-06,
+      "loss": 0.44803494215011597,
+      "step": 1368
+    },
+    {
+      "epoch": 0.7505482456140351,
+      "grad_norm": 0.46357494592666626,
+      "learning_rate": 3.216655746952976e-06,
+      "loss": 0.44369348883628845,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7510964912280702,
+      "grad_norm": 0.44632741808891296,
+      "learning_rate": 3.203339571495887e-06,
+      "loss": 0.4500495195388794,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7516447368421053,
+      "grad_norm": 0.4712839126586914,
+      "learning_rate": 3.190045757461093e-06,
+      "loss": 0.45239824056625366,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7521929824561403,
+      "grad_norm": 0.49454811215400696,
+      "learning_rate": 3.1767743485860514e-06,
+      "loss": 0.4392992854118347,
+      "step": 1372
+    },
+    {
+      "epoch": 0.7527412280701754,
+      "grad_norm": 0.4636279046535492,
+      "learning_rate": 3.1635253885344884e-06,
+      "loss": 0.4610537886619568,
+      "step": 1373
+    },
+    {
+      "epoch": 0.7532894736842105,
+      "grad_norm": 0.49633049964904785,
+      "learning_rate": 3.1502989208962854e-06,
+      "loss": 0.44744524359703064,
+      "step": 1374
+    },
+    {
+      "epoch": 0.7538377192982456,
+      "grad_norm": 0.5043641924858093,
+      "learning_rate": 3.1370949891873213e-06,
+      "loss": 0.4373456835746765,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7543859649122807,
+      "grad_norm": 0.4852113425731659,
+      "learning_rate": 3.123913636849322e-06,
+      "loss": 0.45185375213623047,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7549342105263158,
+      "grad_norm": 0.48131412267684937,
+      "learning_rate": 3.1107549072497324e-06,
+      "loss": 0.4426053464412689,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7554824561403509,
+      "grad_norm": 0.4737791121006012,
+      "learning_rate": 3.097618843681558e-06,
+      "loss": 0.44092023372650146,
+      "step": 1378
+    },
+    {
+      "epoch": 0.7560307017543859,
+      "grad_norm": 0.4885827302932739,
+      "learning_rate": 3.084505489363254e-06,
+      "loss": 0.4620380401611328,
+      "step": 1379
+    },
+    {
+      "epoch": 0.756578947368421,
+      "grad_norm": 0.47208940982818604,
+      "learning_rate": 3.0714148874385376e-06,
+      "loss": 0.4508278965950012,
+      "step": 1380
+    },
+    {
+      "epoch": 0.7571271929824561,
+      "grad_norm": 0.4611866772174835,
+      "learning_rate": 3.058347080976276e-06,
+      "loss": 0.4432736337184906,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7576754385964912,
+      "grad_norm": 0.46872127056121826,
+      "learning_rate": 3.0453021129703465e-06,
+      "loss": 0.4664454460144043,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7582236842105263,
+      "grad_norm": 0.48894163966178894,
+      "learning_rate": 3.0322800263394725e-06,
+      "loss": 0.437408447265625,
+      "step": 1383
+    },
+    {
+      "epoch": 0.7587719298245614,
+      "grad_norm": 0.4994237422943115,
+      "learning_rate": 3.0192808639271065e-06,
+      "loss": 0.4648517966270447,
+      "step": 1384
+    },
+    {
+      "epoch": 0.7593201754385965,
+      "grad_norm": 0.47740432620048523,
+      "learning_rate": 3.0063046685012808e-06,
+      "loss": 0.4363187551498413,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7598684210526315,
+      "grad_norm": 0.480413019657135,
+      "learning_rate": 2.993351482754455e-06,
+      "loss": 0.44096410274505615,
+      "step": 1386
+    },
+    {
+      "epoch": 0.7604166666666666,
+      "grad_norm": 0.4761090874671936,
+      "learning_rate": 2.9804213493033883e-06,
+      "loss": 0.45182913541793823,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7609649122807017,
+      "grad_norm": 0.469472736120224,
+      "learning_rate": 2.967514310689006e-06,
+      "loss": 0.45764875411987305,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7615131578947368,
+      "grad_norm": 0.4679282307624817,
+      "learning_rate": 2.9546304093762357e-06,
+      "loss": 0.4410996437072754,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7620614035087719,
+      "grad_norm": 0.47340869903564453,
+      "learning_rate": 2.9417696877538913e-06,
+      "loss": 0.43983063101768494,
+      "step": 1390
+    },
+    {
+      "epoch": 0.762609649122807,
+      "grad_norm": 0.47689974308013916,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.4503350853919983,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7631578947368421,
+      "grad_norm": 0.5016065835952759,
+      "learning_rate": 2.9161179527542827e-06,
+      "loss": 0.44059401750564575,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7637061403508771,
+      "grad_norm": 0.4677375257015228,
+      "learning_rate": 2.903327023772766e-06,
+      "loss": 0.4515905976295471,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7642543859649122,
+      "grad_norm": 0.47397127747535706,
+      "learning_rate": 2.890559443272906e-06,
+      "loss": 0.44616538286209106,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7648026315789473,
+      "grad_norm": 0.5041709542274475,
+      "learning_rate": 2.877815253260814e-06,
+      "loss": 0.42689597606658936,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7653508771929824,
+      "grad_norm": 0.45834705233573914,
+      "learning_rate": 2.865094495665638e-06,
+      "loss": 0.45505937933921814,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7658991228070176,
+      "grad_norm": 0.460654079914093,
+      "learning_rate": 2.8523972123394405e-06,
+      "loss": 0.444200724363327,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7664473684210527,
+      "grad_norm": 0.4719720780849457,
+      "learning_rate": 2.839723445057049e-06,
+      "loss": 0.4539354741573334,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7669956140350878,
+      "grad_norm": 0.47663259506225586,
+      "learning_rate": 2.827073235515916e-06,
+      "loss": 0.4420117735862732,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7675438596491229,
+      "grad_norm": 0.540690541267395,
+      "learning_rate": 2.814446625335997e-06,
+      "loss": 0.46264857053756714,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7680921052631579,
+      "grad_norm": 0.48664623498916626,
+      "learning_rate": 2.801843656059602e-06,
+      "loss": 0.44607076048851013,
+      "step": 1401
+    },
+    {
+      "epoch": 0.768640350877193,
+      "grad_norm": 0.46868008375167847,
+      "learning_rate": 2.7892643691512555e-06,
+      "loss": 0.4463055729866028,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7691885964912281,
+      "grad_norm": 0.7354031205177307,
+      "learning_rate": 2.7767088059975732e-06,
+      "loss": 0.44075483083724976,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7697368421052632,
+      "grad_norm": 0.45407670736312866,
+      "learning_rate": 2.764177007907113e-06,
+      "loss": 0.4547209143638611,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7702850877192983,
+      "grad_norm": 0.4852718412876129,
+      "learning_rate": 2.7516690161102376e-06,
+      "loss": 0.45378145575523376,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7708333333333334,
+      "grad_norm": 0.4868534505367279,
+      "learning_rate": 2.7391848717590074e-06,
+      "loss": 0.43858057260513306,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7713815789473685,
+      "grad_norm": 0.4446372389793396,
+      "learning_rate": 2.726724615927003e-06,
+      "loss": 0.4574624001979828,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7719298245614035,
+      "grad_norm": 0.4679673910140991,
+      "learning_rate": 2.714288289609217e-06,
+      "loss": 0.44254884123802185,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7724780701754386,
+      "grad_norm": 0.4617109000682831,
+      "learning_rate": 2.701875933721909e-06,
+      "loss": 0.44806107878685,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7730263157894737,
+      "grad_norm": 0.45996034145355225,
+      "learning_rate": 2.6894875891024796e-06,
+      "loss": 0.4340808391571045,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7735745614035088,
+      "grad_norm": 0.4849942624568939,
+      "learning_rate": 2.677123296509334e-06,
+      "loss": 0.453971266746521,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7741228070175439,
+      "grad_norm": 0.4761378765106201,
+      "learning_rate": 2.6647830966217323e-06,
+      "loss": 0.4519736170768738,
+      "step": 1412
+    },
+    {
+      "epoch": 0.774671052631579,
+      "grad_norm": 0.4653758406639099,
+      "learning_rate": 2.6524670300396827e-06,
+      "loss": 0.4523582458496094,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7752192982456141,
+      "grad_norm": 0.489674836397171,
+      "learning_rate": 2.6401751372837815e-06,
+      "loss": 0.4477754235267639,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7757675438596491,
+      "grad_norm": 0.4749486446380615,
+      "learning_rate": 2.627907458795096e-06,
+      "loss": 0.43215465545654297,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7763157894736842,
+      "grad_norm": 0.4954405426979065,
+      "learning_rate": 2.6156640349350282e-06,
+      "loss": 0.44484657049179077,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7768640350877193,
+      "grad_norm": 0.4541327953338623,
+      "learning_rate": 2.603444905985184e-06,
+      "loss": 0.4564289450645447,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7774122807017544,
+      "grad_norm": 0.46660396456718445,
+      "learning_rate": 2.5912501121472287e-06,
+      "loss": 0.4459143877029419,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7779605263157895,
+      "grad_norm": 0.4727344512939453,
+      "learning_rate": 2.5790796935427744e-06,
+      "loss": 0.4507994055747986,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7785087719298246,
+      "grad_norm": 0.48547977209091187,
+      "learning_rate": 2.5669336902132236e-06,
+      "loss": 0.45748457312583923,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7790570175438597,
+      "grad_norm": 0.4882795214653015,
+      "learning_rate": 2.554812142119665e-06,
+      "loss": 0.4455465078353882,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7796052631578947,
+      "grad_norm": 0.47181499004364014,
+      "learning_rate": 2.542715089142723e-06,
+      "loss": 0.45293498039245605,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7801535087719298,
+      "grad_norm": 0.46868982911109924,
+      "learning_rate": 2.53064257108243e-06,
+      "loss": 0.4384647607803345,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7807017543859649,
+      "grad_norm": 0.5356707572937012,
+      "learning_rate": 2.518594627658092e-06,
+      "loss": 0.45708268880844116,
+      "step": 1424
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 0.4938961863517761,
+      "learning_rate": 2.506571298508176e-06,
