diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14582 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2077,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00048152160828217165,
+      "grad_norm": 1.4487173557281494,
+      "learning_rate": 0.0,
+      "loss": 1.7129,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009630432165643433,
+      "grad_norm": 2.391146183013916,
+      "learning_rate": 3.1746031746031746e-06,
+      "loss": 2.3104,
+      "step": 2
+    },
+    {
+      "epoch": 0.001444564824846515,
+      "grad_norm": 0.6623585820198059,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 1.2074,
+      "step": 3
+    },
+    {
+      "epoch": 0.0019260864331286866,
+      "grad_norm": 2.9669229984283447,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 1.7921,
+      "step": 4
+    },
+    {
+      "epoch": 0.002407608041410858,
+      "grad_norm": 1.4202985763549805,
+      "learning_rate": 1.2698412698412699e-05,
+      "loss": 1.5058,
+      "step": 5
+    },
+    {
+      "epoch": 0.00288912964969303,
+      "grad_norm": 0.908497154712677,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 1.3187,
+      "step": 6
+    },
+    {
+      "epoch": 0.0033706512579752016,
+      "grad_norm": 4.113708019256592,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 2.1155,
+      "step": 7
+    },
+    {
+      "epoch": 0.003852172866257373,
+      "grad_norm": 2.203523635864258,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 1.577,
+      "step": 8
+    },
+    {
+      "epoch": 0.004333694474539545,
+      "grad_norm": 3.0024302005767822,
+      "learning_rate": 2.5396825396825397e-05,
+      "loss": 1.7602,
+      "step": 9
+    },
+    {
+      "epoch": 0.004815216082821716,
+      "grad_norm": 2.271422863006592,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 1.6359,
+      "step": 10
+    },
+    {
+      "epoch": 0.005296737691103889,
+      "grad_norm": 1.6946759223937988,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 1.4716,
+      "step": 11
+    },
+    {
+      "epoch": 0.00577825929938606,
+      "grad_norm": 1.5421451330184937,
+      "learning_rate": 3.492063492063492e-05,
+      "loss": 0.9894,
+      "step": 12
+    },
+    {
+      "epoch": 0.006259780907668232,
+      "grad_norm": 2.4348111152648926,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 1.326,
+      "step": 13
+    },
+    {
+      "epoch": 0.006741302515950403,
+      "grad_norm": 2.774260997772217,
+      "learning_rate": 4.126984126984127e-05,
+      "loss": 1.3238,
+      "step": 14
+    },
+    {
+      "epoch": 0.007222824124232575,
+      "grad_norm": 1.2383824586868286,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 1.3033,
+      "step": 15
+    },
+    {
+      "epoch": 0.007704345732514746,
+      "grad_norm": 2.3136370182037354,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 1.9822,
+      "step": 16
+    },
+    {
+      "epoch": 0.008185867340796918,
+      "grad_norm": 0.5560957193374634,
+      "learning_rate": 5.0793650793650794e-05,
+      "loss": 1.3964,
+      "step": 17
+    },
+    {
+      "epoch": 0.00866738894907909,
+      "grad_norm": 1.1794217824935913,
+      "learning_rate": 5.396825396825397e-05,
+      "loss": 1.2084,
+      "step": 18
+    },
+    {
+      "epoch": 0.009148910557361261,
+      "grad_norm": 0.8785364627838135,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 1.1667,
+      "step": 19
+    },
+    {
+      "epoch": 0.009630432165643433,
+      "grad_norm": 2.39178466796875,
+      "learning_rate": 6.0317460317460316e-05,
+      "loss": 1.8301,
+      "step": 20
+    },
+    {
+      "epoch": 0.010111953773925604,
+      "grad_norm": 1.0043989419937134,
+      "learning_rate": 6.349206349206349e-05,
+      "loss": 1.346,
+      "step": 21
+    },
+    {
+      "epoch": 0.010593475382207777,
+      "grad_norm": 1.1557499170303345,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.4663,
+      "step": 22
+    },
+    {
+      "epoch": 0.011074996990489949,
+      "grad_norm": 0.5397685170173645,
+      "learning_rate": 6.984126984126984e-05,
+      "loss": 0.9156,
+      "step": 23
+    },
+    {
+      "epoch": 0.01155651859877212,
+      "grad_norm": 0.86281418800354,
+      "learning_rate": 7.301587301587302e-05,
+      "loss": 0.9897,
+      "step": 24
+    },
+    {
+      "epoch": 0.012038040207054292,
+      "grad_norm": 2.705343723297119,
+      "learning_rate": 7.619047619047618e-05,
+      "loss": 1.6137,
+      "step": 25
+    },
+    {
+      "epoch": 0.012519561815336464,
+      "grad_norm": 0.8081772327423096,
+      "learning_rate": 7.936507936507937e-05,
+      "loss": 0.9733,
+      "step": 26
+    },
+    {
+      "epoch": 0.013001083423618635,
+      "grad_norm": 0.5619446039199829,
+      "learning_rate": 8.253968253968255e-05,
+      "loss": 1.0055,
+      "step": 27
+    },
+    {
+      "epoch": 0.013482605031900807,
+      "grad_norm": 0.9184542298316956,
+      "learning_rate": 8.571428571428571e-05,
+      "loss": 0.7692,
+      "step": 28
+    },
+    {
+      "epoch": 0.013964126640182978,
+      "grad_norm": 0.7189937829971313,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 0.9745,
+      "step": 29
+    },
+    {
+      "epoch": 0.01444564824846515,
+      "grad_norm": 1.1548259258270264,
+      "learning_rate": 9.206349206349206e-05,
+      "loss": 1.0737,
+      "step": 30
+    },
+    {
+      "epoch": 0.014927169856747321,
+      "grad_norm": 1.7942982912063599,
+      "learning_rate": 9.523809523809524e-05,
+      "loss": 0.8184,
+      "step": 31
+    },
+    {
+      "epoch": 0.015408691465029493,
+      "grad_norm": 0.9638906121253967,
+      "learning_rate": 9.841269841269841e-05,
+      "loss": 1.2128,
+      "step": 32
+    },
+    {
+      "epoch": 0.015890213073311666,
+      "grad_norm": 0.9639189839363098,
+      "learning_rate": 0.00010158730158730159,
+      "loss": 1.0256,
+      "step": 33
+    },
+    {
+      "epoch": 0.016371734681593836,
+      "grad_norm": 0.6873184442520142,
+      "learning_rate": 0.00010476190476190477,
+      "loss": 1.048,
+      "step": 34
+    },
+    {
+      "epoch": 0.01685325628987601,
+      "grad_norm": 1.1122071743011475,
+      "learning_rate": 0.00010793650793650794,
+      "loss": 0.6637,
+      "step": 35
+    },
+    {
+      "epoch": 0.01733477789815818,
+      "grad_norm": 0.5951284766197205,
+      "learning_rate": 0.00011111111111111112,
+      "loss": 1.3751,
+      "step": 36
+    },
+    {
+      "epoch": 0.017816299506440352,
+      "grad_norm": 0.6848040819168091,
+      "learning_rate": 0.00011428571428571428,
+      "loss": 0.6586,
+      "step": 37
+    },
+    {
+      "epoch": 0.018297821114722522,
+      "grad_norm": 0.9294832944869995,
+      "learning_rate": 0.00011746031746031746,
+      "loss": 1.2054,
+      "step": 38
+    },
+    {
+      "epoch": 0.018779342723004695,
+      "grad_norm": 0.853848934173584,
+      "learning_rate": 0.00012063492063492063,
+      "loss": 0.6379,
+      "step": 39
+    },
+    {
+      "epoch": 0.019260864331286865,
+      "grad_norm": 1.1381187438964844,
+      "learning_rate": 0.0001238095238095238,
+      "loss": 1.2677,
+      "step": 40
+    },
+    {
+      "epoch": 0.01974238593956904,
+      "grad_norm": 0.812929093837738,
+      "learning_rate": 0.00012698412698412698,
+      "loss": 1.2011,
+      "step": 41
+    },
+    {
+      "epoch": 0.020223907547851208,
+      "grad_norm": 0.9746362566947937,
+      "learning_rate": 0.00013015873015873017,
+      "loss": 1.1164,
+      "step": 42
+    },
+    {
+      "epoch": 0.02070542915613338,
+      "grad_norm": 0.5625665783882141,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.9418,
+      "step": 43
+    },
+    {
+      "epoch": 0.021186950764415555,
+      "grad_norm": 0.7534303665161133,
+      "learning_rate": 0.0001365079365079365,
+      "loss": 0.8973,
+      "step": 44
+    },
+    {
+      "epoch": 0.021668472372697724,
+      "grad_norm": 0.8771131038665771,
+      "learning_rate": 0.00013968253968253967,
+      "loss": 0.9622,
+      "step": 45
+    },
+    {
+      "epoch": 0.022149993980979898,
+      "grad_norm": 1.096232295036316,
+      "learning_rate": 0.00014285714285714287,
+      "loss": 1.3677,
+      "step": 46
+    },
+    {
+      "epoch": 0.022631515589262068,
+      "grad_norm": 2.257794141769409,
+      "learning_rate": 0.00014603174603174603,
+      "loss": 1.0777,
+      "step": 47
+    },
+    {
+      "epoch": 0.02311303719754424,
+      "grad_norm": 0.9336938858032227,
+      "learning_rate": 0.00014920634920634923,
+      "loss": 1.215,
+      "step": 48
+    },
+    {
+      "epoch": 0.02359455880582641,
+      "grad_norm": 2.0201683044433594,
+      "learning_rate": 0.00015238095238095237,
+      "loss": 0.974,
+      "step": 49
+    },
+    {
+      "epoch": 0.024076080414108584,
+      "grad_norm": 0.36621397733688354,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 0.5777,
+      "step": 50
+    },
+    {
+      "epoch": 0.024557602022390754,
+      "grad_norm": 0.9629436135292053,
+      "learning_rate": 0.00015873015873015873,
+      "loss": 0.9346,
+      "step": 51
+    },
+    {
+      "epoch": 0.025039123630672927,
+      "grad_norm": 1.607643485069275,
+      "learning_rate": 0.00016190476190476192,
+      "loss": 1.8116,
+      "step": 52
+    },
+    {
+      "epoch": 0.025520645238955097,
+      "grad_norm": 0.7304220199584961,
+      "learning_rate": 0.0001650793650793651,
+      "loss": 0.7118,
+      "step": 53
+    },
+    {
+      "epoch": 0.02600216684723727,
+      "grad_norm": 0.6517884135246277,
+      "learning_rate": 0.00016825396825396826,
+      "loss": 0.6906,
+      "step": 54
+    },
+    {
+      "epoch": 0.02648368845551944,
+      "grad_norm": 0.833574652671814,
+      "learning_rate": 0.00017142857142857143,
+      "loss": 1.3414,
+      "step": 55
+    },
+    {
+      "epoch": 0.026965210063801613,
+      "grad_norm": 1.3700557947158813,
+      "learning_rate": 0.00017460317460317462,
+      "loss": 0.7933,
+      "step": 56
+    },
+    {
+      "epoch": 0.027446731672083786,
+      "grad_norm": 1.6920475959777832,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 1.2288,
+      "step": 57
+    },
+    {
+      "epoch": 0.027928253280365956,
+      "grad_norm": 0.6903666853904724,
+      "learning_rate": 0.00018095238095238095,
+      "loss": 0.7095,
+      "step": 58
+    },
+    {
+      "epoch": 0.02840977488864813,
+      "grad_norm": 1.1906098127365112,
+      "learning_rate": 0.00018412698412698412,
+      "loss": 0.9009,
+      "step": 59
+    },
+    {
+      "epoch": 0.0288912964969303,
+      "grad_norm": 1.0149602890014648,
+      "learning_rate": 0.00018730158730158731,
+      "loss": 0.6004,
+      "step": 60
+    },
+    {
+      "epoch": 0.029372818105212473,
+      "grad_norm": 0.5224968791007996,
+      "learning_rate": 0.00019047619047619048,
+      "loss": 0.9209,
+      "step": 61
+    },
+    {
+      "epoch": 0.029854339713494642,
+      "grad_norm": 1.0166723728179932,
+      "learning_rate": 0.00019365079365079365,
+      "loss": 0.7085,
+      "step": 62
+    },
+    {
+      "epoch": 0.030335861321776816,
+      "grad_norm": 0.8379011750221252,
+      "learning_rate": 0.00019682539682539682,
+      "loss": 0.7444,
+      "step": 63
+    },
+    {
+      "epoch": 0.030817382930058985,
+      "grad_norm": 1.1669416427612305,
+      "learning_rate": 0.0002,
+      "loss": 1.12,
+      "step": 64
+    },
+    {
+      "epoch": 0.03129890453834116,
+      "grad_norm": 1.1994937658309937,
+      "learning_rate": 0.00019999987833918285,
+      "loss": 1.0237,
+      "step": 65
+    },
+    {
+      "epoch": 0.03178042614662333,
+      "grad_norm": 0.520425021648407,
+      "learning_rate": 0.00019999951335702735,
+      "loss": 1.0711,
+      "step": 66
+    },
+    {
+      "epoch": 0.0322619477549055,
+      "grad_norm": 1.4933056831359863,
+      "learning_rate": 0.00019999890505442158,
+      "loss": 0.9168,
+      "step": 67
+    },
+    {
+      "epoch": 0.03274346936318767,
+      "grad_norm": 1.2318795919418335,
+      "learning_rate": 0.0001999980534328457,
+      "loss": 1.1426,
+      "step": 68
+    },
+    {
+      "epoch": 0.033224990971469845,
+      "grad_norm": 1.2501200437545776,
+      "learning_rate": 0.0001999969584943719,
+      "loss": 1.14,
+      "step": 69
+    },
+    {
+      "epoch": 0.03370651257975202,
+      "grad_norm": 1.9055863618850708,
+      "learning_rate": 0.00019999562024166438,
+      "loss": 0.8023,
+      "step": 70
+    },
+    {
+      "epoch": 0.03418803418803419,
+      "grad_norm": 0.5718879699707031,
+      "learning_rate": 0.0001999940386779794,
+      "loss": 0.3126,
+      "step": 71
+    },
+    {
+      "epoch": 0.03466955579631636,
+      "grad_norm": 0.7291958928108215,
+      "learning_rate": 0.00019999221380716527,
+      "loss": 0.7172,
+      "step": 72
+    },
+    {
+      "epoch": 0.03515107740459853,
+      "grad_norm": 1.2963175773620605,
+      "learning_rate": 0.00019999014563366226,
+      "loss": 0.9218,
+      "step": 73
+    },
+    {
+      "epoch": 0.035632599012880704,
+      "grad_norm": 1.3575979471206665,
+      "learning_rate": 0.00019998783416250268,
+      "loss": 1.0466,
+      "step": 74
+    },
+    {
+      "epoch": 0.03611412062116288,
+      "grad_norm": 1.1832976341247559,
+      "learning_rate": 0.0001999852793993109,
+      "loss": 0.8877,
+      "step": 75
+    },
+    {
+      "epoch": 0.036595642229445044,
+      "grad_norm": 0.5200212597846985,
+      "learning_rate": 0.00019998248135030315,
+      "loss": 0.9763,
+      "step": 76
+    },
+    {
+      "epoch": 0.03707716383772722,
+      "grad_norm": 0.32084816694259644,
+      "learning_rate": 0.00019997944002228774,
+      "loss": 0.8581,
+      "step": 77
+    },
+    {
+      "epoch": 0.03755868544600939,
+      "grad_norm": 1.1948219537734985,
+      "learning_rate": 0.00019997615542266482,
+      "loss": 0.709,
+      "step": 78
+    },
+    {
+      "epoch": 0.038040207054291564,
+      "grad_norm": 0.6765559911727905,
+      "learning_rate": 0.00019997262755942655,
+      "loss": 0.6078,
+      "step": 79
+    },
+    {
+      "epoch": 0.03852172866257373,
+      "grad_norm": 1.0067569017410278,
+      "learning_rate": 0.000199968856441157,
+      "loss": 0.7802,
+      "step": 80
+    },
+    {
+      "epoch": 0.0390032502708559,
+      "grad_norm": 0.9474171996116638,
+      "learning_rate": 0.0001999648420770321,
+      "loss": 0.748,
+      "step": 81
+    },
+    {
+      "epoch": 0.03948477187913808,
+      "grad_norm": 0.7215222716331482,
+      "learning_rate": 0.0001999605844768197,
+      "loss": 1.2064,
+      "step": 82
+    },
+    {
+      "epoch": 0.03996629348742025,
+      "grad_norm": 0.8327376246452332,
+      "learning_rate": 0.00019995608365087946,
+      "loss": 1.214,
+      "step": 83
+    },
+    {
+      "epoch": 0.040447815095702416,
+      "grad_norm": 0.7590318918228149,
+      "learning_rate": 0.0001999513396101628,
+      "loss": 0.9192,
+      "step": 84
+    },
+    {
+      "epoch": 0.04092933670398459,
+      "grad_norm": 1.0768215656280518,
+      "learning_rate": 0.00019994635236621306,
+      "loss": 0.509,
+      "step": 85
+    },
+    {
+      "epoch": 0.04141085831226676,
+      "grad_norm": 0.596147358417511,
+      "learning_rate": 0.00019994112193116528,
+      "loss": 0.9004,
+      "step": 86
+    },
+    {
+      "epoch": 0.041892379920548936,
+      "grad_norm": 0.8125873804092407,
+      "learning_rate": 0.00019993564831774618,
+      "loss": 1.2089,
+      "step": 87
+    },
+    {
+      "epoch": 0.04237390152883111,
+      "grad_norm": 2.323214054107666,
+      "learning_rate": 0.00019992993153927432,
+      "loss": 0.9856,
+      "step": 88
+    },
+    {
+      "epoch": 0.042855423137113276,
+      "grad_norm": 1.0737050771713257,
+      "learning_rate": 0.00019992397160965982,
+      "loss": 0.5664,
+      "step": 89
+    },
+    {
+      "epoch": 0.04333694474539545,
+      "grad_norm": 0.5349381566047668,
+      "learning_rate": 0.0001999177685434045,
+      "loss": 1.1484,
+      "step": 90
+    },
+    {
+      "epoch": 0.04381846635367762,
+      "grad_norm": 1.065415382385254,
+      "learning_rate": 0.00019991132235560176,
+      "loss": 1.3118,
+      "step": 91
+    },
+    {
+      "epoch": 0.044299987961959796,
+      "grad_norm": 0.8875418305397034,
+      "learning_rate": 0.00019990463306193652,
+      "loss": 1.0626,
+      "step": 92
+    },
+    {
+      "epoch": 0.04478150957024196,
+      "grad_norm": 0.6135897040367126,
+      "learning_rate": 0.00019989770067868533,
+      "loss": 0.7007,
+      "step": 93
+    },
+    {
+      "epoch": 0.045263031178524135,
+      "grad_norm": 1.3404655456542969,
+      "learning_rate": 0.00019989052522271622,
+      "loss": 1.2922,
+      "step": 94
+    },
+    {
+      "epoch": 0.04574455278680631,
+      "grad_norm": 0.8707893490791321,
+      "learning_rate": 0.00019988310671148848,
+      "loss": 0.9461,
+      "step": 95
+    },
+    {
+      "epoch": 0.04622607439508848,
+      "grad_norm": 0.8725842833518982,
+      "learning_rate": 0.00019987544516305311,
+      "loss": 1.2371,
+      "step": 96
+    },
+    {
+      "epoch": 0.04670759600337065,
+      "grad_norm": 0.83966064453125,
+      "learning_rate": 0.00019986754059605222,
+      "loss": 0.6515,
+      "step": 97
+    },
+    {
+      "epoch": 0.04718911761165282,
+      "grad_norm": 0.6398138403892517,
+      "learning_rate": 0.00019985939302971938,
+      "loss": 0.8191,
+      "step": 98
+    },
+    {
+      "epoch": 0.047670639219934995,
+      "grad_norm": 0.7741346955299377,
+      "learning_rate": 0.00019985100248387933,
+      "loss": 1.1037,
+      "step": 99
+    },
+    {
+      "epoch": 0.04815216082821717,
+      "grad_norm": 1.0956766605377197,
+      "learning_rate": 0.00019984236897894816,
+      "loss": 0.5513,
+      "step": 100
+    },
+    {
+      "epoch": 0.04863368243649934,
+      "grad_norm": 0.7802196741104126,
+      "learning_rate": 0.000199833492535933,
+      "loss": 0.5401,
+      "step": 101
+    },
+    {
+      "epoch": 0.04911520404478151,
+      "grad_norm": 0.8452202677726746,
+      "learning_rate": 0.00019982437317643217,
+      "loss": 0.9785,
+      "step": 102
+    },
+    {
+      "epoch": 0.04959672565306368,
+      "grad_norm": 1.0651404857635498,
+      "learning_rate": 0.00019981501092263503,
+      "loss": 1.1675,
+      "step": 103
+    },
+    {
+      "epoch": 0.050078247261345854,
+      "grad_norm": 1.3210868835449219,
+      "learning_rate": 0.00019980540579732196,
+      "loss": 0.7879,
+      "step": 104
+    },
+    {
+      "epoch": 0.05055976886962803,
+      "grad_norm": 1.8965524435043335,
+      "learning_rate": 0.00019979555782386434,
+      "loss": 1.0122,
+      "step": 105
+    },
+    {
+      "epoch": 0.051041290477910194,
+      "grad_norm": 0.71902996301651,
+      "learning_rate": 0.00019978546702622443,
+      "loss": 1.0812,
+      "step": 106
+    },
+    {
+      "epoch": 0.05152281208619237,
+      "grad_norm": 0.6849554181098938,
+      "learning_rate": 0.00019977513342895532,
+      "loss": 0.3471,
+      "step": 107
+    },
+    {
+      "epoch": 0.05200433369447454,
+      "grad_norm": 1.114531397819519,
+      "learning_rate": 0.00019976455705720083,
+      "loss": 0.8091,
+      "step": 108
+    },
+    {
+      "epoch": 0.05248585530275671,
+      "grad_norm": 0.9422350525856018,
+      "learning_rate": 0.0001997537379366956,
+      "loss": 1.328,
+      "step": 109
+    },
+    {
+      "epoch": 0.05296737691103888,
+      "grad_norm": 0.8205842971801758,
+      "learning_rate": 0.00019974267609376494,
+      "loss": 1.325,
+      "step": 110
+    },
+    {
+      "epoch": 0.05344889851932105,
+      "grad_norm": 1.2657097578048706,
+      "learning_rate": 0.00019973137155532462,
+      "loss": 0.8736,
+      "step": 111
+    },
+    {
+      "epoch": 0.053930420127603226,
+      "grad_norm": 1.1228113174438477,
+      "learning_rate": 0.00019971982434888107,
+      "loss": 1.0789,
+      "step": 112
+    },
+    {
+      "epoch": 0.0544119417358854,
+      "grad_norm": 1.5230176448822021,
+      "learning_rate": 0.00019970803450253114,
+      "loss": 1.3641,
+      "step": 113
+    },
+    {
+      "epoch": 0.05489346334416757,
+      "grad_norm": 1.4824843406677246,
+      "learning_rate": 0.0001996960020449621,
+      "loss": 1.023,
+      "step": 114
+    },
+    {
+      "epoch": 0.05537498495244974,
+      "grad_norm": 0.8390150666236877,
+      "learning_rate": 0.00019968372700545145,
+      "loss": 1.0051,
+      "step": 115
+    },
+    {
+      "epoch": 0.05585650656073191,
+      "grad_norm": 0.5034816265106201,
+      "learning_rate": 0.00019967120941386709,
+      "loss": 0.8851,
+      "step": 116
+    },
+    {
+      "epoch": 0.056338028169014086,
+      "grad_norm": 0.5197919011116028,
+      "learning_rate": 0.000199658449300667,
+      "loss": 0.9949,
+      "step": 117
+    },
+    {
+      "epoch": 0.05681954977729626,
+      "grad_norm": 0.7896221280097961,
+      "learning_rate": 0.00019964544669689928,
+      "loss": 1.3479,
+      "step": 118
+    },
+    {
+      "epoch": 0.057301071385578425,
+      "grad_norm": 0.9110464453697205,
+      "learning_rate": 0.00019963220163420214,
+      "loss": 1.4343,
+      "step": 119
+    },
+    {
+      "epoch": 0.0577825929938606,
+      "grad_norm": 0.8352183699607849,
+      "learning_rate": 0.0001996187141448036,
+      "loss": 0.5157,
+      "step": 120
+    },
+    {
+      "epoch": 0.05826411460214277,
+      "grad_norm": 0.6002696752548218,
+      "learning_rate": 0.0001996049842615217,
+      "loss": 0.8888,
+      "step": 121
+    },
+    {
+      "epoch": 0.058745636210424945,
+      "grad_norm": 1.1759757995605469,
+      "learning_rate": 0.0001995910120177642,
+      "loss": 1.1579,
+      "step": 122
+    },
+    {
+      "epoch": 0.05922715781870711,
+      "grad_norm": 0.6822336912155151,
+      "learning_rate": 0.00019957679744752859,
+      "loss": 0.8854,
+      "step": 123
+    },
+    {
+      "epoch": 0.059708679426989285,
+      "grad_norm": 1.1303515434265137,
+      "learning_rate": 0.00019956234058540195,
+      "loss": 0.9317,
+      "step": 124
+    },
+    {
+      "epoch": 0.06019020103527146,
+      "grad_norm": 1.6187382936477661,
+      "learning_rate": 0.00019954764146656105,
+      "loss": 0.6916,
+      "step": 125
+    },
+    {
+      "epoch": 0.06067172264355363,
+      "grad_norm": 0.7719541788101196,
+      "learning_rate": 0.00019953270012677195,
+      "loss": 0.6124,
+      "step": 126
+    },
+    {
+      "epoch": 0.0611532442518358,
+      "grad_norm": 0.6049287915229797,
+      "learning_rate": 0.00019951751660239015,
+      "loss": 0.8589,
+      "step": 127
+    },
+    {
+      "epoch": 0.06163476586011797,
+      "grad_norm": 1.2466943264007568,
+      "learning_rate": 0.00019950209093036052,
+      "loss": 0.7661,
+      "step": 128
+    },
+    {
+      "epoch": 0.062116287468400144,
+      "grad_norm": 0.6226364970207214,
+      "learning_rate": 0.000199486423148217,
+      "loss": 0.7367,
+      "step": 129
+    },
+    {
+      "epoch": 0.06259780907668232,
+      "grad_norm": 0.714819610118866,
+      "learning_rate": 0.00019947051329408276,
+      "loss": 0.6814,
+      "step": 130
+    },
+    {
+      "epoch": 0.06307933068496449,
+      "grad_norm": 0.8032063841819763,
+      "learning_rate": 0.00019945436140666981,
+      "loss": 1.4286,
+      "step": 131
+    },
+    {
+      "epoch": 0.06356085229324666,
+      "grad_norm": 0.940105676651001,
+      "learning_rate": 0.0001994379675252793,
+      "loss": 1.2501,
+      "step": 132
+    },
+    {
+      "epoch": 0.06404237390152884,
+      "grad_norm": 1.491990327835083,
+      "learning_rate": 0.00019942133168980103,
+      "loss": 1.5854,
+      "step": 133
+    },
+    {
+      "epoch": 0.064523895509811,
+      "grad_norm": 0.8070606589317322,
+      "learning_rate": 0.00019940445394071355,
+      "loss": 0.9551,
+      "step": 134
+    },
+    {
+      "epoch": 0.06500541711809317,
+      "grad_norm": 0.8355477452278137,
+      "learning_rate": 0.0001993873343190842,
+      "loss": 0.7561,
+      "step": 135
+    },
+    {
+      "epoch": 0.06548693872637534,
+      "grad_norm": 0.6875390410423279,
+      "learning_rate": 0.00019936997286656855,
+      "loss": 1.025,
+      "step": 136
+    },
+    {
+      "epoch": 0.06596846033465752,
+      "grad_norm": 0.6399025321006775,
+      "learning_rate": 0.00019935236962541092,
+      "loss": 0.9071,
+      "step": 137
+    },
+    {
+      "epoch": 0.06644998194293969,
+      "grad_norm": 0.6599445343017578,
+      "learning_rate": 0.00019933452463844376,
+      "loss": 0.9648,
+      "step": 138
+    },
+    {
+      "epoch": 0.06693150355122186,
+      "grad_norm": 1.112597942352295,
+      "learning_rate": 0.00019931643794908772,
+      "loss": 0.7063,
+      "step": 139
+    },
+    {
+      "epoch": 0.06741302515950404,
+      "grad_norm": 0.9359776973724365,
+      "learning_rate": 0.00019929810960135172,
+      "loss": 0.9689,
+      "step": 140
+    },
+    {
+      "epoch": 0.06789454676778621,
+      "grad_norm": 1.7165433168411255,
+      "learning_rate": 0.00019927953963983254,
+      "loss": 1.0222,
+      "step": 141
+    },
+    {
+      "epoch": 0.06837606837606838,
+      "grad_norm": 1.3653531074523926,
+      "learning_rate": 0.00019926072810971492,
+      "loss": 1.159,
+      "step": 142
+    },
+    {
+      "epoch": 0.06885758998435054,
+      "grad_norm": 1.2677444219589233,
+      "learning_rate": 0.00019924167505677137,
+      "loss": 1.4072,
+      "step": 143
+    },
+    {
+      "epoch": 0.06933911159263272,
+      "grad_norm": 2.713906764984131,
+      "learning_rate": 0.00019922238052736215,
+      "loss": 0.8661,
+      "step": 144
+    },
+    {
+      "epoch": 0.06982063320091489,
+      "grad_norm": 1.47153902053833,
+      "learning_rate": 0.00019920284456843498,
+      "loss": 1.2206,
+      "step": 145
+    },
+    {
+      "epoch": 0.07030215480919706,
+      "grad_norm": 1.4640045166015625,
+      "learning_rate": 0.00019918306722752505,
+      "loss": 1.3151,
+      "step": 146
+    },
+    {
+      "epoch": 0.07078367641747924,
+      "grad_norm": 1.0816256999969482,
+      "learning_rate": 0.00019916304855275497,
+      "loss": 0.8234,
+      "step": 147
+    },
+    {
+      "epoch": 0.07126519802576141,
+      "grad_norm": 1.3785395622253418,
+      "learning_rate": 0.00019914278859283445,
+      "loss": 0.9928,
+      "step": 148
+    },
+    {
+      "epoch": 0.07174671963404358,
+      "grad_norm": 1.7482737302780151,
+      "learning_rate": 0.0001991222873970604,
+      "loss": 0.3996,
+      "step": 149
+    },
+    {
+      "epoch": 0.07222824124232576,
+      "grad_norm": 0.5450143814086914,
+      "learning_rate": 0.00019910154501531663,
+      "loss": 0.6338,
+      "step": 150
+    },
+    {
+      "epoch": 0.07270976285060791,
+      "grad_norm": 0.9865407943725586,
+      "learning_rate": 0.0001990805614980739,
+      "loss": 0.803,
+      "step": 151
+    },
+    {
+      "epoch": 0.07319128445889009,
+      "grad_norm": 0.5290705561637878,
+      "learning_rate": 0.00019905933689638955,
+      "loss": 0.8526,
+      "step": 152
+    },
+    {
+      "epoch": 0.07367280606717226,
+      "grad_norm": 0.8891838788986206,
+      "learning_rate": 0.00019903787126190772,
+      "loss": 0.7215,
+      "step": 153
+    },
+    {
+      "epoch": 0.07415432767545443,
+      "grad_norm": 1.012518048286438,
+      "learning_rate": 0.00019901616464685888,
+      "loss": 0.8488,
+      "step": 154
+    },
+    {
+      "epoch": 0.07463584928373661,
+      "grad_norm": 1.3221955299377441,
+      "learning_rate": 0.00019899421710405996,
+      "loss": 1.1629,
+      "step": 155
+    },
+    {
+      "epoch": 0.07511737089201878,
+      "grad_norm": 1.0022908449172974,
+      "learning_rate": 0.00019897202868691407,
+      "loss": 0.9271,
+      "step": 156
+    },
+    {
+      "epoch": 0.07559889250030095,
+      "grad_norm": 1.7466130256652832,
+      "learning_rate": 0.00019894959944941038,
+      "loss": 0.703,
+      "step": 157
+    },
+    {
+      "epoch": 0.07608041410858313,
+      "grad_norm": 0.8413887619972229,
+      "learning_rate": 0.0001989269294461242,
+      "loss": 0.8749,
+      "step": 158
+    },
+    {
+      "epoch": 0.0765619357168653,
+      "grad_norm": 0.904451310634613,
+      "learning_rate": 0.0001989040187322164,
+      "loss": 0.9528,
+      "step": 159
+    },
+    {
+      "epoch": 0.07704345732514746,
+      "grad_norm": 0.7697594165802002,
+      "learning_rate": 0.00019888086736343384,
+      "loss": 0.9407,
+      "step": 160
+    },
+    {
+      "epoch": 0.07752497893342963,
+      "grad_norm": 1.4493168592453003,
+      "learning_rate": 0.0001988574753961087,
+      "loss": 0.7736,
+      "step": 161
+    },
+    {
+      "epoch": 0.0780065005417118,
+      "grad_norm": 0.8093467354774475,
+      "learning_rate": 0.00019883384288715874,
+      "loss": 0.7555,
+      "step": 162
+    },
+    {
+      "epoch": 0.07848802214999398,
+      "grad_norm": 1.269743800163269,
+      "learning_rate": 0.000198809969894087,
+      "loss": 1.3101,
+      "step": 163
+    },
+    {
+      "epoch": 0.07896954375827615,
+      "grad_norm": 0.961480438709259,
+      "learning_rate": 0.0001987858564749816,
+      "loss": 1.049,
+      "step": 164
+    },
+    {
+      "epoch": 0.07945106536655833,
+      "grad_norm": 0.8676347136497498,
+      "learning_rate": 0.00019876150268851572,
+      "loss": 1.0772,
+      "step": 165
+    },
+    {
+      "epoch": 0.0799325869748405,
+      "grad_norm": 1.016195297241211,
+      "learning_rate": 0.00019873690859394737,
+      "loss": 0.971,
+      "step": 166
+    },
+    {
+      "epoch": 0.08041410858312267,
+      "grad_norm": 1.0543973445892334,
+      "learning_rate": 0.0001987120742511193,
+      "loss": 0.8215,
+      "step": 167
+    },
+    {
+      "epoch": 0.08089563019140483,
+      "grad_norm": 0.9103195071220398,
+      "learning_rate": 0.0001986869997204589,
+      "loss": 0.4004,
+      "step": 168
+    },
+    {
+      "epoch": 0.081377151799687,
+      "grad_norm": 1.200688362121582,
+      "learning_rate": 0.00019866168506297788,
+      "loss": 0.855,
+      "step": 169
+    },
+    {
+      "epoch": 0.08185867340796918,
+      "grad_norm": 0.5053849816322327,
+      "learning_rate": 0.00019863613034027224,
+      "loss": 0.7865,
+      "step": 170
+    },
+    {
+      "epoch": 0.08234019501625135,
+      "grad_norm": 2.139782190322876,
+      "learning_rate": 0.00019861033561452223,
+      "loss": 0.7017,
+      "step": 171
+    },
+    {
+      "epoch": 0.08282171662453353,
+      "grad_norm": 1.6163843870162964,
+      "learning_rate": 0.00019858430094849195,
+      "loss": 0.9894,
+      "step": 172
+    },
+    {
+      "epoch": 0.0833032382328157,
+      "grad_norm": 0.6523648500442505,
+      "learning_rate": 0.0001985580264055294,
+      "loss": 0.5497,
+      "step": 173
+    },
+    {
+      "epoch": 0.08378475984109787,
+      "grad_norm": 0.8324865102767944,
+      "learning_rate": 0.00019853151204956616,
+      "loss": 0.7934,
+      "step": 174
+    },
+    {
+      "epoch": 0.08426628144938005,
+      "grad_norm": 1.4201933145523071,
+      "learning_rate": 0.00019850475794511749,
+      "loss": 0.3068,
+      "step": 175
+    },
+    {
+      "epoch": 0.08474780305766222,
+      "grad_norm": 0.9500930309295654,
+      "learning_rate": 0.00019847776415728185,
+      "loss": 0.6451,
+      "step": 176
+    },
+    {
+      "epoch": 0.08522932466594438,
+      "grad_norm": 0.7073138356208801,
+      "learning_rate": 0.000198450530751741,
+      "loss": 1.1907,
+      "step": 177
+    },
+    {
+      "epoch": 0.08571084627422655,
+      "grad_norm": 0.6987308859825134,
+      "learning_rate": 0.00019842305779475968,
+      "loss": 0.5881,
+      "step": 178
+    },
+    {
+      "epoch": 0.08619236788250872,
+      "grad_norm": 0.8138691186904907,
+      "learning_rate": 0.00019839534535318558,
+      "loss": 1.0889,
+      "step": 179
+    },
+    {
+      "epoch": 0.0866738894907909,
+      "grad_norm": 0.7772369980812073,
+      "learning_rate": 0.00019836739349444899,
+      "loss": 0.7811,
+      "step": 180
+    },
+    {
+      "epoch": 0.08715541109907307,
+      "grad_norm": 0.8350958228111267,
+      "learning_rate": 0.00019833920228656292,
+      "loss": 1.0841,
+      "step": 181
+    },
+    {
+      "epoch": 0.08763693270735524,
+      "grad_norm": 1.1595239639282227,
+      "learning_rate": 0.0001983107717981226,
+      "loss": 1.0653,
+      "step": 182
+    },
+    {
+      "epoch": 0.08811845431563742,
+      "grad_norm": 0.6263972520828247,
+      "learning_rate": 0.00019828210209830562,
+      "loss": 0.5907,
+      "step": 183
+    },
+    {
+      "epoch": 0.08859997592391959,
+      "grad_norm": 0.4528619349002838,
+      "learning_rate": 0.00019825319325687154,
+      "loss": 0.6648,
+      "step": 184
+    },
+    {
+      "epoch": 0.08908149753220176,
+      "grad_norm": 1.5184047222137451,
+      "learning_rate": 0.00019822404534416182,
+      "loss": 0.7916,
+      "step": 185
+    },
+    {
+      "epoch": 0.08956301914048392,
+      "grad_norm": 0.7816428542137146,
+      "learning_rate": 0.00019819465843109963,
+      "loss": 0.9215,
+      "step": 186
+    },
+    {
+      "epoch": 0.0900445407487661,
+      "grad_norm": 0.7619443535804749,
+      "learning_rate": 0.00019816503258918969,
+      "loss": 0.447,
+      "step": 187
+    },
+    {
+      "epoch": 0.09052606235704827,
+      "grad_norm": 1.0862969160079956,
+      "learning_rate": 0.00019813516789051808,
+      "loss": 0.6545,
+      "step": 188
+    },
+    {
+      "epoch": 0.09100758396533044,
+      "grad_norm": 1.1096842288970947,
+      "learning_rate": 0.0001981050644077521,
+      "loss": 0.9364,
+      "step": 189
+    },
+    {
+      "epoch": 0.09148910557361262,
+      "grad_norm": 1.102784514427185,
+      "learning_rate": 0.00019807472221414002,
+      "loss": 0.5648,
+      "step": 190
+    },
+    {
+      "epoch": 0.09197062718189479,
+      "grad_norm": 0.9051916599273682,
+      "learning_rate": 0.00019804414138351094,
+      "loss": 1.0847,
+      "step": 191
+    },
+    {
+      "epoch": 0.09245214879017696,
+      "grad_norm": 1.1457304954528809,
+      "learning_rate": 0.00019801332199027467,
+      "loss": 1.2033,
+      "step": 192
+    },
+    {
+      "epoch": 0.09293367039845914,
+      "grad_norm": 0.9964216351509094,
+      "learning_rate": 0.00019798226410942146,
+      "loss": 0.8802,
+      "step": 193
+    },
+    {
+      "epoch": 0.0934151920067413,
+      "grad_norm": 1.1561510562896729,
+      "learning_rate": 0.00019795096781652182,
+      "loss": 1.1658,
+      "step": 194
+    },
+    {
+      "epoch": 0.09389671361502347,
+      "grad_norm": 1.1975256204605103,
+      "learning_rate": 0.00019791943318772643,
+      "loss": 1.4838,
+      "step": 195
+    },
+    {
+      "epoch": 0.09437823522330564,
+      "grad_norm": 0.8118493556976318,
+      "learning_rate": 0.00019788766029976587,
+      "loss": 0.8335,
+      "step": 196
+    },
+    {
+      "epoch": 0.09485975683158782,
+      "grad_norm": 0.6374335885047913,
+      "learning_rate": 0.0001978556492299504,
+      "loss": 0.6217,
+      "step": 197
+    },
+    {
+      "epoch": 0.09534127843986999,
+      "grad_norm": 0.48891177773475647,
+      "learning_rate": 0.00019782340005616996,
+      "loss": 0.5794,
+      "step": 198
+    },
+    {
+      "epoch": 0.09582280004815216,
+      "grad_norm": 0.6944652199745178,
+      "learning_rate": 0.0001977909128568937,
+      "loss": 0.6489,
+      "step": 199
+    },
+    {
+      "epoch": 0.09630432165643434,
+      "grad_norm": 1.1781611442565918,
+      "learning_rate": 0.00019775818771117,
+      "loss": 0.7355,
+      "step": 200
+    },
+    {
+      "epoch": 0.09678584326471651,
+      "grad_norm": 0.6658651232719421,
+      "learning_rate": 0.00019772522469862626,
+      "loss": 0.6897,
+      "step": 201
+    },
+    {
+      "epoch": 0.09726736487299868,
+      "grad_norm": 0.9195631146430969,
+      "learning_rate": 0.00019769202389946863,
+      "loss": 0.8157,
+      "step": 202
+    },
+    {
+      "epoch": 0.09774888648128084,
+      "grad_norm": 0.7591468095779419,
+      "learning_rate": 0.0001976585853944818,
+      "loss": 0.8502,
+      "step": 203
+    },
+    {
+      "epoch": 0.09823040808956301,
+      "grad_norm": 0.5556705594062805,
+      "learning_rate": 0.0001976249092650289,
+      "loss": 0.6078,
+      "step": 204
+    },
+    {
+      "epoch": 0.09871192969784519,
+      "grad_norm": 0.5372620820999146,
+      "learning_rate": 0.00019759099559305124,
+      "loss": 0.6814,
+      "step": 205
+    },
+    {
+      "epoch": 0.09919345130612736,
+      "grad_norm": 0.6747313141822815,
+      "learning_rate": 0.00019755684446106812,
+      "loss": 0.693,
+      "step": 206
+    },
+    {
+      "epoch": 0.09967497291440953,
+      "grad_norm": 0.9310837388038635,
+      "learning_rate": 0.00019752245595217662,
+      "loss": 1.0744,
+      "step": 207
+    },
+    {
+      "epoch": 0.10015649452269171,
+      "grad_norm": 0.7573024034500122,
+      "learning_rate": 0.00019748783015005144,
+      "loss": 0.7726,
+      "step": 208
+    },
+    {
+      "epoch": 0.10063801613097388,
+      "grad_norm": 0.8931111693382263,
+      "learning_rate": 0.00019745296713894465,
+      "loss": 0.542,
+      "step": 209
+    },
+    {
+      "epoch": 0.10111953773925605,
+      "grad_norm": 0.8912571668624878,
+      "learning_rate": 0.00019741786700368548,
+      "loss": 0.4853,
+      "step": 210
+    },
+    {
+      "epoch": 0.10160105934753821,
+      "grad_norm": 0.8952215313911438,
+      "learning_rate": 0.00019738252982968017,
+      "loss": 0.6625,
+      "step": 211
+    },
+    {
+      "epoch": 0.10208258095582039,
+      "grad_norm": 0.5370394587516785,
+      "learning_rate": 0.00019734695570291168,
+      "loss": 0.6617,
+      "step": 212
+    },
+    {
+      "epoch": 0.10256410256410256,
+      "grad_norm": 0.8716063499450684,
+      "learning_rate": 0.00019731114470993962,
+      "loss": 0.7222,
+      "step": 213
+    },
+    {
+      "epoch": 0.10304562417238473,
+      "grad_norm": 1.1294581890106201,
+      "learning_rate": 0.0001972750969378998,
+      "loss": 0.8233,
+      "step": 214
+    },
+    {
+      "epoch": 0.10352714578066691,
+      "grad_norm": 0.9793927669525146,
+      "learning_rate": 0.00019723881247450434,
+      "loss": 0.7701,
+      "step": 215
+    },
+    {
+      "epoch": 0.10400866738894908,
+      "grad_norm": 1.0255509614944458,
+      "learning_rate": 0.0001972022914080411,
+      "loss": 1.3642,
+      "step": 216
+    },
+    {
+      "epoch": 0.10449018899723125,
+      "grad_norm": 0.6845418214797974,
+      "learning_rate": 0.00019716553382737379,
+      "loss": 0.3581,
+      "step": 217
+    },
+    {
+      "epoch": 0.10497171060551343,
+      "grad_norm": 1.35993492603302,
+      "learning_rate": 0.00019712853982194152,
+      "loss": 0.996,
+      "step": 218
+    },
+    {
+      "epoch": 0.1054532322137956,
+      "grad_norm": 0.62435382604599,
+      "learning_rate": 0.00019709130948175876,
+      "loss": 0.6824,
+      "step": 219
+    },
+    {
+      "epoch": 0.10593475382207776,
+      "grad_norm": 0.5951329469680786,
+      "learning_rate": 0.0001970538428974149,
+      "loss": 0.8215,
+      "step": 220
+    },
+    {
+      "epoch": 0.10641627543035993,
+      "grad_norm": 0.7611098885536194,
+      "learning_rate": 0.00019701614016007436,
+      "loss": 0.9625,
+      "step": 221
+    },
+    {
+      "epoch": 0.1068977970386421,
+      "grad_norm": 1.2601478099822998,
+      "learning_rate": 0.00019697820136147597,
+      "loss": 0.8209,
+      "step": 222
+    },
+    {
+      "epoch": 0.10737931864692428,
+      "grad_norm": 1.0291566848754883,
+      "learning_rate": 0.00019694002659393305,
+      "loss": 1.0147,
+      "step": 223
+    },
+    {
+      "epoch": 0.10786084025520645,
+      "grad_norm": 0.6460703015327454,
+      "learning_rate": 0.0001969016159503331,
+      "loss": 1.0697,
+      "step": 224
+    },
+    {
+      "epoch": 0.10834236186348863,
+      "grad_norm": 0.6381105780601501,
+      "learning_rate": 0.00019686296952413747,
+      "loss": 0.4803,
+      "step": 225
+    },
+    {
+      "epoch": 0.1088238834717708,
+      "grad_norm": 1.399173378944397,
+      "learning_rate": 0.0001968240874093813,
+      "loss": 0.654,
+      "step": 226
+    },
+    {
+      "epoch": 0.10930540508005297,
+      "grad_norm": 1.1444861888885498,
+      "learning_rate": 0.00019678496970067325,
+      "loss": 0.4746,
+      "step": 227
+    },
+    {
+      "epoch": 0.10978692668833515,
+      "grad_norm": 0.8985688090324402,
+      "learning_rate": 0.0001967456164931951,
+      "loss": 0.4273,
+      "step": 228
+    },
+    {
+      "epoch": 0.1102684482966173,
+      "grad_norm": 0.747041642665863,
+      "learning_rate": 0.0001967060278827017,
+      "loss": 0.4249,
+      "step": 229
+    },
+    {
+      "epoch": 0.11074996990489948,
+      "grad_norm": 1.1315629482269287,
+      "learning_rate": 0.00019666620396552076,
+      "loss": 0.7375,
+      "step": 230
+    },
+    {
+      "epoch": 0.11123149151318165,
+      "grad_norm": 1.8923050165176392,
+      "learning_rate": 0.00019662614483855246,
+      "loss": 0.9888,
+      "step": 231
+    },
+    {
+      "epoch": 0.11171301312146383,
+      "grad_norm": 2.3420629501342773,
+      "learning_rate": 0.00019658585059926934,
+      "loss": 0.5667,
+      "step": 232
+    },
+    {
+      "epoch": 0.112194534729746,
+      "grad_norm": 0.8595489263534546,
+      "learning_rate": 0.00019654532134571594,
+      "loss": 0.8782,
+      "step": 233
+    },
+    {
+      "epoch": 0.11267605633802817,
+      "grad_norm": 0.9815763235092163,
+      "learning_rate": 0.00019650455717650878,
+      "loss": 0.5611,
+      "step": 234
+    },
+    {
+      "epoch": 0.11315757794631034,
+      "grad_norm": 0.5129497647285461,
+      "learning_rate": 0.00019646355819083589,
+      "loss": 0.5094,
+      "step": 235
+    },
+    {
+      "epoch": 0.11363909955459252,
+      "grad_norm": 1.0263252258300781,
+      "learning_rate": 0.0001964223244884566,
+      "loss": 0.634,
+      "step": 236
+    },
+    {
+      "epoch": 0.11412062116287468,
+      "grad_norm": 1.8378629684448242,
+      "learning_rate": 0.00019638085616970153,
+      "loss": 0.7657,
+      "step": 237
+    },
+    {
+      "epoch": 0.11460214277115685,
+      "grad_norm": 1.0862857103347778,
+      "learning_rate": 0.00019633915333547202,
+      "loss": 0.8037,
+      "step": 238
+    },
+    {
+      "epoch": 0.11508366437943902,
+      "grad_norm": 0.7772231101989746,
+      "learning_rate": 0.00019629721608724004,
+      "loss": 0.6409,
+      "step": 239
+    },
+    {
+      "epoch": 0.1155651859877212,
+      "grad_norm": 0.6412681341171265,
+      "learning_rate": 0.0001962550445270481,
+      "loss": 0.4673,
+      "step": 240
+    },
+    {
+      "epoch": 0.11604670759600337,
+      "grad_norm": 1.4209880828857422,
+      "learning_rate": 0.00019621263875750864,
+      "loss": 0.5923,
+      "step": 241
+    },
+    {
+      "epoch": 0.11652822920428554,
+      "grad_norm": 0.7568668127059937,
+      "learning_rate": 0.00019616999888180406,
+      "loss": 0.3972,
+      "step": 242
+    },
+    {
+      "epoch": 0.11700975081256772,
+      "grad_norm": 1.2214607000350952,
+      "learning_rate": 0.0001961271250036865,
+      "loss": 0.8049,
+      "step": 243
+    },
+    {
+      "epoch": 0.11749127242084989,
+      "grad_norm": 0.8679167032241821,
+      "learning_rate": 0.0001960840172274773,
+      "loss": 0.5002,
+      "step": 244
+    },
+    {
+      "epoch": 0.11797279402913206,
+      "grad_norm": 0.5739406943321228,
+      "learning_rate": 0.00019604067565806704,
+      "loss": 0.4761,
+      "step": 245
+    },
+    {
+      "epoch": 0.11845431563741422,
+      "grad_norm": 0.822028636932373,
+      "learning_rate": 0.00019599710040091512,
+      "loss": 0.6141,
+      "step": 246
+    },
+    {
+      "epoch": 0.1189358372456964,
+      "grad_norm": 0.6882205009460449,
+      "learning_rate": 0.00019595329156204955,
+      "loss": 0.174,
+      "step": 247
+    },
+    {
+      "epoch": 0.11941735885397857,
+      "grad_norm": 1.7359886169433594,
+      "learning_rate": 0.00019590924924806676,
+      "loss": 0.777,
+      "step": 248
+    },
+    {
+      "epoch": 0.11989888046226074,
+      "grad_norm": 0.9156647324562073,
+      "learning_rate": 0.0001958649735661312,
+      "loss": 0.9562,
+      "step": 249
+    },
+    {
+      "epoch": 0.12038040207054292,
+      "grad_norm": 0.8275269865989685,
+      "learning_rate": 0.00019582046462397515,
+      "loss": 0.6719,
+      "step": 250
+    },
+    {
+      "epoch": 0.12086192367882509,
+      "grad_norm": 1.1353651285171509,
+      "learning_rate": 0.00019577572252989854,
+      "loss": 1.0057,
+      "step": 251
+    },
+    {
+      "epoch": 0.12134344528710726,
+      "grad_norm": 1.2231826782226562,
+      "learning_rate": 0.00019573074739276858,
+      "loss": 1.0401,
+      "step": 252
+    },
+    {
+      "epoch": 0.12182496689538944,
+      "grad_norm": 1.7041336297988892,
+      "learning_rate": 0.00019568553932201947,
+      "loss": 0.631,
+      "step": 253
+    },
+    {
+      "epoch": 0.1223064885036716,
+      "grad_norm": 0.949766218662262,
+      "learning_rate": 0.00019564009842765225,
+      "loss": 1.0427,
+      "step": 254
+    },
+    {
+      "epoch": 0.12278801011195377,
+      "grad_norm": 0.9756664633750916,
+      "learning_rate": 0.00019559442482023444,
+      "loss": 1.0868,
+      "step": 255
+    },
+    {
+      "epoch": 0.12326953172023594,
+      "grad_norm": 2.625432014465332,
+      "learning_rate": 0.0001955485186108998,
+      "loss": 0.9394,
+      "step": 256
+    },
+    {
+      "epoch": 0.12375105332851812,
+      "grad_norm": 0.9411224722862244,
+      "learning_rate": 0.00019550237991134805,
+      "loss": 0.5891,
+      "step": 257
+    },
+    {
+      "epoch": 0.12423257493680029,
+      "grad_norm": 0.8636229634284973,
+      "learning_rate": 0.00019545600883384467,
+      "loss": 0.9391,
+      "step": 258
+    },
+    {
+      "epoch": 0.12471409654508246,
+      "grad_norm": 0.5954817533493042,
+      "learning_rate": 0.0001954094054912205,
+      "loss": 0.7277,
+      "step": 259
+    },
+    {
+      "epoch": 0.12519561815336464,
+      "grad_norm": 0.6612427234649658,
+      "learning_rate": 0.00019536256999687157,
+      "loss": 0.4814,
+      "step": 260
+    },
+    {
+      "epoch": 0.1256771397616468,
+      "grad_norm": 0.7988054752349854,
+      "learning_rate": 0.00019531550246475876,
+      "loss": 0.5545,
+      "step": 261
+    },
+    {
+      "epoch": 0.12615866136992898,
+      "grad_norm": 1.6089789867401123,
+      "learning_rate": 0.00019526820300940756,
+      "loss": 1.0021,
+      "step": 262
+    },
+    {
+      "epoch": 0.12664018297821114,
+      "grad_norm": 1.084684133529663,
+      "learning_rate": 0.00019522067174590778,
+      "loss": 1.0537,
+      "step": 263
+    },
+    {
+      "epoch": 0.12712170458649333,
+      "grad_norm": 1.3705826997756958,
+      "learning_rate": 0.00019517290878991324,
+      "loss": 0.7344,
+      "step": 264
+    },
+    {
+      "epoch": 0.1276032261947755,
+      "grad_norm": 1.0224491357803345,
+      "learning_rate": 0.0001951249142576416,
+      "loss": 1.4591,
+      "step": 265
+    },
+    {
+      "epoch": 0.12808474780305767,
+      "grad_norm": 1.1117572784423828,
+      "learning_rate": 0.00019507668826587387,
+      "loss": 0.9157,
+      "step": 266
+    },
+    {
+      "epoch": 0.12856626941133983,
+      "grad_norm": 1.0106638669967651,
+      "learning_rate": 0.0001950282309319544,
+      "loss": 0.998,
+      "step": 267
+    },
+    {
+      "epoch": 0.129047791019622,
+      "grad_norm": 1.1685928106307983,
+      "learning_rate": 0.0001949795423737903,
+      "loss": 1.0733,
+      "step": 268
+    },
+    {
+      "epoch": 0.12952931262790418,
+      "grad_norm": 0.8869622349739075,
+      "learning_rate": 0.00019493062270985144,
+      "loss": 0.9227,
+      "step": 269
+    },
+    {
+      "epoch": 0.13001083423618634,
+      "grad_norm": 0.8877792358398438,
+      "learning_rate": 0.00019488147205916985,
+      "loss": 0.7803,
+      "step": 270
+    },
+    {
+      "epoch": 0.13049235584446853,
+      "grad_norm": 0.8910313844680786,
+      "learning_rate": 0.00019483209054133976,
+      "loss": 0.7102,
+      "step": 271
+    },
+    {
+      "epoch": 0.1309738774527507,
+      "grad_norm": 1.707604169845581,
+      "learning_rate": 0.00019478247827651708,
+      "loss": 0.6851,
+      "step": 272
+    },
+    {
+      "epoch": 0.13145539906103287,
+      "grad_norm": 1.1673096418380737,
+      "learning_rate": 0.00019473263538541914,
+      "loss": 0.5218,
+      "step": 273
+    },
+    {
+      "epoch": 0.13193692066931503,
+      "grad_norm": 1.01649808883667,
+      "learning_rate": 0.00019468256198932455,
+      "loss": 0.6644,
+      "step": 274
+    },
+    {
+      "epoch": 0.13241844227759722,
+      "grad_norm": 0.7683111429214478,
+      "learning_rate": 0.00019463225821007268,
+      "loss": 0.979,
+      "step": 275
+    },
+    {
+      "epoch": 0.13289996388587938,
+      "grad_norm": 0.6378679871559143,
+      "learning_rate": 0.00019458172417006347,
+      "loss": 0.6165,
+      "step": 276
+    },
+    {
+      "epoch": 0.13338148549416154,
+      "grad_norm": 0.6471127271652222,
+      "learning_rate": 0.00019453095999225726,
+      "loss": 0.5204,
+      "step": 277
+    },
+    {
+      "epoch": 0.13386300710244373,
+      "grad_norm": 0.7345598936080933,
+      "learning_rate": 0.0001944799658001742,
+      "loss": 0.4905,
+      "step": 278
+    },
+    {
+      "epoch": 0.13434452871072589,
+      "grad_norm": 0.670501172542572,
+      "learning_rate": 0.00019442874171789418,
+      "loss": 0.4426,
+      "step": 279
+    },
+    {
+      "epoch": 0.13482605031900807,
+      "grad_norm": 0.7248306274414062,
+      "learning_rate": 0.00019437728787005657,
+      "loss": 1.3919,
+      "step": 280
+    },
+    {
+      "epoch": 0.13530757192729023,
+      "grad_norm": 1.8311445713043213,
+      "learning_rate": 0.00019432560438185963,
+      "loss": 0.5238,
+      "step": 281
+    },
+    {
+      "epoch": 0.13578909353557242,
+      "grad_norm": 0.7017165422439575,
+      "learning_rate": 0.00019427369137906046,
+      "loss": 0.8092,
+      "step": 282
+    },
+    {
+      "epoch": 0.13627061514385458,
+      "grad_norm": 1.079423427581787,
+      "learning_rate": 0.00019422154898797472,
+      "loss": 0.6991,
+      "step": 283
+    },
+    {
+      "epoch": 0.13675213675213677,
+      "grad_norm": 0.9701410531997681,
+      "learning_rate": 0.00019416917733547603,
+      "loss": 1.4933,
+      "step": 284
+    },
+    {
+      "epoch": 0.13723365836041893,
+      "grad_norm": 1.552991509437561,
+      "learning_rate": 0.00019411657654899597,
+      "loss": 0.7102,
+      "step": 285
+    },
+    {
+      "epoch": 0.13771517996870108,
+      "grad_norm": 0.9199215769767761,
+      "learning_rate": 0.0001940637467565237,
+      "loss": 0.7313,
+      "step": 286
+    },
+    {
+      "epoch": 0.13819670157698327,
+      "grad_norm": 1.1101661920547485,
+      "learning_rate": 0.00019401068808660546,
+      "loss": 0.8939,
+      "step": 287
+    },
+    {
+      "epoch": 0.13867822318526543,
+      "grad_norm": 0.817594051361084,
+      "learning_rate": 0.0001939574006683445,
+      "loss": 0.7946,
+      "step": 288
+    },
+    {
+      "epoch": 0.13915974479354762,
+      "grad_norm": 0.8862625360488892,
+      "learning_rate": 0.00019390388463140065,
+      "loss": 0.4525,
+      "step": 289
+    },
+    {
+      "epoch": 0.13964126640182978,
+      "grad_norm": 0.799997091293335,
+      "learning_rate": 0.00019385014010598998,
+      "loss": 0.4477,
+      "step": 290
+    },
+    {
+      "epoch": 0.14012278801011196,
+      "grad_norm": 0.8338310718536377,
+      "learning_rate": 0.00019379616722288456,
+      "loss": 0.787,
+      "step": 291
+    },
+    {
+      "epoch": 0.14060430961839412,
+      "grad_norm": 0.9072525501251221,
+      "learning_rate": 0.0001937419661134121,
+      "loss": 0.8017,
+      "step": 292
+    },
+    {
+      "epoch": 0.14108583122667628,
+      "grad_norm": 0.46390998363494873,
+      "learning_rate": 0.0001936875369094556,
+      "loss": 0.6444,
+      "step": 293
+    },
+    {
+      "epoch": 0.14156735283495847,
+      "grad_norm": 0.9819871187210083,
+      "learning_rate": 0.0001936328797434531,
+      "loss": 0.533,
+      "step": 294
+    },
+    {
+      "epoch": 0.14204887444324063,
+      "grad_norm": 0.9907609224319458,
+      "learning_rate": 0.00019357799474839735,
+      "loss": 1.1361,
+      "step": 295
+    },
+    {
+      "epoch": 0.14253039605152282,
+      "grad_norm": 1.1553535461425781,
+      "learning_rate": 0.00019352288205783536,
+      "loss": 0.5737,
+      "step": 296
+    },
+    {
+      "epoch": 0.14301191765980498,
+      "grad_norm": 0.8127950429916382,
+      "learning_rate": 0.00019346754180586825,
+      "loss": 0.4669,
+      "step": 297
+    },
+    {
+      "epoch": 0.14349343926808716,
+      "grad_norm": 0.869855523109436,
+      "learning_rate": 0.00019341197412715082,
+      "loss": 0.9356,
+      "step": 298
+    },
+    {
+      "epoch": 0.14397496087636932,
+      "grad_norm": 0.6495091319084167,
+      "learning_rate": 0.00019335617915689128,
+      "loss": 0.6174,
+      "step": 299
+    },
+    {
+      "epoch": 0.1444564824846515,
+      "grad_norm": 0.7985929846763611,
+      "learning_rate": 0.00019330015703085082,
+      "loss": 0.5252,
+      "step": 300
+    },
+    {
+      "epoch": 0.14493800409293367,
+      "grad_norm": 0.5880833864212036,
+      "learning_rate": 0.00019324390788534343,
+      "loss": 0.3703,
+      "step": 301
+    },
+    {
+      "epoch": 0.14541952570121583,
+      "grad_norm": 1.1846781969070435,
+      "learning_rate": 0.00019318743185723546,
+      "loss": 0.6423,
+      "step": 302
+    },
+    {
+      "epoch": 0.14590104730949802,
+      "grad_norm": 1.47828209400177,
+      "learning_rate": 0.00019313072908394525,
+      "loss": 1.1003,
+      "step": 303
+    },
+    {
+      "epoch": 0.14638256891778018,
+      "grad_norm": 1.1608823537826538,
+      "learning_rate": 0.00019307379970344294,
+      "loss": 0.8453,
+      "step": 304
+    },
+    {
+      "epoch": 0.14686409052606236,
+      "grad_norm": 0.9130098819732666,
+      "learning_rate": 0.00019301664385425004,
+      "loss": 0.6183,
+      "step": 305
+    },
+    {
+      "epoch": 0.14734561213434452,
+      "grad_norm": 0.8722681999206543,
+      "learning_rate": 0.0001929592616754391,
+      "loss": 0.6744,
+      "step": 306
+    },
+    {
+      "epoch": 0.1478271337426267,
+      "grad_norm": 0.7916250228881836,
+      "learning_rate": 0.00019290165330663336,
+      "loss": 0.7351,
+      "step": 307
+    },
+    {
+      "epoch": 0.14830865535090887,
+      "grad_norm": 0.604591965675354,
+      "learning_rate": 0.00019284381888800647,
+      "loss": 0.7715,
+      "step": 308
+    },
+    {
+      "epoch": 0.14879017695919106,
+      "grad_norm": 0.7270418405532837,
+      "learning_rate": 0.00019278575856028206,
+      "loss": 0.94,
+      "step": 309
+    },
+    {
+      "epoch": 0.14927169856747322,
+      "grad_norm": 0.7614173293113708,
+      "learning_rate": 0.00019272747246473345,
+      "loss": 0.4351,
+      "step": 310
+    },
+    {
+      "epoch": 0.14975322017575537,
+      "grad_norm": 0.7510772347450256,
+      "learning_rate": 0.00019266896074318334,
+      "loss": 0.6094,
+      "step": 311
+    },
+    {
+      "epoch": 0.15023474178403756,
+      "grad_norm": 0.6468477845191956,
+      "learning_rate": 0.00019261022353800344,
+      "loss": 0.61,
+      "step": 312
+    },
+    {
+      "epoch": 0.15071626339231972,
+      "grad_norm": 1.057340145111084,
+      "learning_rate": 0.00019255126099211402,
+      "loss": 0.7782,
+      "step": 313
+    },
+    {
+      "epoch": 0.1511977850006019,
+      "grad_norm": 0.5399413704872131,
+      "learning_rate": 0.00019249207324898376,
+      "loss": 0.6912,
+      "step": 314
+    },
+    {
+      "epoch": 0.15167930660888407,
+      "grad_norm": 1.0437463521957397,
+      "learning_rate": 0.0001924326604526292,
+      "loss": 0.7274,
+      "step": 315
+    },
+    {
+      "epoch": 0.15216082821716626,
+      "grad_norm": 0.7240994572639465,
+      "learning_rate": 0.00019237302274761458,
+      "loss": 0.505,
+      "step": 316
+    },
+    {
+      "epoch": 0.15264234982544841,
+      "grad_norm": 1.8743011951446533,
+      "learning_rate": 0.0001923131602790513,
+      "loss": 0.9457,
+      "step": 317
+    },
+    {
+      "epoch": 0.1531238714337306,
+      "grad_norm": 1.084415078163147,
+      "learning_rate": 0.00019225307319259768,
+      "loss": 1.2172,
+      "step": 318
+    },
+    {
+      "epoch": 0.15360539304201276,
+      "grad_norm": 0.8688251972198486,
+      "learning_rate": 0.00019219276163445862,
+      "loss": 0.6531,
+      "step": 319
+    },
+    {
+      "epoch": 0.15408691465029492,
+      "grad_norm": 0.9400373101234436,
+      "learning_rate": 0.00019213222575138522,
+      "loss": 1.4234,
+      "step": 320
+    },
+    {
+      "epoch": 0.1545684362585771,
+      "grad_norm": 1.1395434141159058,
+      "learning_rate": 0.00019207146569067435,
+      "loss": 0.45,
+      "step": 321
+    },
+    {
+      "epoch": 0.15504995786685927,
+      "grad_norm": 1.3613590002059937,
+      "learning_rate": 0.00019201048160016838,
+      "loss": 0.9647,
+      "step": 322
+    },
+    {
+      "epoch": 0.15553147947514145,
+      "grad_norm": 0.7267064452171326,
+      "learning_rate": 0.00019194927362825478,
+      "loss": 0.9339,
+      "step": 323
+    },
+    {
+      "epoch": 0.1560130010834236,
+      "grad_norm": 1.1009801626205444,
+      "learning_rate": 0.00019188784192386587,
+      "loss": 0.9611,
+      "step": 324
+    },
+    {
+      "epoch": 0.1564945226917058,
+      "grad_norm": 0.6520840525627136,
+      "learning_rate": 0.00019182618663647817,
+      "loss": 0.3283,
+      "step": 325
+    },
+    {
+      "epoch": 0.15697604429998796,
+      "grad_norm": 1.0760483741760254,
+      "learning_rate": 0.0001917643079161124,
+      "loss": 0.6973,
+      "step": 326
+    },
+    {
+      "epoch": 0.15745756590827015,
+      "grad_norm": 1.0936267375946045,
+      "learning_rate": 0.00019170220591333283,
+      "loss": 0.3996,
+      "step": 327
+    },
+    {
+      "epoch": 0.1579390875165523,
+      "grad_norm": 1.3938161134719849,
+      "learning_rate": 0.00019163988077924713,
+      "loss": 1.0365,
+      "step": 328
+    },
+    {
+      "epoch": 0.15842060912483447,
+      "grad_norm": 0.8608887195587158,
+      "learning_rate": 0.00019157733266550575,
+      "loss": 0.6497,
+      "step": 329
+    },
+    {
+      "epoch": 0.15890213073311665,
+      "grad_norm": 0.8093037605285645,
+      "learning_rate": 0.00019151456172430183,
+      "loss": 0.4926,
+      "step": 330
+    },
+    {
+      "epoch": 0.1593836523413988,
+      "grad_norm": 1.156674861907959,
+      "learning_rate": 0.0001914515681083707,
+      "loss": 0.9478,
+      "step": 331
+    },
+    {
+      "epoch": 0.159865173949681,
+      "grad_norm": 0.639973521232605,
+      "learning_rate": 0.00019138835197098937,
+      "loss": 0.7327,
+      "step": 332
+    },
+    {
+      "epoch": 0.16034669555796316,
+      "grad_norm": 1.1579989194869995,
+      "learning_rate": 0.00019132491346597643,
+      "loss": 0.8106,
+      "step": 333
+    },
+    {
+      "epoch": 0.16082821716624535,
+      "grad_norm": 0.8499979376792908,
+      "learning_rate": 0.00019126125274769145,
+      "loss": 0.3414,
+      "step": 334
+    },
+    {
+      "epoch": 0.1613097387745275,
+      "grad_norm": 1.4322779178619385,
+      "learning_rate": 0.00019119736997103476,
+      "loss": 0.6139,
+      "step": 335
+    },
+    {
+      "epoch": 0.16179126038280967,
+      "grad_norm": 0.696955144405365,
+      "learning_rate": 0.000191133265291447,
+      "loss": 0.615,
+      "step": 336
+    },
+    {
+      "epoch": 0.16227278199109185,
+      "grad_norm": 0.7599828243255615,
+      "learning_rate": 0.00019106893886490864,
+      "loss": 0.7996,
+      "step": 337
+    },
+    {
+      "epoch": 0.162754303599374,
+      "grad_norm": 0.5892730951309204,
+      "learning_rate": 0.00019100439084793989,
+      "loss": 0.7155,
+      "step": 338
+    },
+    {
+      "epoch": 0.1632358252076562,
+      "grad_norm": 1.2862459421157837,
+      "learning_rate": 0.00019093962139759998,
+      "loss": 1.0595,
+      "step": 339
+    },
+    {
+      "epoch": 0.16371734681593836,
+      "grad_norm": 0.6903579235076904,
+      "learning_rate": 0.000190874630671487,
+      "loss": 0.6998,
+      "step": 340
+    },
+    {
+      "epoch": 0.16419886842422055,
+      "grad_norm": 0.6340907216072083,
+      "learning_rate": 0.00019080941882773745,
+      "loss": 0.8251,
+      "step": 341
+    },
+    {
+      "epoch": 0.1646803900325027,
+      "grad_norm": 0.7512362003326416,
+      "learning_rate": 0.00019074398602502584,
+      "loss": 0.7369,
+      "step": 342
+    },
+    {
+      "epoch": 0.1651619116407849,
+      "grad_norm": 1.1983510255813599,
+      "learning_rate": 0.00019067833242256442,
+      "loss": 0.8641,
+      "step": 343
+    },
+    {
+      "epoch": 0.16564343324906705,
+      "grad_norm": 0.7607635259628296,
+      "learning_rate": 0.0001906124581801025,
+      "loss": 0.406,
+      "step": 344
+    },
+    {
+      "epoch": 0.1661249548573492,
+      "grad_norm": 0.5438935160636902,
+      "learning_rate": 0.0001905463634579264,
+      "loss": 0.5055,
+      "step": 345
+    },
+    {
+      "epoch": 0.1666064764656314,
+      "grad_norm": 1.1291062831878662,
+      "learning_rate": 0.00019048004841685888,
+      "loss": 1.1312,
+      "step": 346
+    },
+    {
+      "epoch": 0.16708799807391356,
+      "grad_norm": 1.3860900402069092,
+      "learning_rate": 0.00019041351321825883,
+      "loss": 1.0954,
+      "step": 347
+    },
+    {
+      "epoch": 0.16756951968219574,
+      "grad_norm": 1.0058181285858154,
+      "learning_rate": 0.00019034675802402068,
+      "loss": 1.1283,
+      "step": 348
+    },
+    {
+      "epoch": 0.1680510412904779,
+      "grad_norm": 0.8692428469657898,
+      "learning_rate": 0.00019027978299657436,
+      "loss": 0.3789,
+      "step": 349
+    },
+    {
+      "epoch": 0.1685325628987601,
+      "grad_norm": 1.0767672061920166,
+      "learning_rate": 0.00019021258829888456,
+      "loss": 1.1003,
+      "step": 350
+    },
+    {
+      "epoch": 0.16901408450704225,
+      "grad_norm": 0.8247532248497009,
+      "learning_rate": 0.00019014517409445052,
+      "loss": 0.3231,
+      "step": 351
+    },
+    {
+      "epoch": 0.16949560611532444,
+      "grad_norm": 1.0215187072753906,
+      "learning_rate": 0.00019007754054730554,
+      "loss": 0.88,
+      "step": 352
+    },
+    {
+      "epoch": 0.1699771277236066,
+      "grad_norm": 0.8507590889930725,
+      "learning_rate": 0.00019000968782201675,
+      "loss": 0.37,
+      "step": 353
+    },
+    {
+      "epoch": 0.17045864933188876,
+      "grad_norm": 0.7044029831886292,
+      "learning_rate": 0.00018994161608368448,
+      "loss": 1.2317,
+      "step": 354
+    },
+    {
+      "epoch": 0.17094017094017094,
+      "grad_norm": 0.7668285369873047,
+      "learning_rate": 0.00018987332549794196,
+      "loss": 0.9292,
+      "step": 355
+    },
+    {
+      "epoch": 0.1714216925484531,
+      "grad_norm": 0.5195323824882507,
+      "learning_rate": 0.00018980481623095502,
+      "loss": 0.6176,
+      "step": 356
+    },
+    {
+      "epoch": 0.1719032141567353,
+      "grad_norm": 0.7647833824157715,
+      "learning_rate": 0.00018973608844942148,
+      "loss": 0.5043,
+      "step": 357
+    },
+    {
+      "epoch": 0.17238473576501745,
+      "grad_norm": 0.7875663638114929,
+      "learning_rate": 0.00018966714232057094,
+      "loss": 0.726,
+      "step": 358
+    },
+    {
+      "epoch": 0.17286625737329964,
+      "grad_norm": 0.6289336681365967,
+      "learning_rate": 0.00018959797801216418,
+      "loss": 0.2639,
+      "step": 359
+    },
+    {
+      "epoch": 0.1733477789815818,
+      "grad_norm": 1.9089537858963013,
+      "learning_rate": 0.000189528595692493,
+      "loss": 0.7418,
+      "step": 360
+    },
+    {
+      "epoch": 0.17382930058986398,
+      "grad_norm": 0.8167737722396851,
+      "learning_rate": 0.00018945899553037956,
+      "loss": 0.7737,
+      "step": 361
+    },
+    {
+      "epoch": 0.17431082219814614,
+      "grad_norm": 0.7543114423751831,
+      "learning_rate": 0.00018938917769517613,
+      "loss": 0.5564,
+      "step": 362
+    },
+    {
+      "epoch": 0.1747923438064283,
+      "grad_norm": 1.0456900596618652,
+      "learning_rate": 0.00018931914235676458,
+      "loss": 0.98,
+      "step": 363
+    },
+    {
+      "epoch": 0.1752738654147105,
+      "grad_norm": 0.49631309509277344,
+      "learning_rate": 0.00018924888968555606,
+      "loss": 0.6173,
+      "step": 364
+    },
+    {
+      "epoch": 0.17575538702299265,
+      "grad_norm": 0.449246883392334,
+      "learning_rate": 0.00018917841985249055,
+      "loss": 0.3526,
+      "step": 365
+    },
+    {
+      "epoch": 0.17623690863127484,
+      "grad_norm": 1.2787824869155884,
+      "learning_rate": 0.0001891077330290363,
+      "loss": 1.2029,
+      "step": 366
+    },
+    {
+      "epoch": 0.176718430239557,
+      "grad_norm": 1.1600645780563354,
+      "learning_rate": 0.00018903682938718977,
+      "loss": 1.1068,
+      "step": 367
+    },
+    {
+      "epoch": 0.17719995184783918,
+      "grad_norm": 0.5357136130332947,
+      "learning_rate": 0.00018896570909947475,
+      "loss": 0.5062,
+      "step": 368
+    },
+    {
+      "epoch": 0.17768147345612134,
+      "grad_norm": 0.872021496295929,
+      "learning_rate": 0.00018889437233894234,
+      "loss": 0.6622,
+      "step": 369
+    },
+    {
+      "epoch": 0.17816299506440353,
+      "grad_norm": 1.1199880838394165,
+      "learning_rate": 0.0001888228192791703,
+      "loss": 1.1557,
+      "step": 370
+    },
+    {
+      "epoch": 0.1786445166726857,
+      "grad_norm": 1.332701563835144,
+      "learning_rate": 0.00018875105009426272,
+      "loss": 0.6267,
+      "step": 371
+    },
+    {
+      "epoch": 0.17912603828096785,
+      "grad_norm": 1.0211436748504639,
+      "learning_rate": 0.00018867906495884955,
+      "loss": 1.06,
+      "step": 372
+    },
+    {
+      "epoch": 0.17960755988925003,
+      "grad_norm": 1.0804567337036133,
+      "learning_rate": 0.0001886068640480862,
+      "loss": 0.8207,
+      "step": 373
+    },
+    {
+      "epoch": 0.1800890814975322,
+      "grad_norm": 1.1362806558609009,
+      "learning_rate": 0.00018853444753765306,
+      "loss": 0.8978,
+      "step": 374
+    },
+    {
+      "epoch": 0.18057060310581438,
+      "grad_norm": 0.8540496826171875,
+      "learning_rate": 0.00018846181560375525,
+      "loss": 0.7343,
+      "step": 375
+    },
+    {
+      "epoch": 0.18105212471409654,
+      "grad_norm": 0.9461526274681091,
+      "learning_rate": 0.0001883889684231219,
+      "loss": 0.7977,
+      "step": 376
+    },
+    {
+      "epoch": 0.18153364632237873,
+      "grad_norm": 1.1165428161621094,
+      "learning_rate": 0.000188315906173006,
+      "loss": 0.8063,
+      "step": 377
+    },
+    {
+      "epoch": 0.1820151679306609,
+      "grad_norm": 0.5946139097213745,
+      "learning_rate": 0.0001882426290311838,
+      "loss": 0.4324,
+      "step": 378
+    },
+    {
+      "epoch": 0.18249668953894305,
+      "grad_norm": 1.2596697807312012,
+      "learning_rate": 0.00018816913717595445,
+      "loss": 1.0515,
+      "step": 379
+    },
+    {
+      "epoch": 0.18297821114722523,
+      "grad_norm": 0.8724560737609863,
+      "learning_rate": 0.00018809543078613953,
+      "loss": 0.4745,
+      "step": 380
+    },
+    {
+      "epoch": 0.1834597327555074,
+      "grad_norm": 1.4185947179794312,
+      "learning_rate": 0.00018802151004108263,
+      "loss": 0.3437,
+      "step": 381
+    },
+    {
+      "epoch": 0.18394125436378958,
+      "grad_norm": 1.0995190143585205,
+      "learning_rate": 0.0001879473751206489,
+      "loss": 1.0071,
+      "step": 382
+    },
+    {
+      "epoch": 0.18442277597207174,
+      "grad_norm": 0.5417447090148926,
+      "learning_rate": 0.00018787302620522467,
+      "loss": 0.3357,
+      "step": 383
+    },
+    {
+      "epoch": 0.18490429758035393,
+      "grad_norm": 0.9899541735649109,
+      "learning_rate": 0.00018779846347571693,
+      "loss": 0.1654,
+      "step": 384
+    },
+    {
+      "epoch": 0.1853858191886361,
+      "grad_norm": 0.5713137984275818,
+      "learning_rate": 0.0001877236871135529,
+      "loss": 0.3284,
+      "step": 385
+    },
+    {
+      "epoch": 0.18586734079691827,
+      "grad_norm": 1.2448984384536743,
+      "learning_rate": 0.00018764869730067968,
+      "loss": 1.2335,
+      "step": 386
+    },
+    {
+      "epoch": 0.18634886240520043,
+      "grad_norm": 0.8601797223091125,
+      "learning_rate": 0.0001875734942195637,
+      "loss": 0.4766,
+      "step": 387
+    },
+    {
+      "epoch": 0.1868303840134826,
+      "grad_norm": 0.8154586553573608,
+      "learning_rate": 0.0001874980780531903,
+      "loss": 0.7463,
+      "step": 388
+    },
+    {
+      "epoch": 0.18731190562176478,
+      "grad_norm": 0.9013132452964783,
+      "learning_rate": 0.00018742244898506337,
+      "loss": 0.6605,
+      "step": 389
+    },
+    {
+      "epoch": 0.18779342723004694,
+      "grad_norm": 1.5045801401138306,
+      "learning_rate": 0.00018734660719920475,
+      "loss": 1.1213,
+      "step": 390
+    },
+    {
+      "epoch": 0.18827494883832913,
+      "grad_norm": 1.017081379890442,
+      "learning_rate": 0.00018727055288015397,
+      "loss": 1.1526,
+      "step": 391
+    },
+    {
+      "epoch": 0.18875647044661129,
+      "grad_norm": 0.8374575972557068,
+      "learning_rate": 0.00018719428621296764,
+      "loss": 1.0737,
+      "step": 392
+    },
+    {
+      "epoch": 0.18923799205489347,
+      "grad_norm": 1.0993142127990723,
+      "learning_rate": 0.00018711780738321897,
+      "loss": 0.3277,
+      "step": 393
+    },
+    {
+      "epoch": 0.18971951366317563,
+      "grad_norm": 0.9795992970466614,
+      "learning_rate": 0.00018704111657699758,
+      "loss": 0.3837,
+      "step": 394
+    },
+    {
+      "epoch": 0.19020103527145782,
+      "grad_norm": 0.5359293818473816,
+      "learning_rate": 0.0001869642139809088,
+      "loss": 0.85,
+      "step": 395
+    },
+    {
+      "epoch": 0.19068255687973998,
+      "grad_norm": 0.6480598449707031,
+      "learning_rate": 0.00018688709978207323,
+      "loss": 0.7276,
+      "step": 396
+    },
+    {
+      "epoch": 0.19116407848802214,
+      "grad_norm": 0.9133846759796143,
+      "learning_rate": 0.00018680977416812644,
+      "loss": 1.0814,
+      "step": 397
+    },
+    {
+      "epoch": 0.19164560009630432,
+      "grad_norm": 0.8055482506752014,
+      "learning_rate": 0.00018673223732721837,
+      "loss": 0.5734,
+      "step": 398
+    },
+    {
+      "epoch": 0.19212712170458648,
+      "grad_norm": 0.83376145362854,
+      "learning_rate": 0.0001866544894480129,
+      "loss": 1.0693,
+      "step": 399
+    },
+    {
+      "epoch": 0.19260864331286867,
+      "grad_norm": 0.6174197196960449,
+      "learning_rate": 0.00018657653071968747,
+      "loss": 0.8912,
+      "step": 400
+    },
+    {
+      "epoch": 0.19309016492115083,
+      "grad_norm": 0.9090850353240967,
+      "learning_rate": 0.00018649836133193253,
+      "loss": 0.6061,
+      "step": 401
+    },
+    {
+      "epoch": 0.19357168652943302,
+      "grad_norm": 0.7542705535888672,
+      "learning_rate": 0.00018641998147495112,
+      "loss": 0.3132,
+      "step": 402
+    },
+    {
+      "epoch": 0.19405320813771518,
+      "grad_norm": 0.8012394309043884,
+      "learning_rate": 0.00018634139133945837,
+      "loss": 0.4904,
+      "step": 403
+    },
+    {
+      "epoch": 0.19453472974599736,
+      "grad_norm": 1.5305569171905518,
+      "learning_rate": 0.00018626259111668105,
+      "loss": 1.2355,
+      "step": 404
+    },
+    {
+      "epoch": 0.19501625135427952,
+      "grad_norm": 0.6317727565765381,
+      "learning_rate": 0.00018618358099835723,
+      "loss": 0.7488,
+      "step": 405
+    },
+    {
+      "epoch": 0.19549777296256168,
+      "grad_norm": 1.183491826057434,
+      "learning_rate": 0.00018610436117673555,
+      "loss": 0.484,
+      "step": 406
+    },
+    {
+      "epoch": 0.19597929457084387,
+      "grad_norm": 1.0415927171707153,
+      "learning_rate": 0.00018602493184457505,
+      "loss": 0.4665,
+      "step": 407
+    },
+    {
+      "epoch": 0.19646081617912603,
+      "grad_norm": 0.8337236046791077,
+      "learning_rate": 0.00018594529319514437,
+      "loss": 0.3315,
+      "step": 408
+    },
+    {
+      "epoch": 0.19694233778740822,
+      "grad_norm": 1.061383605003357,
+      "learning_rate": 0.00018586544542222169,
+      "loss": 1.101,
+      "step": 409
+    },
+    {
+      "epoch": 0.19742385939569038,
+      "grad_norm": 1.9225060939788818,
+      "learning_rate": 0.00018578538872009384,
+      "loss": 0.6274,
+      "step": 410
+    },
+    {
+      "epoch": 0.19790538100397256,
+      "grad_norm": 0.5955725312232971,
+      "learning_rate": 0.00018570512328355612,
+      "loss": 0.5071,
+      "step": 411
+    },
+    {
+      "epoch": 0.19838690261225472,
+      "grad_norm": 1.295961618423462,
+      "learning_rate": 0.00018562464930791167,
+      "loss": 0.9883,
+      "step": 412
+    },
+    {
+      "epoch": 0.1988684242205369,
+      "grad_norm": 1.068457007408142,
+      "learning_rate": 0.00018554396698897116,
+      "loss": 0.5021,
+      "step": 413
+    },
+    {
+      "epoch": 0.19934994582881907,
+      "grad_norm": 0.5233125686645508,
+      "learning_rate": 0.00018546307652305205,
+      "loss": 0.6233,
+      "step": 414
+    },
+    {
+      "epoch": 0.19983146743710123,
+      "grad_norm": 0.9043152928352356,
+      "learning_rate": 0.00018538197810697842,
+      "loss": 0.5477,
+      "step": 415
+    },
+    {
+      "epoch": 0.20031298904538342,
+      "grad_norm": 0.6947689056396484,
+      "learning_rate": 0.0001853006719380802,
+      "loss": 0.7851,
+      "step": 416
+    },
+    {
+      "epoch": 0.20079451065366558,
+      "grad_norm": 1.0613230466842651,
+      "learning_rate": 0.00018521915821419284,
+      "loss": 0.6935,
+      "step": 417
+    },
+    {
+      "epoch": 0.20127603226194776,
+      "grad_norm": 0.9804971218109131,
+      "learning_rate": 0.00018513743713365698,
+      "loss": 0.7748,
+      "step": 418
+    },
+    {
+      "epoch": 0.20175755387022992,
+      "grad_norm": 0.9837954640388489,
+      "learning_rate": 0.00018505550889531765,
+      "loss": 0.4582,
+      "step": 419
+    },
+    {
+      "epoch": 0.2022390754785121,
+      "grad_norm": 0.8446076512336731,
+      "learning_rate": 0.00018497337369852395,
+      "loss": 0.1956,
+      "step": 420
+    },
+    {
+      "epoch": 0.20272059708679427,
+      "grad_norm": 0.9201738238334656,
+      "learning_rate": 0.0001848910317431286,
+      "loss": 0.6636,
+      "step": 421
+    },
+    {
+      "epoch": 0.20320211869507643,
+      "grad_norm": 1.523264765739441,
+      "learning_rate": 0.00018480848322948739,
+      "loss": 0.7119,
+      "step": 422
+    },
+    {
+      "epoch": 0.20368364030335862,
+      "grad_norm": 1.1211214065551758,
+      "learning_rate": 0.00018472572835845873,
+      "loss": 0.8146,
+      "step": 423
+    },
+    {
+      "epoch": 0.20416516191164077,
+      "grad_norm": 0.6903687715530396,
+      "learning_rate": 0.00018464276733140306,
+      "loss": 0.6589,
+      "step": 424
+    },
+    {
+      "epoch": 0.20464668351992296,
+      "grad_norm": 1.2048624753952026,
+      "learning_rate": 0.0001845596003501826,
+      "loss": 0.7029,
+      "step": 425
+    },
+    {
+      "epoch": 0.20512820512820512,
+      "grad_norm": 0.6592060327529907,
+      "learning_rate": 0.00018447622761716057,
+      "loss": 0.492,
+      "step": 426
+    },
+    {
+      "epoch": 0.2056097267364873,
+      "grad_norm": 0.8045766353607178,
+      "learning_rate": 0.00018439264933520084,
+      "loss": 0.9407,
+      "step": 427
+    },
+    {
+      "epoch": 0.20609124834476947,
+      "grad_norm": 0.957320511341095,
+      "learning_rate": 0.00018430886570766747,
+      "loss": 0.9598,
+      "step": 428
+    },
+    {
+      "epoch": 0.20657276995305165,
+      "grad_norm": 0.9459810256958008,
+      "learning_rate": 0.0001842248769384242,
+      "loss": 0.8971,
+      "step": 429
+    },
+    {
+      "epoch": 0.20705429156133381,
+      "grad_norm": 1.4347730875015259,
+      "learning_rate": 0.00018414068323183375,
+      "loss": 0.737,
+      "step": 430
+    },
+    {
+      "epoch": 0.20753581316961597,
+      "grad_norm": 1.069244623184204,
+      "learning_rate": 0.00018405628479275775,
+      "loss": 0.5817,
+      "step": 431
+    },
+    {
+      "epoch": 0.20801733477789816,
+      "grad_norm": 0.9500924348831177,
+      "learning_rate": 0.00018397168182655583,
+      "loss": 0.4447,
+      "step": 432
+    },
+    {
+      "epoch": 0.20849885638618032,
+      "grad_norm": 1.1861311197280884,
+      "learning_rate": 0.00018388687453908527,
+      "loss": 0.4656,
+      "step": 433
+    },
+    {
+      "epoch": 0.2089803779944625,
+      "grad_norm": 1.3185081481933594,
+      "learning_rate": 0.00018380186313670058,
+      "loss": 0.9314,
+      "step": 434
+    },
+    {
+      "epoch": 0.20946189960274467,
+      "grad_norm": 1.0046930313110352,
+      "learning_rate": 0.00018371664782625287,
+      "loss": 0.3958,
+      "step": 435
+    },
+    {
+      "epoch": 0.20994342121102685,
+      "grad_norm": 1.3395354747772217,
+      "learning_rate": 0.00018363122881508945,
+      "loss": 0.8353,
+      "step": 436
+    },
+    {
+      "epoch": 0.210424942819309,
+      "grad_norm": 1.279977798461914,
+      "learning_rate": 0.00018354560631105328,
+      "loss": 0.2531,
+      "step": 437
+    },
+    {
+      "epoch": 0.2109064644275912,
+      "grad_norm": 1.3670711517333984,
+      "learning_rate": 0.00018345978052248233,
+      "loss": 0.6959,
+      "step": 438
+    },
+    {
+      "epoch": 0.21138798603587336,
+      "grad_norm": 1.0435469150543213,
+      "learning_rate": 0.00018337375165820944,
+      "loss": 1.0629,
+      "step": 439
+    },
+    {
+      "epoch": 0.21186950764415552,
+      "grad_norm": 1.0792940855026245,
+      "learning_rate": 0.00018328751992756137,
+      "loss": 0.7146,
+      "step": 440
+    },
+    {
+      "epoch": 0.2123510292524377,
+      "grad_norm": 1.0037693977355957,
+      "learning_rate": 0.0001832010855403586,
+      "loss": 0.5551,
+      "step": 441
+    },
+    {
+      "epoch": 0.21283255086071987,
+      "grad_norm": 0.7013088464736938,
+      "learning_rate": 0.0001831144487069147,
+      "loss": 0.7306,
+      "step": 442
+    },
+    {
+      "epoch": 0.21331407246900205,
+      "grad_norm": 1.636138916015625,
+      "learning_rate": 0.0001830276096380358,
+      "loss": 0.3383,
+      "step": 443
+    },
+    {
+      "epoch": 0.2137955940772842,
+      "grad_norm": 1.0589203834533691,
+      "learning_rate": 0.0001829405685450202,
+      "loss": 0.965,
+      "step": 444
+    },
+    {
+      "epoch": 0.2142771156855664,
+      "grad_norm": 1.0804965496063232,
+      "learning_rate": 0.00018285332563965765,
+      "loss": 1.0993,
+      "step": 445
+    },
+    {
+      "epoch": 0.21475863729384856,
+      "grad_norm": 0.742942214012146,
+      "learning_rate": 0.00018276588113422905,
+      "loss": 0.5167,
+      "step": 446
+    },
+    {
+      "epoch": 0.21524015890213075,
+      "grad_norm": 0.7951419353485107,
+      "learning_rate": 0.00018267823524150575,
+      "loss": 0.9815,
+      "step": 447
+    },
+    {
+      "epoch": 0.2157216805104129,
+      "grad_norm": 0.8592374920845032,
+      "learning_rate": 0.00018259038817474923,
+      "loss": 0.9137,
+      "step": 448
+    },
+    {
+      "epoch": 0.21620320211869506,
+      "grad_norm": 0.5096921920776367,
+      "learning_rate": 0.0001825023401477104,
+      "loss": 0.4994,
+      "step": 449
+    },
+    {
+      "epoch": 0.21668472372697725,
+      "grad_norm": 0.9474820494651794,
+      "learning_rate": 0.0001824140913746291,
+      "loss": 0.7438,
+      "step": 450
+    },
+    {
+      "epoch": 0.2171662453352594,
+      "grad_norm": 0.4747283160686493,
+      "learning_rate": 0.00018232564207023376,
+      "loss": 0.3613,
+      "step": 451
+    },
+    {
+      "epoch": 0.2176477669435416,
+      "grad_norm": 1.2512322664260864,
+      "learning_rate": 0.00018223699244974064,
+      "loss": 0.5614,
+      "step": 452
+    },
+    {
+      "epoch": 0.21812928855182376,
+      "grad_norm": 0.7003946900367737,
+      "learning_rate": 0.00018214814272885343,
+      "loss": 0.7942,
+      "step": 453
+    },
+    {
+      "epoch": 0.21861081016010595,
+      "grad_norm": 0.7767881155014038,
+      "learning_rate": 0.00018205909312376276,
+      "loss": 0.8337,
+      "step": 454
+    },
+    {
+      "epoch": 0.2190923317683881,
+      "grad_norm": 1.2129909992218018,
+      "learning_rate": 0.00018196984385114554,
+      "loss": 0.6863,
+      "step": 455
+    },
+    {
+      "epoch": 0.2195738533766703,
+      "grad_norm": 0.5771981477737427,
+      "learning_rate": 0.0001818803951281646,
+      "loss": 0.5401,
+      "step": 456
+    },
+    {
+      "epoch": 0.22005537498495245,
+      "grad_norm": 1.2848843336105347,
+      "learning_rate": 0.000181790747172468,
+      "loss": 0.7414,
+      "step": 457
+    },
+    {
+      "epoch": 0.2205368965932346,
+      "grad_norm": 0.7299485206604004,
+      "learning_rate": 0.00018170090020218864,
+      "loss": 0.5679,
+      "step": 458
+    },
+    {
+      "epoch": 0.2210184182015168,
+      "grad_norm": 0.48027756810188293,
+      "learning_rate": 0.00018161085443594365,
+      "loss": 0.2938,
+      "step": 459
+    },
+    {
+      "epoch": 0.22149993980979896,
+      "grad_norm": 0.7984429001808167,
+      "learning_rate": 0.00018152061009283382,
+      "loss": 0.3071,
+      "step": 460
+    },
+    {
+      "epoch": 0.22198146141808114,
+      "grad_norm": 1.3941313028335571,
+      "learning_rate": 0.00018143016739244314,
+      "loss": 0.9465,
+      "step": 461
+    },
+    {
+      "epoch": 0.2224629830263633,
+      "grad_norm": 1.314907193183899,
+      "learning_rate": 0.0001813395265548383,
+      "loss": 0.7413,
+      "step": 462
+    },
+    {
+      "epoch": 0.2229445046346455,
+      "grad_norm": 0.863312304019928,
+      "learning_rate": 0.00018124868780056814,
+      "loss": 0.7547,
+      "step": 463
+    },
+    {
+      "epoch": 0.22342602624292765,
+      "grad_norm": 1.0457203388214111,
+      "learning_rate": 0.0001811576513506629,
+      "loss": 0.6511,
+      "step": 464
+    },
+    {
+      "epoch": 0.2239075478512098,
+      "grad_norm": 1.8019721508026123,
+      "learning_rate": 0.00018106641742663397,
+      "loss": 0.8317,
+      "step": 465
+    },
+    {
+      "epoch": 0.224389069459492,
+      "grad_norm": 0.7878347039222717,
+      "learning_rate": 0.00018097498625047328,
+      "loss": 0.3196,
+      "step": 466
+    },
+    {
+      "epoch": 0.22487059106777416,
+      "grad_norm": 1.159590721130371,
+      "learning_rate": 0.00018088335804465258,
+      "loss": 0.6533,
+      "step": 467
+    },
+    {
+      "epoch": 0.22535211267605634,
+      "grad_norm": 0.7787841558456421,
+      "learning_rate": 0.00018079153303212318,
+      "loss": 0.4519,
+      "step": 468
+    },
+    {
+      "epoch": 0.2258336342843385,
+      "grad_norm": 1.8181616067886353,
+      "learning_rate": 0.0001806995114363152,
+      "loss": 0.8562,
+      "step": 469
+    },
+    {
+      "epoch": 0.2263151558926207,
+      "grad_norm": 0.6238754987716675,
+      "learning_rate": 0.00018060729348113707,
+      "loss": 0.4416,
+      "step": 470
+    },
+    {
+      "epoch": 0.22679667750090285,
+      "grad_norm": 0.5132167339324951,
+      "learning_rate": 0.00018051487939097505,
+      "loss": 0.4001,
+      "step": 471
+    },
+    {
+      "epoch": 0.22727819910918504,
+      "grad_norm": 0.9573964476585388,
+      "learning_rate": 0.00018042226939069255,
+      "loss": 0.7387,
+      "step": 472
+    },
+    {
+      "epoch": 0.2277597207174672,
+      "grad_norm": 0.6595173478126526,
+      "learning_rate": 0.00018032946370562982,
+      "loss": 0.6063,
+      "step": 473
+    },
+    {
+      "epoch": 0.22824124232574936,
+      "grad_norm": 0.8158476948738098,
+      "learning_rate": 0.00018023646256160313,
+      "loss": 0.5302,
+      "step": 474
+    },
+    {
+      "epoch": 0.22872276393403154,
+      "grad_norm": 0.6084806323051453,
+      "learning_rate": 0.00018014326618490437,
+      "loss": 0.394,
+      "step": 475
+    },
+    {
+      "epoch": 0.2292042855423137,
+      "grad_norm": 0.9932984113693237,
+      "learning_rate": 0.0001800498748023005,
+      "loss": 0.8495,
+      "step": 476
+    },
+    {
+      "epoch": 0.2296858071505959,
+      "grad_norm": 0.49222004413604736,
+      "learning_rate": 0.000179956288641033,
+      "loss": 0.4108,
+      "step": 477
+    },
+    {
+      "epoch": 0.23016732875887805,
+      "grad_norm": 0.6978711485862732,
+      "learning_rate": 0.00017986250792881718,
+      "loss": 0.3108,
+      "step": 478
+    },
+    {
+      "epoch": 0.23064885036716024,
+      "grad_norm": 1.0181289911270142,
+      "learning_rate": 0.00017976853289384184,
+      "loss": 1.0249,
+      "step": 479
+    },
+    {
+      "epoch": 0.2311303719754424,
+      "grad_norm": 1.5912185907363892,
+      "learning_rate": 0.00017967436376476855,
+      "loss": 1.1765,
+      "step": 480
+    },
+    {
+      "epoch": 0.23161189358372458,
+      "grad_norm": 1.0102556943893433,
+      "learning_rate": 0.0001795800007707312,
+      "loss": 0.4972,
+      "step": 481
+    },
+    {
+      "epoch": 0.23209341519200674,
+      "grad_norm": 0.4210749566555023,
+      "learning_rate": 0.00017948544414133534,
+      "loss": 0.289,
+      "step": 482
+    },
+    {
+      "epoch": 0.2325749368002889,
+      "grad_norm": 0.4447524845600128,
+      "learning_rate": 0.00017939069410665773,
+      "loss": 0.2973,
+      "step": 483
+    },
+    {
+      "epoch": 0.2330564584085711,
+      "grad_norm": 1.2053577899932861,
+      "learning_rate": 0.0001792957508972457,
+      "loss": 0.52,
+      "step": 484
+    },
+    {
+      "epoch": 0.23353798001685325,
+      "grad_norm": 1.1216098070144653,
+      "learning_rate": 0.00017920061474411658,
+      "loss": 0.4747,
+      "step": 485
+    },
+    {
+      "epoch": 0.23401950162513543,
+      "grad_norm": 2.1695163249969482,
+      "learning_rate": 0.00017910528587875729,
+      "loss": 0.7822,
+      "step": 486
+    },
+    {
+      "epoch": 0.2345010232334176,
+      "grad_norm": 0.5691484212875366,
+      "learning_rate": 0.00017900976453312352,
+      "loss": 0.3004,
+      "step": 487
+    },
+    {
+      "epoch": 0.23498254484169978,
+      "grad_norm": 1.2666643857955933,
+      "learning_rate": 0.00017891405093963938,
+      "loss": 0.5522,
+      "step": 488
+    },
+    {
+      "epoch": 0.23546406644998194,
+      "grad_norm": 1.0269070863723755,
+      "learning_rate": 0.00017881814533119675,
+      "loss": 0.9501,
+      "step": 489
+    },
+    {
+      "epoch": 0.23594558805826413,
+      "grad_norm": 1.3407925367355347,
+      "learning_rate": 0.00017872204794115474,
+      "loss": 0.498,
+      "step": 490
+    },
+    {
+      "epoch": 0.2364271096665463,
+      "grad_norm": 0.6913367509841919,
+      "learning_rate": 0.0001786257590033391,
+      "loss": 0.46,
+      "step": 491
+    },
+    {
+      "epoch": 0.23690863127482845,
+      "grad_norm": 1.1869349479675293,
+      "learning_rate": 0.00017852927875204163,
+      "loss": 0.7829,
+      "step": 492
+    },
+    {
+      "epoch": 0.23739015288311063,
+      "grad_norm": 0.6723718047142029,
+      "learning_rate": 0.00017843260742201963,
+      "loss": 0.3372,
+      "step": 493
+    },
+    {
+      "epoch": 0.2378716744913928,
+      "grad_norm": 0.679757297039032,
+      "learning_rate": 0.00017833574524849535,
+      "loss": 0.9048,
+      "step": 494
+    },
+    {
+      "epoch": 0.23835319609967498,
+      "grad_norm": 1.0547089576721191,
+      "learning_rate": 0.00017823869246715553,
+      "loss": 0.6047,
+      "step": 495
+    },
+    {
+      "epoch": 0.23883471770795714,
+      "grad_norm": 1.2114295959472656,
+      "learning_rate": 0.00017814144931415043,
+      "loss": 0.9538,
+      "step": 496
+    },
+    {
+      "epoch": 0.23931623931623933,
+      "grad_norm": 1.396855354309082,
+      "learning_rate": 0.0001780440160260938,
+      "loss": 1.0792,
+      "step": 497
+    },
+    {
+      "epoch": 0.23979776092452149,
+      "grad_norm": 0.9748196601867676,
+      "learning_rate": 0.00017794639284006184,
+      "loss": 1.0788,
+      "step": 498
+    },
+    {
+      "epoch": 0.24027928253280367,
+      "grad_norm": 1.2064623832702637,
+      "learning_rate": 0.0001778485799935929,
+      "loss": 0.7759,
+      "step": 499
+    },
+    {
+      "epoch": 0.24076080414108583,
+      "grad_norm": 1.0788531303405762,
+      "learning_rate": 0.00017775057772468679,
+      "loss": 0.9132,
+      "step": 500
+    },
+    {
+      "epoch": 0.241242325749368,
+      "grad_norm": 0.8036178350448608,
+      "learning_rate": 0.00017765238627180424,
+      "loss": 0.6564,
+      "step": 501
+    },
+    {
+      "epoch": 0.24172384735765018,
+      "grad_norm": 0.7760926485061646,
+      "learning_rate": 0.00017755400587386632,
+      "loss": 0.638,
+      "step": 502
+    },
+    {
+      "epoch": 0.24220536896593234,
+      "grad_norm": 0.6797692775726318,
+      "learning_rate": 0.00017745543677025378,
+      "loss": 0.4243,
+      "step": 503
+    },
+    {
+      "epoch": 0.24268689057421453,
+      "grad_norm": 0.9898531436920166,
+      "learning_rate": 0.00017735667920080661,
+      "loss": 0.6076,
+      "step": 504
+    },
+    {
+      "epoch": 0.24316841218249668,
+      "grad_norm": 0.7974802255630493,
+      "learning_rate": 0.0001772577334058233,
+      "loss": 0.6662,
+      "step": 505
+    },
+    {
+      "epoch": 0.24364993379077887,
+      "grad_norm": 1.0043370723724365,
+      "learning_rate": 0.00017715859962606043,
+      "loss": 0.865,
+      "step": 506
+    },
+    {
+      "epoch": 0.24413145539906103,
+      "grad_norm": 0.6671229600906372,
+      "learning_rate": 0.00017705927810273187,
+      "loss": 0.2871,
+      "step": 507
+    },
+    {
+      "epoch": 0.2446129770073432,
+      "grad_norm": 1.1656029224395752,
+      "learning_rate": 0.00017695976907750844,
+      "loss": 0.5763,
+      "step": 508
+    },
+    {
+      "epoch": 0.24509449861562538,
+      "grad_norm": 1.1035945415496826,
+      "learning_rate": 0.00017686007279251706,
+      "loss": 0.3747,
+      "step": 509
+    },
+    {
+      "epoch": 0.24557602022390754,
+      "grad_norm": 0.8782248497009277,
+      "learning_rate": 0.00017676018949034045,
+      "loss": 0.7433,
+      "step": 510
+    },
+    {
+      "epoch": 0.24605754183218972,
+      "grad_norm": 0.902968168258667,
+      "learning_rate": 0.0001766601194140162,
+      "loss": 0.6472,
+      "step": 511
+    },
+    {
+      "epoch": 0.24653906344047188,
+      "grad_norm": 1.9581364393234253,
+      "learning_rate": 0.0001765598628070365,
+      "loss": 0.7606,
+      "step": 512
+    },
+    {
+      "epoch": 0.24702058504875407,
+      "grad_norm": 0.7049884796142578,
+      "learning_rate": 0.00017645941991334732,
+      "loss": 0.6779,
+      "step": 513
+    },
+    {
+      "epoch": 0.24750210665703623,
+      "grad_norm": 0.9359114170074463,
+      "learning_rate": 0.00017635879097734804,
+      "loss": 0.5744,
+      "step": 514
+    },
+    {
+      "epoch": 0.24798362826531842,
+      "grad_norm": 1.0945175886154175,
+      "learning_rate": 0.00017625797624389055,
+      "loss": 1.2865,
+      "step": 515
+    },
+    {
+      "epoch": 0.24846514987360058,
+      "grad_norm": 0.6059966087341309,
+      "learning_rate": 0.00017615697595827897,
+      "loss": 0.5536,
+      "step": 516
+    },
+    {
+      "epoch": 0.24894667148188274,
+      "grad_norm": 0.6404644846916199,
+      "learning_rate": 0.0001760557903662688,
+      "loss": 0.454,
+      "step": 517
+    },
+    {
+      "epoch": 0.24942819309016492,
+      "grad_norm": 1.5128916501998901,
+      "learning_rate": 0.00017595441971406648,
+      "loss": 1.4014,
+      "step": 518
+    },
+    {