+      "loss": 0.45950478315353394,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7817982456140351,
+      "grad_norm": 0.46023717522621155,
+      "learning_rate": 2.4945726231901535e-06,
+      "loss": 0.45347079634666443,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7823464912280702,
+      "grad_norm": 0.4642326235771179,
+      "learning_rate": 2.482598641180393e-06,
+      "loss": 0.44673532247543335,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7828947368421053,
+      "grad_norm": 0.49191659688949585,
+      "learning_rate": 2.470649391874017e-06,
+      "loss": 0.45227012038230896,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7834429824561403,
+      "grad_norm": 0.48225194215774536,
+      "learning_rate": 2.4587249145847757e-06,
+      "loss": 0.4629409909248352,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7839912280701754,
+      "grad_norm": 0.4709365665912628,
+      "learning_rate": 2.446825248544913e-06,
+      "loss": 0.44721364974975586,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7845394736842105,
+      "grad_norm": 0.4582255780696869,
+      "learning_rate": 2.434950432905053e-06,
+      "loss": 0.444847047328949,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7850877192982456,
+      "grad_norm": 0.47319498658180237,
+      "learning_rate": 2.4231005067340507e-06,
+      "loss": 0.4396522045135498,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7856359649122807,
+      "grad_norm": 0.47419747710227966,
+      "learning_rate": 2.411275509018878e-06,
+      "loss": 0.4477517008781433,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7861842105263158,
+      "grad_norm": 0.485272616147995,
+      "learning_rate": 2.3994754786644925e-06,
+      "loss": 0.43748196959495544,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7867324561403509,
+      "grad_norm": 0.48166120052337646,
+      "learning_rate": 2.387700454493703e-06,
+      "loss": 0.45969265699386597,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7872807017543859,
+      "grad_norm": 0.48745638132095337,
+      "learning_rate": 2.3759504752470463e-06,
+      "loss": 0.4459717273712158,
+      "step": 1436
+    },
+    {
+      "epoch": 0.787828947368421,
+      "grad_norm": 0.4823463261127472,
+      "learning_rate": 2.3642255795826654e-06,
+      "loss": 0.4502348303794861,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7883771929824561,
+      "grad_norm": 0.4792110025882721,
+      "learning_rate": 2.3525258060761734e-06,
+      "loss": 0.46335679292678833,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7889254385964912,
+      "grad_norm": 0.487899512052536,
+      "learning_rate": 2.3408511932205256e-06,
+      "loss": 0.46185988187789917,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.45009350776672363,
+      "learning_rate": 2.329201779425909e-06,
+      "loss": 0.4611009359359741,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7900219298245614,
+      "grad_norm": 0.4828180968761444,
+      "learning_rate": 2.3175776030195917e-06,
+      "loss": 0.45473694801330566,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7905701754385965,
+      "grad_norm": 0.4800177812576294,
+      "learning_rate": 2.3059787022458137e-06,
+      "loss": 0.43865078687667847,
+      "step": 1442
+    },
+    {
+      "epoch": 0.7911184210526315,
+      "grad_norm": 0.45250406861305237,
+      "learning_rate": 2.2944051152656588e-06,
+      "loss": 0.458981454372406,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7916666666666666,
+      "grad_norm": 0.4827856421470642,
+      "learning_rate": 2.2828568801569286e-06,
+      "loss": 0.4438856840133667,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7922149122807017,
+      "grad_norm": 0.4579552710056305,
+      "learning_rate": 2.271334034914009e-06,
+      "loss": 0.45474669337272644,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7927631578947368,
+      "grad_norm": 0.4525872468948364,
+      "learning_rate": 2.2598366174477527e-06,
+      "loss": 0.44240206480026245,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7933114035087719,
+      "grad_norm": 0.4654140770435333,
+      "learning_rate": 2.248364665585362e-06,
+      "loss": 0.4465499818325043,
+      "step": 1447
+    },
+    {
+      "epoch": 0.793859649122807,
+      "grad_norm": 0.4569734036922455,
+      "learning_rate": 2.236918217070244e-06,
+      "loss": 0.43812352418899536,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7944078947368421,
+      "grad_norm": 0.4839644730091095,
+      "learning_rate": 2.225497309561907e-06,
+      "loss": 0.4401845335960388,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7949561403508771,
+      "grad_norm": 0.49678677320480347,
+      "learning_rate": 2.2141019806358254e-06,
+      "loss": 0.4653699994087219,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7955043859649122,
+      "grad_norm": 0.48518845438957214,
+      "learning_rate": 2.2027322677833186e-06,
+      "loss": 0.44742369651794434,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7960526315789473,
+      "grad_norm": 0.46840372681617737,
+      "learning_rate": 2.191388208411421e-06,
+      "loss": 0.4364306628704071,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7966008771929824,
+      "grad_norm": 0.45687833428382874,
+      "learning_rate": 2.1800698398427754e-06,
+      "loss": 0.46097299456596375,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7971491228070176,
+      "grad_norm": 0.49113285541534424,
+      "learning_rate": 2.1687771993155006e-06,
+      "loss": 0.43846258521080017,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7976973684210527,
+      "grad_norm": 0.4738796651363373,
+      "learning_rate": 2.1575103239830554e-06,
+      "loss": 0.44409045577049255,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7982456140350878,
+      "grad_norm": 0.457188218832016,
+      "learning_rate": 2.146269250914147e-06,
+      "loss": 0.4441969394683838,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7987938596491229,
+      "grad_norm": 0.4704115092754364,
+      "learning_rate": 2.1350540170925795e-06,
+      "loss": 0.4427245557308197,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7993421052631579,
+      "grad_norm": 0.4717205762863159,
+      "learning_rate": 2.123864659417144e-06,
+      "loss": 0.42821723222732544,
+      "step": 1458
+    },
+    {
+      "epoch": 0.799890350877193,
+      "grad_norm": 0.46827948093414307,
+      "learning_rate": 2.1127012147015047e-06,
+      "loss": 0.4556030035018921,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8004385964912281,
+      "grad_norm": 0.49792206287384033,
+      "learning_rate": 2.101563719674071e-06,
+      "loss": 0.4553588330745697,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8009868421052632,
+      "grad_norm": 0.46498462557792664,
+      "learning_rate": 2.0904522109778667e-06,
+      "loss": 0.45313581824302673,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8015350877192983,
+      "grad_norm": 0.4481178820133209,
+      "learning_rate": 2.0793667251704296e-06,
+      "loss": 0.4471014738082886,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8020833333333334,
+      "grad_norm": 0.46041181683540344,
+      "learning_rate": 2.068307298723671e-06,
+      "loss": 0.45470839738845825,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8026315789473685,
+      "grad_norm": 0.4539787471294403,
+      "learning_rate": 2.0572739680237718e-06,
+      "loss": 0.45658087730407715,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8031798245614035,
+      "grad_norm": 0.458933562040329,
+      "learning_rate": 2.046266769371059e-06,
+      "loss": 0.460864782333374,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8037280701754386,
+      "grad_norm": 0.47983378171920776,
+      "learning_rate": 2.0352857389798765e-06,
+      "loss": 0.44127601385116577,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8042763157894737,
+      "grad_norm": 0.4826447665691376,
+      "learning_rate": 2.0243309129784727e-06,
+      "loss": 0.4633498787879944,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8048245614035088,
+      "grad_norm": 0.588762640953064,
+      "learning_rate": 2.01340232740889e-06,
+      "loss": 0.4519672989845276,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8053728070175439,
+      "grad_norm": 0.4594285786151886,
+      "learning_rate": 2.0025000182268297e-06,
+      "loss": 0.4531112313270569,
+      "step": 1469
+    },
+    {
+      "epoch": 0.805921052631579,
+      "grad_norm": 0.4814896285533905,
+      "learning_rate": 1.9916240213015458e-06,
+      "loss": 0.44493603706359863,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8064692982456141,
+      "grad_norm": 0.48193618655204773,
+      "learning_rate": 1.9807743724157292e-06,
+      "loss": 0.42968878149986267,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8070175438596491,
+      "grad_norm": 0.4833047688007355,
+      "learning_rate": 1.9699511072653733e-06,
+      "loss": 0.43403005599975586,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8075657894736842,
+      "grad_norm": 0.46741804480552673,
+      "learning_rate": 1.9591542614596715e-06,
+      "loss": 0.4390289783477783,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8081140350877193,
+      "grad_norm": 0.47882014513015747,
+      "learning_rate": 1.948383870520901e-06,
+      "loss": 0.43885254859924316,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8086622807017544,
+      "grad_norm": 0.4652293622493744,
+      "learning_rate": 1.937639969884293e-06,
+      "loss": 0.45536062121391296,
+      "step": 1475
+    },
+    {
+      "epoch": 0.8092105263157895,
+      "grad_norm": 0.48102349042892456,
+      "learning_rate": 1.926922594897932e-06,
+      "loss": 0.4474731385707855,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8097587719298246,
+      "grad_norm": 0.4696721136569977,
+      "learning_rate": 1.9162317808226228e-06,
+      "loss": 0.4380807876586914,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8103070175438597,
+      "grad_norm": 0.45643603801727295,
+      "learning_rate": 1.9055675628317926e-06,
+      "loss": 0.4397779405117035,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8108552631578947,
+      "grad_norm": 0.45802298188209534,
+      "learning_rate": 1.894929976011356e-06,
+      "loss": 0.4323575496673584,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8114035087719298,
+      "grad_norm": 0.4565486013889313,
+      "learning_rate": 1.8843190553596168e-06,
+      "loss": 0.42069071531295776,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8119517543859649,
+      "grad_norm": 0.49252307415008545,
+      "learning_rate": 1.8737348357871477e-06,
+      "loss": 0.4390181005001068,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 0.4725061058998108,
+      "learning_rate": 1.863177352116664e-06,
+      "loss": 0.4410583972930908,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8130482456140351,
+      "grad_norm": 0.4714313745498657,
+      "learning_rate": 1.8526466390829213e-06,
+      "loss": 0.43576177954673767,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8135964912280702,
+      "grad_norm": 0.4661250412464142,
+      "learning_rate": 1.8421427313326046e-06,
+      "loss": 0.4356091618537903,
+      "step": 1484
+    },
+    {
+      "epoch": 0.8141447368421053,
+      "grad_norm": 0.48059454560279846,
+      "learning_rate": 1.831665663424198e-06,
+      "loss": 0.44501131772994995,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8146929824561403,
+      "grad_norm": 0.46940878033638,
+      "learning_rate": 1.8212154698278906e-06,
+      "loss": 0.4514904022216797,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8152412280701754,
+      "grad_norm": 0.5343154668807983,
+      "learning_rate": 1.8107921849254495e-06,
+      "loss": 0.4419550895690918,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8157894736842105,
+      "grad_norm": 0.4681127071380615,
+      "learning_rate": 1.8003958430101087e-06,
+      "loss": 0.46082767844200134,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8163377192982456,
+      "grad_norm": 0.47236642241477966,
+      "learning_rate": 1.7900264782864552e-06,
+      "loss": 0.44913601875305176,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8168859649122807,
+      "grad_norm": 0.4529493451118469,
+      "learning_rate": 1.7796841248703277e-06,
+      "loss": 0.45329952239990234,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8174342105263158,
+      "grad_norm": 0.46615317463874817,
+      "learning_rate": 1.7693688167886947e-06,
+      "loss": 0.4398888945579529,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8179824561403509,
+      "grad_norm": 0.4682173430919647,
+      "learning_rate": 1.7590805879795358e-06,
+      "loss": 0.43605130910873413,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8185307017543859,
+      "grad_norm": 0.44849905371665955,
+      "learning_rate": 1.7488194722917484e-06,
+      "loss": 0.44999951124191284,
+      "step": 1493
+    },
+    {
+      "epoch": 0.819078947368421,
+      "grad_norm": 0.49639979004859924,
+      "learning_rate": 1.7385855034850185e-06,
+      "loss": 0.4317818582057953,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8196271929824561,
+      "grad_norm": 0.48981213569641113,
+      "learning_rate": 1.7283787152297194e-06,
+      "loss": 0.43808650970458984,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8201754385964912,
+      "grad_norm": 0.4760289788246155,
+      "learning_rate": 1.7181991411067989e-06,
+      "loss": 0.4499431550502777,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8207236842105263,
+      "grad_norm": 0.4522970914840698,
+      "learning_rate": 1.7080468146076745e-06,
+      "loss": 0.45054715871810913,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8212719298245614,
+      "grad_norm": 0.4766053855419159,
+      "learning_rate": 1.6979217691341054e-06,
+      "loss": 0.45946937799453735,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8218201754385965,
+      "grad_norm": 0.485436350107193,
+      "learning_rate": 1.687824037998107e-06,
+      "loss": 0.44628486037254333,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8223684210526315,
+      "grad_norm": 0.47168204188346863,
+      "learning_rate": 1.677753654421821e-06,
+      "loss": 0.4408082365989685,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8229166666666666,
+      "grad_norm": 0.478395015001297,
+      "learning_rate": 1.667710651537412e-06,
+      "loss": 0.4404899477958679,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8234649122807017,
+      "grad_norm": 0.4617457687854767,
+      "learning_rate": 1.6576950623869682e-06,
+      "loss": 0.4409525394439697,
+      "step": 1502
+    },
+    {
+      "epoch": 0.8240131578947368,
+      "grad_norm": 0.48547491431236267,
+      "learning_rate": 1.6477069199223839e-06,
+      "loss": 0.45365771651268005,
+      "step": 1503
+    },
+    {
+      "epoch": 0.8245614035087719,
+      "grad_norm": 0.46786943078041077,
+      "learning_rate": 1.6377462570052438e-06,
+      "loss": 0.44676321744918823,
+      "step": 1504
+    },
+    {
+      "epoch": 0.825109649122807,
+      "grad_norm": 0.4777480959892273,
+      "learning_rate": 1.6278131064067349e-06,
+      "loss": 0.44479823112487793,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8256578947368421,
+      "grad_norm": 0.4573366940021515,
+      "learning_rate": 1.6179075008075162e-06,
+      "loss": 0.440067857503891,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8262061403508771,
+      "grad_norm": 0.44294530153274536,
+      "learning_rate": 1.6080294727976288e-06,
+      "loss": 0.42653411626815796,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8267543859649122,
+      "grad_norm": 0.4719429016113281,
+      "learning_rate": 1.598179054876382e-06,
+      "loss": 0.46551793813705444,
+      "step": 1508
+    },
+    {
+      "epoch": 0.8273026315789473,
+      "grad_norm": 0.47400152683258057,
+      "learning_rate": 1.5883562794522423e-06,
+      "loss": 0.4564603269100189,
+      "step": 1509
+    },
+    {
+      "epoch": 0.8278508771929824,
+      "grad_norm": 0.4437645971775055,
+      "learning_rate": 1.5785611788427336e-06,
+      "loss": 0.4511438310146332,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8283991228070176,
+      "grad_norm": 0.47115546464920044,
+      "learning_rate": 1.568793785274323e-06,
+      "loss": 0.4462664723396301,
+      "step": 1511
+    },
+    {
+      "epoch": 0.8289473684210527,
+      "grad_norm": 0.4500909149646759,
+      "learning_rate": 1.559054130882327e-06,
+      "loss": 0.44786590337753296,
+      "step": 1512
+    },
+    {
+      "epoch": 0.8294956140350878,
+      "grad_norm": 0.4785374701023102,
+      "learning_rate": 1.549342247710799e-06,
+      "loss": 0.4540751874446869,
+      "step": 1513
+    },
+    {
+      "epoch": 0.8300438596491229,
+      "grad_norm": 0.4943836033344269,
+      "learning_rate": 1.5396581677124123e-06,
+      "loss": 0.4656469523906708,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8305921052631579,
+      "grad_norm": 0.4699077010154724,
+      "learning_rate": 1.5300019227483809e-06,
+      "loss": 0.4323770999908447,
+      "step": 1515
+    },
+    {
+      "epoch": 0.831140350877193,
+      "grad_norm": 0.48153141140937805,
+      "learning_rate": 1.5203735445883284e-06,
+      "loss": 0.4336276054382324,
+      "step": 1516
+    },
+    {
+      "epoch": 0.8316885964912281,
+      "grad_norm": 0.49159079790115356,
+      "learning_rate": 1.5107730649101948e-06,
+      "loss": 0.45548614859580994,
+      "step": 1517
+    },
+    {
+      "epoch": 0.8322368421052632,
+      "grad_norm": 0.49155518412590027,
+      "learning_rate": 1.5012005153001463e-06,
+      "loss": 0.4504159390926361,
+      "step": 1518
+    },
+    {
+      "epoch": 0.8327850877192983,
+      "grad_norm": 0.4667416214942932,
+      "learning_rate": 1.4916559272524422e-06,
+      "loss": 0.4470210075378418,
+      "step": 1519
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.46270886063575745,
+      "learning_rate": 1.4821393321693523e-06,
+      "loss": 0.4300989508628845,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8338815789473685,
+      "grad_norm": 0.4680396616458893,
+      "learning_rate": 1.4726507613610496e-06,
+      "loss": 0.44302886724472046,
+      "step": 1521
+    },
+    {
+      "epoch": 0.8344298245614035,
+      "grad_norm": 0.48521971702575684,
+      "learning_rate": 1.4631902460455005e-06,
+      "loss": 0.43804043531417847,
+      "step": 1522
+    },
+    {
+      "epoch": 0.8349780701754386,
+      "grad_norm": 0.4627128839492798,
+      "learning_rate": 1.4537578173483736e-06,
+      "loss": 0.4490930438041687,
+      "step": 1523
+    },
+    {
+      "epoch": 0.8355263157894737,
+      "grad_norm": 0.4822083115577698,
+      "learning_rate": 1.444353506302928e-06,
+      "loss": 0.4367152452468872,
+      "step": 1524
+    },
+    {
+      "epoch": 0.8360745614035088,
+      "grad_norm": 0.5016523599624634,
+      "learning_rate": 1.4349773438499115e-06,
+      "loss": 0.4512786865234375,
+      "step": 1525
+    },
+    {
+      "epoch": 0.8366228070175439,
+      "grad_norm": 0.4675818681716919,
+      "learning_rate": 1.4256293608374627e-06,
+      "loss": 0.43894022703170776,
+      "step": 1526
+    },
+    {
+      "epoch": 0.837171052631579,
+      "grad_norm": 0.4565467834472656,
+      "learning_rate": 1.4163095880210131e-06,
+      "loss": 0.43741410970687866,
+      "step": 1527
+    },
+    {
+      "epoch": 0.8377192982456141,
+      "grad_norm": 0.45361560583114624,
+      "learning_rate": 1.4070180560631708e-06,
+      "loss": 0.45772549510002136,
+      "step": 1528
+    },
+    {
+      "epoch": 0.8382675438596491,
+      "grad_norm": 0.4735342860221863,
+      "learning_rate": 1.397754795533639e-06,
+      "loss": 0.44077205657958984,
+      "step": 1529
+    },
+    {
+      "epoch": 0.8388157894736842,
+      "grad_norm": 0.48649099469184875,
+      "learning_rate": 1.3885198369091057e-06,
+      "loss": 0.4449783265590668,
+      "step": 1530
+    },
+    {
+      "epoch": 0.8393640350877193,
+      "grad_norm": 0.4635789096355438,
+      "learning_rate": 1.3793132105731388e-06,
+      "loss": 0.4570218324661255,
+      "step": 1531
+    },
+    {
+      "epoch": 0.8399122807017544,
+      "grad_norm": 0.49008285999298096,
+      "learning_rate": 1.3701349468160908e-06,
+      "loss": 0.435431569814682,
+      "step": 1532
+    },
+    {
+      "epoch": 0.8404605263157895,
+      "grad_norm": 0.4439835846424103,
+      "learning_rate": 1.360985075835004e-06,
+      "loss": 0.44734299182891846,
+      "step": 1533
+    },
+    {
+      "epoch": 0.8410087719298246,
+      "grad_norm": 0.46893739700317383,
+      "learning_rate": 1.3518636277335085e-06,
+      "loss": 0.45212846994400024,
+      "step": 1534
+    },
+    {
+      "epoch": 0.8415570175438597,
+      "grad_norm": 0.4498082995414734,
+      "learning_rate": 1.3427706325217137e-06,
+      "loss": 0.4477176070213318,
+      "step": 1535
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.4555201828479767,
+      "learning_rate": 1.333706120116126e-06,
+      "loss": 0.44618985056877136,
+      "step": 1536
+    },
+    {
+      "epoch": 0.8426535087719298,
+      "grad_norm": 0.47258657217025757,
+      "learning_rate": 1.324670120339535e-06,
+      "loss": 0.45410165190696716,
+      "step": 1537
+    },
+    {
+      "epoch": 0.8432017543859649,
+      "grad_norm": 0.457660973072052,
+      "learning_rate": 1.3156626629209224e-06,
+      "loss": 0.45003390312194824,
+      "step": 1538
+    },
+    {
+      "epoch": 0.84375,
+      "grad_norm": 0.46227186918258667,
+      "learning_rate": 1.3066837774953667e-06,
+      "loss": 0.4546055495738983,
+      "step": 1539
+    },
+    {
+      "epoch": 0.8442982456140351,
+      "grad_norm": 0.4478152394294739,