+      "epoch": 0.24990971469844708,
+      "grad_norm": 0.662467896938324,
+      "learning_rate": 0.00017585286424832874,
+      "loss": 0.4347,
+      "step": 519
+    },
+    {
+      "epoch": 0.25039123630672927,
+      "grad_norm": 0.6688117980957031,
+      "learning_rate": 0.00017575112421616202,
+      "loss": 0.7102,
+      "step": 520
+    },
+    {
+      "epoch": 0.25087275791501146,
+      "grad_norm": 0.5434272885322571,
+      "learning_rate": 0.0001756491998651218,
+      "loss": 0.4888,
+      "step": 521
+    },
+    {
+      "epoch": 0.2513542795232936,
+      "grad_norm": 0.6772214770317078,
+      "learning_rate": 0.0001755470914432121,
+      "loss": 0.7237,
+      "step": 522
+    },
+    {
+      "epoch": 0.2518358011315758,
+      "grad_norm": 0.49257853627204895,
+      "learning_rate": 0.0001754447991988848,
+      "loss": 0.3216,
+      "step": 523
+    },
+    {
+      "epoch": 0.25231732273985796,
+      "grad_norm": 0.9488145709037781,
+      "learning_rate": 0.00017534232338103903,
+      "loss": 1.2589,
+      "step": 524
+    },
+    {
+      "epoch": 0.25279884434814015,
+      "grad_norm": 0.6488147377967834,
+      "learning_rate": 0.0001752396642390207,
+      "loss": 0.845,
+      "step": 525
+    },
+    {
+      "epoch": 0.2532803659564223,
+      "grad_norm": 0.9418458342552185,
+      "learning_rate": 0.00017513682202262163,
+      "loss": 0.9113,
+      "step": 526
+    },
+    {
+      "epoch": 0.25376188756470447,
+      "grad_norm": 0.8072599768638611,
+      "learning_rate": 0.00017503379698207918,
+      "loss": 0.4421,
+      "step": 527
+    },
+    {
+      "epoch": 0.25424340917298666,
+      "grad_norm": 2.4496564865112305,
+      "learning_rate": 0.00017493058936807562,
+      "loss": 0.7209,
+      "step": 528
+    },
+    {
+      "epoch": 0.2547249307812688,
+      "grad_norm": 0.8409088850021362,
+      "learning_rate": 0.00017482719943173739,
+      "loss": 0.5874,
+      "step": 529
+    },
+    {
+      "epoch": 0.255206452389551,
+      "grad_norm": 0.7664375901222229,
+      "learning_rate": 0.00017472362742463455,
+      "loss": 0.5683,
+      "step": 530
+    },
+    {
+      "epoch": 0.25568797399783316,
+      "grad_norm": 1.121518850326538,
+      "learning_rate": 0.0001746198735987802,
+      "loss": 0.957,
+      "step": 531
+    },
+    {
+      "epoch": 0.25616949560611535,
+      "grad_norm": 0.6444967985153198,
+      "learning_rate": 0.00017451593820662988,
+      "loss": 0.2528,
+      "step": 532
+    },
+    {
+      "epoch": 0.2566510172143975,
+      "grad_norm": 1.1670219898223877,
+      "learning_rate": 0.00017441182150108086,
+      "loss": 0.8635,
+      "step": 533
+    },
+    {
+      "epoch": 0.25713253882267967,
+      "grad_norm": 0.790825605392456,
+      "learning_rate": 0.0001743075237354716,
+      "loss": 0.4617,
+      "step": 534
+    },
+    {
+      "epoch": 0.25761406043096186,
+      "grad_norm": 0.7336540222167969,
+      "learning_rate": 0.00017420304516358113,
+      "loss": 0.6611,
+      "step": 535
+    },
+    {
+      "epoch": 0.258095582039244,
+      "grad_norm": 1.0129753351211548,
+      "learning_rate": 0.00017409838603962843,
+      "loss": 0.868,
+      "step": 536
+    },
+    {
+      "epoch": 0.2585771036475262,
+      "grad_norm": 0.9752351641654968,
+      "learning_rate": 0.00017399354661827178,
+      "loss": 0.6151,
+      "step": 537
+    },
+    {
+      "epoch": 0.25905862525580836,
+      "grad_norm": 2.983139753341675,
+      "learning_rate": 0.00017388852715460819,
+      "loss": 0.5691,
+      "step": 538
+    },
+    {
+      "epoch": 0.25954014686409055,
+      "grad_norm": 0.6391310095787048,
+      "learning_rate": 0.00017378332790417273,
+      "loss": 0.4331,
+      "step": 539
+    },
+    {
+      "epoch": 0.2600216684723727,
+      "grad_norm": 0.9141309261322021,
+      "learning_rate": 0.00017367794912293794,
+      "loss": 0.8207,
+      "step": 540
+    },
+    {
+      "epoch": 0.26050319008065487,
+      "grad_norm": 0.5318431854248047,
+      "learning_rate": 0.00017357239106731317,
+      "loss": 0.3676,
+      "step": 541
+    },
+    {
+      "epoch": 0.26098471168893705,
+      "grad_norm": 1.2545955181121826,
+      "learning_rate": 0.00017346665399414405,
+      "loss": 0.9302,
+      "step": 542
+    },
+    {
+      "epoch": 0.2614662332972192,
+      "grad_norm": 0.7107487320899963,
+      "learning_rate": 0.00017336073816071168,
+      "loss": 0.3593,
+      "step": 543
+    },
+    {
+      "epoch": 0.2619477549055014,
+      "grad_norm": 3.153602123260498,
+      "learning_rate": 0.00017325464382473226,
+      "loss": 0.6631,
+      "step": 544
+    },
+    {
+      "epoch": 0.26242927651378356,
+      "grad_norm": 0.750834584236145,
+      "learning_rate": 0.00017314837124435622,
+      "loss": 0.7142,
+      "step": 545
+    },
+    {
+      "epoch": 0.26291079812206575,
+      "grad_norm": 0.7795765995979309,
+      "learning_rate": 0.00017304192067816782,
+      "loss": 0.3292,
+      "step": 546
+    },
+    {
+      "epoch": 0.2633923197303479,
+      "grad_norm": 0.5292965173721313,
+      "learning_rate": 0.00017293529238518422,
+      "loss": 0.2838,
+      "step": 547
+    },
+    {
+      "epoch": 0.26387384133863007,
+      "grad_norm": 0.7363600134849548,
+      "learning_rate": 0.0001728284866248552,
+      "loss": 0.7192,
+      "step": 548
+    },
+    {
+      "epoch": 0.26435536294691225,
+      "grad_norm": 1.5265337228775024,
+      "learning_rate": 0.00017272150365706224,
+      "loss": 0.4173,
+      "step": 549
+    },
+    {
+      "epoch": 0.26483688455519444,
+      "grad_norm": 0.7974181771278381,
+      "learning_rate": 0.00017261434374211802,
+      "loss": 0.5197,
+      "step": 550
+    },
+    {
+      "epoch": 0.2653184061634766,
+      "grad_norm": 0.7714865803718567,
+      "learning_rate": 0.00017250700714076586,
+      "loss": 0.5794,
+      "step": 551
+    },
+    {
+      "epoch": 0.26579992777175876,
+      "grad_norm": 0.587791919708252,
+      "learning_rate": 0.00017239949411417888,
+      "loss": 0.6678,
+      "step": 552
+    },
+    {
+      "epoch": 0.26628144938004095,
+      "grad_norm": 1.730128288269043,
+      "learning_rate": 0.0001722918049239596,
+      "loss": 1.2,
+      "step": 553
+    },
+    {
+      "epoch": 0.2667629709883231,
+      "grad_norm": 0.7081428170204163,
+      "learning_rate": 0.00017218393983213902,
+      "loss": 0.3536,
+      "step": 554
+    },
+    {
+      "epoch": 0.26724449259660527,
+      "grad_norm": 0.8215841054916382,
+      "learning_rate": 0.00017207589910117634,
+      "loss": 0.4589,
+      "step": 555
+    },
+    {
+      "epoch": 0.26772601420488745,
+      "grad_norm": 1.0128780603408813,
+      "learning_rate": 0.00017196768299395797,
+      "loss": 0.7296,
+      "step": 556
+    },
+    {
+      "epoch": 0.26820753581316964,
+      "grad_norm": 1.796431303024292,
+      "learning_rate": 0.00017185929177379714,
+      "loss": 0.7464,
+      "step": 557
+    },
+    {
+      "epoch": 0.26868905742145177,
+      "grad_norm": 1.8734339475631714,
+      "learning_rate": 0.00017175072570443312,
+      "loss": 1.0069,
+      "step": 558
+    },
+    {
+      "epoch": 0.26917057902973396,
+      "grad_norm": 0.860616147518158,
+      "learning_rate": 0.00017164198505003066,
+      "loss": 0.5769,
+      "step": 559
+    },
+    {
+      "epoch": 0.26965210063801615,
+      "grad_norm": 0.49917736649513245,
+      "learning_rate": 0.0001715330700751793,
+      "loss": 0.2335,
+      "step": 560
+    },
+    {
+      "epoch": 0.2701336222462983,
+      "grad_norm": 0.6714743375778198,
+      "learning_rate": 0.00017142398104489273,
+      "loss": 0.4088,
+      "step": 561
+    },
+    {
+      "epoch": 0.27061514385458046,
+      "grad_norm": 0.6164880990982056,
+      "learning_rate": 0.00017131471822460814,
+      "loss": 0.7426,
+      "step": 562
+    },
+    {
+      "epoch": 0.27109666546286265,
+      "grad_norm": 0.5979893803596497,
+      "learning_rate": 0.00017120528188018565,
+      "loss": 0.2952,
+      "step": 563
+    },
+    {
+      "epoch": 0.27157818707114484,
+      "grad_norm": 0.6977973580360413,
+      "learning_rate": 0.00017109567227790754,
+      "loss": 0.4368,
+      "step": 564
+    },
+    {
+      "epoch": 0.27205970867942697,
+      "grad_norm": 0.7051186561584473,
+      "learning_rate": 0.00017098588968447766,
+      "loss": 0.8569,
+      "step": 565
+    },
+    {
+      "epoch": 0.27254123028770916,
+      "grad_norm": 0.5577298998832703,
+      "learning_rate": 0.00017087593436702084,
+      "loss": 0.4247,
+      "step": 566
+    },
+    {
+      "epoch": 0.27302275189599134,
+      "grad_norm": 0.9759942293167114,
+      "learning_rate": 0.00017076580659308222,
+      "loss": 1.0734,
+      "step": 567
+    },
+    {
+      "epoch": 0.27350427350427353,
+      "grad_norm": 0.5310359001159668,
+      "learning_rate": 0.00017065550663062634,
+      "loss": 0.3319,
+      "step": 568
+    },
+    {
+      "epoch": 0.27398579511255566,
+      "grad_norm": 0.8164540529251099,
+      "learning_rate": 0.00017054503474803702,
+      "loss": 0.4834,
+      "step": 569
+    },
+    {
+      "epoch": 0.27446731672083785,
+      "grad_norm": 0.9512844085693359,
+      "learning_rate": 0.00017043439121411618,
+      "loss": 0.4703,
+      "step": 570
+    },
+    {
+      "epoch": 0.27494883832912004,
+      "grad_norm": 1.0269114971160889,
+      "learning_rate": 0.0001703235762980835,
+      "loss": 0.3504,
+      "step": 571
+    },
+    {
+      "epoch": 0.27543035993740217,
+      "grad_norm": 0.6352307200431824,
+      "learning_rate": 0.00017021259026957567,
+      "loss": 0.7479,
+      "step": 572
+    },
+    {
+      "epoch": 0.27591188154568436,
+      "grad_norm": 1.2709698677062988,
+      "learning_rate": 0.00017010143339864562,
+      "loss": 0.8125,
+      "step": 573
+    },
+    {
+      "epoch": 0.27639340315396654,
+      "grad_norm": 1.1885913610458374,
+      "learning_rate": 0.0001699901059557621,
+      "loss": 0.7269,
+      "step": 574
+    },
+    {
+      "epoch": 0.27687492476224873,
+      "grad_norm": 1.061423420906067,
+      "learning_rate": 0.00016987860821180895,
+      "loss": 0.8678,
+      "step": 575
+    },
+    {
+      "epoch": 0.27735644637053086,
+      "grad_norm": 1.2436217069625854,
+      "learning_rate": 0.00016976694043808416,
+      "loss": 1.0843,
+      "step": 576
+    },
+    {
+      "epoch": 0.27783796797881305,
+      "grad_norm": 1.5231688022613525,
+      "learning_rate": 0.00016965510290629972,
+      "loss": 0.6489,
+      "step": 577
+    },
+    {
+      "epoch": 0.27831948958709524,
+      "grad_norm": 0.8685100674629211,
+      "learning_rate": 0.00016954309588858044,
+      "loss": 0.5554,
+      "step": 578
+    },
+    {
+      "epoch": 0.27880101119537737,
+      "grad_norm": 0.8666234612464905,
+      "learning_rate": 0.00016943091965746366,
+      "loss": 0.6095,
+      "step": 579
+    },
+    {
+      "epoch": 0.27928253280365956,
+      "grad_norm": 1.7214361429214478,
+      "learning_rate": 0.00016931857448589845,
+      "loss": 1.1025,
+      "step": 580
+    },
+    {
+      "epoch": 0.27976405441194174,
+      "grad_norm": 0.4393554627895355,
+      "learning_rate": 0.00016920606064724488,
+      "loss": 0.4294,
+      "step": 581
+    },
+    {
+      "epoch": 0.28024557602022393,
+      "grad_norm": 0.4375576972961426,
+      "learning_rate": 0.00016909337841527344,
+      "loss": 0.5308,
+      "step": 582
+    },
+    {
+      "epoch": 0.28072709762850606,
+      "grad_norm": 0.9495024085044861,
+      "learning_rate": 0.00016898052806416444,
+      "loss": 0.6907,
+      "step": 583
+    },
+    {
+      "epoch": 0.28120861923678825,
+      "grad_norm": 0.7171463370323181,
+      "learning_rate": 0.00016886750986850718,
+      "loss": 0.3165,
+      "step": 584
+    },
+    {
+      "epoch": 0.28169014084507044,
+      "grad_norm": 0.6095327138900757,
+      "learning_rate": 0.00016875432410329934,
+      "loss": 0.7329,
+      "step": 585
+    },
+    {
+      "epoch": 0.28217166245335257,
+      "grad_norm": 0.9908806085586548,
+      "learning_rate": 0.0001686409710439464,
+      "loss": 0.5553,
+      "step": 586
+    },
+    {
+      "epoch": 0.28265318406163475,
+      "grad_norm": 0.9707655906677246,
+      "learning_rate": 0.00016852745096626088,
+      "loss": 0.8345,
+      "step": 587
+    },
+    {
+      "epoch": 0.28313470566991694,
+      "grad_norm": 0.6714180707931519,
+      "learning_rate": 0.0001684137641464617,
+      "loss": 0.6241,
+      "step": 588
+    },
+    {
+      "epoch": 0.28361622727819913,
+      "grad_norm": 1.1676234006881714,
+      "learning_rate": 0.0001682999108611735,
+      "loss": 0.3852,
+      "step": 589
+    },
+    {
+      "epoch": 0.28409774888648126,
+      "grad_norm": 0.5526348948478699,
+      "learning_rate": 0.00016818589138742587,
+      "loss": 0.5523,
+      "step": 590
+    },
+    {
+      "epoch": 0.28457927049476345,
+      "grad_norm": 0.7403177618980408,
+      "learning_rate": 0.00016807170600265296,
+      "loss": 0.5595,
+      "step": 591
+    },
+    {
+      "epoch": 0.28506079210304563,
+      "grad_norm": 1.745147705078125,
+      "learning_rate": 0.00016795735498469246,
+      "loss": 0.3464,
+      "step": 592
+    },
+    {
+      "epoch": 0.2855423137113278,
+      "grad_norm": 1.051491141319275,
+      "learning_rate": 0.00016784283861178513,
+      "loss": 0.6082,
+      "step": 593
+    },
+    {
+      "epoch": 0.28602383531960995,
+      "grad_norm": 1.1357232332229614,
+      "learning_rate": 0.00016772815716257412,
+      "loss": 0.4084,
+      "step": 594
+    },
+    {
+      "epoch": 0.28650535692789214,
+      "grad_norm": 1.4268995523452759,
+      "learning_rate": 0.00016761331091610416,
+      "loss": 0.8438,
+      "step": 595
+    },
+    {
+      "epoch": 0.28698687853617433,
+      "grad_norm": 0.7449113130569458,
+      "learning_rate": 0.00016749830015182107,
+      "loss": 0.5502,
+      "step": 596
+    },
+    {
+      "epoch": 0.28746840014445646,
+      "grad_norm": 0.5473132729530334,
+      "learning_rate": 0.00016738312514957086,
+      "loss": 0.2627,
+      "step": 597
+    },
+    {
+      "epoch": 0.28794992175273865,
+      "grad_norm": 0.5442512035369873,
+      "learning_rate": 0.00016726778618959926,
+      "loss": 0.4934,
+      "step": 598
+    },
+    {
+      "epoch": 0.28843144336102083,
+      "grad_norm": 0.7760728001594543,
+      "learning_rate": 0.00016715228355255093,
+      "loss": 0.6557,
+      "step": 599
+    },
+    {
+      "epoch": 0.288912964969303,
+      "grad_norm": 2.4835331439971924,
+      "learning_rate": 0.00016703661751946874,
+      "loss": 1.2639,
+      "step": 600
+    },
+    {
+      "epoch": 0.28939448657758515,
+      "grad_norm": 0.7905292510986328,
+      "learning_rate": 0.00016692078837179318,
+      "loss": 0.778,
+      "step": 601
+    },
+    {
+      "epoch": 0.28987600818586734,
+      "grad_norm": 0.8698615431785583,
+      "learning_rate": 0.00016680479639136163,
+      "loss": 0.7836,
+      "step": 602
+    },
+    {
+      "epoch": 0.2903575297941495,
+      "grad_norm": 0.9689038395881653,
+      "learning_rate": 0.0001666886418604077,
+      "loss": 0.3091,
+      "step": 603
+    },
+    {
+      "epoch": 0.29083905140243166,
+      "grad_norm": 0.6081171035766602,
+      "learning_rate": 0.0001665723250615604,
+      "loss": 0.5046,
+      "step": 604
+    },
+    {
+      "epoch": 0.29132057301071385,
+      "grad_norm": 1.1759047508239746,
+      "learning_rate": 0.00016645584627784381,
+      "loss": 0.8711,
+      "step": 605
+    },
+    {
+      "epoch": 0.29180209461899603,
+      "grad_norm": 0.8777217268943787,
+      "learning_rate": 0.0001663392057926759,
+      "loss": 0.6645,
+      "step": 606
+    },
+    {
+      "epoch": 0.2922836162272782,
+      "grad_norm": 0.5921443700790405,
+      "learning_rate": 0.00016622240388986824,
+      "loss": 0.3308,
+      "step": 607
+    },
+    {
+      "epoch": 0.29276513783556035,
+      "grad_norm": 1.0305765867233276,
+      "learning_rate": 0.0001661054408536251,
+      "loss": 1.1618,
+      "step": 608
+    },
+    {
+      "epoch": 0.29324665944384254,
+      "grad_norm": 0.7973262667655945,
+      "learning_rate": 0.00016598831696854288,
+      "loss": 0.5259,
+      "step": 609
+    },
+    {
+      "epoch": 0.2937281810521247,
+      "grad_norm": 0.7054857015609741,
+      "learning_rate": 0.00016587103251960937,
+      "loss": 0.6195,
+      "step": 610
+    },
+    {
+      "epoch": 0.2942097026604069,
+      "grad_norm": 0.850437581539154,
+      "learning_rate": 0.00016575358779220294,
+      "loss": 0.8113,
+      "step": 611
+    },
+    {
+      "epoch": 0.29469122426868904,
+      "grad_norm": 1.504550576210022,
+      "learning_rate": 0.00016563598307209204,
+      "loss": 1.0547,
+      "step": 612
+    },
+    {
+      "epoch": 0.29517274587697123,
+      "grad_norm": 0.4776376485824585,
+      "learning_rate": 0.0001655182186454344,
+      "loss": 0.2646,
+      "step": 613
+    },
+    {
+      "epoch": 0.2956542674852534,
+      "grad_norm": 0.8611229062080383,
+      "learning_rate": 0.00016540029479877638,
+      "loss": 1.0374,
+      "step": 614
+    },
+    {
+      "epoch": 0.29613578909353555,
+      "grad_norm": 0.8100734353065491,
+      "learning_rate": 0.00016528221181905217,
+      "loss": 0.2605,
+      "step": 615
+    },
+    {
+      "epoch": 0.29661731070181774,
+      "grad_norm": 1.2284023761749268,
+      "learning_rate": 0.00016516396999358322,
+      "loss": 0.8769,
+      "step": 616
+    },
+    {
+      "epoch": 0.2970988323100999,
+      "grad_norm": 0.7563827037811279,
+      "learning_rate": 0.00016504556961007748,
+      "loss": 0.6459,
+      "step": 617
+    },
+    {
+      "epoch": 0.2975803539183821,
+      "grad_norm": 0.7285965085029602,
+      "learning_rate": 0.00016492701095662866,
+      "loss": 0.4741,
+      "step": 618
+    },
+    {
+      "epoch": 0.29806187552666424,
+      "grad_norm": 0.8569052219390869,
+      "learning_rate": 0.00016480829432171564,
+      "loss": 0.6236,
+      "step": 619
+    },
+    {
+      "epoch": 0.29854339713494643,
+      "grad_norm": 0.6136265397071838,
+      "learning_rate": 0.0001646894199942017,
+      "loss": 0.6324,
+      "step": 620
+    },
+    {
+      "epoch": 0.2990249187432286,
+      "grad_norm": 0.8613694310188293,
+      "learning_rate": 0.0001645703882633338,
+      "loss": 0.8755,
+      "step": 621
+    },
+    {
+      "epoch": 0.29950644035151075,
+      "grad_norm": 0.5661609768867493,
+      "learning_rate": 0.00016445119941874183,
+      "loss": 0.3262,
+      "step": 622
+    },
+    {
+      "epoch": 0.29998796195979294,
+      "grad_norm": 0.8413470983505249,
+      "learning_rate": 0.00016433185375043809,
+      "loss": 0.6209,
+      "step": 623
+    },
+    {
+      "epoch": 0.3004694835680751,
+      "grad_norm": 1.3993409872055054,
+      "learning_rate": 0.00016421235154881638,
+      "loss": 0.9886,
+      "step": 624
+    },
+    {
+      "epoch": 0.3009510051763573,
+      "grad_norm": 0.6332157850265503,
+      "learning_rate": 0.00016409269310465146,
+      "loss": 0.3571,
+      "step": 625
+    },
+    {
+      "epoch": 0.30143252678463944,
+      "grad_norm": 1.0084850788116455,
+      "learning_rate": 0.00016397287870909813,
+      "loss": 1.2828,
+      "step": 626
+    },
+    {
+      "epoch": 0.30191404839292163,
+      "grad_norm": 0.9696491360664368,
+      "learning_rate": 0.00016385290865369079,
+      "loss": 1.0937,
+      "step": 627
+    },
+    {
+      "epoch": 0.3023955700012038,
+      "grad_norm": 1.2192230224609375,
+      "learning_rate": 0.00016373278323034255,
+      "loss": 0.4202,
+      "step": 628
+    },
+    {
+      "epoch": 0.30287709160948595,
+      "grad_norm": 0.7007105350494385,
+      "learning_rate": 0.0001636125027313445,
+      "loss": 0.4055,
+      "step": 629
+    },
+    {
+      "epoch": 0.30335861321776814,
+      "grad_norm": 1.2077027559280396,
+      "learning_rate": 0.00016349206744936518,
+      "loss": 0.8379,
+      "step": 630
+    },
+    {
+      "epoch": 0.3038401348260503,
+      "grad_norm": 1.1231609582901,
+      "learning_rate": 0.00016337147767744967,
+      "loss": 1.1825,
+      "step": 631
+    },
+    {
+      "epoch": 0.3043216564343325,
+      "grad_norm": 1.050195336341858,
+      "learning_rate": 0.0001632507337090189,
+      "loss": 0.6219,
+      "step": 632
+    },
+    {
+      "epoch": 0.30480317804261464,
+      "grad_norm": 0.8502358198165894,
+      "learning_rate": 0.0001631298358378692,
+      "loss": 0.7966,
+      "step": 633
+    },
+    {
+      "epoch": 0.30528469965089683,
+      "grad_norm": 0.923303484916687,
+      "learning_rate": 0.00016300878435817113,
+      "loss": 0.8952,
+      "step": 634
+    },
+    {
+      "epoch": 0.305766221259179,
+      "grad_norm": 0.8884153366088867,
+      "learning_rate": 0.00016288757956446918,
+      "loss": 0.6296,
+      "step": 635
+    },
+    {
+      "epoch": 0.3062477428674612,
+      "grad_norm": 1.4516477584838867,
+      "learning_rate": 0.00016276622175168083,
+      "loss": 0.4871,
+      "step": 636
+    },
+    {
+      "epoch": 0.30672926447574334,
+      "grad_norm": 1.6598066091537476,
+      "learning_rate": 0.0001626447112150959,
+      "loss": 1.208,
+      "step": 637
+    },
+    {
+      "epoch": 0.3072107860840255,
+      "grad_norm": 0.6673110723495483,
+      "learning_rate": 0.00016252304825037576,
+      "loss": 0.7643,
+      "step": 638
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.602836012840271,
+      "learning_rate": 0.0001624012331535528,
+      "loss": 1.1832,
+      "step": 639
+    },
+    {
+      "epoch": 0.30817382930058984,
+      "grad_norm": 0.8961763978004456,
+      "learning_rate": 0.00016227926622102947,
+      "loss": 0.5228,
+      "step": 640
+    },
+    {
+      "epoch": 0.30865535090887203,
+      "grad_norm": 0.9909448623657227,
+      "learning_rate": 0.00016215714774957772,
+      "loss": 0.9294,
+      "step": 641
+    },
+    {
+      "epoch": 0.3091368725171542,
+      "grad_norm": 0.9084739685058594,
+      "learning_rate": 0.00016203487803633822,
+      "loss": 0.6734,
+      "step": 642
+    },
+    {
+      "epoch": 0.3096183941254364,
+      "grad_norm": 0.7972694635391235,
+      "learning_rate": 0.00016191245737881956,
+      "loss": 0.4454,
+      "step": 643
+    },
+    {
+      "epoch": 0.31009991573371853,
+      "grad_norm": 0.909772515296936,
+      "learning_rate": 0.00016178988607489777,
+      "loss": 0.5753,
+      "step": 644
+    },
+    {
+      "epoch": 0.3105814373420007,
+      "grad_norm": 1.612001895904541,
+      "learning_rate": 0.00016166716442281528,
+      "loss": 1.2121,
+      "step": 645
+    },
+    {
+      "epoch": 0.3110629589502829,
+      "grad_norm": 1.3951966762542725,
+      "learning_rate": 0.0001615442927211805,
+      "loss": 0.6069,
+      "step": 646
+    },
+    {
+      "epoch": 0.31154448055856504,
+      "grad_norm": 0.5090633630752563,
+      "learning_rate": 0.0001614212712689668,
+      "loss": 0.4586,
+      "step": 647
+    },
+    {
+      "epoch": 0.3120260021668472,
+      "grad_norm": 1.3730465173721313,
+      "learning_rate": 0.00016129810036551198,
+      "loss": 1.0391,
+      "step": 648
+    },
+    {
+      "epoch": 0.3125075237751294,
+      "grad_norm": 1.2121727466583252,
+      "learning_rate": 0.00016117478031051755,
+      "loss": 1.0782,
+      "step": 649
+    },
+    {
+      "epoch": 0.3129890453834116,
+      "grad_norm": 1.1098741292953491,
+      "learning_rate": 0.00016105131140404787,
+      "loss": 0.642,
+      "step": 650
+    },
+    {
+      "epoch": 0.31347056699169373,
+      "grad_norm": 1.337183952331543,
+      "learning_rate": 0.00016092769394652947,
+      "loss": 0.6911,
+      "step": 651
+    },
+    {
+      "epoch": 0.3139520885999759,
+      "grad_norm": 0.5842844247817993,
+      "learning_rate": 0.0001608039282387504,
+      "loss": 0.2357,
+      "step": 652
+    },
+    {
+      "epoch": 0.3144336102082581,
+      "grad_norm": 1.0377110242843628,
+      "learning_rate": 0.00016068001458185936,
+      "loss": 0.6098,
+      "step": 653
+    },
+    {
+      "epoch": 0.3149151318165403,
+      "grad_norm": 0.868000328540802,
+      "learning_rate": 0.0001605559532773651,
+      "loss": 0.689,
+      "step": 654
+    },
+    {
+      "epoch": 0.3153966534248224,
+      "grad_norm": 0.9237671494483948,
+      "learning_rate": 0.00016043174462713566,
+      "loss": 0.2829,
+      "step": 655
+    },
+    {
+      "epoch": 0.3158781750331046,
+      "grad_norm": 1.0178114175796509,
+      "learning_rate": 0.00016030738893339753,
+      "loss": 0.5194,
+      "step": 656
+    },
+    {
+      "epoch": 0.3163596966413868,
+      "grad_norm": 0.6489741802215576,
+      "learning_rate": 0.00016018288649873497,
+      "loss": 0.6068,
+      "step": 657
+    },
+    {
+      "epoch": 0.31684121824966893,
+      "grad_norm": 0.8735011219978333,
+      "learning_rate": 0.0001600582376260894,
+      "loss": 0.3923,
+      "step": 658
+    },
+    {
+      "epoch": 0.3173227398579511,
+      "grad_norm": 0.7427851557731628,
+      "learning_rate": 0.00015993344261875847,
+      "loss": 0.6916,
+      "step": 659
+    },
+    {
+      "epoch": 0.3178042614662333,
+      "grad_norm": 0.6694474220275879,
+      "learning_rate": 0.00015980850178039547,
+      "loss": 0.6211,
+      "step": 660
+    },
+    {
+      "epoch": 0.3182857830745155,
+      "grad_norm": 0.7267779111862183,
+      "learning_rate": 0.00015968341541500842,
+      "loss": 0.8095,
+      "step": 661
+    },
+    {
+      "epoch": 0.3187673046827976,
+      "grad_norm": 0.7212939262390137,
+      "learning_rate": 0.00015955818382695953,
+      "loss": 0.4024,
+      "step": 662
+    },
+    {
+      "epoch": 0.3192488262910798,
+      "grad_norm": 1.4665344953536987,
+      "learning_rate": 0.00015943280732096438,
+      "loss": 1.3903,
+      "step": 663
+    },
+    {
+      "epoch": 0.319730347899362,
+      "grad_norm": 0.7396234273910522,
+      "learning_rate": 0.00015930728620209113,
+      "loss": 0.6563,
+      "step": 664
+    },
+    {
+      "epoch": 0.32021186950764413,
+      "grad_norm": 1.8860375881195068,
+      "learning_rate": 0.00015918162077575976,
+      "loss": 0.8506,
+      "step": 665
+    },
+    {
+      "epoch": 0.3206933911159263,
+      "grad_norm": 0.8444084525108337,
+      "learning_rate": 0.00015905581134774153,
+      "loss": 0.5888,
+      "step": 666
+    },
+    {
+      "epoch": 0.3211749127242085,
+      "grad_norm": 0.8889416456222534,
+      "learning_rate": 0.0001589298582241579,
+      "loss": 0.9857,
+      "step": 667
+    },
+    {
+      "epoch": 0.3216564343324907,
+      "grad_norm": 0.9507708549499512,
+      "learning_rate": 0.00015880376171148014,
+      "loss": 0.4902,
+      "step": 668
+    },
+    {
+      "epoch": 0.3221379559407728,
+      "grad_norm": 0.7778554558753967,
+      "learning_rate": 0.00015867752211652831,
+      "loss": 0.6368,
+      "step": 669
+    },
+    {
+      "epoch": 0.322619477549055,
+      "grad_norm": 0.9680811762809753,
+      "learning_rate": 0.00015855113974647068,
+      "loss": 1.2206,
+      "step": 670
+    },
+    {
+      "epoch": 0.3231009991573372,
+      "grad_norm": 1.222985029220581,
+      "learning_rate": 0.0001584246149088229,
+      "loss": 0.5694,
+      "step": 671
+    },
+    {
+      "epoch": 0.32358252076561933,
+      "grad_norm": 0.7617347240447998,
+      "learning_rate": 0.0001582979479114472,
+      "loss": 0.6872,
+      "step": 672
+    },
+    {
+      "epoch": 0.3240640423739015,
+      "grad_norm": 0.8455665111541748,
+      "learning_rate": 0.0001581711390625519,
+      "loss": 0.4931,
+      "step": 673
+    },
+    {
+      "epoch": 0.3245455639821837,
+      "grad_norm": 1.143609642982483,
+      "learning_rate": 0.0001580441886706903,
+      "loss": 0.8739,
+      "step": 674
+    },
+    {
+      "epoch": 0.3250270855904659,
+      "grad_norm": 0.8333857655525208,
+      "learning_rate": 0.00015791709704476015,
+      "loss": 0.6653,
+      "step": 675
+    },
+    {
+      "epoch": 0.325508607198748,
+      "grad_norm": 0.41765791177749634,
+      "learning_rate": 0.00015778986449400292,
+      "loss": 0.3178,
+      "step": 676
+    },
+    {
+      "epoch": 0.3259901288070302,
+      "grad_norm": 0.5796922445297241,
+      "learning_rate": 0.00015766249132800292,
+      "loss": 0.5722,
+      "step": 677
+    },
+    {
+      "epoch": 0.3264716504153124,
+      "grad_norm": 0.763243556022644,
+      "learning_rate": 0.00015753497785668663,
+      "loss": 0.3656,
+      "step": 678
+    },
+    {
+      "epoch": 0.3269531720235946,
+      "grad_norm": 1.2675962448120117,
+      "learning_rate": 0.00015740732439032187,
+      "loss": 0.4063,
+      "step": 679
+    },
+    {
+      "epoch": 0.3274346936318767,
+      "grad_norm": 0.7724534273147583,
+      "learning_rate": 0.00015727953123951716,
+      "loss": 0.9757,
+      "step": 680
+    },
+    {
+      "epoch": 0.3279162152401589,
+      "grad_norm": 0.8847770690917969,
+      "learning_rate": 0.00015715159871522086,
+      "loss": 0.5965,
+      "step": 681
+    },
+    {
+      "epoch": 0.3283977368484411,
+      "grad_norm": 0.6064730882644653,
+      "learning_rate": 0.00015702352712872056,
+      "loss": 0.4475,
+      "step": 682
+    },
+    {
+      "epoch": 0.3288792584567232,
+      "grad_norm": 0.9881625175476074,
+      "learning_rate": 0.00015689531679164204,
+      "loss": 0.8257,
+      "step": 683
+    },
+    {
+      "epoch": 0.3293607800650054,
+      "grad_norm": 1.2793527841567993,
+      "learning_rate": 0.00015676696801594886,
+      "loss": 0.9258,
+      "step": 684
+    },
+    {
+      "epoch": 0.3298423016732876,
+      "grad_norm": 0.9854462146759033,
+      "learning_rate": 0.00015663848111394132,
+      "loss": 1.0733,
+      "step": 685
+    },
+    {
+      "epoch": 0.3303238232815698,
+      "grad_norm": 1.0031025409698486,
+      "learning_rate": 0.00015650985639825585,
+      "loss": 1.077,
+      "step": 686
+    },
+    {
+      "epoch": 0.3308053448898519,
+      "grad_norm": 0.9065315127372742,
+      "learning_rate": 0.00015638109418186424,
+      "loss": 0.2222,
+      "step": 687
+    },
+    {
+      "epoch": 0.3312868664981341,
+      "grad_norm": 0.8304358124732971,
+      "learning_rate": 0.00015625219477807277,
+      "loss": 0.7335,
+      "step": 688
+    },
+    {
+      "epoch": 0.3317683881064163,
+      "grad_norm": 0.6953612565994263,
+      "learning_rate": 0.00015612315850052166,
+      "loss": 0.8908,
+      "step": 689
+    },
+    {
+      "epoch": 0.3322499097146984,
+      "grad_norm": 1.183565616607666,
+      "learning_rate": 0.00015599398566318396,
+      "loss": 0.6687,
+      "step": 690
+    },
+    {
+      "epoch": 0.3327314313229806,
+      "grad_norm": 0.7502872943878174,
+      "learning_rate": 0.00015586467658036524,
+      "loss": 0.4798,
+      "step": 691
+    },
+    {
+      "epoch": 0.3332129529312628,
+      "grad_norm": 1.2281885147094727,
+      "learning_rate": 0.00015573523156670244,
+      "loss": 1.0472,
+      "step": 692
+    },
+    {
+      "epoch": 0.333694474539545,
+      "grad_norm": 2.2051644325256348,
+      "learning_rate": 0.0001556056509371633,
+      "loss": 0.9149,
+      "step": 693
+    },
+    {
+      "epoch": 0.3341759961478271,
+      "grad_norm": 0.5925647616386414,
+      "learning_rate": 0.00015547593500704547,
+      "loss": 0.8149,
+      "step": 694
+    },
+    {
+      "epoch": 0.3346575177561093,
+      "grad_norm": 0.5184139013290405,
+      "learning_rate": 0.00015534608409197592,
+      "loss": 0.3174,
+      "step": 695
+    },
+    {
+      "epoch": 0.3351390393643915,
+      "grad_norm": 0.702053964138031,
+      "learning_rate": 0.00015521609850791004,
+      "loss": 0.4663,
+      "step": 696
+    },
+    {
+      "epoch": 0.3356205609726737,
+      "grad_norm": 0.6669120788574219,
+      "learning_rate": 0.0001550859785711308,
+      "loss": 0.4877,
+      "step": 697
+    },
+    {
+      "epoch": 0.3361020825809558,
+      "grad_norm": 1.3044205904006958,
+      "learning_rate": 0.0001549557245982482,
+      "loss": 0.9579,
+      "step": 698
+    },
+    {
+      "epoch": 0.336583604189238,
+      "grad_norm": 0.6670746207237244,
+      "learning_rate": 0.00015482533690619837,
+      "loss": 0.1644,
+      "step": 699
+    },
+    {
+      "epoch": 0.3370651257975202,
+      "grad_norm": 1.218701720237732,
+      "learning_rate": 0.00015469481581224272,
+      "loss": 0.517,
+      "step": 700
+    },
+    {
+      "epoch": 0.3375466474058023,
+      "grad_norm": 0.6596776843070984,
+      "learning_rate": 0.0001545641616339673,
+      "loss": 0.3598,
+      "step": 701
+    },
+    {
+      "epoch": 0.3380281690140845,
+      "grad_norm": 0.7192044854164124,
+      "learning_rate": 0.00015443337468928206,
+      "loss": 0.3021,
+      "step": 702
+    },
+    {
+      "epoch": 0.3385096906223667,
+      "grad_norm": 0.8185523152351379,
+      "learning_rate": 0.00015430245529641986,
+      "loss": 0.331,
+      "step": 703
+    },
+    {
+      "epoch": 0.3389912122306489,
+      "grad_norm": 0.8126874566078186,
+      "learning_rate": 0.00015417140377393596,
+      "loss": 0.7754,
+      "step": 704
+    },
+    {
+      "epoch": 0.339472733838931,
+      "grad_norm": 0.8206513524055481,
+      "learning_rate": 0.00015404022044070704,
+      "loss": 0.4213,
+      "step": 705
+    },
+    {
+      "epoch": 0.3399542554472132,
+      "grad_norm": 0.9104949831962585,
+      "learning_rate": 0.00015390890561593052,
+      "loss": 0.6292,
+      "step": 706
+    },
+    {
+      "epoch": 0.3404357770554954,
+      "grad_norm": 0.8065645694732666,
+      "learning_rate": 0.0001537774596191238,
+      "loss": 0.5776,
+      "step": 707
+    },
+    {
+      "epoch": 0.3409172986637775,
+      "grad_norm": 0.8759814500808716,
+      "learning_rate": 0.00015364588277012344,
+      "loss": 0.6147,
+      "step": 708
+    },
+    {
+      "epoch": 0.3413988202720597,
+      "grad_norm": 0.8460003137588501,
+      "learning_rate": 0.00015351417538908435,
+      "loss": 1.1585,
+      "step": 709
+    },
+    {
+      "epoch": 0.3418803418803419,
+      "grad_norm": 0.5036705136299133,
+      "learning_rate": 0.0001533823377964791,
+      "loss": 0.5074,
+      "step": 710
+    },
+    {
+      "epoch": 0.3423618634886241,
+      "grad_norm": 0.9990852475166321,
+      "learning_rate": 0.00015325037031309704,
+      "loss": 0.817,
+      "step": 711
+    },
+    {
+      "epoch": 0.3428433850969062,
+      "grad_norm": 1.0572148561477661,
+      "learning_rate": 0.00015311827326004363,
+      "loss": 0.5744,
+      "step": 712
+    },
+    {
+      "epoch": 0.3433249067051884,
+      "grad_norm": 0.6696701049804688,
+      "learning_rate": 0.0001529860469587396,
+      "loss": 0.5898,
+      "step": 713
+    },
+    {
+      "epoch": 0.3438064283134706,
+      "grad_norm": 1.029279351234436,
+      "learning_rate": 0.00015285369173092015,
+      "loss": 0.4406,
+      "step": 714
+    },
+    {
+      "epoch": 0.3442879499217527,
+      "grad_norm": 1.0727728605270386,
+      "learning_rate": 0.00015272120789863413,
+      "loss": 0.7744,
+      "step": 715
+    },
+    {
+      "epoch": 0.3447694715300349,
+      "grad_norm": 0.7367604374885559,
+      "learning_rate": 0.00015258859578424342,
+      "loss": 0.7764,
+      "step": 716
+    },
+    {
+      "epoch": 0.3452509931383171,
+      "grad_norm": 0.877487063407898,
+      "learning_rate": 0.00015245585571042194,
+      "loss": 0.3476,
+      "step": 717
+    },
+    {
+      "epoch": 0.3457325147465993,
+      "grad_norm": 0.7964932918548584,
+      "learning_rate": 0.00015232298800015506,
+      "loss": 0.399,
+      "step": 718
+    },
+    {
+      "epoch": 0.3462140363548814,
+      "grad_norm": 0.7974869608879089,
+      "learning_rate": 0.00015218999297673862,
+      "loss": 0.3338,
+      "step": 719
+    },
+    {
+      "epoch": 0.3466955579631636,
+      "grad_norm": 0.8630387783050537,
+      "learning_rate": 0.0001520568709637783,
+      "loss": 0.7174,
+      "step": 720
+    },
+    {
+      "epoch": 0.3471770795714458,
+      "grad_norm": 0.797203540802002,
+      "learning_rate": 0.00015192362228518875,
+      "loss": 0.4063,
+      "step": 721
+    },
+    {
+      "epoch": 0.34765860117972797,
+      "grad_norm": 0.870797336101532,
+      "learning_rate": 0.00015179024726519284,
+      "loss": 0.8056,
+      "step": 722
+    },
+    {
+      "epoch": 0.3481401227880101,
+      "grad_norm": 0.7476208806037903,
+      "learning_rate": 0.00015165674622832085,
+      "loss": 0.4459,
+      "step": 723
+    },
+    {
+      "epoch": 0.3486216443962923,
+      "grad_norm": 1.1456966400146484,
+      "learning_rate": 0.0001515231194994097,
+      "loss": 1.2802,
+      "step": 724
+    },
+    {
+      "epoch": 0.34910316600457447,
+      "grad_norm": 0.7962654232978821,
+      "learning_rate": 0.00015138936740360207,
+      "loss": 0.8159,
+      "step": 725
+    },
+    {
+      "epoch": 0.3495846876128566,
+      "grad_norm": 0.5400972366333008,
+      "learning_rate": 0.00015125549026634585,
+      "loss": 0.3333,
+      "step": 726
+    },
+    {
+      "epoch": 0.3500662092211388,
+      "grad_norm": 0.683711051940918,
+      "learning_rate": 0.00015112148841339295,
+      "loss": 0.648,
+      "step": 727
+    },
+    {
+      "epoch": 0.350547730829421,
+      "grad_norm": 1.0952285528182983,
+      "learning_rate": 0.000150987362170799,
+      "loss": 0.8084,
+      "step": 728
+    },
+    {
+      "epoch": 0.35102925243770317,
+      "grad_norm": 1.6250587701797485,
+      "learning_rate": 0.00015085311186492206,
+      "loss": 0.8142,
+      "step": 729
+    },
+    {
+      "epoch": 0.3515107740459853,
+      "grad_norm": 1.0683465003967285,
+      "learning_rate": 0.00015071873782242223,
+      "loss": 0.6077,
+      "step": 730
+    },
+    {
+      "epoch": 0.3519922956542675,
+      "grad_norm": 1.0285414457321167,
+      "learning_rate": 0.0001505842403702606,
+      "loss": 0.7088,
+      "step": 731
+    },
+    {
+      "epoch": 0.35247381726254967,
+      "grad_norm": 0.675561785697937,
+      "learning_rate": 0.00015044961983569856,
+      "loss": 0.4079,
+      "step": 732
+    },
+    {
+      "epoch": 0.3529553388708318,
+      "grad_norm": 0.5926306247711182,
+      "learning_rate": 0.00015031487654629702,
+      "loss": 0.249,
+      "step": 733
+    },
+    {
+      "epoch": 0.353436860479114,
+      "grad_norm": 0.6946805119514465,
+      "learning_rate": 0.00015018001082991553,
+      "loss": 0.2278,
+      "step": 734
+    },
+    {
+      "epoch": 0.3539183820873962,
+      "grad_norm": 1.2133913040161133,
+      "learning_rate": 0.0001500450230147116,
+      "loss": 0.6548,
+      "step": 735
+    },
+    {
+      "epoch": 0.35439990369567836,
+      "grad_norm": 1.1130623817443848,
+      "learning_rate": 0.00014990991342913974,
+      "loss": 0.87,
+      "step": 736
+    },
+    {
+      "epoch": 0.3548814253039605,
+      "grad_norm": 1.3685212135314941,
+      "learning_rate": 0.00014977468240195084,
+      "loss": 0.8236,
+      "step": 737
+    },
+    {
+      "epoch": 0.3553629469122427,
+      "grad_norm": 1.4713952541351318,
+      "learning_rate": 0.0001496393302621912,
+      "loss": 1.1213,
+      "step": 738
+    },
+    {
+      "epoch": 0.35584446852052487,
+      "grad_norm": 0.9908416867256165,
+      "learning_rate": 0.00014950385733920188,
+      "loss": 0.5959,
+      "step": 739
+    },
+    {
+      "epoch": 0.35632599012880706,
+      "grad_norm": 1.0449920892715454,
+      "learning_rate": 0.00014936826396261783,
+      "loss": 0.735,
+      "step": 740
+    },
+    {
+      "epoch": 0.3568075117370892,
+      "grad_norm": 1.2540510892868042,
+      "learning_rate": 0.00014923255046236705,
+      "loss": 0.6025,
+      "step": 741
+    },
+    {
+      "epoch": 0.3572890333453714,
+      "grad_norm": 0.9532543420791626,
+      "learning_rate": 0.00014909671716866984,
+      "loss": 0.622,
+      "step": 742
+    },
+    {
+      "epoch": 0.35777055495365356,
+      "grad_norm": 1.1329494714736938,
+      "learning_rate": 0.00014896076441203802,
+      "loss": 1.3104,
+      "step": 743
+    },
+    {
+      "epoch": 0.3582520765619357,
+      "grad_norm": 0.6802012324333191,
+      "learning_rate": 0.000148824692523274,
+      "loss": 0.7314,
+      "step": 744
+    },
+    {
+      "epoch": 0.3587335981702179,
+      "grad_norm": 0.8869780898094177,
+      "learning_rate": 0.0001486885018334702,
+      "loss": 0.9224,
+      "step": 745
+    },
+    {
+      "epoch": 0.35921511977850007,
+      "grad_norm": 1.5476834774017334,
+      "learning_rate": 0.00014855219267400797,
+      "loss": 0.7926,
+      "step": 746
+    },
+    {
+      "epoch": 0.35969664138678226,
+      "grad_norm": 0.7869364023208618,
+      "learning_rate": 0.00014841576537655705,
+      "loss": 0.4486,
+      "step": 747
+    },
+    {
+      "epoch": 0.3601781629950644,
+      "grad_norm": 0.6606510281562805,
+      "learning_rate": 0.00014827922027307451,
+      "loss": 0.4667,
+      "step": 748
+    },
+    {
+      "epoch": 0.3606596846033466,
+      "grad_norm": 0.6294459700584412,
+      "learning_rate": 0.00014814255769580415,
+      "loss": 0.4268,
+      "step": 749
+    },
+    {
+      "epoch": 0.36114120621162876,
+      "grad_norm": 1.0464389324188232,
+      "learning_rate": 0.00014800577797727558,
+      "loss": 0.8922,
+      "step": 750
+    },
+    {
+      "epoch": 0.3616227278199109,
+      "grad_norm": 0.8546883463859558,
+      "learning_rate": 0.00014786888145030343,
+      "loss": 1.1779,
+      "step": 751
+    },
+    {
+      "epoch": 0.3621042494281931,
+      "grad_norm": 0.7854865789413452,
+      "learning_rate": 0.0001477318684479866,
+      "loss": 0.1934,
+      "step": 752
+    },
+    {
+      "epoch": 0.36258577103647527,
+      "grad_norm": 2.1393556594848633,
+      "learning_rate": 0.00014759473930370736,
+      "loss": 0.7357,
+      "step": 753
+    },
+    {
+      "epoch": 0.36306729264475746,
+      "grad_norm": 0.7535027861595154,
+      "learning_rate": 0.0001474574943511306,
+      "loss": 0.6884,
+      "step": 754
+    },
+    {
+      "epoch": 0.3635488142530396,
+      "grad_norm": 1.1644819974899292,
+      "learning_rate": 0.0001473201339242029,
+      "loss": 1.1809,
+      "step": 755
+    },
+    {
+      "epoch": 0.3640303358613218,
+      "grad_norm": 1.5013142824172974,
+      "learning_rate": 0.000147182658357152,
+      "loss": 0.6,
+      "step": 756
+    },
+    {
+      "epoch": 0.36451185746960396,
+      "grad_norm": 0.8900628685951233,
+      "learning_rate": 0.00014704506798448566,
+      "loss": 0.5416,
+      "step": 757
+    },
+    {
+      "epoch": 0.3649933790778861,
+      "grad_norm": 0.8392864465713501,
+      "learning_rate": 0.00014690736314099101,
+      "loss": 0.711,
+      "step": 758
+    },
+    {
+      "epoch": 0.3654749006861683,
+      "grad_norm": 0.5832529067993164,
+      "learning_rate": 0.00014676954416173373,
+      "loss": 0.5333,
+      "step": 759
+    },
+    {
+      "epoch": 0.36595642229445047,
+      "grad_norm": 1.227352261543274,
+      "learning_rate": 0.00014663161138205724,
+      "loss": 1.1198,
+      "step": 760
+    },
+    {
+      "epoch": 0.36643794390273265,
+      "grad_norm": 0.8971624970436096,
+      "learning_rate": 0.00014649356513758176,
+      "loss": 0.429,
+      "step": 761
+    },
+    {
+      "epoch": 0.3669194655110148,
+      "grad_norm": 0.860392153263092,
+      "learning_rate": 0.00014635540576420374,
+      "loss": 0.6527,
+      "step": 762
+    },
+    {
+      "epoch": 0.367400987119297,
+      "grad_norm": 0.34480923414230347,
+      "learning_rate": 0.0001462171335980948,
+      "loss": 0.3004,
+      "step": 763
+    },
+    {
+      "epoch": 0.36788250872757916,
+      "grad_norm": 1.0565292835235596,
+      "learning_rate": 0.00014607874897570105,
+      "loss": 0.3944,
+      "step": 764
+    },
+    {
+      "epoch": 0.36836403033586135,
+      "grad_norm": 0.7156208753585815,
+      "learning_rate": 0.0001459402522337422,
+      "loss": 0.1887,
+      "step": 765
+    },
+    {
+      "epoch": 0.3688455519441435,
+      "grad_norm": 1.2437664270401,
+      "learning_rate": 0.00014580164370921078,
+      "loss": 0.4478,
+      "step": 766
+    },
+    {
+      "epoch": 0.36932707355242567,
+      "grad_norm": 0.6799991726875305,
+      "learning_rate": 0.0001456629237393713,
+      "loss": 0.4307,
+      "step": 767
+    },
+    {
+      "epoch": 0.36980859516070785,
+      "grad_norm": 0.487223744392395,
+      "learning_rate": 0.00014552409266175952,
+      "loss": 0.1922,
+      "step": 768
+    },
+    {
+      "epoch": 0.37029011676899,
+      "grad_norm": 0.73237544298172,
+      "learning_rate": 0.00014538515081418142,
+      "loss": 0.6541,
+      "step": 769
+    },
+    {
+      "epoch": 0.3707716383772722,
+      "grad_norm": 1.2179851531982422,
+      "learning_rate": 0.00014524609853471264,
+      "loss": 0.8882,
+      "step": 770
+    },
+    {
+      "epoch": 0.37125315998555436,
+      "grad_norm": 1.2429336309432983,
+      "learning_rate": 0.00014510693616169741,
+      "loss": 0.6963,
+      "step": 771
+    },
+    {
+      "epoch": 0.37173468159383655,
+      "grad_norm": 0.764161229133606,
+      "learning_rate": 0.0001449676640337479,
+      "loss": 0.8483,
+      "step": 772
+    },
+    {
+      "epoch": 0.3722162032021187,
+      "grad_norm": 0.7411964535713196,
+      "learning_rate": 0.00014482828248974335,
+      "loss": 0.3896,
+      "step": 773
+    },
+    {
+      "epoch": 0.37269772481040087,
+      "grad_norm": 0.7775552868843079,
+      "learning_rate": 0.00014468879186882916,
+      "loss": 0.7667,
+      "step": 774
+    },
+    {
+      "epoch": 0.37317924641868305,
+      "grad_norm": 0.5666738748550415,
+      "learning_rate": 0.00014454919251041622,
+      "loss": 0.3198,
+      "step": 775
+    },
+    {
+      "epoch": 0.3736607680269652,
+      "grad_norm": 1.0107135772705078,
+      "learning_rate": 0.00014440948475418,
+      "loss": 0.8955,
+      "step": 776
+    },
+    {
+      "epoch": 0.37414228963524737,
+      "grad_norm": 0.7883855104446411,
+      "learning_rate": 0.00014426966894005966,
+      "loss": 0.5721,
+      "step": 777
+    },
+    {
+      "epoch": 0.37462381124352956,
+      "grad_norm": 0.8659805059432983,