+      "learning_rate": 1.2977334936039454e-06,
+      "loss": 0.43452781438827515,
+      "step": 1540
+    },
+    {
+      "epoch": 0.8448464912280702,
+      "grad_norm": 0.4613502621650696,
+      "learning_rate": 1.28881184069363e-06,
+      "loss": 0.44267815351486206,
+      "step": 1541
+    },
+    {
+      "epoch": 0.8453947368421053,
+      "grad_norm": 0.5708100199699402,
+      "learning_rate": 1.2799188481171931e-06,
+      "loss": 0.4500243365764618,
+      "step": 1542
+    },
+    {
+      "epoch": 0.8459429824561403,
+      "grad_norm": 0.48053842782974243,
+      "learning_rate": 1.2710545451331203e-06,
+      "loss": 0.42664283514022827,
+      "step": 1543
+    },
+    {
+      "epoch": 0.8464912280701754,
+      "grad_norm": 0.45715469121932983,
+      "learning_rate": 1.262218960905498e-06,
+      "loss": 0.4444156587123871,
+      "step": 1544
+    },
+    {
+      "epoch": 0.8470394736842105,
+      "grad_norm": 0.4761993885040283,
+      "learning_rate": 1.2534121245039343e-06,
+      "loss": 0.4384285807609558,
+      "step": 1545
+    },
+    {
+      "epoch": 0.8475877192982456,
+      "grad_norm": 0.4799306094646454,
+      "learning_rate": 1.2446340649034517e-06,
+      "loss": 0.4414188265800476,
+      "step": 1546
+    },
+    {
+      "epoch": 0.8481359649122807,
+      "grad_norm": 0.4641325771808624,
+      "learning_rate": 1.2358848109843936e-06,
+      "loss": 0.45236217975616455,
+      "step": 1547
+    },
+    {
+      "epoch": 0.8486842105263158,
+      "grad_norm": 0.4714052677154541,
+      "learning_rate": 1.2271643915323316e-06,
+      "loss": 0.44093751907348633,
+      "step": 1548
+    },
+    {
+      "epoch": 0.8492324561403509,
+      "grad_norm": 0.4963100850582123,
+      "learning_rate": 1.2184728352379715e-06,
+      "loss": 0.44910988211631775,
+      "step": 1549
+    },
+    {
+      "epoch": 0.8497807017543859,
+      "grad_norm": 0.4910122752189636,
+      "learning_rate": 1.209810170697061e-06,
+      "loss": 0.4461553692817688,
+      "step": 1550
+    },
+    {
+      "epoch": 0.850328947368421,
+      "grad_norm": 0.474320650100708,
+      "learning_rate": 1.201176426410282e-06,
+      "loss": 0.4358775019645691,
+      "step": 1551
+    },
+    {
+      "epoch": 0.8508771929824561,
+      "grad_norm": 0.4582522511482239,
+      "learning_rate": 1.1925716307831792e-06,
+      "loss": 0.45600447058677673,
+      "step": 1552
+    },
+    {
+      "epoch": 0.8514254385964912,
+      "grad_norm": 0.45246538519859314,
+      "learning_rate": 1.1839958121260464e-06,
+      "loss": 0.44419237971305847,
+      "step": 1553
+    },
+    {
+      "epoch": 0.8519736842105263,
+      "grad_norm": 0.4695712625980377,
+      "learning_rate": 1.175448998653842e-06,
+      "loss": 0.4544498324394226,
+      "step": 1554
+    },
+    {
+      "epoch": 0.8525219298245614,
+      "grad_norm": 0.46356385946273804,
+      "learning_rate": 1.1669312184860982e-06,
+      "loss": 0.45395684242248535,
+      "step": 1555
+    },
+    {
+      "epoch": 0.8530701754385965,
+      "grad_norm": 0.4739828109741211,
+      "learning_rate": 1.1584424996468269e-06,
+      "loss": 0.4454335570335388,
+      "step": 1556
+    },
+    {
+      "epoch": 0.8536184210526315,
+      "grad_norm": 0.4642820358276367,
+      "learning_rate": 1.1499828700644212e-06,
+      "loss": 0.445566862821579,
+      "step": 1557
+    },
+    {
+      "epoch": 0.8541666666666666,
+      "grad_norm": 0.5376269817352295,
+      "learning_rate": 1.1415523575715758e-06,
+      "loss": 0.44570326805114746,
+      "step": 1558
+    },
+    {
+      "epoch": 0.8547149122807017,
+      "grad_norm": 0.47028419375419617,
+      "learning_rate": 1.1331509899051828e-06,
+      "loss": 0.4468782842159271,
+      "step": 1559
+    },
+    {
+      "epoch": 0.8552631578947368,
+      "grad_norm": 0.4601708650588989,
+      "learning_rate": 1.124778794706245e-06,
+      "loss": 0.4463876485824585,
+      "step": 1560
+    },
+    {
+      "epoch": 0.8558114035087719,
+      "grad_norm": 0.4608907699584961,
+      "learning_rate": 1.116435799519795e-06,
+      "loss": 0.4424273669719696,
+      "step": 1561
+    },
+    {
+      "epoch": 0.856359649122807,
+      "grad_norm": 0.4768998622894287,
+      "learning_rate": 1.1081220317947871e-06,
+      "loss": 0.4271972179412842,
+      "step": 1562
+    },
+    {
+      "epoch": 0.8569078947368421,
+      "grad_norm": 0.44866839051246643,
+      "learning_rate": 1.0998375188840148e-06,
+      "loss": 0.4346953332424164,
+      "step": 1563
+    },
+    {
+      "epoch": 0.8574561403508771,
+      "grad_norm": 0.48162323236465454,
+      "learning_rate": 1.0915822880440309e-06,
+      "loss": 0.4489680528640747,
+      "step": 1564
+    },
+    {
+      "epoch": 0.8580043859649122,
+      "grad_norm": 0.45125117897987366,
+      "learning_rate": 1.0833563664350355e-06,
+      "loss": 0.4487599730491638,
+      "step": 1565
+    },
+    {
+      "epoch": 0.8585526315789473,
+      "grad_norm": 0.47349071502685547,
+      "learning_rate": 1.0751597811208104e-06,
+      "loss": 0.4338189363479614,
+      "step": 1566
+    },
+    {
+      "epoch": 0.8591008771929824,
+      "grad_norm": 0.4895882308483124,
+      "learning_rate": 1.066992559068616e-06,
+      "loss": 0.44717928767204285,
+      "step": 1567
+    },
+    {
+      "epoch": 0.8596491228070176,
+      "grad_norm": 0.4759470224380493,
+      "learning_rate": 1.0588547271491033e-06,
+      "loss": 0.42937859892845154,
+      "step": 1568
+    },
+    {
+      "epoch": 0.8601973684210527,
+      "grad_norm": 0.4551708400249481,
+      "learning_rate": 1.0507463121362283e-06,
+      "loss": 0.43980643153190613,
+      "step": 1569
+    },
+    {
+      "epoch": 0.8607456140350878,
+      "grad_norm": 0.4665743112564087,
+      "learning_rate": 1.0426673407071674e-06,
+      "loss": 0.4547693729400635,
+      "step": 1570
+    },
+    {
+      "epoch": 0.8612938596491229,
+      "grad_norm": 0.46718016266822815,
+      "learning_rate": 1.0346178394422203e-06,
+      "loss": 0.42919105291366577,
+      "step": 1571
+    },
+    {
+      "epoch": 0.8618421052631579,
+      "grad_norm": 0.48040589690208435,
+      "learning_rate": 1.0265978348247319e-06,
+      "loss": 0.4436418414115906,
+      "step": 1572
+    },
+    {
+      "epoch": 0.862390350877193,
+      "grad_norm": 0.46300461888313293,
+      "learning_rate": 1.0186073532410046e-06,
+      "loss": 0.4462057054042816,
+      "step": 1573
+    },
+    {
+      "epoch": 0.8629385964912281,
+      "grad_norm": 0.46362096071243286,
+      "learning_rate": 1.0106464209802013e-06,
+      "loss": 0.44162052869796753,
+      "step": 1574
+    },
+    {
+      "epoch": 0.8634868421052632,
+      "grad_norm": 0.4580335319042206,
+      "learning_rate": 1.0027150642342664e-06,
+      "loss": 0.4479283094406128,
+      "step": 1575
+    },
+    {
+      "epoch": 0.8640350877192983,
+      "grad_norm": 0.48166683316230774,
+      "learning_rate": 9.94813309097844e-07,
+      "loss": 0.44494980573654175,
+      "step": 1576
+    },
+    {
+      "epoch": 0.8645833333333334,
+      "grad_norm": 0.4515690803527832,
+      "learning_rate": 9.869411815681861e-07,
+      "loss": 0.4551948308944702,
+      "step": 1577
+    },
+    {
+      "epoch": 0.8651315789473685,
+      "grad_norm": 0.4830446243286133,
+      "learning_rate": 9.790987075450652e-07,
+      "loss": 0.4476730227470398,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8656798245614035,
+      "grad_norm": 0.4780424237251282,
+      "learning_rate": 9.712859128306906e-07,
+      "loss": 0.4410339295864105,
+      "step": 1579
+    },
+    {
+      "epoch": 0.8662280701754386,
+      "grad_norm": 0.46548083424568176,
+      "learning_rate": 9.635028231296328e-07,
+      "loss": 0.4511341452598572,
+      "step": 1580
+    },
+    {
+      "epoch": 0.8667763157894737,
+      "grad_norm": 0.47102057933807373,
+      "learning_rate": 9.557494640487197e-07,
+      "loss": 0.4316350221633911,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8673245614035088,
+      "grad_norm": 0.4755796194076538,
+      "learning_rate": 9.480258610969739e-07,
+      "loss": 0.43706297874450684,
+      "step": 1582
+    },
+    {
+      "epoch": 0.8678728070175439,
+      "grad_norm": 0.46922212839126587,
+      "learning_rate": 9.403320396855153e-07,
+      "loss": 0.44665640592575073,
+      "step": 1583
+    },
+    {
+      "epoch": 0.868421052631579,
+      "grad_norm": 0.44689786434173584,
+      "learning_rate": 9.326680251274778e-07,
+      "loss": 0.43797099590301514,
+      "step": 1584
+    },
+    {
+      "epoch": 0.8689692982456141,
+      "grad_norm": 0.4606018364429474,
+      "learning_rate": 9.250338426379301e-07,
+      "loss": 0.4488024115562439,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8695175438596491,
+      "grad_norm": 0.46374639868736267,
+      "learning_rate": 9.174295173337965e-07,
+      "loss": 0.44695234298706055,
+      "step": 1586
+    },
+    {
+      "epoch": 0.8700657894736842,
+      "grad_norm": 0.45890888571739197,
+      "learning_rate": 9.098550742337598e-07,
+      "loss": 0.4423515200614929,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8706140350877193,
+      "grad_norm": 0.472668319940567,
+      "learning_rate": 9.023105382581976e-07,
+      "loss": 0.4320164620876312,
+      "step": 1588
+    },
+    {
+      "epoch": 0.8711622807017544,
+      "grad_norm": 0.4817354679107666,
+      "learning_rate": 8.94795934229089e-07,
+      "loss": 0.43965157866477966,
+      "step": 1589
+    },
+    {
+      "epoch": 0.8717105263157895,
+      "grad_norm": 0.4685409963130951,
+      "learning_rate": 8.873112868699329e-07,
+      "loss": 0.4492000341415405,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8722587719298246,
+      "grad_norm": 0.46073803305625916,
+      "learning_rate": 8.798566208056669e-07,
+      "loss": 0.4389263987541199,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8728070175438597,
+      "grad_norm": 0.536866307258606,
+      "learning_rate": 8.724319605625942e-07,
+      "loss": 0.4503793716430664,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8733552631578947,
+      "grad_norm": 0.46781212091445923,
+      "learning_rate": 8.650373305682968e-07,
+      "loss": 0.4543306827545166,
+      "step": 1593
+    },
+    {
+      "epoch": 0.8739035087719298,
+      "grad_norm": 0.45546168088912964,
+      "learning_rate": 8.576727551515473e-07,
+      "loss": 0.45232880115509033,
+      "step": 1594
+    },
+    {
+      "epoch": 0.8744517543859649,
+      "grad_norm": 0.46312063932418823,
+      "learning_rate": 8.503382585422482e-07,
+      "loss": 0.4270923137664795,
+      "step": 1595
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.4830167889595032,
+      "learning_rate": 8.430338648713332e-07,
+      "loss": 0.4483802914619446,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8755482456140351,
+      "grad_norm": 0.4733735918998718,
+      "learning_rate": 8.357595981706934e-07,
+      "loss": 0.4472507834434509,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8760964912280702,
+      "grad_norm": 0.46488088369369507,
+      "learning_rate": 8.285154823731101e-07,
+      "loss": 0.45538365840911865,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8766447368421053,
+      "grad_norm": 0.4749467074871063,
+      "learning_rate": 8.213015413121584e-07,
+      "loss": 0.44120535254478455,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8771929824561403,
+      "grad_norm": 0.47514861822128296,
+      "learning_rate": 8.141177987221394e-07,
+      "loss": 0.4361691176891327,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8777412280701754,
+      "grad_norm": 0.47822439670562744,
+      "learning_rate": 8.069642782379994e-07,
+      "loss": 0.4516603648662567,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8782894736842105,
+      "grad_norm": 0.4655228853225708,
+      "learning_rate": 7.998410033952497e-07,
+      "loss": 0.43014663457870483,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8788377192982456,
+      "grad_norm": 0.4764757752418518,
+      "learning_rate": 7.927479976298957e-07,
+      "loss": 0.4478931427001953,
+      "step": 1603
+    },
+    {
+      "epoch": 0.8793859649122807,
+      "grad_norm": 0.46074727177619934,
+      "learning_rate": 7.856852842783546e-07,
+      "loss": 0.44595688581466675,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8799342105263158,
+      "grad_norm": 0.4583260416984558,
+      "learning_rate": 7.786528865773779e-07,
+      "loss": 0.4490140676498413,
+      "step": 1605
+    },
+    {
+      "epoch": 0.8804824561403509,
+      "grad_norm": 0.4848959445953369,
+      "learning_rate": 7.716508276639756e-07,
+      "loss": 0.44253939390182495,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8810307017543859,
+      "grad_norm": 0.4748147130012512,
+      "learning_rate": 7.646791305753476e-07,
+      "loss": 0.45834192633628845,
+      "step": 1607
+    },
+    {
+      "epoch": 0.881578947368421,
+      "grad_norm": 0.4740658402442932,
+      "learning_rate": 7.577378182487927e-07,
+      "loss": 0.43871140480041504,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8821271929824561,
+      "grad_norm": 0.4656429886817932,
+      "learning_rate": 7.508269135216495e-07,
+      "loss": 0.43657320737838745,
+      "step": 1609
+    },
+    {
+      "epoch": 0.8826754385964912,
+      "grad_norm": 0.5353845357894897,
+      "learning_rate": 7.439464391312102e-07,
+      "loss": 0.45518967509269714,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8832236842105263,
+      "grad_norm": 0.4724985659122467,
+      "learning_rate": 7.370964177146511e-07,
+      "loss": 0.43806204199790955,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8837719298245614,
+      "grad_norm": 0.44743263721466064,
+      "learning_rate": 7.302768718089548e-07,
+      "loss": 0.4636176526546478,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8843201754385965,
+      "grad_norm": 0.45549145340919495,
+      "learning_rate": 7.234878238508358e-07,
+      "loss": 0.4512561559677124,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8848684210526315,
+      "grad_norm": 0.4739094078540802,
+      "learning_rate": 7.167292961766726e-07,
+      "loss": 0.44614070653915405,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8854166666666666,
+      "grad_norm": 0.5969752669334412,
+      "learning_rate": 7.100013110224313e-07,
+      "loss": 0.44239649176597595,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8859649122807017,
+      "grad_norm": 0.46936947107315063,
+      "learning_rate": 7.033038905235845e-07,
+      "loss": 0.44706910848617554,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8865131578947368,
+      "grad_norm": 0.4682731032371521,
+      "learning_rate": 6.966370567150515e-07,
+      "loss": 0.4575401544570923,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8870614035087719,
+      "grad_norm": 0.4480043351650238,
+      "learning_rate": 6.900008315311147e-07,
+      "loss": 0.4444824159145355,
+      "step": 1618
+    },
+    {
+      "epoch": 0.887609649122807,
+      "grad_norm": 0.4583923816680908,
+      "learning_rate": 6.833952368053565e-07,
+      "loss": 0.44810950756073,
+      "step": 1619
+    },
+    {
+      "epoch": 0.8881578947368421,
+      "grad_norm": 0.4623442590236664,
+      "learning_rate": 6.768202942705837e-07,
+      "loss": 0.4399760663509369,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8887061403508771,
+      "grad_norm": 0.4465370774269104,
+      "learning_rate": 6.702760255587504e-07,
+      "loss": 0.45121389627456665,
+      "step": 1621
+    },
+    {
+      "epoch": 0.8892543859649122,
+      "grad_norm": 0.4720335006713867,
+      "learning_rate": 6.637624522008934e-07,
+      "loss": 0.4361550211906433,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8898026315789473,
+      "grad_norm": 0.4659505784511566,
+      "learning_rate": 6.572795956270661e-07,
+      "loss": 0.45230960845947266,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8903508771929824,
+      "grad_norm": 0.4665406346321106,
+      "learning_rate": 6.508274771662526e-07,
+      "loss": 0.43126940727233887,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8908991228070176,
+      "grad_norm": 0.46323761343955994,
+      "learning_rate": 6.444061180463135e-07,
+      "loss": 0.43909797072410583,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8914473684210527,
+      "grad_norm": 0.44538217782974243,
+      "learning_rate": 6.380155393939092e-07,
+      "loss": 0.4441220760345459,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8919956140350878,
+      "grad_norm": 0.46254977583885193,
+      "learning_rate": 6.316557622344266e-07,
+      "loss": 0.4325595796108246,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8925438596491229,
+      "grad_norm": 0.434124618768692,
+      "learning_rate": 6.253268074919139e-07,
+      "loss": 0.4580111503601074,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8930921052631579,
+      "grad_norm": 0.4571579396724701,
+      "learning_rate": 6.190286959890157e-07,
+      "loss": 0.45825982093811035,
+      "step": 1629
+    },
+    {
+      "epoch": 0.893640350877193,
+      "grad_norm": 0.46914803981781006,
+      "learning_rate": 6.127614484468991e-07,
+      "loss": 0.4454300105571747,
+      "step": 1630
+    },
+    {
+      "epoch": 0.8941885964912281,
+      "grad_norm": 0.4706801474094391,
+      "learning_rate": 6.065250854851834e-07,
+      "loss": 0.4399760663509369,
+      "step": 1631
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.45345693826675415,
+      "learning_rate": 6.003196276218815e-07,
+      "loss": 0.42961764335632324,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8952850877192983,
+      "grad_norm": 0.4495810270309448,
+      "learning_rate": 5.941450952733219e-07,
+      "loss": 0.43326735496520996,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8958333333333334,
+      "grad_norm": 0.45630401372909546,
+      "learning_rate": 5.880015087540858e-07,
+      "loss": 0.4345873296260834,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8963815789473685,
+      "grad_norm": 0.4689438045024872,
+      "learning_rate": 5.818888882769469e-07,
+      "loss": 0.4507316052913666,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8969298245614035,
+      "grad_norm": 0.46279099583625793,
+      "learning_rate": 5.758072539527937e-07,
+      "loss": 0.4242330491542816,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8974780701754386,
+      "grad_norm": 0.4933059513568878,
+      "learning_rate": 5.697566257905685e-07,
+      "loss": 0.4421001374721527,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8980263157894737,
+      "grad_norm": 0.4722570776939392,
+      "learning_rate": 5.63737023697205e-07,
+      "loss": 0.43669748306274414,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8985745614035088,
+      "grad_norm": 0.4798142611980438,
+      "learning_rate": 5.577484674775535e-07,
+      "loss": 0.4351981580257416,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8991228070175439,
+      "grad_norm": 0.4653114080429077,
+      "learning_rate": 5.517909768343255e-07,
+      "loss": 0.44025319814682007,
+      "step": 1640
+    },
+    {
+      "epoch": 0.899671052631579,
+      "grad_norm": 0.4510413408279419,
+      "learning_rate": 5.458645713680255e-07,
+      "loss": 0.44666802883148193,
+      "step": 1641
+    },
+    {
+      "epoch": 0.9002192982456141,
+      "grad_norm": 0.4883483946323395,
+      "learning_rate": 5.399692705768844e-07,
+      "loss": 0.43601131439208984,
+      "step": 1642
+    },
+    {
+      "epoch": 0.9007675438596491,
+      "grad_norm": 0.501358151435852,
+      "learning_rate": 5.341050938567971e-07,
+      "loss": 0.44397610425949097,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9013157894736842,
+      "grad_norm": 0.4486638605594635,
+      "learning_rate": 5.28272060501256e-07,
+      "loss": 0.43776631355285645,
+      "step": 1644
+    },
+    {
+      "epoch": 0.9018640350877193,
+      "grad_norm": 0.4545532166957855,
+      "learning_rate": 5.224701897012941e-07,
+      "loss": 0.44379138946533203,
+      "step": 1645
+    },
+    {
+      "epoch": 0.9024122807017544,
+      "grad_norm": 0.46450698375701904,
+      "learning_rate": 5.166995005454167e-07,
+      "loss": 0.4377807080745697,
+      "step": 1646
+    },
+    {
+      "epoch": 0.9029605263157895,
+      "grad_norm": 0.46801090240478516,
+      "learning_rate": 5.109600120195368e-07,
+      "loss": 0.4503188133239746,
+      "step": 1647
+    },
+    {
+      "epoch": 0.9035087719298246,
+      "grad_norm": 0.479351282119751,
+      "learning_rate": 5.052517430069204e-07,
+      "loss": 0.44086822867393494,
+      "step": 1648
+    },
+    {
+      "epoch": 0.9040570175438597,
+      "grad_norm": 0.46688783168792725,
+      "learning_rate": 4.995747122881134e-07,
+      "loss": 0.4453203082084656,
+      "step": 1649
+    },
+    {
+      "epoch": 0.9046052631578947,
+      "grad_norm": 0.4640738368034363,
+      "learning_rate": 4.939289385408896e-07,
+      "loss": 0.4435104727745056,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9051535087719298,
+      "grad_norm": 0.490331768989563,
+      "learning_rate": 4.883144403401852e-07,
+      "loss": 0.44754600524902344,
+      "step": 1651
+    },
+    {
+      "epoch": 0.9057017543859649,
+      "grad_norm": 0.4848006069660187,
+      "learning_rate": 4.827312361580383e-07,
+      "loss": 0.44139939546585083,
+      "step": 1652
+    },
+    {
+      "epoch": 0.90625,
+      "grad_norm": 0.48309126496315,
+      "learning_rate": 4.771793443635254e-07,
+      "loss": 0.4521448016166687,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9067982456140351,
+      "grad_norm": 0.46978119015693665,
+      "learning_rate": 4.716587832227071e-07,