+      "learning_rate": 0.0001441297454082573,
+      "loss": 1.0827,
+      "step": 778
+    },
+    {
+      "epoch": 0.37510533285181175,
+      "grad_norm": 0.8902775049209595,
+      "learning_rate": 0.00014398971449923722,
+      "loss": 0.3986,
+      "step": 779
+    },
+    {
+      "epoch": 0.3755868544600939,
+      "grad_norm": 0.9566343426704407,
+      "learning_rate": 0.00014384957655372483,
+      "loss": 0.622,
+      "step": 780
+    },
+    {
+      "epoch": 0.37606837606837606,
+      "grad_norm": 0.7155655026435852,
+      "learning_rate": 0.00014370933191270617,
+      "loss": 0.5707,
+      "step": 781
+    },
+    {
+      "epoch": 0.37654989767665825,
+      "grad_norm": 1.2331191301345825,
+      "learning_rate": 0.0001435689809174267,
+      "loss": 0.8049,
+      "step": 782
+    },
+    {
+      "epoch": 0.37703141928494044,
+      "grad_norm": 1.137040615081787,
+      "learning_rate": 0.0001434285239093908,
+      "loss": 1.2335,
+      "step": 783
+    },
+    {
+      "epoch": 0.37751294089322257,
+      "grad_norm": 1.0161875486373901,
+      "learning_rate": 0.00014328796123036071,
+      "loss": 0.665,
+      "step": 784
+    },
+    {
+      "epoch": 0.37799446250150476,
+      "grad_norm": 0.5084791779518127,
+      "learning_rate": 0.0001431472932223559,
+      "loss": 0.3809,
+      "step": 785
+    },
+    {
+      "epoch": 0.37847598410978694,
+      "grad_norm": 0.9150707721710205,
+      "learning_rate": 0.00014300652022765207,
+      "loss": 0.4742,
+      "step": 786
+    },
+    {
+      "epoch": 0.3789575057180691,
+      "grad_norm": 0.849323034286499,
+      "learning_rate": 0.00014286564258878033,
+      "loss": 0.7684,
+      "step": 787
+    },
+    {
+      "epoch": 0.37943902732635126,
+      "grad_norm": 1.380608320236206,
+      "learning_rate": 0.00014272466064852644,
+      "loss": 0.522,
+      "step": 788
+    },
+    {
+      "epoch": 0.37992054893463345,
+      "grad_norm": 1.035540223121643,
+      "learning_rate": 0.00014258357474993,
+      "loss": 1.0147,
+      "step": 789
+    },
+    {
+      "epoch": 0.38040207054291564,
+      "grad_norm": 1.7030549049377441,
+      "learning_rate": 0.0001424423852362835,
+      "loss": 0.541,
+      "step": 790
+    },
+    {
+      "epoch": 0.38088359215119777,
+      "grad_norm": 0.6918346285820007,
+      "learning_rate": 0.00014230109245113158,
+      "loss": 0.7923,
+      "step": 791
+    },
+    {
+      "epoch": 0.38136511375947996,
+      "grad_norm": 0.7446008324623108,
+      "learning_rate": 0.00014215969673827018,
+      "loss": 0.3233,
+      "step": 792
+    },
+    {
+      "epoch": 0.38184663536776214,
+      "grad_norm": 0.4108048975467682,
+      "learning_rate": 0.00014201819844174564,
+      "loss": 0.1844,
+      "step": 793
+    },
+    {
+      "epoch": 0.3823281569760443,
+      "grad_norm": 1.1727381944656372,
+      "learning_rate": 0.0001418765979058539,
+      "loss": 0.8246,
+      "step": 794
+    },
+    {
+      "epoch": 0.38280967858432646,
+      "grad_norm": 0.8327621221542358,
+      "learning_rate": 0.00014173489547513973,
+      "loss": 0.9647,
+      "step": 795
+    },
+    {
+      "epoch": 0.38329120019260865,
+      "grad_norm": 0.9883455634117126,
+      "learning_rate": 0.00014159309149439582,
+      "loss": 0.6689,
+      "step": 796
+    },
+    {
+      "epoch": 0.38377272180089084,
+      "grad_norm": 0.9439548254013062,
+      "learning_rate": 0.00014145118630866187,
+      "loss": 0.3843,
+      "step": 797
+    },
+    {
+      "epoch": 0.38425424340917297,
+      "grad_norm": 1.062584638595581,
+      "learning_rate": 0.000141309180263224,
+      "loss": 0.2714,
+      "step": 798
+    },
+    {
+      "epoch": 0.38473576501745516,
+      "grad_norm": 1.0796810388565063,
+      "learning_rate": 0.0001411670737036135,
+      "loss": 0.4017,
+      "step": 799
+    },
+    {
+      "epoch": 0.38521728662573734,
+      "grad_norm": 0.6542953252792358,
+      "learning_rate": 0.0001410248669756065,
+      "loss": 0.3069,
+      "step": 800
+    },
+    {
+      "epoch": 0.3856988082340195,
+      "grad_norm": 1.2995156049728394,
+      "learning_rate": 0.00014088256042522264,
+      "loss": 0.7217,
+      "step": 801
+    },
+    {
+      "epoch": 0.38618032984230166,
+      "grad_norm": 0.58650803565979,
+      "learning_rate": 0.00014074015439872458,
+      "loss": 0.4809,
+      "step": 802
+    },
+    {
+      "epoch": 0.38666185145058385,
+      "grad_norm": 0.7780939340591431,
+      "learning_rate": 0.00014059764924261703,
+      "loss": 0.3629,
+      "step": 803
+    },
+    {
+      "epoch": 0.38714337305886604,
+      "grad_norm": 1.2356970310211182,
+      "learning_rate": 0.00014045504530364584,
+      "loss": 0.3341,
+      "step": 804
+    },
+    {
+      "epoch": 0.38762489466714817,
+      "grad_norm": 1.130676507949829,
+      "learning_rate": 0.00014031234292879725,
+      "loss": 0.5205,
+      "step": 805
+    },
+    {
+      "epoch": 0.38810641627543035,
+      "grad_norm": 0.8669336438179016,
+      "learning_rate": 0.00014016954246529696,
+      "loss": 0.7471,
+      "step": 806
+    },
+    {
+      "epoch": 0.38858793788371254,
+      "grad_norm": 1.1832457780838013,
+      "learning_rate": 0.00014002664426060942,
+      "loss": 0.4838,
+      "step": 807
+    },
+    {
+      "epoch": 0.38906945949199473,
+      "grad_norm": 0.7935713529586792,
+      "learning_rate": 0.00013988364866243693,
+      "loss": 0.883,
+      "step": 808
+    },
+    {
+      "epoch": 0.38955098110027686,
+      "grad_norm": 0.6437901258468628,
+      "learning_rate": 0.00013974055601871868,
+      "loss": 0.4664,
+      "step": 809
+    },
+    {
+      "epoch": 0.39003250270855905,
+      "grad_norm": 1.122780680656433,
+      "learning_rate": 0.00013959736667762998,
+      "loss": 0.7229,
+      "step": 810
+    },
+    {
+      "epoch": 0.39051402431684124,
+      "grad_norm": 0.9974272847175598,
+      "learning_rate": 0.00013945408098758156,
+      "loss": 0.5652,
+      "step": 811
+    },
+    {
+      "epoch": 0.39099554592512337,
+      "grad_norm": 0.5831164121627808,
+      "learning_rate": 0.0001393106992972184,
+      "loss": 0.4848,
+      "step": 812
+    },
+    {
+      "epoch": 0.39147706753340555,
+      "grad_norm": 0.643200159072876,
+      "learning_rate": 0.00013916722195541926,
+      "loss": 0.3277,
+      "step": 813
+    },
+    {
+      "epoch": 0.39195858914168774,
+      "grad_norm": 1.4757206439971924,
+      "learning_rate": 0.00013902364931129557,
+      "loss": 0.6434,
+      "step": 814
+    },
+    {
+      "epoch": 0.39244011074996993,
+      "grad_norm": 2.269386053085327,
+      "learning_rate": 0.00013887998171419058,
+      "loss": 0.9765,
+      "step": 815
+    },
+    {
+      "epoch": 0.39292163235825206,
+      "grad_norm": 1.381133794784546,
+      "learning_rate": 0.00013873621951367862,
+      "loss": 1.3941,
+      "step": 816
+    },
+    {
+      "epoch": 0.39340315396653425,
+      "grad_norm": 0.5003638863563538,
+      "learning_rate": 0.00013859236305956425,
+      "loss": 0.5137,
+      "step": 817
+    },
+    {
+      "epoch": 0.39388467557481643,
+      "grad_norm": 0.9194219708442688,
+      "learning_rate": 0.00013844841270188132,
+      "loss": 0.7313,
+      "step": 818
+    },
+    {
+      "epoch": 0.39436619718309857,
+      "grad_norm": 0.6109256148338318,
+      "learning_rate": 0.00013830436879089228,
+      "loss": 0.7035,
+      "step": 819
+    },
+    {
+      "epoch": 0.39484771879138075,
+      "grad_norm": 0.5414918065071106,
+      "learning_rate": 0.00013816023167708704,
+      "loss": 0.241,
+      "step": 820
+    },
+    {
+      "epoch": 0.39532924039966294,
+      "grad_norm": 0.564078688621521,
+      "learning_rate": 0.00013801600171118244,
+      "loss": 0.2047,
+      "step": 821
+    },
+    {
+      "epoch": 0.3958107620079451,
+      "grad_norm": 1.235252857208252,
+      "learning_rate": 0.00013787167924412112,
+      "loss": 0.7474,
+      "step": 822
+    },
+    {
+      "epoch": 0.39629228361622726,
+      "grad_norm": 0.5138280987739563,
+      "learning_rate": 0.0001377272646270709,
+      "loss": 0.2388,
+      "step": 823
+    },
+    {
+      "epoch": 0.39677380522450945,
+      "grad_norm": 0.6855552196502686,
+      "learning_rate": 0.00013758275821142382,
+      "loss": 0.3218,
+      "step": 824
+    },
+    {
+      "epoch": 0.39725532683279163,
+      "grad_norm": 1.0833673477172852,
+      "learning_rate": 0.00013743816034879523,
+      "loss": 0.3818,
+      "step": 825
+    },
+    {
+      "epoch": 0.3977368484410738,
+      "grad_norm": 0.8923762440681458,
+      "learning_rate": 0.000137293471391023,
+      "loss": 0.8322,
+      "step": 826
+    },
+    {
+      "epoch": 0.39821837004935595,
+      "grad_norm": 1.838834524154663,
+      "learning_rate": 0.00013714869169016667,
+      "loss": 0.3636,
+      "step": 827
+    },
+    {
+      "epoch": 0.39869989165763814,
+      "grad_norm": 0.7193182706832886,
+      "learning_rate": 0.00013700382159850656,
+      "loss": 0.3181,
+      "step": 828
+    },
+    {
+      "epoch": 0.3991814132659203,
+      "grad_norm": 0.867734432220459,
+      "learning_rate": 0.00013685886146854297,
+      "loss": 0.3646,
+      "step": 829
+    },
+    {
+      "epoch": 0.39966293487420246,
+      "grad_norm": 1.1769030094146729,
+      "learning_rate": 0.00013671381165299525,
+      "loss": 0.6471,
+      "step": 830
+    },
+    {
+      "epoch": 0.40014445648248465,
+      "grad_norm": 2.466710329055786,
+      "learning_rate": 0.00013656867250480098,
+      "loss": 0.5789,
+      "step": 831
+    },
+    {
+      "epoch": 0.40062597809076683,
+      "grad_norm": 0.5317124724388123,
+      "learning_rate": 0.00013642344437711512,
+      "loss": 0.2964,
+      "step": 832
+    },
+    {
+      "epoch": 0.401107499699049,
+      "grad_norm": 1.1287859678268433,
+      "learning_rate": 0.00013627812762330912,
+      "loss": 0.8241,
+      "step": 833
+    },
+    {
+      "epoch": 0.40158902130733115,
+      "grad_norm": 0.7147384285926819,
+      "learning_rate": 0.00013613272259697007,
+      "loss": 0.5656,
+      "step": 834
+    },
+    {
+      "epoch": 0.40207054291561334,
+      "grad_norm": 1.6891028881072998,
+      "learning_rate": 0.00013598722965189986,
+      "loss": 0.9899,
+      "step": 835
+    },
+    {
+      "epoch": 0.4025520645238955,
+      "grad_norm": 0.943759560585022,
+      "learning_rate": 0.0001358416491421143,
+      "loss": 0.5832,
+      "step": 836
+    },
+    {
+      "epoch": 0.40303358613217766,
+      "grad_norm": 0.8546479344367981,
+      "learning_rate": 0.00013569598142184225,
+      "loss": 0.7991,
+      "step": 837
+    },
+    {
+      "epoch": 0.40351510774045984,
+      "grad_norm": 0.9846085906028748,
+      "learning_rate": 0.00013555022684552483,
+      "loss": 0.9692,
+      "step": 838
+    },
+    {
+      "epoch": 0.40399662934874203,
+      "grad_norm": 1.969257116317749,
+      "learning_rate": 0.00013540438576781441,
+      "loss": 0.8052,
+      "step": 839
+    },
+    {
+      "epoch": 0.4044781509570242,
+      "grad_norm": 1.115907907485962,
+      "learning_rate": 0.0001352584585435739,
+      "loss": 1.2649,
+      "step": 840
+    },
+    {
+      "epoch": 0.40495967256530635,
+      "grad_norm": 0.6951058506965637,
+      "learning_rate": 0.00013511244552787583,
+      "loss": 0.2709,
+      "step": 841
+    },
+    {
+      "epoch": 0.40544119417358854,
+      "grad_norm": 1.5416226387023926,
+      "learning_rate": 0.00013496634707600147,
+      "loss": 0.7605,
+      "step": 842
+    },
+    {
+      "epoch": 0.4059227157818707,
+      "grad_norm": 1.0881750583648682,
+      "learning_rate": 0.0001348201635434399,
+      "loss": 0.5618,
+      "step": 843
+    },
+    {
+      "epoch": 0.40640423739015286,
+      "grad_norm": 1.3643512725830078,
+      "learning_rate": 0.0001346738952858873,
+      "loss": 1.0733,
+      "step": 844
+    },
+    {
+      "epoch": 0.40688575899843504,
+      "grad_norm": 0.7877455353736877,
+      "learning_rate": 0.000134527542659246,
+      "loss": 1.6297,
+      "step": 845
+    },
+    {
+      "epoch": 0.40736728060671723,
+      "grad_norm": 0.5141893029212952,
+      "learning_rate": 0.00013438110601962362,
+      "loss": 0.1867,
+      "step": 846
+    },
+    {
+      "epoch": 0.4078488022149994,
+      "grad_norm": 0.5263735055923462,
+      "learning_rate": 0.00013423458572333214,
+      "loss": 0.7638,
+      "step": 847
+    },
+    {
+      "epoch": 0.40833032382328155,
+      "grad_norm": 0.5297187566757202,
+      "learning_rate": 0.0001340879821268872,
+      "loss": 0.4511,
+      "step": 848
+    },
+    {
+      "epoch": 0.40881184543156374,
+      "grad_norm": 1.2766536474227905,
+      "learning_rate": 0.000133941295587007,
+      "loss": 0.4998,
+      "step": 849
+    },
+    {
+      "epoch": 0.4092933670398459,
+      "grad_norm": 0.8083428740501404,
+      "learning_rate": 0.00013379452646061164,
+      "loss": 0.1786,
+      "step": 850
+    },
+    {
+      "epoch": 0.4097748886481281,
+      "grad_norm": 0.8990474343299866,
+      "learning_rate": 0.0001336476751048222,
+      "loss": 0.99,
+      "step": 851
+    },
+    {
+      "epoch": 0.41025641025641024,
+      "grad_norm": 0.793449878692627,
+      "learning_rate": 0.00013350074187695979,
+      "loss": 0.5605,
+      "step": 852
+    },
+    {
+      "epoch": 0.41073793186469243,
+      "grad_norm": 0.7886202931404114,
+      "learning_rate": 0.00013335372713454467,
+      "loss": 0.294,
+      "step": 853
+    },
+    {
+      "epoch": 0.4112194534729746,
+      "grad_norm": 0.8802816867828369,
+      "learning_rate": 0.0001332066312352956,
+      "loss": 0.6326,
+      "step": 854
+    },
+    {
+      "epoch": 0.41170097508125675,
+      "grad_norm": 0.8170877695083618,
+      "learning_rate": 0.00013305945453712868,
+      "loss": 0.6033,
+      "step": 855
+    },
+    {
+      "epoch": 0.41218249668953894,
+      "grad_norm": 0.5515619516372681,
+      "learning_rate": 0.0001329121973981567,
+      "loss": 0.5479,
+      "step": 856
+    },
+    {
+      "epoch": 0.4126640182978211,
+      "grad_norm": 1.2902063131332397,
+      "learning_rate": 0.00013276486017668807,
+      "loss": 0.6442,
+      "step": 857
+    },
+    {
+      "epoch": 0.4131455399061033,
+      "grad_norm": 0.8257691860198975,
+      "learning_rate": 0.0001326174432312262,
+      "loss": 1.1458,
+      "step": 858
+    },
+    {
+      "epoch": 0.41362706151438544,
+      "grad_norm": 0.5979056358337402,
+      "learning_rate": 0.00013246994692046836,
+      "loss": 0.5462,
+      "step": 859
+    },
+    {
+      "epoch": 0.41410858312266763,
+      "grad_norm": 0.6814475655555725,
+      "learning_rate": 0.000132322371603305,
+      "loss": 0.6534,
+      "step": 860
+    },
+    {
+      "epoch": 0.4145901047309498,
+      "grad_norm": 0.6793029308319092,
+      "learning_rate": 0.0001321747176388188,
+      "loss": 0.2851,
+      "step": 861
+    },
+    {
+      "epoch": 0.41507162633923195,
+      "grad_norm": 0.5598545670509338,
+      "learning_rate": 0.00013202698538628376,
+      "loss": 0.7152,
+      "step": 862
+    },
+    {
+      "epoch": 0.41555314794751413,
+      "grad_norm": 0.7994054555892944,
+      "learning_rate": 0.00013187917520516448,
+      "loss": 0.6246,
+      "step": 863
+    },
+    {
+      "epoch": 0.4160346695557963,
+      "grad_norm": 0.9414712190628052,
+      "learning_rate": 0.00013173128745511508,
+      "loss": 1.0983,
+      "step": 864
+    },
+    {
+      "epoch": 0.4165161911640785,
+      "grad_norm": 0.8112841844558716,
+      "learning_rate": 0.0001315833224959784,
+      "loss": 0.5866,
+      "step": 865
+    },
+    {
+      "epoch": 0.41699771277236064,
+      "grad_norm": 0.9308454990386963,
+      "learning_rate": 0.00013143528068778525,
+      "loss": 1.0087,
+      "step": 866
+    },
+    {
+      "epoch": 0.4174792343806428,
+      "grad_norm": 0.592212438583374,
+      "learning_rate": 0.00013128716239075338,
+      "loss": 0.3319,
+      "step": 867
+    },
+    {
+      "epoch": 0.417960755988925,
+      "grad_norm": 0.8218088150024414,
+      "learning_rate": 0.00013113896796528664,
+      "loss": 0.8615,
+      "step": 868
+    },
+    {
+      "epoch": 0.4184422775972072,
+      "grad_norm": 0.7835381627082825,
+      "learning_rate": 0.00013099069777197412,
+      "loss": 0.8362,
+      "step": 869
+    },
+    {
+      "epoch": 0.41892379920548933,
+      "grad_norm": 0.8324255347251892,
+      "learning_rate": 0.0001308423521715893,
+      "loss": 0.9373,
+      "step": 870
+    },
+    {
+      "epoch": 0.4194053208137715,
+      "grad_norm": 0.7907895445823669,
+      "learning_rate": 0.00013069393152508906,
+      "loss": 0.5857,
+      "step": 871
+    },
+    {
+      "epoch": 0.4198868424220537,
+      "grad_norm": 0.7639227509498596,
+      "learning_rate": 0.00013054543619361303,
+      "loss": 0.3794,
+      "step": 872
+    },
+    {
+      "epoch": 0.42036836403033584,
+      "grad_norm": 1.056784749031067,
+      "learning_rate": 0.0001303968665384824,
+      "loss": 1.1836,
+      "step": 873
+    },
+    {
+      "epoch": 0.420849885638618,
+      "grad_norm": 0.6656973958015442,
+      "learning_rate": 0.00013024822292119934,
+      "loss": 0.2145,
+      "step": 874
+    },
+    {
+      "epoch": 0.4213314072469002,
+      "grad_norm": 0.7051798105239868,
+      "learning_rate": 0.0001300995057034459,
+      "loss": 0.5826,
+      "step": 875
+    },
+    {
+      "epoch": 0.4218129288551824,
+      "grad_norm": 0.9368335604667664,
+      "learning_rate": 0.00012995071524708325,
+      "loss": 0.5162,
+      "step": 876
+    },
+    {
+      "epoch": 0.42229445046346453,
+      "grad_norm": 0.608950138092041,
+      "learning_rate": 0.00012980185191415074,
+      "loss": 0.2055,
+      "step": 877
+    },
+    {
+      "epoch": 0.4227759720717467,
+      "grad_norm": 0.8486785292625427,
+      "learning_rate": 0.0001296529160668651,
+      "loss": 0.1395,
+      "step": 878
+    },
+    {
+      "epoch": 0.4232574936800289,
+      "grad_norm": 0.6813993453979492,
+      "learning_rate": 0.00012950390806761944,
+      "loss": 0.8436,
+      "step": 879
+    },
+    {
+      "epoch": 0.42373901528831104,
+      "grad_norm": 0.5258020758628845,
+      "learning_rate": 0.0001293548282789825,
+      "loss": 0.1877,
+      "step": 880
+    },
+    {
+      "epoch": 0.4242205368965932,
+      "grad_norm": 0.9190642833709717,
+      "learning_rate": 0.00012920567706369758,
+      "loss": 1.1206,
+      "step": 881
+    },
+    {
+      "epoch": 0.4247020585048754,
+      "grad_norm": 0.545403003692627,
+      "learning_rate": 0.00012905645478468192,
+      "loss": 0.3437,
+      "step": 882
+    },
+    {
+      "epoch": 0.4251835801131576,
+      "grad_norm": 0.7194792032241821,
+      "learning_rate": 0.00012890716180502564,
+      "loss": 0.3147,
+      "step": 883
+    },
+    {
+      "epoch": 0.42566510172143973,
+      "grad_norm": 0.661845862865448,
+      "learning_rate": 0.00012875779848799078,
+      "loss": 1.0886,
+      "step": 884
+    },
+    {
+      "epoch": 0.4261466233297219,
+      "grad_norm": 1.2993658781051636,
+      "learning_rate": 0.00012860836519701063,
+      "loss": 0.9139,
+      "step": 885
+    },
+    {
+      "epoch": 0.4266281449380041,
+      "grad_norm": 0.7186025381088257,
+      "learning_rate": 0.00012845886229568873,
+      "loss": 0.5141,
+      "step": 886
+    },
+    {
+      "epoch": 0.42710966654628624,
+      "grad_norm": 0.9918057322502136,
+      "learning_rate": 0.00012830929014779797,
+      "loss": 1.1286,
+      "step": 887
+    },
+    {
+      "epoch": 0.4275911881545684,
+      "grad_norm": 1.213967204093933,
+      "learning_rate": 0.0001281596491172797,
+      "loss": 0.3991,
+      "step": 888
+    },
+    {
+      "epoch": 0.4280727097628506,
+      "grad_norm": 1.1254137754440308,
+      "learning_rate": 0.00012800993956824303,
+      "loss": 1.0619,
+      "step": 889
+    },
+    {
+      "epoch": 0.4285542313711328,
+      "grad_norm": 0.8088738322257996,
+      "learning_rate": 0.00012786016186496358,
+      "loss": 0.4409,
+      "step": 890
+    },
+    {
+      "epoch": 0.42903575297941493,
+      "grad_norm": 0.8131324648857117,
+      "learning_rate": 0.000127710316371883,
+      "loss": 0.4324,
+      "step": 891
+    },
+    {
+      "epoch": 0.4295172745876971,
+      "grad_norm": 0.5402470231056213,
+      "learning_rate": 0.0001275604034536077,
+      "loss": 0.2812,
+      "step": 892
+    },
+    {
+      "epoch": 0.4299987961959793,
+      "grad_norm": 0.4589332938194275,
+      "learning_rate": 0.0001274104234749083,
+      "loss": 0.2145,
+      "step": 893
+    },
+    {
+      "epoch": 0.4304803178042615,
+      "grad_norm": 1.0071089267730713,
+      "learning_rate": 0.00012726037680071853,
+      "loss": 0.6196,
+      "step": 894
+    },
+    {
+      "epoch": 0.4309618394125436,
+      "grad_norm": 0.6019487380981445,
+      "learning_rate": 0.00012711026379613434,
+      "loss": 0.4071,
+      "step": 895
+    },
+    {
+      "epoch": 0.4314433610208258,
+      "grad_norm": 0.4118281602859497,
+      "learning_rate": 0.00012696008482641325,
+      "loss": 0.3439,
+      "step": 896
+    },
+    {
+      "epoch": 0.431924882629108,
+      "grad_norm": 0.6263596415519714,
+      "learning_rate": 0.00012680984025697313,
+      "loss": 0.3853,
+      "step": 897
+    },
+    {
+      "epoch": 0.43240640423739013,
+      "grad_norm": 1.7972133159637451,
+      "learning_rate": 0.00012665953045339152,
+      "loss": 0.6076,
+      "step": 898
+    },
+    {
+      "epoch": 0.4328879258456723,
+      "grad_norm": 0.615397572517395,
+      "learning_rate": 0.0001265091557814047,
+      "loss": 0.6319,
+      "step": 899
+    },
+    {
+      "epoch": 0.4333694474539545,
+      "grad_norm": 1.5493266582489014,
+      "learning_rate": 0.00012635871660690676,
+      "loss": 0.4057,
+      "step": 900
+    },
+    {
+      "epoch": 0.4338509690622367,
+      "grad_norm": 0.7660396099090576,
+      "learning_rate": 0.0001262082132959488,
+      "loss": 0.5197,
+      "step": 901
+    },
+    {
+      "epoch": 0.4343324906705188,
+      "grad_norm": 0.9497624635696411,
+      "learning_rate": 0.00012605764621473792,
+      "loss": 0.6527,
+      "step": 902
+    },
+    {
+      "epoch": 0.434814012278801,
+      "grad_norm": 0.5582836866378784,
+      "learning_rate": 0.00012590701572963642,
+      "loss": 0.5082,
+      "step": 903
+    },
+    {
+      "epoch": 0.4352955338870832,
+      "grad_norm": 1.3667628765106201,
+      "learning_rate": 0.00012575632220716078,
+      "loss": 0.8073,
+      "step": 904
+    },
+    {
+      "epoch": 0.43577705549536533,
+      "grad_norm": 0.6490867733955383,
+      "learning_rate": 0.000125605566013981,
+      "loss": 0.7873,
+      "step": 905
+    },
+    {
+      "epoch": 0.4362585771036475,
+      "grad_norm": 0.7977167963981628,
+      "learning_rate": 0.00012545474751691953,
+      "loss": 0.8709,
+      "step": 906
+    },
+    {
+      "epoch": 0.4367400987119297,
+      "grad_norm": 0.9856142401695251,
+      "learning_rate": 0.00012530386708295036,
+      "loss": 0.9742,
+      "step": 907
+    },
+    {
+      "epoch": 0.4372216203202119,
+      "grad_norm": 0.665690541267395,
+      "learning_rate": 0.00012515292507919829,
+      "loss": 0.4871,
+      "step": 908
+    },
+    {
+      "epoch": 0.437703141928494,
+      "grad_norm": 0.4781898558139801,
+      "learning_rate": 0.0001250019218729378,
+      "loss": 0.4098,
+      "step": 909
+    },
+    {
+      "epoch": 0.4381846635367762,
+      "grad_norm": 0.7730495929718018,
+      "learning_rate": 0.00012485085783159238,
+      "loss": 0.2263,
+      "step": 910
+    },
+    {
+      "epoch": 0.4386661851450584,
+      "grad_norm": 0.9093911647796631,
+      "learning_rate": 0.00012469973332273354,
+      "loss": 0.4428,
+      "step": 911
+    },
+    {
+      "epoch": 0.4391477067533406,
+      "grad_norm": 0.7920442819595337,
+      "learning_rate": 0.00012454854871407994,
+      "loss": 0.4661,
+      "step": 912
+    },
+    {
+      "epoch": 0.4396292283616227,
+      "grad_norm": 0.8402988910675049,
+      "learning_rate": 0.00012439730437349635,
+      "loss": 0.4356,
+      "step": 913
+    },
+    {
+      "epoch": 0.4401107499699049,
+      "grad_norm": 0.9629238247871399,
+      "learning_rate": 0.00012424600066899302,
+      "loss": 0.6917,
+      "step": 914
+    },
+    {
+      "epoch": 0.4405922715781871,
+      "grad_norm": 1.3128741979599,
+      "learning_rate": 0.00012409463796872464,
+      "loss": 0.6547,
+      "step": 915
+    },
+    {
+      "epoch": 0.4410737931864692,
+      "grad_norm": 0.8441032767295837,
+      "learning_rate": 0.0001239432166409893,
+      "loss": 0.8729,
+      "step": 916
+    },
+    {
+      "epoch": 0.4415553147947514,
+      "grad_norm": 0.6667619347572327,
+      "learning_rate": 0.00012379173705422795,
+      "loss": 0.5423,
+      "step": 917
+    },
+    {
+      "epoch": 0.4420368364030336,
+      "grad_norm": 0.8423894643783569,
+      "learning_rate": 0.00012364019957702315,
+      "loss": 0.7282,
+      "step": 918
+    },
+    {
+      "epoch": 0.4425183580113158,
+      "grad_norm": 2.0424094200134277,
+      "learning_rate": 0.00012348860457809838,
+      "loss": 1.1008,
+      "step": 919
+    },
+    {
+      "epoch": 0.4429998796195979,
+      "grad_norm": 0.6044926047325134,
+      "learning_rate": 0.00012333695242631705,
+      "loss": 0.8077,
+      "step": 920
+    },
+    {
+      "epoch": 0.4434814012278801,
+      "grad_norm": 0.8025707602500916,
+      "learning_rate": 0.0001231852434906817,
+      "loss": 0.7705,
+      "step": 921
+    },
+    {
+      "epoch": 0.4439629228361623,
+      "grad_norm": 0.8117222785949707,
+      "learning_rate": 0.00012303347814033292,
+      "loss": 0.4655,
+      "step": 922
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.715910017490387,
+      "learning_rate": 0.0001228816567445487,
+      "loss": 0.6408,
+      "step": 923
+    },
+    {
+      "epoch": 0.4449259660527266,
+      "grad_norm": 0.7318637371063232,
+      "learning_rate": 0.0001227297796727433,
+      "loss": 0.4922,
+      "step": 924
+    },
+    {
+      "epoch": 0.4454074876610088,
+      "grad_norm": 0.8235260844230652,
+      "learning_rate": 0.00012257784729446656,
+      "loss": 0.7587,
+      "step": 925
+    },
+    {
+      "epoch": 0.445889009269291,
+      "grad_norm": 0.7627338171005249,
+      "learning_rate": 0.00012242585997940275,
+      "loss": 0.391,
+      "step": 926
+    },
+    {
+      "epoch": 0.4463705308775731,
+      "grad_norm": 0.5861409306526184,
+      "learning_rate": 0.0001222738180973699,
+      "loss": 0.5866,
+      "step": 927
+    },
+    {
+      "epoch": 0.4468520524858553,
+      "grad_norm": 0.6915172934532166,
+      "learning_rate": 0.00012212172201831885,
+      "loss": 0.3472,
+      "step": 928
+    },
+    {
+      "epoch": 0.4473335740941375,
+      "grad_norm": 0.6583700776100159,
+      "learning_rate": 0.00012196957211233222,
+      "loss": 0.5164,
+      "step": 929
+    },
+    {
+      "epoch": 0.4478150957024196,
+      "grad_norm": 0.8904280662536621,
+      "learning_rate": 0.00012181736874962371,
+      "loss": 1.3786,
+      "step": 930
+    },
+    {
+      "epoch": 0.4482966173107018,
+      "grad_norm": 0.9074341058731079,
+      "learning_rate": 0.00012166511230053696,
+      "loss": 0.5478,
+      "step": 931
+    },
+    {
+      "epoch": 0.448778138918984,
+      "grad_norm": 0.6648260354995728,
+      "learning_rate": 0.00012151280313554486,
+      "loss": 1.1953,
+      "step": 932
+    },
+    {
+      "epoch": 0.4492596605272662,
+      "grad_norm": 1.5189319849014282,
+      "learning_rate": 0.00012136044162524858,
+      "loss": 1.042,
+      "step": 933
+    },
+    {
+      "epoch": 0.4497411821355483,
+      "grad_norm": 0.5307234525680542,
+      "learning_rate": 0.00012120802814037663,
+      "loss": 0.3387,
+      "step": 934
+    },
+    {
+      "epoch": 0.4502227037438305,
+      "grad_norm": 0.6986709237098694,
+      "learning_rate": 0.00012105556305178399,
+      "loss": 0.3857,
+      "step": 935
+    },
+    {
+      "epoch": 0.4507042253521127,
+      "grad_norm": 0.5282310843467712,
+      "learning_rate": 0.00012090304673045123,
+      "loss": 0.3448,
+      "step": 936
+    },
+    {
+      "epoch": 0.4511857469603949,
+      "grad_norm": 1.0674173831939697,
+      "learning_rate": 0.00012075047954748353,
+      "loss": 0.9065,
+      "step": 937
+    },
+    {
+      "epoch": 0.451667268568677,
+      "grad_norm": 1.6909977197647095,
+      "learning_rate": 0.00012059786187410984,
+      "loss": 0.7831,
+      "step": 938
+    },
+    {
+      "epoch": 0.4521487901769592,
+      "grad_norm": 1.0591840744018555,
+      "learning_rate": 0.000120445194081682,
+      "loss": 0.5693,
+      "step": 939
+    },
+    {
+      "epoch": 0.4526303117852414,
+      "grad_norm": 0.5961440801620483,
+      "learning_rate": 0.00012029247654167379,
+      "loss": 0.2809,
+      "step": 940
+    },
+    {
+      "epoch": 0.4531118333935235,
+      "grad_norm": 0.8247968554496765,
+      "learning_rate": 0.00012013970962568002,
+      "loss": 0.4512,
+      "step": 941
+    },
+    {
+      "epoch": 0.4535933550018057,
+      "grad_norm": 1.0450423955917358,
+      "learning_rate": 0.00011998689370541562,
+      "loss": 0.8942,
+      "step": 942
+    },
+    {
+      "epoch": 0.4540748766100879,
+      "grad_norm": 1.201209545135498,
+      "learning_rate": 0.00011983402915271478,
+      "loss": 0.4332,
+      "step": 943
+    },
+    {
+      "epoch": 0.4545563982183701,
+      "grad_norm": 0.8509264588356018,
+      "learning_rate": 0.00011968111633953007,
+      "loss": 0.3682,
+      "step": 944
+    },
+    {
+      "epoch": 0.4550379198266522,
+      "grad_norm": 0.7999116778373718,
+      "learning_rate": 0.0001195281556379314,
+      "loss": 0.5891,
+      "step": 945
+    },
+    {
+      "epoch": 0.4555194414349344,
+      "grad_norm": 0.8322030305862427,
+      "learning_rate": 0.0001193751474201053,
+      "loss": 0.2077,
+      "step": 946
+    },
+    {
+      "epoch": 0.4560009630432166,
+      "grad_norm": 1.6226719617843628,
+      "learning_rate": 0.00011922209205835382,
+      "loss": 0.2934,
+      "step": 947
+    },
+    {
+      "epoch": 0.4564824846514987,
+      "grad_norm": 1.2562118768692017,
+      "learning_rate": 0.0001190689899250938,
+      "loss": 0.7436,
+      "step": 948
+    },
+    {
+      "epoch": 0.4569640062597809,
+      "grad_norm": 0.7772266864776611,
+      "learning_rate": 0.00011891584139285582,
+      "loss": 0.8018,
+      "step": 949
+    },
+    {
+      "epoch": 0.4574455278680631,
+      "grad_norm": 0.8317371606826782,
+      "learning_rate": 0.00011876264683428344,
+      "loss": 0.4488,
+      "step": 950
+    },
+    {
+      "epoch": 0.45792704947634527,
+      "grad_norm": 0.3852919042110443,
+      "learning_rate": 0.00011860940662213211,
+      "loss": 0.1071,
+      "step": 951
+    },
+    {
+      "epoch": 0.4584085710846274,
+      "grad_norm": 1.2257070541381836,
+      "learning_rate": 0.00011845612112926843,
+      "loss": 0.8347,
+      "step": 952
+    },
+    {
+      "epoch": 0.4588900926929096,
+      "grad_norm": 0.7787636518478394,
+      "learning_rate": 0.00011830279072866921,
+      "loss": 0.6742,
+      "step": 953
+    },
+    {
+      "epoch": 0.4593716143011918,
+      "grad_norm": 1.6153544187545776,
+      "learning_rate": 0.00011814941579342044,
+      "loss": 0.432,
+      "step": 954
+    },
+    {
+      "epoch": 0.45985313590947396,
+      "grad_norm": 0.6438875198364258,
+      "learning_rate": 0.00011799599669671654,
+      "loss": 0.2599,
+      "step": 955
+    },
+    {
+      "epoch": 0.4603346575177561,
+      "grad_norm": 1.2926790714263916,
+      "learning_rate": 0.00011784253381185937,
+      "loss": 0.811,
+      "step": 956
+    },
+    {
+      "epoch": 0.4608161791260383,
+      "grad_norm": 0.5644267797470093,
+      "learning_rate": 0.0001176890275122573,
+      "loss": 0.4718,
+      "step": 957
+    },
+    {
+      "epoch": 0.46129770073432047,
+      "grad_norm": 0.6094907522201538,
+      "learning_rate": 0.0001175354781714244,
+      "loss": 0.216,
+      "step": 958
+    },
+    {
+      "epoch": 0.4617792223426026,
+      "grad_norm": 0.8137814402580261,
+      "learning_rate": 0.0001173818861629794,
+      "loss": 0.9217,
+      "step": 959
+    },
+    {
+      "epoch": 0.4622607439508848,
+      "grad_norm": 0.5632073879241943,
+      "learning_rate": 0.00011722825186064494,
+      "loss": 0.4477,
+      "step": 960
+    },
+    {
+      "epoch": 0.462742265559167,
+      "grad_norm": 0.6808025240898132,
+      "learning_rate": 0.00011707457563824646,
+      "loss": 0.3814,
+      "step": 961
+    },
+    {
+      "epoch": 0.46322378716744916,
+      "grad_norm": 0.7273874878883362,
+      "learning_rate": 0.00011692085786971149,
+      "loss": 0.3307,
+      "step": 962
+    },
+    {
+      "epoch": 0.4637053087757313,
+      "grad_norm": 1.1415742635726929,
+      "learning_rate": 0.00011676709892906858,
+      "loss": 0.3101,
+      "step": 963
+    },
+    {
+      "epoch": 0.4641868303840135,
+      "grad_norm": 0.8693737983703613,
+      "learning_rate": 0.00011661329919044656,
+      "loss": 0.7078,
+      "step": 964
+    },
+    {
+      "epoch": 0.46466835199229567,
+      "grad_norm": 0.4996722638607025,
+      "learning_rate": 0.00011645945902807341,
+      "loss": 0.1454,
+      "step": 965
+    },
+    {
+      "epoch": 0.4651498736005778,
+      "grad_norm": 0.9881248474121094,
+      "learning_rate": 0.00011630557881627553,
+      "loss": 0.7298,
+      "step": 966
+    },
+    {
+      "epoch": 0.46563139520886,
+      "grad_norm": 0.7234969139099121,
+      "learning_rate": 0.0001161516589294768,
+      "loss": 0.7709,
+      "step": 967
+    },
+    {
+      "epoch": 0.4661129168171422,
+      "grad_norm": 1.0357717275619507,
+      "learning_rate": 0.00011599769974219757,
+      "loss": 0.5054,
+      "step": 968
+    },
+    {
+      "epoch": 0.46659443842542436,
+      "grad_norm": 1.1411643028259277,
+      "learning_rate": 0.0001158437016290539,
+      "loss": 1.0258,
+      "step": 969
+    },
+    {
+      "epoch": 0.4670759600337065,
+      "grad_norm": 0.5702272653579712,
+      "learning_rate": 0.00011568966496475649,
+      "loss": 0.3137,
+      "step": 970
+    },
+    {
+      "epoch": 0.4675574816419887,
+      "grad_norm": 0.838179886341095,
+      "learning_rate": 0.00011553559012410984,
+      "loss": 0.4455,
+      "step": 971
+    },
+    {
+      "epoch": 0.46803900325027087,
+      "grad_norm": 0.9521105885505676,
+      "learning_rate": 0.00011538147748201138,
+      "loss": 0.3627,
+      "step": 972
+    },
+    {
+      "epoch": 0.468520524858553,
+      "grad_norm": 0.6624534130096436,
+      "learning_rate": 0.00011522732741345053,
+      "loss": 0.6825,
+      "step": 973
+    },
+    {
+      "epoch": 0.4690020464668352,
+      "grad_norm": 0.5236545205116272,
+      "learning_rate": 0.00011507314029350776,
+      "loss": 0.3712,
+      "step": 974
+    },
+    {
+      "epoch": 0.4694835680751174,
+      "grad_norm": 0.49228930473327637,
+      "learning_rate": 0.00011491891649735366,
+      "loss": 0.5079,
+      "step": 975
+    },
+    {
+      "epoch": 0.46996508968339956,
+      "grad_norm": 1.3148273229599,
+      "learning_rate": 0.00011476465640024814,
+      "loss": 1.0293,
+      "step": 976
+    },
+    {
+      "epoch": 0.4704466112916817,
+      "grad_norm": 0.534582257270813,
+      "learning_rate": 0.00011461036037753934,
+      "loss": 0.2893,
+      "step": 977
+    },
+    {
+      "epoch": 0.4709281328999639,
+      "grad_norm": 0.7397232055664062,
+      "learning_rate": 0.00011445602880466288,
+      "loss": 0.2611,
+      "step": 978
+    },
+    {
+      "epoch": 0.47140965450824607,
+      "grad_norm": 0.5304082036018372,
+      "learning_rate": 0.00011430166205714088,
+      "loss": 0.7689,
+      "step": 979
+    },
+    {
+      "epoch": 0.47189117611652825,
+      "grad_norm": 0.6593488454818726,
+      "learning_rate": 0.00011414726051058102,
+      "loss": 0.4602,
+      "step": 980
+    },
+    {
+      "epoch": 0.4723726977248104,
+      "grad_norm": 1.089796543121338,
+      "learning_rate": 0.0001139928245406757,
+      "loss": 0.4331,
+      "step": 981
+    },
+    {
+      "epoch": 0.4728542193330926,
+      "grad_norm": 0.6848332285881042,
+      "learning_rate": 0.00011383835452320097,
+      "loss": 0.4966,
+      "step": 982
+    },
+    {
+      "epoch": 0.47333574094137476,
+      "grad_norm": 0.784115195274353,
+      "learning_rate": 0.00011368385083401585,
+      "loss": 1.1322,
+      "step": 983
+    },
+    {
+      "epoch": 0.4738172625496569,
+      "grad_norm": 1.0432567596435547,
+      "learning_rate": 0.00011352931384906125,
+      "loss": 0.578,
+      "step": 984
+    },
+    {
+      "epoch": 0.4742987841579391,
+      "grad_norm": 0.6884013414382935,
+      "learning_rate": 0.00011337474394435908,
+      "loss": 0.8005,
+      "step": 985
+    },
+    {
+      "epoch": 0.47478030576622127,
+      "grad_norm": 0.7809309959411621,
+      "learning_rate": 0.00011322014149601136,
+      "loss": 0.656,
+      "step": 986
+    },
+    {
+      "epoch": 0.47526182737450345,
+      "grad_norm": 0.5823318958282471,
+      "learning_rate": 0.00011306550688019926,
+      "loss": 0.6732,
+      "step": 987
+    },
+    {
+      "epoch": 0.4757433489827856,
+      "grad_norm": 0.6410465240478516,
+      "learning_rate": 0.0001129108404731823,
+      "loss": 0.6143,
+      "step": 988
+    },
+    {
+      "epoch": 0.4762248705910678,
+      "grad_norm": 1.332196593284607,
+      "learning_rate": 0.0001127561426512973,
+      "loss": 1.0384,
+      "step": 989
+    },
+    {
+      "epoch": 0.47670639219934996,
+      "grad_norm": 0.7239651679992676,
+      "learning_rate": 0.0001126014137909575,
+      "loss": 0.378,
+      "step": 990
+    },
+    {
+      "epoch": 0.4771879138076321,
+      "grad_norm": 0.5169118046760559,
+      "learning_rate": 0.00011244665426865174,
+      "loss": 0.5249,
+      "step": 991
+    },
+    {
+      "epoch": 0.4776694354159143,
+      "grad_norm": 1.1747828722000122,
+      "learning_rate": 0.00011229186446094338,
+      "loss": 0.5218,
+      "step": 992
+    },
+    {
+      "epoch": 0.47815095702419647,
+      "grad_norm": 0.5271717309951782,
+      "learning_rate": 0.00011213704474446951,
+      "loss": 0.6091,
+      "step": 993
+    },
+    {
+      "epoch": 0.47863247863247865,
+      "grad_norm": 1.4207522869110107,
+      "learning_rate": 0.00011198219549594,
+      "loss": 0.8363,
+      "step": 994
+    },
+    {
+      "epoch": 0.4791140002407608,
+      "grad_norm": 0.704620897769928,
+      "learning_rate": 0.00011182731709213659,
+      "loss": 0.3071,
+      "step": 995
+    },
+    {
+      "epoch": 0.47959552184904297,
+      "grad_norm": 0.775363564491272,
+      "learning_rate": 0.00011167240990991192,
+      "loss": 0.248,
+      "step": 996
+    },
+    {
+      "epoch": 0.48007704345732516,
+      "grad_norm": 0.6383013725280762,
+      "learning_rate": 0.00011151747432618871,
+      "loss": 0.6298,
+      "step": 997
+    },
+    {
+      "epoch": 0.48055856506560735,
+      "grad_norm": 0.826794445514679,
+      "learning_rate": 0.00011136251071795871,
+      "loss": 0.6077,
+      "step": 998
+    },
+    {
+      "epoch": 0.4810400866738895,
+      "grad_norm": 1.2073599100112915,
+      "learning_rate": 0.00011120751946228197,
+      "loss": 0.7025,
+      "step": 999
+    },
+    {
+      "epoch": 0.48152160828217166,
+      "grad_norm": 0.6839677691459656,
+      "learning_rate": 0.00011105250093628565,
+      "loss": 0.608,
+      "step": 1000
+    },
+    {
+      "epoch": 0.48200312989045385,
+      "grad_norm": 0.6721844673156738,
+      "learning_rate": 0.00011089745551716344,
+      "loss": 0.6139,
+      "step": 1001
+    },
+    {
+      "epoch": 0.482484651498736,
+      "grad_norm": 1.1474583148956299,
+      "learning_rate": 0.00011074238358217437,
+      "loss": 1.0103,
+      "step": 1002
+    },
+    {
+      "epoch": 0.48296617310701817,
+      "grad_norm": 0.670769989490509,
+      "learning_rate": 0.00011058728550864197,
+      "loss": 0.2139,
+      "step": 1003
+    },
+    {
+      "epoch": 0.48344769471530036,
+      "grad_norm": 0.5323492884635925,
+      "learning_rate": 0.00011043216167395344,
+      "loss": 0.3595,
+      "step": 1004
+    },
+    {
+      "epoch": 0.48392921632358255,
+      "grad_norm": 0.8688573837280273,
+      "learning_rate": 0.00011027701245555865,
+      "loss": 0.6961,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4844107379318647,
+      "grad_norm": 1.572704553604126,
+      "learning_rate": 0.00011012183823096917,
+      "loss": 0.2322,
+      "step": 1006
+    },
+    {
+      "epoch": 0.48489225954014686,
+      "grad_norm": 0.6855084896087646,
+      "learning_rate": 0.00010996663937775751,
+      "loss": 0.2161,
+      "step": 1007
+    },
+    {
+      "epoch": 0.48537378114842905,
+      "grad_norm": 0.722327470779419,
+      "learning_rate": 0.000109811416273556,
+      "loss": 1.2421,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4858553027567112,
+      "grad_norm": 0.6337584257125854,
+      "learning_rate": 0.00010965616929605609,
+      "loss": 0.8193,
+      "step": 1009
+    },
+    {
+      "epoch": 0.48633682436499337,
+      "grad_norm": 0.7821305394172668,
+      "learning_rate": 0.0001095008988230072,
+      "loss": 0.4506,
+      "step": 1010
+    },
+    {
+      "epoch": 0.48681834597327556,
+      "grad_norm": 1.3698606491088867,
+      "learning_rate": 0.00010934560523221602,
+      "loss": 0.7479,
+      "step": 1011
+    },
+    {
+      "epoch": 0.48729986758155774,
+      "grad_norm": 0.9158796072006226,
+      "learning_rate": 0.00010919028890154543,
+      "loss": 0.5681,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4877813891898399,
+      "grad_norm": 0.808804452419281,
+      "learning_rate": 0.00010903495020891375,
+      "loss": 0.7309,
+      "step": 1013
+    },
+    {
+      "epoch": 0.48826291079812206,
+      "grad_norm": 0.6452478170394897,
+      "learning_rate": 0.00010887958953229349,
+      "loss": 1.1787,
+      "step": 1014
+    },
+    {
+      "epoch": 0.48874443240640425,
+      "grad_norm": 0.7447563409805298,
+      "learning_rate": 0.00010872420724971088,
+      "loss": 0.5845,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4892259540146864,
+      "grad_norm": 0.6775797009468079,
+      "learning_rate": 0.0001085688037392446,
+      "loss": 0.3645,
+      "step": 1016
+    },
+    {
+      "epoch": 0.48970747562296857,
+      "grad_norm": 0.8377703428268433,
+      "learning_rate": 0.000108413379379025,
+      "loss": 0.4498,
+      "step": 1017
+    },
+    {
+      "epoch": 0.49018899723125076,
+      "grad_norm": 0.3953412175178528,
+      "learning_rate": 0.00010825793454723325,
+      "loss": 0.4478,
+      "step": 1018
+    },
+    {
+      "epoch": 0.49067051883953294,
+      "grad_norm": 0.7678600549697876,
+      "learning_rate": 0.00010810246962210018,
+      "loss": 0.978,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4911520404478151,
+      "grad_norm": 1.129772424697876,
+      "learning_rate": 0.00010794698498190557,
+      "loss": 0.5298,
+      "step": 1020
+    },
+    {
+      "epoch": 0.49163356205609726,
+      "grad_norm": 0.7634359002113342,
+      "learning_rate": 0.00010779148100497722,
+      "loss": 0.5905,
+      "step": 1021
+    },
+    {
+      "epoch": 0.49211508366437945,
+      "grad_norm": 0.6913365125656128,
+      "learning_rate": 0.00010763595806968996,
+      "loss": 0.9734,
+      "step": 1022
+    },
+    {
+      "epoch": 0.49259660527266164,
+      "grad_norm": 1.0433826446533203,
+      "learning_rate": 0.00010748041655446473,
+      "loss": 0.9034,
+      "step": 1023
+    },
+    {
+      "epoch": 0.49307812688094377,
+      "grad_norm": 0.9241974949836731,
+      "learning_rate": 0.00010732485683776768,
+      "loss": 0.9534,
+      "step": 1024
+    },
+    {
+      "epoch": 0.49355964848922596,
+      "grad_norm": 0.8278858661651611,
+      "learning_rate": 0.00010716927929810925,
+      "loss": 0.6806,
+      "step": 1025
+    },
+    {
+      "epoch": 0.49404117009750814,
+      "grad_norm": 1.0138304233551025,
+      "learning_rate": 0.00010701368431404326,
+      "loss": 0.5995,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4945226917057903,
+      "grad_norm": 0.5914368629455566,
+      "learning_rate": 0.00010685807226416598,
+      "loss": 0.3246,
+      "step": 1027
+    },
+    {
+      "epoch": 0.49500421331407246,
+      "grad_norm": 0.81244295835495,
+      "learning_rate": 0.00010670244352711518,
+      "loss": 0.4668,
+      "step": 1028
+    },
+    {
+      "epoch": 0.49548573492235465,
+      "grad_norm": 0.7371711134910583,
+      "learning_rate": 0.00010654679848156925,
+      "loss": 0.3986,
+      "step": 1029
+    },
+    {
+      "epoch": 0.49596725653063684,
+      "grad_norm": 0.7539998292922974,
+      "learning_rate": 0.00010639113750624625,
+      "loss": 0.3003,
+      "step": 1030
+    },
+    {
+      "epoch": 0.49644877813891897,
+      "grad_norm": 0.879912793636322,
+      "learning_rate": 0.00010623546097990303,
+      "loss": 0.844,
+      "step": 1031
+    },
+    {
+      "epoch": 0.49693029974720115,
+      "grad_norm": 0.8178533315658569,
+      "learning_rate": 0.00010607976928133423,
+      "loss": 0.253,
+      "step": 1032
+    },
+    {
+      "epoch": 0.49741182135548334,
+      "grad_norm": 0.8457581400871277,
+      "learning_rate": 0.00010592406278937144,
+      "loss": 0.4384,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4978933429637655,
+      "grad_norm": 1.146519660949707,
+      "learning_rate": 0.00010576834188288226,
+      "loss": 0.4484,
+      "step": 1034
+    },
+    {
+      "epoch": 0.49837486457204766,
+      "grad_norm": 0.8387516736984253,
+      "learning_rate": 0.00010561260694076935,
+      "loss": 0.6428,
+      "step": 1035
+    },
+    {
+      "epoch": 0.49885638618032985,
+      "grad_norm": 0.5453752279281616,
+      "learning_rate": 0.00010545685834196948,
+      "loss": 0.4481,
+      "step": 1036
+    },
+    {
+      "epoch": 0.49933790778861203,
+      "grad_norm": 1.3885741233825684,
+      "learning_rate": 0.00010530109646545272,
+      "loss": 0.5443,