+      "loss": 0.43720659613609314,
+      "step": 1654
+    },
+    {
+      "epoch": 0.9073464912280702,
+      "grad_norm": 0.47674959897994995,
+      "learning_rate": 4.6616957089856143e-07,
+      "loss": 0.4457699954509735,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9078947368421053,
+      "grad_norm": 0.4758906960487366,
+      "learning_rate": 4.6071172545092414e-07,
+      "loss": 0.4408598840236664,
+      "step": 1656
+    },
+    {
+      "epoch": 0.9084429824561403,
+      "grad_norm": 0.4686860740184784,
+      "learning_rate": 4.552852648364414e-07,
+      "loss": 0.4513329267501831,
+      "step": 1657
+    },
+    {
+      "epoch": 0.9089912280701754,
+      "grad_norm": 0.4459623396396637,
+      "learning_rate": 4.4989020690849315e-07,
+      "loss": 0.443990558385849,
+      "step": 1658
+    },
+    {
+      "epoch": 0.9095394736842105,
+      "grad_norm": 0.46643540263175964,
+      "learning_rate": 4.44526569417143e-07,
+      "loss": 0.443215548992157,
+      "step": 1659
+    },
+    {
+      "epoch": 0.9100877192982456,
+      "grad_norm": 0.47309648990631104,
+      "learning_rate": 4.391943700090839e-07,
+      "loss": 0.4382048547267914,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9106359649122807,
+      "grad_norm": 0.4757229685783386,
+      "learning_rate": 4.338936262275717e-07,
+      "loss": 0.443150132894516,
+      "step": 1661
+    },
+    {
+      "epoch": 0.9111842105263158,
+      "grad_norm": 0.48588597774505615,
+      "learning_rate": 4.286243555123737e-07,
+      "loss": 0.43010038137435913,
+      "step": 1662
+    },
+    {
+      "epoch": 0.9117324561403509,
+      "grad_norm": 0.4803837239742279,
+      "learning_rate": 4.2338657519970903e-07,
+      "loss": 0.44262608885765076,
+      "step": 1663
+    },
+    {
+      "epoch": 0.9122807017543859,
+      "grad_norm": 0.45542028546333313,
+      "learning_rate": 4.1818030252218976e-07,
+      "loss": 0.45058876276016235,
+      "step": 1664
+    },
+    {
+      "epoch": 0.912828947368421,
+      "grad_norm": 0.46372881531715393,
+      "learning_rate": 4.1300555460876415e-07,
+      "loss": 0.4500831365585327,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9133771929824561,
+      "grad_norm": 0.4932945668697357,
+      "learning_rate": 4.0786234848466775e-07,
+      "loss": 0.43038302659988403,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9139254385964912,
+      "grad_norm": 0.4446896016597748,
+      "learning_rate": 4.0275070107135605e-07,
+      "loss": 0.4467035233974457,
+      "step": 1667
+    },
+    {
+      "epoch": 0.9144736842105263,
+      "grad_norm": 0.47738441824913025,
+      "learning_rate": 3.976706291864596e-07,
+      "loss": 0.4430491328239441,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9150219298245614,
+      "grad_norm": 0.4746912121772766,
+      "learning_rate": 3.926221495437199e-07,
+      "loss": 0.4459044337272644,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9155701754385965,
+      "grad_norm": 0.4573158919811249,
+      "learning_rate": 3.8760527875293943e-07,
+      "loss": 0.4443172216415405,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9161184210526315,
+      "grad_norm": 0.46117162704467773,
+      "learning_rate": 3.826200333199237e-07,
+      "loss": 0.4394051432609558,
+      "step": 1671
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 0.463273823261261,
+      "learning_rate": 3.7766642964643165e-07,
+      "loss": 0.44520771503448486,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9172149122807017,
+      "grad_norm": 0.47583577036857605,
+      "learning_rate": 3.7274448403011975e-07,
+      "loss": 0.4473021924495697,
+      "step": 1673
+    },
+    {
+      "epoch": 0.9177631578947368,
+      "grad_norm": 0.47658321261405945,
+      "learning_rate": 3.678542126644813e-07,
+      "loss": 0.43678218126296997,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9183114035087719,
+      "grad_norm": 0.4604431688785553,
+      "learning_rate": 3.6299563163880836e-07,
+      "loss": 0.45016664266586304,
+      "step": 1675
+    },
+    {
+      "epoch": 0.918859649122807,
+      "grad_norm": 0.4587794542312622,
+      "learning_rate": 3.5816875693812316e-07,
+      "loss": 0.4506627917289734,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9194078947368421,
+      "grad_norm": 0.4732038080692291,
+      "learning_rate": 3.5337360444313354e-07,
+      "loss": 0.4495949149131775,
+      "step": 1677
+    },
+    {
+      "epoch": 0.9199561403508771,
+      "grad_norm": 0.4442287087440491,
+      "learning_rate": 3.486101899301797e-07,
+      "loss": 0.4385918974876404,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9205043859649122,
+      "grad_norm": 0.46998754143714905,
+      "learning_rate": 3.438785290711855e-07,
+      "loss": 0.44074422121047974,
+      "step": 1679
+    },
+    {
+      "epoch": 0.9210526315789473,
+      "grad_norm": 0.5042349696159363,
+      "learning_rate": 3.3917863743359815e-07,
+      "loss": 0.45951956510543823,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9216008771929824,
+      "grad_norm": 0.4709184467792511,
+      "learning_rate": 3.345105304803431e-07,
+      "loss": 0.4513796269893646,
+      "step": 1681
+    },
+    {
+      "epoch": 0.9221491228070176,
+      "grad_norm": 0.48918354511260986,
+      "learning_rate": 3.298742235697749e-07,
+      "loss": 0.43227851390838623,
+      "step": 1682
+    },
+    {
+      "epoch": 0.9226973684210527,
+      "grad_norm": 0.5025581121444702,
+      "learning_rate": 3.252697319556231e-07,
+      "loss": 0.43624621629714966,
+      "step": 1683
+    },
+    {
+      "epoch": 0.9232456140350878,
+      "grad_norm": 0.4808434844017029,
+      "learning_rate": 3.206970707869406e-07,
+      "loss": 0.4348883628845215,
+      "step": 1684
+    },
+    {
+      "epoch": 0.9237938596491229,
+      "grad_norm": 0.4720389246940613,
+      "learning_rate": 3.1615625510806104e-07,
+      "loss": 0.4509449601173401,
+      "step": 1685
+    },
+    {
+      "epoch": 0.9243421052631579,
+      "grad_norm": 0.4458601474761963,
+      "learning_rate": 3.116472998585396e-07,
+      "loss": 0.4543895423412323,
+      "step": 1686
+    },
+    {
+      "epoch": 0.924890350877193,
+      "grad_norm": 0.4926338493824005,
+      "learning_rate": 3.0717021987310967e-07,
+      "loss": 0.4404997229576111,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9254385964912281,
+      "grad_norm": 0.4627749025821686,
+      "learning_rate": 3.0272502988163645e-07,
+      "loss": 0.4535558223724365,
+      "step": 1688
+    },
+    {
+      "epoch": 0.9259868421052632,
+      "grad_norm": 0.4702151119709015,
+      "learning_rate": 2.983117445090622e-07,
+      "loss": 0.44108468294143677,
+      "step": 1689
+    },
+    {
+      "epoch": 0.9265350877192983,
+      "grad_norm": 0.46197929978370667,
+      "learning_rate": 2.939303782753611e-07,
+      "loss": 0.4439297616481781,
+      "step": 1690
+    },
+    {
+      "epoch": 0.9270833333333334,
+      "grad_norm": 0.45058155059814453,
+      "learning_rate": 2.8958094559549345e-07,
+      "loss": 0.4373398721218109,
+      "step": 1691
+    },
+    {
+      "epoch": 0.9276315789473685,
+      "grad_norm": 0.45918381214141846,
+      "learning_rate": 2.8526346077935253e-07,
+      "loss": 0.447954386472702,
+      "step": 1692
+    },
+    {
+      "epoch": 0.9281798245614035,
+      "grad_norm": 0.4705754518508911,
+      "learning_rate": 2.8097793803172456e-07,
+      "loss": 0.43893513083457947,
+      "step": 1693
+    },
+    {
+      "epoch": 0.9287280701754386,
+      "grad_norm": 0.4503750205039978,
+      "learning_rate": 2.767243914522377e-07,
+      "loss": 0.44709891080856323,
+      "step": 1694
+    },
+    {
+      "epoch": 0.9292763157894737,
+      "grad_norm": 0.44616571068763733,
+      "learning_rate": 2.725028350353176e-07,
+      "loss": 0.4397805631160736,
+      "step": 1695
+    },
+    {
+      "epoch": 0.9298245614035088,
+      "grad_norm": 0.49138715863227844,
+      "learning_rate": 2.6831328267013625e-07,
+      "loss": 0.44462379813194275,
+      "step": 1696
+    },
+    {
+      "epoch": 0.9303728070175439,
+      "grad_norm": 0.46928104758262634,
+      "learning_rate": 2.6415574814057765e-07,
+      "loss": 0.44445496797561646,
+      "step": 1697
+    },
+    {
+      "epoch": 0.930921052631579,
+      "grad_norm": 0.4646775424480438,
+      "learning_rate": 2.6003024512517683e-07,
+      "loss": 0.4469691216945648,
+      "step": 1698
+    },
+    {
+      "epoch": 0.9314692982456141,
+      "grad_norm": 0.44269227981567383,
+      "learning_rate": 2.5593678719708746e-07,
+      "loss": 0.4480845630168915,
+      "step": 1699
+    },
+    {
+      "epoch": 0.9320175438596491,
+      "grad_norm": 0.458169549703598,
+      "learning_rate": 2.5187538782403653e-07,
+      "loss": 0.4372929036617279,
+      "step": 1700
+    },
+    {
+      "epoch": 0.9325657894736842,
+      "grad_norm": 0.4619954526424408,
+      "learning_rate": 2.4784606036826974e-07,
+      "loss": 0.45119795203208923,
+      "step": 1701
+    },
+    {
+      "epoch": 0.9331140350877193,
+      "grad_norm": 0.45698124170303345,
+      "learning_rate": 2.4384881808651727e-07,
+      "loss": 0.4426797032356262,
+      "step": 1702
+    },
+    {
+      "epoch": 0.9336622807017544,
+      "grad_norm": 0.4546036720275879,
+      "learning_rate": 2.398836741299482e-07,
+      "loss": 0.43738052248954773,
+      "step": 1703
+    },
+    {
+      "epoch": 0.9342105263157895,
+      "grad_norm": 0.44701454043388367,
+      "learning_rate": 2.3595064154412374e-07,
+      "loss": 0.44920939207077026,
+      "step": 1704
+    },
+    {
+      "epoch": 0.9347587719298246,
+      "grad_norm": 0.46189531683921814,
+      "learning_rate": 2.3204973326895753e-07,
+      "loss": 0.43537482619285583,
+      "step": 1705
+    },
+    {
+      "epoch": 0.9353070175438597,
+      "grad_norm": 0.44235292077064514,
+      "learning_rate": 2.2818096213867657e-07,
+      "loss": 0.4314463138580322,
+      "step": 1706
+    },
+    {
+      "epoch": 0.9358552631578947,
+      "grad_norm": 0.44796082377433777,
+      "learning_rate": 2.2434434088176694e-07,
+      "loss": 0.4354679584503174,
+      "step": 1707
+    },
+    {
+      "epoch": 0.9364035087719298,
+      "grad_norm": 0.4733355641365051,
+      "learning_rate": 2.205398821209459e-07,
+      "loss": 0.4397640824317932,
+      "step": 1708
+    },
+    {
+      "epoch": 0.9369517543859649,
+      "grad_norm": 0.4604198634624481,
+      "learning_rate": 2.167675983731099e-07,
+      "loss": 0.43632054328918457,
+      "step": 1709
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.4596974849700928,
+      "learning_rate": 2.1302750204930112e-07,
+      "loss": 0.43812295794487,
+      "step": 1710
+    },
+    {
+      "epoch": 0.9380482456140351,
+      "grad_norm": 0.44256773591041565,
+      "learning_rate": 2.0931960545466311e-07,