+      "step": 1037
+    },
+    {
+      "epoch": 0.49981942939689417,
+      "grad_norm": 0.8055309653282166,
+      "learning_rate": 0.0001051453216902214,
+      "loss": 0.5387,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5003009510051764,
+      "grad_norm": 0.8343227505683899,
+      "learning_rate": 0.00010498953439530925,
+      "loss": 0.7728,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5007824726134585,
+      "grad_norm": 0.5583564639091492,
+      "learning_rate": 0.00010483373495978046,
+      "loss": 0.4572,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5012639942217407,
+      "grad_norm": 1.2023993730545044,
+      "learning_rate": 0.00010467792376272877,
+      "loss": 0.8647,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5017455158300229,
+      "grad_norm": 0.6801705360412598,
+      "learning_rate": 0.00010452210118327652,
+      "loss": 0.4475,
+      "step": 1042
+    },
+    {
+      "epoch": 0.502227037438305,
+      "grad_norm": 1.6109269857406616,
+      "learning_rate": 0.00010436626760057378,
+      "loss": 0.6505,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5027085590465872,
+      "grad_norm": 0.7384167313575745,
+      "learning_rate": 0.00010421042339379732,
+      "loss": 0.2938,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5031900806548694,
+      "grad_norm": 0.7244082689285278,
+      "learning_rate": 0.00010405456894214987,
+      "loss": 0.1328,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5036716022631516,
+      "grad_norm": 1.1768090724945068,
+      "learning_rate": 0.00010389870462485902,
+      "loss": 1.5518,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5041531238714337,
+      "grad_norm": 0.6785743832588196,
+      "learning_rate": 0.00010374283082117635,
+      "loss": 0.2975,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5046346454797159,
+      "grad_norm": 0.9195048213005066,
+      "learning_rate": 0.00010358694791037653,
+      "loss": 0.6257,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5051161670879981,
+      "grad_norm": 1.71181058883667,
+      "learning_rate": 0.00010343105627175644,
+      "loss": 0.6551,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5055976886962803,
+      "grad_norm": 0.998651385307312,
+      "learning_rate": 0.00010327515628463415,
+      "loss": 0.823,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5060792103045624,
+      "grad_norm": 0.828630805015564,
+      "learning_rate": 0.00010311924832834808,
+      "loss": 0.9087,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5065607319128446,
+      "grad_norm": 1.2712898254394531,
+      "learning_rate": 0.00010296333278225599,
+      "loss": 0.4819,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5070422535211268,
+      "grad_norm": 1.1045798063278198,
+      "learning_rate": 0.00010280741002573413,
+      "loss": 0.2973,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5075237751294089,
+      "grad_norm": 0.6555344462394714,
+      "learning_rate": 0.00010265148043817632,
+      "loss": 0.4144,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5080052967376911,
+      "grad_norm": 0.92887943983078,
+      "learning_rate": 0.00010249554439899298,
+      "loss": 0.3468,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5084868183459733,
+      "grad_norm": 0.6355475187301636,
+      "learning_rate": 0.00010233960228761022,
+      "loss": 0.5131,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5089683399542555,
+      "grad_norm": 0.5725042223930359,
+      "learning_rate": 0.00010218365448346893,
+      "loss": 0.5927,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5094498615625376,
+      "grad_norm": 1.2848678827285767,
+      "learning_rate": 0.00010202770136602388,
+      "loss": 0.7266,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5099313831708198,
+      "grad_norm": 0.6311907768249512,
+      "learning_rate": 0.00010187174331474271,
+      "loss": 0.2647,
+      "step": 1059
+    },
+    {
+      "epoch": 0.510412904779102,
+      "grad_norm": 0.8230004906654358,
+      "learning_rate": 0.00010171578070910512,
+      "loss": 0.4257,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5108944263873841,
+      "grad_norm": 0.9461697340011597,
+      "learning_rate": 0.00010155981392860185,
+      "loss": 0.3434,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5113759479956663,
+      "grad_norm": 0.7990442514419556,
+      "learning_rate": 0.00010140384335273386,
+      "loss": 0.6928,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5118574696039485,
+      "grad_norm": 0.9711220860481262,
+      "learning_rate": 0.00010124786936101127,
+      "loss": 0.4676,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5123389912122307,
+      "grad_norm": 0.7467138171195984,
+      "learning_rate": 0.00010109189233295255,
+      "loss": 0.7543,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5128205128205128,
+      "grad_norm": 1.2974790334701538,
+      "learning_rate": 0.00010093591264808358,
+      "loss": 0.6458,
+      "step": 1065
+    },
+    {
+      "epoch": 0.513302034428795,
+      "grad_norm": 0.8935269117355347,
+      "learning_rate": 0.00010077993068593663,
+      "loss": 0.581,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5137835560370771,
+      "grad_norm": 1.0466784238815308,
+      "learning_rate": 0.00010062394682604963,
+      "loss": 0.5093,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5142650776453593,
+      "grad_norm": 0.6737858057022095,
+      "learning_rate": 0.00010046796144796497,
+      "loss": 0.5222,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5147465992536415,
+      "grad_norm": 0.957423746585846,
+      "learning_rate": 0.0001003119749312289,
+      "loss": 0.506,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5152281208619237,
+      "grad_norm": 0.963703989982605,
+      "learning_rate": 0.00010015598765539031,
+      "loss": 0.6979,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5157096424702059,
+      "grad_norm": 0.8813979029655457,
+      "learning_rate": 0.0001,
+      "loss": 0.5174,
+      "step": 1071
+    },
+    {
+      "epoch": 0.516191164078488,
+      "grad_norm": 0.6864508390426636,
+      "learning_rate": 9.984401234460971e-05,
+      "loss": 0.5415,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5166726856867702,
+      "grad_norm": 1.0436968803405762,
+      "learning_rate": 9.968802506877111e-05,
+      "loss": 1.0143,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5171542072950523,
+      "grad_norm": 0.4878181517124176,
+      "learning_rate": 9.953203855203504e-05,
+      "loss": 0.266,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5176357289033345,
+      "grad_norm": 0.9048622846603394,
+      "learning_rate": 9.93760531739504e-05,
+      "loss": 0.7538,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5181172505116167,
+      "grad_norm": 1.1710916757583618,
+      "learning_rate": 9.922006931406338e-05,
+      "loss": 0.8256,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5185987721198989,
+      "grad_norm": 0.8252659440040588,
+      "learning_rate": 9.906408735191643e-05,
+      "loss": 0.2878,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5190802937281811,
+      "grad_norm": 0.8352766633033752,
+      "learning_rate": 9.890810766704745e-05,
+      "loss": 0.6413,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5195618153364632,
+      "grad_norm": 1.7258776426315308,
+      "learning_rate": 9.875213063898875e-05,
+      "loss": 0.6046,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5200433369447454,
+      "grad_norm": 0.6900729537010193,
+      "learning_rate": 9.859615664726615e-05,
+      "loss": 0.6857,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5205248585530275,
+      "grad_norm": 1.3947724103927612,
+      "learning_rate": 9.844018607139818e-05,
+      "loss": 0.6629,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5210063801613097,
+      "grad_norm": 0.6548985242843628,
+      "learning_rate": 9.828421929089493e-05,
+      "loss": 0.2404,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5214879017695919,
+      "grad_norm": 1.1727840900421143,
+      "learning_rate": 9.812825668525733e-05,
+      "loss": 0.668,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5219694233778741,
+      "grad_norm": 0.7604357004165649,
+      "learning_rate": 9.797229863397615e-05,
+      "loss": 0.5043,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5224509449861563,
+      "grad_norm": 0.792493999004364,
+      "learning_rate": 9.781634551653108e-05,
+      "loss": 0.4905,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5229324665944384,
+      "grad_norm": 0.8167349696159363,
+      "learning_rate": 9.766039771238982e-05,
+      "loss": 0.9703,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5234139882027206,
+      "grad_norm": 0.9371551275253296,
+      "learning_rate": 9.750445560100706e-05,
+      "loss": 0.7786,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5238955098110027,
+      "grad_norm": 1.1863070726394653,
+      "learning_rate": 9.73485195618237e-05,
+      "loss": 1.097,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5243770314192849,
+      "grad_norm": 0.9151334166526794,
+      "learning_rate": 9.719258997426588e-05,
+      "loss": 0.6039,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5248585530275671,
+      "grad_norm": 0.6455950140953064,
+      "learning_rate": 9.703666721774402e-05,
+      "loss": 0.2652,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5253400746358493,
+      "grad_norm": 0.9584352374076843,
+      "learning_rate": 9.688075167165194e-05,
+      "loss": 0.5744,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5258215962441315,
+      "grad_norm": 0.9051257967948914,
+      "learning_rate": 9.672484371536586e-05,
+      "loss": 0.4059,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5263031178524137,
+      "grad_norm": 0.840788722038269,
+      "learning_rate": 9.656894372824358e-05,
+      "loss": 1.0904,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5267846394606958,
+      "grad_norm": 0.787898063659668,
+      "learning_rate": 9.64130520896235e-05,
+      "loss": 0.4965,
+      "step": 1094
+    },
+    {
+      "epoch": 0.527266161068978,
+      "grad_norm": 0.5363286733627319,
+      "learning_rate": 9.625716917882367e-05,
+      "loss": 0.4016,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5277476826772601,
+      "grad_norm": 0.6209604740142822,
+      "learning_rate": 9.6101295375141e-05,
+      "loss": 0.9046,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5282292042855423,
+      "grad_norm": 0.7012305855751038,
+      "learning_rate": 9.594543105785013e-05,
+      "loss": 0.6374,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5287107258938245,
+      "grad_norm": 0.6587700247764587,
+      "learning_rate": 9.578957660620267e-05,
+      "loss": 0.3757,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5291922475021067,
+      "grad_norm": 1.1536179780960083,
+      "learning_rate": 9.563373239942623e-05,
+      "loss": 0.3881,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5296737691103889,
+      "grad_norm": 0.8401791453361511,
+      "learning_rate": 9.547789881672348e-05,
+      "loss": 0.5374,
+      "step": 1100
+    },
+    {
+      "epoch": 0.530155290718671,
+      "grad_norm": 1.215343713760376,
+      "learning_rate": 9.532207623727126e-05,
+      "loss": 0.9501,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5306368123269531,
+      "grad_norm": 0.7829989194869995,
+      "learning_rate": 9.516626504021957e-05,
+      "loss": 0.5429,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5311183339352353,
+      "grad_norm": 1.4323217868804932,
+      "learning_rate": 9.501046560469079e-05,
+      "loss": 0.7029,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5315998555435175,
+      "grad_norm": 0.8599506616592407,
+      "learning_rate": 9.485467830977864e-05,
+      "loss": 1.1013,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5320813771517997,
+      "grad_norm": 1.4327293634414673,
+      "learning_rate": 9.469890353454732e-05,
+      "loss": 0.2591,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5325628987600819,
+      "grad_norm": 0.6792353987693787,
+      "learning_rate": 9.454314165803054e-05,
+      "loss": 0.2762,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5330444203683641,
+      "grad_norm": 0.7597439885139465,
+      "learning_rate": 9.438739305923067e-05,
+      "loss": 0.5865,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5335259419766462,
+      "grad_norm": 1.0125542879104614,
+      "learning_rate": 9.423165811711777e-05,
+      "loss": 0.9981,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5340074635849283,
+      "grad_norm": 1.0826612710952759,
+      "learning_rate": 9.407593721062859e-05,
+      "loss": 0.4732,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5344889851932105,
+      "grad_norm": 0.7020871639251709,
+      "learning_rate": 9.39202307186658e-05,
+      "loss": 0.3723,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5349705068014927,
+      "grad_norm": 0.6123396754264832,
+      "learning_rate": 9.3764539020097e-05,
+      "loss": 0.5444,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5354520284097749,
+      "grad_norm": 0.9989874958992004,
+      "learning_rate": 9.360886249375376e-05,
+      "loss": 0.9759,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5359335500180571,
+      "grad_norm": 0.4249228239059448,
+      "learning_rate": 9.345320151843078e-05,
+      "loss": 0.3224,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5364150716263393,
+      "grad_norm": 1.9821633100509644,
+      "learning_rate": 9.329755647288485e-05,
+      "loss": 1.0542,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5368965932346214,
+      "grad_norm": 0.9637462496757507,
+      "learning_rate": 9.314192773583403e-05,
+      "loss": 0.5969,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5373781148429035,
+      "grad_norm": 0.8755899667739868,
+      "learning_rate": 9.298631568595674e-05,
+      "loss": 0.5683,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5378596364511857,
+      "grad_norm": 0.6031029224395752,
+      "learning_rate": 9.283072070189075e-05,
+      "loss": 0.624,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5383411580594679,
+      "grad_norm": 0.8484808802604675,
+      "learning_rate": 9.267514316223234e-05,
+      "loss": 0.5871,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5388226796677501,
+      "grad_norm": 0.7090550065040588,
+      "learning_rate": 9.251958344553528e-05,
+      "loss": 0.3945,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5393042012760323,
+      "grad_norm": 0.906046450138092,
+      "learning_rate": 9.23640419303101e-05,
+      "loss": 0.572,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5397857228843145,
+      "grad_norm": 0.7741056084632874,
+      "learning_rate": 9.220851899502283e-05,
+      "loss": 1.0599,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5402672444925966,
+      "grad_norm": 0.6731898784637451,
+      "learning_rate": 9.205301501809448e-05,
+      "loss": 0.4439,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5407487661008787,
+      "grad_norm": 0.40701019763946533,
+      "learning_rate": 9.189753037789987e-05,
+      "loss": 0.5468,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5412302877091609,
+      "grad_norm": 0.5402950048446655,
+      "learning_rate": 9.174206545276677e-05,
+      "loss": 0.7252,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5417118093174431,
+      "grad_norm": 0.8427053093910217,
+      "learning_rate": 9.158662062097501e-05,
+      "loss": 1.004,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5421933309257253,
+      "grad_norm": 0.36185023188591003,
+      "learning_rate": 9.143119626075542e-05,
+      "loss": 0.137,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5426748525340075,
+      "grad_norm": 1.0534988641738892,
+      "learning_rate": 9.127579275028914e-05,
+      "loss": 1.3426,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5431563741422897,
+      "grad_norm": 0.47852763533592224,
+      "learning_rate": 9.112041046770653e-05,
+      "loss": 0.5912,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5436378957505718,
+      "grad_norm": 1.3245254755020142,
+      "learning_rate": 9.096504979108629e-05,
+      "loss": 0.5478,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5441194173588539,
+      "grad_norm": 0.6596415638923645,
+      "learning_rate": 9.080971109845458e-05,
+      "loss": 0.665,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5446009389671361,
+      "grad_norm": 1.0128172636032104,
+      "learning_rate": 9.0654394767784e-05,
+      "loss": 0.523,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5450824605754183,
+      "grad_norm": 0.9200664758682251,
+      "learning_rate": 9.049910117699281e-05,
+      "loss": 0.4824,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5455639821837005,
+      "grad_norm": 0.8327468633651733,
+      "learning_rate": 9.034383070394393e-05,
+      "loss": 0.5497,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5460455037919827,
+      "grad_norm": 0.7922326326370239,
+      "learning_rate": 9.0188583726444e-05,
+      "loss": 1.0697,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5465270254002649,
+      "grad_norm": 0.9038323760032654,
+      "learning_rate": 9.00333606222425e-05,
+      "loss": 0.572,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5470085470085471,
+      "grad_norm": 0.9835309386253357,
+      "learning_rate": 8.987816176903082e-05,
+      "loss": 0.5335,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5474900686168291,
+      "grad_norm": 0.7787536978721619,
+      "learning_rate": 8.972298754444136e-05,
+      "loss": 0.7036,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5479715902251113,
+      "grad_norm": 0.6886593699455261,
+      "learning_rate": 8.956783832604654e-05,
+      "loss": 0.2535,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5484531118333935,
+      "grad_norm": 1.0839778184890747,
+      "learning_rate": 8.941271449135806e-05,
+      "loss": 1.2139,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5489346334416757,
+      "grad_norm": 1.1864848136901855,
+      "learning_rate": 8.925761641782567e-05,
+      "loss": 1.1105,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5494161550499579,
+      "grad_norm": 0.4651387631893158,
+      "learning_rate": 8.910254448283659e-05,
+      "loss": 0.2631,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5498976766582401,
+      "grad_norm": 1.0268776416778564,
+      "learning_rate": 8.894749906371439e-05,
+      "loss": 0.6797,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5503791982665223,
+      "grad_norm": 0.5834671258926392,
+      "learning_rate": 8.87924805377181e-05,
+      "loss": 0.3967,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5508607198748043,
+      "grad_norm": 0.7293856143951416,
+      "learning_rate": 8.863748928204131e-05,
+      "loss": 0.2256,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5513422414830865,
+      "grad_norm": 1.627150058746338,
+      "learning_rate": 8.848252567381131e-05,
+      "loss": 0.6562,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5518237630913687,
+      "grad_norm": 0.8662706613540649,
+      "learning_rate": 8.83275900900881e-05,
+      "loss": 0.4498,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5523052846996509,
+      "grad_norm": 0.8959720730781555,
+      "learning_rate": 8.817268290786343e-05,
+      "loss": 0.6327,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5527868063079331,
+      "grad_norm": 0.780279815196991,
+      "learning_rate": 8.801780450406002e-05,
+      "loss": 0.3141,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5532683279162153,
+      "grad_norm": 0.8042910099029541,
+      "learning_rate": 8.786295525553053e-05,
+      "loss": 0.1242,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5537498495244975,
+      "grad_norm": 0.8482663631439209,
+      "learning_rate": 8.770813553905664e-05,
+      "loss": 0.2934,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5542313711327795,
+      "grad_norm": 1.3287039995193481,
+      "learning_rate": 8.755334573134829e-05,
+      "loss": 0.5532,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5547128927410617,
+      "grad_norm": 0.7641162872314453,
+      "learning_rate": 8.739858620904251e-05,
+      "loss": 0.512,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5551944143493439,
+      "grad_norm": 0.44970107078552246,
+      "learning_rate": 8.724385734870271e-05,
+      "loss": 0.4683,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5556759359576261,
+      "grad_norm": 0.8389862179756165,
+      "learning_rate": 8.708915952681769e-05,
+      "loss": 0.353,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5561574575659083,
+      "grad_norm": 0.7373841404914856,
+      "learning_rate": 8.693449311980074e-05,
+      "loss": 0.7819,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5566389791741905,
+      "grad_norm": 1.1287013292312622,
+      "learning_rate": 8.677985850398866e-05,
+      "loss": 0.5839,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5571205007824727,
+      "grad_norm": 0.9452033638954163,
+      "learning_rate": 8.662525605564093e-05,
+      "loss": 0.534,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5576020223907547,
+      "grad_norm": 0.9100724458694458,
+      "learning_rate": 8.647068615093875e-05,
+      "loss": 1.1361,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5580835439990369,
+      "grad_norm": 1.0564773082733154,
+      "learning_rate": 8.631614916598419e-05,
+      "loss": 0.6164,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5585650656073191,
+      "grad_norm": 1.761771559715271,
+      "learning_rate": 8.616164547679906e-05,
+      "loss": 0.3506,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5590465872156013,
+      "grad_norm": 1.167173147201538,
+      "learning_rate": 8.600717545932435e-05,
+      "loss": 1.0166,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5595281088238835,
+      "grad_norm": 0.8327739238739014,
+      "learning_rate": 8.5852739489419e-05,
+      "loss": 0.1438,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5600096304321657,
+      "grad_norm": 0.8960033655166626,
+      "learning_rate": 8.569833794285915e-05,
+      "loss": 0.8653,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5604911520404479,
+      "grad_norm": 0.8796095848083496,
+      "learning_rate": 8.554397119533714e-05,
+      "loss": 1.2771,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5609726736487299,
+      "grad_norm": 0.594007670879364,
+      "learning_rate": 8.538963962246069e-05,
+      "loss": 0.3621,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5614541952570121,
+      "grad_norm": 0.8897623419761658,
+      "learning_rate": 8.523534359975189e-05,
+      "loss": 0.5407,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5619357168652943,
+      "grad_norm": 0.7416290044784546,
+      "learning_rate": 8.508108350264635e-05,
+      "loss": 0.3464,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5624172384735765,
+      "grad_norm": 0.941420316696167,
+      "learning_rate": 8.492685970649228e-05,
+      "loss": 0.432,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5628987600818587,
+      "grad_norm": 1.5289616584777832,
+      "learning_rate": 8.477267258654949e-05,
+      "loss": 0.7616,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5633802816901409,
+      "grad_norm": 1.0448668003082275,
+      "learning_rate": 8.461852251798866e-05,
+      "loss": 0.9855,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5638618032984231,
+      "grad_norm": 0.48141905665397644,
+      "learning_rate": 8.44644098758902e-05,
+      "loss": 0.2598,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5643433249067051,
+      "grad_norm": 0.6879386305809021,
+      "learning_rate": 8.431033503524354e-05,
+      "loss": 0.3566,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5648248465149873,
+      "grad_norm": 1.2092243432998657,
+      "learning_rate": 8.415629837094611e-05,
+      "loss": 0.8088,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5653063681232695,
+      "grad_norm": 0.7214942574501038,
+      "learning_rate": 8.400230025780243e-05,
+      "loss": 0.395,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5657878897315517,
+      "grad_norm": 1.370095133781433,
+      "learning_rate": 8.384834107052321e-05,
+      "loss": 0.6134,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5662694113398339,
+      "grad_norm": 1.073525071144104,
+      "learning_rate": 8.369442118372447e-05,
+      "loss": 0.8196,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5667509329481161,
+      "grad_norm": 0.705037534236908,
+      "learning_rate": 8.35405409719266e-05,
+      "loss": 0.6361,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5672324545563983,
+      "grad_norm": 0.7596774697303772,
+      "learning_rate": 8.338670080955349e-05,
+      "loss": 0.438,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5677139761646804,
+      "grad_norm": 0.9082719087600708,
+      "learning_rate": 8.323290107093143e-05,
+      "loss": 0.6048,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5681954977729625,
+      "grad_norm": 0.7221473455429077,
+      "learning_rate": 8.307914213028856e-05,
+      "loss": 0.6149,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5686770193812447,
+      "grad_norm": 0.6236855983734131,
+      "learning_rate": 8.292542436175356e-05,
+      "loss": 0.5865,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5691585409895269,
+      "grad_norm": 1.6320685148239136,
+      "learning_rate": 8.277174813935508e-05,
+      "loss": 1.1466,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5696400625978091,
+      "grad_norm": 0.8854456543922424,
+      "learning_rate": 8.261811383702061e-05,
+      "loss": 0.5219,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5701215842060913,
+      "grad_norm": 0.8134288787841797,
+      "learning_rate": 8.246452182857562e-05,
+      "loss": 0.4383,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5706031058143735,
+      "grad_norm": 1.0532677173614502,
+      "learning_rate": 8.231097248774274e-05,
+      "loss": 0.9061,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5710846274226556,
+      "grad_norm": 0.7403393983840942,
+      "learning_rate": 8.215746618814067e-05,
+      "loss": 0.2132,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5715661490309377,
+      "grad_norm": 1.1494709253311157,
+      "learning_rate": 8.200400330328348e-05,
+      "loss": 0.866,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5720476706392199,
+      "grad_norm": 1.0485625267028809,
+      "learning_rate": 8.185058420657957e-05,
+      "loss": 1.1284,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5725291922475021,
+      "grad_norm": 1.1191109418869019,
+      "learning_rate": 8.16972092713308e-05,
+      "loss": 0.6806,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5730107138557843,
+      "grad_norm": 1.6901990175247192,
+      "learning_rate": 8.154387887073158e-05,
+      "loss": 0.4902,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5734922354640665,
+      "grad_norm": 0.799044132232666,
+      "learning_rate": 8.139059337786792e-05,
+      "loss": 0.5245,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5739737570723487,
+      "grad_norm": 0.6634275913238525,
+      "learning_rate": 8.12373531657166e-05,
+      "loss": 0.7991,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5744552786806308,
+      "grad_norm": 0.5595707893371582,
+      "learning_rate": 8.108415860714418e-05,
+      "loss": 0.2768,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5749368002889129,
+      "grad_norm": 1.764378547668457,
+      "learning_rate": 8.093101007490622e-05,
+      "loss": 1.0182,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5754183218971951,
+      "grad_norm": 0.7943925261497498,
+      "learning_rate": 8.077790794164619e-05,
+      "loss": 0.4515,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5758998435054773,
+      "grad_norm": 0.6726824045181274,
+      "learning_rate": 8.062485257989471e-05,
+      "loss": 0.6998,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5763813651137595,
+      "grad_norm": 0.791667103767395,
+      "learning_rate": 8.047184436206864e-05,
+      "loss": 0.7586,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5768628867220417,
+      "grad_norm": 0.7900064587593079,
+      "learning_rate": 8.031888366046998e-05,
+      "loss": 0.5003,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5773444083303239,
+      "grad_norm": 0.6237706542015076,
+      "learning_rate": 8.016597084728526e-05,
+      "loss": 0.6402,
+      "step": 1199
+    },
+    {
+      "epoch": 0.577825929938606,
+      "grad_norm": 0.612813413143158,
+      "learning_rate": 8.001310629458443e-05,
+      "loss": 0.8667,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5783074515468881,
+      "grad_norm": 0.8505640625953674,
+      "learning_rate": 7.986029037432002e-05,
+      "loss": 0.9555,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5787889731551703,
+      "grad_norm": 0.801115095615387,
+      "learning_rate": 7.970752345832623e-05,
+      "loss": 0.4727,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5792704947634525,
+      "grad_norm": 0.49146944284439087,
+      "learning_rate": 7.9554805918318e-05,
+      "loss": 0.2439,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5797520163717347,
+      "grad_norm": 0.7376623749732971,
+      "learning_rate": 7.940213812589018e-05,
+      "loss": 0.807,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5802335379800169,
+      "grad_norm": 0.5820401906967163,
+      "learning_rate": 7.92495204525165e-05,
+      "loss": 0.4257,
+      "step": 1205
+    },
+    {
+      "epoch": 0.580715059588299,
+      "grad_norm": 0.8996889591217041,
+      "learning_rate": 7.909695326954878e-05,
+      "loss": 0.5803,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5811965811965812,
+      "grad_norm": 0.5629736185073853,
+      "learning_rate": 7.894443694821602e-05,
+      "loss": 0.3024,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5816781028048633,
+      "grad_norm": 0.8386406302452087,
+      "learning_rate": 7.879197185962339e-05,
+      "loss": 0.5131,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5821596244131455,
+      "grad_norm": 1.4311405420303345,
+      "learning_rate": 7.863955837475144e-05,
+      "loss": 0.5596,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5826411460214277,
+      "grad_norm": 0.671789824962616,
+      "learning_rate": 7.848719686445515e-05,
+      "loss": 0.374,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5831226676297099,
+      "grad_norm": 0.8818476796150208,
+      "learning_rate": 7.833488769946306e-05,
+      "loss": 0.6616,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5836041892379921,
+      "grad_norm": 0.9540754556655884,
+      "learning_rate": 7.818263125037633e-05,
+      "loss": 0.4891,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5840857108462743,
+      "grad_norm": 0.7437126040458679,
+      "learning_rate": 7.803042788766777e-05,
+      "loss": 0.3433,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5845672324545564,
+      "grad_norm": 0.3874172568321228,
+      "learning_rate": 7.787827798168115e-05,
+      "loss": 0.2215,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5850487540628385,
+      "grad_norm": 0.6588778495788574,
+      "learning_rate": 7.772618190263009e-05,
+      "loss": 0.5484,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5855302756711207,
+      "grad_norm": 0.536056399345398,
+      "learning_rate": 7.757414002059726e-05,
+      "loss": 0.4488,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5860117972794029,
+      "grad_norm": 1.0876915454864502,
+      "learning_rate": 7.742215270553349e-05,
+      "loss": 0.4387,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5864933188876851,
+      "grad_norm": 1.0314831733703613,
+      "learning_rate": 7.727022032725672e-05,
+      "loss": 0.6554,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5869748404959673,
+      "grad_norm": 0.8363491296768188,
+      "learning_rate": 7.711834325545135e-05,
+      "loss": 0.483,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5874563621042495,
+      "grad_norm": 0.8176609873771667,
+      "learning_rate": 7.696652185966711e-05,
+      "loss": 0.6103,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5879378837125316,
+      "grad_norm": 0.7678027153015137,
+      "learning_rate": 7.681475650931834e-05,
+      "loss": 0.5092,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5884194053208138,
+      "grad_norm": 0.6872639656066895,
+      "learning_rate": 7.666304757368297e-05,
+      "loss": 0.7324,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5889009269290959,
+      "grad_norm": 0.47040292620658875,
+      "learning_rate": 7.651139542190164e-05,
+      "loss": 0.6611,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5893824485373781,
+      "grad_norm": 0.6442480087280273,
+      "learning_rate": 7.635980042297687e-05,
+      "loss": 0.3392,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5898639701456603,
+      "grad_norm": 0.7973917126655579,
+      "learning_rate": 7.620826294577208e-05,
+      "loss": 0.4293,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5903454917539425,
+      "grad_norm": 1.671976923942566,
+      "learning_rate": 7.605678335901071e-05,
+      "loss": 0.5356,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5908270133622247,
+      "grad_norm": 1.2067997455596924,
+      "learning_rate": 7.59053620312754e-05,
+      "loss": 0.8919,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5913085349705068,
+      "grad_norm": 0.5278024077415466,
+      "learning_rate": 7.575399933100697e-05,
+      "loss": 0.4646,
+      "step": 1228
+    },
+    {
+      "epoch": 0.591790056578789,
+      "grad_norm": 0.8876713514328003,
+      "learning_rate": 7.560269562650368e-05,
+      "loss": 0.6154,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5922715781870711,
+      "grad_norm": 0.7381996512413025,
+      "learning_rate": 7.54514512859201e-05,
+      "loss": 0.3608,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5927530997953533,
+      "grad_norm": 0.8265383839607239,
+      "learning_rate": 7.530026667726645e-05,
+      "loss": 0.4787,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5932346214036355,
+      "grad_norm": 0.8717719316482544,
+      "learning_rate": 7.51491421684076e-05,
+      "loss": 0.9259,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5937161430119177,
+      "grad_norm": 1.1018342971801758,
+      "learning_rate": 7.49980781270622e-05,
+      "loss": 0.5792,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5941976646201999,
+      "grad_norm": 0.6729315519332886,
+      "learning_rate": 7.484707492080172e-05,
+      "loss": 0.7996,
+      "step": 1234
+    },
+    {
+      "epoch": 0.594679186228482,
+      "grad_norm": 0.9117011427879333,
+      "learning_rate": 7.469613291704962e-05,
+      "loss": 0.9377,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5951607078367642,
+      "grad_norm": 0.7894290089607239,
+      "learning_rate": 7.45452524830805e-05,
+      "loss": 0.7289,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5956422294450463,
+      "grad_norm": 0.4750572443008423,
+      "learning_rate": 7.439443398601903e-05,
+      "loss": 0.5718,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5961237510533285,
+      "grad_norm": 0.7789412140846252,
+      "learning_rate": 7.424367779283926e-05,
+      "loss": 0.9233,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5966052726616107,
+      "grad_norm": 0.7270252704620361,
+      "learning_rate": 7.409298427036364e-05,
+      "loss": 0.3033,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5970867942698929,
+      "grad_norm": 1.0719108581542969,
+      "learning_rate": 7.39423537852621e-05,
+      "loss": 0.4617,
+      "step": 1240
+    },
+    {
+      "epoch": 0.597568315878175,
+      "grad_norm": 0.5281317234039307,
+      "learning_rate": 7.379178670405123e-05,
+      "loss": 0.3342,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5980498374864572,
+      "grad_norm": 0.40090054273605347,
+      "learning_rate": 7.364128339309326e-05,
+      "loss": 0.3611,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5985313590947394,
+      "grad_norm": 0.7400622367858887,
+      "learning_rate": 7.349084421859533e-05,
+      "loss": 0.3046,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5990128807030215,
+      "grad_norm": 0.7203524708747864,
+      "learning_rate": 7.334046954660852e-05,
+      "loss": 0.34,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5994944023113037,
+      "grad_norm": 0.7717897891998291,
+      "learning_rate": 7.31901597430269e-05,
+      "loss": 0.7379,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5999759239195859,
+      "grad_norm": 0.8210568428039551,
+      "learning_rate": 7.303991517358678e-05,
+      "loss": 0.8208,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6004574455278681,
+      "grad_norm": 0.8457664251327515,
+      "learning_rate": 7.288973620386568e-05,
+      "loss": 0.8886,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6009389671361502,
+      "grad_norm": 0.5993969440460205,
+      "learning_rate": 7.273962319928151e-05,
+      "loss": 0.488,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6014204887444324,
+      "grad_norm": 0.5078853964805603,
+      "learning_rate": 7.258957652509171e-05,
+      "loss": 0.4295,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6019020103527146,
+      "grad_norm": 0.6125525236129761,
+      "learning_rate": 7.24395965463923e-05,
+      "loss": 0.151,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6023835319609967,
+      "grad_norm": 0.6394007802009583,
+      "learning_rate": 7.228968362811702e-05,
+      "loss": 0.4192,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6028650535692789,
+      "grad_norm": 0.8358025550842285,
+      "learning_rate": 7.21398381350364e-05,
+      "loss": 0.6631,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6033465751775611,
+      "grad_norm": 0.5983412265777588,
+      "learning_rate": 7.199006043175698e-05,
+      "loss": 0.6831,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6038280967858433,
+      "grad_norm": 0.5821589231491089,
+      "learning_rate": 7.184035088272028e-05,
+      "loss": 0.4347,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6043096183941254,
+      "grad_norm": 0.6282423734664917,
+      "learning_rate": 7.169070985220208e-05,
+      "loss": 0.2918,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6047911400024076,
+      "grad_norm": 0.7748804092407227,
+      "learning_rate": 7.154113770431132e-05,
+      "loss": 0.5475,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6052726616106898,
+      "grad_norm": 0.6467021107673645,
+      "learning_rate": 7.13916348029894e-05,
+      "loss": 0.6028,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6057541832189719,
+      "grad_norm": 0.7658000588417053,
+      "learning_rate": 7.124220151200926e-05,
+      "loss": 0.4896,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6062357048272541,
+      "grad_norm": 0.5645423531532288,
+      "learning_rate": 7.10928381949744e-05,
+      "loss": 0.3842,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6067172264355363,
+      "grad_norm": 1.0179091691970825,
+      "learning_rate": 7.094354521531807e-05,
+      "loss": 0.8221,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6071987480438185,
+      "grad_norm": 0.9218524098396301,
+      "learning_rate": 7.079432293630244e-05,
+      "loss": 1.0428,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6076802696521006,
+      "grad_norm": 1.1272860765457153,
+      "learning_rate": 7.064517172101753e-05,
+      "loss": 1.1758,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6081617912603828,
+      "grad_norm": 1.2278001308441162,
+      "learning_rate": 7.04960919323806e-05,
+      "loss": 0.5739,
+      "step": 1263
+    },
+    {
+      "epoch": 0.608643312868665,
+      "grad_norm": 0.9171412587165833,
+      "learning_rate": 7.034708393313493e-05,
+      "loss": 1.0,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6091248344769472,
+      "grad_norm": 1.4739869832992554,
+      "learning_rate": 7.019814808584928e-05,
+      "loss": 1.0325,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6096063560852293,
+      "grad_norm": 0.9023488163948059,
+      "learning_rate": 7.004928475291678e-05,
+      "loss": 0.527,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6100878776935115,
+      "grad_norm": 1.3271821737289429,
+      "learning_rate": 6.990049429655412e-05,
+      "loss": 0.6692,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6105693993017937,
+      "grad_norm": 0.9032428860664368,
+      "learning_rate": 6.97517770788007e-05,
+      "loss": 0.7722,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6110509209100758,
+      "grad_norm": 0.7868967652320862,
+      "learning_rate": 6.960313346151761e-05,
+      "loss": 0.6579,
+      "step": 1269
+    },
+    {
+      "epoch": 0.611532442518358,
+      "grad_norm": 1.430017113685608,
+      "learning_rate": 6.9454563806387e-05,
+      "loss": 1.0602,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6120139641266402,
+      "grad_norm": 0.6199007630348206,
+      "learning_rate": 6.930606847491094e-05,
+      "loss": 0.5871,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6124954857349224,
+      "grad_norm": 0.9371126294136047,
+      "learning_rate": 6.915764782841072e-05,
+      "loss": 0.8398,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6129770073432045,
+      "grad_norm": 0.8416410684585571,
+      "learning_rate": 6.900930222802588e-05,
+      "loss": 0.5942,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6134585289514867,
+      "grad_norm": 0.6848739385604858,
+      "learning_rate": 6.886103203471337e-05,
+      "loss": 0.7256,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6139400505597689,
+      "grad_norm": 0.899040162563324,
+      "learning_rate": 6.871283760924665e-05,
+      "loss": 0.5853,
+      "step": 1275
+    },
+    {
+      "epoch": 0.614421572168051,
+      "grad_norm": 0.7222660183906555,
+      "learning_rate": 6.856471931221478e-05,
+      "loss": 0.7026,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6149030937763332,
+      "grad_norm": 0.5651343464851379,
+      "learning_rate": 6.841667750402162e-05,
+      "loss": 0.3303,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.5582093000411987,
+      "learning_rate": 6.826871254488496e-05,
+      "loss": 0.3963,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6158661369928976,
+      "grad_norm": 0.6273393034934998,
+      "learning_rate": 6.812082479483553e-05,
+      "loss": 0.1722,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6163476586011797,
+      "grad_norm": 0.5864158868789673,
+      "learning_rate": 6.797301461371625e-05,
+      "loss": 0.2222,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6168291802094619,
+      "grad_norm": 0.4145785868167877,
+      "learning_rate": 6.782528236118124e-05,
+      "loss": 0.1757,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6173107018177441,
+      "grad_norm": 0.8258092403411865,
+      "learning_rate": 6.767762839669503e-05,
+      "loss": 0.4112,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6177922234260262,
+      "grad_norm": 0.7576352953910828,
+      "learning_rate": 6.753005307953167e-05,
+      "loss": 0.4947,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6182737450343084,
+      "grad_norm": 3.720597743988037,
+      "learning_rate": 6.738255676877381e-05,
+      "loss": 0.3222,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6187552666425906,
+      "grad_norm": 0.43584874272346497,
+      "learning_rate": 6.723513982331195e-05,
+      "loss": 0.3454,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6192367882508728,
+      "grad_norm": 1.3589578866958618,
+      "learning_rate": 6.708780260184333e-05,
+      "loss": 0.4681,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6197183098591549,
+      "grad_norm": 0.5593907237052917,
+      "learning_rate": 6.694054546287132e-05,
+      "loss": 0.4205,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6201998314674371,
+      "grad_norm": 0.9470499157905579,
+      "learning_rate": 6.679336876470441e-05,
+      "loss": 0.7325,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6206813530757193,
+      "grad_norm": 1.2402819395065308,
+      "learning_rate": 6.664627286545535e-05,
+      "loss": 0.5052,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6211628746840014,
+      "grad_norm": 0.6698610782623291,
+      "learning_rate": 6.649925812304025e-05,
+      "loss": 0.4292,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6216443962922836,
+      "grad_norm": 0.7191325426101685,
+      "learning_rate": 6.635232489517782e-05,
+      "loss": 0.669,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6221259179005658,
+      "grad_norm": 0.5654041171073914,
+      "learning_rate": 6.620547353938836e-05,
+      "loss": 0.4625,
+      "step": 1292
+    },
+    {
+      "epoch": 0.622607439508848,
+      "grad_norm": 0.7504422068595886,
+      "learning_rate": 6.605870441299302e-05,
+      "loss": 0.3405,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6230889611171301,
+      "grad_norm": 1.1788744926452637,
+      "learning_rate": 6.591201787311285e-05,
+      "loss": 1.0483,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6235704827254123,
+      "grad_norm": 0.8092876076698303,
+      "learning_rate": 6.57654142766679e-05,
+      "loss": 0.8794,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6240520043336945,
+      "grad_norm": 0.7599140405654907,
+      "learning_rate": 6.561889398037643e-05,
+      "loss": 0.3576,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6245335259419766,
+      "grad_norm": 1.1310917139053345,
+      "learning_rate": 6.547245734075403e-05,
+      "loss": 0.4158,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6250150475502588,
+      "grad_norm": 0.5297826528549194,
+      "learning_rate": 6.532610471411274e-05,
+      "loss": 0.3243,
+      "step": 1298
+    },
+    {