+      "loss": 0.4327971339225769,
+      "step": 1711
+    },
+    {
+      "epoch": 0.9385964912280702,
+      "grad_norm": 0.4798187017440796,
+      "learning_rate": 2.0564392078839647e-07,
+      "loss": 0.4406653344631195,
+      "step": 1712
+    },
+    {
+      "epoch": 0.9391447368421053,
+      "grad_norm": 0.486609548330307,
+      "learning_rate": 2.0200046014372644e-07,
+      "loss": 0.43356356024742126,
+      "step": 1713
+    },
+    {
+      "epoch": 0.9396929824561403,
+      "grad_norm": 0.6000531911849976,
+      "learning_rate": 1.9838923550785872e-07,
+      "loss": 0.4366445541381836,
+      "step": 1714
+    },
+    {
+      "epoch": 0.9402412280701754,
+      "grad_norm": 0.4627332389354706,
+      "learning_rate": 1.9481025876193826e-07,
+      "loss": 0.44243139028549194,
+      "step": 1715
+    },
+    {
+      "epoch": 0.9407894736842105,
+      "grad_norm": 0.7296271324157715,
+      "learning_rate": 1.91263541681016e-07,
+      "loss": 0.439893901348114,
+      "step": 1716
+    },
+    {
+      "epoch": 0.9413377192982456,
+      "grad_norm": 0.4577433466911316,
+      "learning_rate": 1.877490959340045e-07,
+      "loss": 0.43834924697875977,
+      "step": 1717
+    },
+    {
+      "epoch": 0.9418859649122807,
+      "grad_norm": 0.4576733112335205,
+      "learning_rate": 1.8426693308364108e-07,
+      "loss": 0.43985801935195923,
+      "step": 1718
+    },
+    {
+      "epoch": 0.9424342105263158,
+      "grad_norm": 0.4688190817832947,
+      "learning_rate": 1.8081706458645154e-07,
+      "loss": 0.4420456886291504,
+      "step": 1719
+    },
+    {
+      "epoch": 0.9429824561403509,
+      "grad_norm": 0.45228761434555054,
+      "learning_rate": 1.7739950179271103e-07,
+      "loss": 0.4464713931083679,
+      "step": 1720
+    },
+    {
+      "epoch": 0.9435307017543859,
+      "grad_norm": 0.4589843451976776,
+      "learning_rate": 1.7401425594640753e-07,
+      "loss": 0.4457239508628845,
+      "step": 1721
+    },
+    {
+      "epoch": 0.944078947368421,
+      "grad_norm": 0.45316648483276367,
+      "learning_rate": 1.7066133818520402e-07,
+      "loss": 0.45121461153030396,
+      "step": 1722
+    },
+    {
+      "epoch": 0.9446271929824561,
+      "grad_norm": 0.4696495532989502,
+      "learning_rate": 1.6734075954040195e-07,
+      "loss": 0.4448394179344177,
+      "step": 1723
+    },
+    {
+      "epoch": 0.9451754385964912,
+      "grad_norm": 0.4616459906101227,
+      "learning_rate": 1.6405253093690344e-07,
+      "loss": 0.4420175552368164,
+      "step": 1724
+    },
+    {
+      "epoch": 0.9457236842105263,
+      "grad_norm": 0.4860699474811554,
+      "learning_rate": 1.6079666319318122e-07,
+      "loss": 0.4343988299369812,
+      "step": 1725
+    },
+    {
+      "epoch": 0.9462719298245614,
+      "grad_norm": 0.4608551561832428,
+      "learning_rate": 1.5757316702123326e-07,
+      "loss": 0.4510036110877991,
+      "step": 1726
+    },
+    {
+      "epoch": 0.9468201754385965,
+      "grad_norm": 0.4521878957748413,
+      "learning_rate": 1.543820530265594e-07,
+      "loss": 0.4449629783630371,
+      "step": 1727
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.4613654315471649,
+      "learning_rate": 1.5122333170811577e-07,
+      "loss": 0.4393472671508789,
+      "step": 1728
+    },
+    {
+      "epoch": 0.9479166666666666,
+      "grad_norm": 0.4632744789123535,
+      "learning_rate": 1.4809701345828487e-07,
+      "loss": 0.4483092725276947,
+      "step": 1729
+    },
+    {
+      "epoch": 0.9484649122807017,
+      "grad_norm": 0.4722960293292999,
+      "learning_rate": 1.4500310856284337e-07,
+      "loss": 0.4436339735984802,
+      "step": 1730
+    },
+    {
+      "epoch": 0.9490131578947368,
+      "grad_norm": 0.4626271426677704,
+      "learning_rate": 1.419416272009233e-07,
+      "loss": 0.4514046609401703,
+      "step": 1731
+    },
+    {
+      "epoch": 0.9495614035087719,
+      "grad_norm": 0.46035781502723694,
+      "learning_rate": 1.3891257944498416e-07,
+      "loss": 0.44555479288101196,
+      "step": 1732
+    },
+    {
+      "epoch": 0.950109649122807,
+      "grad_norm": 0.4788576066493988,
+      "learning_rate": 1.3591597526077527e-07,
+      "loss": 0.43546605110168457,
+      "step": 1733
+    },
+    {
+      "epoch": 0.9506578947368421,
+      "grad_norm": 0.4717090129852295,
+      "learning_rate": 1.3295182450730472e-07,
+      "loss": 0.4481337070465088,
+      "step": 1734
+    },
+    {
+      "epoch": 0.9512061403508771,
+      "grad_norm": 0.4677014648914337,
+      "learning_rate": 1.3002013693680816e-07,
+      "loss": 0.44089481234550476,
+      "step": 1735
+    },
+    {
+      "epoch": 0.9517543859649122,
+      "grad_norm": 0.5226771831512451,
+      "learning_rate": 1.2712092219471228e-07,
+      "loss": 0.4421740174293518,
+      "step": 1736
+    },
+    {
+      "epoch": 0.9523026315789473,
+      "grad_norm": 0.46488237380981445,
+      "learning_rate": 1.2425418981960813e-07,
+      "loss": 0.4298960566520691,
+      "step": 1737
+    },
+    {
+      "epoch": 0.9528508771929824,
+      "grad_norm": 0.473090797662735,
+      "learning_rate": 1.214199492432211e-07,
+      "loss": 0.4425022006034851,
+      "step": 1738
+    },
+    {
+      "epoch": 0.9533991228070176,
+      "grad_norm": 0.4689166843891144,
+      "learning_rate": 1.1861820979036986e-07,
+      "loss": 0.44378551840782166,
+      "step": 1739
+    },
+    {
+      "epoch": 0.9539473684210527,
+      "grad_norm": 0.47417959570884705,
+      "learning_rate": 1.1584898067894867e-07,
+      "loss": 0.4403073787689209,
+      "step": 1740
+    },
+    {
+      "epoch": 0.9544956140350878,
+      "grad_norm": 0.4821755886077881,
+      "learning_rate": 1.131122710198873e-07,
+      "loss": 0.4455302655696869,
+      "step": 1741
+    },
+    {
+      "epoch": 0.9550438596491229,
+      "grad_norm": 0.4931686520576477,
+      "learning_rate": 1.1040808981712448e-07,
+      "loss": 0.43022045493125916,
+      "step": 1742
+    },
+    {
+      "epoch": 0.9555921052631579,
+      "grad_norm": 0.4758283197879791,
+      "learning_rate": 1.0773644596758003e-07,
+      "loss": 0.4449695348739624,
+      "step": 1743
+    },
+    {
+      "epoch": 0.956140350877193,
+      "grad_norm": 0.4499591588973999,
+      "learning_rate": 1.0509734826112395e-07,
+      "loss": 0.44729599356651306,
+      "step": 1744
+    },
+    {
+      "epoch": 0.9566885964912281,
+      "grad_norm": 0.4610680937767029,
+      "learning_rate": 1.0249080538054512e-07,
+      "loss": 0.4448246359825134,
+      "step": 1745
+    },
+    {
+      "epoch": 0.9572368421052632,
+      "grad_norm": 0.4604075849056244,
+      "learning_rate": 9.991682590152707e-08,
+      "loss": 0.4426057040691376,
+      "step": 1746
+    },
+    {
+      "epoch": 0.9577850877192983,
+      "grad_norm": 0.5067770481109619,
+      "learning_rate": 9.737541829261787e-08,
+      "loss": 0.4603898525238037,
+      "step": 1747
+    },
+    {
+      "epoch": 0.9583333333333334,
+      "grad_norm": 0.4828222692012787,
+      "learning_rate": 9.486659091520244e-08,
+      "loss": 0.44565194845199585,
+      "step": 1748
+    },
+    {
+      "epoch": 0.9588815789473685,
+      "grad_norm": 0.46997013688087463,
+      "learning_rate": 9.239035202347147e-08,
+      "loss": 0.44362831115722656,
+      "step": 1749
+    },
+    {
+      "epoch": 0.9594298245614035,
+      "grad_norm": 0.5518832802772522,
+      "learning_rate": 8.994670976440357e-08,
+      "loss": 0.44277042150497437,
+      "step": 1750
+    },
+    {
+      "epoch": 0.9599780701754386,
+      "grad_norm": 0.4646471440792084,
+      "learning_rate": 8.753567217772652e-08,
+      "loss": 0.44326919317245483,
+      "step": 1751
+    },
+    {
+      "epoch": 0.9605263157894737,
+      "grad_norm": 0.46489813923835754,
+      "learning_rate": 8.515724719589835e-08,
+      "loss": 0.4298398196697235,
+      "step": 1752
+    },
+    {
+      "epoch": 0.9610745614035088,
+      "grad_norm": 0.9936460852622986,
+      "learning_rate": 8.281144264408291e-08,
+      "loss": 0.44022315740585327,
+      "step": 1753
+    },
+    {
+      "epoch": 0.9616228070175439,
+      "grad_norm": 0.462695449590683,
+      "learning_rate": 8.04982662401188e-08,
+      "loss": 0.442985862493515,
+      "step": 1754
+    },
+    {
+      "epoch": 0.962171052631579,
+      "grad_norm": 0.4724031984806061,
+      "learning_rate": 7.821772559449381e-08,
+      "loss": 0.4336352050304413,
+      "step": 1755
+    },
+    {
+      "epoch": 0.9627192982456141,
+      "grad_norm": 0.4420538544654846,
+      "learning_rate": 7.596982821032495e-08,
+      "loss": 0.4361079931259155,
+      "step": 1756
+    },
+    {
+      "epoch": 0.9632675438596491,
+      "grad_norm": 0.4517374336719513,
+      "learning_rate": 7.375458148332959e-08,
+      "loss": 0.4529159963130951,
+      "step": 1757
+    },
+    {
+      "epoch": 0.9638157894736842,
+      "grad_norm": 0.4620855450630188,
+      "learning_rate": 7.157199270180326e-08,
+      "loss": 0.4500415027141571,
+      "step": 1758
+    },
+    {
+      "epoch": 0.9643640350877193,
+      "grad_norm": 0.47779715061187744,
+      "learning_rate": 6.942206904659521e-08,
+      "loss": 0.44449353218078613,
+      "step": 1759
+    },
+    {
+      "epoch": 0.9649122807017544,
+      "grad_norm": 0.47931793332099915,
+      "learning_rate": 6.730481759108287e-08,
+      "loss": 0.4509713053703308,
+      "step": 1760
+    },
+    {
+      "epoch": 0.9654605263157895,
+      "grad_norm": 0.464688777923584,
+      "learning_rate": 6.522024530115079e-08,
+      "loss": 0.4385107755661011,
+      "step": 1761
+    },
+    {
+      "epoch": 0.9660087719298246,
+      "grad_norm": 0.4776732325553894,
+      "learning_rate": 6.316835903516949e-08,
+      "loss": 0.45002496242523193,
+      "step": 1762
+    },
+    {
+      "epoch": 0.9665570175438597,
+      "grad_norm": 0.4563760757446289,
+      "learning_rate": 6.114916554396555e-08,
+      "loss": 0.4582784175872803,
+      "step": 1763
+    },
+    {
+      "epoch": 0.9671052631578947,
+      "grad_norm": 0.46798089146614075,
+      "learning_rate": 5.9162671470809343e-08,
+      "loss": 0.43207693099975586,
+      "step": 1764
+    },
+    {
+      "epoch": 0.9676535087719298,
+      "grad_norm": 0.5528680086135864,
+      "learning_rate": 5.720888335138508e-08,
+      "loss": 0.43807539343833923,
+      "step": 1765
+    },
+    {
+      "epoch": 0.9682017543859649,
+      "grad_norm": 0.5016975402832031,
+      "learning_rate": 5.528780761377306e-08,
+      "loss": 0.46113985776901245,
+      "step": 1766
+    },
+    {
+      "epoch": 0.96875,
+      "grad_norm": 0.4595453143119812,
+      "learning_rate": 5.339945057842966e-08,
+      "loss": 0.4475017488002777,
+      "step": 1767
+    },
+    {
+      "epoch": 0.9692982456140351,
+      "grad_norm": 0.49467697739601135,
+      "learning_rate": 5.15438184581607e-08,
+      "loss": 0.4496666193008423,
+      "step": 1768
+    },
+    {
+      "epoch": 0.9698464912280702,
+      "grad_norm": 0.4644295573234558,
+      "learning_rate": 4.9720917358111464e-08,
+      "loss": 0.44429510831832886,
+      "step": 1769
+    },
+    {
+      "epoch": 0.9703947368421053,
+      "grad_norm": 0.4744158089160919,
+      "learning_rate": 4.7930753275733376e-08,
+      "loss": 0.4378962516784668,
+      "step": 1770
+    },
+    {
+      "epoch": 0.9709429824561403,
+      "grad_norm": 0.4634189009666443,
+      "learning_rate": 4.617333210077513e-08,
+      "loss": 0.4459246098995209,
+      "step": 1771
+    },
+    {
+      "epoch": 0.9714912280701754,
+      "grad_norm": 0.4679155647754669,
+      "learning_rate": 4.444865961525824e-08,
+      "loss": 0.4505774974822998,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9720394736842105,
+      "grad_norm": 0.488528311252594,
+      "learning_rate": 4.275674149345821e-08,
+      "loss": 0.4513344466686249,
+      "step": 1773
+    },
+    {
+      "epoch": 0.9725877192982456,
+      "grad_norm": 0.4505665898323059,
+      "learning_rate": 4.109758330188895e-08,
+      "loss": 0.44093236327171326,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9731359649122807,
+      "grad_norm": 0.4865361750125885,
+      "learning_rate": 3.94711904992795e-08,
+      "loss": 0.4391603171825409,
+      "step": 1775
+    },
+    {
+      "epoch": 0.9736842105263158,
+      "grad_norm": 0.46886420249938965,
+      "learning_rate": 3.787756843656287e-08,
+      "loss": 0.4503663182258606,
+      "step": 1776
+    },
+    {
+      "epoch": 0.9742324561403509,
+      "grad_norm": 0.46422988176345825,
+      "learning_rate": 3.6316722356848354e-08,
+      "loss": 0.4432747960090637,
+      "step": 1777
+    },
+    {
+      "epoch": 0.9747807017543859,
+      "grad_norm": 0.4637066423892975,
+      "learning_rate": 3.4788657395414813e-08,
+      "loss": 0.4446980953216553,
+      "step": 1778
+    },
+    {
+      "epoch": 0.975328947368421,
+      "grad_norm": 0.48918765783309937,
+      "learning_rate": 3.3293378579688505e-08,
+      "loss": 0.43319782614707947,
+      "step": 1779
+    },
+    {
+      "epoch": 0.9758771929824561,
+      "grad_norm": 0.4590325951576233,
+      "learning_rate": 3.1830890829226416e-08,
+      "loss": 0.44327086210250854,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9764254385964912,
+      "grad_norm": 0.450764536857605,
+      "learning_rate": 3.0401198955702926e-08,
+      "loss": 0.44975000619888306,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9769736842105263,
+      "grad_norm": 0.45482075214385986,
+      "learning_rate": 2.9004307662887642e-08,
+      "loss": 0.44603753089904785,
+      "step": 1782
+    },
+    {
+      "epoch": 0.9775219298245614,
+      "grad_norm": 0.45557013154029846,
+      "learning_rate": 2.76402215466387e-08,
+      "loss": 0.4406141936779022,
+      "step": 1783
+    },
+    {
+      "epoch": 0.9780701754385965,
+      "grad_norm": 0.48736274242401123,
+      "learning_rate": 2.6308945094880577e-08,
+      "loss": 0.4546322524547577,
+      "step": 1784
+    },
+    {
+      "epoch": 0.9786184210526315,
+      "grad_norm": 0.4766358435153961,
+      "learning_rate": 2.5010482687595206e-08,
+      "loss": 0.43386587500572205,
+      "step": 1785
+    },
+    {
+      "epoch": 0.9791666666666666,
+      "grad_norm": 0.5122347474098206,
+      "learning_rate": 2.3744838596799767e-08,
+      "loss": 0.42556485533714294,
+      "step": 1786
+    },
+    {
+      "epoch": 0.9797149122807017,
+      "grad_norm": 0.4523909389972687,
+      "learning_rate": 2.251201698654115e-08,
+      "loss": 0.4424014985561371,
+      "step": 1787
+    },
+    {
+      "epoch": 0.9802631578947368,
+      "grad_norm": 0.48024672269821167,
+      "learning_rate": 2.1312021912875957e-08,
+      "loss": 0.44508880376815796,
+      "step": 1788
+    },
+    {
+      "epoch": 0.9808114035087719,
+      "grad_norm": 0.47597408294677734,
+      "learning_rate": 2.01448573238594e-08,
+      "loss": 0.4251779317855835,
+      "step": 1789
+    },
+    {
+      "epoch": 0.981359649122807,
+      "grad_norm": 0.49238336086273193,
+      "learning_rate": 1.901052705953532e-08,
+      "loss": 0.4298170208930969,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9819078947368421,
+      "grad_norm": 0.4651866853237152,
+      "learning_rate": 1.7909034851917306e-08,
+      "loss": 0.42286503314971924,
+      "step": 1791
+    },
+    {
+      "epoch": 0.9824561403508771,
+      "grad_norm": 0.4722592830657959,
+      "learning_rate": 1.684038432498092e-08,
+      "loss": 0.428172767162323,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9830043859649122,
+      "grad_norm": 0.4582705795764923,
+      "learning_rate": 1.5804578994652598e-08,
+      "loss": 0.4320406913757324,
+      "step": 1793
+    },
+    {
+      "epoch": 0.9835526315789473,
+      "grad_norm": 0.46447935700416565,
+      "learning_rate": 1.4801622268791893e-08,
+      "loss": 0.44131776690483093,
+      "step": 1794
+    },
+    {
+      "epoch": 0.9841008771929824,
+      "grad_norm": 0.4600572884082794,
+      "learning_rate": 1.3831517447188137e-08,
+      "loss": 0.46617740392684937,
+      "step": 1795
+    },
+    {
+      "epoch": 0.9846491228070176,
+      "grad_norm": 0.46110910177230835,
+      "learning_rate": 1.2894267721543786e-08,
+      "loss": 0.4432889223098755,
+      "step": 1796
+    },
+    {
+      "epoch": 0.9851973684210527,
+      "grad_norm": 0.45410728454589844,
+      "learning_rate": 1.1989876175468872e-08,
+      "loss": 0.42493683099746704,
+      "step": 1797
+    },
+    {
+      "epoch": 0.9857456140350878,
+      "grad_norm": 0.47796985507011414,
+      "learning_rate": 1.1118345784463247e-08,
+      "loss": 0.42538389563560486,
+      "step": 1798
+    },
+    {
+      "epoch": 0.9862938596491229,
+      "grad_norm": 0.4609275460243225,
+      "learning_rate": 1.0279679415916565e-08,
+      "loss": 0.45832300186157227,
+      "step": 1799
+    },
+    {
+      "epoch": 0.9868421052631579,
+      "grad_norm": 0.4755944013595581,
+      "learning_rate": 9.473879829091648e-09,
+      "loss": 0.4499424397945404,
+      "step": 1800
+    },
+    {
+      "epoch": 0.987390350877193,
+      "grad_norm": 0.4787357449531555,
+      "learning_rate": 8.700949675115588e-09,
+      "loss": 0.44248414039611816,
+      "step": 1801
+    },
+    {
+      "epoch": 0.9879385964912281,
+      "grad_norm": 0.45411446690559387,
+      "learning_rate": 7.960891496977541e-09,
+      "loss": 0.4389973282814026,
+      "step": 1802
+    },
+    {
+      "epoch": 0.9884868421052632,
+      "grad_norm": 0.4596308469772339,
+      "learning_rate": 7.253707729513171e-09,
+      "loss": 0.4519648253917694,
+      "step": 1803
+    },
+    {
+      "epoch": 0.9890350877192983,
+      "grad_norm": 0.4609762728214264,
+      "learning_rate": 6.579400699397998e-09,
+      "loss": 0.44419950246810913,
+      "step": 1804
+    },
+    {
+      "epoch": 0.9895833333333334,
+      "grad_norm": 0.4751090705394745,
+      "learning_rate": 5.937972625145172e-09,
+      "loss": 0.46066397428512573,
+      "step": 1805
+    },
+    {
+      "epoch": 0.9901315789473685,
+      "grad_norm": 0.4669820964336395,
+      "learning_rate": 5.329425617088824e-09,
+      "loss": 0.43845662474632263,
+      "step": 1806
+    },
+    {
+      "epoch": 0.9906798245614035,
+      "grad_norm": 0.46809566020965576,
+      "learning_rate": 4.753761677388502e-09,
+      "loss": 0.4518412947654724,
+      "step": 1807
+    },
+    {
+      "epoch": 0.9912280701754386,
+      "grad_norm": 0.4599871039390564,
+      "learning_rate": 4.2109827000103024e-09,
+      "loss": 0.4281280040740967,
+      "step": 1808
+    },
+    {
+      "epoch": 0.9917763157894737,
+      "grad_norm": 0.4735415279865265,
+      "learning_rate": 3.7010904707313057e-09,
+      "loss": 0.44415637850761414,
+      "step": 1809
+    },
+    {
+      "epoch": 0.9923245614035088,
+      "grad_norm": 0.4610520005226135,
+      "learning_rate": 3.224086667128479e-09,
+      "loss": 0.44374585151672363,
+      "step": 1810
+    },
+    {
+      "epoch": 0.9928728070175439,
+      "grad_norm": 0.4630661606788635,
+      "learning_rate": 2.7799728585709005e-09,
+      "loss": 0.43956732749938965,
+      "step": 1811
+    },
+    {
+      "epoch": 0.993421052631579,
+      "grad_norm": 0.46847841143608093,
+      "learning_rate": 2.3687505062208737e-09,
+      "loss": 0.4338926672935486,
+      "step": 1812
+    },
+    {
+      "epoch": 0.9939692982456141,
+      "grad_norm": 0.4724494218826294,
+      "learning_rate": 1.9904209630250417e-09,
+      "loss": 0.4389767050743103,
+      "step": 1813
+    },
+    {
+      "epoch": 0.9945175438596491,
+      "grad_norm": 0.47051912546157837,
+      "learning_rate": 1.644985473709948e-09,
+      "loss": 0.4498752951622009,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9950657894736842,
+      "grad_norm": 0.44187289476394653,
+      "learning_rate": 1.3324451747798173e-09,
+      "loss": 0.44881471991539,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9956140350877193,
+      "grad_norm": 0.4773845076560974,
+      "learning_rate": 1.0528010945098921e-09,
+      "loss": 0.44216442108154297,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9961622807017544,
+      "grad_norm": 0.44168299436569214,
+      "learning_rate": 8.060541529464339e-10,
+      "loss": 0.45296916365623474,
+      "step": 1817
+    },
+    {
+      "epoch": 0.9967105263157895,
+      "grad_norm": 0.4654878079891205,
+      "learning_rate": 5.922051619033919e-10,
+      "loss": 0.44899439811706543,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9972587719298246,
+      "grad_norm": 0.47827789187431335,
+      "learning_rate": 4.112548249546322e-10,
+      "loss": 0.44498586654663086,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9978070175438597,
+      "grad_norm": 0.48996180295944214,
+      "learning_rate": 2.632037374383778e-10,
+      "loss": 0.4426143169403076,
+      "step": 1820
+    },
+    {
+      "epoch": 0.9983552631578947,
+      "grad_norm": 0.4770708680152893,
+      "learning_rate": 1.480523864516581e-10,
+      "loss": 0.44537267088890076,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9989035087719298,
+      "grad_norm": 0.5157729387283325,
+      "learning_rate": 6.580115084919847e-11,
+      "loss": 0.4398488402366638,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9994517543859649,
+      "grad_norm": 0.48881420493125916,
+      "learning_rate": 1.6450301243420285e-11,
+      "loss": 0.437910258769989,
+      "step": 1823
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4549266993999481,
+      "learning_rate": 0.0,
+      "loss": 0.4407366216182709,
+      "step": 1824
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1824,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1659437929638186e+20,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}