+      "epoch": 0.625496569158541,
+      "grad_norm": 0.8771138191223145,
+      "learning_rate": 6.517983645656014e-05,
+      "loss": 0.3289,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6259780907668232,
+      "grad_norm": 0.9111340641975403,
+      "learning_rate": 6.503365292399857e-05,
+      "loss": 0.8497,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6264596123751053,
+      "grad_norm": 0.7726771235466003,
+      "learning_rate": 6.488755447212418e-05,
+      "loss": 0.6634,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6269411339833875,
+      "grad_norm": 0.7417446970939636,
+      "learning_rate": 6.474154145642612e-05,
+      "loss": 1.17,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6274226555916697,
+      "grad_norm": 0.7755337953567505,
+      "learning_rate": 6.459561423218561e-05,
+      "loss": 1.199,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6279041771999518,
+      "grad_norm": 0.4433456361293793,
+      "learning_rate": 6.444977315447521e-05,
+      "loss": 0.3119,
+      "step": 1304
+    },
+    {
+      "epoch": 0.628385698808234,
+      "grad_norm": 0.6270627975463867,
+      "learning_rate": 6.430401857815776e-05,
+      "loss": 0.415,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6288672204165162,
+      "grad_norm": 0.9268920421600342,
+      "learning_rate": 6.415835085788575e-05,
+      "loss": 0.6387,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6293487420247984,
+      "grad_norm": 0.47612449526786804,
+      "learning_rate": 6.401277034810017e-05,
+      "loss": 0.1866,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6298302636330806,
+      "grad_norm": 0.8117198348045349,
+      "learning_rate": 6.386727740302994e-05,
+      "loss": 0.6716,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6303117852413627,
+      "grad_norm": 1.017600417137146,
+      "learning_rate": 6.37218723766909e-05,
+      "loss": 0.2563,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6307933068496449,
+      "grad_norm": 1.1226799488067627,
+      "learning_rate": 6.357655562288488e-05,
+      "loss": 0.8567,
+      "step": 1310
+    },
+    {
+      "epoch": 0.631274828457927,
+      "grad_norm": 1.2738351821899414,
+      "learning_rate": 6.343132749519902e-05,
+      "loss": 0.9565,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6317563500662092,
+      "grad_norm": 1.0473432540893555,
+      "learning_rate": 6.328618834700474e-05,
+      "loss": 0.8701,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6322378716744914,
+      "grad_norm": 1.9390569925308228,
+      "learning_rate": 6.314113853145703e-05,
+      "loss": 0.4867,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6327193932827736,
+      "grad_norm": 0.5643951296806335,
+      "learning_rate": 6.299617840149349e-05,
+      "loss": 0.184,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6332009148910558,
+      "grad_norm": 0.5072335004806519,
+      "learning_rate": 6.285130830983339e-05,
+      "loss": 0.1524,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6336824364993379,
+      "grad_norm": 0.8295088410377502,
+      "learning_rate": 6.270652860897704e-05,
+      "loss": 0.6265,
+      "step": 1316
+    },
+    {
+      "epoch": 0.63416395810762,
+      "grad_norm": 0.7741579413414001,
+      "learning_rate": 6.25618396512048e-05,
+      "loss": 0.5177,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6346454797159022,
+      "grad_norm": 0.6545950770378113,
+      "learning_rate": 6.24172417885762e-05,
+      "loss": 0.7957,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6351270013241844,
+      "grad_norm": 0.7442547082901001,
+      "learning_rate": 6.227273537292911e-05,
+      "loss": 0.8306,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6356085229324666,
+      "grad_norm": 0.6095592975616455,
+      "learning_rate": 6.212832075587891e-05,
+      "loss": 0.442,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6360900445407488,
+      "grad_norm": 1.4066251516342163,
+      "learning_rate": 6.19839982888176e-05,
+      "loss": 0.5071,
+      "step": 1321
+    },
+    {
+      "epoch": 0.636571566149031,
+      "grad_norm": 0.7725120782852173,
+      "learning_rate": 6.183976832291296e-05,
+      "loss": 0.3988,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6370530877573131,
+      "grad_norm": 0.5813517570495605,
+      "learning_rate": 6.169563120910775e-05,
+      "loss": 0.5359,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6375346093655953,
+      "grad_norm": 0.9773880839347839,
+      "learning_rate": 6.155158729811867e-05,
+      "loss": 0.5878,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6380161309738774,
+      "grad_norm": 0.6472037434577942,
+      "learning_rate": 6.140763694043578e-05,
+      "loss": 0.3894,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6384976525821596,
+      "grad_norm": 0.5015087723731995,
+      "learning_rate": 6.126378048632139e-05,
+      "loss": 0.5544,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6389791741904418,
+      "grad_norm": 0.7416954636573792,
+      "learning_rate": 6.112001828580944e-05,
+      "loss": 0.8504,
+      "step": 1327
+    },
+    {
+      "epoch": 0.639460695798724,
+      "grad_norm": 0.4432564973831177,
+      "learning_rate": 6.0976350688704455e-05,
+      "loss": 0.1723,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6399422174070062,
+      "grad_norm": 0.6768189072608948,
+      "learning_rate": 6.083277804458072e-05,
+      "loss": 0.7732,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6404237390152883,
+      "grad_norm": 1.28372061252594,
+      "learning_rate": 6.068930070278159e-05,
+      "loss": 0.7025,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6409052606235704,
+      "grad_norm": 0.7883098125457764,
+      "learning_rate": 6.054591901241846e-05,
+      "loss": 0.4218,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6413867822318526,
+      "grad_norm": 1.1983596086502075,
+      "learning_rate": 6.040263332237002e-05,
+      "loss": 0.9044,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6418683038401348,
+      "grad_norm": 0.5475990772247314,
+      "learning_rate": 6.025944398128137e-05,
+      "loss": 0.5043,
+      "step": 1333
+    },
+    {
+      "epoch": 0.642349825448417,
+      "grad_norm": 1.6600192785263062,
+      "learning_rate": 6.011635133756309e-05,
+      "loss": 0.5005,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6428313470566992,
+      "grad_norm": 0.6498377919197083,
+      "learning_rate": 5.99733557393906e-05,
+      "loss": 0.7945,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6433128686649814,
+      "grad_norm": 0.9499074816703796,
+      "learning_rate": 5.983045753470308e-05,
+      "loss": 0.9912,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6437943902732635,
+      "grad_norm": 0.6833046078681946,
+      "learning_rate": 5.96876570712028e-05,
+      "loss": 0.985,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6442759118815456,
+      "grad_norm": 0.7049830555915833,
+      "learning_rate": 5.954495469635417e-05,
+      "loss": 0.5277,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6447574334898278,
+      "grad_norm": 0.9056922197341919,
+      "learning_rate": 5.940235075738296e-05,
+      "loss": 0.6231,
+      "step": 1339
+    },
+    {
+      "epoch": 0.64523895509811,
+      "grad_norm": 0.8221293091773987,
+      "learning_rate": 5.925984560127542e-05,
+      "loss": 0.8753,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6457204767063922,
+      "grad_norm": 0.7707056999206543,
+      "learning_rate": 5.911743957477739e-05,
+      "loss": 0.8813,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6462019983146744,
+      "grad_norm": 0.6258026957511902,
+      "learning_rate": 5.897513302439355e-05,
+      "loss": 0.4881,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6466835199229566,
+      "grad_norm": 0.7180628776550293,
+      "learning_rate": 5.883292629638651e-05,
+      "loss": 0.7696,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6471650415312387,
+      "grad_norm": 0.7205024361610413,
+      "learning_rate": 5.869081973677604e-05,
+      "loss": 0.4843,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6476465631395208,
+      "grad_norm": 0.9404179453849792,
+      "learning_rate": 5.8548813691338134e-05,
+      "loss": 0.8674,
+      "step": 1345
+    },
+    {
+      "epoch": 0.648128084747803,
+      "grad_norm": 0.9139581322669983,
+      "learning_rate": 5.84069085056042e-05,
+      "loss": 0.9279,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6486096063560852,
+      "grad_norm": 0.902281641960144,
+      "learning_rate": 5.826510452486027e-05,
+      "loss": 0.6238,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6490911279643674,
+      "grad_norm": 0.8244741559028625,
+      "learning_rate": 5.81234020941461e-05,
+      "loss": 0.4828,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6495726495726496,
+      "grad_norm": 0.6052896976470947,
+      "learning_rate": 5.798180155825437e-05,
+      "loss": 0.5735,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6500541711809318,
+      "grad_norm": 0.6693307757377625,
+      "learning_rate": 5.784030326172981e-05,
+      "loss": 0.5781,
+      "step": 1350
+    },
+    {
+      "epoch": 0.650535692789214,
+      "grad_norm": 0.7676788568496704,
+      "learning_rate": 5.7698907548868395e-05,
+      "loss": 0.5507,
+      "step": 1351
+    },
+    {
+      "epoch": 0.651017214397496,
+      "grad_norm": 0.5379273295402527,
+      "learning_rate": 5.755761476371653e-05,
+      "loss": 0.5197,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6514987360057782,
+      "grad_norm": 0.7525188326835632,
+      "learning_rate": 5.741642525007003e-05,
+      "loss": 0.9615,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6519802576140604,
+      "grad_norm": 1.112884521484375,
+      "learning_rate": 5.727533935147359e-05,
+      "loss": 0.7473,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6524617792223426,
+      "grad_norm": 1.1446647644042969,
+      "learning_rate": 5.713435741121975e-05,
+      "loss": 0.3788,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6529433008306248,
+      "grad_norm": 1.2720409631729126,
+      "learning_rate": 5.699347977234799e-05,
+      "loss": 0.616,
+      "step": 1356
+    },
+    {
+      "epoch": 0.653424822438907,
+      "grad_norm": 1.1350654363632202,
+      "learning_rate": 5.685270677764412e-05,
+      "loss": 0.6315,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6539063440471892,
+      "grad_norm": 0.6351329684257507,
+      "learning_rate": 5.671203876963931e-05,
+      "loss": 0.6546,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6543878656554712,
+      "grad_norm": 0.8298540711402893,
+      "learning_rate": 5.657147609060924e-05,
+      "loss": 0.7425,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6548693872637534,
+      "grad_norm": 0.930432915687561,
+      "learning_rate": 5.643101908257333e-05,
+      "loss": 0.6273,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6553509088720356,
+      "grad_norm": 0.4794197082519531,
+      "learning_rate": 5.629066808729385e-05,
+      "loss": 0.1168,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6558324304803178,
+      "grad_norm": 0.9514833092689514,
+      "learning_rate": 5.6150423446275144e-05,
+      "loss": 0.3497,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6563139520886,
+      "grad_norm": 0.6880574822425842,
+      "learning_rate": 5.601028550076277e-05,
+      "loss": 0.1992,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6567954736968822,
+      "grad_norm": 0.6361804008483887,
+      "learning_rate": 5.587025459174271e-05,
+      "loss": 0.5376,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6572769953051644,
+      "grad_norm": 0.7879760265350342,
+      "learning_rate": 5.573033105994038e-05,
+      "loss": 0.5568,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6577585169134464,
+      "grad_norm": 0.8786200284957886,
+      "learning_rate": 5.559051524582002e-05,
+      "loss": 0.7973,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6582400385217286,
+      "grad_norm": 0.602081835269928,
+      "learning_rate": 5.5450807489583777e-05,
+      "loss": 0.9517,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6587215601300108,
+      "grad_norm": 0.4655710458755493,
+      "learning_rate": 5.531120813117085e-05,
+      "loss": 0.3658,
+      "step": 1368
+    },
+    {
+      "epoch": 0.659203081738293,
+      "grad_norm": 0.7563959956169128,
+      "learning_rate": 5.517171751025667e-05,
+      "loss": 0.6331,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6596846033465752,
+      "grad_norm": 0.630258321762085,
+      "learning_rate": 5.5032335966252103e-05,
+      "loss": 0.3159,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6601661249548574,
+      "grad_norm": 0.7362279295921326,
+      "learning_rate": 5.489306383830258e-05,
+      "loss": 0.2588,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6606476465631396,
+      "grad_norm": 0.608984649181366,
+      "learning_rate": 5.475390146528738e-05,
+      "loss": 0.2641,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6611291681714216,
+      "grad_norm": 0.6702209115028381,
+      "learning_rate": 5.461484918581858e-05,
+      "loss": 0.3605,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6616106897797038,
+      "grad_norm": 1.0780071020126343,
+      "learning_rate": 5.4475907338240494e-05,
+      "loss": 0.3744,
+      "step": 1374
+    },
+    {
+      "epoch": 0.662092211387986,
+      "grad_norm": 1.0941399335861206,
+      "learning_rate": 5.43370762606287e-05,
+      "loss": 0.6251,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6625737329962682,
+      "grad_norm": 0.6190235018730164,
+      "learning_rate": 5.4198356290789276e-05,
+      "loss": 0.6111,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6630552546045504,
+      "grad_norm": 1.441415548324585,
+      "learning_rate": 5.405974776625785e-05,
+      "loss": 0.4304,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6635367762128326,
+      "grad_norm": 0.496945858001709,
+      "learning_rate": 5.392125102429899e-05,
+      "loss": 0.52,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6640182978211148,
+      "grad_norm": 0.6923816800117493,
+      "learning_rate": 5.378286640190522e-05,
+      "loss": 0.4308,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6644998194293968,
+      "grad_norm": 0.938606858253479,
+      "learning_rate": 5.364459423579629e-05,
+      "loss": 0.3205,
+      "step": 1380
+    },
+    {
+      "epoch": 0.664981341037679,
+      "grad_norm": 0.6108879446983337,
+      "learning_rate": 5.350643486241825e-05,
+      "loss": 0.2478,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6654628626459612,
+      "grad_norm": 0.5518016815185547,
+      "learning_rate": 5.33683886179428e-05,
+      "loss": 0.4628,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6659443842542434,
+      "grad_norm": 0.5168072581291199,
+      "learning_rate": 5.3230455838266266e-05,
+      "loss": 0.2749,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6664259058625256,
+      "grad_norm": 0.7263458967208862,
+      "learning_rate": 5.309263685900898e-05,
+      "loss": 0.2829,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6669074274708078,
+      "grad_norm": 0.574898898601532,
+      "learning_rate": 5.295493201551433e-05,
+      "loss": 0.2551,
+      "step": 1385
+    },
+    {
+      "epoch": 0.66738894907909,
+      "grad_norm": 0.9553223848342896,
+      "learning_rate": 5.281734164284802e-05,
+      "loss": 0.6584,
+      "step": 1386
+    },
+    {
+      "epoch": 0.667870470687372,
+      "grad_norm": 1.0642074346542358,
+      "learning_rate": 5.26798660757971e-05,
+      "loss": 0.9263,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6683519922956542,
+      "grad_norm": 0.9214754104614258,
+      "learning_rate": 5.2542505648869434e-05,
+      "loss": 0.6928,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6688335139039364,
+      "grad_norm": 0.7935864329338074,
+      "learning_rate": 5.240526069629265e-05,
+      "loss": 0.7917,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6693150355122186,
+      "grad_norm": 0.8348840475082397,
+      "learning_rate": 5.22681315520134e-05,
+      "loss": 1.1228,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6697965571205008,
+      "grad_norm": 1.2822461128234863,
+      "learning_rate": 5.213111854969661e-05,
+      "loss": 0.4916,
+      "step": 1391
+    },
+    {
+      "epoch": 0.670278078728783,
+      "grad_norm": 0.6782784461975098,
+      "learning_rate": 5.199422202272448e-05,
+      "loss": 0.3989,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6707596003370652,
+      "grad_norm": 1.1376097202301025,
+      "learning_rate": 5.185744230419589e-05,
+      "loss": 0.6257,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6712411219453474,
+      "grad_norm": 1.124526858329773,
+      "learning_rate": 5.172077972692553e-05,
+      "loss": 0.7778,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6717226435536294,
+      "grad_norm": 0.7318700551986694,
+      "learning_rate": 5.1584234623442974e-05,
+      "loss": 0.7091,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6722041651619116,
+      "grad_norm": 0.3620985448360443,
+      "learning_rate": 5.1447807325992025e-05,
+      "loss": 0.1584,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6726856867701938,
+      "grad_norm": 0.6356948614120483,
+      "learning_rate": 5.13114981665298e-05,
+      "loss": 0.9857,
+      "step": 1397
+    },
+    {
+      "epoch": 0.673167208378476,
+      "grad_norm": 0.6141039729118347,
+      "learning_rate": 5.117530747672603e-05,
+      "loss": 0.7072,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6736487299867582,
+      "grad_norm": 0.4410860240459442,
+      "learning_rate": 5.103923558796203e-05,
+      "loss": 0.5948,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6741302515950404,
+      "grad_norm": 0.7041406631469727,
+      "learning_rate": 5.090328283133019e-05,
+      "loss": 0.3836,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6746117732033226,
+      "grad_norm": 0.596027135848999,
+      "learning_rate": 5.0767449537632986e-05,
+      "loss": 0.2445,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6750932948116046,
+      "grad_norm": 0.6667171716690063,
+      "learning_rate": 5.06317360373822e-05,
+      "loss": 0.4582,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6755748164198868,
+      "grad_norm": 0.6862946152687073,
+      "learning_rate": 5.049614266079813e-05,
+      "loss": 0.3222,
+      "step": 1403
+    },
+    {
+      "epoch": 0.676056338028169,
+      "grad_norm": 0.5457684397697449,
+      "learning_rate": 5.036066973780882e-05,
+      "loss": 0.5482,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6765378596364512,
+      "grad_norm": 0.7717635631561279,
+      "learning_rate": 5.022531759804918e-05,
+      "loss": 0.5445,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6770193812447334,
+      "grad_norm": 0.994358241558075,
+      "learning_rate": 5.009008657086025e-05,
+      "loss": 0.826,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6775009028530156,
+      "grad_norm": 0.8234153985977173,
+      "learning_rate": 4.9954976985288395e-05,
+      "loss": 0.3035,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6779824244612978,
+      "grad_norm": 1.1973003149032593,
+      "learning_rate": 4.981998917008448e-05,
+      "loss": 0.406,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6784639460695798,
+      "grad_norm": 0.8367493748664856,
+      "learning_rate": 4.9685123453703e-05,
+      "loss": 0.7669,
+      "step": 1409
+    },
+    {
+      "epoch": 0.678945467677862,
+      "grad_norm": 0.4213549494743347,
+      "learning_rate": 4.955038016430149e-05,
+      "loss": 0.1896,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6794269892861442,
+      "grad_norm": 0.8429536819458008,
+      "learning_rate": 4.9415759629739455e-05,
+      "loss": 0.532,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6799085108944264,
+      "grad_norm": 1.1326082944869995,
+      "learning_rate": 4.928126217757782e-05,
+      "loss": 1.2761,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6803900325027086,
+      "grad_norm": 0.42613062262535095,
+      "learning_rate": 4.914688813507797e-05,
+      "loss": 0.6146,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6808715541109908,
+      "grad_norm": 0.6152105927467346,
+      "learning_rate": 4.901263782920105e-05,
+      "loss": 0.6455,
+      "step": 1414
+    },
+    {
+      "epoch": 0.681353075719273,
+      "grad_norm": 0.9997090101242065,
+      "learning_rate": 4.887851158660706e-05,
+      "loss": 0.307,
+      "step": 1415
+    },
+    {
+      "epoch": 0.681834597327555,
+      "grad_norm": 0.8020671606063843,
+      "learning_rate": 4.8744509733654184e-05,
+      "loss": 0.7692,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6823161189358372,
+      "grad_norm": 1.184843897819519,
+      "learning_rate": 4.861063259639793e-05,
+      "loss": 0.714,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6827976405441194,
+      "grad_norm": 0.5545733571052551,
+      "learning_rate": 4.847688050059033e-05,
+      "loss": 0.3846,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6832791621524016,
+      "grad_norm": 0.7227141857147217,
+      "learning_rate": 4.8343253771679155e-05,
+      "loss": 0.4455,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6837606837606838,
+      "grad_norm": 0.6209794878959656,
+      "learning_rate": 4.82097527348072e-05,
+      "loss": 1.0001,
+      "step": 1420
+    },
+    {
+      "epoch": 0.684242205368966,
+      "grad_norm": 0.47615864872932434,
+      "learning_rate": 4.8076377714811284e-05,
+      "loss": 0.5838,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6847237269772481,
+      "grad_norm": 0.4865884780883789,
+      "learning_rate": 4.7943129036221735e-05,
+      "loss": 0.419,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6852052485855302,
+      "grad_norm": 0.5061058402061462,
+      "learning_rate": 4.781000702326142e-05,
+      "loss": 0.2144,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6856867701938124,
+      "grad_norm": 0.9223092794418335,
+      "learning_rate": 4.767701199984497e-05,
+      "loss": 0.4185,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6861682918020946,
+      "grad_norm": 0.6437268853187561,
+      "learning_rate": 4.7544144289578066e-05,
+      "loss": 0.4223,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6866498134103768,
+      "grad_norm": 0.6153153777122498,
+      "learning_rate": 4.7411404215756594e-05,
+      "loss": 0.335,
+      "step": 1426
+    },
+    {
+      "epoch": 0.687131335018659,
+      "grad_norm": 1.2971034049987793,
+      "learning_rate": 4.7278792101365866e-05,
+      "loss": 0.7627,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6876128566269412,
+      "grad_norm": 0.7124690413475037,
+      "learning_rate": 4.714630826907985e-05,
+      "loss": 0.6546,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6880943782352233,
+      "grad_norm": 1.0929278135299683,
+      "learning_rate": 4.701395304126038e-05,
+      "loss": 0.8798,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6885758998435054,
+      "grad_norm": 0.6752956509590149,
+      "learning_rate": 4.6881726739956375e-05,
+      "loss": 0.5457,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6890574214517876,
+      "grad_norm": 0.9237945675849915,
+      "learning_rate": 4.6749629686902984e-05,
+      "loss": 1.2754,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6895389430600698,
+      "grad_norm": 0.6456115245819092,
+      "learning_rate": 4.661766220352097e-05,
+      "loss": 0.2924,
+      "step": 1432
+    },
+    {
+      "epoch": 0.690020464668352,
+      "grad_norm": 1.0163204669952393,
+      "learning_rate": 4.64858246109157e-05,
+      "loss": 1.2191,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6905019862766342,
+      "grad_norm": 0.9082103967666626,
+      "learning_rate": 4.63541172298766e-05,
+      "loss": 0.7933,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6909835078849164,
+      "grad_norm": 0.5109544992446899,
+      "learning_rate": 4.622254038087622e-05,
+      "loss": 0.3987,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6914650294931985,
+      "grad_norm": 0.88107830286026,
+      "learning_rate": 4.60910943840695e-05,
+      "loss": 0.586,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6919465511014807,
+      "grad_norm": 0.6991325616836548,
+      "learning_rate": 4.5959779559292985e-05,
+      "loss": 0.3088,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6924280727097628,
+      "grad_norm": 0.7649497985839844,
+      "learning_rate": 4.582859622606406e-05,
+      "loss": 0.5081,
+      "step": 1438
+    },
+    {
+      "epoch": 0.692909594318045,
+      "grad_norm": 1.080644965171814,
+      "learning_rate": 4.569754470358014e-05,
+      "loss": 0.8645,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6933911159263272,
+      "grad_norm": 0.8446236252784729,
+      "learning_rate": 4.556662531071796e-05,
+      "loss": 0.3782,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6938726375346094,
+      "grad_norm": 0.8029199242591858,
+      "learning_rate": 4.54358383660327e-05,
+      "loss": 0.6237,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6943541591428916,
+      "grad_norm": 0.8086987137794495,
+      "learning_rate": 4.530518418775733e-05,
+      "loss": 0.4498,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6948356807511737,
+      "grad_norm": 1.1346222162246704,
+      "learning_rate": 4.5174663093801674e-05,
+      "loss": 0.7626,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6953172023594559,
+      "grad_norm": 0.5477440357208252,
+      "learning_rate": 4.504427540175181e-05,
+      "loss": 0.3377,
+      "step": 1444
+    },
+    {
+      "epoch": 0.695798723967738,
+      "grad_norm": 0.37069225311279297,
+      "learning_rate": 4.491402142886922e-05,
+      "loss": 0.2361,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6962802455760202,
+      "grad_norm": 1.2140967845916748,
+      "learning_rate": 4.4783901492089984e-05,
+      "loss": 0.7769,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6967617671843024,
+      "grad_norm": 0.42369186878204346,
+      "learning_rate": 4.465391590802407e-05,
+      "loss": 0.403,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6972432887925846,
+      "grad_norm": 1.4413726329803467,
+      "learning_rate": 4.4524064992954516e-05,
+      "loss": 0.9888,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6977248104008668,
+      "grad_norm": 0.7362139821052551,
+      "learning_rate": 4.4394349062836736e-05,
+      "loss": 0.7057,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6982063320091489,
+      "grad_norm": 0.700434148311615,
+      "learning_rate": 4.4264768433297565e-05,
+      "loss": 0.385,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6986878536174311,
+      "grad_norm": 1.0879504680633545,
+      "learning_rate": 4.4135323419634766e-05,
+      "loss": 1.1829,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6991693752257132,
+      "grad_norm": 1.2308809757232666,
+      "learning_rate": 4.4006014336816035e-05,
+      "loss": 1.0978,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6996508968339954,
+      "grad_norm": 0.6011996865272522,
+      "learning_rate": 4.387684149947837e-05,
+      "loss": 0.362,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7001324184422776,
+      "grad_norm": 1.436378002166748,
+      "learning_rate": 4.374780522192726e-05,
+      "loss": 1.3376,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7006139400505598,
+      "grad_norm": 0.7376883625984192,
+      "learning_rate": 4.3618905818135805e-05,
+      "loss": 0.3268,
+      "step": 1455
+    },
+    {
+      "epoch": 0.701095461658842,
+      "grad_norm": 0.6043660044670105,
+      "learning_rate": 4.349014360174417e-05,
+      "loss": 0.6669,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7015769832671241,
+      "grad_norm": 0.7776713371276855,
+      "learning_rate": 4.336151888605871e-05,
+      "loss": 0.4488,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7020585048754063,
+      "grad_norm": 0.7803006172180176,
+      "learning_rate": 4.323303198405117e-05,
+      "loss": 0.8023,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7025400264836884,
+      "grad_norm": 0.7615090608596802,
+      "learning_rate": 4.310468320835796e-05,
+      "loss": 0.5186,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7030215480919706,
+      "grad_norm": 0.9116225242614746,
+      "learning_rate": 4.297647287127946e-05,
+      "loss": 0.3674,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7035030697002528,
+      "grad_norm": 0.9148691892623901,
+      "learning_rate": 4.284840128477913e-05,
+      "loss": 0.5605,
+      "step": 1461
+    },
+    {
+      "epoch": 0.703984591308535,
+      "grad_norm": 0.7424203157424927,
+      "learning_rate": 4.2720468760482854e-05,
+      "loss": 0.4504,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7044661129168172,
+      "grad_norm": 0.6610391139984131,
+      "learning_rate": 4.2592675609678135e-05,
+      "loss": 0.3636,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7049476345250993,
+      "grad_norm": 0.6278050541877747,
+      "learning_rate": 4.24650221433134e-05,
+      "loss": 0.529,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7054291561333815,
+      "grad_norm": 0.5832677483558655,
+      "learning_rate": 4.2337508671997086e-05,
+      "loss": 0.3057,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7059106777416636,
+      "grad_norm": 0.7446826696395874,
+      "learning_rate": 4.221013550599707e-05,
+      "loss": 0.3268,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7063921993499458,
+      "grad_norm": 0.9066232442855835,
+      "learning_rate": 4.208290295523984e-05,
+      "loss": 0.624,
+      "step": 1467
+    },
+    {
+      "epoch": 0.706873720958228,
+      "grad_norm": 1.0170692205429077,
+      "learning_rate": 4.1955811329309746e-05,
+      "loss": 0.6293,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7073552425665102,
+      "grad_norm": 0.7351876497268677,
+      "learning_rate": 4.182886093744813e-05,
+      "loss": 0.9277,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7078367641747924,
+      "grad_norm": 0.7814710140228271,
+      "learning_rate": 4.170205208855281e-05,
+      "loss": 0.4777,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7083182857830745,
+      "grad_norm": 0.8738664984703064,
+      "learning_rate": 4.157538509117714e-05,
+      "loss": 0.7923,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7087998073913567,
+      "grad_norm": 0.9012139439582825,
+      "learning_rate": 4.144886025352934e-05,
+      "loss": 0.6325,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7092813289996388,
+      "grad_norm": 1.0267746448516846,
+      "learning_rate": 4.13224778834717e-05,
+      "loss": 0.5129,
+      "step": 1473
+    },
+    {
+      "epoch": 0.709762850607921,
+      "grad_norm": 0.795587956905365,
+      "learning_rate": 4.1196238288519874e-05,
+      "loss": 0.4198,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7102443722162032,
+      "grad_norm": 0.9339183568954468,
+      "learning_rate": 4.107014177584211e-05,
+      "loss": 0.4369,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7107258938244854,
+      "grad_norm": 0.6808327436447144,
+      "learning_rate": 4.094418865225853e-05,
+      "loss": 0.3405,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7112074154327676,
+      "grad_norm": 0.9060590863227844,
+      "learning_rate": 4.081837922424027e-05,
+      "loss": 0.43,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7116889370410497,
+      "grad_norm": 0.4485381543636322,
+      "learning_rate": 4.069271379790891e-05,
+      "loss": 0.2424,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7121704586493319,
+      "grad_norm": 0.6939108967781067,
+      "learning_rate": 4.0567192679035636e-05,
+      "loss": 0.4309,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7126519802576141,
+      "grad_norm": 0.8871076107025146,
+      "learning_rate": 4.044181617304048e-05,
+      "loss": 0.3657,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7131335018658962,
+      "grad_norm": 0.6109139323234558,
+      "learning_rate": 4.03165845849916e-05,
+      "loss": 0.3523,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7136150234741784,
+      "grad_norm": 0.755155622959137,
+      "learning_rate": 4.019149821960455e-05,
+      "loss": 0.5116,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7140965450824606,
+      "grad_norm": 0.684121310710907,
+      "learning_rate": 4.006655738124152e-05,
+      "loss": 0.4508,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7145780666907428,
+      "grad_norm": 0.7652056813240051,
+      "learning_rate": 3.9941762373910586e-05,
+      "loss": 0.4199,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7150595882990249,
+      "grad_norm": 0.8583548069000244,
+      "learning_rate": 3.9817113501265016e-05,
+      "loss": 0.6258,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7155411099073071,
+      "grad_norm": 0.8159501552581787,
+      "learning_rate": 3.9692611066602516e-05,
+      "loss": 0.6959,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7160226315155893,
+      "grad_norm": 0.7192055583000183,
+      "learning_rate": 3.956825537286436e-05,
+      "loss": 0.2499,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7165041531238714,
+      "grad_norm": 0.5423712134361267,
+      "learning_rate": 3.944404672263494e-05,
+      "loss": 0.4917,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7169856747321536,
+      "grad_norm": 0.9170364737510681,
+      "learning_rate": 3.931998541814069e-05,
+      "loss": 0.898,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7174671963404358,
+      "grad_norm": 1.1975452899932861,
+      "learning_rate": 3.919607176124966e-05,
+      "loss": 1.2809,
+      "step": 1490
+    },
+    {
+      "epoch": 0.717948717948718,
+      "grad_norm": 0.5959516167640686,
+      "learning_rate": 3.9072306053470566e-05,
+      "loss": 0.3478,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7184302395570001,
+      "grad_norm": 0.7946900129318237,
+      "learning_rate": 3.8948688595952164e-05,
+      "loss": 0.1685,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7189117611652823,
+      "grad_norm": 1.519437551498413,
+      "learning_rate": 3.882521968948246e-05,
+      "loss": 0.7702,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7193932827735645,
+      "grad_norm": 0.7796205878257751,
+      "learning_rate": 3.8701899634488014e-05,
+      "loss": 0.6952,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7198748043818466,
+      "grad_norm": 0.7571779489517212,
+      "learning_rate": 3.857872873103322e-05,
+      "loss": 0.3655,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7203563259901288,
+      "grad_norm": 0.4898691177368164,
+      "learning_rate": 3.8455707278819507e-05,
+      "loss": 0.371,
+      "step": 1496
+    },
+    {
+      "epoch": 0.720837847598411,
+      "grad_norm": 0.4889354407787323,
+      "learning_rate": 3.833283557718471e-05,
+      "loss": 0.2633,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7213193692066932,
+      "grad_norm": 0.5309864282608032,
+      "learning_rate": 3.821011392510228e-05,
+      "loss": 0.2295,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7218008908149753,
+      "grad_norm": 0.5448015928268433,
+      "learning_rate": 3.808754262118046e-05,
+      "loss": 0.2642,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7222824124232575,
+      "grad_norm": 0.9109936952590942,
+      "learning_rate": 3.796512196366182e-05,
+      "loss": 0.3738,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7227639340315397,
+      "grad_norm": 0.8303773999214172,
+      "learning_rate": 3.784285225042229e-05,
+      "loss": 0.7826,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7232454556398218,
+      "grad_norm": 0.9534183144569397,
+      "learning_rate": 3.772073377897052e-05,
+      "loss": 0.6641,
+      "step": 1502
+    },
+    {
+      "epoch": 0.723726977248104,
+      "grad_norm": 1.6140865087509155,
+      "learning_rate": 3.7598766846447184e-05,
+      "loss": 0.5608,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7242084988563862,
+      "grad_norm": 0.47576579451560974,
+      "learning_rate": 3.747695174962423e-05,
+      "loss": 0.4865,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7246900204646683,
+      "grad_norm": 0.8235217928886414,
+      "learning_rate": 3.7355288784904116e-05,
+      "loss": 0.4982,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7251715420729505,
+      "grad_norm": 0.9987104535102844,
+      "learning_rate": 3.7233778248319176e-05,
+      "loss": 0.4883,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7256530636812327,
+      "grad_norm": 0.8933877348899841,
+      "learning_rate": 3.7112420435530845e-05,
+      "loss": 0.511,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7261345852895149,
+      "grad_norm": 0.5606909394264221,
+      "learning_rate": 3.69912156418289e-05,
+      "loss": 0.2032,
+      "step": 1508
+    },
+    {
+      "epoch": 0.726616106897797,
+      "grad_norm": 0.7086384296417236,
+      "learning_rate": 3.687016416213084e-05,
+      "loss": 0.5728,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7270976285060792,
+      "grad_norm": 0.6372523903846741,
+      "learning_rate": 3.674926629098113e-05,
+      "loss": 0.1722,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7275791501143614,
+      "grad_norm": 0.7619569301605225,
+      "learning_rate": 3.6628522322550394e-05,
+      "loss": 0.2178,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7280606717226435,
+      "grad_norm": 0.811999499797821,
+      "learning_rate": 3.6507932550634846e-05,
+      "loss": 0.5692,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7285421933309257,
+      "grad_norm": 0.6715248227119446,
+      "learning_rate": 3.638749726865552e-05,
+      "loss": 0.1933,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7290237149392079,
+      "grad_norm": 0.4963141679763794,
+      "learning_rate": 3.6267216769657485e-05,
+      "loss": 0.4058,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7295052365474901,
+      "grad_norm": 0.7091789245605469,
+      "learning_rate": 3.6147091346309224e-05,
+      "loss": 0.846,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7299867581557722,
+      "grad_norm": 0.6115385293960571,
+      "learning_rate": 3.602712129090189e-05,
+      "loss": 0.3185,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7304682797640544,
+      "grad_norm": 0.5440830588340759,
+      "learning_rate": 3.590730689534857e-05,
+      "loss": 0.3759,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7309498013723366,
+      "grad_norm": 0.5895819664001465,
+      "learning_rate": 3.578764845118362e-05,
+      "loss": 0.3927,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7314313229806187,
+      "grad_norm": 1.0050193071365356,
+      "learning_rate": 3.566814624956194e-05,
+      "loss": 0.4598,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7319128445889009,
+      "grad_norm": 0.67363440990448,
+      "learning_rate": 3.554880058125819e-05,
+      "loss": 0.5574,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7323943661971831,
+      "grad_norm": 1.1256099939346313,
+      "learning_rate": 3.5429611736666235e-05,
+      "loss": 0.6292,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7328758878054653,
+      "grad_norm": 0.5500608086585999,
+      "learning_rate": 3.53105800057983e-05,
+      "loss": 0.2614,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7333574094137475,
+      "grad_norm": 0.6335020065307617,
+      "learning_rate": 3.519170567828435e-05,
+      "loss": 0.5855,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7338389310220296,
+      "grad_norm": 1.045906901359558,
+      "learning_rate": 3.507298904337134e-05,
+      "loss": 0.763,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7343204526303118,
+      "grad_norm": 0.7786596417427063,
+      "learning_rate": 3.495443038992253e-05,
+      "loss": 0.4323,
+      "step": 1525
+    },
+    {
+      "epoch": 0.734801974238594,
+      "grad_norm": 0.8979135155677795,
+      "learning_rate": 3.4836030006416775e-05,
+      "loss": 0.424,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7352834958468761,
+      "grad_norm": 0.8464924097061157,
+      "learning_rate": 3.471778818094785e-05,
+      "loss": 0.6309,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7357650174551583,
+      "grad_norm": 0.38327470421791077,
+      "learning_rate": 3.459970520122364e-05,
+      "loss": 0.2277,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7362465390634405,
+      "grad_norm": 1.0083869695663452,
+      "learning_rate": 3.44817813545656e-05,
+      "loss": 0.4736,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7367280606717227,
+      "grad_norm": 0.7795249819755554,
+      "learning_rate": 3.4364016927907974e-05,
+      "loss": 0.329,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7372095822800048,
+      "grad_norm": 0.6777287125587463,
+      "learning_rate": 3.424641220779711e-05,
+      "loss": 0.5657,
+      "step": 1531
+    },
+    {
+      "epoch": 0.737691103888287,
+      "grad_norm": 0.6228256821632385,
+      "learning_rate": 3.412896748039067e-05,
+      "loss": 0.4603,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7381726254965691,
+      "grad_norm": 0.5707159042358398,
+      "learning_rate": 3.401168303145713e-05,
+      "loss": 0.4364,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7386541471048513,
+      "grad_norm": 0.6516470909118652,
+      "learning_rate": 3.3894559146374924e-05,
+      "loss": 0.3653,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7391356687131335,
+      "grad_norm": 0.8824191093444824,
+      "learning_rate": 3.37775961101318e-05,
+      "loss": 0.8059,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7396171903214157,
+      "grad_norm": 0.6240821480751038,
+      "learning_rate": 3.366079420732413e-05,
+      "loss": 0.5367,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7400987119296979,
+      "grad_norm": 1.1601349115371704,
+      "learning_rate": 3.3544153722156216e-05,
+      "loss": 0.5936,
+      "step": 1537
+    },
+    {
+      "epoch": 0.74058023353798,
+      "grad_norm": 0.4957769811153412,
+      "learning_rate": 3.3427674938439594e-05,
+      "loss": 0.5746,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7410617551462622,
+      "grad_norm": 0.3493504226207733,
+      "learning_rate": 3.3311358139592317e-05,
+      "loss": 0.3965,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7415432767545443,
+      "grad_norm": 0.7727856636047363,
+      "learning_rate": 3.319520360863837e-05,
+      "loss": 0.5924,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7420247983628265,
+      "grad_norm": 0.8370104432106018,
+      "learning_rate": 3.3079211628206854e-05,
+      "loss": 0.3911,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7425063199711087,
+      "grad_norm": 1.2629445791244507,
+      "learning_rate": 3.296338248053129e-05,
+      "loss": 0.3831,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7429878415793909,
+      "grad_norm": 0.7771987915039062,
+      "learning_rate": 3.2847716447449096e-05,
+      "loss": 0.5615,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7434693631876731,
+      "grad_norm": 0.7107818722724915,
+      "learning_rate": 3.2732213810400745e-05,
+      "loss": 0.4814,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7439508847959552,
+      "grad_norm": 0.7308077812194824,
+      "learning_rate": 3.261687485042915e-05,
+      "loss": 0.4592,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7444324064042374,
+      "grad_norm": 1.0506657361984253,
+      "learning_rate": 3.250169984817897e-05,
+      "loss": 0.5339,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7449139280125195,
+      "grad_norm": 1.354854702949524,
+      "learning_rate": 3.238668908389586e-05,
+      "loss": 0.7969,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7453954496208017,
+      "grad_norm": 0.7732512354850769,
+      "learning_rate": 3.227184283742591e-05,
+      "loss": 0.786,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7458769712290839,
+      "grad_norm": 1.486000895500183,
+      "learning_rate": 3.215716138821488e-05,
+      "loss": 0.3588,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7463584928373661,
+      "grad_norm": 0.7014599442481995,
+      "learning_rate": 3.204264501530756e-05,
+      "loss": 0.5279,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7468400144456483,
+      "grad_norm": 1.0592234134674072,
+      "learning_rate": 3.192829399734706e-05,
+      "loss": 0.7584,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7473215360539304,
+      "grad_norm": 0.7656548023223877,
+      "learning_rate": 3.181410861257413e-05,
+      "loss": 0.8036,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7478030576622126,
+      "grad_norm": 2.293597936630249,
+      "learning_rate": 3.170008913882656e-05,
+      "loss": 0.6918,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7482845792704947,
+      "grad_norm": 0.6488149762153625,
+      "learning_rate": 3.1586235853538325e-05,
+      "loss": 0.7967,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7487661008787769,
+      "grad_norm": 1.1204566955566406,
+      "learning_rate": 3.1472549033739126e-05,
+      "loss": 0.4146,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7492476224870591,
+      "grad_norm": 0.6003812551498413,
+      "learning_rate": 3.1359028956053615e-05,
+      "loss": 0.3162,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7497291440953413,
+      "grad_norm": 1.8639508485794067,
+      "learning_rate": 3.1245675896700685e-05,
+      "loss": 1.1739,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7502106657036235,
+      "grad_norm": 0.7486677169799805,
+      "learning_rate": 3.113249013149284e-05,
+      "loss": 0.1952,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7506921873119056,
+      "grad_norm": 1.239945888519287,
+      "learning_rate": 3.101947193583557e-05,
+      "loss": 0.754,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7511737089201878,
+      "grad_norm": 1.0496931076049805,
+      "learning_rate": 3.0906621584726546e-05,
+      "loss": 0.7861,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7516552305284699,
+      "grad_norm": 0.981391191482544,
+      "learning_rate": 3.079393935275513e-05,
+      "loss": 0.4328,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7521367521367521,
+      "grad_norm": 1.4165148735046387,
+      "learning_rate": 3.068142551410155e-05,
+      "loss": 0.398,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7526182737450343,
+      "grad_norm": 0.9278653860092163,
+      "learning_rate": 3.0569080342536347e-05,
+      "loss": 0.6272,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7530997953533165,
+      "grad_norm": 1.465111255645752,
+      "learning_rate": 3.0456904111419572e-05,
+      "loss": 1.0547,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7535813169615987,
+      "grad_norm": 0.7694311738014221,
+      "learning_rate": 3.034489709370033e-05,
+      "loss": 0.7232,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7540628385698809,
+      "grad_norm": 0.9846645593643188,
+      "learning_rate": 3.0233059561915855e-05,
+      "loss": 0.5923,
+      "step": 1566
+    },
+    {
+      "epoch": 0.754544360178163,
+      "grad_norm": 0.9460967779159546,
+      "learning_rate": 3.01213917881911e-05,
+      "loss": 0.8232,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7550258817864451,
+      "grad_norm": 0.761855959892273,
+      "learning_rate": 3.0009894044237907e-05,
+      "loss": 0.6175,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7555074033947273,
+      "grad_norm": 0.5922994017601013,
+      "learning_rate": 2.9898566601354418e-05,
+      "loss": 0.6428,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7559889250030095,
+      "grad_norm": 0.5083173513412476,
+      "learning_rate": 2.9787409730424374e-05,
+      "loss": 0.4254,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7564704466112917,
+      "grad_norm": 0.6428760886192322,
+      "learning_rate": 2.96764237019165e-05,
+      "loss": 0.4674,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7569519682195739,
+      "grad_norm": 0.906851053237915,
+      "learning_rate": 2.9565608785883815e-05,
+      "loss": 0.9182,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7574334898278561,
+      "grad_norm": 0.6919524073600769,
+      "learning_rate": 2.9454965251962973e-05,
+      "loss": 0.7283,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7579150114361382,
+      "grad_norm": 0.5480821132659912,
+      "learning_rate": 2.9344493369373637e-05,
+      "loss": 0.4671,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7583965330444203,
+      "grad_norm": 0.929789125919342,
+      "learning_rate": 2.9234193406917833e-05,
+      "loss": 0.5857,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7588780546527025,
+      "grad_norm": 1.0483143329620361,
+      "learning_rate": 2.912406563297916e-05,
+      "loss": 0.7638,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7593595762609847,
+      "grad_norm": 0.7521671056747437,
+      "learning_rate": 2.901411031552236e-05,
+      "loss": 0.5069,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7598410978692669,
+      "grad_norm": 0.6959591507911682,
+      "learning_rate": 2.8904327722092495e-05,
+      "loss": 0.898,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7603226194775491,
+      "grad_norm": 0.9434571862220764,
+      "learning_rate": 2.879471811981437e-05,
+      "loss": 0.8069,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7608041410858313,
+      "grad_norm": 0.9961397051811218,
+      "learning_rate": 2.868528177539187e-05,
+      "loss": 0.8701,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7612856626941134,
+      "grad_norm": 0.8834352493286133,
+      "learning_rate": 2.8576018955107285e-05,
+      "loss": 0.7611,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7617671843023955,
+      "grad_norm": 0.7454970479011536,
+      "learning_rate": 2.8466929924820705e-05,
+      "loss": 0.9264,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7622487059106777,
+      "grad_norm": 0.5560579299926758,
+      "learning_rate": 2.8358014949969334e-05,
+      "loss": 0.6513,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7627302275189599,
+      "grad_norm": 0.793477475643158,
+      "learning_rate": 2.8249274295566864e-05,
+      "loss": 0.4191,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7632117491272421,
+      "grad_norm": 0.7971614003181458,
+      "learning_rate": 2.8140708226202884e-05,
+      "loss": 0.607,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7636932707355243,
+      "grad_norm": 0.9233664870262146,
+      "learning_rate": 2.803231700604204e-05,
+      "loss": 0.3923,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7641747923438065,
+      "grad_norm": 0.8491899967193604,
+      "learning_rate": 2.7924100898823702e-05,
+      "loss": 0.6279,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7646563139520886,
+      "grad_norm": 0.3680364787578583,
+      "learning_rate": 2.7816060167861002e-05,
+      "loss": 0.4774,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7651378355603707,
+      "grad_norm": 0.42586857080459595,
+      "learning_rate": 2.7708195076040445e-05,
+      "loss": 0.234,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7656193571686529,
+      "grad_norm": 0.4964386820793152,
+      "learning_rate": 2.760050588582114e-05,
+      "loss": 0.7201,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7661008787769351,
+      "grad_norm": 0.886875569820404,
+      "learning_rate": 2.749299285923417e-05,
+      "loss": 0.6147,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7665824003852173,
+      "grad_norm": 0.7478306293487549,
+      "learning_rate": 2.7385656257881997e-05,
+      "loss": 0.5431,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7670639219934995,
+      "grad_norm": 0.4899425208568573,
+      "learning_rate": 2.7278496342937788e-05,
+      "loss": 0.1444,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7675454436017817,
+      "grad_norm": 1.0289299488067627,
+      "learning_rate": 2.717151337514482e-05,
+      "loss": 0.5972,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7680269652100638,
+      "grad_norm": 0.7863545417785645,
+      "learning_rate": 2.7064707614815776e-05,
+      "loss": 0.4758,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7685084868183459,
+      "grad_norm": 0.6140004992485046,
+      "learning_rate": 2.6958079321832185e-05,
+      "loss": 0.2617,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7689900084266281,
+      "grad_norm": 0.5038211941719055,
+      "learning_rate": 2.6851628755643776e-05,
+      "loss": 0.1513,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7694715300349103,
+      "grad_norm": 0.7092880010604858,
+      "learning_rate": 2.6745356175267765e-05,
+      "loss": 0.3773,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7699530516431925,
+      "grad_norm": 0.9827279448509216,
+      "learning_rate": 2.6639261839288343e-05,
+      "loss": 0.5516,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7704345732514747,
+      "grad_norm": 0.9941007494926453,
+      "learning_rate": 2.6533346005855987e-05,
+      "loss": 1.1843,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7709160948597569,
+      "grad_norm": 0.5808009505271912,
+      "learning_rate": 2.6427608932686843e-05,
+      "loss": 0.3892,
+      "step": 1601
+    },
+    {
+      "epoch": 0.771397616468039,
+      "grad_norm": 0.8638581037521362,
+      "learning_rate": 2.6322050877062064e-05,
+      "loss": 0.3914,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7718791380763211,
+      "grad_norm": 0.8344993591308594,
+      "learning_rate": 2.6216672095827266e-05,
+      "loss": 0.6955,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7723606596846033,
+      "grad_norm": 0.7511135339736938,
+      "learning_rate": 2.6111472845391827e-05,
+      "loss": 0.6755,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7728421812928855,
+      "grad_norm": 0.6375028491020203,
+      "learning_rate": 2.6006453381728236e-05,
+      "loss": 0.1404,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7733237029011677,
+      "grad_norm": 0.7611637711524963,
+      "learning_rate": 2.5901613960371585e-05,
+      "loss": 0.377,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7738052245094499,
+      "grad_norm": 2.2889697551727295,
+      "learning_rate": 2.5796954836418884e-05,
+      "loss": 0.8719,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7742867461177321,
+      "grad_norm": 0.8444758057594299,
+      "learning_rate": 2.569247626452842e-05,
+      "loss": 0.5739,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7747682677260143,
+      "grad_norm": 1.26003897190094,
+      "learning_rate": 2.558817849891918e-05,
+      "loss": 0.433,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7752497893342963,
+      "grad_norm": 0.647497296333313,
+      "learning_rate": 2.548406179337015e-05,
+      "loss": 0.547,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7757313109425785,
+      "grad_norm": 0.31829091906547546,
+      "learning_rate": 2.5380126401219807e-05,
+      "loss": 0.2283,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7762128325508607,
+      "grad_norm": 0.9248780012130737,
+      "learning_rate": 2.527637257536547e-05,
+      "loss": 0.3302,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7766943541591429,
+      "grad_norm": 0.8120497465133667,
+      "learning_rate": 2.517280056826262e-05,
+      "loss": 0.4713,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7771758757674251,
+      "grad_norm": 0.47866806387901306,
+      "learning_rate": 2.5069410631924385e-05,
+      "loss": 0.2786,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7776573973757073,
+      "grad_norm": 0.48474082350730896,
+      "learning_rate": 2.4966203017920818e-05,
+      "loss": 0.2915,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7781389189839895,
+      "grad_norm": 0.8543782830238342,
+      "learning_rate": 2.4863177977378392e-05,
+      "loss": 0.795,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7786204405922715,
+      "grad_norm": 0.4673691987991333,
+      "learning_rate": 2.4760335760979312e-05,
+      "loss": 0.6174,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7791019622005537,
+      "grad_norm": 0.7312822937965393,
+      "learning_rate": 2.4657676618960944e-05,
+      "loss": 0.2589,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7795834838088359,
+      "grad_norm": 0.8223809003829956,
+      "learning_rate": 2.455520080111522e-05,
+      "loss": 0.7372,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7800650054171181,
+      "grad_norm": 0.8796128034591675,
+      "learning_rate": 2.4452908556787912e-05,
+      "loss": 0.5219,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7805465270254003,
+      "grad_norm": 0.8800025582313538,
+      "learning_rate": 2.4350800134878203e-05,
+      "loss": 0.5718,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7810280486336825,
+      "grad_norm": 0.3794774115085602,
+      "learning_rate": 2.4248875783837987e-05,
+      "loss": 0.1713,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7815095702419647,
+      "grad_norm": 1.717668890953064,
+      "learning_rate": 2.414713575167129e-05,
+      "loss": 0.2789,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7819910918502467,
+      "grad_norm": 0.5892308354377747,
+      "learning_rate": 2.4045580285933557e-05,
+      "loss": 0.3826,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7824726134585289,
+      "grad_norm": 1.4412480592727661,
+      "learning_rate": 2.3944209633731242e-05,
+      "loss": 0.9213,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7829541350668111,
+      "grad_norm": 0.9920421838760376,
+      "learning_rate": 2.3843024041721053e-05,
+      "loss": 0.2247,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7834356566750933,
+      "grad_norm": 1.3514714241027832,
+      "learning_rate": 2.3742023756109456e-05,
+      "loss": 1.0385,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7839171782833755,
+      "grad_norm": 0.6163105368614197,
+      "learning_rate": 2.3641209022651976e-05,
+      "loss": 0.6583,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7843986998916577,
+      "grad_norm": 0.7857782244682312,
+      "learning_rate": 2.3540580086652675e-05,
+      "loss": 0.4282,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7848802214999399,
+      "grad_norm": 0.8993891477584839,
+      "learning_rate": 2.344013719296353e-05,
+      "loss": 0.35,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7853617431082219,
+      "grad_norm": 0.7018194794654846,
+      "learning_rate": 2.3339880585983842e-05,
+      "loss": 0.3266,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7858432647165041,
+      "grad_norm": 0.35083359479904175,
+      "learning_rate": 2.3239810509659597e-05,
+      "loss": 0.3474,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7863247863247863,
+      "grad_norm": 0.9153875708580017,
+      "learning_rate": 2.313992720748295e-05,
+      "loss": 0.5287,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7868063079330685,
+      "grad_norm": 0.8885396718978882,
+      "learning_rate": 2.304023092249159e-05,
+      "loss": 0.7783,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7872878295413507,
+      "grad_norm": 1.1704275608062744,
+      "learning_rate": 2.2940721897268136e-05,
+      "loss": 0.6887,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7877693511496329,
+      "grad_norm": 0.5215083360671997,
+      "learning_rate": 2.2841400373939592e-05,
+      "loss": 0.4906,
+      "step": 1636
+    },
+    {
+      "epoch": 0.788250872757915,
+      "grad_norm": 0.9289473295211792,
+      "learning_rate": 2.274226659417671e-05,
+      "loss": 1.1291,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7887323943661971,
+      "grad_norm": 0.5596996545791626,
+      "learning_rate": 2.2643320799193402e-05,
+      "loss": 0.2618,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7892139159744793,
+      "grad_norm": 0.8898619413375854,
+      "learning_rate": 2.2544563229746218e-05,
+      "loss": 0.6691,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7896954375827615,
+      "grad_norm": 0.943631649017334,
+      "learning_rate": 2.2445994126133708e-05,
+      "loss": 1.1104,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7901769591910437,
+      "grad_norm": 0.8219139575958252,
+      "learning_rate": 2.234761372819577e-05,
+      "loss": 0.7676,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7906584807993259,
+      "grad_norm": 0.8116326928138733,
+      "learning_rate": 2.2249422275313214e-05,
+      "loss": 0.434,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7911400024076081,
+      "grad_norm": 0.5445429682731628,
+      "learning_rate": 2.215142000640714e-05,
+      "loss": 0.252,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7916215240158903,
+      "grad_norm": 0.751610279083252,
+      "learning_rate": 2.2053607159938195e-05,
+      "loss": 0.5864,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7921030456241723,
+      "grad_norm": 0.7384713292121887,
+      "learning_rate": 2.1955983973906236e-05,
+      "loss": 0.3833,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7925845672324545,
+      "grad_norm": 0.6013317108154297,
+      "learning_rate": 2.1858550685849578e-05,
+      "loss": 0.5739,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7930660888407367,
+      "grad_norm": 1.1576766967773438,
+      "learning_rate": 2.17613075328445e-05,
+      "loss": 0.425,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7935476104490189,
+      "grad_norm": 1.0560667514801025,
+      "learning_rate": 2.1664254751504642e-05,
+      "loss": 0.5623,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7940291320573011,
+      "grad_norm": 0.37843796610832214,
+      "learning_rate": 2.1567392577980393e-05,
+      "loss": 0.1924,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7945106536655833,
+      "grad_norm": 1.0658668279647827,
+      "learning_rate": 2.1470721247958404e-05,
+      "loss": 1.2051,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7949921752738655,
+      "grad_norm": 1.119051456451416,
+      "learning_rate": 2.137424099666091e-05,
+      "loss": 0.7042,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7954736968821476,
+      "grad_norm": 0.3540979027748108,
+      "learning_rate": 2.1277952058845284e-05,
+      "loss": 0.1893,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7959552184904297,
+      "grad_norm": 0.6795416474342346,
+      "learning_rate": 2.118185466880327e-05,
+      "loss": 1.0059,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7964367400987119,
+      "grad_norm": 0.9295594692230225,
+      "learning_rate": 2.1085949060360654e-05,
+      "loss": 0.6773,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7969182617069941,
+      "grad_norm": 0.669072687625885,
+      "learning_rate": 2.0990235466876517e-05,
+      "loss": 0.5336,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7973997833152763,
+      "grad_norm": 0.5881422162055969,
+      "learning_rate": 2.089471412124274e-05,
+      "loss": 0.5242,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7978813049235585,
+      "grad_norm": 0.7635432481765747,
+      "learning_rate": 2.079938525588342e-05,
+      "loss": 0.2237,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7983628265318407,
+      "grad_norm": 0.6804701089859009,
+      "learning_rate": 2.0704249102754324e-05,
+      "loss": 0.6416,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7988443481401228,
+      "grad_norm": 0.4592357575893402,
+      "learning_rate": 2.0609305893342278e-05,
+      "loss": 0.4071,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7993258697484049,
+      "grad_norm": 0.7150890827178955,
+      "learning_rate": 2.0514555858664663e-05,
+      "loss": 0.9323,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7998073913566871,
+      "grad_norm": 0.5522563457489014,
+      "learning_rate": 2.0419999229268805e-05,
+      "loss": 0.3048,
+      "step": 1661
+    },
+    {
+      "epoch": 0.8002889129649693,
+      "grad_norm": 1.1032781600952148,
+      "learning_rate": 2.032563623523147e-05,
+      "loss": 0.6891,
+      "step": 1662
+    },
+    {
+      "epoch": 0.8007704345732515,
+      "grad_norm": 0.8238122463226318,
+      "learning_rate": 2.0231467106158186e-05,
+      "loss": 0.281,
+      "step": 1663
+    },
+    {
+      "epoch": 0.8012519561815337,
+      "grad_norm": 0.7528010010719299,
+      "learning_rate": 2.0137492071182863e-05,
+      "loss": 0.6063,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8017334777898159,
+      "grad_norm": 1.0722914934158325,
+      "learning_rate": 2.0043711358967043e-05,
+      "loss": 1.0133,
+      "step": 1665
+    },
+    {
+      "epoch": 0.802214999398098,
+      "grad_norm": 0.44027334451675415,
+      "learning_rate": 1.9950125197699508e-05,
+      "loss": 0.2833,
+      "step": 1666
+    },
+    {
+      "epoch": 0.8026965210063801,
+      "grad_norm": 0.9600662589073181,
+      "learning_rate": 1.985673381509565e-05,
+      "loss": 0.3938,
+      "step": 1667
+    },
+    {
+      "epoch": 0.8031780426146623,
+      "grad_norm": 0.6282616853713989,
+      "learning_rate": 1.9763537438396894e-05,
+      "loss": 0.6348,
+      "step": 1668
+    },
+    {
+      "epoch": 0.8036595642229445,
+      "grad_norm": 0.5281164646148682,
+      "learning_rate": 1.96705362943702e-05,
+      "loss": 0.5014,
+      "step": 1669
+    },
+    {
+      "epoch": 0.8041410858312267,
+      "grad_norm": 0.6858304142951965,
+      "learning_rate": 1.9577730609307454e-05,
+      "loss": 0.5066,
+      "step": 1670
+    },
+    {
+      "epoch": 0.8046226074395089,
+      "grad_norm": 0.6477140784263611,
+      "learning_rate": 1.9485120609024975e-05,
+      "loss": 0.445,
+      "step": 1671
+    },
+    {
+      "epoch": 0.805104129047791,
+      "grad_norm": 0.6046128273010254,
+      "learning_rate": 1.9392706518862935e-05,
+      "loss": 0.2247,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8055856506560732,
+      "grad_norm": 0.8289013504981995,
+      "learning_rate": 1.9300488563684804e-05,
+      "loss": 0.2918,
+      "step": 1673
+    },
+    {
+      "epoch": 0.8060671722643553,
+      "grad_norm": 0.46251946687698364,
+      "learning_rate": 1.920846696787684e-05,
+      "loss": 0.6147,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8065486938726375,
+      "grad_norm": 0.5980027914047241,
+      "learning_rate": 1.9116641955347446e-05,
+      "loss": 0.3733,
+      "step": 1675
+    },
+    {
+      "epoch": 0.8070302154809197,
+      "grad_norm": 0.6853753924369812,
+      "learning_rate": 1.9025013749526767e-05,
+      "loss": 0.5122,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8075117370892019,
+      "grad_norm": 1.0705738067626953,
+      "learning_rate": 1.8933582573366036e-05,
+      "loss": 0.9123,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8079932586974841,
+      "grad_norm": 0.9407148957252502,
+      "learning_rate": 1.8842348649337116e-05,
+      "loss": 0.9415,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8084747803057662,
+      "grad_norm": 1.0631650686264038,
+      "learning_rate": 1.875131219943187e-05,
+      "loss": 1.1339,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8089563019140484,
+      "grad_norm": 0.6693783402442932,
+      "learning_rate": 1.8660473445161663e-05,
+      "loss": 0.2972,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8094378235223305,
+      "grad_norm": 0.8969826102256775,
+      "learning_rate": 1.856983260755686e-05,
+      "loss": 0.5933,
+      "step": 1681
+    },
+    {
+      "epoch": 0.8099193451306127,
+      "grad_norm": 1.0234383344650269,
+      "learning_rate": 1.8479389907166223e-05,
+      "loss": 0.3956,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8104008667388949,
+      "grad_norm": 0.9167569875717163,
+      "learning_rate": 1.8389145564056387e-05,
+      "loss": 0.4642,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8108823883471771,
+      "grad_norm": 0.5638496279716492,
+      "learning_rate": 1.829909979781137e-05,
+      "loss": 0.3361,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8113639099554593,
+      "grad_norm": 0.7476749420166016,
+      "learning_rate": 1.820925282753201e-05,
+      "loss": 0.5014,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8118454315637414,
+      "grad_norm": 0.7180486917495728,
+      "learning_rate": 1.8119604871835437e-05,
+      "loss": 0.5145,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8123269531720236,
+      "grad_norm": 0.5081984996795654,
+      "learning_rate": 1.8030156148854492e-05,
+      "loss": 0.3872,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8128084747803057,
+      "grad_norm": 0.9229518175125122,
+      "learning_rate": 1.7940906876237284e-05,
+      "loss": 0.3678,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8132899963885879,
+      "grad_norm": 0.8328425288200378,
+      "learning_rate": 1.78518572711466e-05,
+      "loss": 0.5457,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8137715179968701,
+      "grad_norm": 1.213712215423584,
+      "learning_rate": 1.776300755025939e-05,
+      "loss": 1.061,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8142530396051523,
+      "grad_norm": 1.0858510732650757,
+      "learning_rate": 1.767435792976626e-05,
+      "loss": 0.7261,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8147345612134345,
+      "grad_norm": 0.6739503145217896,
+      "learning_rate": 1.7585908625370905e-05,
+      "loss": 0.5279,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8152160828217166,
+      "grad_norm": 0.8865543007850647,
+      "learning_rate": 1.749765985228963e-05,
+      "loss": 0.9244,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8156976044299988,
+      "grad_norm": 0.7478240728378296,
+      "learning_rate": 1.740961182525077e-05,
+      "loss": 0.2868,
+      "step": 1694
+    },
+    {
+      "epoch": 0.816179126038281,
+      "grad_norm": 1.0323337316513062,
+      "learning_rate": 1.7321764758494252e-05,
+      "loss": 1.0011,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8166606476465631,
+      "grad_norm": 0.9194225668907166,
+      "learning_rate": 1.7234118865770987e-05,
+      "loss": 0.9113,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8171421692548453,
+      "grad_norm": 0.8026208877563477,
+      "learning_rate": 1.7146674360342373e-05,
+      "loss": 0.4512,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8176236908631275,
+      "grad_norm": 0.4755818545818329,
+      "learning_rate": 1.7059431454979824e-05,
+      "loss": 0.694,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8181052124714097,
+      "grad_norm": 1.2281359434127808,
+      "learning_rate": 1.6972390361964195e-05,
+      "loss": 1.1019,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8185867340796918,
+      "grad_norm": 0.7719329595565796,
+      "learning_rate": 1.688555129308531e-05,
+      "loss": 0.5681,
+      "step": 1700
+    },
+    {
+      "epoch": 0.819068255687974,
+      "grad_norm": 1.0498366355895996,
+      "learning_rate": 1.6798914459641434e-05,
+      "loss": 0.7296,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8195497772962562,
+      "grad_norm": 0.6994045376777649,
+      "learning_rate": 1.6712480072438662e-05,
+      "loss": 0.9328,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8200312989045383,
+      "grad_norm": 0.9230305552482605,
+      "learning_rate": 1.6626248341790596e-05,
+      "loss": 1.1862,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8205128205128205,
+      "grad_norm": 0.6558325886726379,
+      "learning_rate": 1.6540219477517684e-05,
+      "loss": 0.2061,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8209943421211027,
+      "grad_norm": 0.6145660877227783,
+      "learning_rate": 1.6454393688946767e-05,
+      "loss": 0.4734,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8214758637293849,
+      "grad_norm": 0.8060041666030884,
+      "learning_rate": 1.6368771184910557e-05,
+      "loss": 0.6261,
+      "step": 1706
+    },
+    {
+      "epoch": 0.821957385337667,
+      "grad_norm": 0.7207950949668884,
+      "learning_rate": 1.6283352173747145e-05,
+      "loss": 0.7611,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8224389069459492,
+      "grad_norm": 0.5801020860671997,
+      "learning_rate": 1.619813686329946e-05,
+      "loss": 0.7966,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8229204285542314,
+      "grad_norm": 0.5637111067771912,
+      "learning_rate": 1.611312546091476e-05,
+      "loss": 0.7279,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8234019501625135,
+      "grad_norm": 1.224278211593628,
+      "learning_rate": 1.6028318173444202e-05,
+      "loss": 1.3064,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8238834717707957,
+      "grad_norm": 0.832797646522522,
+      "learning_rate": 1.594371520724226e-05,
+      "loss": 0.6784,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8243649933790779,
+      "grad_norm": 0.7406448721885681,
+      "learning_rate": 1.5859316768166244e-05,
+      "loss": 0.5418,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8248465149873601,
+      "grad_norm": 0.9135259985923767,
+      "learning_rate": 1.5775123061575836e-05,
+      "loss": 1.03,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8253280365956422,
+      "grad_norm": 0.486418217420578,
+      "learning_rate": 1.569113429233252e-05,
+      "loss": 0.2692,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8258095582039244,
+      "grad_norm": 0.7394134998321533,
+      "learning_rate": 1.5607350664799157e-05,
+      "loss": 0.6892,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8262910798122066,
+      "grad_norm": 0.7706437110900879,
+      "learning_rate": 1.552377238283943e-05,
+      "loss": 0.2832,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8267726014204887,
+      "grad_norm": 0.9024636745452881,
+      "learning_rate": 1.5440399649817385e-05,
+      "loss": 0.3443,
+      "step": 1717
+    },
+    {
+      "epoch": 0.8272541230287709,
+      "grad_norm": 0.6336109042167664,
+      "learning_rate": 1.5357232668596933e-05,
+      "loss": 0.6889,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8277356446370531,
+      "grad_norm": 0.3393593728542328,
+      "learning_rate": 1.5274271641541295e-05,
+      "loss": 0.1545,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8282171662453353,
+      "grad_norm": 0.6772252321243286,
+      "learning_rate": 1.5191516770512649e-05,
+      "loss": 0.7456,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8286986878536174,
+      "grad_norm": 0.7937234044075012,
+      "learning_rate": 1.5108968256871437e-05,
+      "loss": 0.349,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8291802094618996,
+      "grad_norm": 1.5913159847259521,
+      "learning_rate": 1.5026626301476087e-05,
+      "loss": 0.9463,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8296617310701818,
+      "grad_norm": 0.6769959330558777,
+      "learning_rate": 1.4944491104682379e-05,
+      "loss": 0.5529,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8301432526784639,
+      "grad_norm": 1.4310240745544434,
+      "learning_rate": 1.4862562866343034e-05,
+      "loss": 0.5971,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8306247742867461,
+      "grad_norm": 0.9801859259605408,
+      "learning_rate": 1.4780841785807164e-05,
+      "loss": 0.7973,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8311062958950283,
+      "grad_norm": 0.6424405574798584,
+      "learning_rate": 1.4699328061919848e-05,
+      "loss": 0.2604,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8315878175033105,
+      "grad_norm": 1.0231726169586182,
+      "learning_rate": 1.4618021893021605e-05,
+      "loss": 0.3021,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8320693391115926,
+      "grad_norm": 0.6480558514595032,
+      "learning_rate": 1.453692347694794e-05,
+      "loss": 0.4345,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8325508607198748,
+      "grad_norm": 0.8369768857955933,
+      "learning_rate": 1.4456033011028835e-05,
+      "loss": 0.3847,
+      "step": 1729
+    },
+    {
+      "epoch": 0.833032382328157,
+      "grad_norm": 0.6749899983406067,
+      "learning_rate": 1.437535069208833e-05,
+      "loss": 0.2968,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8335139039364391,
+      "grad_norm": 0.6067473888397217,
+      "learning_rate": 1.4294876716443906e-05,
+      "loss": 0.3409,
+      "step": 1731
+    },
+    {
+      "epoch": 0.8339954255447213,
+      "grad_norm": 0.641484797000885,
+      "learning_rate": 1.4214611279906187e-05,
+      "loss": 0.2532,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8344769471530035,
+      "grad_norm": 0.698528528213501,
+      "learning_rate": 1.4134554577778337e-05,
+      "loss": 0.3922,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8349584687612857,
+      "grad_norm": 0.823616087436676,
+      "learning_rate": 1.4054706804855634e-05,
+      "loss": 0.8489,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8354399903695678,
+      "grad_norm": 0.7121361494064331,
+      "learning_rate": 1.3975068155424976e-05,
+      "loss": 0.6374,
+      "step": 1735
+    },
+    {
+      "epoch": 0.83592151197785,
+      "grad_norm": 1.6407108306884766,
+      "learning_rate": 1.3895638823264446e-05,
+      "loss": 0.5144,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8364030335861322,
+      "grad_norm": 1.073464035987854,
+      "learning_rate": 1.3816419001642777e-05,
+      "loss": 0.6168,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8368845551944144,
+      "grad_norm": 1.1071691513061523,
+      "learning_rate": 1.3737408883318948e-05,
+      "loss": 0.5041,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8373660768026965,
+      "grad_norm": 0.9471048712730408,
+      "learning_rate": 1.365860866054165e-05,
+      "loss": 0.3752,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8378475984109787,
+      "grad_norm": 0.42994487285614014,
+      "learning_rate": 1.358001852504891e-05,
+      "loss": 0.2467,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8383291200192609,
+      "grad_norm": 1.1875547170639038,
+      "learning_rate": 1.3501638668067485e-05,
+      "loss": 0.8259,
+      "step": 1741
+    },
+    {
+      "epoch": 0.838810641627543,
+      "grad_norm": 1.273987889289856,
+      "learning_rate": 1.3423469280312562e-05,
+      "loss": 0.478,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8392921632358252,
+      "grad_norm": 0.571315586566925,
+      "learning_rate": 1.3345510551987128e-05,
+      "loss": 0.2148,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8397736848441074,
+      "grad_norm": 0.632285475730896,
+      "learning_rate": 1.326776267278167e-05,
+      "loss": 0.3969,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8402552064523896,
+      "grad_norm": 1.3262178897857666,
+      "learning_rate": 1.3190225831873581e-05,
+      "loss": 0.4584,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8407367280606717,
+      "grad_norm": 0.6027417778968811,
+      "learning_rate": 1.3112900217926782e-05,
+      "loss": 0.6441,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8412182496689539,
+      "grad_norm": 0.8074085712432861,
+      "learning_rate": 1.3035786019091223e-05,
+      "loss": 0.5024,
+      "step": 1747
+    },
+    {
+      "epoch": 0.841699771277236,
+      "grad_norm": 0.7299503684043884,
+      "learning_rate": 1.2958883423002422e-05,
+      "loss": 1.0229,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8421812928855182,
+      "grad_norm": 0.6064135432243347,
+      "learning_rate": 1.288219261678103e-05,
+      "loss": 0.3824,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8426628144938004,
+      "grad_norm": 0.9795793294906616,
+      "learning_rate": 1.2805713787032381e-05,
+      "loss": 0.3306,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8431443361020826,
+      "grad_norm": 0.3210099935531616,
+      "learning_rate": 1.2729447119846016e-05,
+      "loss": 0.2856,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8436258577103648,
+      "grad_norm": 1.0421223640441895,
+      "learning_rate": 1.265339280079525e-05,
+      "loss": 0.7998,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8441073793186469,
+      "grad_norm": 0.8377938270568848,
+      "learning_rate": 1.257755101493665e-05,
+      "loss": 0.6302,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8445889009269291,
+      "grad_norm": 0.3239783048629761,
+      "learning_rate": 1.2501921946809714e-05,
+      "loss": 0.1868,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8450704225352113,
+      "grad_norm": 0.5518014430999756,
+      "learning_rate": 1.2426505780436326e-05,
+      "loss": 0.3083,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8455519441434934,
+      "grad_norm": 0.7590487003326416,
+      "learning_rate": 1.2351302699320332e-05,
+      "loss": 0.7756,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8460334657517756,
+      "grad_norm": 0.8196136951446533,
+      "learning_rate": 1.2276312886447106e-05,
+      "loss": 0.734,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8465149873600578,
+      "grad_norm": 0.9058529138565063,
+      "learning_rate": 1.2201536524283074e-05,
+      "loss": 0.5072,
+      "step": 1758
+    },
+    {
+      "epoch": 0.84699650896834,
+      "grad_norm": 0.389612078666687,
+      "learning_rate": 1.2126973794775343e-05,
+      "loss": 0.308,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8474780305766221,
+      "grad_norm": 0.9301180243492126,
+      "learning_rate": 1.2052624879351104e-05,
+      "loss": 0.3945,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8479595521849043,
+      "grad_norm": 0.7932952046394348,
+      "learning_rate": 1.1978489958917382e-05,
+      "loss": 0.651,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8484410737931865,
+      "grad_norm": 0.8034958839416504,
+      "learning_rate": 1.1904569213860472e-05,
+      "loss": 0.7516,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8489225954014686,
+      "grad_norm": 0.584695041179657,
+      "learning_rate": 1.1830862824045552e-05,
+      "loss": 0.8093,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8494041170097508,
+      "grad_norm": 1.2081027030944824,
+      "learning_rate": 1.1757370968816217e-05,
+      "loss": 0.6173,
+      "step": 1764
+    },
+    {
+      "epoch": 0.849885638618033,
+      "grad_norm": 0.9123238921165466,
+      "learning_rate": 1.1684093826994024e-05,
+      "loss": 0.2586,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8503671602263152,
+      "grad_norm": 0.805109977722168,
+      "learning_rate": 1.1611031576878117e-05,
+      "loss": 0.519,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8508486818345973,
+      "grad_norm": 0.49961966276168823,
+      "learning_rate": 1.1538184396244778e-05,
+      "loss": 0.2066,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8513302034428795,
+      "grad_norm": 0.7591211199760437,
+      "learning_rate": 1.146555246234694e-05,
+      "loss": 0.3607,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8518117250511617,
+      "grad_norm": 1.0066025257110596,
+      "learning_rate": 1.1393135951913824e-05,
+      "loss": 0.6076,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8522932466594438,
+      "grad_norm": 0.8413494229316711,
+      "learning_rate": 1.132093504115046e-05,
+      "loss": 0.5581,
+      "step": 1770
+    },
+    {
+      "epoch": 0.852774768267726,
+      "grad_norm": 0.9132838249206543,
+      "learning_rate": 1.1248949905737283e-05,
+      "loss": 0.3262,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8532562898760082,
+      "grad_norm": 0.4182647466659546,
+      "learning_rate": 1.1177180720829694e-05,
+      "loss": 0.1514,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8537378114842904,
+      "grad_norm": 0.4618658125400543,
+      "learning_rate": 1.1105627661057671e-05,
+      "loss": 0.4407,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8542193330925725,
+      "grad_norm": 0.7735852003097534,
+      "learning_rate": 1.103429090052528e-05,
+      "loss": 0.7609,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8547008547008547,
+      "grad_norm": 0.5562105178833008,
+      "learning_rate": 1.096317061281027e-05,
+      "loss": 0.5554,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8551823763091368,
+      "grad_norm": 1.005146861076355,
+      "learning_rate": 1.0892266970963704e-05,
+      "loss": 0.8242,
+      "step": 1776
+    },
+    {
+      "epoch": 0.855663897917419,
+      "grad_norm": 0.9059540033340454,
+      "learning_rate": 1.082158014750948e-05,
+      "loss": 0.8464,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8561454195257012,
+      "grad_norm": 0.5801354646682739,
+      "learning_rate": 1.0751110314443958e-05,
+      "loss": 0.4619,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8566269411339834,
+      "grad_norm": 0.5400471091270447,
+      "learning_rate": 1.0680857643235431e-05,
+      "loss": 0.2479,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8571084627422656,
+      "grad_norm": 0.6659761071205139,
+      "learning_rate": 1.0610822304823887e-05,
+      "loss": 0.5576,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8575899843505478,
+      "grad_norm": 0.9168156981468201,
+      "learning_rate": 1.0541004469620452e-05,
+      "loss": 0.6976,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8580715059588299,
+      "grad_norm": 0.9955235719680786,
+      "learning_rate": 1.0471404307507016e-05,
+      "loss": 1.502,
+      "step": 1782
+    },
+    {
+      "epoch": 0.858553027567112,
+      "grad_norm": 0.669767439365387,
+      "learning_rate": 1.0402021987835831e-05,
+      "loss": 0.2673,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8590345491753942,
+      "grad_norm": 0.3703511953353882,
+      "learning_rate": 1.0332857679429098e-05,
+      "loss": 0.212,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8595160707836764,
+      "grad_norm": 1.1823803186416626,
+      "learning_rate": 1.0263911550578531e-05,
+      "loss": 0.6145,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8599975923919586,
+      "grad_norm": 0.8142738938331604,
+      "learning_rate": 1.0195183769045013e-05,
+      "loss": 0.2403,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8604791140002408,
+      "grad_norm": 0.8147161602973938,
+      "learning_rate": 1.0126674502058054e-05,
+      "loss": 0.8877,
+      "step": 1787
+    },
+    {
+      "epoch": 0.860960635608523,
+      "grad_norm": 0.6185226440429688,
+      "learning_rate": 1.005838391631555e-05,
+      "loss": 0.3974,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8614421572168051,
+      "grad_norm": 0.7300569415092468,
+      "learning_rate": 9.990312177983263e-06,
+      "loss": 0.4777,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8619236788250872,
+      "grad_norm": 0.8982576131820679,
+      "learning_rate": 9.922459452694466e-06,
+      "loss": 0.6434,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8624052004333694,
+      "grad_norm": 0.5035396218299866,
+      "learning_rate": 9.854825905549503e-06,
+      "loss": 0.5651,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8628867220416516,
+      "grad_norm": 0.6236551403999329,
+      "learning_rate": 9.787411701115456e-06,
+      "loss": 0.2783,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8633682436499338,
+      "grad_norm": 1.1999423503875732,
+      "learning_rate": 9.720217003425647e-06,
+      "loss": 1.0484,
+      "step": 1793
+    },
+    {
+      "epoch": 0.863849765258216,
+      "grad_norm": 0.6424204111099243,
+      "learning_rate": 9.65324197597931e-06,
+      "loss": 0.294,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8643312868664982,
+      "grad_norm": 0.724156379699707,
+      "learning_rate": 9.58648678174121e-06,
+      "loss": 0.3884,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8648128084747803,
+      "grad_norm": 0.47926265001296997,
+      "learning_rate": 9.51995158314113e-06,
+      "loss": 0.5631,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8652943300830624,
+      "grad_norm": 1.520438551902771,
+      "learning_rate": 9.45363654207363e-06,
+      "loss": 0.8433,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8657758516913446,
+      "grad_norm": 0.3574250340461731,
+      "learning_rate": 9.387541819897549e-06,
+      "loss": 0.4286,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8662573732996268,
+      "grad_norm": 1.6910595893859863,
+      "learning_rate": 9.321667577435634e-06,
+      "loss": 0.691,
+      "step": 1799
+    },
+    {
+      "epoch": 0.866738894907909,
+      "grad_norm": 0.5806576013565063,
+      "learning_rate": 9.256013974974175e-06,
+      "loss": 0.3654,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8672204165161912,
+      "grad_norm": 0.661484956741333,
+      "learning_rate": 9.19058117226258e-06,
+      "loss": 0.6508,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8677019381244734,
+      "grad_norm": 0.8972189426422119,
+      "learning_rate": 9.125369328513034e-06,
+      "loss": 0.9418,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8681834597327555,
+      "grad_norm": 1.4898693561553955,
+      "learning_rate": 9.060378602400054e-06,
+      "loss": 0.374,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8686649813410376,
+      "grad_norm": 1.0133569240570068,
+      "learning_rate": 8.995609152060136e-06,
+      "loss": 0.8903,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8691465029493198,
+      "grad_norm": 0.5878236889839172,
+      "learning_rate": 8.931061135091357e-06,
+      "loss": 0.7779,
+      "step": 1805
+    },
+    {
+      "epoch": 0.869628024557602,
+      "grad_norm": 0.7630800008773804,
+      "learning_rate": 8.866734708553015e-06,
+      "loss": 0.8601,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8701095461658842,
+      "grad_norm": 0.45022618770599365,
+      "learning_rate": 8.802630028965242e-06,
+      "loss": 0.2447,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8705910677741664,
+      "grad_norm": 0.34775686264038086,
+      "learning_rate": 8.738747252308555e-06,
+      "loss": 0.1954,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8710725893824486,
+      "grad_norm": 0.6343087553977966,
+      "learning_rate": 8.675086534023591e-06,
+      "loss": 0.3324,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8715541109907307,
+      "grad_norm": 1.0207873582839966,
+      "learning_rate": 8.611648029010643e-06,
+      "loss": 0.3045,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8720356325990128,
+      "grad_norm": 0.7967151403427124,
+      "learning_rate": 8.548431891629316e-06,
+      "loss": 0.2367,
+      "step": 1811
+    },
+    {
+      "epoch": 0.872517154207295,
+      "grad_norm": 0.6096175909042358,
+      "learning_rate": 8.485438275698154e-06,
+      "loss": 0.1747,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8729986758155772,
+      "grad_norm": 1.0279371738433838,
+      "learning_rate": 8.422667334494249e-06,
+      "loss": 0.6618,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8734801974238594,
+      "grad_norm": 0.5840936899185181,
+      "learning_rate": 8.360119220752893e-06,
+      "loss": 0.1349,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8739617190321416,
+      "grad_norm": 0.6559387445449829,
+      "learning_rate": 8.297794086667165e-06,
+      "loss": 0.6161,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8744432406404238,
+      "grad_norm": 0.7474731802940369,
+      "learning_rate": 8.235692083887613e-06,
+      "loss": 0.4046,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8749247622487059,
+      "grad_norm": 1.1205002069473267,
+      "learning_rate": 8.173813363521843e-06,
+      "loss": 0.8628,
+      "step": 1817
+    },
+    {
+      "epoch": 0.875406283856988,
+      "grad_norm": 0.8745394349098206,
+      "learning_rate": 8.112158076134157e-06,
+      "loss": 0.6641,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8758878054652702,
+      "grad_norm": 0.773175060749054,
+      "learning_rate": 8.05072637174522e-06,
+      "loss": 0.5763,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8763693270735524,
+      "grad_norm": 0.4917106032371521,
+      "learning_rate": 7.989518399831641e-06,
+      "loss": 0.2699,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8768508486818346,
+      "grad_norm": 0.6174960732460022,
+      "learning_rate": 7.928534309325675e-06,
+      "loss": 0.6247,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8773323702901168,
+      "grad_norm": 1.1026489734649658,
+      "learning_rate": 7.8677742486148e-06,
+      "loss": 0.7072,
+      "step": 1822
+    },
+    {
+      "epoch": 0.877813891898399,
+      "grad_norm": 0.716529130935669,
+      "learning_rate": 7.807238365541391e-06,
+      "loss": 0.2818,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8782954135066812,
+      "grad_norm": 0.7560042142868042,
+      "learning_rate": 7.746926807402344e-06,
+      "loss": 0.5247,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8787769351149632,
+      "grad_norm": 0.9436668753623962,
+      "learning_rate": 7.686839720948736e-06,
+      "loss": 1.0009,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8792584567232454,
+      "grad_norm": 1.0390268564224243,
+      "learning_rate": 7.6269772523854365e-06,
+      "loss": 0.2738,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8797399783315276,
+      "grad_norm": 0.6893677115440369,
+      "learning_rate": 7.567339547370789e-06,
+      "loss": 0.3845,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8802214999398098,
+      "grad_norm": 1.3264459371566772,
+      "learning_rate": 7.507926751016248e-06,
+      "loss": 0.4929,
+      "step": 1828
+    },
+    {
+      "epoch": 0.880703021548092,
+      "grad_norm": 0.6041299700737,
+      "learning_rate": 7.4487390078859855e-06,
+      "loss": 0.3404,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8811845431563742,
+      "grad_norm": 0.7319620251655579,
+      "learning_rate": 7.389776461996578e-06,
+      "loss": 0.6722,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8816660647646564,
+      "grad_norm": 0.9752634167671204,
+      "learning_rate": 7.331039256816663e-06,
+      "loss": 0.9134,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8821475863729384,
+      "grad_norm": 1.0027194023132324,
+      "learning_rate": 7.27252753526656e-06,
+      "loss": 0.8346,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8826291079812206,
+      "grad_norm": 0.8595365285873413,
+      "learning_rate": 7.214241439717962e-06,
+      "loss": 0.4265,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8831106295895028,
+      "grad_norm": 0.7600406408309937,
+      "learning_rate": 7.1561811119935425e-06,
+      "loss": 0.3207,
+      "step": 1834
+    },
+    {
+      "epoch": 0.883592151197785,
+      "grad_norm": 0.5692286491394043,
+      "learning_rate": 7.098346693366642e-06,
+      "loss": 0.3267,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8840736728060672,
+      "grad_norm": 0.8591275215148926,
+      "learning_rate": 7.0407383245609136e-06,
+      "loss": 0.3607,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8845551944143494,
+      "grad_norm": 0.6609878540039062,
+      "learning_rate": 6.983356145749975e-06,
+      "loss": 0.2951,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8850367160226316,
+      "grad_norm": 0.565071702003479,
+      "learning_rate": 6.9262002965570835e-06,
+      "loss": 0.1949,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8855182376309136,
+      "grad_norm": 0.515185534954071,
+      "learning_rate": 6.869270916054782e-06,
+      "loss": 0.4361,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8859997592391958,
+      "grad_norm": 0.5699068903923035,
+      "learning_rate": 6.812568142764575e-06,
+      "loss": 0.3235,
+      "step": 1840
+    },
+    {
+      "epoch": 0.886481280847478,
+      "grad_norm": 0.7445323467254639,
+      "learning_rate": 6.756092114656587e-06,
+      "loss": 0.622,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8869628024557602,
+      "grad_norm": 0.9684603214263916,
+      "learning_rate": 6.699842969149195e-06,
+      "loss": 0.3984,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8874443240640424,
+      "grad_norm": 0.6813137531280518,
+      "learning_rate": 6.64382084310875e-06,
+      "loss": 0.3141,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8879258456723246,
+      "grad_norm": 1.0037721395492554,
+      "learning_rate": 6.5880258728491905e-06,
+      "loss": 0.4377,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8884073672806068,
+      "grad_norm": 1.097926139831543,
+      "learning_rate": 6.532458194131763e-06,
+      "loss": 0.8514,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.8100180625915527,
+      "learning_rate": 6.477117942164657e-06,
+      "loss": 0.3044,
+      "step": 1846
+    },
+    {
+      "epoch": 0.889370410497171,
+      "grad_norm": 0.5506505370140076,
+      "learning_rate": 6.422005251602658e-06,
+      "loss": 0.2629,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8898519321054532,
+      "grad_norm": 0.5061823725700378,
+      "learning_rate": 6.367120256546888e-06,
+      "loss": 0.2579,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8903334537137354,
+      "grad_norm": 0.8128954768180847,
+      "learning_rate": 6.312463090544396e-06,
+      "loss": 0.4331,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8908149753220176,
+      "grad_norm": 0.6516621708869934,
+      "learning_rate": 6.258033886587911e-06,
+      "loss": 0.8575,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8912964969302998,
+      "grad_norm": 0.6090648174285889,
+      "learning_rate": 6.2038327771154485e-06,
+      "loss": 0.3023,
+      "step": 1851
+    },
+    {
+      "epoch": 0.891778018538582,
+      "grad_norm": 0.5383626818656921,
+      "learning_rate": 6.1498598940100346e-06,
+      "loss": 0.4444,
+      "step": 1852
+    },
+    {
+      "epoch": 0.892259540146864,
+      "grad_norm": 0.432041198015213,
+      "learning_rate": 6.0961153685993646e-06,
+      "loss": 0.2102,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8927410617551462,
+      "grad_norm": 0.5301342010498047,
+      "learning_rate": 6.0425993316554965e-06,
+      "loss": 0.2474,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8932225833634284,
+      "grad_norm": 0.5947489738464355,
+      "learning_rate": 5.989311913394546e-06,
+      "loss": 0.294,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8937041049717106,
+      "grad_norm": 0.46525564789772034,
+      "learning_rate": 5.93625324347632e-06,
+      "loss": 0.1681,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8941856265799928,
+      "grad_norm": 1.2389193773269653,
+      "learning_rate": 5.8834234510040335e-06,
+      "loss": 0.7665,
+      "step": 1857
+    },
+    {
+      "epoch": 0.894667148188275,
+      "grad_norm": 0.48213279247283936,
+      "learning_rate": 5.830822664523994e-06,
+      "loss": 0.0887,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8951486697965572,
+      "grad_norm": 0.721508800983429,
+      "learning_rate": 5.77845101202531e-06,
+      "loss": 0.3518,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8956301914048392,
+      "grad_norm": 0.5365184545516968,
+      "learning_rate": 5.726308620939536e-06,
+      "loss": 0.1631,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8961117130131214,
+      "grad_norm": 0.500117838382721,
+      "learning_rate": 5.674395618140393e-06,
+      "loss": 0.2508,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8965932346214036,
+      "grad_norm": 0.418396919965744,
+      "learning_rate": 5.622712129943453e-06,
+      "loss": 0.1585,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8970747562296858,
+      "grad_norm": 0.5959372520446777,
+      "learning_rate": 5.571258282105829e-06,
+      "loss": 0.7838,
+      "step": 1863
+    },
+    {
+      "epoch": 0.897556277837968,
+      "grad_norm": 0.8515798449516296,
+      "learning_rate": 5.520034199825841e-06,
+      "loss": 1.1867,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8980377994462502,
+      "grad_norm": 0.6227186918258667,
+      "learning_rate": 5.469040007742776e-06,
+      "loss": 0.6046,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8985193210545324,
+      "grad_norm": 1.0233310461044312,
+      "learning_rate": 5.418275829936537e-06,
+      "loss": 1.1337,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8990008426628145,
+      "grad_norm": 1.1459534168243408,
+      "learning_rate": 5.36774178992735e-06,
+      "loss": 0.5849,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8994823642710966,
+      "grad_norm": 0.5045539736747742,
+      "learning_rate": 5.317438010675469e-06,
+      "loss": 0.3194,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8999638858793788,
+      "grad_norm": 0.5089043974876404,
+      "learning_rate": 5.267364614580861e-06,
+      "loss": 0.1604,
+      "step": 1869
+    },
+    {
+      "epoch": 0.900445407487661,
+      "grad_norm": 0.706655740737915,
+      "learning_rate": 5.217521723482943e-06,
+      "loss": 0.3677,
+      "step": 1870
+    },
+    {
+      "epoch": 0.9009269290959432,
+      "grad_norm": 0.5909689664840698,
+      "learning_rate": 5.167909458660258e-06,
+      "loss": 0.5146,
+      "step": 1871
+    },
+    {
+      "epoch": 0.9014084507042254,
+      "grad_norm": 0.9625548720359802,
+      "learning_rate": 5.118527940830165e-06,
+      "loss": 0.3767,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9018899723125076,
+      "grad_norm": 1.1841224431991577,
+      "learning_rate": 5.069377290148602e-06,
+      "loss": 0.4,
+      "step": 1873
+    },
+    {
+      "epoch": 0.9023714939207897,
+      "grad_norm": 0.5126852989196777,
+      "learning_rate": 5.020457626209707e-06,
+      "loss": 0.2379,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9028530155290718,
+      "grad_norm": 0.9060919284820557,
+      "learning_rate": 4.971769068045628e-06,
+      "loss": 1.1507,
+      "step": 1875
+    },
+    {
+      "epoch": 0.903334537137354,
+      "grad_norm": 0.5319393873214722,
+      "learning_rate": 4.923311734126135e-06,
+      "loss": 0.4051,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9038160587456362,
+      "grad_norm": 0.4319634735584259,
+      "learning_rate": 4.875085742358432e-06,
+      "loss": 0.5979,
+      "step": 1877
+    },
+    {
+      "epoch": 0.9042975803539184,
+      "grad_norm": 0.5022990107536316,
+      "learning_rate": 4.827091210086776e-06,
+      "loss": 0.2293,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9047791019622006,
+      "grad_norm": 0.6433436870574951,
+      "learning_rate": 4.779328254092252e-06,
+      "loss": 0.7774,
+      "step": 1879
+    },
+    {
+      "epoch": 0.9052606235704828,
+      "grad_norm": 0.859475314617157,
+      "learning_rate": 4.731796990592452e-06,
+      "loss": 0.4518,
+      "step": 1880
+    },
+    {
+      "epoch": 0.905742145178765,
+      "grad_norm": 0.5635149478912354,
+      "learning_rate": 4.68449753524125e-06,
+      "loss": 0.6374,
+      "step": 1881
+    },
+    {
+      "epoch": 0.906223666787047,
+      "grad_norm": 0.5391207933425903,
+      "learning_rate": 4.637430003128429e-06,
+      "loss": 0.5627,
+      "step": 1882
+    },
+    {
+      "epoch": 0.9067051883953292,
+      "grad_norm": 1.4014676809310913,
+      "learning_rate": 4.5905945087794996e-06,
+      "loss": 0.6463,
+      "step": 1883
+    },
+    {
+      "epoch": 0.9071867100036114,
+      "grad_norm": 0.8849884271621704,
+      "learning_rate": 4.543991166155337e-06,
+      "loss": 0.8515,
+      "step": 1884
+    },
+    {
+      "epoch": 0.9076682316118936,
+      "grad_norm": 0.7712661623954773,
+      "learning_rate": 4.497620088651966e-06,
+      "loss": 0.3442,
+      "step": 1885
+    },
+    {
+      "epoch": 0.9081497532201758,
+      "grad_norm": 0.4782671332359314,
+      "learning_rate": 4.451481389100232e-06,
+      "loss": 0.5196,
+      "step": 1886
+    },
+    {
+      "epoch": 0.908631274828458,
+      "grad_norm": 0.9418804049491882,
+      "learning_rate": 4.405575179765586e-06,
+      "loss": 0.4203,
+      "step": 1887
+    },
+    {
+      "epoch": 0.9091127964367401,
+      "grad_norm": 0.6803520321846008,
+      "learning_rate": 4.359901572347758e-06,
+      "loss": 0.9799,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9095943180450222,
+      "grad_norm": 0.8574094772338867,
+      "learning_rate": 4.314460677980537e-06,
+      "loss": 0.4231,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9100758396533044,
+      "grad_norm": 0.4506451487541199,
+      "learning_rate": 4.269252607231422e-06,
+      "loss": 0.269,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9105573612615866,
+      "grad_norm": 0.6377048492431641,
+      "learning_rate": 4.224277470101445e-06,
+      "loss": 0.447,
+      "step": 1891
+    },
+    {
+      "epoch": 0.9110388828698688,
+      "grad_norm": 0.6049963235855103,
+      "learning_rate": 4.179535376024857e-06,
+      "loss": 0.4994,
+      "step": 1892
+    },
+    {
+      "epoch": 0.911520404478151,
+      "grad_norm": 0.8465968370437622,
+      "learning_rate": 4.135026433868827e-06,
+      "loss": 0.6467,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9120019260864332,
+      "grad_norm": 0.8001445531845093,
+      "learning_rate": 4.090750751933248e-06,
+      "loss": 0.9881,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9124834476947153,
+      "grad_norm": 0.8922566175460815,
+      "learning_rate": 4.046708437950464e-06,
+      "loss": 0.5415,
+      "step": 1895
+    },
+    {
+      "epoch": 0.9129649693029974,
+      "grad_norm": 1.0354160070419312,
+      "learning_rate": 4.0028995990849084e-06,
+      "loss": 0.4927,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9134464909112796,
+      "grad_norm": 0.4661247432231903,
+      "learning_rate": 3.95932434193299e-06,
+      "loss": 0.1731,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9139280125195618,
+      "grad_norm": 0.596748948097229,
+      "learning_rate": 3.915982772522719e-06,
+      "loss": 0.4255,
+      "step": 1898
+    },
+    {
+      "epoch": 0.914409534127844,
+      "grad_norm": 0.9924038648605347,
+      "learning_rate": 3.872874996313513e-06,
+      "loss": 1.0724,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9148910557361262,
+      "grad_norm": 0.7841641306877136,
+      "learning_rate": 3.830001118195936e-06,
+      "loss": 0.3032,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9153725773444084,
+      "grad_norm": 0.9350888729095459,
+      "learning_rate": 3.787361242491394e-06,
+      "loss": 0.4085,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9158540989526905,
+      "grad_norm": 0.3341303765773773,
+      "learning_rate": 3.744955472951928e-06,
+      "loss": 0.1407,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9163356205609726,
+      "grad_norm": 1.7002257108688354,
+      "learning_rate": 3.702783912759955e-06,
+      "loss": 0.5921,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9168171421692548,
+      "grad_norm": 0.5047416687011719,
+      "learning_rate": 3.660846664528006e-06,
+      "loss": 0.3457,
+      "step": 1904
+    },
+    {
+      "epoch": 0.917298663777537,
+      "grad_norm": 2.0711960792541504,
+      "learning_rate": 3.6191438302984772e-06,
+      "loss": 0.947,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9177801853858192,
+      "grad_norm": 0.5945473909378052,
+      "learning_rate": 3.577675511543388e-06,
+      "loss": 0.4444,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9182617069941014,
+      "grad_norm": 0.8059927225112915,
+      "learning_rate": 3.5364418091641373e-06,
+      "loss": 1.0539,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9187432286023836,
+      "grad_norm": 0.6776530146598816,
+      "learning_rate": 3.495442823491224e-06,
+      "loss": 0.5971,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9192247502106657,
+      "grad_norm": 1.1483100652694702,
+      "learning_rate": 3.4546786542840605e-06,
+      "loss": 0.6773,
+      "step": 1909
+    },
+    {
+      "epoch": 0.9197062718189479,
+      "grad_norm": 0.8443659543991089,
+      "learning_rate": 3.4141494007306816e-06,
+      "loss": 0.3389,
+      "step": 1910
+    },
+    {
+      "epoch": 0.92018779342723,
+      "grad_norm": 0.8820102214813232,
+      "learning_rate": 3.373855161447548e-06,
+      "loss": 1.1183,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9206693150355122,
+      "grad_norm": 0.750338613986969,
+      "learning_rate": 3.333796034479242e-06,
+      "loss": 0.3541,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9211508366437944,
+      "grad_norm": 0.4099019169807434,
+      "learning_rate": 3.293972117298294e-06,
+      "loss": 0.1908,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9216323582520766,
+      "grad_norm": 0.8014377951622009,
+      "learning_rate": 3.2543835068049255e-06,
+      "loss": 0.3878,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9221138798603588,
+      "grad_norm": 0.7586486339569092,
+      "learning_rate": 3.21503029932676e-06,
+      "loss": 0.4422,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9225954014686409,
+      "grad_norm": 0.32165953516960144,
+      "learning_rate": 3.1759125906186793e-06,
+      "loss": 0.1944,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.6342649459838867,
+      "learning_rate": 3.137030475862535e-06,
+      "loss": 0.7593,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9235584446852052,
+      "grad_norm": 0.6746546030044556,
+      "learning_rate": 3.098384049666925e-06,
+      "loss": 0.2116,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9240399662934874,
+      "grad_norm": 0.5435047149658203,
+      "learning_rate": 3.059973406066963e-06,
+      "loss": 0.3394,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9245214879017696,
+      "grad_norm": 0.5165433883666992,
+      "learning_rate": 3.0217986385240537e-06,
+      "loss": 0.2825,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9250030095100518,
+      "grad_norm": 0.8822990655899048,
+      "learning_rate": 2.983859839925662e-06,
+      "loss": 0.6913,
+      "step": 1921
+    },
+    {
+      "epoch": 0.925484531118334,
+      "grad_norm": 1.065269112586975,
+      "learning_rate": 2.94615710258509e-06,
+      "loss": 0.9278,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9259660527266161,
+      "grad_norm": 1.2459629774093628,
+      "learning_rate": 2.908690518241275e-06,
+      "loss": 0.6308,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9264475743348983,
+      "grad_norm": 0.7520962953567505,
+      "learning_rate": 2.8714601780584937e-06,
+      "loss": 0.3369,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9269290959431804,
+      "grad_norm": 0.5629241466522217,
+      "learning_rate": 2.834466172626238e-06,
+      "loss": 0.2532,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9274106175514626,
+      "grad_norm": 0.9873923659324646,
+      "learning_rate": 2.7977085919589254e-06,
+      "loss": 0.4126,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9278921391597448,
+      "grad_norm": 0.7369255423545837,
+      "learning_rate": 2.76118752549569e-06,
+      "loss": 0.169,
+      "step": 1927
+    },
+    {
+      "epoch": 0.928373660768027,
+      "grad_norm": 0.8374475240707397,
+      "learning_rate": 2.7249030621001924e-06,
+      "loss": 0.7864,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9288551823763092,
+      "grad_norm": 0.8158120512962341,
+      "learning_rate": 2.688855290060399e-06,
+      "loss": 0.6308,
+      "step": 1929
+    },
+    {
+      "epoch": 0.9293367039845913,
+      "grad_norm": 0.8472880721092224,
+      "learning_rate": 2.653044297088314e-06,
+      "loss": 0.9682,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9298182255928735,
+      "grad_norm": 0.49577999114990234,
+      "learning_rate": 2.6174701703198468e-06,
+      "loss": 0.3384,
+      "step": 1931
+    },
+    {
+      "epoch": 0.9302997472011556,
+      "grad_norm": 0.7291809916496277,
+      "learning_rate": 2.5821329963145347e-06,
+      "loss": 0.6925,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9307812688094378,
+      "grad_norm": 0.7352073788642883,
+      "learning_rate": 2.547032861055376e-06,
+      "loss": 0.3493,
+      "step": 1933
+    },
+    {
+      "epoch": 0.93126279041772,
+      "grad_norm": 0.6518670320510864,
+      "learning_rate": 2.5121698499485757e-06,
+      "loss": 0.7209,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9317443120260022,
+      "grad_norm": 0.9821415543556213,
+      "learning_rate": 2.4775440478233993e-06,
+      "loss": 0.8259,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9322258336342844,
+      "grad_norm": 0.6522502899169922,
+      "learning_rate": 2.4431555389319074e-06,
+      "loss": 0.4079,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9327073552425665,
+      "grad_norm": 0.7061333060264587,
+      "learning_rate": 2.4090044069487784e-06,
+      "loss": 0.7641,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9331888768508487,
+      "grad_norm": 0.49332860112190247,
+      "learning_rate": 2.3750907349711084e-06,
+      "loss": 0.2268,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9336703984591308,
+      "grad_norm": 1.0159000158309937,
+      "learning_rate": 2.3414146055182106e-06,
+      "loss": 0.785,
+      "step": 1939
+    },
+    {
+      "epoch": 0.934151920067413,
+      "grad_norm": 0.6537801027297974,
+      "learning_rate": 2.307976100531384e-06,
+      "loss": 0.2548,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9346334416756952,
+      "grad_norm": 0.5711776614189148,
+      "learning_rate": 2.274775301373744e-06,
+      "loss": 0.4644,
+      "step": 1941
+    },
+    {
+      "epoch": 0.9351149632839774,
+      "grad_norm": 0.6775410175323486,
+      "learning_rate": 2.241812288830003e-06,
+      "loss": 0.8405,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9355964848922596,
+      "grad_norm": 0.4718037247657776,
+      "learning_rate": 2.2090871431063253e-06,
+      "loss": 0.2021,
+      "step": 1943
+    },
+    {
+      "epoch": 0.9360780065005417,
+      "grad_norm": 0.9879113435745239,
+      "learning_rate": 2.176599943830071e-06,
+      "loss": 0.5937,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9365595281088239,
+      "grad_norm": 0.8017491698265076,
+      "learning_rate": 2.144350770049597e-06,
+      "loss": 0.3747,
+      "step": 1945
+    },
+    {
+      "epoch": 0.937041049717106,
+      "grad_norm": 0.752362847328186,
+      "learning_rate": 2.112339700234156e-06,
+      "loss": 0.7747,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9375225713253882,
+      "grad_norm": 0.8939694166183472,
+      "learning_rate": 2.0805668122735767e-06,
+      "loss": 0.3778,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9380040929336704,
+      "grad_norm": 0.6203170418739319,
+      "learning_rate": 2.0490321834781833e-06,
+      "loss": 0.3018,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9384856145419526,
+      "grad_norm": 0.8364834785461426,
+      "learning_rate": 2.0177358905785537e-06,
+      "loss": 0.5277,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9389671361502347,
+      "grad_norm": 0.6893821358680725,
+      "learning_rate": 1.986678009725329e-06,
+      "loss": 0.5612,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9394486577585169,
+      "grad_norm": 0.5284529328346252,
+      "learning_rate": 1.955858616489059e-06,
+      "loss": 0.7037,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9399301793667991,
+      "grad_norm": 0.9045463800430298,
+      "learning_rate": 1.9252777858599915e-06,
+      "loss": 0.7651,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9404117009750813,
+      "grad_norm": 0.7330710291862488,
+      "learning_rate": 1.8949355922479151e-06,
+      "loss": 0.4345,
+      "step": 1953
+    },
+    {
+      "epoch": 0.9408932225833634,
+      "grad_norm": 0.9409630298614502,
+      "learning_rate": 1.8648321094819287e-06,
+      "loss": 0.7097,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9413747441916456,
+      "grad_norm": 0.9952296018600464,
+      "learning_rate": 1.8349674108103288e-06,
+      "loss": 0.5895,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9418562657999278,
+      "grad_norm": 0.7514768838882446,
+      "learning_rate": 1.8053415689003872e-06,
+      "loss": 0.566,
+      "step": 1956
+    },
+    {
+      "epoch": 0.94233778740821,
+      "grad_norm": 0.7092472314834595,
+      "learning_rate": 1.7759546558381967e-06,
+      "loss": 0.2969,
+      "step": 1957
+    },
+    {
+      "epoch": 0.9428193090164921,
+      "grad_norm": 0.7995561361312866,
+      "learning_rate": 1.7468067431284707e-06,
+      "loss": 0.286,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9433008306247743,
+      "grad_norm": 1.1525226831436157,
+      "learning_rate": 1.7178979016943764e-06,
+      "loss": 0.5039,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9437823522330565,
+      "grad_norm": 0.40452930331230164,
+      "learning_rate": 1.6892282018773908e-06,
+      "loss": 0.2762,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9442638738413386,
+      "grad_norm": 0.7908893823623657,
+      "learning_rate": 1.6607977134370789e-06,
+      "loss": 0.6573,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9447453954496208,
+      "grad_norm": 0.7641875743865967,
+      "learning_rate": 1.6326065055510043e-06,
+      "loss": 0.5544,
+      "step": 1962
+    },
+    {
+      "epoch": 0.945226917057903,
+      "grad_norm": 0.8250647187232971,
+      "learning_rate": 1.6046546468144407e-06,
+      "loss": 0.6286,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9457084386661851,
+      "grad_norm": 0.6536379456520081,
+      "learning_rate": 1.576942205240317e-06,
+      "loss": 0.361,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9461899602744673,
+      "grad_norm": 0.6074386835098267,
+      "learning_rate": 1.5494692482590057e-06,
+      "loss": 0.4025,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9466714818827495,
+      "grad_norm": 0.6296261548995972,
+      "learning_rate": 1.522235842718156e-06,
+      "loss": 0.5711,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9471530034910317,
+      "grad_norm": 0.8962925672531128,
+      "learning_rate": 1.4952420548825285e-06,
+      "loss": 0.4009,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9476345250993138,
+      "grad_norm": 0.6276614665985107,
+      "learning_rate": 1.468487950433839e-06,
+      "loss": 0.4574,
+      "step": 1968
+    },
+    {
+      "epoch": 0.948116046707596,
+      "grad_norm": 0.5954809188842773,
+      "learning_rate": 1.441973594470636e-06,
+      "loss": 0.4849,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9485975683158782,
+      "grad_norm": 1.2679002285003662,
+      "learning_rate": 1.415699051508068e-06,
+      "loss": 0.5457,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9490790899241603,
+      "grad_norm": 0.5073424577713013,
+      "learning_rate": 1.3896643854777847e-06,
+      "loss": 0.4817,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9495606115324425,
+      "grad_norm": 0.6933119893074036,
+      "learning_rate": 1.3638696597277679e-06,
+      "loss": 0.3137,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9500421331407247,
+      "grad_norm": 0.6285628080368042,
+      "learning_rate": 1.3383149370221449e-06,
+      "loss": 0.5385,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9505236547490069,
+      "grad_norm": 1.3554258346557617,
+      "learning_rate": 1.313000279541121e-06,
+      "loss": 0.4391,
+      "step": 1974
+    },
+    {
+      "epoch": 0.951005176357289,
+      "grad_norm": 0.6554133892059326,
+      "learning_rate": 1.287925748880703e-06,
+      "loss": 0.4671,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9514866979655712,
+      "grad_norm": 1.4406400918960571,
+      "learning_rate": 1.2630914060526522e-06,
+      "loss": 0.8691,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9519682195738534,
+      "grad_norm": 0.5234692692756653,
+      "learning_rate": 1.2384973114843101e-06,
+      "loss": 0.5122,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9524497411821355,
+      "grad_norm": 0.3509373664855957,
+      "learning_rate": 1.2141435250184185e-06,
+      "loss": 0.0802,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9529312627904177,
+      "grad_norm": 0.504723072052002,
+      "learning_rate": 1.1900301059130093e-06,
+      "loss": 0.5345,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9534127843986999,
+      "grad_norm": 0.5742740631103516,
+      "learning_rate": 1.1661571128412596e-06,
+      "loss": 0.4383,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9538943060069821,
+      "grad_norm": 0.576581597328186,
+      "learning_rate": 1.142524603891315e-06,
+      "loss": 0.5606,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9543758276152642,
+      "grad_norm": 0.7667593955993652,
+      "learning_rate": 1.1191326365661892e-06,
+      "loss": 0.5136,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9548573492235464,
+      "grad_norm": 0.4355958104133606,
+      "learning_rate": 1.0959812677835968e-06,
+      "loss": 0.2748,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9553388708318286,
+      "grad_norm": 0.7036843299865723,
+      "learning_rate": 1.0730705538758322e-06,
+      "loss": 0.5994,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9558203924401107,
+      "grad_norm": 0.5516580939292908,
+      "learning_rate": 1.0504005505896141e-06,
+      "loss": 0.4582,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9563019140483929,
+      "grad_norm": 0.5592047572135925,
+      "learning_rate": 1.0279713130859514e-06,
+      "loss": 0.5129,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9567834356566751,
+      "grad_norm": 0.5601054430007935,
+      "learning_rate": 1.005782895940055e-06,
+      "loss": 0.8754,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9572649572649573,
+      "grad_norm": 0.7053203582763672,
+      "learning_rate": 9.838353531411272e-07,
+      "loss": 0.4834,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9577464788732394,
+      "grad_norm": 0.4365988075733185,
+      "learning_rate": 9.62128738092294e-07,
+      "loss": 0.5507,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9582280004815216,
+      "grad_norm": 0.4267100989818573,
+      "learning_rate": 9.406631036104508e-07,
+      "loss": 0.2282,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9587095220898038,
+      "grad_norm": 0.9049471020698547,
+      "learning_rate": 9.194385019261287e-07,
+      "loss": 0.7224,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9591910436980859,
+      "grad_norm": 0.8407077789306641,
+      "learning_rate": 8.984549846833612e-07,
+      "loss": 0.407,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9596725653063681,
+      "grad_norm": 0.5468326210975647,
+      "learning_rate": 8.777126029396065e-07,
+      "loss": 0.2352,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9601540869146503,
+      "grad_norm": 0.5015347003936768,
+      "learning_rate": 8.572114071655479e-07,
+      "loss": 0.2969,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9606356085229325,
+      "grad_norm": 0.6514408588409424,
+      "learning_rate": 8.369514472450379e-07,
+      "loss": 0.2341,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9611171301312147,
+      "grad_norm": 0.6202306151390076,
+      "learning_rate": 8.169327724749543e-07,
+      "loss": 0.6092,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9615986517394968,
+      "grad_norm": 0.596602737903595,
+      "learning_rate": 7.971554315650442e-07,
+      "loss": 0.5926,
+      "step": 1997
+    },
+    {
+      "epoch": 0.962080173347779,
+      "grad_norm": 1.011766791343689,
+      "learning_rate": 7.776194726378583e-07,
+      "loss": 0.8328,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9625616949560611,
+      "grad_norm": 0.7206372022628784,
+      "learning_rate": 7.583249432286277e-07,
+      "loss": 0.4498,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9630432165643433,
+      "grad_norm": 0.48031672835350037,
+      "learning_rate": 7.392718902850981e-07,
+      "loss": 0.306,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9635247381726255,
+      "grad_norm": 0.7958242297172546,
+      "learning_rate": 7.204603601674853e-07,
+      "loss": 0.3053,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9640062597809077,
+      "grad_norm": 1.0383106470108032,
+      "learning_rate": 7.018903986483083e-07,
+      "loss": 0.7788,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9644877813891899,
+      "grad_norm": 0.9940798282623291,
+      "learning_rate": 6.835620509122897e-07,
+      "loss": 0.5206,
+      "step": 2003
+    },
+    {
+      "epoch": 0.964969302997472,
+      "grad_norm": 0.8290696144104004,
+      "learning_rate": 6.65475361556267e-07,
+      "loss": 0.452,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9654508246057542,
+      "grad_norm": 0.6911272406578064,
+      "learning_rate": 6.47630374589081e-07,
+      "loss": 0.5752,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9659323462140363,
+      "grad_norm": 0.6400201320648193,
+      "learning_rate": 6.300271334314434e-07,
+      "loss": 0.3883,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9664138678223185,
+      "grad_norm": 0.5815731287002563,
+      "learning_rate": 6.126656809158359e-07,
+      "loss": 0.6059,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9668953894306007,
+      "grad_norm": 0.5034971237182617,
+      "learning_rate": 5.955460592864337e-07,
+      "loss": 0.5949,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9673769110388829,
+      "grad_norm": 0.64183509349823,
+      "learning_rate": 5.78668310198982e-07,
+      "loss": 0.7758,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9678584326471651,
+      "grad_norm": 0.5774924159049988,
+      "learning_rate": 5.620324747207084e-07,
+      "loss": 0.2339,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9683399542554472,
+      "grad_norm": 0.6001867055892944,
+      "learning_rate": 5.456385933301777e-07,
+      "loss": 0.2502,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9688214758637294,
+      "grad_norm": 0.5865710377693176,
+      "learning_rate": 5.294867059172592e-07,
+      "loss": 0.2,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9693029974720115,
+      "grad_norm": 0.47389328479766846,
+      "learning_rate": 5.135768517829819e-07,
+      "loss": 0.4203,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9697845190802937,
+      "grad_norm": 0.6576399803161621,
+      "learning_rate": 4.979090696394795e-07,
+      "loss": 0.7084,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9702660406885759,
+      "grad_norm": 0.9264883995056152,
+      "learning_rate": 4.824833976098453e-07,
+      "loss": 0.3988,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9707475622968581,
+      "grad_norm": 1.5536925792694092,
+      "learning_rate": 4.6729987322807757e-07,
+      "loss": 0.9253,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9712290839051403,
+      "grad_norm": 0.6507259607315063,
+      "learning_rate": 4.523585334389679e-07,
+      "loss": 0.806,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9717106055134224,
+      "grad_norm": 0.9468237161636353,
+      "learning_rate": 4.3765941459804614e-07,
+      "loss": 0.4927,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9721921271217046,
+      "grad_norm": 0.6918371915817261,
+      "learning_rate": 4.232025524714356e-07,
+      "loss": 0.1446,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9726736487299867,
+      "grad_norm": 0.9984461665153503,
+      "learning_rate": 4.0898798223582e-07,
+      "loss": 0.5403,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9731551703382689,
+      "grad_norm": 0.6529189348220825,
+      "learning_rate": 3.950157384783104e-07,
+      "loss": 0.5878,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9736366919465511,
+      "grad_norm": 0.6561586260795593,
+      "learning_rate": 3.8128585519640046e-07,
+      "loss": 0.3575,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9741182135548333,
+      "grad_norm": 0.8341250419616699,
+      "learning_rate": 3.677983657978779e-07,
+      "loss": 0.7522,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9745997351631155,
+      "grad_norm": 0.7429258823394775,
+      "learning_rate": 3.545533031007131e-07,
+      "loss": 0.4519,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9750812567713976,
+      "grad_norm": 0.771535336971283,
+      "learning_rate": 3.415506993330153e-07,
+      "loss": 0.2805,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9755627783796798,
+      "grad_norm": 1.0194131135940552,
+      "learning_rate": 3.2879058613292105e-07,
+      "loss": 0.2996,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9760442999879619,
+      "grad_norm": 0.5492636561393738,
+      "learning_rate": 3.1627299454856095e-07,
+      "loss": 0.3451,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9765258215962441,
+      "grad_norm": 0.6983237862586975,
+      "learning_rate": 3.0399795503793793e-07,
+      "loss": 0.8118,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9770073432045263,
+      "grad_norm": 0.5727863311767578,
+      "learning_rate": 2.9196549746888235e-07,
+      "loss": 0.1593,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9774888648128085,
+      "grad_norm": 0.5290704965591431,
+      "learning_rate": 2.801756511189524e-07,
+      "loss": 0.1473,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9779703864210907,
+      "grad_norm": 0.647200882434845,
+      "learning_rate": 2.686284446754006e-07,
+      "loss": 0.2473,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9784519080293728,
+      "grad_norm": 0.8799815773963928,
+      "learning_rate": 2.573239062350963e-07,
+      "loss": 0.9664,
+      "step": 2032
+    },
+    {
+      "epoch": 0.978933429637655,
+      "grad_norm": 0.8440240025520325,
+      "learning_rate": 2.4626206330440326e-07,
+      "loss": 1.0026,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9794149512459371,
+      "grad_norm": 0.7759560942649841,
+      "learning_rate": 2.3544294279918e-07,
+      "loss": 0.4478,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9798964728542193,
+      "grad_norm": 0.4280017614364624,
+      "learning_rate": 2.2486657104471286e-07,
+      "loss": 0.6288,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9803779944625015,
+      "grad_norm": 0.7536220550537109,
+      "learning_rate": 2.1453297377557191e-07,
+      "loss": 0.3241,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9808595160707837,
+      "grad_norm": 0.8434854745864868,
+      "learning_rate": 2.044421761356552e-07,
+      "loss": 0.345,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9813410376790659,
+      "grad_norm": 1.095892310142517,
+      "learning_rate": 1.9459420267804452e-07,
+      "loss": 1.0977,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9818225592873481,
+      "grad_norm": 1.2428241968154907,
+      "learning_rate": 1.8498907736499426e-07,
+      "loss": 0.7439,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9823040808956301,
+      "grad_norm": 1.0324878692626953,
+      "learning_rate": 1.7562682356786487e-07,
+      "loss": 0.4408,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9827856025039123,
+      "grad_norm": 0.7506169676780701,
+      "learning_rate": 1.665074640670228e-07,
+      "loss": 0.8056,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9832671241121945,
+      "grad_norm": 0.48109811544418335,
+      "learning_rate": 1.576310210518517e-07,
+      "loss": 0.2994,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9837486457204767,
+      "grad_norm": 0.7194755673408508,
+      "learning_rate": 1.489975161206636e-07,
+      "loss": 0.572,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9842301673287589,
+      "grad_norm": 0.6078580617904663,
+      "learning_rate": 1.406069702806323e-07,
+      "loss": 0.2252,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9847116889370411,
+      "grad_norm": 1.0418319702148438,
+      "learning_rate": 1.324594039477822e-07,
+      "loss": 0.9186,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9851932105453233,
+      "grad_norm": 0.6754590272903442,
+      "learning_rate": 1.2455483694689962e-07,
+      "loss": 0.9269,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9856747321536053,
+      "grad_norm": 0.754641592502594,
+      "learning_rate": 1.1689328851151038e-07,
+      "loss": 0.9815,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9861562537618875,
+      "grad_norm": 0.5206847786903381,
+      "learning_rate": 1.0947477728381339e-07,
+      "loss": 0.8868,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9866377753701697,
+      "grad_norm": 0.599271297454834,
+      "learning_rate": 1.0229932131465836e-07,
+      "loss": 0.3564,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9871192969784519,
+      "grad_norm": 0.6615093350410461,
+      "learning_rate": 9.536693806347919e-08,
+      "loss": 0.5842,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9876008185867341,
+      "grad_norm": 0.693345308303833,
+      "learning_rate": 8.867764439826065e-08,
+      "loss": 0.3433,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9880823401950163,
+      "grad_norm": 0.5857682824134827,
+      "learning_rate": 8.223145659550513e-08,
+      "loss": 0.6968,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9885638618032985,
+      "grad_norm": 0.8526471853256226,
+      "learning_rate": 7.602839034017706e-08,
+      "loss": 0.6844,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9890453834115805,
+      "grad_norm": 0.7888714671134949,
+      "learning_rate": 7.006846072568074e-08,
+      "loss": 0.3751,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9895269050198627,
+      "grad_norm": 0.969866156578064,
+      "learning_rate": 6.435168225381594e-08,
+      "loss": 0.5924,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9900084266281449,
+      "grad_norm": 0.3491519093513489,
+      "learning_rate": 5.887806883474456e-08,
+      "loss": 0.1858,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9904899482364271,
+      "grad_norm": 0.8504678606987,
+      "learning_rate": 5.364763378694626e-08,
+      "loss": 0.4558,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9909714698447093,
+      "grad_norm": 0.9942469596862793,
+      "learning_rate": 4.8660389837207334e-08,
+      "loss": 0.8354,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9914529914529915,
+      "grad_norm": 0.7020037770271301,
+      "learning_rate": 4.391634912056519e-08,
+      "loss": 1.1202,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9919345130612737,
+      "grad_norm": 0.4268772602081299,
+      "learning_rate": 3.9415523180297286e-08,
+      "loss": 0.1628,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9924160346695557,
+      "grad_norm": 0.5923699140548706,
+      "learning_rate": 3.515792296789888e-08,
+      "loss": 0.1115,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9928975562778379,
+      "grad_norm": 0.8240038156509399,
+      "learning_rate": 3.114355884301645e-08,
+      "loss": 0.7954,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9933790778861201,
+      "grad_norm": 0.8708813190460205,
+      "learning_rate": 2.7372440573469883e-08,
+      "loss": 0.4003,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9938605994944023,
+      "grad_norm": 0.6724766492843628,
+      "learning_rate": 2.384457733520806e-08,
+      "loss": 0.225,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9943421211026845,
+      "grad_norm": 0.7021605372428894,
+      "learning_rate": 2.0559977712297785e-08,
+      "loss": 0.2706,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9948236427109667,
+      "grad_norm": 0.7431803941726685,
+      "learning_rate": 1.7518649696857126e-08,
+      "loss": 0.6236,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9953051643192489,
+      "grad_norm": 0.4876454174518585,
+      "learning_rate": 1.4720600689110963e-08,
+      "loss": 0.596,
+      "step": 2067
+    },
+    {
+      "epoch": 0.995786685927531,
+      "grad_norm": 0.9122818112373352,
+      "learning_rate": 1.216583749731326e-08,
+      "loss": 0.5965,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9962682075358131,
+      "grad_norm": 0.8722514510154724,
+      "learning_rate": 9.854366337758159e-09,
+      "loss": 0.8851,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9967497291440953,
+      "grad_norm": 0.9117795825004578,
+      "learning_rate": 7.786192834746686e-09,
+      "loss": 0.8438,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9972312507523775,
+      "grad_norm": 0.7254170179367065,
+      "learning_rate": 5.961322020608951e-09,
+      "loss": 0.4586,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9977127723606597,
+      "grad_norm": 0.433788001537323,
+      "learning_rate": 4.3797583356264275e-09,
+      "loss": 0.1281,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9981942939689419,
+      "grad_norm": 0.6030900478363037,
+      "learning_rate": 3.0415056281096755e-09,
+      "loss": 0.371,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9986758155772241,
+      "grad_norm": 0.9900925755500793,
+      "learning_rate": 1.9465671543095197e-09,
+      "loss": 0.7417,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9991573371855061,
+      "grad_norm": 0.774526834487915,
+      "learning_rate": 1.094945578439255e-09,
+      "loss": 0.6423,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9996388587937883,
+      "grad_norm": 0.49781298637390137,
+      "learning_rate": 4.866429726857469e-10,
+      "loss": 0.582,
+      "step": 2076
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6113485097885132,
+      "learning_rate": 1.2166081717612797e-10,
+      "loss": 0.4426,
+      "step": 2077
+    },
+    {
+      "epoch": 1.0,
+      "step": 2077,
+      "total_flos": 4.4863387190181914e+17,
+      "train_loss": 0.6395983156794302,
+      "train_runtime": 14822.4278,
+      "train_samples_per_second": 2.242,
+      "train_steps_per_second": 0.14
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2077,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2400000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.4863387190181914e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}